diff options
Diffstat (limited to 'Build')
57 files changed, 31929 insertions, 0 deletions
diff --git a/Build/source/texk/web2c/luatexdir/luapplib/Makefile.orig b/Build/source/texk/web2c/luatexdir/luapplib/Makefile.orig new file mode 100644 index 00000000000..c6c1a73dd63 --- /dev/null +++ b/Build/source/texk/web2c/luatexdir/luapplib/Makefile.orig @@ -0,0 +1,84 @@ +
+ifeq ($(ARCH),)
+ ARCH=x64
+endif
+
+UNAME:=$(shell uname)
+ifeq ($(UNAME),Linux)
+ BUILDSUFFIX=_ux_$(ARCH)
+ SOEXT=.so
+ EXEEXT=
+ STATICCOPTS=
+# SHAREDCOPTS=-fPIC
+ LFLAGS=
+ LIBS=-lm
+else
+ BUILDSUFFIX=_win_$(ARCH)
+ SOEXT=.dll
+ EXEEXT=.exe
+ STATICCOPTS=
+# SHAREDCOPTS=-DPPDLL
+ LFLAGS = -Wl,--output-def,$@.def,--out-implib,$@.a
+ LIBS=
+endif
+
+CC=gcc
+W=-Wall -Wunused -Wimplicit -Wreturn-type -Wdeclaration-after-statement -Wno-unknown-pragmas -Wmissing-prototypes -Wmissing-declarations -Wparentheses -Wswitch -Wtrigraphs -Wcast-qual -Wcast-align -Wwrite-strings -Wold-style-definition -Wpointer-arith
+#W=-Wall
+CFLAGS=$(W) -O2
+
+OBJEXT=$(BUILDSUFFIX).o
+
+PPUTILS=utilmem utillog utilnumber utiliof utilmd5 utilsha utilcrypt
+PPFILTERS=utilbasexx utilflate utillzw utilfpred
+PPMODULES=ppheap pparray ppdict ppstream ppcrypt ppxref ppload
+
+OUTDIR=build
+UTILSDIR=util
+
+PPSTATICUTILSOBJ=$(patsubst %, $(OUTDIR)/%$(OBJEXT), $(PPUTILS))
+PPSTATICFILTERSOBJ=$(patsubst %, $(OUTDIR)/%$(OBJEXT), $(PPFILTERS))
+PPSTATICOBJ=$(patsubst %, $(OUTDIR)/%$(OBJEXT), $(PPMODULES))
+PPSTATICLIB=$(OUTDIR)/libpp.static$(BUILDSUFFIX).a
+
+# dependencies
+FLATEDIR=zlib
+FLATELIB=$(FLATEDIR)/libz.static$(BUILDSUFFIX).a
+PPSTATICDEPS=$(FLATELIB)
+
+# test programm
+PPTEST1=$(OUTDIR)/pptest1$(EXEEXT)
+PPTEST2=$(OUTDIR)/pptest2$(EXEEXT)
+
+# includes
+INC=-I $(UTILSDIR)
+AUX=-I $(FLATEDIR)
+
+default: md $(PPSTATICLIB) $(PPTEST1) $(PPTEST2)
+
+md:
+ @[ -d $(OUTDIR) ] || mkdir -p $(OUTDIR)
+
+$(PPSTATICOBJ): $(OUTDIR)/%$(OBJEXT): %.c %.h
+ $(CC) $(CFLAGS) $(STATICCOPTS) $(INC) -o $@ -c $<
+
+$(PPSTATICUTILSOBJ): $(OUTDIR)/%$(OBJEXT): $(UTILSDIR)/%.c $(UTILSDIR)/%.h
+ $(CC) $(CFLAGS) $(STATICCOPTS) $(INC) -o $@ -c $<
+
+$(PPSTATICFILTERSOBJ): $(OUTDIR)/%$(OBJEXT): $(UTILSDIR)/%.c $(UTILSDIR)/%.h
+ $(CC) $(CFLAGS) $(STATICCOPTS) $(INC) $(AUX) -o $@ -c $<
+
+$(PPSTATICLIB): $(PPSTATICOBJ) $(PPSTATICUTILSOBJ) $(PPSTATICFILTERSOBJ)
+ ar rcs $@ $^
+ strip -N print_hex $@
+
+$(PPTEST1): $(PPSTATICLIB) pptest1.c
+ $(CC) $(CFLAGS) $(STATICCOPTS) -o $(OUTDIR)/pptest1$(OBJEXT) -c pptest1.c
+ $(CC) -static-libgcc $(CFLAGS) -o $(PPTEST1) $(OUTDIR)/pptest1$(OBJEXT) $(PPSTATICLIB) $(PPSTATICDEPS) $(LIBS)
+
+$(PPTEST2): $(PPSTATICLIB) pptest2.c
+ $(CC) $(CFLAGS) $(STATICCOPTS) -o $(OUTDIR)/pptest2$(OBJEXT) -c pptest2.c
+ $(CC) -static-libgcc $(CFLAGS) -o $(PPTEST2) $(OUTDIR)/pptest2$(OBJEXT) $(PPSTATICLIB) $(PPSTATICDEPS) $(LIBS)
+
+clean:
+ rm -f $(OUTDIR)/*$(OBJEXT) $(OUTDIR)/*.a $(OUTDIR)/*$(EXEEXT)
diff --git a/Build/source/texk/web2c/luatexdir/luapplib/ppapi.h b/Build/source/texk/web2c/luatexdir/luapplib/ppapi.h new file mode 100644 index 00000000000..6d591ca05fb --- /dev/null +++ b/Build/source/texk/web2c/luatexdir/luapplib/ppapi.h @@ -0,0 +1,391 @@ + +#ifndef PP_API_H +#define PP_API_H + +#include <stdint.h> +#include <stddef.h> +#include <string.h> + +#include "ppconf.h" + +#define pplib_version "v0.97" +#define pplib_author "p.jackowski@gust.org.pl" + +/* types */ + +typedef int64_t ppint; +typedef size_t ppuint; // machine word + +typedef double ppnum; +typedef char * ppname; +typedef char * ppstring; + +typedef struct { + size_t size; + int flags; +} _ppname; + +typedef struct { + size_t size; + int flags; +} _ppstring; + +typedef struct ppobj ppobj; +typedef struct ppref ppref; + + +#if defined __arm__ || defined __ARM__ || defined ARM || defined __ARM || defined __arm || defined __ARM_ARCH ||defined __aarch64__ +typedef struct { + ppobj *data; + size_t size; + ppnum PPARRAY_ALIGNMENT; +} pparray; +#else +typedef struct { + ppobj *data; + size_t size; +} pparray; +#endif + + +#if defined __arm__ || defined __ARM__ || defined ARM || defined __ARM || defined __arm || defined __ARM_ARCH ||defined __aarch64__ +typedef struct { + ppobj *data; + ppname *keys; + size_t size; + ppnum PPDICT_ALIGNMENT; +} ppdict; + +#else +typedef struct { + ppobj *data; + ppname *keys; + size_t size; +} ppdict; +#endif + + +typedef struct { + ppdict *dict; + void *input, *I; + size_t offset; + size_t length; + ppstring cryptkey; + int flags; +} ppstream; + +#define PPSTREAM_COMPRESSED (1<<0) +#define PPSTREAM_ENCRYPTED_AES (1<<1) +#define PPSTREAM_ENCRYPTED_RC4 (1<<2) +#define PPSTREAM_ENCRYPTED (PPSTREAM_ENCRYPTED_AES|PPSTREAM_ENCRYPTED_RC4) +#define PPSTREAM_ENCRYPTED_OWN (1<<3) + +#define ppstream_compressed(stream) ((stream)->flags & PPSTREAM_COMPRESSED) +#define ppstream_encrypted(stream) ((stream)->flags & PPSTREAM_ENCRYPTED) + +typedef enum { + PPNONE = 0, + PPNULL, + PPBOOL, + PPINT, + PPNUM, + PPNAME, + PPSTRING, + PPARRAY, + PPDICT, + PPSTREAM, + PPREF +} ppobjtp; + +PPDEF extern const char * ppobj_kind[]; + +struct ppobj { + ppobjtp type; + union { + ppint integer; + ppnum number; + ppname name; + ppstring string; + pparray *array; + ppdict *dict; + ppstream *stream; + ppref *ref; + void *any; + }; +}; + +typedef struct ppxref ppxref; + +struct ppref { + ppobj object; + ppuint number, version; + size_t offset; + size_t length; + ppxref *xref; +}; + +typedef struct ppdoc ppdoc; + +/* object */ + +#define ppobj_get_null(o) ((o)->type == PPNULL ? 1 : 0) +#define ppobj_get_bool(o, v) ((o)->type == PPBOOL ? ((v = ((o)->integer != 0)), 1) : 0) +#define ppobj_get_int(o, v) ((o)->type == PPINT ? ((v = (o)->integer), 1) : 0) +#define ppobj_get_uint(o, v) ((o)->type == PPINT && (o)->integer >= 0 ? ((v = (ppuint)((o)->integer)), 1) : 0) +#define ppobj_get_num(o, v) ((o)->type == PPNUM ? ((v = (o)->number), 1) : (((o)->type == PPINT ? ((v = (ppnum)((o)->integer)), 1) : 0))) +#define ppobj_get_name(o) ((o)->type == PPNAME ? (o)->name : NULL) +#define ppobj_get_string(o) ((o)->type == PPSTRING ? (o)->string : NULL) +#define ppobj_get_array(o) ((o)->type == PPARRAY ? (o)->array : NULL) +#define ppobj_get_dict(o) ((o)->type == PPDICT ? (o)->dict : NULL) +#define ppobj_get_stream(o) ((o)->type == PPSTREAM ? (o)->stream : NULL) +#define ppobj_get_ref(o) ((o)->type == PPREF ? (o)->ref : NULL) + +#define ppobj_rget_obj(o) ((o)->type == PPREF ? ppref_obj((o)->ref) : o) +#define ppobj_rget_null(o) ((o)->type == PPNULL ? 1 : ((o)->type == PPREF ? ppobj_get_null(ppref_obj((o)->ref)) : 0)) +#define ppobj_rget_bool(o, v) ((o)->type == PPBOOL ? ((v = ((o)->integer != 0)), 1) : ((o)->type == PPREF ? ppobj_get_bool(ppref_obj((o)->ref), v) : 0)) +#define ppobj_rget_int(o, v) ((o)->type == PPINT ? ((v = (o)->integer), 1) : ((o)->type == PPREF ? ppobj_get_int(ppref_obj((o)->ref), v) : 0)) +#define ppobj_rget_uint(o, v) ((o)->type == PPINT && (o)->integer >= 0 ? ((v = (ppuint)((o)->integer)), 1) : ((o)->type == PPREF ? ppobj_get_uint(ppref_obj((o)->ref), v) : 0)) +#define ppobj_rget_num(o, v) ((o)->type == PPNUM ? ((v = (o)->number), 1) : (((o)->type == PPINT ? ((v = (ppnum)((o)->integer)), 1) : ((o)->type == PPREF ? ppobj_get_num(ppref_obj((o)->ref), v) : 0)))) +#define ppobj_rget_name(o) ((o)->type == PPNAME ? (o)->name : ((o)->type == PPREF ? ppobj_get_name(ppref_obj((o)->ref)) : NULL)) +#define ppobj_rget_string(o) ((o)->type == PPSTRING ? (o)->string : ((o)->type == PPREF ? ppobj_get_string(ppref_obj((o)->ref)) : NULL)) +#define ppobj_rget_array(o) ((o)->type == PPARRAY ? (o)->array : ((o)->type == PPREF ? ppobj_get_array(ppref_obj((o)->ref)) : NULL)) +#define ppobj_rget_dict(o) ((o)->type == PPDICT ? (o)->dict : ((o)->type == PPREF ? ppobj_get_dict(ppref_obj((o)->ref)) : NULL)) +#define ppobj_rget_stream(o) ((o)->type == PPSTREAM ? (o)->stream : ((o)->type == PPREF ? ppobj_get_stream(ppref_obj((o)->ref)) : NULL)) +#define ppobj_rget_ref(o) ((o)->type == PPREF ? (o)->ref : ((o)->type == PPREF ? ppobj_get_ref(ppref_obj((o)->ref)) : NULL)) + +#define ppobj_get_bool_value(o) ((o)->type == PPBOOL ? ((o)->integer != 0) : 0) +#define ppobj_get_int_value(o) ((o)->type == PPINT ? (o)->integer : 0) +#define ppobj_get_num_value(o) ((o)->type == PPNUM ? (o)->number : ((o)->type == PPINT ? (ppnum)((o)->integer) : 0.0)) + +/* name */ + +#define ppname_is(name, s) (memcmp(name, s, sizeof("" s) - 1) == 0) +#define ppname_eq(name, n) (memcmp(name, s, ppname_size(name)) == 0) + +#if defined __arm__ || defined __ARM__ || defined ARM || defined __ARM || defined __arm || defined __ARM_ARCH ||defined __aarch64__ +#define _ppname_ghost(name) (((const _ppname *)((void *)name)) - 1) +#else +#define _ppname_ghost(name) (((const _ppname *)(name)) - 1) +#endif + +#define ppname_size(name) (_ppname_ghost(name)->size) +#define ppname_exec(name) (_ppname_ghost(name)->flags & PPNAME_EXEC) + +#define PPNAME_ENCODED (1 << 0) +#define PPNAME_DECODED (1 << 1) +#define PPNAME_EXEC (1 << 1) + +PPAPI ppname ppname_decoded (ppname name); +PPAPI ppname ppname_encoded (ppname name); + +/* string */ + +#if defined __arm__ || defined __ARM__ || defined ARM || defined __ARM || defined __arm || defined __ARM_ARCH ||defined __aarch64__ +#define _ppstring_ghost(string) (((const _ppstring *)((void *)string)) - 1) +#else +#define _ppstring_ghost(string) (((const _ppstring *)(string)) - 1) +#endif + +#define ppstring_size(string) (_ppstring_ghost(string)->size) + +#define PPSTRING_ENCODED (1 << 0) +#define PPSTRING_DECODED (1 << 1) +//#define PPSTRING_EXEC (1 << 2) // postscript only +#define PPSTRING_PLAIN 0 +#define PPSTRING_BASE16 (1 << 3) +#define PPSTRING_BASE85 (1 << 4) +#define PPSTRING_UTF16BE (1 << 5) +#define PPSTRING_UTF16LE (1 << 6) + +#define ppstring_type(string) (_ppstring_ghost(string)->flags & (PPSTRING_BASE16|PPSTRING_BASE85)) +#define ppstring_hex(string) (_ppstring_ghost(string)->flags & PPSTRING_BASE16) +#define ppstring_utf(string) (_ppstring_ghost(string)->flags & (PPSTRING_UTF16BE|PPSTRING_UTF16LE)) + +PPAPI ppstring ppstring_decoded (ppstring string); +PPAPI ppstring ppstring_encoded (ppstring string); + +/* array */ + +#define pparray_size(array) ((array)->size) +#define pparray_at(array, index) ((array)->data + index) + +#define pparray_first(array, index, obj) ((index) = 0, (obj) = pparray_at(array, 0)) +#define pparray_next(index, obj) (++(index), ++(obj)) + +#define pparray_get(array, index) (index < (array)->size ? pparray_at(array, index) : NULL) + +PPAPI ppobj * pparray_get_obj (pparray *array, size_t index); +PPAPI int pparray_get_bool (pparray *array, size_t index, int *v); +PPAPI int pparray_get_int (pparray *array, size_t index, ppint *v); +PPAPI int pparray_get_uint (pparray *array, size_t index, ppuint *v); +PPAPI int pparray_get_num (pparray *array, size_t index, ppnum *v); +PPAPI ppname pparray_get_name (pparray *array, size_t index); +PPAPI ppstring pparray_get_string (pparray *array, size_t index); +PPAPI pparray * pparray_get_array (pparray *array, size_t index); +PPAPI ppdict * pparray_get_dict (pparray *array, size_t index); +//PPAPI ppstream * pparray_get_stream (pparray *array, size_t index); +PPAPI ppref * pparray_get_ref (pparray *array, size_t index); + +PPAPI ppobj * pparray_rget_obj (pparray *array, size_t index); +PPAPI int pparray_rget_bool (pparray *array, size_t index, int *v); +PPAPI int pparray_rget_int (pparray *array, size_t index, ppint *v); +PPAPI int pparray_rget_uint (pparray *array, size_t index, ppuint *v); +PPAPI int pparray_rget_num (pparray *array, size_t index, ppnum *v); +PPAPI ppname pparray_rget_name (pparray *array, size_t index); +PPAPI ppstring pparray_rget_string (pparray *array, size_t index); +PPAPI pparray * pparray_rget_array (pparray *array, size_t index); +PPAPI ppdict * pparray_rget_dict (pparray *array, size_t index); +PPAPI ppstream * pparray_rget_stream (pparray *array, size_t index); +PPAPI ppref * pparray_rget_ref (pparray *array, size_t index); + +/* dict */ + +#define ppdict_size(dict) ((dict)->size) +#define ppdict_at(dict, index) ((dict)->data + index) +#define ppdict_key(dict, index) ((dict)->keys[index]) + +PPAPI ppobj * ppdict_get_obj (ppdict *dict, const char *name); +PPAPI int ppdict_get_bool (ppdict *dict, const char *name, int *v); +PPAPI int ppdict_get_int (ppdict *dict, const char *name, ppint *v); +PPAPI int ppdict_get_uint (ppdict *dict, const char *name, ppuint *v); +PPAPI int ppdict_get_num (ppdict *dict, const char *name, ppnum *v); +PPAPI ppname ppdict_get_name (ppdict *dict, const char *name); +PPAPI ppstring ppdict_get_string (ppdict *dict, const char *name); +PPAPI pparray * ppdict_get_array (ppdict *dict, const char *name); +PPAPI ppdict * ppdict_get_dict (ppdict *dict, const char *name); +//PPAPI ppstream * ppdict_get_stream (ppdict *dict, const char *name); +PPAPI ppref * ppdict_get_ref (ppdict *dict, const char *name); + +PPAPI ppobj * ppdict_rget_obj (ppdict *dict, const char *name); +PPAPI int ppdict_rget_bool (ppdict *dict, const char *name, int *v); +PPAPI int ppdict_rget_int (ppdict *dict, const char *name, ppint *v); +PPAPI int ppdict_rget_uint (ppdict *dict, const char *name, ppuint *v); +PPAPI int ppdict_rget_num (ppdict *dict, const char *name, ppnum *v); +PPAPI ppname ppdict_rget_name (ppdict *dict, const char *name); +PPAPI ppstring ppdict_rget_string (ppdict *dict, const char *name); +PPAPI pparray * ppdict_rget_array (ppdict *dict, const char *name); +PPAPI ppdict * ppdict_rget_dict (ppdict *dict, const char *name); +PPAPI ppstream * ppdict_rget_stream (ppdict *dict, const char *name); +PPAPI ppref * ppdict_rget_ref (ppdict *dict, const char *name); + +#define ppdict_first(dict, pkey, obj) (pkey = (dict)->keys, obj = (dict)->data) +#define ppdict_next(pkey, obj) (++(pkey), ++(obj)) + +/* stream */ + +#define ppstream_dict(stream) ((stream)->dict) + +PPAPI uint8_t * ppstream_first (ppstream *stream, size_t *size, int decode); +PPAPI uint8_t * ppstream_next (ppstream *stream, size_t *size); +PPAPI uint8_t * ppstream_all (ppstream *stream, size_t *size, int decode); +PPAPI void ppstream_done (ppstream *stream); + +PPAPI void ppstream_init_buffers (void); +PPAPI void ppstream_free_buffers (void); + +/* ref */ + +#define ppref_obj(ref) (&(ref)->object) + +/* xref */ + +PPAPI ppxref * ppdoc_xref (ppdoc *pdf); +PPAPI ppxref * ppxref_prev (ppxref *xref); +PPAPI ppdict * ppxref_trailer (ppxref *xref); +PPAPI ppdict * ppxref_catalog (ppxref *xref); +PPAPI ppdict * ppxref_info (ppxref *xref); +PPAPI ppref * ppxref_pages (ppxref *xref); +PPAPI ppref * ppxref_find (ppxref *xref, ppuint refnumber); + +/* doc */ + +PPAPI ppdoc * ppdoc_load (const char *filename); +PPAPI ppdoc * ppdoc_mem (const void *data, size_t size); +PPAPI void ppdoc_free (ppdoc *pdf); + +#define ppdoc_trailer(pdf) ppxref_trailer(ppdoc_xref(pdf)) +#define ppdoc_catalog(pdf) ppxref_catalog(ppdoc_xref(pdf)) +#define ppdoc_info(pdf) ppxref_info(ppdoc_xref(pdf)) +#define ppdoc_pages(pdf) ppxref_pages(ppdoc_xref(pdf)) + +PPAPI ppuint ppdoc_page_count (ppdoc *pdf); +PPAPI ppref * ppdoc_page (ppdoc *pdf, ppuint index); +PPAPI ppref * ppdoc_first_page (ppdoc *pdf); +PPAPI ppref * ppdoc_next_page (ppdoc *pdf); + +PPAPI ppstream * ppcontents_first (ppdict *dict); +PPAPI ppstream * ppcontents_next (ppdict *dict, ppstream *stream); + +/* crypt */ + +typedef enum { + PPCRYPT_NONE = 0, + PPCRYPT_DONE = 1, + PPCRYPT_FAIL = -1, + PPCRYPT_PASS = -2 +} ppcrypt_status; + +PPAPI ppcrypt_status ppdoc_crypt_status (ppdoc *pdf); +PPAPI ppcrypt_status ppdoc_crypt_pass (ppdoc *pdf, const void *userpass, size_t userpasslength, const void *ownerpass, size_t ownerpasslength); + +/* permission flags, effect in Acrobat File -> Properties -> Security tab */ + +PPAPI ppint ppdoc_permissions (ppdoc *pdf); + +#define PPDOC_ALLOW_PRINT (1<<2) // printing +#define PPDOC_ALLOW_MODIFY (1<<3) // filling form fields, signing, creating template pages +#define PPDOC_ALLOW_COPY (1<<4) // copying, copying for accessibility +#define PPDOC_ALLOW_ANNOTS (1<<5) // filling form fields, copying, signing +#define PPDOC_ALLOW_EXTRACT (1<<9) // contents copying for accessibility +#define PPDOC_ALLOW_ASSEMBLY (1<<10) // (no effect) +#define PPDOC_ALLOW_PRINT_HIRES (1<<11) // (no effect) + +/* context */ + +typedef struct ppcontext ppcontext; + +PPAPI ppcontext * ppcontext_new (void); +PPAPI void ppcontext_done (ppcontext *context); +PPAPI void ppcontext_free (ppcontext *context); + +/* contents parser */ + +PPAPI ppobj * ppcontents_first_op (ppcontext *context, ppstream *stream, size_t *psize, ppname *pname); +PPAPI ppobj * ppcontents_next_op (ppcontext *context, ppstream *stream, size_t *psize, ppname *pname); +PPAPI ppobj * ppcontents_parse (ppcontext *context, ppstream *stream, size_t *psize); + +/* boxes and transforms */ + +typedef struct { + ppnum lx, ly, rx, ry; +} pprect; + +PPAPI pprect * pparray_to_rect (pparray *array, pprect *rect); +PPAPI pprect * ppdict_get_rect (ppdict *dict, const char *name, pprect *rect); +PPAPI pprect * ppdict_get_box (ppdict *dict, const char *name, pprect *rect); + +typedef struct { + ppnum xx, xy, yx, yy, x, y; +} ppmatrix; + +PPAPI ppmatrix * pparray_to_matrix (pparray *array, ppmatrix *matrix); +PPAPI ppmatrix * ppdict_get_matrix (ppdict *dict, const char *name, ppmatrix *matrix); + +/* logger */ + +typedef void (*pplogger_callback) (const char *message, void *alien); +PPAPI void pplog_callback (pplogger_callback logger, void *alien); +PPAPI int pplog_prefix (const char *prefix); + +/* version */ + +PPAPI const char * ppdoc_version_string (ppdoc *pdf); +PPAPI int ppdoc_version_number (ppdoc *pdf, int *minor); + +/* doc info */ + +PPAPI size_t ppdoc_file_size (ppdoc *pdf); +PPAPI ppuint ppdoc_objects (ppdoc *pdf); +PPAPI size_t ppdoc_memory (ppdoc *pdf, size_t *waste); + +#endif diff --git a/Build/source/texk/web2c/luatexdir/luapplib/ppapi.h-OK b/Build/source/texk/web2c/luatexdir/luapplib/ppapi.h-OK new file mode 100644 index 00000000000..eec45c8c2a5 --- /dev/null +++ b/Build/source/texk/web2c/luatexdir/luapplib/ppapi.h-OK @@ -0,0 +1,360 @@ + +#ifndef PP_API_H +#define PP_API_H + +#include <stdint.h> +#include <stddef.h> +#include <string.h> + +#include "ppconf.h" + +#define pplib_version "v0.97" +#define pplib_author "p.jackowski@gust.org.pl" + +/* types */ + +typedef int64_t ppint; +typedef size_t ppuint; // machine word + +typedef double ppnum; +typedef char * ppname; +typedef char * ppstring; + +typedef struct { + size_t size; + int flags; +} _ppname; + +typedef struct { + size_t size; + int flags; +} _ppstring; + +typedef struct ppobj ppobj; +typedef struct ppref ppref; + +typedef struct { + ppobj *data; + size_t size; +} pparray; + +typedef struct { + ppobj *data; + ppname *keys; + size_t size; +} ppdict; + +typedef struct { + ppdict *dict; + void *input, *I; + size_t offset; + size_t length; + ppstring cryptkey; + int flags; +} ppstream; + +#define PPSTREAM_COMPRESSED (1<<0) +#define PPSTREAM_ENCRYPTED_AES (1<<1) +#define PPSTREAM_ENCRYPTED_RC4 (1<<2) +#define PPSTREAM_ENCRYPTED (PPSTREAM_ENCRYPTED_AES|PPSTREAM_ENCRYPTED_RC4) +#define PPSTREAM_ENCRYPTED_OWN (1<<3) + +#define ppstream_compressed(stream) ((stream)->flags & PPSTREAM_COMPRESSED) +#define ppstream_encrypted(stream) ((stream)->flags & PPSTREAM_ENCRYPTED) + +typedef enum { + PPNONE = 0, + PPNULL, + PPBOOL, + PPINT, + PPNUM, + PPNAME, + PPSTRING, + PPARRAY, + PPDICT, + PPSTREAM, + PPREF +} ppobjtp; + +PPDEF extern const char * ppobj_kind[]; + +struct ppobj { + ppobjtp type; + union { + ppint integer; + ppnum number; + ppname name; + ppstring string; + pparray *array; + ppdict *dict; + ppstream *stream; + ppref *ref; + void *any; + }; +}; + +typedef struct ppxref ppxref; + +struct ppref { + ppobj object; + ppuint number, version; + size_t offset; + size_t length; + ppxref *xref; +}; + +typedef struct ppdoc ppdoc; + +/* object */ + +#define ppobj_get_null(o) ((o)->type == PPNULL ? 1 : 0) +#define ppobj_get_bool(o, v) ((o)->type == PPBOOL ? ((v = ((o)->integer != 0)), 1) : 0) +#define ppobj_get_int(o, v) ((o)->type == PPINT ? ((v = (o)->integer), 1) : 0) +#define ppobj_get_uint(o, v) ((o)->type == PPINT && (o)->integer >= 0 ? ((v = (ppuint)((o)->integer)), 1) : 0) +#define ppobj_get_num(o, v) ((o)->type == PPNUM ? ((v = (o)->number), 1) : (((o)->type == PPINT ? ((v = (ppnum)((o)->integer)), 1) : 0))) +#define ppobj_get_name(o) ((o)->type == PPNAME ? (o)->name : NULL) +#define ppobj_get_string(o) ((o)->type == PPSTRING ? (o)->string : NULL) +#define ppobj_get_array(o) ((o)->type == PPARRAY ? (o)->array : NULL) +#define ppobj_get_dict(o) ((o)->type == PPDICT ? (o)->dict : NULL) +#define ppobj_get_stream(o) ((o)->type == PPSTREAM ? (o)->stream : NULL) +#define ppobj_get_ref(o) ((o)->type == PPREF ? (o)->ref : NULL) + +#define ppobj_rget_obj(o) ((o)->type == PPREF ? ppref_obj((o)->ref) : o) +#define ppobj_rget_null(o) ((o)->type == PPNULL ? 1 : ((o)->type == PPREF ? ppobj_get_null(ppref_obj((o)->ref)) : 0)) +#define ppobj_rget_bool(o, v) ((o)->type == PPBOOL ? ((v = ((o)->integer != 0)), 1) : ((o)->type == PPREF ? ppobj_get_bool(ppref_obj((o)->ref), v) : 0)) +#define ppobj_rget_int(o, v) ((o)->type == PPINT ? ((v = (o)->integer), 1) : ((o)->type == PPREF ? ppobj_get_int(ppref_obj((o)->ref), v) : 0)) +#define ppobj_rget_uint(o, v) ((o)->type == PPINT && (o)->integer >= 0 ? ((v = (ppuint)((o)->integer)), 1) : ((o)->type == PPREF ? ppobj_get_uint(ppref_obj((o)->ref), v) : 0)) +#define ppobj_rget_num(o, v) ((o)->type == PPNUM ? ((v = (o)->number), 1) : (((o)->type == PPINT ? ((v = (ppnum)((o)->integer)), 1) : ((o)->type == PPREF ? ppobj_get_num(ppref_obj((o)->ref), v) : 0)))) +#define ppobj_rget_name(o) ((o)->type == PPNAME ? (o)->name : ((o)->type == PPREF ? ppobj_get_name(ppref_obj((o)->ref)) : NULL)) +#define ppobj_rget_string(o) ((o)->type == PPSTRING ? (o)->string : ((o)->type == PPREF ? ppobj_get_string(ppref_obj((o)->ref)) : NULL)) +#define ppobj_rget_array(o) ((o)->type == PPARRAY ? (o)->array : ((o)->type == PPREF ? ppobj_get_array(ppref_obj((o)->ref)) : NULL)) +#define ppobj_rget_dict(o) ((o)->type == PPDICT ? (o)->dict : ((o)->type == PPREF ? ppobj_get_dict(ppref_obj((o)->ref)) : NULL)) +#define ppobj_rget_stream(o) ((o)->type == PPSTREAM ? (o)->stream : ((o)->type == PPREF ? ppobj_get_stream(ppref_obj((o)->ref)) : NULL)) +#define ppobj_rget_ref(o) ((o)->type == PPREF ? (o)->ref : ((o)->type == PPREF ? ppobj_get_ref(ppref_obj((o)->ref)) : NULL)) + +#define ppobj_get_bool_value(o) ((o)->type == PPBOOL ? ((o)->integer != 0) : 0) +#define ppobj_get_int_value(o) ((o)->type == PPINT ? (o)->integer : 0) +#define ppobj_get_num_value(o) ((o)->type == PPNUM ? (o)->number : ((o)->type == PPINT ? (ppnum)((o)->integer) : 0.0)) + +/* name */ + +#define ppname_is(name, s) (memcmp(name, s, sizeof("" s) - 1) == 0) +#define ppname_eq(name, n) (memcmp(name, s, ppname_size(name)) == 0) + +#define _ppname_ghost(name) (((const _ppname *)(name)) - 1) +#define ppname_size(name) (_ppname_ghost(name)->size) +#define ppname_exec(name) (_ppname_ghost(name)->flags & PPNAME_EXEC) + +#define PPNAME_ENCODED (1 << 0) +#define PPNAME_DECODED (1 << 1) +#define PPNAME_EXEC (1 << 1) + +PPAPI ppname ppname_decoded (ppname name); +PPAPI ppname ppname_encoded (ppname name); + +/* string */ + +#define _ppstring_ghost(string) (((const _ppstring *)(string)) - 1) +#define ppstring_size(string) (_ppstring_ghost(string)->size) + +#define PPSTRING_ENCODED (1 << 0) +#define PPSTRING_DECODED (1 << 1) +//#define PPSTRING_EXEC (1 << 2) // postscript only +#define PPSTRING_PLAIN 0 +#define PPSTRING_BASE16 (1 << 3) +#define PPSTRING_BASE85 (1 << 4) +#define PPSTRING_UTF16BE (1 << 5) +#define PPSTRING_UTF16LE (1 << 6) + +#define ppstring_type(string) (_ppstring_ghost(string)->flags & (PPSTRING_BASE16|PPSTRING_BASE85)) +#define ppstring_hex(string) (_ppstring_ghost(string)->flags & PPSTRING_BASE16) +#define ppstring_utf(string) (_ppstring_ghost(string)->flags & (PPSTRING_UTF16BE|PPSTRING_UTF16LE)) + +PPAPI ppstring ppstring_decoded (ppstring string); +PPAPI ppstring ppstring_encoded (ppstring string); + +/* array */ + +#define pparray_size(array) ((array)->size) +#define pparray_at(array, index) ((array)->data + index) + +#define pparray_first(array, index, obj) ((index) = 0, (obj) = pparray_at(array, 0)) +#define pparray_next(index, obj) (++(index), ++(obj)) + +#define pparray_get(array, index) (index < (array)->size ? pparray_at(array, index) : NULL) + +PPAPI ppobj * pparray_get_obj (pparray *array, size_t index); +PPAPI int pparray_get_bool (pparray *array, size_t index, int *v); +PPAPI int pparray_get_int (pparray *array, size_t index, ppint *v); +PPAPI int pparray_get_uint (pparray *array, size_t index, ppuint *v); +PPAPI int pparray_get_num (pparray *array, size_t index, ppnum *v); +PPAPI ppname pparray_get_name (pparray *array, size_t index); +PPAPI ppstring pparray_get_string (pparray *array, size_t index); +PPAPI pparray * pparray_get_array (pparray *array, size_t index); +PPAPI ppdict * pparray_get_dict (pparray *array, size_t index); +//PPAPI ppstream * pparray_get_stream (pparray *array, size_t index); +PPAPI ppref * pparray_get_ref (pparray *array, size_t index); + +PPAPI ppobj * pparray_rget_obj (pparray *array, size_t index); +PPAPI int pparray_rget_bool (pparray *array, size_t index, int *v); +PPAPI int pparray_rget_int (pparray *array, size_t index, ppint *v); +PPAPI int pparray_rget_uint (pparray *array, size_t index, ppuint *v); +PPAPI int pparray_rget_num (pparray *array, size_t index, ppnum *v); +PPAPI ppname pparray_rget_name (pparray *array, size_t index); +PPAPI ppstring pparray_rget_string (pparray *array, size_t index); +PPAPI pparray * pparray_rget_array (pparray *array, size_t index); +PPAPI ppdict * pparray_rget_dict (pparray *array, size_t index); +PPAPI ppstream * pparray_rget_stream (pparray *array, size_t index); +PPAPI ppref * pparray_rget_ref (pparray *array, size_t index); + +/* dict */ + +#define ppdict_size(dict) ((dict)->size) +#define ppdict_at(dict, index) ((dict)->data + index) +#define ppdict_key(dict, index) ((dict)->keys[index]) + +PPAPI ppobj * ppdict_get_obj (ppdict *dict, const char *name); +PPAPI int ppdict_get_bool (ppdict *dict, const char *name, int *v); +PPAPI int ppdict_get_int (ppdict *dict, const char *name, ppint *v); +PPAPI int ppdict_get_uint (ppdict *dict, const char *name, ppuint *v); +PPAPI int ppdict_get_num (ppdict *dict, const char *name, ppnum *v); +PPAPI ppname ppdict_get_name (ppdict *dict, const char *name); +PPAPI ppstring ppdict_get_string (ppdict *dict, const char *name); +PPAPI pparray * ppdict_get_array (ppdict *dict, const char *name); +PPAPI ppdict * ppdict_get_dict (ppdict *dict, const char *name); +//PPAPI ppstream * ppdict_get_stream (ppdict *dict, const char *name); +PPAPI ppref * ppdict_get_ref (ppdict *dict, const char *name); + +PPAPI ppobj * ppdict_rget_obj (ppdict *dict, const char *name); +PPAPI int ppdict_rget_bool (ppdict *dict, const char *name, int *v); +PPAPI int ppdict_rget_int (ppdict *dict, const char *name, ppint *v); +PPAPI int ppdict_rget_uint (ppdict *dict, const char *name, ppuint *v); +PPAPI int ppdict_rget_num (ppdict *dict, const char *name, ppnum *v); +PPAPI ppname ppdict_rget_name (ppdict *dict, const char *name); +PPAPI ppstring ppdict_rget_string (ppdict *dict, const char *name); +PPAPI pparray * ppdict_rget_array (ppdict *dict, const char *name); +PPAPI ppdict * ppdict_rget_dict (ppdict *dict, const char *name); +PPAPI ppstream * ppdict_rget_stream (ppdict *dict, const char *name); +PPAPI ppref * ppdict_rget_ref (ppdict *dict, const char *name); + +#define ppdict_first(dict, pkey, obj) (pkey = (dict)->keys, obj = (dict)->data) +#define ppdict_next(pkey, obj) (++(pkey), ++(obj)) + +/* stream */ + +#define ppstream_dict(stream) ((stream)->dict) + +PPAPI uint8_t * ppstream_first (ppstream *stream, size_t *size, int decode); +PPAPI uint8_t * ppstream_next (ppstream *stream, size_t *size); +PPAPI uint8_t * ppstream_all (ppstream *stream, size_t *size, int decode); +PPAPI void ppstream_done (ppstream *stream); + +PPAPI void ppstream_init_buffers (void); +PPAPI void ppstream_free_buffers (void); + +/* ref */ + +#define ppref_obj(ref) (&(ref)->object) + +/* xref */ + +PPAPI ppxref * ppdoc_xref (ppdoc *pdf); +PPAPI ppxref * ppxref_prev (ppxref *xref); +PPAPI ppdict * ppxref_trailer (ppxref *xref); +PPAPI ppdict * ppxref_catalog (ppxref *xref); +PPAPI ppdict * ppxref_info (ppxref *xref); +PPAPI ppref * ppxref_pages (ppxref *xref); +PPAPI ppref * ppxref_find (ppxref *xref, ppuint refnumber); + +/* doc */ + +PPAPI ppdoc * ppdoc_load (const char *filename); +PPAPI ppdoc * ppdoc_mem (const void *data, size_t size); +PPAPI void ppdoc_free (ppdoc *pdf); + +#define ppdoc_trailer(pdf) ppxref_trailer(ppdoc_xref(pdf)) +#define ppdoc_catalog(pdf) ppxref_catalog(ppdoc_xref(pdf)) +#define ppdoc_info(pdf) ppxref_info(ppdoc_xref(pdf)) +#define ppdoc_pages(pdf) ppxref_pages(ppdoc_xref(pdf)) + +PPAPI ppuint ppdoc_page_count (ppdoc *pdf); +PPAPI ppref * ppdoc_page (ppdoc *pdf, ppuint index); +PPAPI ppref * ppdoc_first_page (ppdoc *pdf); +PPAPI ppref * ppdoc_next_page (ppdoc *pdf); + +PPAPI ppstream * ppcontents_first (ppdict *dict); +PPAPI ppstream * ppcontents_next (ppdict *dict, ppstream *stream); + +/* crypt */ + +typedef enum { + PPCRYPT_NONE = 0, + PPCRYPT_DONE = 1, + PPCRYPT_FAIL = -1, + PPCRYPT_PASS = -2 +} ppcrypt_status; + +PPAPI ppcrypt_status ppdoc_crypt_status (ppdoc *pdf); +PPAPI ppcrypt_status ppdoc_crypt_pass (ppdoc *pdf, const void *userpass, size_t userpasslength, const void *ownerpass, size_t ownerpasslength); + +/* permission flags, effect in Acrobat File -> Properties -> Security tab */ + +PPAPI ppint ppdoc_permissions (ppdoc *pdf); + +#define PPDOC_ALLOW_PRINT (1<<2) // printing +#define PPDOC_ALLOW_MODIFY (1<<3) // filling form fields, signing, creating template pages +#define PPDOC_ALLOW_COPY (1<<4) // copying, copying for accessibility +#define PPDOC_ALLOW_ANNOTS (1<<5) // filling form fields, copying, signing +#define PPDOC_ALLOW_EXTRACT (1<<9) // contents copying for accessibility +#define PPDOC_ALLOW_ASSEMBLY (1<<10) // (no effect) +#define PPDOC_ALLOW_PRINT_HIRES (1<<11) // (no effect) + +/* context */ + +typedef struct ppcontext ppcontext; + +PPAPI ppcontext * ppcontext_new (void); +PPAPI void ppcontext_done (ppcontext *context); +PPAPI void ppcontext_free (ppcontext *context); + +/* contents parser */ + +PPAPI ppobj * ppcontents_first_op (ppcontext *context, ppstream *stream, size_t *psize, ppname *pname); +PPAPI ppobj * ppcontents_next_op (ppcontext *context, ppstream *stream, size_t *psize, ppname *pname); +PPAPI ppobj * ppcontents_parse (ppcontext *context, ppstream *stream, size_t *psize); + +/* boxes and transforms */ + +typedef struct { + ppnum lx, ly, rx, ry; +} pprect; + +PPAPI pprect * pparray_to_rect (pparray *array, pprect *rect); +PPAPI pprect * ppdict_get_rect (ppdict *dict, const char *name, pprect *rect); +PPAPI pprect * ppdict_get_box (ppdict *dict, const char *name, pprect *rect); + +typedef struct { + ppnum xx, xy, yx, yy, x, y; +} ppmatrix; + +PPAPI ppmatrix * pparray_to_matrix (pparray *array, ppmatrix *matrix); +PPAPI ppmatrix * ppdict_get_matrix (ppdict *dict, const char *name, ppmatrix *matrix); + +/* logger */ + +typedef void (*pplogger_callback) (const char *message, void *alien); +PPAPI void pplog_callback (pplogger_callback logger, void *alien); +PPAPI int pplog_prefix (const char *prefix); + +/* version */ + +PPAPI const char * ppdoc_version_string (ppdoc *pdf); +PPAPI int ppdoc_version_number (ppdoc *pdf, int *minor); + +/* doc info */ + +PPAPI size_t ppdoc_file_size (ppdoc *pdf); +PPAPI ppuint ppdoc_objects (ppdoc *pdf); +PPAPI size_t ppdoc_memory (ppdoc *pdf, size_t *waste); + +#endif diff --git a/Build/source/texk/web2c/luatexdir/luapplib/pparray.c b/Build/source/texk/web2c/luatexdir/luapplib/pparray.c new file mode 100644 index 00000000000..25e8c495038 --- /dev/null +++ b/Build/source/texk/web2c/luatexdir/luapplib/pparray.c @@ -0,0 +1,146 @@ + +#include "pplib.h" + +pparray * pparray_create (const ppobj *stackpos, size_t size, ppheap **pheap) +{ + pparray *array; + ppobj *data; + array = (pparray *)ppheap_take(pheap, sizeof(pparray) + size * sizeof(ppobj)); + array->size = size; + array->data = data = (ppobj *)(array + 1); + memcpy(data, stackpos, size * sizeof(ppobj)); + return array; +} + +ppobj * pparray_get_obj (pparray *array, size_t index) +{ + return pparray_get(array, index); +} + +ppobj * pparray_rget_obj (pparray *array, size_t index) +{ + ppobj *obj; + return (obj = pparray_get(array, index)) != NULL ? ppobj_rget_obj(obj) : NULL; +} + +int pparray_get_bool (pparray *array, size_t index, int *v) +{ + ppobj *obj; + return (obj = pparray_get(array, index)) != NULL ? ppobj_get_bool(obj, *v) : 0; +} + +int pparray_rget_bool (pparray *array, size_t index, int *v) +{ + ppobj *obj; + return (obj = pparray_get(array, index)) != NULL ? ppobj_rget_bool(obj, *v) : 0; +} + +int pparray_get_int (pparray *array, size_t index, ppint *v) +{ + ppobj *obj; + return (obj = pparray_get(array, index)) != NULL ? ppobj_get_int(obj, *v) : 0; +} + +int pparray_rget_int (pparray *array, size_t index, ppint *v) +{ + ppobj *obj; + return (obj = pparray_get(array, index)) != NULL ? ppobj_rget_int(obj, *v) : 0; +} + +int pparray_get_uint (pparray *array, size_t index, ppuint *v) +{ + ppobj *obj; + return (obj = pparray_get(array, index)) != NULL ? ppobj_get_uint(obj, *v) : 0; +} + +int pparray_rget_uint (pparray *array, size_t index, ppuint *v) +{ + ppobj *obj; + return (obj = pparray_get(array, index)) != NULL ? ppobj_rget_uint(obj, *v) : 0; +} + +int pparray_get_num (pparray *array, size_t index, ppnum *v) +{ + ppobj *obj; + return (obj = pparray_get(array, index)) != NULL ? ppobj_get_num(obj, *v) : 0; +} + +int pparray_rget_num (pparray *array, size_t index, ppnum *v) +{ + ppobj *obj; + return (obj = pparray_get(array, index)) != NULL ? ppobj_rget_num(obj, *v) : 0; +} + +ppname pparray_get_name (pparray *array, size_t index) +{ + ppobj *obj; + return (obj = pparray_get(array, index)) != NULL ? ppobj_get_name(obj) : NULL; +} + +ppname pparray_rget_name (pparray *array, size_t index) +{ + ppobj *obj; + return (obj = pparray_get(array, index)) != NULL ? ppobj_rget_name(obj) : NULL; +} + +ppstring pparray_get_string (pparray *array, size_t index) +{ + ppobj *obj; + return (obj = pparray_get(array, index)) != NULL ? ppobj_get_string(obj) : NULL; +} + +ppstring pparray_rget_string (pparray *array, size_t index) +{ + ppobj *obj; + return (obj = pparray_get(array, index)) != NULL ? ppobj_rget_string(obj) : NULL; +} + +pparray * pparray_get_array (pparray *array, size_t index) +{ + ppobj *obj; + return (obj = pparray_get(array, index)) != NULL ? ppobj_get_array(obj) : NULL; +} + +pparray * pparray_rget_array (pparray *array, size_t index) +{ + ppobj *obj; + return (obj = pparray_get(array, index)) != NULL ? ppobj_rget_array(obj) : NULL; +} + +ppdict * pparray_get_dict (pparray *array, size_t index) +{ + ppobj *obj; + return (obj = pparray_get(array, index)) != NULL ? ppobj_get_dict(obj) : NULL; +} + +ppdict * pparray_rget_dict (pparray *array, size_t index) +{ + ppobj *obj; + return (obj = pparray_get(array, index)) != NULL ? ppobj_rget_dict(obj) : NULL; +} + +/* +ppstream * pparray_get_stream (pparray *array, size_t index) +{ + ppobj *obj; + return (obj = pparray_get(array, index)) != NULL ? ppobj_get_stream(obj) : NULL; +} +*/ + +ppstream * pparray_rget_stream (pparray *array, size_t index) +{ + ppobj *obj; + return (obj = pparray_get(array, index)) != NULL ? ppobj_rget_stream(obj) : NULL; +} + +ppref * pparray_get_ref (pparray *array, size_t index) +{ + ppobj *obj; + return (obj = pparray_get(array, index)) != NULL ? ppobj_get_ref(obj) : NULL; +} + +ppref * pparray_rget_ref (pparray *array, size_t index) +{ + ppobj *obj; + return (obj = pparray_get(array, index)) != NULL ? ppobj_rget_ref(obj) : NULL; +} diff --git a/Build/source/texk/web2c/luatexdir/luapplib/pparray.h b/Build/source/texk/web2c/luatexdir/luapplib/pparray.h new file mode 100644 index 00000000000..6fdd8b8143f --- /dev/null +++ b/Build/source/texk/web2c/luatexdir/luapplib/pparray.h @@ -0,0 +1,7 @@ + +#ifndef PP_ARRAY_H +#define PP_ARRAY_H + +pparray * pparray_create (const ppobj *stack, size_t size, ppheap **pheap); + +#endif
\ No newline at end of file diff --git a/Build/source/texk/web2c/luatexdir/luapplib/ppconf.h b/Build/source/texk/web2c/luatexdir/luapplib/ppconf.h new file mode 100644 index 00000000000..57edc205a54 --- /dev/null +++ b/Build/source/texk/web2c/luatexdir/luapplib/ppconf.h @@ -0,0 +1,76 @@ + +#ifndef PP_CONF_H +#define PP_CONF_H + +/* +Aux flags: + PPDLL -- indicates a part of a shared library + PPEXE -- indicates a host program using shared library functions +*/ + +#if defined(_WIN32) || defined(_WIN64) +# ifdef PPDLL +# define PPAPI __declspec(dllexport) +# define PPDEF __declspec(dllexport) +# else +# ifdef PPEXE +# define PPAPI __declspec(dllimport) +# define PPDEF +# else +# define PPAPI +# define PPDEF +# endif +# endif +#else +# define PPAPI +# define PPDEF +#endif + +/* platform vs integers */ + +#if defined(_WIN32) || defined(WIN32) +# ifdef _MSC_VER +# if defined(_M_64) || defined(_WIN64) +# define MSVC64 +# else +# define MSVC32 +# endif +# else +# if defined(__MINGW64__) +# define MINGW64 +# else +# if defined(__MINGW32__) +# define MINGW32 +# endif +# endif +# endif +#endif + +#if defined(_WIN64) || defined(__MINGW32__) +# define PPINT64F "%I64d" +# define PPUINT64F "%I64u" +#else +# define PPINT64F "%lld" +# define PPUINT64F "%llu" +#endif + +#if defined(MSVC64) +# define PPINT(N) N##I64 +# define PPUINT(N) N##UI64 +# define PPINTF PPINT64F +# define PPUINTF PPUINT64F +#elif defined(MINGW64) +# define PPINT(N) N##LL +# define PPUINT(N) N##ULL +# define PPINTF PPINT64F +# define PPUINTF PPUINT64F +#else // 32bit or sane 64bit (LP64, where long is long indeed) +# define PPINT(N) N##L +# define PPUINT(N) N##UL +# define PPINTF "%ld" +# define PPUINTF "%lu" +#endif + +#define PPSIZEF PPUINTF + +#endif
\ No newline at end of file diff --git a/Build/source/texk/web2c/luatexdir/luapplib/ppcrypt.c b/Build/source/texk/web2c/luatexdir/luapplib/ppcrypt.c new file mode 100644 index 00000000000..780ec78be12 --- /dev/null +++ b/Build/source/texk/web2c/luatexdir/luapplib/ppcrypt.c @@ -0,0 +1,628 @@ + +#include "utilmd5.h" +#include "utilsha.h" + +#include "pplib.h" + +/* crypt struct */ + +#define CRYPT_AES (1<<0) +#define CRYPT_RC4 (1<<1) +#define CRYPT_MD (1<<2) +#define CRYPT_NOMD (1<<3) + +static ppcrypt * ppcrypt_create (ppheap **pheap) +{ + ppcrypt *crypt; + crypt = (ppcrypt *)ppheap_take(pheap, sizeof(ppcrypt)); + memset(crypt, 0, sizeof(ppcrypt)); + return crypt; +} + +static int ppcrypt_type (ppcrypt *crypt, ppname cryptname, ppuint *length, int *cryptflags) +{ + ppdict *filterdict; + ppname filtertype; + int cryptmd = 0, default256 = 0; + + if (crypt->map == NULL || (filterdict = ppdict_rget_dict(crypt->map, cryptname)) == NULL) + return 0; + if ((filtertype = ppdict_get_name(filterdict, "CFM")) == NULL) + return 0; + *cryptflags = 0; + if (ppname_is(filtertype, "V2")) + *cryptflags |= CRYPT_RC4; + else if (ppname_is(filtertype, "AESV2")) + *cryptflags |= CRYPT_AES; + else if (ppname_is(filtertype, "AESV3")) + *cryptflags |= CRYPT_AES, default256 = 1; + else + return 0; + /* pdf spec page. 134: /Length is said to be optional bit-length of the key, but it seems to be a mistake, as Acrobat + produces /Length key with bytes lengths, opposite to /Length key of the main encrypt dict. */ + if (length != NULL) + if (!ppdict_get_uint(filterdict, "Length", length)) + *length = (*cryptflags & CRYPT_RC4) ? 5 : (default256 ? 32 : 16); + /* one of metadata flags is set iff there is an explicit EncryptMetadata key */ + if (ppdict_get_bool(filterdict, "EncryptMetadata", &cryptmd)) + *cryptflags |= (cryptmd ? CRYPT_MD : CRYPT_NOMD); + return 1; +} + +static const uint8_t padding_string[] = { + 0x28, 0xBF, 0x4E, 0x5E, 0x4E, 0x75, 0x8A, 0x41, 0x64, 0x00, 0x4E, 0x56, 0xFF, 0xFA, 0x01, 0x08, + 0x2E, 0x2E, 0x00, 0xB6, 0xD0, 0x68, 0x3E, 0x80, 0x2F, 0x0C, 0xA9, 0xFE, 0x64, 0x53, 0x69, 0x7A +}; + +static void ppcrypt_set_userpass (ppcrypt *crypt, const void *userpass, size_t userpasslength) +{ + crypt->userpasslength = userpasslength > 32 ? 32 : userpasslength; + memcpy(crypt->userpass, userpass, crypt->userpasslength); + memcpy(crypt->userpass + crypt->userpasslength, padding_string, 32 - crypt->userpasslength); + crypt->flags |= PPCRYPT_USER_PASSWORD; +} + +static void ppcrypt_set_ownerpass (ppcrypt *crypt, const void *ownerpass, size_t ownerpasslength) +{ + crypt->ownerpasslength = ownerpasslength > 32 ? 32 : ownerpasslength; + memcpy(crypt->ownerpass, ownerpass, crypt->ownerpasslength); + memcpy(crypt->ownerpass + crypt->ownerpasslength, padding_string, 32 - crypt->ownerpasslength); + crypt->flags |= PPCRYPT_OWNER_PASSWORD; +} + +/* retrieving user password from owner password and owner key (variant < 5) */ + +static void ppcrypt_retrieve_userpass (ppcrypt *crypt, const void *ownerkey, size_t ownerkeysize) +{ + uint8_t temp[16], rc4key[32], rc4key2[32]; + uint8_t i; + ppuint k; + + md5init(); + md5add(crypt->ownerpass, 32); + md5put(rc4key); + if (crypt->algorithm_revision >= 3) + { + for (i = 0; i < 50; ++i) + { + pplib_md5(rc4key, 16, temp); + memcpy(rc4key, temp, 16); + } + } + rc4_decode_data(ownerkey, ownerkeysize, crypt->userpass, rc4key, crypt->filekeylength); + if (crypt->algorithm_revision >= 3) + { + for (i = 1; i <= 19; ++i) + { + for (k = 0; k < crypt->filekeylength; ++k) + rc4key2[k] = rc4key[k] ^ i; + rc4_decode_data(crypt->userpass, 32, crypt->userpass, rc4key2, crypt->filekeylength); + } + } + //crypt->userpasslength = 32; + for (crypt->userpasslength = 0; crypt->userpasslength < 32; ++crypt->userpasslength) + if (memcmp(&crypt->userpass[crypt->userpasslength], padding_string, 32 - crypt->userpasslength) == 0) + break; + crypt->flags |= PPCRYPT_USER_PASSWORD; +} + +/* generating file key; pdf spec p. 125 */ + +static void ppcrypt_filekey (ppcrypt *crypt, const void *ownerkey, size_t ownerkeysize, const void *id, size_t idsize) +{ + uint32_t p; + uint8_t permissions[4], temp[16]; + int i; + + md5init(); + md5add(crypt->userpass, 32); + md5add(ownerkey, ownerkeysize); + p = (uint32_t)crypt->permissions; + permissions[0] = get_byte1(p); + permissions[1] = get_byte2(p); + permissions[2] = get_byte3(p); + permissions[3] = get_byte4(p); + md5add(permissions, 4); + md5add(id, idsize); + if (crypt->algorithm_revision >= 4 && (crypt->flags & PPCRYPT_NO_METADATA)) + md5add("\xFF\xFF\xFF\xFF", 4); + md5put(crypt->filekey); + if (crypt->algorithm_revision >= 3) + { + for (i = 0; i < 50; ++i) + { + pplib_md5(crypt->filekey, (size_t)crypt->filekeylength, temp); + memcpy(crypt->filekey, temp, 16); + } + } +} + +/* generating userkey for comparison with /U; requires a general file key and id; pdf spec page 126-127 */ + +static void ppcrypt_userkey (ppcrypt *crypt, const void *id, size_t idsize, uint8_t *password_hash) +{ + uint8_t rc4key2[32]; + uint8_t i; + ppuint k; + + if (crypt->algorithm_revision <= 2) + { + rc4_encode_data(padding_string, 32, password_hash, crypt->filekey, crypt->filekeylength); + } + else + { + md5init(); + md5add(padding_string, 32); + md5add(id, idsize); + md5put(password_hash); + rc4_encode_data(password_hash, 16, password_hash, crypt->filekey, crypt->filekeylength); + for (i = 1; i <= 19; ++i) + { + for (k = 0; k < crypt->filekeylength; ++k) + rc4key2[k] = crypt->filekey[k] ^ i; + rc4_encode_data(password_hash, 16, password_hash, rc4key2, crypt->filekeylength); + } + for (i = 16; i < 32; ++i) + password_hash[i] = password_hash[i - 16] ^ i; /* arbitrary 16-bytes padding */ + } +} + +/* validating /Perms key (pdf 1.7, /V 5 /R 5 crypt) */ + +static const uint8_t nulliv[16] = {0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0}; /* AES-256 initialization vector */ + +static ppcrypt_status ppcrypt_authenticate_perms (ppcrypt *crypt, ppstring perms) +{ /* decode /Perms string overriding crypt setup (should match anyway) */ + uint8_t permsdata[16]; + //int64_t p; + //int i; + + aes_decode_data(perms, ppstring_size(perms), permsdata, crypt->filekey, crypt->filekeylength, nulliv, AES_NULL_PADDING); + + if (permsdata[9] != 'a' || permsdata[10] != 'd' || permsdata[11] != 'b') + return PPCRYPT_FAIL; + + // do not update permissions flags; they seem to be different inside crypt string + //for (p = 0, i = 0; i < 8; ++i) + // p = p + (permsdata[i] << (i << 3)); /* low order bytes first */ + //crypt->permissions = (ppint)(int32_t)(p & 0x00000000FFFFFFFFLL); /* unset bits 33..64, treat as 32-bit signed int */ + + if (permsdata[8] == 'T') + crypt->flags &= ~PPCRYPT_NO_METADATA; + else if (permsdata[8] == 'F') + crypt->flags |= PPCRYPT_NO_METADATA; + + return PPCRYPT_DONE; +} + +ppcrypt_status ppdoc_crypt_init (ppdoc *pdf, const void *userpass, size_t userpasslength, const void *ownerpass, size_t ownerpasslength) +{ + ppcrypt *crypt; + ppdict *trailer, *encrypt; + ppobj *obj; + ppname name, *pkey; + ppstring userkey, ownerkey, userkey_e = NULL, ownerkey_e = NULL; + size_t hashlength; + pparray *idarray; + ppstring id = NULL, perms = NULL; + int cryptflags, encryptmd; + size_t strkeylength, stmkeylength; + + uint8_t password_hash[32]; /* /U and /O are 48 bytes strings for AES-256, but here we use only 32 */ + uint8_t *validation_salt, *key_salt; + + /* Every xref could theoretically have a separate encryption info. Not clarified in pdf spec but it seems that the top + level xref encryption info is the one to be applied to all objects in all xrefs, including older. */ + trailer = ppxref_trailer(pdf->xref); + if ((obj = ppdict_get_obj(trailer, "Encrypt")) == NULL) + return PPCRYPT_NONE; + /* Typically this is all done early, before loading body, so if /Encrypt is indirect reference, it points nothing. We have to load it here. */ + obj = ppobj_preloaded(pdf, obj); + if (obj->type != PPDICT) + return PPCRYPT_FAIL; + encrypt = obj->dict; + for (ppdict_first(encrypt, pkey, obj); *pkey != NULL; ppdict_next(pkey, obj)) + (void)ppobj_preloaded(pdf, obj); + + if ((name = ppdict_get_name(encrypt, "Filter")) != NULL && !ppname_is(name, "Standard")) + return PPCRYPT_FAIL; + + if ((crypt = pdf->crypt) == NULL) + crypt = pdf->crypt = ppcrypt_create(&pdf->heap); + if (!ppdict_get_uint(encrypt, "V", &crypt->algorithm_variant)) + crypt->algorithm_variant = 0; + if (crypt->algorithm_variant < 1 || crypt->algorithm_variant > 5) + return PPCRYPT_FAIL; + if (!ppdict_get_uint(encrypt, "R", &crypt->algorithm_revision)) + return PPCRYPT_FAIL; + if (crypt->algorithm_revision >= 3) + crypt->flags |= PPCRYPT_OBSCURITY; + if (!ppdict_get_int(encrypt, "P", &crypt->permissions)) + return PPCRYPT_FAIL; + if ((userkey = ppdict_get_string(encrypt, "U")) == NULL || (ownerkey = ppdict_get_string(encrypt, "O")) == NULL) + return PPCRYPT_FAIL; + userkey = ppstring_decoded(userkey); + ownerkey = ppstring_decoded(ownerkey); + /* for some reason acrobat pads /O and /U to 127 bytes with NULL, so we don't check the exact length but ensure the minimal */ + hashlength = crypt->algorithm_variant < 5 ? 32 : 48; + if (ppstring_size(userkey) < hashlength || ppstring_size(ownerkey) < hashlength) + return PPCRYPT_FAIL; + if (crypt->algorithm_variant < 5) + { // get first string from /ID (must not be ref) + if ((idarray = ppdict_get_array(trailer, "ID")) == NULL || (id = pparray_get_string(idarray, 0)) == NULL) + return PPCRYPT_FAIL; + id = ppstring_decoded(id); + } + else + { + if ((userkey_e = ppdict_get_string(encrypt, "UE")) == NULL || (ownerkey_e = ppdict_get_string(encrypt, "OE")) == NULL) + return PPCRYPT_FAIL; + userkey_e = ppstring_decoded(userkey_e); + ownerkey_e = ppstring_decoded(ownerkey_e); + if (ppstring_size(userkey_e) < 32 || ppstring_size(ownerkey_e) < 32) + return PPCRYPT_FAIL; + if ((perms = ppdict_get_string(encrypt, "Perms")) == NULL) + return PPCRYPT_FAIL; + perms = ppstring_decoded(perms); + if (ppstring_size(perms) != 16) + return PPCRYPT_FAIL; + } + + switch (crypt->algorithm_revision) + { + case 1: + crypt->filekeylength = 5; + crypt->flags |= PPCRYPT_RC4; + break; + case 2: case 3: + if (ppdict_get_uint(encrypt, "Length", &crypt->filekeylength)) + crypt->filekeylength >>= 3; /* 40..256 bits, 5..32 bytes*/ + else + crypt->filekeylength = 5; /* 40 bits, 5 bytes */ + crypt->flags |= PPCRYPT_RC4; + break; + case 4: case 5: + if ((crypt->map = ppdict_rget_dict(encrypt, "CF")) == NULL) + return PPCRYPT_FAIL; + for (ppdict_first(crypt->map, pkey, obj); *pkey != NULL; ppdict_next(pkey, obj)) + (void)ppobj_preloaded(pdf, obj); + /* /EncryptMetadata relevant only for version >=4, may be also provided in crypt filter dictionary; which takes a precedence then? + we assume that if there is an explicit EncryptMetadata key, it overrides main encrypt dict flag or default flag (the default is true, + meaning that Metadata stream is encrypted as others) */ + if (ppdict_get_bool(encrypt, "EncryptMetadata", &encryptmd) && !encryptmd) + crypt->flags |= PPCRYPT_NO_METADATA; + + strkeylength = stmkeylength = 0; + /* streams filter */ + if ((name = ppdict_get_name(encrypt, "StmF")) != NULL && ppcrypt_type(crypt, name, &stmkeylength, &cryptflags)) + { + if (cryptflags & CRYPT_AES) + crypt->flags |= PPCRYPT_STREAM_AES; + else if (cryptflags & CRYPT_RC4) + crypt->flags |= PPCRYPT_STREAM_RC4; + if (cryptflags & CRYPT_NOMD) + crypt->flags |= PPCRYPT_NO_METADATA; + else if (cryptflags & CRYPT_MD) + crypt->flags &= ~PPCRYPT_NO_METADATA; + } /* else identity */ + /* strings filter */ + if ((name = ppdict_get_name(encrypt, "StrF")) != NULL && ppcrypt_type(crypt, name, &strkeylength, &cryptflags)) + { + if (cryptflags & CRYPT_AES) + crypt->flags |= PPCRYPT_STRING_AES; + else if (cryptflags & CRYPT_RC4) + crypt->flags |= PPCRYPT_STRING_RC4; + } /* else identity */ + + /* /Length of encrypt dict is irrelevant here, theoretically every crypt filter may have own length... It means that we should + actually keep a different file key for streams and strings. But it leads to nonsense, as /U and /O entries refers to a single + keylength, without a distinction for strings/streams. So we have to assume /Length is consistent. To expose the limitation: */ + if ((crypt->flags & PPCRYPT_STREAM) && (crypt->flags & PPCRYPT_STRING)) + if (strkeylength != stmkeylength) + return PPCRYPT_FAIL; + crypt->filekeylength = stmkeylength ? stmkeylength : strkeylength; + if ((crypt->flags & PPCRYPT_STREAM) || (crypt->flags & PPCRYPT_STRING)) + if (crypt->filekeylength == 0) + return PPCRYPT_FAIL; + break; + default: + return PPCRYPT_FAIL; + } + + /* password */ + + if (userpass != NULL) + { + ppcrypt_set_userpass(crypt, userpass, userpasslength); + } + else if (ownerpass != NULL) + { + if (crypt->algorithm_variant < 5) // fetch user password from owner password + ppcrypt_retrieve_userpass(crypt, ownerkey, ppstring_size(ownerkey)); + else // open the document using owner password + ppcrypt_set_ownerpass(crypt, ownerpass, ownerpasslength); + } + else + { + return PPCRYPT_FAIL; + } + + if (crypt->algorithm_variant < 5) + { /* authenticate by comparing a generated vs present /U entry; depending on variant 16 or 32 bytes to compare */ + ppcrypt_filekey(crypt, ownerkey, ppstring_size(ownerkey), id, ppstring_size(id)); + ppcrypt_userkey(crypt, id, ppstring_size(id), password_hash); /* needs file key so comes after key generation */ + if (memcmp(userkey, password_hash, (crypt->algorithm_revision >= 3 ? 16 : 32)) == 0) + return PPCRYPT_DONE; + return PPCRYPT_PASS; + } + if (crypt->flags & PPCRYPT_USER_PASSWORD) + { + validation_salt = (uint8_t *)userkey + 32; + key_salt = validation_salt + 8; + sha256init(); + sha256add(crypt->userpass, crypt->userpasslength); + sha256add(validation_salt, 8); + sha256put(password_hash); + if (memcmp(userkey, password_hash, 32) != 0) + return PPCRYPT_PASS; + sha256init(); + sha256add(crypt->userpass, crypt->userpasslength); + sha256add(key_salt, 8); + sha256put(password_hash); + aes_decode_data(userkey_e, 32, crypt->filekey, password_hash, 32, nulliv, AES_NULL_PADDING); + return ppcrypt_authenticate_perms(crypt, perms); + } + if (crypt->flags & PPCRYPT_OWNER_PASSWORD) + { + validation_salt = (uint8_t *)ownerkey + 32; + key_salt = validation_salt + 8; + sha256init(); + sha256add(crypt->ownerpass, crypt->ownerpasslength); + sha256add(validation_salt, 8); + sha256add(userkey, 48); + sha256put(password_hash); + if (memcmp(ownerkey, password_hash, 32) != 0) + return PPCRYPT_PASS; + sha256init(); + sha256add(crypt->ownerpass, crypt->ownerpasslength); + sha256add(key_salt, 8); + sha256add(userkey, 48); + sha256put(password_hash); + aes_decode_data(ownerkey_e, 32, crypt->filekey, password_hash, 32, nulliv, AES_NULL_PADDING); + return ppcrypt_authenticate_perms(crypt, perms); + } + return PPCRYPT_FAIL; // should never get here +} + +/* decrypting strings */ + +/* +Since strings are generally rare, but might occur in mass (name trees). We generate decryption key when needed. +All strings within the same reference are crypted with the same key. Both RC4 and AES algorithms expands +the crypt key in some way and the result of expansion is the same for the same crypt key. Instead of recreating +the ky for every string, we backup the initial decryption state. +*/ + +static void ppcrypt_strkey (ppcrypt *crypt, ppref *ref, int aes) +{ + if (crypt->cryptkeylength > 0) + { /* crypt key already generated, just reinitialize crypt states */ + if (aes) + { /* aes codecs that works on c-strings do not modify aes_state flags at all, so we actually don't need to revitalize the state, + we only rewrite an initialization vector, which is modified during crypt procedure */ + } + else + { /* rc4 crypt map is modified during crypt procedure, so here we reinitialize rc4 bytes map */ + rc4_map_restore(&crypt->rc4state, &crypt->rc4copy); + } + return; + } + + if (crypt->algorithm_variant < 5) + { + crypt->filekey[crypt->filekeylength + 0] = get_byte1(ref->number); + crypt->filekey[crypt->filekeylength + 1] = get_byte2(ref->number); + crypt->filekey[crypt->filekeylength + 2] = get_byte3(ref->number); + crypt->filekey[crypt->filekeylength + 3] = get_byte1(ref->version); + crypt->filekey[crypt->filekeylength + 4] = get_byte2(ref->version); + + if (aes) + { + crypt->filekey[crypt->filekeylength + 5] = 0x73; + crypt->filekey[crypt->filekeylength + 6] = 0x41; + crypt->filekey[crypt->filekeylength + 7] = 0x6C; + crypt->filekey[crypt->filekeylength + 8] = 0x54; + } + + pplib_md5(crypt->filekey, crypt->filekeylength + (aes ? 9 : 5), crypt->cryptkey); + crypt->cryptkeylength = crypt->filekeylength + 5 >= 16 ? 16 : crypt->filekeylength + 5; + } + else + { + memcpy(crypt->cryptkey, crypt->filekey, 32); + crypt->cryptkeylength = 32; + } + + if (aes) + { + aes_decode_initialize(&crypt->aesstate, &crypt->aeskeyblock, crypt->cryptkey, crypt->cryptkeylength, NULL); + aes_pdf_mode(&crypt->aesstate); + } + else + { + rc4_state_initialize(&crypt->rc4state, &crypt->rc4map, crypt->cryptkey, crypt->cryptkeylength); + rc4_map_save(&crypt->rc4state, &crypt->rc4copy); + } +} + +int ppstring_decrypt (ppcrypt *crypt, const void *input, size_t size, void *output, size_t *newsize) +{ + int aes, rc4; + aes = crypt->flags & PPCRYPT_STRING_AES; + rc4 = crypt->flags & PPCRYPT_STRING_RC4; + if (aes || rc4) + { + ppcrypt_strkey(crypt, crypt->ref, aes); + if (aes) + *newsize = aes_decode_state_data(&crypt->aesstate, input, size, output); + else // if (rc4) + *newsize = rc4_decode_state_data(&crypt->rc4state, input, size, output); + return 1; + } + return 0; // identity crypt +} + +/* decrypting streams */ + +/* +Streams are decrypted everytime when accessing the stream data. We need to be able to get or make +the key for decryption as long as the stream is alive. And to get the key we need the reference +number and version, plus document crypt info. First thought was to keep the reference to which +the stream belongs; stream->ref and accessing the crypt info stream->ref->xref->pdf->crypt. +It would be ok as long as absolutelly nothing happens with ref and crypt. At some point pplib +may drift into rewriting support, which would imply ref/xref/crypt/pdf structures modifications. +So I feel better with generating a crypt key for every stream in encrypted document, paying a cost +of pplib_md5() for all streams, not necessarily those actually read. + +Key generation is the same as for strings, but different for distinct encryption methods (rc4 vs aes). +Since streams and strings might theoretically be encrypted with different filters. No reason to cacche +decryption state here. +*/ + +static ppstring ppcrypt_stmkey (ppcrypt *crypt, ppref *ref, int aes, ppheap **pheap) +{ + ppstring cryptkeystring; + //if (crypt->cryptkeylength > 0) + // return; + + if (crypt->algorithm_variant < 5) + { + crypt->filekey[crypt->filekeylength + 0] = get_byte1(ref->number); + crypt->filekey[crypt->filekeylength + 1] = get_byte2(ref->number); + crypt->filekey[crypt->filekeylength + 2] = get_byte3(ref->number); + crypt->filekey[crypt->filekeylength + 3] = get_byte1(ref->version); + crypt->filekey[crypt->filekeylength + 4] = get_byte2(ref->version); + + if (aes) + { + crypt->filekey[crypt->filekeylength + 5] = 0x73; + crypt->filekey[crypt->filekeylength + 6] = 0x41; + crypt->filekey[crypt->filekeylength + 7] = 0x6C; + crypt->filekey[crypt->filekeylength + 8] = 0x54; + } + + pplib_md5(crypt->filekey, crypt->filekeylength + (aes ? 9 : 5), crypt->cryptkey); + crypt->cryptkeylength = crypt->filekeylength + 5 >= 16 ? 16 : crypt->filekeylength + 5; // how about 256bits AES?? + } + else + { // we could actually generate this string once, but.. aes itself is way more expensive that we can earn here + memcpy(crypt->cryptkey, crypt->filekey, 32); // just for the record + crypt->cryptkeylength = 32; + } + cryptkeystring = ppstring_internal(crypt->cryptkey, crypt->cryptkeylength, pheap); + return ppstring_decoded(cryptkeystring); +} + +void ppstream_info (ppstream *stream, ppdoc *pdf) +{ // just after the stream creation + ppdict *dict; + ppobj *fobj, *pobj; + pparray *farray; + ppname fname, owncryptfilter, tname; + ppcrypt *crypt; + ppref *ref; + size_t i; + int owncrypt, cflags; + + dict = stream->dict; + ppdict_rget_uint(dict, "Length", &stream->length); + + if ((fobj = ppdict_get_obj(dict, "Filter")) != NULL) + { + fobj = ppobj_preloaded(pdf, fobj); + switch (fobj->type) + { + case PPNAME: + stream->flags |= PPSTREAM_COMPRESSED; + break; + case PPARRAY: + if (fobj->array->size > 0) + stream->flags |= PPSTREAM_COMPRESSED; + break; + default: + break; + } + } + if ((crypt = pdf->crypt) == NULL || (ref = crypt->ref) == NULL) + return; + owncrypt = 0; + owncryptfilter = NULL; + if (fobj != NULL) + { + if ((pobj = ppdict_get_obj(dict, "DecodeParms")) != NULL) + pobj = ppobj_preloaded(pdf, pobj); + switch (fobj->type) + { + case PPNAME: + if (ppname_is(fobj->name, "Crypt")) + { + owncrypt = 1; + if (pobj != NULL && pobj->type == PPDICT) // /Type /CryptFilterDecodeParms + owncryptfilter = ppdict_get_name(pobj->dict, "Name"); + } + break; + case PPARRAY: + farray = fobj->array; + for (i = 0; i < farray->size; ++i) + { + if ((fname = pparray_get_name(farray, i)) != NULL && ppname_is(fname, "Crypt")) + { + owncrypt = 1; + if (pobj != NULL && pobj->type == PPARRAY && (pobj = pparray_get_obj(pobj->array, i)) != NULL) + { + pobj = ppobj_preloaded(pdf, pobj); + if (pobj != NULL && pobj->type == PPDICT) // /Type /CryptFilterDecodeParms + owncryptfilter = ppdict_get_name(pobj->dict, "Name"); + } + break; + } + } + break; + default: + break; + } + } + + if (owncrypt) + { + stream->flags |= PPSTREAM_ENCRYPTED_OWN; + /* Seems a common habit to use just /Crypt filter name with no params, which defaults to /Identity. + A real example with uncompressed metadata: <</Filter[/Crypt]/Length 4217/Subtype/XML/Type/Metadata>> */ + if (owncryptfilter != NULL && !ppname_is(owncryptfilter, "Identity")) + { + if (crypt->map != NULL && ppcrypt_type(crypt, owncryptfilter, NULL, &cflags)) + { + if (cflags & CRYPT_AES) + stream->flags |= PPSTREAM_ENCRYPTED_AES; + else if (cflags & CRYPT_RC4) + stream->flags |= PPSTREAM_ENCRYPTED_RC4; + } + } + } + else + { + if ((crypt->flags & PPCRYPT_NO_METADATA) && (tname = ppdict_get_name(dict, "Type")) != NULL && ppname_is(tname, "Metadata")) + ; /* special treatment of metadata stream; we assume that explicit /Filter /Crypt setup overrides document level setup of EncryptMetadata. */ + else + { + if (crypt->flags & PPCRYPT_STREAM_RC4) + stream->flags |= PPSTREAM_ENCRYPTED_RC4; + else if (crypt->flags & PPCRYPT_STREAM_AES) + stream->flags |= PPSTREAM_ENCRYPTED_AES; + } + } + + /* finally, if the stream is encrypted with non-identity crypt (implicit or explicit), make and save the crypt key */ + if (stream->flags & PPSTREAM_ENCRYPTED) + stream->cryptkey = ppcrypt_stmkey(crypt, ref, ((stream->flags & PPSTREAM_ENCRYPTED_AES) != 0), &pdf->heap); +} diff --git a/Build/source/texk/web2c/luatexdir/luapplib/ppcrypt.c.orig b/Build/source/texk/web2c/luatexdir/luapplib/ppcrypt.c.orig new file mode 100644 index 00000000000..4428a1ccf08 --- /dev/null +++ b/Build/source/texk/web2c/luatexdir/luapplib/ppcrypt.c.orig @@ -0,0 +1,628 @@ + +#include "utilmd5.h" +#include "utilsha.h" + +#include "pplib.h" + +/* crypt struct */ + +#define CRYPT_AES (1<<0) +#define CRYPT_RC4 (1<<1) +#define CRYPT_MD (1<<2) +#define CRYPT_NOMD (1<<3) + +static ppcrypt * ppcrypt_create (ppheap **pheap) +{ + ppcrypt *crypt; + crypt = ppheap_take(pheap, sizeof(ppcrypt)); + memset(crypt, 0, sizeof(ppcrypt)); + return crypt; +} + +static int ppcrypt_type (ppcrypt *crypt, ppname cryptname, ppuint *length, int *cryptflags) +{ + ppdict *filterdict; + ppname filtertype; + int cryptmd = 0, default256 = 0; + + if (crypt->map == NULL || (filterdict = ppdict_rget_dict(crypt->map, cryptname)) == NULL) + return 0; + if ((filtertype = ppdict_get_name(filterdict, "CFM")) == NULL) + return 0; + *cryptflags = 0; + if (ppname_is(filtertype, "V2")) + *cryptflags |= CRYPT_RC4; + else if (ppname_is(filtertype, "AESV2")) + *cryptflags |= CRYPT_AES; + else if (ppname_is(filtertype, "AESV3")) + *cryptflags |= CRYPT_AES, default256 = 1; + else + return 0; + /* pdf spec page. 134: /Length is said to be optional bit-length of the key, but it seems to be a mistake, as Acrobat + produces /Length key with bytes lengths, opposite to /Length key of the main encrypt dict. */ + if (length != NULL) + if (!ppdict_get_uint(filterdict, "Length", length)) + *length = (*cryptflags & CRYPT_RC4) ? 5 : (default256 ? 32 : 16); + /* one of metadata flags is set iff there is an explicit EncryptMetadata key */ + if (ppdict_get_bool(filterdict, "EncryptMetadata", &cryptmd)) + *cryptflags |= (cryptmd ? CRYPT_MD : CRYPT_NOMD); + return 1; +} + +static const uint8_t padding_string[] = { + 0x28, 0xBF, 0x4E, 0x5E, 0x4E, 0x75, 0x8A, 0x41, 0x64, 0x00, 0x4E, 0x56, 0xFF, 0xFA, 0x01, 0x08, + 0x2E, 0x2E, 0x00, 0xB6, 0xD0, 0x68, 0x3E, 0x80, 0x2F, 0x0C, 0xA9, 0xFE, 0x64, 0x53, 0x69, 0x7A +}; + +static void ppcrypt_set_userpass (ppcrypt *crypt, const void *userpass, size_t userpasslength) +{ + crypt->userpasslength = userpasslength > 32 ? 32 : userpasslength; + memcpy(crypt->userpass, userpass, crypt->userpasslength); + memcpy(crypt->userpass + crypt->userpasslength, padding_string, 32 - crypt->userpasslength); + crypt->flags |= PPCRYPT_USER_PASSWORD; +} + +static void ppcrypt_set_ownerpass (ppcrypt *crypt, const void *ownerpass, size_t ownerpasslength) +{ + crypt->ownerpasslength = ownerpasslength > 32 ? 32 : ownerpasslength; + memcpy(crypt->ownerpass, ownerpass, crypt->ownerpasslength); + memcpy(crypt->ownerpass + crypt->ownerpasslength, padding_string, 32 - crypt->ownerpasslength); + crypt->flags |= PPCRYPT_OWNER_PASSWORD; +} + +/* retrieving user password from owner password and owner key (variant < 5) */ + +static void ppcrypt_retrieve_userpass (ppcrypt *crypt, const void *ownerkey, size_t ownerkeysize) +{ + uint8_t temp[16], rc4key[32], rc4key2[32]; + uint8_t i; + ppuint k; + + md5init(); + md5add(crypt->ownerpass, 32); + md5put(rc4key); + if (crypt->algorithm_revision >= 3) + { + for (i = 0; i < 50; ++i) + { + md5(rc4key, 16, temp); + memcpy(rc4key, temp, 16); + } + } + rc4_decode_data(ownerkey, ownerkeysize, crypt->userpass, rc4key, crypt->filekeylength); + if (crypt->algorithm_revision >= 3) + { + for (i = 1; i <= 19; ++i) + { + for (k = 0; k < crypt->filekeylength; ++k) + rc4key2[k] = rc4key[k] ^ i; + rc4_decode_data(crypt->userpass, 32, crypt->userpass, rc4key2, crypt->filekeylength); + } + } + //crypt->userpasslength = 32; + for (crypt->userpasslength = 0; crypt->userpasslength < 32; ++crypt->userpasslength) + if (memcmp(&crypt->userpass[crypt->userpasslength], padding_string, 32 - crypt->userpasslength) == 0) + break; + crypt->flags |= PPCRYPT_USER_PASSWORD; +} + +/* generating file key; pdf spec p. 125 */ + +static void ppcrypt_filekey (ppcrypt *crypt, const void *ownerkey, size_t ownerkeysize, const void *id, size_t idsize) +{ + uint32_t p; + uint8_t permissions[4], temp[16]; + int i; + + md5init(); + md5add(crypt->userpass, 32); + md5add(ownerkey, ownerkeysize); + p = (uint32_t)crypt->permissions; + permissions[0] = get_byte1(p); + permissions[1] = get_byte2(p); + permissions[2] = get_byte3(p); + permissions[3] = get_byte4(p); + md5add(permissions, 4); + md5add(id, idsize); + if (crypt->algorithm_revision >= 4 && (crypt->flags & PPCRYPT_NO_METADATA)) + md5add("\xFF\xFF\xFF\xFF", 4); + md5put(crypt->filekey); + if (crypt->algorithm_revision >= 3) + { + for (i = 0; i < 50; ++i) + { + md5(crypt->filekey, (size_t)crypt->filekeylength, temp); + memcpy(crypt->filekey, temp, 16); + } + } +} + +/* generating userkey for comparison with /U; requires a general file key and id; pdf spec page 126-127 */ + +static void ppcrypt_userkey (ppcrypt *crypt, const void *id, size_t idsize, uint8_t *password_hash) +{ + uint8_t rc4key2[32]; + uint8_t i; + ppuint k; + + if (crypt->algorithm_revision <= 2) + { + rc4_encode_data(padding_string, 32, password_hash, crypt->filekey, crypt->filekeylength); + } + else + { + md5init(); + md5add(padding_string, 32); + md5add(id, idsize); + md5put(password_hash); + rc4_encode_data(password_hash, 16, password_hash, crypt->filekey, crypt->filekeylength); + for (i = 1; i <= 19; ++i) + { + for (k = 0; k < crypt->filekeylength; ++k) + rc4key2[k] = crypt->filekey[k] ^ i; + rc4_encode_data(password_hash, 16, password_hash, rc4key2, crypt->filekeylength); + } + for (i = 16; i < 32; ++i) + password_hash[i] = password_hash[i - 16] ^ i; /* arbitrary 16-bytes padding */ + } +} + +/* validating /Perms key (pdf 1.7, /V 5 /R 5 crypt) */ + +static const uint8_t nulliv[16] = {0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0}; /* AES-256 initialization vector */ + +static ppcrypt_status ppcrypt_authenticate_perms (ppcrypt *crypt, ppstring perms) +{ /* decode /Perms string overriding crypt setup (should match anyway) */ + uint8_t permsdata[16]; + //int64_t p; + //int i; + + aes_decode_data(perms, ppstring_size(perms), permsdata, crypt->filekey, crypt->filekeylength, nulliv, AES_NULL_PADDING); + + if (permsdata[9] != 'a' || permsdata[10] != 'd' || permsdata[11] != 'b') + return PPCRYPT_FAIL; + + // do not update permissions flags; they seem to be different inside crypt string + //for (p = 0, i = 0; i < 8; ++i) + // p = p + (permsdata[i] << (i << 3)); /* low order bytes first */ + //crypt->permissions = (ppint)(int32_t)(p & 0x00000000FFFFFFFFLL); /* unset bits 33..64, treat as 32-bit signed int */ + + if (permsdata[8] == 'T') + crypt->flags &= ~PPCRYPT_NO_METADATA; + else if (permsdata[8] == 'F') + crypt->flags |= PPCRYPT_NO_METADATA; + + return PPCRYPT_DONE; +} + +ppcrypt_status ppdoc_crypt_init (ppdoc *pdf, const void *userpass, size_t userpasslength, const void *ownerpass, size_t ownerpasslength) +{ + ppcrypt *crypt; + ppdict *trailer, *encrypt; + ppobj *obj; + ppname name, *pkey; + ppstring userkey, ownerkey, userkey_e = NULL, ownerkey_e = NULL; + size_t hashlength; + pparray *idarray; + ppstring id = NULL, perms = NULL; + int cryptflags, encryptmd; + size_t strkeylength, stmkeylength; + + uint8_t password_hash[32]; /* /U and /O are 48 bytes strings for AES-256, but here we use only 32 */ + uint8_t *validation_salt, *key_salt; + + /* Every xref could theoretically have a separate encryption info. Not clarified in pdf spec but it seems that the top + level xref encryption info is the one to be applied to all objects in all xrefs, including older. */ + trailer = ppxref_trailer(pdf->xref); + if ((obj = ppdict_get_obj(trailer, "Encrypt")) == NULL) + return PPCRYPT_NONE; + /* Typically this is all done early, before loading body, so if /Encrypt is indirect reference, it points nothing. We have to load it here. */ + obj = ppobj_preloaded(pdf, obj); + if (obj->type != PPDICT) + return PPCRYPT_FAIL; + encrypt = obj->dict; + for (ppdict_first(encrypt, pkey, obj); *pkey != NULL; ppdict_next(pkey, obj)) + (void)ppobj_preloaded(pdf, obj); + + if ((name = ppdict_get_name(encrypt, "Filter")) != NULL && !ppname_is(name, "Standard")) + return PPCRYPT_FAIL; + + if ((crypt = pdf->crypt) == NULL) + crypt = pdf->crypt = ppcrypt_create(&pdf->heap); + if (!ppdict_get_uint(encrypt, "V", &crypt->algorithm_variant)) + crypt->algorithm_variant = 0; + if (crypt->algorithm_variant < 1 || crypt->algorithm_variant > 5) + return PPCRYPT_FAIL; + if (!ppdict_get_uint(encrypt, "R", &crypt->algorithm_revision)) + return PPCRYPT_FAIL; + if (crypt->algorithm_revision >= 3) + crypt->flags |= PPCRYPT_OBSCURITY; + if (!ppdict_get_int(encrypt, "P", &crypt->permissions)) + return PPCRYPT_FAIL; + if ((userkey = ppdict_get_string(encrypt, "U")) == NULL || (ownerkey = ppdict_get_string(encrypt, "O")) == NULL) + return PPCRYPT_FAIL; + userkey = ppstring_decoded(userkey); + ownerkey = ppstring_decoded(ownerkey); + /* for some reason acrobat pads /O and /U to 127 bytes with NULL, so we don't check the exact length but ensure the minimal */ + hashlength = crypt->algorithm_variant < 5 ? 32 : 48; + if (ppstring_size(userkey) < hashlength || ppstring_size(ownerkey) < hashlength) + return PPCRYPT_FAIL; + if (crypt->algorithm_variant < 5) + { // get first string from /ID (must not be ref) + if ((idarray = ppdict_get_array(trailer, "ID")) == NULL || (id = pparray_get_string(idarray, 0)) == NULL) + return PPCRYPT_FAIL; + id = ppstring_decoded(id); + } + else + { + if ((userkey_e = ppdict_get_string(encrypt, "UE")) == NULL || (ownerkey_e = ppdict_get_string(encrypt, "OE")) == NULL) + return PPCRYPT_FAIL; + userkey_e = ppstring_decoded(userkey_e); + ownerkey_e = ppstring_decoded(ownerkey_e); + if (ppstring_size(userkey_e) < 32 || ppstring_size(ownerkey_e) < 32) + return PPCRYPT_FAIL; + if ((perms = ppdict_get_string(encrypt, "Perms")) == NULL) + return PPCRYPT_FAIL; + perms = ppstring_decoded(perms); + if (ppstring_size(perms) != 16) + return PPCRYPT_FAIL; + } + + switch (crypt->algorithm_revision) + { + case 1: + crypt->filekeylength = 5; + crypt->flags |= PPCRYPT_RC4; + break; + case 2: case 3: + if (ppdict_get_uint(encrypt, "Length", &crypt->filekeylength)) + crypt->filekeylength >>= 3; /* 40..256 bits, 5..32 bytes*/ + else + crypt->filekeylength = 5; /* 40 bits, 5 bytes */ + crypt->flags |= PPCRYPT_RC4; + break; + case 4: case 5: + if ((crypt->map = ppdict_rget_dict(encrypt, "CF")) == NULL) + return PPCRYPT_FAIL; + for (ppdict_first(crypt->map, pkey, obj); *pkey != NULL; ppdict_next(pkey, obj)) + (void)ppobj_preloaded(pdf, obj); + /* /EncryptMetadata relevant only for version >=4, may be also provided in crypt filter dictionary; which takes a precedence then? + we assume that if there is an explicit EncryptMetadata key, it overrides main encrypt dict flag or default flag (the default is true, + meaning that Metadata stream is encrypted as others) */ + if (ppdict_get_bool(encrypt, "EncryptMetadata", &encryptmd) && !encryptmd) + crypt->flags |= PPCRYPT_NO_METADATA; + + strkeylength = stmkeylength = 0; + /* streams filter */ + if ((name = ppdict_get_name(encrypt, "StmF")) != NULL && ppcrypt_type(crypt, name, &stmkeylength, &cryptflags)) + { + if (cryptflags & CRYPT_AES) + crypt->flags |= PPCRYPT_STREAM_AES; + else if (cryptflags & CRYPT_RC4) + crypt->flags |= PPCRYPT_STREAM_RC4; + if (cryptflags & CRYPT_NOMD) + crypt->flags |= PPCRYPT_NO_METADATA; + else if (cryptflags & CRYPT_MD) + crypt->flags &= ~PPCRYPT_NO_METADATA; + } /* else identity */ + /* strings filter */ + if ((name = ppdict_get_name(encrypt, "StrF")) != NULL && ppcrypt_type(crypt, name, &strkeylength, &cryptflags)) + { + if (cryptflags & CRYPT_AES) + crypt->flags |= PPCRYPT_STRING_AES; + else if (cryptflags & CRYPT_RC4) + crypt->flags |= PPCRYPT_STRING_RC4; + } /* else identity */ + + /* /Length of encrypt dict is irrelevant here, theoretically every crypt filter may have own length... It means that we should + actually keep a different file key for streams and strings. But it leads to nonsense, as /U and /O entries refers to a single + keylength, without a distinction for strings/streams. So we have to assume /Length is consistent. To expose the limitation: */ + if ((crypt->flags & PPCRYPT_STREAM) && (crypt->flags & PPCRYPT_STRING)) + if (strkeylength != stmkeylength) + return PPCRYPT_FAIL; + crypt->filekeylength = stmkeylength ? stmkeylength : strkeylength; + if ((crypt->flags & PPCRYPT_STREAM) || (crypt->flags & PPCRYPT_STRING)) + if (crypt->filekeylength == 0) + return PPCRYPT_FAIL; + break; + default: + return PPCRYPT_FAIL; + } + + /* password */ + + if (userpass != NULL) + { + ppcrypt_set_userpass(crypt, userpass, userpasslength); + } + else if (ownerpass != NULL) + { + if (crypt->algorithm_variant < 5) // fetch user password from owner password + ppcrypt_retrieve_userpass(crypt, ownerkey, ppstring_size(ownerkey)); + else // open the document using owner password + ppcrypt_set_ownerpass(crypt, ownerpass, ownerpasslength); + } + else + { + return PPCRYPT_FAIL; + } + + if (crypt->algorithm_variant < 5) + { /* authenticate by comparing a generated vs present /U entry; depending on variant 16 or 32 bytes to compare */ + ppcrypt_filekey(crypt, ownerkey, ppstring_size(ownerkey), id, ppstring_size(id)); + ppcrypt_userkey(crypt, id, ppstring_size(id), password_hash); /* needs file key so comes after key generation */ + if (memcmp(userkey, password_hash, (crypt->algorithm_revision >= 3 ? 16 : 32)) == 0) + return PPCRYPT_DONE; + return PPCRYPT_PASS; + } + if (crypt->flags & PPCRYPT_USER_PASSWORD) + { + validation_salt = (uint8_t *)userkey + 32; + key_salt = validation_salt + 8; + sha256init(); + sha256add(crypt->userpass, crypt->userpasslength); + sha256add(validation_salt, 8); + sha256put(password_hash); + if (memcmp(userkey, password_hash, 32) != 0) + return PPCRYPT_PASS; + sha256init(); + sha256add(crypt->userpass, crypt->userpasslength); + sha256add(key_salt, 8); + sha256put(password_hash); + aes_decode_data(userkey_e, 32, crypt->filekey, password_hash, 32, nulliv, AES_NULL_PADDING); + return ppcrypt_authenticate_perms(crypt, perms); + } + if (crypt->flags & PPCRYPT_OWNER_PASSWORD) + { + validation_salt = (uint8_t *)ownerkey + 32; + key_salt = validation_salt + 8; + sha256init(); + sha256add(crypt->ownerpass, crypt->ownerpasslength); + sha256add(validation_salt, 8); + sha256add(userkey, 48); + sha256put(password_hash); + if (memcmp(ownerkey, password_hash, 32) != 0) + return PPCRYPT_PASS; + sha256init(); + sha256add(crypt->ownerpass, crypt->ownerpasslength); + sha256add(key_salt, 8); + sha256add(userkey, 48); + sha256put(password_hash); + aes_decode_data(ownerkey_e, 32, crypt->filekey, password_hash, 32, nulliv, AES_NULL_PADDING); + return ppcrypt_authenticate_perms(crypt, perms); + } + return PPCRYPT_FAIL; // should never get here +} + +/* decrypting strings */ + +/* +Since strings are generally rare, but might occur in mass (name trees). We generate decryption key when needed. +All strings within the same reference are crypted with the same key. Both RC4 and AES algorithms expands +the crypt key in some way and the result of expansion is the same for the same crypt key. Instead of recreating +the ky for every string, we backup the initial decryption state. +*/ + +static void ppcrypt_strkey (ppcrypt *crypt, ppref *ref, int aes) +{ + if (crypt->cryptkeylength > 0) + { /* crypt key already generated, just reinitialize crypt states */ + if (aes) + { /* aes codecs that works on c-strings do not modify aes_state flags at all, so we actually don't need to revitalize the state, + we only rewrite an initialization vector, which is modified during crypt procedure */ + } + else + { /* rc4 crypt map is modified during crypt procedure, so here we reinitialize rc4 bytes map */ + rc4_map_restore(&crypt->rc4state, &crypt->rc4copy); + } + return; + } + + if (crypt->algorithm_variant < 5) + { + crypt->filekey[crypt->filekeylength + 0] = get_byte1(ref->number); + crypt->filekey[crypt->filekeylength + 1] = get_byte2(ref->number); + crypt->filekey[crypt->filekeylength + 2] = get_byte3(ref->number); + crypt->filekey[crypt->filekeylength + 3] = get_byte1(ref->version); + crypt->filekey[crypt->filekeylength + 4] = get_byte2(ref->version); + + if (aes) + { + crypt->filekey[crypt->filekeylength + 5] = 0x73; + crypt->filekey[crypt->filekeylength + 6] = 0x41; + crypt->filekey[crypt->filekeylength + 7] = 0x6C; + crypt->filekey[crypt->filekeylength + 8] = 0x54; + } + + md5(crypt->filekey, crypt->filekeylength + (aes ? 9 : 5), crypt->cryptkey); + crypt->cryptkeylength = crypt->filekeylength + 5 >= 16 ? 16 : crypt->filekeylength + 5; + } + else + { + memcpy(crypt->cryptkey, crypt->filekey, 32); + crypt->cryptkeylength = 32; + } + + if (aes) + { + aes_decode_initialize(&crypt->aesstate, &crypt->aeskeyblock, crypt->cryptkey, crypt->cryptkeylength, NULL); + aes_pdf_mode(&crypt->aesstate); + } + else + { + rc4_state_initialize(&crypt->rc4state, &crypt->rc4map, crypt->cryptkey, crypt->cryptkeylength); + rc4_map_save(&crypt->rc4state, &crypt->rc4copy); + } +} + +int ppstring_decrypt (ppcrypt *crypt, const void *input, size_t size, void *output, size_t *newsize) +{ + int aes, rc4; + aes = crypt->flags & PPCRYPT_STRING_AES; + rc4 = crypt->flags & PPCRYPT_STRING_RC4; + if (aes || rc4) + { + ppcrypt_strkey(crypt, crypt->ref, aes); + if (aes) + *newsize = aes_decode_state_data(&crypt->aesstate, input, size, output); + else // if (rc4) + *newsize = rc4_decode_state_data(&crypt->rc4state, input, size, output); + return 1; + } + return 0; // identity crypt +} + +/* decrypting streams */ + +/* +Streams are decrypted everytime when accessing the stream data. We need to be able to get or make +the key for decryption as long as the stream is alive. And to get the key we need the reference +number and version, plus document crypt info. First thought was to keep the reference to which +the stream belongs; stream->ref and accessing the crypt info stream->ref->xref->pdf->crypt. +It would be ok as long as absolutelly nothing happens with ref and crypt. At some point pplib +may drift into rewriting support, which would imply ref/xref/crypt/pdf structures modifications. +So I feel better with generating a crypt key for every stream in encrypted document, paying a cost +of md5() for all streams, not necessarily those actually read. + +Key generation is the same as for strings, but different for distinct encryption methods (rc4 vs aes). +Since streams and strings might theoretically be encrypted with different filters. No reason to cacche +decryption state here. +*/ + +static ppstring ppcrypt_stmkey (ppcrypt *crypt, ppref *ref, int aes, ppheap **pheap) +{ + ppstring cryptkeystring; + //if (crypt->cryptkeylength > 0) + // return; + + if (crypt->algorithm_variant < 5) + { + crypt->filekey[crypt->filekeylength + 0] = get_byte1(ref->number); + crypt->filekey[crypt->filekeylength + 1] = get_byte2(ref->number); + crypt->filekey[crypt->filekeylength + 2] = get_byte3(ref->number); + crypt->filekey[crypt->filekeylength + 3] = get_byte1(ref->version); + crypt->filekey[crypt->filekeylength + 4] = get_byte2(ref->version); + + if (aes) + { + crypt->filekey[crypt->filekeylength + 5] = 0x73; + crypt->filekey[crypt->filekeylength + 6] = 0x41; + crypt->filekey[crypt->filekeylength + 7] = 0x6C; + crypt->filekey[crypt->filekeylength + 8] = 0x54; + } + + md5(crypt->filekey, crypt->filekeylength + (aes ? 9 : 5), crypt->cryptkey); + crypt->cryptkeylength = crypt->filekeylength + 5 >= 16 ? 16 : crypt->filekeylength + 5; // how about 256bits AES?? + } + else + { // we could actually generate this string once, but.. aes itself is way more expensive that we can earn here + memcpy(crypt->cryptkey, crypt->filekey, 32); // just for the record + crypt->cryptkeylength = 32; + } + cryptkeystring = ppstring_internal(crypt->cryptkey, crypt->cryptkeylength, pheap); + return ppstring_decoded(cryptkeystring); +} + +void ppstream_info (ppstream *stream, ppdoc *pdf) +{ // just after the stream creation + ppdict *dict; + ppobj *fobj, *pobj; + pparray *farray; + ppname fname, owncryptfilter, tname; + ppcrypt *crypt; + ppref *ref; + size_t i; + int owncrypt, cflags; + + dict = stream->dict; + ppdict_rget_uint(dict, "Length", &stream->length); + + if ((fobj = ppdict_get_obj(dict, "Filter")) != NULL) + { + fobj = ppobj_preloaded(pdf, fobj); + switch (fobj->type) + { + case PPNAME: + stream->flags |= PPSTREAM_COMPRESSED; + break; + case PPARRAY: + if (fobj->array->size > 0) + stream->flags |= PPSTREAM_COMPRESSED; + break; + default: + break; + } + } + if ((crypt = pdf->crypt) == NULL || (ref = crypt->ref) == NULL) + return; + owncrypt = 0; + owncryptfilter = NULL; + if (fobj != NULL) + { + if ((pobj = ppdict_get_obj(dict, "DecodeParms")) != NULL) + pobj = ppobj_preloaded(pdf, pobj); + switch (fobj->type) + { + case PPNAME: + if (ppname_is(fobj->name, "Crypt")) + { + owncrypt = 1; + if (pobj != NULL && pobj->type == PPDICT) // /Type /CryptFilterDecodeParms + owncryptfilter = ppdict_get_name(pobj->dict, "Name"); + } + break; + case PPARRAY: + farray = fobj->array; + for (i = 0; i < farray->size; ++i) + { + if ((fname = pparray_get_name(farray, i)) != NULL && ppname_is(fname, "Crypt")) + { + owncrypt = 1; + if (pobj != NULL && pobj->type == PPARRAY && (pobj = pparray_get_obj(pobj->array, i)) != NULL) + { + pobj = ppobj_preloaded(pdf, pobj); + if (pobj != NULL && pobj->type == PPDICT) // /Type /CryptFilterDecodeParms + owncryptfilter = ppdict_get_name(pobj->dict, "Name"); + } + break; + } + } + break; + default: + break; + } + } + + if (owncrypt) + { + stream->flags |= PPSTREAM_ENCRYPTED_OWN; + /* Seems a common habit to use just /Crypt filter name with no params, which defaults to /Identity. + A real example with uncompressed metadata: <</Filter[/Crypt]/Length 4217/Subtype/XML/Type/Metadata>> */ + if (owncryptfilter != NULL && !ppname_is(owncryptfilter, "Identity")) + { + if (crypt->map != NULL && ppcrypt_type(crypt, owncryptfilter, NULL, &cflags)) + { + if (cflags & CRYPT_AES) + stream->flags |= PPSTREAM_ENCRYPTED_AES; + else if (cflags & CRYPT_RC4) + stream->flags |= PPSTREAM_ENCRYPTED_RC4; + } + } + } + else + { + if ((crypt->flags & PPCRYPT_NO_METADATA) && (tname = ppdict_get_name(dict, "Type")) != NULL && ppname_is(tname, "Metadata")) + ; /* special treatment of metadata stream; we assume that explicit /Filter /Crypt setup overrides document level setup of EncryptMetadata. */ + else + { + if (crypt->flags & PPCRYPT_STREAM_RC4) + stream->flags |= PPSTREAM_ENCRYPTED_RC4; + else if (crypt->flags & PPCRYPT_STREAM_AES) + stream->flags |= PPSTREAM_ENCRYPTED_AES; + } + } + + /* finally, if the stream is encrypted with non-identity crypt (implicit or explicit), make and save the crypt key */ + if (stream->flags & PPSTREAM_ENCRYPTED) + stream->cryptkey = ppcrypt_stmkey(crypt, ref, ((stream->flags & PPSTREAM_ENCRYPTED_AES) != 0), &pdf->heap); +} diff --git a/Build/source/texk/web2c/luatexdir/luapplib/ppcrypt.h b/Build/source/texk/web2c/luatexdir/luapplib/ppcrypt.h new file mode 100644 index 00000000000..fc74cfe37ed --- /dev/null +++ b/Build/source/texk/web2c/luatexdir/luapplib/ppcrypt.h @@ -0,0 +1,61 @@ + +#ifndef PP_CRYPT_H +#define PP_CRYPT_H + +#include "ppfilter.h" +#include "utilcrypt.h" +#include "utilcryptdef.h" + +typedef struct { + ppuint algorithm_variant; /* /V entry of encrypt dict */ + ppuint algorithm_revision; /* /R entry of encrypt dict */ + ppint permissions; /* /P entry of encrypt dict */ + ppdict *map; /* /CF filters map of encrypt dict */ + uint8_t userpass[32]; /* padded user password */ + size_t userpasslength; /* the length of unpadded user password */ + uint8_t ownerpass[32]; /* padded owner password */ + size_t ownerpasslength; /* the length of unpadded owner password */ + uint8_t filekey[32+5+4]; /* generated file key with extra space of 5..9 bytes for salt */ + size_t filekeylength; /* key length; usually 5, 16 or 32 bytes */ + uint8_t cryptkey[32]; /* final crypt key for a given reference */ + size_t cryptkeylength; /* final crypt key length; usually keylength + 5 */ + ppref *ref; /* currently loaded ref (each ref may have a different key) */ + union { /* cached crypt states for strings encrypted/decrypted with the same key */ + struct { + rc4_state rc4state; + rc4_map rc4map; + rc4_map rc4copy; + }; + struct { + aes_state aesstate; + aes_keyblock aeskeyblock; + uint8_t ivcopy[16]; + }; + }; + int flags; +} ppcrypt; + +#define PPCRYPT_NO_METADATA (1<<0) +#define PPCRYPT_USER_PASSWORD (1<<1) +#define PPCRYPT_OWNER_PASSWORD (1<<2) +#define PPCRYPT_STREAM_RC4 (1<<3) +#define PPCRYPT_STRING_RC4 (1<<4) +#define PPCRYPT_STREAM_AES (1<<5) +#define PPCRYPT_STRING_AES (1<<6) +#define PPCRYPT_OBSCURITY (1<<7) + +#define PPCRYPT_STREAM (PPCRYPT_STREAM_AES|PPCRYPT_STREAM_RC4) +#define PPCRYPT_STRING (PPCRYPT_STRING_AES|PPCRYPT_STRING_RC4) +#define PPCRYPT_RC4 (PPCRYPT_STREAM_RC4|PPCRYPT_STRING_RC4) +#define PPCRYPT_AES (PPCRYPT_STREAM_AES|PPCRYPT_STRING_AES) + +ppcrypt_status ppdoc_crypt_init (ppdoc *pdf, const void *userpass, size_t userpasslength, const void *ownerpass, size_t ownerpasslength); +int ppstring_decrypt (ppcrypt *crypt, const void *input, size_t size, void *output, size_t *newsize); + +#define ppcrypt_start_ref(crypt, r) ((crypt)->ref = r, (crypt)->cryptkeylength = 0) +#define ppcrypt_end_ref(crypt) ((crypt)->ref = NULL, (crypt)->cryptkeylength = 0) +#define ppcrypt_ref(pdf, crypt) ((crypt = (pdf)->crypt) != NULL && crypt->ref != NULL) + +void ppstream_info (ppstream *stream, ppdoc *pdf); + +#endif
\ No newline at end of file diff --git a/Build/source/texk/web2c/luatexdir/luapplib/ppdict.c b/Build/source/texk/web2c/luatexdir/luapplib/ppdict.c new file mode 100644 index 00000000000..f04403d8350 --- /dev/null +++ b/Build/source/texk/web2c/luatexdir/luapplib/ppdict.c @@ -0,0 +1,163 @@ + +#include "pplib.h" + +ppdict * ppdict_create (const ppobj *stackpos, size_t size, ppheap **pheap) +{ + ppdict *dict; + ppobj *data; + ppname *pkey; + size_t i; + size >>= 1; + dict = (ppdict *)ppheap_take(pheap, sizeof(ppdict) + size * sizeof(ppobj) + (size + 1) * sizeof(ppname *)); // ? + size * sizeof(ppdict_map_node) + dict->size = 0; + dict->data = data = (ppobj *)(dict + 1); + dict->keys = pkey = (ppname *)(dict->data + size); + for (i = 0; i < size; ++i, stackpos += 2) + { + if (stackpos->type != PPNAME) // we need this check at lest for trailer hack + continue; + *pkey = stackpos->name; + *data = *(stackpos + 1); + ++pkey, ++data, ++dict->size; + } + *pkey = NULL; // sentinel for convinient iteration + return dict; +} + +ppobj * ppdict_get_obj (ppdict *dict, const char *name) +{ // some lookup? will see + ppname *pkey; + ppobj *obj; + for (ppdict_first(dict, pkey, obj); *pkey != NULL; ppdict_next(pkey, obj)) + if (strcmp(*pkey, name) == 0) // not ppname_eq() or ppname_is()!! + return obj; + return NULL; +} + +ppobj * ppdict_rget_obj (ppdict *dict, const char *name) +{ + ppobj *obj; + return (obj = ppdict_get_obj(dict, name)) != NULL ? ppobj_rget_obj(obj) : NULL; +} + +int ppdict_get_bool (ppdict *dict, const char *name, int *v) +{ + ppobj *obj; + return (obj = ppdict_get_obj(dict, name)) != NULL ? ppobj_get_bool(obj, *v) : 0; +} + +int ppdict_rget_bool (ppdict *dict, const char *name, int *v) +{ + ppobj *obj; + return (obj = ppdict_get_obj(dict, name)) != NULL ? ppobj_rget_bool(obj, *v) : 0; +} + +int ppdict_get_int (ppdict *dict, const char *name, ppint *v) +{ + ppobj *obj; + return (obj = ppdict_get_obj(dict, name)) != NULL ? ppobj_get_int(obj, *v) : 0; +} + +int ppdict_rget_int (ppdict *dict, const char *name, ppint *v) +{ + ppobj *obj; + return (obj = ppdict_get_obj(dict, name)) != NULL ? ppobj_rget_int(obj, *v) : 0; +} + +int ppdict_get_uint (ppdict *dict, const char *name, ppuint *v) +{ + ppobj *obj; + return (obj = ppdict_get_obj(dict, name)) != NULL ? ppobj_get_uint(obj, *v) : 0; +} + +int ppdict_rget_uint (ppdict *dict, const char *name, ppuint *v) +{ + ppobj *obj; + return (obj = ppdict_get_obj(dict, name)) != NULL ? ppobj_rget_uint(obj, *v) : 0; +} + +int ppdict_get_num (ppdict *dict, const char *name, ppnum *v) +{ + ppobj *obj; + return (obj = ppdict_get_obj(dict, name)) != NULL ? ppobj_get_num(obj, *v) : 0; +} + +int ppdict_rget_num (ppdict *dict, const char *name, ppnum *v) +{ + ppobj *obj; + return (obj = ppdict_get_obj(dict, name)) != NULL ? ppobj_rget_num(obj, *v) : 0; +} + +ppname ppdict_get_name (ppdict *dict, const char *name) +{ + ppobj *obj; + return (obj = ppdict_get_obj(dict, name)) != NULL ? ppobj_get_name(obj) : NULL; +} + +ppname ppdict_rget_name (ppdict *dict, const char *name) +{ + ppobj *obj; + return (obj = ppdict_get_obj(dict, name)) != NULL ? ppobj_rget_name(obj) : NULL; +} + +ppstring ppdict_get_string (ppdict *dict, const char *name) +{ + ppobj *obj; + return (obj = ppdict_get_obj(dict, name)) != NULL ? ppobj_get_string(obj) : NULL; +} + +ppstring ppdict_rget_string (ppdict *dict, const char *name) +{ + ppobj *obj; + return (obj = ppdict_get_obj(dict, name)) != NULL ? ppobj_rget_string(obj) : NULL; +} + +pparray * ppdict_get_array (ppdict *dict, const char *name) +{ + ppobj *obj; + return (obj = ppdict_get_obj(dict, name)) != NULL ? ppobj_get_array(obj) : NULL; +} + +pparray * ppdict_rget_array (ppdict *dict, const char *name) +{ + ppobj *obj; + return (obj = ppdict_get_obj(dict, name)) != NULL ? ppobj_rget_array(obj) : NULL; +} + +ppdict * ppdict_get_dict (ppdict *dict, const char *name) +{ + ppobj *obj; + return (obj = ppdict_get_obj(dict, name)) != NULL ? ppobj_get_dict(obj) : NULL; +} + +ppdict * ppdict_rget_dict (ppdict *dict, const char *name) +{ + ppobj *obj; + return (obj = ppdict_get_obj(dict, name)) != NULL ? ppobj_rget_dict(obj) : NULL; +} + +/* +ppstream * ppdict_get_stream (ppdict *dict, const char *name) +{ + ppobj *obj; + return (obj = ppdict_get_obj(dict, name)) != NULL ? ppobj_get_stream(obj) : NULL; +} +*/ + +ppstream * ppdict_rget_stream (ppdict *dict, const char *name) +{ + ppobj *obj; + return (obj = ppdict_get_obj(dict, name)) != NULL ? ppobj_rget_stream(obj) : NULL; +} + +ppref * ppdict_get_ref (ppdict *dict, const char *name) +{ + ppobj *obj; + return (obj = ppdict_get_obj(dict, name)) != NULL ? ppobj_get_ref(obj) : NULL; +} + +ppref * ppdict_rget_ref (ppdict *dict, const char *name) +{ + ppobj *obj; + return (obj = ppdict_get_obj(dict, name)) != NULL ? ppobj_rget_ref(obj) : NULL; +} diff --git a/Build/source/texk/web2c/luatexdir/luapplib/ppdict.h b/Build/source/texk/web2c/luatexdir/luapplib/ppdict.h new file mode 100644 index 00000000000..5a808b772c1 --- /dev/null +++ b/Build/source/texk/web2c/luatexdir/luapplib/ppdict.h @@ -0,0 +1,7 @@ + +#ifndef PP_DICT_H +#define PP_DICT_H + +ppdict * ppdict_create (const ppobj *stack, size_t size, ppheap **pheap); + +#endif
\ No newline at end of file diff --git a/Build/source/texk/web2c/luatexdir/luapplib/ppfilter.h b/Build/source/texk/web2c/luatexdir/luapplib/ppfilter.h new file mode 100644 index 00000000000..583aa8cf48f --- /dev/null +++ b/Build/source/texk/web2c/luatexdir/luapplib/ppfilter.h @@ -0,0 +1,10 @@ + +#ifndef PP_FILTER_H +#define PP_FILTER_H + +#include "utilbasexx.h" +#include "utilflate.h" +#include "utillzw.h" +#include "utilfpred.h" + +#endif
\ No newline at end of file diff --git a/Build/source/texk/web2c/luatexdir/luapplib/ppheap.c b/Build/source/texk/web2c/luatexdir/luapplib/ppheap.c new file mode 100644 index 00000000000..f05e1b7a5df --- /dev/null +++ b/Build/source/texk/web2c/luatexdir/luapplib/ppheap.c @@ -0,0 +1,424 @@ + +#include "pplib.h" + +#define PPHEAP_BUFFER 0xFFFF +#define PPHEAP_WASTE 0x00FF + +#define ppheap_head(heap) ((uint8_t *)((heap) + 1)) + + +#if defined __arm__ || defined __ARM__ || defined ARM || defined __ARM || defined __arm || defined __ARM_ARCH ||defined __aarch64__ +# define PPHEAP_ARCH_ARM +# define PPHEAP_NEED_ALIGNMENT +#endif + + + +#ifdef PPHEAP_NEED_ALIGNMENT +/* Tests has shown that long double seems to work: */ +/* for 32bit aligned_data has algn: 64 as ppxref and ppref, */ +/* the other algns divide algn: 64, so it's ok.*/ +/* Hopefully it's ok for aarch64 too */ + +/* These data are stored in ppheap.c.001t.tu */ +/* made with gcc -fdump-tree-all-fdump-tree-all */ + + +/* @2586 identifier_node strg: ppxref lngt: 6 */ +/* @2565 record_type name: @2586 size: @679 algn: 64 */ +/* tag : struct flds: @2587 */ + + +/* @2672 identifier_node strg: ppxsec lngt: 6 */ +/* @2648 type_decl name: @2672 type: @2623 scpe: @221 */ +/* srcp: ppxref.h:16 chain: @2673 */ +/* @2623 record_type name: @2648 unql: @2649 size: @593 */ +/* algn: 32 tag : struct flds: @2650 */ + + +/* @2642 identifier_node strg: ppstream lngt: 8 */ +/* @2615 type_decl name: @2642 type: @2643 scpe: @221 */ +/* srcp: ppapi.h:54 chain: @2644 */ +/* @2643 record_type name: @2615 unql: @2614 size: @634 */ +/* algn: 32 tag : struct flds: @2641 */ + + +/* @2765 identifier_node strg: ppkids lngt: 6 */ +/* @2743 type_decl name: @2765 type: @2766 scpe: @221 */ +/* srcp: ppload.h:16 chain: @2767 */ +/* @2766 record_type name: @2743 unql: @2742 size: @19 */ +/* algn: 32 tag : struct flds: @2764 */ + + +/* @2625 identifier_node strg: ppdoc lngt: 5 */ +/* @2605 record_type name: @2625 size: @2626 algn: 32 */ +/* tag : struct flds: @2627 */ + +/* @2526 identifier_node strg: ppref lngt: 5 */ +/* @2513 record_type name: @2526 size: @662 algn: 64 */ +/* tag : struct flds: @2527 */ + + +/* @2595 identifier_node strg: ppdict lngt: 6 */ +/* @2576 type_decl name: @2595 type: @2596 scpe: @221 */ +/* srcp: ppapi.h:45 chain: @2597 */ +/* @2596 record_type name: @2576 unql: @2575 size: @593 */ +/* algn: 32 tag : struct flds: @2594 */ + + +/* @2769 identifier_node strg: ppcrypt_status lngt: 14 */ +/* @2752 type_decl name: @2769 type: @2770 scpe: @221 */ +/* srcp: ppapi.h:295 chain: @2771 */ +/* @2770 enumeral_type name: @2752 unql: @2655 size: @5 */ +/* algn: 32 prec: 32 sign: signed */ +/* min : @6 max : @7 csts: @2681 */ + + +/* @2558 identifier_node strg: pparray lngt: 7 */ +/* @2541 type_decl name: @2558 type: @2559 scpe: @221 */ +/* srcp: ppapi.h:39 chain: @2560 */ +/* @2559 record_type name: @2541 unql: @2540 size: @19 */ +/* algn: 32 tag : struct flds: @2557 */ + + +/* @2817 identifier_node strg: aligned_data */ +/* @2801 type_decl name: @2817 type: @2818 scpe: @221 */ +/* srcp: ppheap.c:22 chain: @2819 */ +/* @2818 real_type name: @2801 unql: @99 size: @19 */ +/* algn: 64 prec: 64 */ + + +typedef long double aligned_data; + + +#define ALIGN_BUFF_BUCKET_SIZE 0x3000 /* heuristic value, found by running few tests */ +#ifdef __SIZEOF_POINTER__ +#define SIZE_OF_POINTER __SIZEOF_POINTER__ +#else +#define SIZE_OF_POINTER (sizeof(void *)) +#endif + +typedef struct _simplereg { + size_t bucket_pos; + size_t bucket_size; + size_t heap_instance; + aligned_data **align_data_set ; +} simplereg; + +/* By default static vars are initialized to NULL, but to be clear.. */ +static simplereg *align_set = NULL ; + +static void align_init_set(void){ + size_t size ; + + if (align_set) { + align_set->heap_instance++; + return ; + } + + align_set = malloc(sizeof(simplereg)); + if (!align_set) { + fprintf(stderr,"! fatal error: unable to setup master register for aligned pointers\n"); + exit(EXIT_FAILURE); + } + + size = SIZE_OF_POINTER*ALIGN_BUFF_BUCKET_SIZE; + align_set->align_data_set = malloc(size); + if (!align_set->align_data_set) { + fprintf(stderr,"! fatal error: unable to setup register for aligned pointers\n"); + exit(EXIT_FAILURE); + } + + align_set->bucket_pos = 0; + align_set->bucket_size = ALIGN_BUFF_BUCKET_SIZE; + align_set->heap_instance = 1; + memset(align_set->align_data_set, 0UL,size); +} + +static void align_save_into_set(aligned_data *p){ + if (align_set->bucket_pos >= align_set->bucket_size) { + size_t new_size; + aligned_data **align_data_set_new; + + if(!align_set->align_data_set){ + fprintf(stderr,"! fatal error: unable to save aligned pointer,corrupted set\n"); + exit(EXIT_FAILURE); + } + new_size = (ALIGN_BUFF_BUCKET_SIZE+align_set->bucket_size)*SIZE_OF_POINTER; + align_data_set_new = malloc(new_size); + if (!align_data_set_new) { + fprintf(stderr,"! fatal error: unable to save aligned pointer\n"); + exit(EXIT_FAILURE); + } + memset(align_data_set_new,0,new_size); + memcpy(align_data_set_new, align_set->align_data_set, align_set->bucket_size*SIZE_OF_POINTER); + free(align_set->align_data_set); + align_set->align_data_set = align_data_set_new ; + align_set->bucket_size += ALIGN_BUFF_BUCKET_SIZE; + } + if (align_set->bucket_pos>align_set->bucket_size){ + fprintf(stderr,"! fatal error: unable to save aligned pointer, wrong position\n"); + exit(EXIT_FAILURE); + } + align_set->align_data_set[align_set->bucket_pos] = p ; + align_set->bucket_pos++; + +} + +static void align_free_set(void){ + /* We don't know what heap does with its data, so free here is not secure */ + + if(align_set){ + if (align_set->heap_instance>1) { + align_set->heap_instance--; + } else if (align_set->heap_instance ==1) { + if (align_set->align_data_set){ + size_t p; + for(p=1;p<align_set->bucket_pos;p++){ + if (align_set->align_data_set[p]) { + free(align_set->align_data_set[p]); + } + } + free(align_set->align_data_set); + align_set->align_data_set = NULL; + } + align_set->heap_instance=0; + free(align_set); + align_set = NULL ; + } + } +} + +#endif /* PPHEAP_NEED_ALIGNMENT */ + + +static ppheap * ppheap_create (size_t size) +{ + ppheap *heap; + heap = (ppheap *)pp_malloc(sizeof(ppheap) + size * sizeof(uint8_t)); + heap->size = size; + heap->space = size; + heap->data = ppheap_head(heap); + heap->prev = NULL; + return heap; +} + +ppheap * ppheap_new (void) +{ +#ifdef PPHEAP_NEED_ALIGNMENT + align_init_set(); +#endif + return ppheap_create(PPHEAP_BUFFER); +} + +void ppheap_free (ppheap *heap) +{ + ppheap *prev; + do + { + prev = heap->prev; + pp_free(heap); + heap = prev; + } while (heap != NULL); +#ifdef PPHEAP_NEED_ALIGNMENT + align_free_set(); +#endif + +} + +void ppheap_renew (ppheap *heap) +{ // free all but first + ppheap *prev; + if ((prev = heap->prev) != NULL) + { + heap->prev = NULL; + ppheap_free(prev); + } + heap->size = heap->space; + heap->data = ppheap_head(heap); +} + +static ppheap * ppheap_insert_top (ppheap **pheap, size_t size) +{ + ppheap *heap; + heap = ppheap_create(size); + heap->prev = (*pheap); + *pheap = heap; + return heap; +} + +static ppheap * ppheap_insert_sub (ppheap **pheap, size_t size) +{ + ppheap *heap; + heap = ppheap_create(size); + heap->prev = (*pheap)->prev; + (*pheap)->prev = heap; + return heap; +} + +void * ppheap_take (ppheap **pheap, size_t size) +{ + ppheap *heap; + uint8_t *data; +#ifdef PPHEAP_NEED_ALIGNMENT + aligned_data *p_aligned_data; +#endif + heap = *pheap; + if (size <= heap->size) + ; + else if (heap->size <= PPHEAP_WASTE && size <= (PPHEAP_BUFFER >> 1)) + heap = ppheap_insert_top(pheap, PPHEAP_BUFFER); + else + heap = ppheap_insert_sub(pheap, size); + data = heap->data; + heap->data += size; + heap->size -= size; +#ifdef PPHEAP_NEED_ALIGNMENT + /* Todo: only if data%sizeof(aligned_data) != 0 */ + p_aligned_data = malloc(size); + memcpy(p_aligned_data,data,size); + align_save_into_set(p_aligned_data); + return (void *)p_aligned_data; +#else + return (void *)data; +#endif + +} + + +/* iof buffer tied to a heap */ + +static ppheap * ppheap_resize (ppheap **pheap, size_t size) +{ + ppheap *heap; + heap = ppheap_create(size); + heap->prev = (*pheap)->prev; + memcpy(heap->data, (*pheap)->data, (*pheap)->space); // (*pheap)->size is irrelevant + pp_free(*pheap); + *pheap = heap; + return heap; +} + +static size_t ppheap_buffer_handler (iof *O, iof_mode mode) +{ + ppheap **pheap, *heap; + size_t objectsize, buffersize, neededsize; + uint8_t *copyfrom; + switch (mode) + { + case IOFWRITE: + /* apparently more space needed then assumed initsize */ + pheap = (ppheap **)O->link; + heap = *pheap; + objectsize = (size_t)(O->buf - heap->data); + buffersize = (size_t)(O->pos - O->buf); + neededsize = objectsize + (buffersize << 1); + if (ppheap_head(heap) < heap->data) + { + if (heap->size <= PPHEAP_WASTE && neededsize <= (PPHEAP_BUFFER >> 1)) + { + heap = ppheap_insert_top(pheap, PPHEAP_BUFFER); + copyfrom = heap->prev->data; + } + else + { + heap = ppheap_insert_sub(pheap, neededsize); + copyfrom = (*pheap)->data; + O->link = (void *)(&(*pheap)->prev); + } + memcpy(heap->data, copyfrom, objectsize + buffersize); + } + else + { /* the buffer was (re)initialized from a new empty heap and occupies its entire space */ + // ASSERT(ppheap_head(heap) == heap->data); + heap = ppheap_resize(pheap, neededsize); + } + O->buf = heap->data + objectsize; + O->pos = O->buf + buffersize; + O->end = heap->data + heap->size; + return (size_t)(O->end - O->pos); + case IOFFLUSH: + return 0; + case IOFCLOSE: + // O->buf = O->pos = O->end = NULL; + // O->link = NULL; + return 0; + default: + break; + } + return 0; +} + +iof * ppheap_buffer (ppheap **pheap, size_t objectsize, size_t initsize) +{ + static iof ppheap_buffer_iof = IOF_WRITER_STRUCT(ppheap_buffer_handler, NULL, NULL, 0, 0); + ppheap *heap; + size_t size; + size = objectsize + initsize; + heap = *pheap; + if (size <= heap->size) + ; + else if (heap->size <= PPHEAP_WASTE && size <= (PPHEAP_BUFFER >> 1)) + { + heap = ppheap_create(PPHEAP_BUFFER); + heap->prev = (*pheap); + *pheap = heap; + } + else + { + heap = ppheap_create(size); + heap->prev = (*pheap)->prev; + (*pheap)->prev = heap; + pheap = &(*pheap)->prev; // passed to ppheap_buffer_iof.link + } + ppheap_buffer_iof.buf = ppheap_buffer_iof.pos = heap->data + objectsize; + ppheap_buffer_iof.end = heap->data + heap->size; + ppheap_buffer_iof.link = pheap; // ASSERT(*pheap == heap) + return &ppheap_buffer_iof; +} + +/* +void * ppheap_buffer_data (iof *O, size_t *psize) +{ + ppheap *heap; + heap = *((ppheap **)(O->link)); + *psize = ppheap_buffer_size(O, heap); + return heap->data; +} +*/ + +void * ppheap_flush (iof *O, size_t *psize) // not from explicit ppheap ** !!! +{ + ppheap *heap; + uint8_t *data; + size_t size; +#ifdef PPHEAP_NEED_ALIGNMENT + aligned_data *p_aligned_data; +#endif + heap = *((ppheap **)(O->link)); + *psize = ppheap_buffer_size(O, heap); + size = *psize; + data = heap->data; +/* heap->data += *psize; + heap->size -= *psize; +*/ + heap->data += size; + heap->size -= size; + // O->buf = O->pos = O->end = NULL; + // O->link = NULL; + // iof_close(O); +#ifdef PPHEAP_NEED_ALIGNMENT + /* Todo: only if data%sizeof(aligned_data) != 0 */ + p_aligned_data = malloc(size); + memcpy(p_aligned_data,data,size); + align_save_into_set(p_aligned_data); + return (void *)p_aligned_data; +#else + return data; + +#endif + + +} + + diff --git a/Build/source/texk/web2c/luatexdir/luapplib/ppheap.h b/Build/source/texk/web2c/luatexdir/luapplib/ppheap.h new file mode 100644 index 00000000000..128f9c56d62 --- /dev/null +++ b/Build/source/texk/web2c/luatexdir/luapplib/ppheap.h @@ -0,0 +1,29 @@ + +#ifndef PP_HEAP_H +#define PP_HEAP_H + +#include "utilmem.h" + +#define pp_malloc util_malloc +#define pp_callic util_calloc +#define pprealloc util_realloc +#define pp_free util_free + +typedef struct ppheap ppheap; +struct ppheap { + size_t size; + size_t space; + uint8_t *data; + ppheap *prev; +}; + +ppheap * ppheap_new (void); +void ppheap_free (ppheap *heap); +void ppheap_renew (ppheap *heap); +void * ppheap_take (ppheap **pheap, size_t size); +iof * ppheap_buffer (ppheap **pheap, size_t objectsize, size_t initsize); +#define ppheap_buffer_size(O, heap) (size_t)((O)->pos - (heap)->data) // == (size_t)(O->buf - heap->data) + (size_t)(O->pos - O->buf); +//void * ppheap_buffer_data (iof *O, size_t *psize); +void * ppheap_flush (iof *O, size_t *psize); + +#endif
\ No newline at end of file diff --git a/Build/source/texk/web2c/luatexdir/luapplib/pplib.h b/Build/source/texk/web2c/luatexdir/luapplib/pplib.h new file mode 100644 index 00000000000..e753cfa0514 --- /dev/null +++ b/Build/source/texk/web2c/luatexdir/luapplib/pplib.h @@ -0,0 +1,22 @@ + +#ifndef PP_LIB_H +#define PP_LIB_H + +#include <stdio.h> +#include <stdlib.h> +#include <string.h> +#include <assert.h> + +#include "utiliof.h" +#include "utillog.h" + +#include "ppapi.h" +#include "ppheap.h" +#include "ppdict.h" +#include "ppstream.h" +#include "pparray.h" +#include "ppcrypt.h" +#include "ppxref.h" +#include "ppload.h" + +#endif
\ No newline at end of file diff --git a/Build/source/texk/web2c/luatexdir/luapplib/ppload.c b/Build/source/texk/web2c/luatexdir/luapplib/ppload.c new file mode 100644 index 00000000000..0ad286e5769 --- /dev/null +++ b/Build/source/texk/web2c/luatexdir/luapplib/ppload.c @@ -0,0 +1,2590 @@ + +#include <utilbasexx.h> + +#include "pplib.h" + +const char * ppobj_kind[] = { "none", "null", "bool", "integer", "number", "name", "string", "array", "dict", "stream", "ref" }; + +#define ignored_char(c) (c == 0x20 || c == 0x0A || c == 0x0D || c == 0x09 || c == 0x00) +#define newline_char(c) (c == 0x0A || c == 0x0D) +#define IGNORED_CHAR_CASE 0x20: case 0x0A: case 0x0D: case 0x09: case 0x00 +#define NEWLINE_CHAR_CASE 0x0A: case 0x0D +#define DIGIT_CHAR_CASE '0': case '1': case '2': case '3': case '4': case '5': case '6': case '7': case '8': case '9' +#define OCTAL_CHAR_CASE '0': case '1': case '2': case '3': case '4': case '5': case '6': case '7' + +#define MAX_INT_DIGITS 32 + +#define PP_LENGTH_UNKNOWN ((size_t)-1) + +static const char * ppref_str (ppuint refnumber, ppuint refversion) +{ + static char buffer[MAX_INT_DIGITS + 1 + MAX_INT_DIGITS + 1 + 1 + 1]; +#if defined(MSVC64)|| defined(MINGW64) + sprintf(buffer, PPUINTF " " PPUINTF " R", refnumber, refversion); +#else + sprintf(buffer, PPUINTF " " PPUINTF " R", (unsigned long)(refnumber), (unsigned long)(refversion)); +#endif + return buffer; +} + +/* name */ + +// pdf name delimiters: 0..32, ()<>[]{}/% +// # treated specially +// .+- are valid part of name; keep in mind names such as -| | |- .notdef ABCDEF+Font etc. +const char ppname_byte_lookup[] = { + 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, + 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, + 0, 1, 1, '#', 1, 0, 1, 1, 0, 0, 1, 1, 1, 1, 1, 0, + 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 0, 1, 0, 1, + 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, + 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 0, 1, 0, 1, 1, + 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, + 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 0, 1, 0, 1, 1, + 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, + 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, + 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, + 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, + 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, + 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, + 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, + 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1 +}; + +#define PPNAME_INIT (8+1) + +#define ppname_flush(O, ghost, siz, flgs) \ + (iof_put(O, '\0'), \ + ghost = (_ppname *)ppheap_flush(O, &siz), \ + ghost->flags = flgs, \ + ghost->size = siz - sizeof(_ppname) - 1, \ + (ppname)(ghost + 1)) + +#define ppname_flush_with_ego(O, ghost, siz, flgs) \ + (iof_put(O, '\0'), \ + iof_ensure(O, sizeof(ppname *)), \ + O->pos += sizeof(ppname *), \ + ghost = (_ppname *)ppheap_flush(O, &siz), \ + ghost->flags = flgs, \ + ghost->size = siz - sizeof(_ppname) - 1 - sizeof(ppname *), \ + (ppname)(ghost + 1)) + + +#if defined __arm__ || defined __ARM__ || defined ARM || defined __ARM || defined __arm || defined __ARM_ARCH ||defined __aarch64__ +#define ppname_set_alter_ego(name, ghost, ego) do {\ + ppname temp;\ + ppname *temp1;\ + temp = (name + (ghost)->size + 1) ; \ + temp1 = (ppname *)((void*)temp); \ + *temp1= ego; \ + }while(0) +#else +#define ppname_set_alter_ego(name, ghost, ego) (*((ppname *)(name + (ghost)->size + 1)) = ego) +#endif + +#if defined __arm__ || defined __ARM__ || defined ARM || defined __ARM || defined __arm || defined __ARM_ARCH ||defined __aarch64__ +#define ppname_get_alter_ego(name) (*((ppname *)( (void*)(name + ppname_size(name) + 1)))) +#else +#define ppname_get_alter_ego(name) (*((ppname *)(name + ppname_size(name) + 1))) +#endif + + + + + +static ppname ppscan_name (iof *I, ppheap **pheap) +{ + int c, decode; + iof *O; + _ppname *ghost1, *ghost2; + ppname encoded, decoded; + size_t size; + const char *p, *e; + uint8_t v1, v2; + O = ppheap_buffer(pheap, sizeof(_ppname), PPNAME_INIT); + for (decode = 0, c = iof_char(I); c >= 0 && ppname_byte_lookup[c]; c = iof_next(I)) + { + if (c == '#') + decode = 1; + iof_put(O, c); + } + if (!decode) + return ppname_flush(O, ghost1, size, 0); + encoded = ppname_flush_with_ego(O, ghost1, size, 0|PPNAME_ENCODED); + O = ppheap_buffer(pheap, sizeof(_ppname), PPNAME_INIT); + for (p = encoded, e = encoded + ghost1->size; p < e; ++p) + { + if (*p == '#' && p + 2 < e ){ + v1 = base16_value(p[1]); + v2 = base16_value(p[2]); + iof_put(O, ((v1<<4)+v2)); + }else + iof_put(O, *p); + } + decoded = ppname_flush_with_ego(O, ghost2, size, 0|PPNAME_DECODED); + ppname_set_alter_ego(encoded, ghost1, decoded); + ppname_set_alter_ego(decoded, ghost2, encoded); + return encoded; +} + +static ppname ppscan_exec (iof *I, ppheap **pheap, int first) +{ + int c, decode; + iof *O; + _ppname *ghost1, *ghost2; + ppname encoded, decoded; + size_t size; + const char *p, *e; + uint8_t v1, v2; + + O = ppheap_buffer(pheap, sizeof(_ppname), PPNAME_INIT); + iof_put(O, first); + for (decode = 0, c = iof_char(I); c >= 0 && ppname_byte_lookup[c]; c = iof_next(I)) + { + if (c == '#') + decode = 1; + iof_put(O, c); + } + if (!decode) + return ppname_flush(O, ghost1, size, PPNAME_EXEC); + encoded = ppname_flush_with_ego(O, ghost1, size, PPNAME_EXEC|PPNAME_ENCODED); + O = ppheap_buffer(pheap, sizeof(_ppname), PPNAME_INIT); + for (p = encoded, e = encoded + ghost1->size; p < e; ++p) + { + if (*p == '#' && p + 2 < e ){ + v1 = base16_value(p[1]); + v2 = base16_value(p[2]); + iof_put(O, ((v1<<4)+v2)); + }else + iof_put(O, *p); + } + decoded = ppname_flush_with_ego(O, ghost2, size, PPNAME_EXEC|PPNAME_DECODED); + ppname_set_alter_ego(encoded, ghost1, decoded); + ppname_set_alter_ego(decoded, ghost2, encoded); + return encoded; +} + +static ppname ppexec_internal (const void *data, size_t size, ppheap **pheap) +{ // used only for artificial 'EI' operator name + iof *O; + _ppname *ghost1; + + O = ppheap_buffer(pheap, sizeof(_ppname), size); + iof_write(O, data, size); + return ppname_flush(O, ghost1, size, PPNAME_EXEC); +} + +ppname ppname_decoded (ppname name) +{ + const _ppname *ghost; + ghost = _ppname_ghost(name); + return (ghost->flags & PPNAME_ENCODED) ? ppname_get_alter_ego(name) : name; +} + +ppname ppname_encoded (ppname name) +{ + const _ppname *ghost; + ghost = _ppname_ghost(name); + return (ghost->flags & PPNAME_DECODED) ? ppname_get_alter_ego(name) : name; +} + +/* string */ + +#define PPSTRING_INIT (16+1) + +#define ppstring_flush(O, ghost, siz, flgs) \ + (iof_put(O, '\0'), \ + ghost = (_ppstring *)ppheap_flush(O, &siz), \ + ghost->flags = flgs, \ + ghost->size = siz - sizeof(_ppstring) - 1, \ + (ppstring)(ghost + 1)) + +#define ppstring_flush_with_ego(O, ghost, siz, flgs) \ + (iof_put(O, '\0'), \ + iof_ensure(O, sizeof(ppstring *)), \ + O->pos += sizeof(ppstring *), \ + ghost = (_ppstring *)ppheap_flush(O, &siz), \ + ghost->flags = flgs, \ + ghost->size = siz - sizeof(_ppstring) - 1 - sizeof(ppstring *), \ + (ppstring)(ghost + 1)) + +#define ppstring_utf16be_bom(decoded) (decoded[0] == ((char)0xFE) && decoded[1] == ((char)0xFF) ) +#define ppstring_utf16le_bom(decoded) (decoded[0] == ((char)0xFF) && decoded[1] == ((char)0xFE)) + +#define ppstring_check_bom(decoded, ghost) ((void)\ + (ghost->size >= 2 ? (ppstring_utf16be_bom(decoded) ? (ghost->flags |= PPSTRING_UTF16BE) : \ + (ppstring_utf16le_bom(decoded) ? (ghost->flags |= PPSTRING_UTF16LE) : 0)) : 0)) + +#define ppstring_check_bom2(decoded, ghost1, ghost2) ((void)\ + (ghost2->size >= 2 ? (ppstring_utf16be_bom(decoded) ? ((ghost1->flags |= PPSTRING_UTF16BE), (ghost2->flags |= PPSTRING_UTF16BE)) : \ + (ppstring_utf16le_bom(decoded) ? ((ghost1->flags |= PPSTRING_UTF16LE), (ghost2->flags |= PPSTRING_UTF16LE)) : 0)) : 0)) + + + +#if defined __arm__ || defined __ARM__ || defined ARM || defined __ARM || defined __arm || defined __ARM_ARCH ||defined __aarch64__ +#define ppstring_set_alter_ego(string, ghost, ego) (*((ppstring *)((void *)(string + (ghost)->size + 1))) = ego) +#else +#define ppstring_set_alter_ego(string, ghost, ego) (*((ppstring *)(string + (ghost)->size + 1)) = ego) +#endif + + +#if defined __arm__ || defined __ARM__ || defined ARM || defined __ARM || defined __arm || defined __ARM_ARCH ||defined __aarch64__ +#define ppstring_get_alter_ego(string) (*((ppstring *)((void *)(string + ppstring_size(string) + 1)))) +#else +#define ppstring_get_alter_ego(string) (*((ppstring *)(string + ppstring_size(string) + 1))) +#endif + + + + +static ppstring ppscan_string (iof *I, ppheap **pheap) +{ + int c, decode, balance; + iof *O; + _ppstring *ghost1, *ghost2; + uint8_t *p, *e; + ppstring encoded, decoded; + size_t size; + O = ppheap_buffer(pheap, sizeof(_ppstring), PPSTRING_INIT); + for (decode = 0, balance = 0, c = iof_char(I); c >= 0; ) + { + switch (c) + { + case '\\': + decode = 1; // unescaping later + iof_put(O, '\\'); + if ((c = iof_next(I)) >= 0) + { + iof_put(O, c); + c = iof_next(I); + } + break; + case '(': // may be unescaped if balanced + ++balance; + iof_put(O, '('); + c = iof_next(I); + break; + case ')': + if (balance == 0) + { + c = IOFEOF; + ++I->pos; + break; + } + --balance; + iof_put(O, ')'); + c = iof_next(I); + break; + default: + iof_put(O, c); + c = iof_next(I); + } + } + if (!decode) + { + encoded = ppstring_flush(O, ghost1, size, 0); + ppstring_check_bom(encoded, ghost1); // any bytes can be there + return encoded; + } + encoded = ppstring_flush_with_ego(O, ghost1, size, 0|PPSTRING_ENCODED); + O = ppheap_buffer(pheap, sizeof(_ppstring), PPSTRING_INIT); + for (p = (uint8_t *)encoded, e = (uint8_t *)encoded + ghost1->size; p < e; ++p) + { + if (*p == '\\') + { + if (++p >= e) + break; + switch (*p) + { + case OCTAL_CHAR_CASE: + c = *p - '0'; + if (++p < e && *p >= '0' && *p <= '7') + { + c = (c << 3) + *p - '0'; + if (++p < e && *p >= '0' && *p <= '7') + c = (c << 3) + *p - '0'; + } + iof_put(O, c); + break; + case 'n': + iof_put(O, '\n'); + break; + case 'r': + iof_put(O, '\r'); + break; + case 't': + iof_put(O, '\t'); + break; + case 'b': + iof_put(O, '\b'); + break; + case 'f': + iof_put(O, '\f'); + break; + case NEWLINE_CHAR_CASE: // not a part of the string, ignore (pdf spec page 55) + break; + case '(': case ')': case '\\': + default: // for enything else backslash is ignored (pdf spec page 54) + iof_put(O, *p); + break; + } + } + else + iof_put(O, *p); + } + decoded = ppstring_flush_with_ego(O, ghost2, size, 0|PPSTRING_DECODED); + ppstring_check_bom2(decoded, ghost1, ghost2); + ppstring_set_alter_ego(encoded, ghost1, decoded); + ppstring_set_alter_ego(decoded, ghost2, encoded); + return encoded; +} + +/* Hex string may contain white characters. If odd number of digits, the last assumed to be '0' */ + +static ppstring ppscan_base16 (iof *I, ppheap **pheap) +{ + int c, v1, v2; + iof *O; + _ppstring *ghost1, *ghost2; + size_t size; + ppstring encoded, decoded; + uint8_t *p, *e; + + O = ppheap_buffer(pheap, sizeof(_ppstring), PPSTRING_INIT); + for (c = iof_char(I); c >= 0 && (base16_digit(c) || ignored_char(c)); c = iof_next(I)) + iof_put(O, c); + if (c == '>') + ++I->pos; + encoded = ppstring_flush_with_ego(O, ghost1, size, PPSTRING_BASE16|PPSTRING_ENCODED); + O = ppheap_buffer(pheap, sizeof(_ppstring), (ghost1->size >> 1) + 1); + for (p = (uint8_t *)encoded, e = (uint8_t *)encoded + ghost1->size; p < e; ++p) + { + if ((v1 = base16_value(*p)) < 0) // ignored + continue; + for (v2 = 0, ++p; p < e && (v2 = base16_value(*p)) < 0; ++p); + iof_put(O, (v1<<4)|v2); + } + decoded = ppstring_flush_with_ego(O, ghost2, size, 0|PPSTRING_DECODED); + ppstring_check_bom2(decoded, ghost1, ghost2); + ppstring_set_alter_ego(encoded, ghost1, decoded); + ppstring_set_alter_ego(decoded, ghost2, encoded); + return encoded; +} + +/* internal use only; binary string */ + +static ppstring ppstring_buffer (iof *O, ppheap **pheap) +{ + _ppstring *ghost1, *ghost2; + ppstring encoded, decoded; + uint8_t *p, *e; + size_t size; + + decoded = ppstring_flush_with_ego(O, ghost2, size, 0|PPSTRING_DECODED); + O = ppheap_buffer(pheap, sizeof(_ppstring), (ghost2->size << 1) + 1); + for (p = (uint8_t *)decoded, e = (uint8_t *)decoded + ghost2->size; p < e; ++p) + iof_set2(O, base16_uc_alphabet[(*p)>>4], base16_uc_alphabet[(*p)&15]); + encoded = ppstring_flush_with_ego(O, ghost1, size, PPSTRING_BASE16|PPSTRING_ENCODED); + ppstring_set_alter_ego(encoded, ghost1, decoded); + ppstring_set_alter_ego(decoded, ghost2, encoded); + return encoded; +} + +ppstring ppstring_internal (const void *data, size_t size, ppheap **pheap) +{ // so far used only for crypt key + iof *O; + + O = ppheap_buffer(pheap, sizeof(_ppstring), size); + iof_write(O, data, size); + return ppstring_buffer(O, pheap); +} + +/* PDF spec says nothing about base85 strings, but streams might be (afair no leading '<~' but present trailing '~>') */ + +static ppstring ppscan_base85 (iof *I, ppheap **pheap) +{ // bawse85 alphabet is 33.117, adobe also hires 'z' and 'y' for compression + int c; + iof *O, B; + _ppstring *ghost1, *ghost2; + size_t size; + ppstring encoded, decoded; + O = ppheap_buffer(pheap, sizeof(_ppstring), PPSTRING_INIT); + for (c = iof_char(I); (c >= '!' && c <= 'u') || c == 'z' || c == 'y'; c = iof_next(I)) + iof_put(O, c); + if (c == '~') + if ((c = iof_next(I)) == '>') + ++I->pos; + encoded = ppstring_flush_with_ego(O, ghost1, size, PPSTRING_BASE85|PPSTRING_ENCODED); + iof_string_reader(&B, encoded, ghost1->size); + O = ppheap_buffer(pheap, sizeof(_ppstring), (ghost1->size * 5 / 4) + 1); + base85_decode(&B, O); + decoded = ppstring_flush_with_ego(O, ghost2, size, 0|PPSTRING_DECODED); + ppstring_check_bom2(decoded, ghost1, ghost2); + ppstring_set_alter_ego(encoded, ghost1, decoded); + ppstring_set_alter_ego(decoded, ghost2, encoded); + return encoded; +} + +/* +Encrypted strings. In case of encrypted strings, we first need to decode the string (saving original form hardly makes sense), +then decrypt the string, and encode it again. +*/ + +const char ppstring_byte_escape[] = { /* -1 escaped with octal, >0 escaped with \\, 0 left intact*/ + -1,-1,-1,-1,-1,-1,-1,-1,'b','t','n',-1,'f','r',-1,-1, + -1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1, + 0, 0, 0, 0, 0, 0, 0, 0,'(',')', 0, 0, 0, 0, 0, 0, + 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, + 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, + 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,'\\', 0, 0, 0, + 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, + 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, + -1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1, + -1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1, + -1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1, + -1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1, + -1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1, + -1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1, + -1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1, + -1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1 +}; + + +static ppstring ppscan_crypt_string (iof *I, ppcrypt *crypt, ppheap **pheap) +{ + int c, b, balance, encode; + iof *O; + _ppstring *ghost1, *ghost2; + ppstring encoded, decoded; + uint8_t *p, *e; + size_t size; + + O = ppheap_buffer(pheap, sizeof(_ppstring), PPSTRING_INIT); + for (balance = 0, encode = 0, c = iof_char(I); c >= 0; ) + { + switch (c) + { + case '\\': + if ((c = iof_next(I)) < 0) + break; + encode = 1; + switch (c) + { + case OCTAL_CHAR_CASE: + b = c - '0'; + if ((c = iof_next(I)) >= 0 && c >= '0' && c <= '7') + { + b = (b << 3) + c - '0'; + if ((c = iof_next(I)) >= 0 && c >= '0' && c <= '7') + { + b = (b << 3) + c - '0'; + c = iof_next(I); + } + } + iof_put(O, b); + // c is set to the next char + break; + case 'n': + iof_put(O, '\n'); + c = iof_next(I); + break; + case 'r': + iof_put(O, '\r'); + c = iof_next(I); + break; + case 't': + iof_put(O, '\t'); + c = iof_next(I); + break; + case 'b': + iof_put(O, '\b'); + c = iof_next(I); + break; + case 'f': + iof_put(O, '\f'); + c = iof_next(I); + break; + case NEWLINE_CHAR_CASE: // not a part of the string, ignore (pdf spec page 55) + c = iof_next(I); + break; + case '(': case ')': case '\\': + default: // for enything else backslash is ignored (pdf spec page 54) + iof_put(O, c); + c = iof_next(I); + break; + } + break; + case '(': + ++balance; + encode = 1; + iof_put(O, '('); + c = iof_next(I); + break; + case ')': + if (balance == 0) + { + c = IOFEOF; + ++I->pos; + } + else + { + --balance; + //encode = 1; + iof_put(O, ')'); + c = iof_next(I); + } + break; + default: + if (ppstring_byte_escape[c] != 0) + encode = 1; + iof_put(O, c); + c = iof_next(I); + } + } + /* decrypt the buffer in place, update size */ + if (ppstring_decrypt(crypt, O->buf, iof_size(O), O->buf, &size)) + O->pos = O->buf + size; + /* make encoded counterpart */ + if (!encode) + { + decoded = ppstring_flush(O, ghost2, size, 0); + ppstring_check_bom(decoded, ghost2); + return decoded; + } + decoded = ppstring_flush_with_ego(O, ghost2, size, PPSTRING_DECODED); + O = ppheap_buffer(pheap, sizeof(_ppstring), ghost2->size); + for (p = (uint8_t *)decoded, e = (uint8_t *)decoded + ghost2->size; p < e; ++p) + { + switch ((b = ppstring_byte_escape[*p])) + { + case 0: + iof_put(O, *p); + break; + case -1: + iof_put4(O, '\\', (c >> 6) + '0', ((c >> 3) & 7) + '0', (c & 7) + '0'); + break; + default: + iof_put2(O, '\\', b); + break; + } + } + encoded = ppstring_flush_with_ego(O, ghost1, size, PPSTRING_ENCODED); + ppstring_check_bom2(decoded, ghost1, ghost2); + ppstring_set_alter_ego(encoded, ghost1, decoded); + ppstring_set_alter_ego(decoded, ghost2, encoded); + return encoded; +} + + +static ppstring ppscan_crypt_base16 (iof *I, ppcrypt *crypt, ppheap **pheap) +{ + int c, v1, v2; + iof *O; + _ppstring *ghost1, *ghost2; + ppstring encoded, decoded; + uint8_t *p, *e; + size_t size; + + O = ppheap_buffer(pheap, sizeof(_ppstring), PPSTRING_INIT); + // base16_decode(I, O); // no info about the last char.. + for (c = iof_char(I); c != '>' && c >= 0; ) + { + if ((v1 = base16_value(c)) < 0) + { + if (ignored_char(c)) + { + c = iof_next(I); + continue; + } + break; + } + do { + c = iof_next(I); + if ((v2 = base16_value(c)) >= 0) + { + c = iof_next(I); + break; + } + if (!ignored_char(c)) // c == '>' || c < 0 or some crap + { + v2 = 0; + break; + } + } while (1); + iof_put(O, (v1 << 4)|v2); + } + if (c == '>') + ++I->pos; + /* decrypt the buffer in place, update size */ + if (ppstring_decrypt(crypt, O->buf, iof_size(O), O->buf, &size)) + O->pos = O->buf + size; + decoded = ppstring_flush_with_ego(O, ghost2, size, PPSTRING_DECODED); + /* recreate an encoded form */ + O = ppheap_buffer(pheap, sizeof(_ppstring), (ghost2->size << 1) + 1); + for (p = (uint8_t *)decoded, e = (uint8_t *)decoded + ghost2->size; p < e; ++p) + iof_set2(O, base16_uc_alphabet[(*p)>>4], base16_uc_alphabet[(*p)&15]); + encoded = ppstring_flush_with_ego(O, ghost1, size, PPSTRING_BASE16|PPSTRING_ENCODED); + ppstring_check_bom2(decoded, ghost1, ghost2); + ppstring_set_alter_ego(encoded, ghost1, decoded); + ppstring_set_alter_ego(decoded, ghost2, encoded); + return encoded; +} + +/* ppstring alter ego switcher */ + +ppstring ppstring_decoded (ppstring string) +{ + const _ppstring *ghost; + ghost = _ppstring_ghost(string); + return (ghost->flags & PPSTRING_ENCODED) ? ppstring_get_alter_ego(string) : string; +} + +ppstring ppstring_encoded (ppstring string) +{ + const _ppstring *ghost; + ghost = _ppstring_ghost(string); + return (ghost->flags & PPSTRING_DECODED) ? ppstring_get_alter_ego(string) : string; +} + +/* scanner stack */ + +#define PPSTACK_BUFFER 512 + +static void ppstack_init (ppstack *stack, ppheap **pheap) +{ + stack->buf = stack->pos = (ppobj *)pp_malloc(PPSTACK_BUFFER * sizeof(ppobj)); + stack->size = 0; + stack->space = PPSTACK_BUFFER; + stack->pheap = pheap; +} + +#define ppstack_free_buffer(stack) (pp_free((stack)->buf)) + +static void ppstack_resize (ppstack *stack) +{ + ppobj *newbuffer; + stack->space <<= 1; + newbuffer = (ppobj *)pp_malloc(stack->space * sizeof(ppobj)); + memcpy(newbuffer, stack->buf, stack->size * sizeof(ppobj)); + ppstack_free_buffer(stack); + stack->buf = newbuffer; + stack->pos = newbuffer + stack->size; +} + +#define ppstack_push(stack) ((void)((stack)->size < (stack)->space || (ppstack_resize(stack), 0)), ++(stack)->size, (stack)->pos++) +#define ppstack_pop(stack, n) ((stack)->size -= (n), (stack)->pos -= (n)) +#define ppstack_at(stack, i) ((stack)->buf + i) +#define ppstack_clear(stack) ((stack)->pos = (stack)->buf, (stack)->size = 0) + +/* scanner commons */ + +#define ppscan_uint(I, u) iof_get_uintlw(I, u) +#define ppread_uint(s, u) string_to_uintlw((const char *)(s), u) + +static ppobj * ppscan_numobj (iof *I, ppobj *obj, int negative) +{ + ppint integer; + ppnum number; + int exponent; + int c; + c = iof_char(I); + iof_scan_integer(I, c, integer); + switch(c) + { + case '.': + { + number = (ppnum)integer; + c = iof_next(I); + iof_scan_fraction(I, c, number, exponent); + double_negative_exp10(number, exponent); + obj->type = PPNUM, obj->number = negative ? -number : number; + break; + } + default: + obj->type = PPINT, obj->integer = negative ? -integer : integer; + break; + } + return obj; +} + +static ppobj * ppscan_numobj_frac (iof *I, ppobj *obj, int negative) +{ + ppnum number; + int c, exponent; + + number = 0.0; + c = iof_next(I); + iof_scan_fraction(I, c, number, exponent); + double_negative_exp10(number, exponent); + obj->type = PPNUM, obj->number = negative ? -number : number; + return obj; +} + +static int ppscan_find (iof *I) +{ // skips whitechars and comments + int c; + for (c = iof_char(I); ; c = iof_next(I)) + { + switch (c) + { + case IGNORED_CHAR_CASE: + break; + case '%': { + do { + if ((c = iof_next(I)) < 0) + return c; + } while (!newline_char(c)); + break; + } + default: + return c; + } + } + return c; // never reached +} + +static int ppscan_keyword (iof *I, const char *keyword, size_t size) +{ + size_t i; + int c; + if (iof_left(I) >= size) + { + if (memcmp(I->pos, keyword, size) != 0) + return 0; + I->pos += size; + return 1; + } + // sticky case, we can't go back + for (i = 0, c = iof_char(I); i < size; ++i, ++keyword, c = iof_next(I)) + if (i != c) + return 0; + return 1; +} + +#define ppscan_key(I, literal) ppscan_keyword(I, "" literal, sizeof(literal) - 1) + +/* objects parser */ + +static ppref * ppref_unresolved (ppheap **pheap, ppuint refnumber, ppuint refversion) +{ + ppref *ref = (ppref *)ppheap_take(pheap, sizeof(ppref)); + memset(ref, 0, sizeof(ppref)); + ref->object.type = PPNONE; + ref->number = refnumber; + ref->version = refversion; + return ref; +} + +#define PPMARK PPNONE + +static ppobj * ppscan_obj (iof *I, ppdoc *pdf, ppxref *xref) +{ + int c; + ppobj *obj; + size_t mark, size; + ppuint refnumber, refversion; + ppref *ref; + ppstack *stack; + ppcrypt *crypt; + + stack = &pdf->stack; + c = iof_char(I); + switch (c) + { + case DIGIT_CHAR_CASE: + return ppscan_numobj(I, ppstack_push(stack), 0); + case '.': + return ppscan_numobj_frac(I, ppstack_push(stack), 0); + case '+': + ++I->pos; + return ppscan_numobj(I, ppstack_push(stack), 0); + case '-': + ++I->pos; + return ppscan_numobj(I, ppstack_push(stack), 1); + case '/': + ++I->pos; + obj = ppstack_push(stack); + obj->type = PPNAME; + obj->name = ppscan_name(I, &pdf->heap); + return obj; + case '(': + ++I->pos; + obj = ppstack_push(stack); + obj->type = PPSTRING; + if (ppcrypt_ref(pdf, crypt)) + obj->string = ppscan_crypt_string(I, crypt, &pdf->heap); + else + obj->string = ppscan_string(I, &pdf->heap); + return obj; + case '[': + mark = stack->size; + obj = ppstack_push(stack); + obj->type = PPMARK; // ppscan_obj() checks types backward for 'R', so set the type immediatelly (reserved for PPARRAY) + obj->any = NULL; + ++I->pos; + for (c = ppscan_find(I); c != ']'; c = ppscan_find(I)) + { + if (ppscan_obj(I, pdf, xref) == NULL) + { // callers assume that NULL returns means nothing pushed + size = stack->size - mark; // pop items AND the obj reserved for array + ppstack_pop(stack, size); + return NULL; + } + } + ++I->pos; + size = stack->size - mark - 1; + obj = ppstack_at(stack, mark); // stack might have been realocated + obj->type = PPARRAY; + obj->array = pparray_create(ppstack_at(stack, mark + 1), size, &pdf->heap); + ppstack_pop(stack, size); // pop array items, leave the array on top + return obj; + case '<': + if ((c = iof_next(I)) == '<') + { + mark = stack->size; + obj = ppstack_push(stack); + obj->type = PPMARK; + obj->any = NULL; + ++I->pos; + for (c = ppscan_find(I); c != '>'; c = ppscan_find(I)) + { + if (ppscan_obj(I, pdf, xref) == NULL) + { + size = stack->size - mark; + ppstack_pop(stack, size); + return NULL; + } + } + if (iof_next(I) == '>') + ++I->pos; + size = stack->size - mark - 1; + obj = ppstack_at(stack, mark); + obj->type = PPDICT; + obj->dict = ppdict_create(ppstack_at(stack, mark + 1), size, &pdf->heap); + ppstack_pop(stack, size); + return obj; + } + obj = ppstack_push(stack); + obj->type = PPSTRING; + if (ppcrypt_ref(pdf, crypt)) + obj->string = ppscan_crypt_base16(I, crypt, &pdf->heap); + else + obj->string = ppscan_base16(I, &pdf->heap); + return obj; + case 'R': + if (stack->size >= 2 && stack->pos[-1].type == PPINT && stack->pos[-2].type == PPINT) + { + ++I->pos; + obj = &stack->pos[-2]; + refnumber = (ppuint)obj->integer; + ppstack_pop(stack, 1); // pop version number, retype obj to a reference + if (xref == NULL || (ref = ppxref_find(xref, refnumber)) == NULL) + { /* pdf spec page 64: unresolvable reference is not an error, should just be treated as a reference to null. + we also need this to read trailer, where refs can't be resolved yet */ + refversion = (obj + 1)->integer; + //if (xref != NULL) + // loggerf("unresolved reference %s", ppref_str(refnumber, refversion)); + ref = ppref_unresolved(stack->pheap, refnumber, refversion); + } + obj->type = PPREF; + obj->ref = ref; + return obj; + } + break; + case 't': + if (iof_next(I) == 'r' && iof_next(I) == 'u' && iof_next(I) == 'e') + { + ++I->pos; + obj = ppstack_push(stack); + obj->type = PPBOOL; + obj->integer = 1; + return obj; + } + break; + case 'f': + if (iof_next(I) == 'a' && iof_next(I) == 'l' && iof_next(I) == 's' && iof_next(I) == 'e') + { + ++I->pos; + obj = ppstack_push(stack); + obj->type = PPBOOL; + obj->integer = 0; + return obj; + } + break; + case 'n': + if (iof_next(I) == 'u' && iof_next(I) == 'l' && iof_next(I) == 'l') + { + ++I->pos; + obj = ppstack_push(stack); + obj->type = PPNULL; + obj->any = NULL; + return obj; + } + break; + } + return NULL; +} + +/* +A variant for contents streams (aka postscript); wise of operators, blind to references. +We are still PDF, so we don't care about postscript specific stuff such as radix numbers +and scientific numbers notation. It takes ppstack * as context (no ppdoc *) to be able +to run contents parser beyond the scope of ppdoc heap. +*/ + +static ppstring ppstring_inline (iof *I, ppdict *imagedict, ppheap **pheap); + +static ppobj * ppscan_psobj (iof *I, ppstack *stack) +{ + int c; + ppobj *obj, *op; + size_t size, mark; + ppname exec; + + c = iof_char(I); + switch (c) + { + case DIGIT_CHAR_CASE: + return ppscan_numobj(I, ppstack_push(stack), 0); + case '.': + return ppscan_numobj_frac(I, ppstack_push(stack), 0); + case '+': + c = iof_next(I); + if (base10_digit(c)) // '+.abc' is probably an executable name, but we are not in postscript + return ppscan_numobj(I, ppstack_push(stack), 0); + else if (c == '.') + return ppscan_numobj_frac(I, ppstack_push(stack), 0); + obj = ppstack_push(stack); + obj->type = PPNAME; + obj->name = ppscan_exec(I, stack->pheap, '+'); + return obj; + case '-': + c = iof_next(I); + if (base10_digit(c)) // ditto, we would handle type1 '-|' '|-' operators though + return ppscan_numobj(I, ppstack_push(stack), 1); + else if (c == '.') + return ppscan_numobj_frac(I, ppstack_push(stack), 1); + obj = ppstack_push(stack); + obj->type = PPNAME; + obj->name = ppscan_exec(I, stack->pheap, '-'); + return obj; + case '/': + ++I->pos; + obj = ppstack_push(stack); + obj->type = PPNAME; + obj->name = ppscan_name(I, stack->pheap); + return obj; + case '(': + ++I->pos; + obj = ppstack_push(stack); + obj->type = PPSTRING; + obj->string = ppscan_string(I, stack->pheap); + return obj; + case '[': + mark = stack->size; + obj = ppstack_push(stack); + obj->type = PPMARK; + obj->any = NULL; + ++I->pos; + for (c = ppscan_find(I); c != ']'; c = ppscan_find(I)) + { + if (ppscan_psobj(I, stack) == NULL) + { + size = stack->size - mark; + ppstack_pop(stack, size); + return NULL; + } + } + ++I->pos; + size = stack->size - mark - 1; + obj = ppstack_at(stack, mark); + obj->type = PPARRAY; + obj->array = pparray_create(ppstack_at(stack, mark + 1), size, stack->pheap); + ppstack_pop(stack, size); + return obj; + case '<': + if ((c = iof_next(I)) == '<') + { + mark = stack->size; + obj = ppstack_push(stack); + obj->type = PPMARK; + obj->any = NULL; + ++I->pos; + for (c = ppscan_find(I); c != '>'; c = ppscan_find(I)) + { + if (ppscan_psobj(I, stack) == NULL) + { + size = stack->size - mark; + ppstack_pop(stack, size); + return NULL; + } + } + if (iof_next(I) == '>') + ++I->pos; + size = stack->size - mark - 1; + obj = ppstack_at(stack, mark); + obj->type = PPDICT; + obj->dict = ppdict_create(ppstack_at(stack, mark + 1), size, stack->pheap); + ppstack_pop(stack, size); + return obj; + } + obj = ppstack_push(stack); + obj->type = PPSTRING; + if (c == '~') + ++I->pos, obj->string = ppscan_base85(I, stack->pheap); + else + obj->string = ppscan_base16(I, stack->pheap); + return obj; + default: + if (c < 0 || !ppname_byte_lookup[c]) + break; // forbid empty names; dead loop otherwise + ++I->pos; + /* true false null practically don't occur in streams so it makes sense to assume that we get an operator name here. + If it happen to be a keyword we could give back those several bytes to the heap but.. heap buffer is tricky enough. */ + exec = ppscan_exec(I, stack->pheap, c); + obj = ppstack_push(stack); + switch (exec[0]) + { + case 't': + if (exec[1] == 'r' && exec[2] == 'u' && exec[3] == 'e' && exec[4] == '\0') + { + obj->type = PPBOOL; + obj->integer = 1; + // todo: drop exec + return obj; + } + break; + case 'f': + if (exec[1] == 'a' && exec[2] == 'l' && exec[3] == 's' && exec[4] == 'e' && exec[5] == '\0') + { + obj->type = PPBOOL; + obj->integer = 0; + // todo: drop exec + return obj; + } + break; + case 'n': + if (exec[1] == 'u' && exec[2] == 'l' && exec[3] == 'l' && exec[4] == '\0') + { + obj->type = PPNULL; + obj->any = NULL; + // todo: drop exec + return obj; + } + break; + case 'B': + /* + Inline images break rules of operand/operator syntax, so 'BI/ID' operators need to be treated as special syntactic keywords. + + BI <keyval pairs> ID<whitechar?><imagedata><whitechar?>EI + + We treat the image as a single syntactic token; BI starts collecting a dict, ID is the beginning of the data. Effectively EI + operator obtains two operands - dict and string. It is ok to put three items onto the stack, callers dont't assume there is just one. + */ + if (exec[1] == 'I' && exec[2] == '\0') + { + ppdict *imagedict; + /* key val pairs -> dict */ + mark = stack->size - 1; + obj->type = PPMARK; + obj->any = NULL; + for (c = ppscan_find(I); ; c = ppscan_find(I)) + { + if ((op = ppscan_psobj(I, stack)) == NULL) + { + size = stack->size - mark; + ppstack_pop(stack, size); + return NULL; + } + if (op->type == PPNAME && ppname_exec(op->name)) + { + if (!ppname_is(op->name, "ID")) + { // weird + size = stack->size - mark; + ppstack_pop(stack, size); + return NULL; + } + break; + } + } + size = stack->size - mark - 1; + obj = ppstack_at(stack, mark); + obj->type = PPDICT; + obj->dict = imagedict = ppdict_create(ppstack_at(stack, mark + 1), size, stack->pheap); + ppstack_pop(stack, size); + /* put image data string */ + obj = ppstack_push(stack); + obj->type = PPSTRING; + obj->string = ppstring_inline(I, imagedict, stack->pheap);; + /* put EI operator name */ + obj = ppstack_push(stack); + obj->type = PPNAME; + obj->name = ppexec_internal("EI", 2, stack->pheap); + return obj; + } + } + obj->type = PPNAME; + obj->name = exec; + return obj; + } + return NULL; +} + +/* +We try to get the exact inline image length from its dict params. If cannot predict the length, we have to scan the input until 'EI'. +I've checked on may examples that it gives the same results but one can never be sure, as 'EI' might happen to be a part of the data. +Stripping white char is also very heuristic; \0 is a white char in PDF and very likely to be a data byte.. weak method. +*/ + +static size_t inline_image_length (ppdict *dict) +{ + ppuint w, h, bpc, colors; + ppname cs; + + if (ppdict_get_uint(dict, "W", &w) && ppdict_get_uint(dict, "H", &h) && ppdict_get_uint(dict, "BPC", &bpc) && (cs = ppdict_get_name(dict, "CS")) != NULL) + { + if (ppname_is(cs, "DeviceGray")) + colors = 1; + else if (ppname_is(cs, "DeviceRGB")) + colors = 3; + else if (ppname_is(cs, "DeviceCMYK")) + colors = 4; + else + return PP_LENGTH_UNKNOWN; + return (w * h * bpc * colors + 7) >> 3; + } + return PP_LENGTH_UNKNOWN; +} + +static ppstring ppstring_inline (iof *I, ppdict *imagedict, ppheap **pheap) +{ + iof *O; + int c, d, e; + size_t length, leftin, leftout, bytes; + + O = ppheap_buffer(pheap, sizeof(_ppstring), PPSTRING_INIT); + c = iof_char(I); + if (ignored_char(c)) + c = iof_next(I); + + length = inline_image_length(imagedict); + if (length != PP_LENGTH_UNKNOWN) + { + while (length > 0 && iof_readable(I) && iof_writable(O)) + { + leftin = iof_left(I); + leftout = iof_left(O); + bytes = length; + if (bytes > leftin) bytes = leftin; + if (bytes > leftout) bytes = leftout; + memcpy(O->pos, I->pos, bytes); + I->pos += bytes; + O->pos += bytes; + length -= bytes; + } + // gobble EI + if (ppscan_find(I) == 'E') + if (iof_next(I) == 'I') + ++I->pos; + } + else + { + while (c >= 0) + { + if (c == 'E') + { + d = iof_next(I); + if (d == 'I') + { + e = iof_next(I); + if (!ppname_byte_lookup[e]) + { /* strip one newline from the end and stop */ + if (O->pos - 2 >= O->buf) // sanity + { + c = *(O->pos - 1); + if (ignored_char(c)) + { + if (c == 0x0A && *(O->pos - 2) == 0x0D) + O->pos -= 2; + else + O->pos -= 1; + } + } + break; + } + iof_put2(O, c, d); + c = e; + } + else + { + iof_put(O, c); + c = d; + } + } + else + { + iof_put(O, c); + c = iof_next(I); + } + } + } + return ppstring_buffer(O, pheap); +} + +/* input reader */ + +/* +PDF input is a pseudo file that either keeps FILE * or data. Reader iof * is a proxy to input +that provides byte-by-byte interface. Our iof structure is capable to link iof_file *input, +but t avoid redundant checks on IOF_DATA flag, here we link iof *I directly to FILE * or mem buffer. +When reading from file we need an internal buffer, which should be kept rather small, as it is +only used to parse xrefs and objects (no streams). We allocate the buffer from a private heap +(not static) to avoid conflicts when processing >1 pdfs at once. Besides, the input buffer may be +needed after loading the document, eg. to access references raw data. +*/ + +#define PPDOC_BUFFER 0xFFF // keep that small, it is only used to parse body objects + +static void ppdoc_reader_init (ppdoc *pdf, iof_file *input) +{ + iof *I; + pdf->input = *input; + input = &pdf->input; + input->refcount = 1; + I = &pdf->reader; + if (input->flags & IOF_DATA) + { + pdf->buffer = NULL; // input iof_file is the buffer + iof_string_reader(I, NULL, 0); // gets IOF_DATA flag + } + else + { + pdf->buffer = (uint8_t *)ppheap_take(&pdf->heap, PPDOC_BUFFER); + iof_setup_file_handle_reader(I, NULL, 0, iof_file_get_fh(input)); // gets IOF_FILE_HANDLE flag and FILE * + I->space = PPDOC_BUFFER; // used on refill + } +} + +/* +Whenever we need to read the input file, we fseek the to the given offset and fread to to the private buffer. +The length we need is not always predictable, in which case PPDOC_BUFFER bytes are read (keep it small). +I->buf = I->pos is set to the beginning, I->end set to the end (end is the first byte one shouldn't read). +*/ + +static iof * ppdoc_reader (ppdoc *pdf, size_t offset, size_t length) +{ + iof_file *input; + iof *I; + input = &pdf->input; + I = &pdf->reader; + if (iof_file_seek(input, offset, SEEK_SET) != 0) + return NULL; + I->flags &= ~IOF_STOPPED; + if (input->flags & IOF_DATA) + { + I->buf = I->pos = input->pos; + I->end = (length == PP_LENGTH_UNKNOWN || I->pos + length >= input->end) ? input->end : (I->pos + length); + } + else + { + I->buf = I->pos = pdf->buffer; // ->buf is actually permanently equal pdf->buffer but we might need some tricks + if (length == PP_LENGTH_UNKNOWN || length > PPDOC_BUFFER) + length = PPDOC_BUFFER; + length = fread(I->buf, 1, length, I->file); + I->end = I->buf + length; + } + return I; +} + +/* The position from the beginning of input +- for data buffer: (pdf->input.pos - pdf->input.buf) + (I->pos - I->buf) + I->buf == pdf->input.pos, so this resolves to (I->pos - pdf->input.buf), independent from I->buf +- for file buffer: ftell(pdf->input.file) - (I->end - I->pos) +*/ + +#define ppdoc_reader_tell(pdf, I) ((size_t)(((pdf)->input.flags & IOF_DATA) ? ((I)->pos - (pdf)->input.buf) : (ftell(iof_file_get_fh(&(pdf)->input)) - ((I)->end - (I)->pos)))) + +/* pdf */ + +#define PPDOC_HEADER 10 // "%PDF-?.??\n" + +static int ppdoc_header (ppdoc *pdf, uint8_t header[PPDOC_HEADER]) +{ + size_t i; + if (memcmp(header, "%PDF-", 5) != 0) + return 0; + for (i = 5; i < PPDOC_HEADER - 1 && !ignored_char(header[i]); ++i) + pdf->version[i - 5] = header[i]; + pdf->version[i - 5] = '\0'; + return 1; +} + +static int ppdoc_tail (ppdoc *pdf, iof_file *input, size_t *pxrefoffset) +{ + int c; + uint8_t tail[4*10], *p, back, tailbytes; + + if (iof_file_seek(input, 0, SEEK_END) != 0) + return 0; + pdf->filesize = (size_t)iof_file_tell(input); + // simple heuristic to avoif fgetc() / fseek(-2) hiccup: keep seeking back by len(startxref) + 1 == 10 + // until a letter found (assuming liberal white characters and tail length) + for (back = 1, tailbytes = 0; ; ++back) + { + if (iof_file_seek(input, -10, SEEK_CUR) != 0) + return 0; + tailbytes += 10; + c = iof_file_getc(input); + tailbytes -= 1; + switch (c) + { + case IGNORED_CHAR_CASE: + case DIGIT_CHAR_CASE: + case '%': case 'E': case 'O': case 'F': + if (back > 4) // 2 should be enough + return 0; + continue; + case 's': case 't': case 'a': case 'r': case 'x': case 'e': case 'f': + if (iof_file_read(tail, 1, tailbytes, input) != tailbytes) + return 0; + tail[tailbytes] = '\0'; + for (p = &tail[0]; ; ++p) + { + if (*p == '\0') + return 0; + if ((c = base10_value(*p)) >= 0) + break; + } + ppread_uint(p, pxrefoffset); + return 1; + default: + return 0; + } + } + return 0; +} + +/* xref/body */ + +static int ppscan_start_entry (iof *I, ppref *ref) +{ + ppuint u; + ppscan_find(I); if (!ppscan_uint(I, &u) || u != ref->number) return 0; + ppscan_find(I); if (!ppscan_uint(I, &u) || u != ref->version) return 0; + ppscan_find(I); if (!ppscan_key(I, "obj")) return 0; + ppscan_find(I); + return 1; +} + +static int ppscan_skip_entry (iof *I) +{ + size_t u; + ppscan_find(I); if (!ppscan_uint(I, &u)) return 0; + ppscan_find(I); if (!ppscan_uint(I, &u)) return 0; + ppscan_find(I); if (!ppscan_key(I, "obj")) return 0; + ppscan_find(I); + return 1; +} + +static int ppscan_start_stream (iof *I, ppdoc *pdf, size_t *streamoffset) +{ + int c; + ppscan_find(I); + if (ppscan_key(I, "stream")) + { // skip 1 or 2 whites (here we shouldn't just gobble all blanks) + c = iof_char(I); + if (ignored_char(c)) + { + c = iof_next(I); + if (ignored_char(c)) + ++I->pos; + } + *streamoffset = ppdoc_reader_tell(pdf, I); + return 1; + } + return 0; +} + +static ppxref * ppxref_load (ppdoc *pdf, size_t xrefoffset); +static ppxref * ppxref_load_chain (ppdoc *pdf, ppxref *xref); + +/* Parsing xref table + + 1 10 // first ref number and refs count + 0000000000 00000 n // 10-digits offset, 5 digits version, type identifier + 0000000000 00000 n // n states for normal I guess + 0000000000 00000 f // f states for free (not used) + ... + +Free entries seem to be a relic of ancient times, completelly useless for us. To avoid parsing xref table twice, +we waste some space on free entries by allocating one plane of refs for each section. Later on we slice sections, +so that effectively free entries are not involved in map. + +Subsequent refs gets number, version and offset. Other fields initialized when parsing PDF body. + +Having xref table loaded, we sort sections for future binary search (xref with objects count == 0 is considered invalid). + +Then we have to deal with the trailer dict. In general, to load objects and resolve references we need a complete chain +of xrefs (not only the top). To load the previous xref, we need its offset, which is given in trailer. So we have to +parse the trailer ignoring references, which might be unresolvable at this point (objects parser makes a dummy check +for xref != NULL on refs resolving ppscan_obj(), which irritates me but I don't want a separate parser for trailer..). +The same applies to xref streams, in which we have parse the trailer not having xref map at all. So the procedure is: + + - load xref map, initialize references, make it ready to search + - parse trailer ignoring references + - get /Prev xref offset and load older xref (linked list via ->prev) + - sort all refs in all xrefs by offset + - parse refs in order resolving references in contained objects + - fix trailer references + +First created xref becomes a pdf->xref (top xref). We link that early to control offsets already read (insane loops?). +*/ + +// Every xref table item "0000000000 00000 n" is said to be terminated with 2-byte EOL but we don't like relying on whites. +#define xref_item_length (10+1+5+1+1) + +static ppxref * ppxref_load_table (iof *I, ppdoc *pdf, size_t xrefoffset) +{ + ppxref *xref; + ppxsec *xrefsection; + ppref *ref; + ppuint first, count, refindex; + uint8_t buffer[xref_item_length + 1]; + const char *p; + const ppobj *obj; + + buffer[xref_item_length] = '\0'; + xref = ppxref_create(pdf, 0, xrefoffset); + if (pdf->xref == NULL) pdf->xref = xref; + for (ppscan_find(I); ppscan_uint(I, &first); ppscan_find(I)) + { + ppscan_find(I); + if (!ppscan_uint(I, &count)) + return NULL; + if (count == 0) // weird + continue; + xref->count += count; + xrefsection = NULL; + ref = (ppref *)ppheap_take(&pdf->heap, count * sizeof(ppref)); + for (refindex = 0; refindex < count; ++refindex, ++ref) + { + ref->xref = xref; + ref->number = first + refindex; + ppscan_find(I); + iof_read(I, buffer, xref_item_length); + switch (buffer[xref_item_length - 1]) + { + case 'n': + if (xrefsection == NULL) + { + xrefsection = ppxref_push_section(xref, &pdf->heap); + xrefsection->first = ref->number; + xrefsection->refs = ref; + } + xrefsection->last = ref->number; + for (p = (const char *)buffer; *p == '0'; ++p); + p = ppread_uint(p, &ref->offset); + for ( ; *p == ' ' || *p == '0'; ++p); + p = ppread_uint(p, &ref->version); + ref->object.type = PPNONE; // init for sanity + ref->object.any = NULL; + ref->length = 0; + break; + case 'f': + default: + --ref; + xrefsection = NULL; + --xref->count; + } + } + } + /* sort section */ + if (!ppxref_sort(xref)) + ; // case of xref->size == 0 handled by ppxref_load_chain() + /* get trailer ignoring refs */ + if (!ppscan_key(I, "trailer")) + return NULL; + ppscan_find(I); + if ((obj = ppscan_obj(I, pdf, NULL)) == NULL) + return NULL; + ppstack_pop(&pdf->stack, 1); + if (obj->type != PPDICT) + return NULL; + xref->trailer = *obj; + return ppxref_load_chain(pdf, xref); +} + +/* Parsing xref stream +First we load the trailer, ignoring references. Dict defines sections and fields lengths: + + /Size % max ref number plus 1 + /Index [ first count first count ... ] % a pair of numbers for every section, defaults to [0 Size] + /W [w1 w2 w3] % fields lengths, 0 states for omitted field + +xref stream data is a continuous stream of binary number triplets. First number is a type: + + 0 - free entry (as 'f' in xref table) + 1 - normal entry, followed by offset an version (as 'n' in xref table) + 2 - compressed entry, followed by parent object stream number and entry index + +0 and 1 are handled as 'n' and 'f' entries in xref table. For type 2 we normally initialize +ref->number and ref->version (the later is implicitly 0). ref->offset is set to 0 (invalid offset), +which is recognized by objects loader. +*/ + +#define XREF_STREAM_MAX_FIELD 4 + +static ppxref * ppxref_load_stream (iof *I, ppdoc *pdf, size_t xrefoffset) +{ + ppxref *xref; + ppxsec *xrefsection; + ppref *ref; + ppobj *obj; + ppstream *xrefstream; + size_t streamoffset; + ppuint w1, w2, w3, w, bufferbytes; + uint8_t buffer[3 * XREF_STREAM_MAX_FIELD], *b; + ppuint first, count, f1, f2, f3; + pparray *fieldwidths, *sectionindices; + ppobj sectionmock[2], *sectionfirst, *sectioncount; + size_t sections, sectionindex, refindex; + + if (!ppscan_skip_entry(I)) + return NULL; + if ((obj = ppscan_obj(I, pdf, NULL)) == NULL) + return NULL; + ppstack_pop(&pdf->stack, 1); + if (obj->type != PPDICT || !ppscan_start_stream(I, pdf, &streamoffset)) + return NULL; + xrefstream = ppstream_create(pdf, obj->dict, streamoffset); + /* All normal streams go through ppstream_info(), but it makes no sense for trailer stream (no crypt allowed, no refs yet). + So we just record the length and informative flag. Here we have to expect that /Length and /Filter are not indirects. */ + if (!ppdict_get_uint(obj->dict, "Length", &xrefstream->length)) + return NULL; + if (ppdict_get_obj(obj->dict, "Filter") != NULL) + xrefstream->flags |= PPSTREAM_COMPRESSED; + if ((fieldwidths = ppdict_get_array(xrefstream->dict, "W")) != NULL) + { + if (!pparray_get_uint(fieldwidths, 0, &w1)) w1 = 0; + if (!pparray_get_uint(fieldwidths, 1, &w2)) w2 = 0; + if (!pparray_get_uint(fieldwidths, 2, &w3)) w3 = 0; + } + else + w1 = w2 = w3 = 0; + if (w1 > XREF_STREAM_MAX_FIELD || w2 > XREF_STREAM_MAX_FIELD || w3 > XREF_STREAM_MAX_FIELD) + return NULL; + bufferbytes = w1 + w2 + w3; + if ((sectionindices = ppdict_get_array(xrefstream->dict, "Index")) != NULL) + { + sections = sectionindices->size >> 1; + sectionfirst = sectionindices->data; + } + else + { + sections = 1; + sectionmock[0].type = PPINT; + sectionmock[0].integer = 0; + sectionmock[1].type = PPINT; + if (!ppdict_get_int(xrefstream->dict, "Size", §ionmock[1].integer)) + sectionmock[1].integer = 0; + sectionfirst = §ionmock[0]; + } + if ((I = ppstream_read(xrefstream, 1, 0)) == NULL) + return NULL; // we fseek() so original I is useless anyway + xref = ppxref_create(pdf, sections, xrefoffset); + if (pdf->xref == NULL) pdf->xref = xref; + xref->trailer.type = PPSTREAM; + xref->trailer.stream = xrefstream; + for (sectionindex = 0; sectionindex < sections; ++sectionindex, sectionfirst += 2) + { + sectioncount = sectionfirst + 1; + if (!ppobj_get_uint(sectionfirst, first) || !ppobj_get_uint(sectioncount, count)) + goto xref_stream_error; + if (count == 0) + continue; + xref->count += count; + xrefsection = NULL; + ref = (ppref *)ppheap_take(&pdf->heap, count * sizeof(ppref)); + for (refindex = 0; refindex < count; ++refindex, ++ref) + { + ref->xref = xref; + ref->number = first + refindex; + if (iof_read(I, buffer, bufferbytes) != bufferbytes) + goto xref_stream_error; + b = buffer; + if (w1 == 0) + f1 = 1; // default type is 1 + else + for (f1 = 0, w = 0; w < w1; f1 = (f1 << 8)|(*b), ++w, ++b); + for (f2 = 0, w = 0; w < w2; f2 = (f2 << 8)|(*b), ++w, ++b); + for (f3 = 0, w = 0; w < w3; f3 = (f3 << 8)|(*b), ++w, ++b); + switch (f1) + { + case 0: + //--ref; + xrefsection = NULL; + --xref->count; + break; + case 1: + if (xrefsection == NULL) + { + xrefsection = ppxref_push_section(xref, &pdf->heap); + xrefsection->first = ref->number; + xrefsection->refs = ref; + } + xrefsection->last = ref->number; + ref->offset = f2; + ref->version = f3; + ref->object.type = PPNONE; + ref->object.any = NULL; + ref->length = 0; + break; + case 2: + if (xrefsection == NULL) + { + xrefsection = ppxref_push_section(xref, &pdf->heap); + xrefsection->first = ref->number; + xrefsection->refs = ref; + } + xrefsection->last = ref->number; + ref->offset = 0; // f2 is parent objstm, f3 is index in parent, both useless + ref->version = 0; // compressed objects has implicit version == 0 + ref->object.type = PPNONE; + ref->object.any = NULL; + ref->length = 0; + break; + default: + goto xref_stream_error; + } + } + } + /* sort sections */ + if (!ppxref_sort(xref)) + ; // case of xref->size == 0 handled by ppxref_load_chain() + /* close the stream _before_ loading prev xref */ + ppstream_done(xrefstream); + /* load prev and return */ + return ppxref_load_chain(pdf, xref); +xref_stream_error: + ppstream_done(xrefstream); + return NULL; +} + +/* +The following procedure loads xref /Prev, links xref->prev and typically returns xref. +Some docs contain empty xref (one section with zero objects) that is actually a proxy +to xref stream referred as /XRefStm (genuine concept of xrefs old/new style xrefs in +the same doc). In case of 0-length xref we ignore the proxy and return the target xref +(otherwise we would need annoying sanity check for xref->size > 0 on every ref search). +*/ + +static ppxref * ppxref_load_chain (ppdoc *pdf, ppxref *xref) +{ + ppdict *trailer; + ppuint xrefoffset; + ppxref *prevxref, *nextxref; + + trailer = ppxref_trailer(xref); + if (!ppdict_get_uint(trailer, "Prev", &xrefoffset)) // XRefStm is useless + return xref; // missing /Prev is obviously ok + for (nextxref = pdf->xref; nextxref != NULL; nextxref = nextxref->prev) + if (nextxref->offset == xrefoffset) // insane + return NULL; + if ((prevxref = ppxref_load(pdf, (size_t)xrefoffset)) == NULL) + return NULL; + if (xref->size > 0) + { + xref->prev = prevxref; + return xref; + } + if (pdf->xref == xref) + pdf->xref = prevxref; + return prevxref; +} + +static ppxref * ppxref_load (ppdoc *pdf, size_t xrefoffset) +{ + iof *I; + if ((I = ppdoc_reader(pdf, xrefoffset, PP_LENGTH_UNKNOWN)) == NULL) + return NULL; + ppscan_find(I); + if (ppscan_key(I, "xref")) + return ppxref_load_table(I, pdf, xrefoffset); + return ppxref_load_stream(I, pdf, xrefoffset); + // iof_close(I) does nothing here +} + +static void ppoffmap_sort (ppref **left, ppref **right) +{ + ppref **l, **r, *t; + ppuint pivot; + l = left, r = right; + pivot = (*(l + ((r - l) / 2)))->offset; + do + { // don't read from pointer! + while ((*l)->offset < pivot) ++l; + while ((*r)->offset > pivot) --r; + if (l <= r) + { + t = *l; + *l = *r; + *r = t; + ++l, --r; + } + } while (l <= r); + if (left < r) + ppoffmap_sort(left, r); + if (l < right) + ppoffmap_sort(l, right); +} + + +static void fix_trailer_references (ppdoc *pdf) +{ + ppxref *xref; + ppdict *trailer; + ppname *pkey; + ppobj *obj; + ppref *ref; + for (xref = pdf->xref; xref != NULL; xref = xref->prev) + { + if ((trailer = ppxref_trailer(xref)) == NULL) + continue; + for (ppdict_first(trailer, pkey, obj); *pkey != NULL; ppdict_next(pkey, obj)) + { // no need to go deeper in structs, all items in trailer except info and root must be direct refs + if (obj->type != PPREF) + continue; + ref = obj->ref; + if (ref->offset == 0) // unresolved? + if ((ref = ppxref_find(xref, ref->number)) != NULL) + obj->ref = ref; // at this moment the reference still points nothing, but should be the one with the proper offset + } + } +} + +/* +Here comes a procedure that loads all entries from all document bodies. We resolve references while +parsing objects and to make resolving correct, we need a complete chain of xref maps, and a knowledge +about possible linearized dict (first offset). So loading refs sorted by offsets makes sense (not sure +if it matters nowadays but we also avoid fseek() by large offsets). + +Here is the proc: + + - create a list of all refs in all bodies + - sort the list by offsets + - for every ref from the sorted list: + - estimate object length to avoid fread-ing more than necessary (not perfect but enough) + - fseek() to the proper offset, fread() entry data or its part + - parse the object with ppscan_obj(I, pdf, xref), where xref is not necessarily top pdf->xref + - save the actual ref->length (not sure if we need that?) + - make a stream if a dict is followed by "stream" keyword, also save the stream offset + - free the list +*/ + +static int ppdoc_load_objstm (ppstream *stream, ppdoc *pdf, ppxref *xref); + +static void ppdoc_load_entries (ppdoc *pdf) +{ + size_t objects, sectionindex, refnumber, offindex; + ppnum linearized; + ppref **offmap, **pref, *ref; + ppxref *xref; + ppxsec *xsec; + ppobj *obj; + ppname type; + int redundant_indirection = 0; + ppcrypt *crypt; + ppstream *stream; + + if ((objects = (size_t)ppdoc_objects(pdf)) == 0) // can't happen + return; + pref = offmap = (ppref **)pp_malloc(objects * sizeof(ppref *)); + objects = 0; // recount refs with offset > 0 + for (xref = pdf->xref; xref != NULL; xref = xref->prev) + for (sectionindex = 0, xsec = xref->sects; sectionindex < xref->size; ++sectionindex, ++xsec) + for (refnumber = xsec->first, ref = xsec->refs; refnumber <= xsec->last; ++refnumber, ++ref) + if (ref->offset > 0) // 0 means compressed or insane + *pref++ = ref, ++objects; + ppoffmap_sort(offmap, offmap + objects - 1); + + crypt = pdf->crypt; + for (offindex = 0, pref = offmap; offindex < objects; ) + { + ref = *pref; + ++pref; + ++offindex; + if (ref->object.type != PPNONE) // might be preloaded already (/Encrypt dict, stream filter dicts, stream /Length..) + continue; + if (offindex < objects) + ref->length = (*pref)->offset - ref->offset; + else + ref->length = pdf->filesize > ref->offset ? pdf->filesize - ref->offset : 0; + if (crypt != NULL) + { + ppcrypt_start_ref(crypt, ref); + obj = ppdoc_load_entry(pdf, ref); + ppcrypt_end_ref(crypt); + } + else + { + obj = ppdoc_load_entry(pdf, ref); + } + switch (obj->type) + { + case PPDICT: /* Check if the object at first offset is linearized dict. We need that to resolve all references properly. */ + if (offindex == 1 && ppdict_get_num(obj->dict, "Linearized", &linearized)) // /Linearized value is a version number, default 1.0 + pdf->flags |= PPDOC_LINEARIZED; + break; + case PPREF: + redundant_indirection = 1; + break; + default: + break; + } + // if pdf->crypt crypt->ref = NULL + } + + /* refs pointngs refs? cut. */ + if (redundant_indirection) + { + for (offindex = 0, pref = offmap; offindex < objects; ++offindex) + { + ref = *pref++; + if (ref->object.type == PPREF) + ref->object = ref->object.ref->object; // doing for all effectively cuts all insane chains + } + } + + /* now handle streams; update stream info (eg. /Length), load pdf 1.5 object streams + we could do earlier but then we would need to struggle with indirects */ + for (offindex = 0, pref = offmap; offindex < objects; ++offindex) + { + ref = *pref++; + obj = &ref->object; + if (obj->type != PPSTREAM) + continue; + stream = obj->stream; + if (crypt != NULL) + { + ppcrypt_start_ref(crypt, ref); + ppstream_info(stream, pdf); + ppcrypt_end_ref(crypt); + } + else + { + ppstream_info(stream, pdf); + } + if (ref->xref->trailer.type == PPSTREAM && (type = ppdict_get_name(stream->dict, "Type")) != NULL && ppname_is(type, "ObjStm")) // somewhat dummy.. + if (!ppdoc_load_objstm(stream, pdf, ref->xref)) + loggerf("invalid objects stream %s at offset " PPSIZEF, ppref_str(ref->number, ref->version), ref->offset); + } + pp_free(offmap); +} + +ppobj * ppdoc_load_entry (ppdoc *pdf, ppref *ref) +{ + iof *I; + size_t length; + ppxref *xref; + ppobj *obj; + ppstack *stack; + size_t streamoffset; + ppref *refref; + ppuint refnumber, refversion; + + length = ref->length > 0 ? ref->length : PP_LENGTH_UNKNOWN; // estimated or unknown + if ((I = ppdoc_reader(pdf, ref->offset, length)) == NULL || !ppscan_start_entry(I, ref)) + { + loggerf("invalid %s offset " PPSIZEF, ppref_str(ref->number, ref->version), ref->offset); + return &ref->object; // PPNONE + } + stack = &pdf->stack; + xref = ref->xref; // to resolve indirects properly + if ((obj = ppscan_obj(I, pdf, xref)) == NULL) + { + loggerf("invalid %s object at offset " PPSIZEF, ppref_str(ref->number, ref->version), ref->offset); + return &ref->object; // PPNONE + } + ref->object = *obj; + ppstack_pop(stack, 1); + obj = &ref->object; + ref->length = ppdoc_reader_tell(pdf, I) - ref->offset; + if (obj->type == PPDICT) + { + if (ppscan_start_stream(I, pdf, &streamoffset)) + { + obj->type = PPSTREAM; + obj->stream = ppstream_create(pdf, obj->dict, streamoffset); + } + } + else if (obj->type == PPINT) + { + ppscan_find(I); + if (ppscan_uint(I, &refversion) && ppscan_find(I) == 'R') + { + refnumber = (ppuint)obj->integer; + if ((refref = ppxref_find(xref, refnumber)) != NULL) + { + obj->type = PPREF; + obj->ref = refref; + } + else + { + obj->type = PPNONE; // as ppref_unresolved() + obj->any = NULL; + } + } + } + return obj; +} + +/* Loading entries from object stream + + /N is the number of contained entries + /First is the offset of the first item + +The stream consists of N pairs of numbers <objnum> <offset> <objnum> <offset> ... +Offsets are ascending (relative to the first), but ref numbers order is arbitrary. +PDF spec says there might be some additional data between objects, so we should obey offsets. +Which means we should basically load the stream at once (may be needed anyway to grab the stream [...]). +*/ + +static int ppdoc_load_objstm (ppstream *stream, ppdoc *pdf, ppxref *xref) +{ + ppdict *dict; // stream dict, actually still on stack + ppref *ref; + ppobj *obj; + ppuint items, firstoffset, offset, objnum, i, invalid = 0; + iof *I; + uint8_t *firstdata, *indexdata; + ppstack *stack; + + dict = stream->dict; + if (!ppdict_rget_uint(dict, "N", &items) || !ppdict_rget_uint(dict, "First", &firstoffset)) + return 0; + if ((I = ppstream_read(stream, 1, 1)) == NULL) + return 0; + firstdata = I->pos + firstoffset; + if (firstdata >= I->end) + goto invalid_objstm; + stack = &pdf->stack; + //if (pdf->crypt != NULL) + // ppcrypt_end_ref(pdf->crypt); // objects are not encrypted, pdf->crypt->ref ensured NULL + for (i = 0; i < items; ++i) + { + ppscan_find(I); + if (!ppscan_uint(I, &objnum)) + goto invalid_objstm; + ppscan_find(I); + if (!ppscan_uint(I, &offset)) + goto invalid_objstm; + if ((ref = ppxref_find_local(xref, objnum)) == NULL || ref->object.type != PPNONE) + { + loggerf("invalid compressed object number " PPUINTF " at position " PPUINTF, objnum, i); + ++invalid; + continue; + } + if (firstdata + offset >= I->end) + { + loggerf("invalid compressed object offset " PPUINTF " at position " PPUINTF, offset, i); + ++invalid; + continue; + } + indexdata = I->pos; // save position + I->pos = firstdata + offset; // go to the object + ppscan_find(I); + if ((obj = ppscan_obj(I, pdf, xref)) != NULL) + { + ref->object = *obj; + ppstack_pop(stack, 1); + // nothing more needed, as obj can never be indirect ref or stream + } + else + { + ++invalid; + loggerf("invalid compressed object %s at stream offset " PPUINTF, ppref_str(objnum, 0), offset); + } + I->pos = indexdata; // restore position and read next from index + } + ppstream_done(stream); + return invalid == 0; +invalid_objstm: + ppstream_done(stream); + return 0; +} + +/* main PDF loader proc */ + +ppcrypt_status ppdoc_crypt_pass (ppdoc *pdf, const void *userpass, size_t userpasslength, const void *ownerpass, size_t ownerpasslength) +{ + switch (pdf->cryptstatus) + { + case PPCRYPT_NONE: + case PPCRYPT_DONE: + case PPCRYPT_FAIL: + break; + case PPCRYPT_PASS: // initial status or really needs password + pdf->cryptstatus = ppdoc_crypt_init(pdf, userpass, userpasslength, ownerpass, ownerpasslength); + switch (pdf->cryptstatus) + { + case PPCRYPT_NONE: + case PPCRYPT_DONE: + ppdoc_load_entries(pdf); + break; + case PPCRYPT_PASS: // user needs to check ppdoc_crypt_status() and recall ppdoc_crypt_pass() with the proper password + case PPCRYPT_FAIL: // hopeless.. + break; + } + break; + } + return pdf->cryptstatus; +} + +static ppdoc * ppdoc_read (ppdoc *pdf, iof_file *input) +{ + uint8_t header[PPDOC_HEADER]; + size_t xrefoffset; + + input = &pdf->input; + if (iof_file_read(header, 1, PPDOC_HEADER, input) != PPDOC_HEADER || !ppdoc_header(pdf, header)) + return NULL; + if (!ppdoc_tail(pdf, input, &xrefoffset)) + return NULL; + if (ppxref_load(pdf, xrefoffset) == NULL) + return NULL; + fix_trailer_references(pdf); // after loading xrefs but before accessing trailer refs (/Encrypt might be a reference) + // check encryption, if any, try empty password + switch (ppdoc_crypt_pass(pdf, "", 0, NULL, 0)) + { + case PPCRYPT_NONE: // no encryption + case PPCRYPT_DONE: // encryption with an empty password + case PPCRYPT_PASS: // the user needs to check ppdoc_crypt_status() and call ppdoc_crypt_pass() + break; + case PPCRYPT_FAIL: // hopeless + //loggerf("decryption failed"); + //return NULL; + break; + } + return pdf; +} + +static void ppdoc_pages_init (ppdoc *pdf); + +static ppdoc * ppdoc_create (iof_file *input) +{ + ppdoc *pdf; + ppheap *heap; + + heap = ppheap_new(); + pdf = (ppdoc *)ppheap_take(&heap, sizeof(ppdoc)); + pdf->flags = 0; + pdf->heap = heap; + pdf->xref = NULL; + pdf->version[0] = '\0'; + pdf->crypt = NULL; + pdf->cryptstatus = PPCRYPT_PASS; // force encryption check on ppdoc_read() -> ppdoc_crypt_pass() + ppstack_init(&pdf->stack, &pdf->heap); + ppdoc_reader_init(pdf, input); + ppdoc_pages_init(pdf); + if (ppdoc_read(pdf, &pdf->input) != NULL) + return pdf; + ppdoc_free(pdf); + return NULL; +} + +ppdoc * ppdoc_load (const char *filename) +{ + FILE *file; + iof_file input; + if ((file = fopen(filename, "rb")) == NULL) + return NULL; + iof_file_init(&input, file); + input.flags |= IOF_CLOSE_FILE; + return ppdoc_create(&input); +} + +ppdoc * ppdoc_mem (const void *data, size_t size) +{ + iof_file input; + iof_file_rdata_init(&input, data, size); + input.flags |= IOF_BUFFER_ALLOC; // todo: 3 modes: borrow, take over, copy? + return ppdoc_create(&input); +} + +void ppdoc_free (ppdoc *pdf) +{ + //iof_file_free(&pdf->input); + iof_file_decref(&pdf->input); + ppstack_free_buffer(&pdf->stack); + ppheap_free(pdf->heap); // last! +} + +ppcrypt_status ppdoc_crypt_status (ppdoc *pdf) +{ + return pdf->cryptstatus; +} + +ppint ppdoc_permissions (ppdoc *pdf) +{ + return pdf->crypt != NULL ? pdf->crypt->permissions : (ppint)0xFFFFFFFFFFFFFFFF; +} + +/* pages access */ + +static pparray * pppage_node (ppdict *dict, ppuint *count, ppname *type) +{ + ppname *pkey, key; + ppobj *obj; + pparray *kids = NULL; + *count = 0; + *type = NULL; + for (ppdict_first(dict, pkey, obj); (key = *pkey) != NULL; ppdict_next(pkey, obj)) + { + switch (key[0]) + { + case 'T': + if (ppname_is(key, "Type")) + *type = ppobj_get_name(obj); + break; + case 'C': + if (ppname_is(key, "Count")) + ppobj_get_uint(obj, *count); + break; + case 'K': + if (ppname_is(key, "Kids")) + kids = ppobj_rget_array(obj); + break; + } + } + return kids; +} + +#define ppname_is_page(type) (type != NULL && ppname_is(type, "Page")) + +ppuint ppdoc_page_count (ppdoc *pdf) +{ + ppref *ref; + ppname type; + ppuint count; + if ((ref = ppxref_pages(pdf->xref)) == NULL) + return 0; + if (pppage_node(ref->object.dict, &count, &type) == NULL) + return ppname_is_page(type) ? 1 : 0; // acrobat and ghostscript accept documents with root /Pages entry being a reference to a sole /Page object + return count; +} + +ppref * ppdoc_page (ppdoc *pdf, ppuint index) +{ + ppdict *dict; + ppuint count; + pparray *kids; + size_t size, i; + ppobj *r, *o; + ppref *ref; + ppname type; + + + if ((ref = ppxref_pages(pdf->xref)) == NULL) + return NULL; + dict = ref->object.dict; + if ((kids = pppage_node(dict, &count, &type)) != NULL) + { + if (index < 1 || index > count) + return NULL; + } + else + { + return index == 1 && ppname_is_page(type) ? ref : NULL; + } +scan_array: + if (index <= count / 2) + { // probably shorter way from the beginning + for (i = 0, size = kids->size, r = pparray_at(kids, 0); i < size; ++i, ++r) + { + if (r->type != PPREF) + return NULL; + o = &r->ref->object; + if (o->type != PPDICT) + return NULL; + dict = o->dict; + if ((kids = pppage_node(dict, &count, &type)) != NULL) + { + if (index <= count) + goto scan_array; + index -= count; + continue; + } + if (index == 1 && ppname_is_page(type)) + return r->ref; + --index; + } + } + else if ((size = kids->size) > 0) // for safe (size-1) + { // probably shorter way from the end + index = count - index + 1; + for (i = 0, r = pparray_at(kids, size - 1); i < size; ++i, --r) + { + if (r->type != PPREF) + return NULL; + o = &r->ref->object; + if (o->type != PPDICT) + return NULL; + dict = o->dict; + if ((kids = pppage_node(dict, &count, &type)) != NULL) + { + if (index <= count) + goto scan_array; + index -= count; + continue; + } + if (index == 1 && ppname_is_page(type)) + return r->ref; + --index; + } + } + return NULL; +} + +/* +Through pages iterator. Iterating over pages tree just on the base of /Kids and /Parent keys +is ineffective, as to get next pageref we need to take parent, find the pageref in /Kids, +take next (or go upper).. Annoying. We use a dedicated stack for pages iterator. This could +actually be done with pdf->stack, but some operations may clear it, so safer to keep it independent +Besides, its depth is constant (set on first use), so no need for allocs. +*/ + +static void ppdoc_pages_init (ppdoc *pdf) +{ + pppages *pages; + pages = &pdf->pages; + pages->root = pages->parent = pages->buffer; + pages->depth = 0; + pages->space = PPPAGES_STACK_DEPTH; +} + +static ppkids * pppages_push (ppdoc *pdf, pparray *kids) +{ + ppkids *newroot, *bounds; + pppages *pages; + pages = &pdf->pages; + if (pages->depth == pages->space) + { + pages->space <<= 1; + newroot = (ppkids *)ppheap_take(&pdf->heap, pages->space * sizeof(ppkids)); + memcpy(newroot, pages->root, pages->depth * sizeof(ppkids)); + pages->root = newroot; + } + bounds = pages->parent = &pages->root[pages->depth++]; + bounds->current = pparray_at(kids, 0); + bounds->sentinel = pparray_at(kids, kids->size); + return bounds; +} + +#define pppages_pop(pages) (--((pages)->parent), --((pages)->depth)) + +static ppref * ppdoc_pages_group_first (ppdoc *pdf, ppref *ref) +{ + ppdict *dict; + pparray *kids; + ppuint count; + ppname type; + + dict = ref->object.dict; // typecheck made by callers + while ((kids = pppage_node(dict, &count, &type)) != NULL) + { + if ((ref = pparray_get_ref(kids, 0)) == NULL || ref->object.type != PPDICT) + return NULL; + pppages_push(pdf, kids); + dict = ref->object.dict; + } + return ppname_is_page(type) ? ref : NULL; +} + +ppref * ppdoc_first_page (ppdoc *pdf) +{ + ppref *ref; + pppages *pages; + if ((ref = ppdoc_pages(pdf)) == NULL) + return NULL; + pages = &pdf->pages; + pages->parent = pages->root; + pages->depth = 0; + return ppdoc_pages_group_first(pdf, ref); +} + +ppref * ppdoc_next_page (ppdoc *pdf) +{ + pppages *pages; + ppkids *bounds; + ppref *ref; + ppobj *obj; + pages = &pdf->pages; + while (pages->depth > 0) + { + bounds = pages->parent; + obj = ++bounds->current; + if (obj < bounds->sentinel) + { + if (obj->type != PPREF) + return NULL; + ref = obj->ref; + if (ref->object.type != PPDICT) + return NULL; + return ppdoc_pages_group_first(pdf, ref); + } + else + { // no next node, go upper + pppages_pop(pages); + } + } + return NULL; +} + +/* context */ + +ppcontext * ppcontext_new (void) +{ + ppheap *heap; + ppcontext *context; + heap = ppheap_new(); + context = (ppcontext *)pp_malloc(sizeof(ppcontext)); // not from priv heap, as we delete it on renew + context->heap = heap; + ppstack_init(&context->stack, &context->heap); + return context; +} + +void ppcontext_done (ppcontext *context) +{ + ppheap_renew(context->heap); + ppstack_clear(&context->stack); +} + +void ppcontext_free (ppcontext *context) +{ + ppstack_free_buffer(&context->stack); + ppheap_free(context->heap); + pp_free(context); +} + +/* page contents streams */ + +//#define ppcontents_first_stream(array) pparray_rget_stream(array, 0) + +static ppstream * ppcontents_first_stream (pparray *array) +{ + size_t i; + ppobj *obj; + ppref *ref; + for (pparray_first(array, i, obj); i < array->size; pparray_next(i, obj)) + if ((ref = ppobj_get_ref(obj)) != NULL && ref->object.type == PPSTREAM) + return ref->object.stream; + return NULL; +} + +static ppstream * ppcontents_next_stream (pparray *array, ppstream *stream) +{ + size_t i; + ppobj *obj; + ppref *ref; + for (pparray_first(array, i, obj); i < array->size; pparray_next(i, obj)) + if ((ref = ppobj_get_ref(obj)) != NULL && ref->object.type == PPSTREAM && ref->object.stream == stream) + if (++i < array->size && (ref = ppobj_get_ref(obj + 1)) != NULL && ref->object.type == PPSTREAM) + return ref->object.stream; + return NULL; +} + +ppstream * ppcontents_first (ppdict *dict) +{ + ppobj *contentsobj; + if ((contentsobj = ppdict_rget_obj(dict, "Contents")) == NULL) + return NULL; + switch (contentsobj->type) + { + case PPARRAY: + return ppcontents_first_stream(contentsobj->array); + case PPSTREAM: + return contentsobj->stream; + default: + break; + } + return NULL; +} + +ppstream * ppcontents_next (ppdict *dict, ppstream *stream) +{ + ppobj *contentsobj; + if ((contentsobj = ppdict_rget_obj(dict, "Contents")) == NULL) + return NULL; + switch (contentsobj->type) + { + case PPARRAY: + return ppcontents_next_stream(contentsobj->array, stream); + case PPSTREAM: + break; + default: + break; + } + return NULL; +} + +static ppobj * ppcontents_op (iof *I, ppstack *stack, size_t *psize, ppname *pname) +{ + ppobj *obj; + ppstack_clear(stack); + do { + if (ppscan_find(I) < 0) + return NULL; + if ((obj = ppscan_psobj(I, stack)) == NULL) + return NULL; + } while (obj->type != PPNAME || !ppname_exec(obj->name)); + *pname = obj->name; + *psize = stack->size - 1; + return stack->buf; +} + +ppobj * ppcontents_first_op (ppcontext *context, ppstream *stream, size_t *psize, ppname *pname) +{ + iof *I; + if ((I = ppstream_read(stream, 1, 0)) == NULL) + return NULL; + return ppcontents_op(I, &context->stack, psize, pname); +} + +ppobj * ppcontents_next_op (ppcontext *context, ppstream *stream, size_t *psize, ppname *pname) +{ + return ppcontents_op(ppstream_iof(stream), &context->stack, psize, pname); +} + +ppobj * ppcontents_parse (ppcontext *context, ppstream *stream, size_t *psize) +{ + iof *I; + ppstack *stack; + ppobj *obj; + stack = &context->stack; + ppstack_clear(stack); + if ((I = ppstream_read(stream, 1, 0)) == NULL) + return NULL; + while (ppscan_find(I) >= 0) + if ((obj = ppscan_psobj(I, stack)) == NULL) + goto error; + *psize = stack->size; + ppstream_done(stream); + return stack->buf; +error: + ppstream_done(stream); + return NULL; +} + +/* boxes */ + +pprect * pparray_to_rect (pparray *array, pprect *rect) +{ + ppobj *obj; + if (array->size != 4) + return NULL; + obj = pparray_at(array, 0); + if (!ppobj_get_num(obj, rect->lx)) return NULL; + obj = pparray_at(array, 1); + if (!ppobj_get_num(obj, rect->ly)) return NULL; + obj = pparray_at(array, 2); + if (!ppobj_get_num(obj, rect->rx)) return NULL; + obj = pparray_at(array, 3); + if (!ppobj_get_num(obj, rect->ry)) return NULL; + return rect; +} + +pprect * ppdict_get_rect (ppdict *dict, const char *name, pprect *rect) +{ + pparray *array; + return (array = ppdict_rget_array(dict, name)) != NULL ? pparray_to_rect(array, rect) : NULL; +} + +pprect * ppdict_get_box (ppdict *dict, const char *name, pprect *rect) +{ + do { + if (ppdict_get_rect(dict, name, rect) != NULL) + return rect; + dict = ppdict_rget_dict(dict, "Parent"); + } while (dict != NULL); + return NULL; +} + +ppmatrix * pparray_to_matrix (pparray *array, ppmatrix *matrix) +{ + ppobj *obj; + if (array->size != 6) + return NULL; + obj = pparray_at(array, 0); + if (!ppobj_get_num(obj, matrix->xx)) return NULL; + obj = pparray_at(array, 1); + if (!ppobj_get_num(obj, matrix->xy)) return NULL; + obj = pparray_at(array, 2); + if (!ppobj_get_num(obj, matrix->yx)) return NULL; + obj = pparray_at(array, 3); + if (!ppobj_get_num(obj, matrix->yy)) return NULL; + obj = pparray_at(array, 4); + if (!ppobj_get_num(obj, matrix->x)) return NULL; + obj = pparray_at(array, 5); + if (!ppobj_get_num(obj, matrix->y)) return NULL; + return matrix; +} + +ppmatrix * ppdict_get_matrix (ppdict *dict, const char *name, ppmatrix *matrix) +{ + pparray *array; + return (array = ppdict_rget_array(dict, name)) != NULL ? pparray_to_matrix(array, matrix) : NULL; +} + +/* logger */ + +void pplog_callback (pplogger_callback logger, void *alien) +{ + logger_callback((logger_function)logger, alien); +} + +int pplog_prefix (const char *prefix) +{ + return logger_prefix(prefix); +} + +/* version */ + +const char * ppdoc_version_string (ppdoc *pdf) +{ + return pdf->version; +} + +int ppdoc_version_number (ppdoc *pdf, int *minor) +{ + *minor = pdf->version[2] - '0'; + return pdf->version[0] - '0'; +} + +/* doc info */ + +size_t ppdoc_file_size (ppdoc *pdf) +{ + return pdf->filesize; +} + +ppuint ppdoc_objects (ppdoc *pdf) +{ + ppuint count; + ppxref *xref; + for (count = 0, xref = pdf->xref; xref != NULL; xref = xref->prev) + count += xref->count; + return count; +} + +size_t ppdoc_memory (ppdoc *pdf, size_t *waste) +{ + size_t used; + ppheap *heap; + used = 0, *waste = 0; + for (heap = pdf->heap; heap != NULL; heap = heap->prev) + { + used += heap->space; + *waste += heap->size; + } + return used; +} diff --git a/Build/source/texk/web2c/luatexdir/luapplib/ppload.c-OK b/Build/source/texk/web2c/luatexdir/luapplib/ppload.c-OK new file mode 100644 index 00000000000..fda726656ab --- /dev/null +++ b/Build/source/texk/web2c/luatexdir/luapplib/ppload.c-OK @@ -0,0 +1,2555 @@ + +#include <utilbasexx.h> + +#include "pplib.h" + +const char * ppobj_kind[] = { "none", "null", "bool", "integer", "number", "name", "string", "array", "dict", "stream", "ref" }; + +#define ignored_char(c) (c == 0x20 || c == 0x0A || c == 0x0D || c == 0x09 || c == 0x00) +#define newline_char(c) (c == 0x0A || c == 0x0D) +#define IGNORED_CHAR_CASE 0x20: case 0x0A: case 0x0D: case 0x09: case 0x00 +#define NEWLINE_CHAR_CASE 0x0A: case 0x0D +#define DIGIT_CHAR_CASE '0': case '1': case '2': case '3': case '4': case '5': case '6': case '7': case '8': case '9' +#define OCTAL_CHAR_CASE '0': case '1': case '2': case '3': case '4': case '5': case '6': case '7' + +#define MAX_INT_DIGITS 32 + +#define PP_LENGTH_UNKNOWN ((size_t)-1) + +static const char * ppref_str (ppuint refnumber, ppuint refversion) +{ + static char buffer[MAX_INT_DIGITS + 1 + MAX_INT_DIGITS + 1 + 1 + 1]; +#if defined(MSVC64)|| defined(MINGW64) + sprintf(buffer, PPUINTF " " PPUINTF " R", refnumber, refversion); +#else + sprintf(buffer, PPUINTF " " PPUINTF " R", (unsigned long)(refnumber), (unsigned long)(refversion)); +#endif + return buffer; +} + +/* name */ + +// pdf name delimiters: 0..32, ()<>[]{}/% +// # treated specially +// .+- are valid part of name; keep in mind names such as -| | |- .notdef ABCDEF+Font etc. +const char ppname_byte_lookup[] = { + 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, + 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, + 0, 1, 1, '#', 1, 0, 1, 1, 0, 0, 1, 1, 1, 1, 1, 0, + 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 0, 1, 0, 1, + 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, + 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 0, 1, 0, 1, 1, + 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, + 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 0, 1, 0, 1, 1, + 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, + 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, + 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, + 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, + 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, + 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, + 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, + 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1 +}; + +#define PPNAME_INIT (8+1) + +#define ppname_flush(O, ghost, siz, flgs) \ + (iof_put(O, '\0'), \ + ghost = (_ppname *)ppheap_flush(O, &siz), \ + ghost->flags = flgs, \ + ghost->size = siz - sizeof(_ppname) - 1, \ + (ppname)(ghost + 1)) + +#define ppname_flush_with_ego(O, ghost, siz, flgs) \ + (iof_put(O, '\0'), \ + iof_ensure(O, sizeof(ppname *)), \ + O->pos += sizeof(ppname *), \ + ghost = (_ppname *)ppheap_flush(O, &siz), \ + ghost->flags = flgs, \ + ghost->size = siz - sizeof(_ppname) - 1 - sizeof(ppname *), \ + (ppname)(ghost + 1)) + +#define ppname_set_alter_ego(name, ghost, ego) (*((ppname *)(name + (ghost)->size + 1)) = ego) +#define ppname_get_alter_ego(name) (*((ppname *)(name + ppname_size(name) + 1))) + +static ppname ppscan_name (iof *I, ppheap **pheap) +{ + int c, decode; + iof *O; + _ppname *ghost1, *ghost2; + ppname encoded, decoded; + size_t size; + const char *p, *e; + uint8_t v1, v2; + O = ppheap_buffer(pheap, sizeof(_ppname), PPNAME_INIT); + for (decode = 0, c = iof_char(I); c >= 0 && ppname_byte_lookup[c]; c = iof_next(I)) + { + if (c == '#') + decode = 1; + iof_put(O, c); + } + if (!decode) + return ppname_flush(O, ghost1, size, 0); + encoded = ppname_flush_with_ego(O, ghost1, size, 0|PPNAME_ENCODED); + O = ppheap_buffer(pheap, sizeof(_ppname), PPNAME_INIT); + for (p = encoded, e = encoded + ghost1->size; p < e; ++p) + { + if (*p == '#' && p + 2 < e ){ + v1 = base16_value(p[1]); + v2 = base16_value(p[2]); + iof_put(O, ((v1<<4)+v2)); + }else + iof_put(O, *p); + } + decoded = ppname_flush_with_ego(O, ghost2, size, 0|PPNAME_DECODED); + ppname_set_alter_ego(encoded, ghost1, decoded); + ppname_set_alter_ego(decoded, ghost2, encoded); + return encoded; +} + +static ppname ppscan_exec (iof *I, ppheap **pheap, int first) +{ + int c, decode; + iof *O; + _ppname *ghost1, *ghost2; + ppname encoded, decoded; + size_t size; + const char *p, *e; + uint8_t v1, v2; + + O = ppheap_buffer(pheap, sizeof(_ppname), PPNAME_INIT); + iof_put(O, first); + for (decode = 0, c = iof_char(I); c >= 0 && ppname_byte_lookup[c]; c = iof_next(I)) + { + if (c == '#') + decode = 1; + iof_put(O, c); + } + if (!decode) + return ppname_flush(O, ghost1, size, PPNAME_EXEC); + encoded = ppname_flush_with_ego(O, ghost1, size, PPNAME_EXEC|PPNAME_ENCODED); + O = ppheap_buffer(pheap, sizeof(_ppname), PPNAME_INIT); + for (p = encoded, e = encoded + ghost1->size; p < e; ++p) + { + if (*p == '#' && p + 2 < e ){ + v1 = base16_value(p[1]); + v2 = base16_value(p[2]); + iof_put(O, ((v1<<4)+v2)); + }else + iof_put(O, *p); + } + decoded = ppname_flush_with_ego(O, ghost2, size, PPNAME_EXEC|PPNAME_DECODED); + ppname_set_alter_ego(encoded, ghost1, decoded); + ppname_set_alter_ego(decoded, ghost2, encoded); + return encoded; +} + +static ppname ppexec_internal (const void *data, size_t size, ppheap **pheap) +{ // used only for artificial 'EI' operator name + iof *O; + _ppname *ghost1; + + O = ppheap_buffer(pheap, sizeof(_ppname), size); + iof_write(O, data, size); + return ppname_flush(O, ghost1, size, PPNAME_EXEC); +} + +ppname ppname_decoded (ppname name) +{ + const _ppname *ghost; + ghost = _ppname_ghost(name); + return (ghost->flags & PPNAME_ENCODED) ? ppname_get_alter_ego(name) : name; +} + +ppname ppname_encoded (ppname name) +{ + const _ppname *ghost; + ghost = _ppname_ghost(name); + return (ghost->flags & PPNAME_DECODED) ? ppname_get_alter_ego(name) : name; +} + +/* string */ + +#define PPSTRING_INIT (16+1) + +#define ppstring_flush(O, ghost, siz, flgs) \ + (iof_put(O, '\0'), \ + ghost = (_ppstring *)ppheap_flush(O, &siz), \ + ghost->flags = flgs, \ + ghost->size = siz - sizeof(_ppstring) - 1, \ + (ppstring)(ghost + 1)) + +#define ppstring_flush_with_ego(O, ghost, siz, flgs) \ + (iof_put(O, '\0'), \ + iof_ensure(O, sizeof(ppstring *)), \ + O->pos += sizeof(ppstring *), \ + ghost = (_ppstring *)ppheap_flush(O, &siz), \ + ghost->flags = flgs, \ + ghost->size = siz - sizeof(_ppstring) - 1 - sizeof(ppstring *), \ + (ppstring)(ghost + 1)) + +#define ppstring_utf16be_bom(decoded) (decoded[0] == ((char)0xFE) && decoded[1] == ((char)0xFF) ) +#define ppstring_utf16le_bom(decoded) (decoded[0] == ((char)0xFF) && decoded[1] == ((char)0xFE)) + +#define ppstring_check_bom(decoded, ghost) ((void)\ + (ghost->size >= 2 ? (ppstring_utf16be_bom(decoded) ? (ghost->flags |= PPSTRING_UTF16BE) : \ + (ppstring_utf16le_bom(decoded) ? (ghost->flags |= PPSTRING_UTF16LE) : 0)) : 0)) + +#define ppstring_check_bom2(decoded, ghost1, ghost2) ((void)\ + (ghost2->size >= 2 ? (ppstring_utf16be_bom(decoded) ? ((ghost1->flags |= PPSTRING_UTF16BE), (ghost2->flags |= PPSTRING_UTF16BE)) : \ + (ppstring_utf16le_bom(decoded) ? ((ghost1->flags |= PPSTRING_UTF16LE), (ghost2->flags |= PPSTRING_UTF16LE)) : 0)) : 0)) + +#define ppstring_set_alter_ego(string, ghost, ego) (*((ppstring *)(string + (ghost)->size + 1)) = ego) +#define ppstring_get_alter_ego(string) (*((ppstring *)(string + ppstring_size(string) + 1))) + +static ppstring ppscan_string (iof *I, ppheap **pheap) +{ + int c, decode, balance; + iof *O; + _ppstring *ghost1, *ghost2; + uint8_t *p, *e; + ppstring encoded, decoded; + size_t size; + O = ppheap_buffer(pheap, sizeof(_ppstring), PPSTRING_INIT); + for (decode = 0, balance = 0, c = iof_char(I); c >= 0; ) + { + switch (c) + { + case '\\': + decode = 1; // unescaping later + iof_put(O, '\\'); + if ((c = iof_next(I)) >= 0) + { + iof_put(O, c); + c = iof_next(I); + } + break; + case '(': // may be unescaped if balanced + ++balance; + iof_put(O, '('); + c = iof_next(I); + break; + case ')': + if (balance == 0) + { + c = IOFEOF; + ++I->pos; + break; + } + --balance; + iof_put(O, ')'); + c = iof_next(I); + break; + default: + iof_put(O, c); + c = iof_next(I); + } + } + if (!decode) + { + encoded = ppstring_flush(O, ghost1, size, 0); + ppstring_check_bom(encoded, ghost1); // any bytes can be there + return encoded; + } + encoded = ppstring_flush_with_ego(O, ghost1, size, 0|PPSTRING_ENCODED); + O = ppheap_buffer(pheap, sizeof(_ppstring), PPSTRING_INIT); + for (p = (uint8_t *)encoded, e = (uint8_t *)encoded + ghost1->size; p < e; ++p) + { + if (*p == '\\') + { + if (++p >= e) + break; + switch (*p) + { + case OCTAL_CHAR_CASE: + c = *p - '0'; + if (++p < e && *p >= '0' && *p <= '7') + { + c = (c << 3) + *p - '0'; + if (++p < e && *p >= '0' && *p <= '7') + c = (c << 3) + *p - '0'; + } + iof_put(O, c); + break; + case 'n': + iof_put(O, '\n'); + break; + case 'r': + iof_put(O, '\r'); + break; + case 't': + iof_put(O, '\t'); + break; + case 'b': + iof_put(O, '\b'); + break; + case 'f': + iof_put(O, '\f'); + break; + case NEWLINE_CHAR_CASE: // not a part of the string, ignore (pdf spec page 55) + break; + case '(': case ')': case '\\': + default: // for enything else backslash is ignored (pdf spec page 54) + iof_put(O, *p); + break; + } + } + else + iof_put(O, *p); + } + decoded = ppstring_flush_with_ego(O, ghost2, size, 0|PPSTRING_DECODED); + ppstring_check_bom2(decoded, ghost1, ghost2); + ppstring_set_alter_ego(encoded, ghost1, decoded); + ppstring_set_alter_ego(decoded, ghost2, encoded); + return encoded; +} + +/* Hex string may contain white characters. If odd number of digits, the last assumed to be '0' */ + +static ppstring ppscan_base16 (iof *I, ppheap **pheap) +{ + int c, v1, v2; + iof *O; + _ppstring *ghost1, *ghost2; + size_t size; + ppstring encoded, decoded; + uint8_t *p, *e; + + O = ppheap_buffer(pheap, sizeof(_ppstring), PPSTRING_INIT); + for (c = iof_char(I); c >= 0 && (base16_digit(c) || ignored_char(c)); c = iof_next(I)) + iof_put(O, c); + if (c == '>') + ++I->pos; + encoded = ppstring_flush_with_ego(O, ghost1, size, PPSTRING_BASE16|PPSTRING_ENCODED); + O = ppheap_buffer(pheap, sizeof(_ppstring), (ghost1->size >> 1) + 1); + for (p = (uint8_t *)encoded, e = (uint8_t *)encoded + ghost1->size; p < e; ++p) + { + if ((v1 = base16_value(*p)) < 0) // ignored + continue; + for (v2 = 0, ++p; p < e && (v2 = base16_value(*p)) < 0; ++p); + iof_put(O, (v1<<4)|v2); + } + decoded = ppstring_flush_with_ego(O, ghost2, size, 0|PPSTRING_DECODED); + ppstring_check_bom2(decoded, ghost1, ghost2); + ppstring_set_alter_ego(encoded, ghost1, decoded); + ppstring_set_alter_ego(decoded, ghost2, encoded); + return encoded; +} + +/* internal use only; binary string */ + +static ppstring ppstring_buffer (iof *O, ppheap **pheap) +{ + _ppstring *ghost1, *ghost2; + ppstring encoded, decoded; + uint8_t *p, *e; + size_t size; + + decoded = ppstring_flush_with_ego(O, ghost2, size, 0|PPSTRING_DECODED); + O = ppheap_buffer(pheap, sizeof(_ppstring), (ghost2->size << 1) + 1); + for (p = (uint8_t *)decoded, e = (uint8_t *)decoded + ghost2->size; p < e; ++p) + iof_set2(O, base16_uc_alphabet[(*p)>>4], base16_uc_alphabet[(*p)&15]); + encoded = ppstring_flush_with_ego(O, ghost1, size, PPSTRING_BASE16|PPSTRING_ENCODED); + ppstring_set_alter_ego(encoded, ghost1, decoded); + ppstring_set_alter_ego(decoded, ghost2, encoded); + return encoded; +} + +ppstring ppstring_internal (const void *data, size_t size, ppheap **pheap) +{ // so far used only for crypt key + iof *O; + + O = ppheap_buffer(pheap, sizeof(_ppstring), size); + iof_write(O, data, size); + return ppstring_buffer(O, pheap); +} + +/* PDF spec says nothing about base85 strings, but streams might be (afair no leading '<~' but present trailing '~>') */ + +static ppstring ppscan_base85 (iof *I, ppheap **pheap) +{ // bawse85 alphabet is 33.117, adobe also hires 'z' and 'y' for compression + int c; + iof *O, B; + _ppstring *ghost1, *ghost2; + size_t size; + ppstring encoded, decoded; + O = ppheap_buffer(pheap, sizeof(_ppstring), PPSTRING_INIT); + for (c = iof_char(I); (c >= '!' && c <= 'u') || c == 'z' || c == 'y'; c = iof_next(I)) + iof_put(O, c); + if (c == '~') + if ((c = iof_next(I)) == '>') + ++I->pos; + encoded = ppstring_flush_with_ego(O, ghost1, size, PPSTRING_BASE85|PPSTRING_ENCODED); + iof_string_reader(&B, encoded, ghost1->size); + O = ppheap_buffer(pheap, sizeof(_ppstring), (ghost1->size * 5 / 4) + 1); + base85_decode(&B, O); + decoded = ppstring_flush_with_ego(O, ghost2, size, 0|PPSTRING_DECODED); + ppstring_check_bom2(decoded, ghost1, ghost2); + ppstring_set_alter_ego(encoded, ghost1, decoded); + ppstring_set_alter_ego(decoded, ghost2, encoded); + return encoded; +} + +/* +Encrypted strings. In case of encrypted strings, we first need to decode the string (saving original form hardly makes sense), +then decrypt the string, and encode it again. +*/ + +const char ppstring_byte_escape[] = { /* -1 escaped with octal, >0 escaped with \\, 0 left intact*/ + -1,-1,-1,-1,-1,-1,-1,-1,'b','t','n',-1,'f','r',-1,-1, + -1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1, + 0, 0, 0, 0, 0, 0, 0, 0,'(',')', 0, 0, 0, 0, 0, 0, + 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, + 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, + 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,'\\', 0, 0, 0, + 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, + 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, + -1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1, + -1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1, + -1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1, + -1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1, + -1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1, + -1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1, + -1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1, + -1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1 +}; + + +static ppstring ppscan_crypt_string (iof *I, ppcrypt *crypt, ppheap **pheap) +{ + int c, b, balance, encode; + iof *O; + _ppstring *ghost1, *ghost2; + ppstring encoded, decoded; + uint8_t *p, *e; + size_t size; + + O = ppheap_buffer(pheap, sizeof(_ppstring), PPSTRING_INIT); + for (balance = 0, encode = 0, c = iof_char(I); c >= 0; ) + { + switch (c) + { + case '\\': + if ((c = iof_next(I)) < 0) + break; + encode = 1; + switch (c) + { + case OCTAL_CHAR_CASE: + b = c - '0'; + if ((c = iof_next(I)) >= 0 && c >= '0' && c <= '7') + { + b = (b << 3) + c - '0'; + if ((c = iof_next(I)) >= 0 && c >= '0' && c <= '7') + { + b = (b << 3) + c - '0'; + c = iof_next(I); + } + } + iof_put(O, b); + // c is set to the next char + break; + case 'n': + iof_put(O, '\n'); + c = iof_next(I); + break; + case 'r': + iof_put(O, '\r'); + c = iof_next(I); + break; + case 't': + iof_put(O, '\t'); + c = iof_next(I); + break; + case 'b': + iof_put(O, '\b'); + c = iof_next(I); + break; + case 'f': + iof_put(O, '\f'); + c = iof_next(I); + break; + case NEWLINE_CHAR_CASE: // not a part of the string, ignore (pdf spec page 55) + c = iof_next(I); + break; + case '(': case ')': case '\\': + default: // for enything else backslash is ignored (pdf spec page 54) + iof_put(O, c); + c = iof_next(I); + break; + } + break; + case '(': + ++balance; + encode = 1; + iof_put(O, '('); + c = iof_next(I); + break; + case ')': + if (balance == 0) + { + c = IOFEOF; + ++I->pos; + } + else + { + --balance; + //encode = 1; + iof_put(O, ')'); + c = iof_next(I); + } + break; + default: + if (ppstring_byte_escape[c] != 0) + encode = 1; + iof_put(O, c); + c = iof_next(I); + } + } + /* decrypt the buffer in place, update size */ + if (ppstring_decrypt(crypt, O->buf, iof_size(O), O->buf, &size)) + O->pos = O->buf + size; + /* make encoded counterpart */ + if (!encode) + { + decoded = ppstring_flush(O, ghost2, size, 0); + ppstring_check_bom(decoded, ghost2); + return decoded; + } + decoded = ppstring_flush_with_ego(O, ghost2, size, PPSTRING_DECODED); + O = ppheap_buffer(pheap, sizeof(_ppstring), ghost2->size); + for (p = (uint8_t *)decoded, e = (uint8_t *)decoded + ghost2->size; p < e; ++p) + { + switch ((b = ppstring_byte_escape[*p])) + { + case 0: + iof_put(O, *p); + break; + case -1: + iof_put4(O, '\\', (c >> 6) + '0', ((c >> 3) & 7) + '0', (c & 7) + '0'); + break; + default: + iof_put2(O, '\\', b); + break; + } + } + encoded = ppstring_flush_with_ego(O, ghost1, size, PPSTRING_ENCODED); + ppstring_check_bom2(decoded, ghost1, ghost2); + ppstring_set_alter_ego(encoded, ghost1, decoded); + ppstring_set_alter_ego(decoded, ghost2, encoded); + return encoded; +} + + +static ppstring ppscan_crypt_base16 (iof *I, ppcrypt *crypt, ppheap **pheap) +{ + int c, v1, v2; + iof *O; + _ppstring *ghost1, *ghost2; + ppstring encoded, decoded; + uint8_t *p, *e; + size_t size; + + O = ppheap_buffer(pheap, sizeof(_ppstring), PPSTRING_INIT); + // base16_decode(I, O); // no info about the last char.. + for (c = iof_char(I); c != '>' && c >= 0; ) + { + if ((v1 = base16_value(c)) < 0) + { + if (ignored_char(c)) + { + c = iof_next(I); + continue; + } + break; + } + do { + c = iof_next(I); + if ((v2 = base16_value(c)) >= 0) + { + c = iof_next(I); + break; + } + if (!ignored_char(c)) // c == '>' || c < 0 or some crap + { + v2 = 0; + break; + } + } while (1); + iof_put(O, (v1 << 4)|v2); + } + if (c == '>') + ++I->pos; + /* decrypt the buffer in place, update size */ + if (ppstring_decrypt(crypt, O->buf, iof_size(O), O->buf, &size)) + O->pos = O->buf + size; + decoded = ppstring_flush_with_ego(O, ghost2, size, PPSTRING_DECODED); + /* recreate an encoded form */ + O = ppheap_buffer(pheap, sizeof(_ppstring), (ghost2->size << 1) + 1); + for (p = (uint8_t *)decoded, e = (uint8_t *)decoded + ghost2->size; p < e; ++p) + iof_set2(O, base16_uc_alphabet[(*p)>>4], base16_uc_alphabet[(*p)&15]); + encoded = ppstring_flush_with_ego(O, ghost1, size, PPSTRING_BASE16|PPSTRING_ENCODED); + ppstring_check_bom2(decoded, ghost1, ghost2); + ppstring_set_alter_ego(encoded, ghost1, decoded); + ppstring_set_alter_ego(decoded, ghost2, encoded); + return encoded; +} + +/* ppstring alter ego switcher */ + +ppstring ppstring_decoded (ppstring string) +{ + const _ppstring *ghost; + ghost = _ppstring_ghost(string); + return (ghost->flags & PPSTRING_ENCODED) ? ppstring_get_alter_ego(string) : string; +} + +ppstring ppstring_encoded (ppstring string) +{ + const _ppstring *ghost; + ghost = _ppstring_ghost(string); + return (ghost->flags & PPSTRING_DECODED) ? ppstring_get_alter_ego(string) : string; +} + +/* scanner stack */ + +#define PPSTACK_BUFFER 512 + +static void ppstack_init (ppstack *stack, ppheap **pheap) +{ + stack->buf = stack->pos = (ppobj *)pp_malloc(PPSTACK_BUFFER * sizeof(ppobj)); + stack->size = 0; + stack->space = PPSTACK_BUFFER; + stack->pheap = pheap; +} + +#define ppstack_free_buffer(stack) (pp_free((stack)->buf)) + +static void ppstack_resize (ppstack *stack) +{ + ppobj *newbuffer; + stack->space <<= 1; + newbuffer = (ppobj *)pp_malloc(stack->space * sizeof(ppobj)); + memcpy(newbuffer, stack->buf, stack->size * sizeof(ppobj)); + ppstack_free_buffer(stack); + stack->buf = newbuffer; + stack->pos = newbuffer + stack->size; +} + +#define ppstack_push(stack) ((void)((stack)->size < (stack)->space || (ppstack_resize(stack), 0)), ++(stack)->size, (stack)->pos++) +#define ppstack_pop(stack, n) ((stack)->size -= (n), (stack)->pos -= (n)) +#define ppstack_at(stack, i) ((stack)->buf + i) +#define ppstack_clear(stack) ((stack)->pos = (stack)->buf, (stack)->size = 0) + +/* scanner commons */ + +#define ppscan_uint(I, u) iof_get_uintlw(I, u) +#define ppread_uint(s, u) string_to_uintlw((const char *)(s), u) + +static ppobj * ppscan_numobj (iof *I, ppobj *obj, int negative) +{ + ppint integer; + ppnum number; + int exponent; + int c; + c = iof_char(I); + iof_scan_integer(I, c, integer); + switch(c) + { + case '.': + { + number = (ppnum)integer; + c = iof_next(I); + iof_scan_fraction(I, c, number, exponent); + double_negative_exp10(number, exponent); + obj->type = PPNUM, obj->number = negative ? -number : number; + break; + } + default: + obj->type = PPINT, obj->integer = negative ? -integer : integer; + break; + } + return obj; +} + +static ppobj * ppscan_numobj_frac (iof *I, ppobj *obj, int negative) +{ + ppnum number; + int c, exponent; + + number = 0.0; + c = iof_next(I); + iof_scan_fraction(I, c, number, exponent); + double_negative_exp10(number, exponent); + obj->type = PPNUM, obj->number = negative ? -number : number; + return obj; +} + +static int ppscan_find (iof *I) +{ // skips whitechars and comments + int c; + for (c = iof_char(I); ; c = iof_next(I)) + { + switch (c) + { + case IGNORED_CHAR_CASE: + break; + case '%': { + do { + if ((c = iof_next(I)) < 0) + return c; + } while (!newline_char(c)); + break; + } + default: + return c; + } + } + return c; // never reached +} + +static int ppscan_keyword (iof *I, const char *keyword, size_t size) +{ + size_t i; + int c; + if (iof_left(I) >= size) + { + if (memcmp(I->pos, keyword, size) != 0) + return 0; + I->pos += size; + return 1; + } + // sticky case, we can't go back + for (i = 0, c = iof_char(I); i < size; ++i, ++keyword, c = iof_next(I)) + if (i != c) + return 0; + return 1; +} + +#define ppscan_key(I, literal) ppscan_keyword(I, "" literal, sizeof(literal) - 1) + +/* objects parser */ + +static ppref * ppref_unresolved (ppheap **pheap, ppuint refnumber, ppuint refversion) +{ + ppref *ref = (ppref *)ppheap_take(pheap, sizeof(ppref)); + memset(ref, 0, sizeof(ppref)); + ref->object.type = PPNONE; + ref->number = refnumber; + ref->version = refversion; + return ref; +} + +#define PPMARK PPNONE + +static ppobj * ppscan_obj (iof *I, ppdoc *pdf, ppxref *xref) +{ + int c; + ppobj *obj; + size_t mark, size; + ppuint refnumber, refversion; + ppref *ref; + ppstack *stack; + ppcrypt *crypt; + + stack = &pdf->stack; + c = iof_char(I); + switch (c) + { + case DIGIT_CHAR_CASE: + return ppscan_numobj(I, ppstack_push(stack), 0); + case '.': + return ppscan_numobj_frac(I, ppstack_push(stack), 0); + case '+': + ++I->pos; + return ppscan_numobj(I, ppstack_push(stack), 0); + case '-': + ++I->pos; + return ppscan_numobj(I, ppstack_push(stack), 1); + case '/': + ++I->pos; + obj = ppstack_push(stack); + obj->type = PPNAME; + obj->name = ppscan_name(I, &pdf->heap); + return obj; + case '(': + ++I->pos; + obj = ppstack_push(stack); + obj->type = PPSTRING; + if (ppcrypt_ref(pdf, crypt)) + obj->string = ppscan_crypt_string(I, crypt, &pdf->heap); + else + obj->string = ppscan_string(I, &pdf->heap); + return obj; + case '[': + mark = stack->size; + obj = ppstack_push(stack); + obj->type = PPMARK; // ppscan_obj() checks types backward for 'R', so set the type immediatelly (reserved for PPARRAY) + obj->any = NULL; + ++I->pos; + for (c = ppscan_find(I); c != ']'; c = ppscan_find(I)) + { + if (ppscan_obj(I, pdf, xref) == NULL) + { // callers assume that NULL returns means nothing pushed + size = stack->size - mark; // pop items AND the obj reserved for array + ppstack_pop(stack, size); + return NULL; + } + } + ++I->pos; + size = stack->size - mark - 1; + obj = ppstack_at(stack, mark); // stack might have been realocated + obj->type = PPARRAY; + obj->array = pparray_create(ppstack_at(stack, mark + 1), size, &pdf->heap); + ppstack_pop(stack, size); // pop array items, leave the array on top + return obj; + case '<': + if ((c = iof_next(I)) == '<') + { + mark = stack->size; + obj = ppstack_push(stack); + obj->type = PPMARK; + obj->any = NULL; + ++I->pos; + for (c = ppscan_find(I); c != '>'; c = ppscan_find(I)) + { + if (ppscan_obj(I, pdf, xref) == NULL) + { + size = stack->size - mark; + ppstack_pop(stack, size); + return NULL; + } + } + if (iof_next(I) == '>') + ++I->pos; + size = stack->size - mark - 1; + obj = ppstack_at(stack, mark); + obj->type = PPDICT; + obj->dict = ppdict_create(ppstack_at(stack, mark + 1), size, &pdf->heap); + ppstack_pop(stack, size); + return obj; + } + obj = ppstack_push(stack); + obj->type = PPSTRING; + if (ppcrypt_ref(pdf, crypt)) + obj->string = ppscan_crypt_base16(I, crypt, &pdf->heap); + else + obj->string = ppscan_base16(I, &pdf->heap); + return obj; + case 'R': + if (stack->size >= 2 && stack->pos[-1].type == PPINT && stack->pos[-2].type == PPINT) + { + ++I->pos; + obj = &stack->pos[-2]; + refnumber = (ppuint)obj->integer; + ppstack_pop(stack, 1); // pop version number, retype obj to a reference + if (xref == NULL || (ref = ppxref_find(xref, refnumber)) == NULL) + { /* pdf spec page 64: unresolvable reference is not an error, should just be treated as a reference to null. + we also need this to read trailer, where refs can't be resolved yet */ + refversion = (obj + 1)->integer; + //if (xref != NULL) + // loggerf("unresolved reference %s", ppref_str(refnumber, refversion)); + ref = ppref_unresolved(stack->pheap, refnumber, refversion); + } + obj->type = PPREF; + obj->ref = ref; + return obj; + } + break; + case 't': + if (iof_next(I) == 'r' && iof_next(I) == 'u' && iof_next(I) == 'e') + { + ++I->pos; + obj = ppstack_push(stack); + obj->type = PPBOOL; + obj->integer = 1; + return obj; + } + break; + case 'f': + if (iof_next(I) == 'a' && iof_next(I) == 'l' && iof_next(I) == 's' && iof_next(I) == 'e') + { + ++I->pos; + obj = ppstack_push(stack); + obj->type = PPBOOL; + obj->integer = 0; + return obj; + } + break; + case 'n': + if (iof_next(I) == 'u' && iof_next(I) == 'l' && iof_next(I) == 'l') + { + ++I->pos; + obj = ppstack_push(stack); + obj->type = PPNULL; + obj->any = NULL; + return obj; + } + break; + } + return NULL; +} + +/* +A variant for contents streams (aka postscript); wise of operators, blind to references. +We are still PDF, so we don't care about postscript specific stuff such as radix numbers +and scientific numbers notation. It takes ppstack * as context (no ppdoc *) to be able +to run contents parser beyond the scope of ppdoc heap. +*/ + +static ppstring ppstring_inline (iof *I, ppdict *imagedict, ppheap **pheap); + +static ppobj * ppscan_psobj (iof *I, ppstack *stack) +{ + int c; + ppobj *obj, *op; + size_t size, mark; + ppname exec; + + c = iof_char(I); + switch (c) + { + case DIGIT_CHAR_CASE: + return ppscan_numobj(I, ppstack_push(stack), 0); + case '.': + return ppscan_numobj_frac(I, ppstack_push(stack), 0); + case '+': + c = iof_next(I); + if (base10_digit(c)) // '+.abc' is probably an executable name, but we are not in postscript + return ppscan_numobj(I, ppstack_push(stack), 0); + else if (c == '.') + return ppscan_numobj_frac(I, ppstack_push(stack), 0); + obj = ppstack_push(stack); + obj->type = PPNAME; + obj->name = ppscan_exec(I, stack->pheap, '+'); + return obj; + case '-': + c = iof_next(I); + if (base10_digit(c)) // ditto, we would handle type1 '-|' '|-' operators though + return ppscan_numobj(I, ppstack_push(stack), 1); + else if (c == '.') + return ppscan_numobj_frac(I, ppstack_push(stack), 1); + obj = ppstack_push(stack); + obj->type = PPNAME; + obj->name = ppscan_exec(I, stack->pheap, '-'); + return obj; + case '/': + ++I->pos; + obj = ppstack_push(stack); + obj->type = PPNAME; + obj->name = ppscan_name(I, stack->pheap); + return obj; + case '(': + ++I->pos; + obj = ppstack_push(stack); + obj->type = PPSTRING; + obj->string = ppscan_string(I, stack->pheap); + return obj; + case '[': + mark = stack->size; + obj = ppstack_push(stack); + obj->type = PPMARK; + obj->any = NULL; + ++I->pos; + for (c = ppscan_find(I); c != ']'; c = ppscan_find(I)) + { + if (ppscan_psobj(I, stack) == NULL) + { + size = stack->size - mark; + ppstack_pop(stack, size); + return NULL; + } + } + ++I->pos; + size = stack->size - mark - 1; + obj = ppstack_at(stack, mark); + obj->type = PPARRAY; + obj->array = pparray_create(ppstack_at(stack, mark + 1), size, stack->pheap); + ppstack_pop(stack, size); + return obj; + case '<': + if ((c = iof_next(I)) == '<') + { + mark = stack->size; + obj = ppstack_push(stack); + obj->type = PPMARK; + obj->any = NULL; + ++I->pos; + for (c = ppscan_find(I); c != '>'; c = ppscan_find(I)) + { + if (ppscan_psobj(I, stack) == NULL) + { + size = stack->size - mark; + ppstack_pop(stack, size); + return NULL; + } + } + if (iof_next(I) == '>') + ++I->pos; + size = stack->size - mark - 1; + obj = ppstack_at(stack, mark); + obj->type = PPDICT; + obj->dict = ppdict_create(ppstack_at(stack, mark + 1), size, stack->pheap); + ppstack_pop(stack, size); + return obj; + } + obj = ppstack_push(stack); + obj->type = PPSTRING; + if (c == '~') + ++I->pos, obj->string = ppscan_base85(I, stack->pheap); + else + obj->string = ppscan_base16(I, stack->pheap); + return obj; + default: + if (c < 0 || !ppname_byte_lookup[c]) + break; // forbid empty names; dead loop otherwise + ++I->pos; + /* true false null practically don't occur in streams so it makes sense to assume that we get an operator name here. + If it happen to be a keyword we could give back those several bytes to the heap but.. heap buffer is tricky enough. */ + exec = ppscan_exec(I, stack->pheap, c); + obj = ppstack_push(stack); + switch (exec[0]) + { + case 't': + if (exec[1] == 'r' && exec[2] == 'u' && exec[3] == 'e' && exec[4] == '\0') + { + obj->type = PPBOOL; + obj->integer = 1; + // todo: drop exec + return obj; + } + break; + case 'f': + if (exec[1] == 'a' && exec[2] == 'l' && exec[3] == 's' && exec[4] == 'e' && exec[5] == '\0') + { + obj->type = PPBOOL; + obj->integer = 0; + // todo: drop exec + return obj; + } + break; + case 'n': + if (exec[1] == 'u' && exec[2] == 'l' && exec[3] == 'l' && exec[4] == '\0') + { + obj->type = PPNULL; + obj->any = NULL; + // todo: drop exec + return obj; + } + break; + case 'B': + /* + Inline images break rules of operand/operator syntax, so 'BI/ID' operators need to be treated as special syntactic keywords. + + BI <keyval pairs> ID<whitechar?><imagedata><whitechar?>EI + + We treat the image as a single syntactic token; BI starts collecting a dict, ID is the beginning of the data. Effectively EI + operator obtains two operands - dict and string. It is ok to put three items onto the stack, callers dont't assume there is just one. + */ + if (exec[1] == 'I' && exec[2] == '\0') + { + ppdict *imagedict; + /* key val pairs -> dict */ + mark = stack->size - 1; + obj->type = PPMARK; + obj->any = NULL; + for (c = ppscan_find(I); ; c = ppscan_find(I)) + { + if ((op = ppscan_psobj(I, stack)) == NULL) + { + size = stack->size - mark; + ppstack_pop(stack, size); + return NULL; + } + if (op->type == PPNAME && ppname_exec(op->name)) + { + if (!ppname_is(op->name, "ID")) + { // weird + size = stack->size - mark; + ppstack_pop(stack, size); + return NULL; + } + break; + } + } + size = stack->size - mark - 1; + obj = ppstack_at(stack, mark); + obj->type = PPDICT; + obj->dict = imagedict = ppdict_create(ppstack_at(stack, mark + 1), size, stack->pheap); + ppstack_pop(stack, size); + /* put image data string */ + obj = ppstack_push(stack); + obj->type = PPSTRING; + obj->string = ppstring_inline(I, imagedict, stack->pheap);; + /* put EI operator name */ + obj = ppstack_push(stack); + obj->type = PPNAME; + obj->name = ppexec_internal("EI", 2, stack->pheap); + return obj; + } + } + obj->type = PPNAME; + obj->name = exec; + return obj; + } + return NULL; +} + +/* +We try to get the exact inline image length from its dict params. If cannot predict the length, we have to scan the input until 'EI'. +I've checked on may examples that it gives the same results but one can never be sure, as 'EI' might happen to be a part of the data. +Stripping white char is also very heuristic; \0 is a white char in PDF and very likely to be a data byte.. weak method. +*/ + +static size_t inline_image_length (ppdict *dict) +{ + ppuint w, h, bpc, colors; + ppname cs; + + if (ppdict_get_uint(dict, "W", &w) && ppdict_get_uint(dict, "H", &h) && ppdict_get_uint(dict, "BPC", &bpc) && (cs = ppdict_get_name(dict, "CS")) != NULL) + { + if (ppname_is(cs, "DeviceGray")) + colors = 1; + else if (ppname_is(cs, "DeviceRGB")) + colors = 3; + else if (ppname_is(cs, "DeviceCMYK")) + colors = 4; + else + return PP_LENGTH_UNKNOWN; + return (w * h * bpc * colors + 7) >> 3; + } + return PP_LENGTH_UNKNOWN; +} + +static ppstring ppstring_inline (iof *I, ppdict *imagedict, ppheap **pheap) +{ + iof *O; + int c, d, e; + size_t length, leftin, leftout, bytes; + + O = ppheap_buffer(pheap, sizeof(_ppstring), PPSTRING_INIT); + c = iof_char(I); + if (ignored_char(c)) + c = iof_next(I); + + length = inline_image_length(imagedict); + if (length != PP_LENGTH_UNKNOWN) + { + while (length > 0 && iof_readable(I) && iof_writable(O)) + { + leftin = iof_left(I); + leftout = iof_left(O); + bytes = length; + if (bytes > leftin) bytes = leftin; + if (bytes > leftout) bytes = leftout; + memcpy(O->pos, I->pos, bytes); + I->pos += bytes; + O->pos += bytes; + length -= bytes; + } + // gobble EI + if (ppscan_find(I) == 'E') + if (iof_next(I) == 'I') + ++I->pos; + } + else + { + while (c >= 0) + { + if (c == 'E') + { + d = iof_next(I); + if (d == 'I') + { + e = iof_next(I); + if (!ppname_byte_lookup[e]) + { /* strip one newline from the end and stop */ + if (O->pos - 2 >= O->buf) // sanity + { + c = *(O->pos - 1); + if (ignored_char(c)) + { + if (c == 0x0A && *(O->pos - 2) == 0x0D) + O->pos -= 2; + else + O->pos -= 1; + } + } + break; + } + iof_put2(O, c, d); + c = e; + } + else + { + iof_put(O, c); + c = d; + } + } + else + { + iof_put(O, c); + c = iof_next(I); + } + } + } + return ppstring_buffer(O, pheap); +} + +/* input reader */ + +/* +PDF input is a pseudo file that either keeps FILE * or data. Reader iof * is a proxy to input +that provides byte-by-byte interface. Our iof structure is capable to link iof_file *input, +but t avoid redundant checks on IOF_DATA flag, here we link iof *I directly to FILE * or mem buffer. +When reading from file we need an internal buffer, which should be kept rather small, as it is +only used to parse xrefs and objects (no streams). We allocate the buffer from a private heap +(not static) to avoid conflicts when processing >1 pdfs at once. Besides, the input buffer may be +needed after loading the document, eg. to access references raw data. +*/ + +#define PPDOC_BUFFER 0xFFF // keep that small, it is only used to parse body objects + +static void ppdoc_reader_init (ppdoc *pdf, iof_file *input) +{ + iof *I; + pdf->input = *input; + input = &pdf->input; + input->refcount = 1; + I = &pdf->reader; + if (input->flags & IOF_DATA) + { + pdf->buffer = NULL; // input iof_file is the buffer + iof_string_reader(I, NULL, 0); // gets IOF_DATA flag + } + else + { + pdf->buffer = (uint8_t *)ppheap_take(&pdf->heap, PPDOC_BUFFER); + iof_setup_file_handle_reader(I, NULL, 0, iof_file_get_fh(input)); // gets IOF_FILE_HANDLE flag and FILE * + I->space = PPDOC_BUFFER; // used on refill + } +} + +/* +Whenever we need to read the input file, we fseek the to the given offset and fread to to the private buffer. +The length we need is not always predictable, in which case PPDOC_BUFFER bytes are read (keep it small). +I->buf = I->pos is set to the beginning, I->end set to the end (end is the first byte one shouldn't read). +*/ + +static iof * ppdoc_reader (ppdoc *pdf, size_t offset, size_t length) +{ + iof_file *input; + iof *I; + input = &pdf->input; + I = &pdf->reader; + if (iof_file_seek(input, offset, SEEK_SET) != 0) + return NULL; + I->flags &= ~IOF_STOPPED; + if (input->flags & IOF_DATA) + { + I->buf = I->pos = input->pos; + I->end = (length == PP_LENGTH_UNKNOWN || I->pos + length >= input->end) ? input->end : (I->pos + length); + } + else + { + I->buf = I->pos = pdf->buffer; // ->buf is actually permanently equal pdf->buffer but we might need some tricks + if (length == PP_LENGTH_UNKNOWN || length > PPDOC_BUFFER) + length = PPDOC_BUFFER; + length = fread(I->buf, 1, length, I->file); + I->end = I->buf + length; + } + return I; +} + +/* The position from the beginning of input +- for data buffer: (pdf->input.pos - pdf->input.buf) + (I->pos - I->buf) + I->buf == pdf->input.pos, so this resolves to (I->pos - pdf->input.buf), independent from I->buf +- for file buffer: ftell(pdf->input.file) - (I->end - I->pos) +*/ + +#define ppdoc_reader_tell(pdf, I) ((size_t)(((pdf)->input.flags & IOF_DATA) ? ((I)->pos - (pdf)->input.buf) : (ftell(iof_file_get_fh(&(pdf)->input)) - ((I)->end - (I)->pos)))) + +/* pdf */ + +#define PPDOC_HEADER 10 // "%PDF-?.??\n" + +static int ppdoc_header (ppdoc *pdf, uint8_t header[PPDOC_HEADER]) +{ + size_t i; + if (memcmp(header, "%PDF-", 5) != 0) + return 0; + for (i = 5; i < PPDOC_HEADER - 1 && !ignored_char(header[i]); ++i) + pdf->version[i - 5] = header[i]; + pdf->version[i - 5] = '\0'; + return 1; +} + +static int ppdoc_tail (ppdoc *pdf, iof_file *input, size_t *pxrefoffset) +{ + int c; + uint8_t tail[4*10], *p, back, tailbytes; + + if (iof_file_seek(input, 0, SEEK_END) != 0) + return 0; + pdf->filesize = (size_t)iof_file_tell(input); + // simple heuristic to avoif fgetc() / fseek(-2) hiccup: keep seeking back by len(startxref) + 1 == 10 + // until a letter found (assuming liberal white characters and tail length) + for (back = 1, tailbytes = 0; ; ++back) + { + if (iof_file_seek(input, -10, SEEK_CUR) != 0) + return 0; + tailbytes += 10; + c = iof_file_getc(input); + tailbytes -= 1; + switch (c) + { + case IGNORED_CHAR_CASE: + case DIGIT_CHAR_CASE: + case '%': case 'E': case 'O': case 'F': + if (back > 4) // 2 should be enough + return 0; + continue; + case 's': case 't': case 'a': case 'r': case 'x': case 'e': case 'f': + if (iof_file_read(tail, 1, tailbytes, input) != tailbytes) + return 0; + tail[tailbytes] = '\0'; + for (p = &tail[0]; ; ++p) + { + if (*p == '\0') + return 0; + if ((c = base10_value(*p)) >= 0) + break; + } + ppread_uint(p, pxrefoffset); + return 1; + default: + return 0; + } + } + return 0; +} + +/* xref/body */ + +static int ppscan_start_entry (iof *I, ppref *ref) +{ + ppuint u; + ppscan_find(I); if (!ppscan_uint(I, &u) || u != ref->number) return 0; + ppscan_find(I); if (!ppscan_uint(I, &u) || u != ref->version) return 0; + ppscan_find(I); if (!ppscan_key(I, "obj")) return 0; + ppscan_find(I); + return 1; +} + +static int ppscan_skip_entry (iof *I) +{ + size_t u; + ppscan_find(I); if (!ppscan_uint(I, &u)) return 0; + ppscan_find(I); if (!ppscan_uint(I, &u)) return 0; + ppscan_find(I); if (!ppscan_key(I, "obj")) return 0; + ppscan_find(I); + return 1; +} + +static int ppscan_start_stream (iof *I, ppdoc *pdf, size_t *streamoffset) +{ + int c; + ppscan_find(I); + if (ppscan_key(I, "stream")) + { // skip 1 or 2 whites (here we shouldn't just gobble all blanks) + c = iof_char(I); + if (ignored_char(c)) + { + c = iof_next(I); + if (ignored_char(c)) + ++I->pos; + } + *streamoffset = ppdoc_reader_tell(pdf, I); + return 1; + } + return 0; +} + +static ppxref * ppxref_load (ppdoc *pdf, size_t xrefoffset); +static ppxref * ppxref_load_chain (ppdoc *pdf, ppxref *xref); + +/* Parsing xref table + + 1 10 // first ref number and refs count + 0000000000 00000 n // 10-digits offset, 5 digits version, type identifier + 0000000000 00000 n // n states for normal I guess + 0000000000 00000 f // f states for free (not used) + ... + +Free entries seem to be a relic of ancient times, completelly useless for us. To avoid parsing xref table twice, +we waste some space on free entries by allocating one plane of refs for each section. Later on we slice sections, +so that effectively free entries are not involved in map. + +Subsequent refs gets number, version and offset. Other fields initialized when parsing PDF body. + +Having xref table loaded, we sort sections for future binary search (xref with objects count == 0 is considered invalid). + +Then we have to deal with the trailer dict. In general, to load objects and resolve references we need a complete chain +of xrefs (not only the top). To load the previous xref, we need its offset, which is given in trailer. So we have to +parse the trailer ignoring references, which might be unresolvable at this point (objects parser makes a dummy check +for xref != NULL on refs resolving ppscan_obj(), which irritates me but I don't want a separate parser for trailer..). +The same applies to xref streams, in which we have parse the trailer not having xref map at all. So the procedure is: + + - load xref map, initialize references, make it ready to search + - parse trailer ignoring references + - get /Prev xref offset and load older xref (linked list via ->prev) + - sort all refs in all xrefs by offset + - parse refs in order resolving references in contained objects + - fix trailer references + +First created xref becomes a pdf->xref (top xref). We link that early to control offsets already read (insane loops?). +*/ + +// Every xref table item "0000000000 00000 n" is said to be terminated with 2-byte EOL but we don't like relying on whites. +#define xref_item_length (10+1+5+1+1) + +static ppxref * ppxref_load_table (iof *I, ppdoc *pdf, size_t xrefoffset) +{ + ppxref *xref; + ppxsec *xrefsection; + ppref *ref; + ppuint first, count, refindex; + uint8_t buffer[xref_item_length + 1]; + const char *p; + const ppobj *obj; + + buffer[xref_item_length] = '\0'; + xref = ppxref_create(pdf, 0, xrefoffset); + if (pdf->xref == NULL) pdf->xref = xref; + for (ppscan_find(I); ppscan_uint(I, &first); ppscan_find(I)) + { + ppscan_find(I); + if (!ppscan_uint(I, &count)) + return NULL; + if (count == 0) // weird + continue; + xref->count += count; + xrefsection = NULL; + ref = (ppref *)ppheap_take(&pdf->heap, count * sizeof(ppref)); + for (refindex = 0; refindex < count; ++refindex, ++ref) + { + ref->xref = xref; + ref->number = first + refindex; + ppscan_find(I); + iof_read(I, buffer, xref_item_length); + switch (buffer[xref_item_length - 1]) + { + case 'n': + if (xrefsection == NULL) + { + xrefsection = ppxref_push_section(xref, &pdf->heap); + xrefsection->first = ref->number; + xrefsection->refs = ref; + } + xrefsection->last = ref->number; + for (p = (const char *)buffer; *p == '0'; ++p); + p = ppread_uint(p, &ref->offset); + for ( ; *p == ' ' || *p == '0'; ++p); + p = ppread_uint(p, &ref->version); + ref->object.type = PPNONE; // init for sanity + ref->object.any = NULL; + ref->length = 0; + break; + case 'f': + default: + --ref; + xrefsection = NULL; + --xref->count; + } + } + } + /* sort section */ + if (!ppxref_sort(xref)) + ; // case of xref->size == 0 handled by ppxref_load_chain() + /* get trailer ignoring refs */ + if (!ppscan_key(I, "trailer")) + return NULL; + ppscan_find(I); + if ((obj = ppscan_obj(I, pdf, NULL)) == NULL) + return NULL; + ppstack_pop(&pdf->stack, 1); + if (obj->type != PPDICT) + return NULL; + xref->trailer = *obj; + return ppxref_load_chain(pdf, xref); +} + +/* Parsing xref stream +First we load the trailer, ignoring references. Dict defines sections and fields lengths: + + /Size % max ref number plus 1 + /Index [ first count first count ... ] % a pair of numbers for every section, defaults to [0 Size] + /W [w1 w2 w3] % fields lengths, 0 states for omitted field + +xref stream data is a continuous stream of binary number triplets. First number is a type: + + 0 - free entry (as 'f' in xref table) + 1 - normal entry, followed by offset an version (as 'n' in xref table) + 2 - compressed entry, followed by parent object stream number and entry index + +0 and 1 are handled as 'n' and 'f' entries in xref table. For type 2 we normally initialize +ref->number and ref->version (the later is implicitly 0). ref->offset is set to 0 (invalid offset), +which is recognized by objects loader. +*/ + +#define XREF_STREAM_MAX_FIELD 4 + +static ppxref * ppxref_load_stream (iof *I, ppdoc *pdf, size_t xrefoffset) +{ + ppxref *xref; + ppxsec *xrefsection; + ppref *ref; + ppobj *obj; + ppstream *xrefstream; + size_t streamoffset; + ppuint w1, w2, w3, w, bufferbytes; + uint8_t buffer[3 * XREF_STREAM_MAX_FIELD], *b; + ppuint first, count, f1, f2, f3; + pparray *fieldwidths, *sectionindices; + ppobj sectionmock[2], *sectionfirst, *sectioncount; + size_t sections, sectionindex, refindex; + + if (!ppscan_skip_entry(I)) + return NULL; + if ((obj = ppscan_obj(I, pdf, NULL)) == NULL) + return NULL; + ppstack_pop(&pdf->stack, 1); + if (obj->type != PPDICT || !ppscan_start_stream(I, pdf, &streamoffset)) + return NULL; + xrefstream = ppstream_create(pdf, obj->dict, streamoffset); + /* All normal streams go through ppstream_info(), but it makes no sense for trailer stream (no crypt allowed, no refs yet). + So we just record the length and informative flag. Here we have to expect that /Length and /Filter are not indirects. */ + if (!ppdict_get_uint(obj->dict, "Length", &xrefstream->length)) + return NULL; + if (ppdict_get_obj(obj->dict, "Filter") != NULL) + xrefstream->flags |= PPSTREAM_COMPRESSED; + if ((fieldwidths = ppdict_get_array(xrefstream->dict, "W")) != NULL) + { + if (!pparray_get_uint(fieldwidths, 0, &w1)) w1 = 0; + if (!pparray_get_uint(fieldwidths, 1, &w2)) w2 = 0; + if (!pparray_get_uint(fieldwidths, 2, &w3)) w3 = 0; + } + else + w1 = w2 = w3 = 0; + if (w1 > XREF_STREAM_MAX_FIELD || w2 > XREF_STREAM_MAX_FIELD || w3 > XREF_STREAM_MAX_FIELD) + return NULL; + bufferbytes = w1 + w2 + w3; + if ((sectionindices = ppdict_get_array(xrefstream->dict, "Index")) != NULL) + { + sections = sectionindices->size >> 1; + sectionfirst = sectionindices->data; + } + else + { + sections = 1; + sectionmock[0].type = PPINT; + sectionmock[0].integer = 0; + sectionmock[1].type = PPINT; + if (!ppdict_get_int(xrefstream->dict, "Size", §ionmock[1].integer)) + sectionmock[1].integer = 0; + sectionfirst = §ionmock[0]; + } + if ((I = ppstream_read(xrefstream, 1, 0)) == NULL) + return NULL; // we fseek() so original I is useless anyway + xref = ppxref_create(pdf, sections, xrefoffset); + if (pdf->xref == NULL) pdf->xref = xref; + xref->trailer.type = PPSTREAM; + xref->trailer.stream = xrefstream; + for (sectionindex = 0; sectionindex < sections; ++sectionindex, sectionfirst += 2) + { + sectioncount = sectionfirst + 1; + if (!ppobj_get_uint(sectionfirst, first) || !ppobj_get_uint(sectioncount, count)) + goto xref_stream_error; + if (count == 0) + continue; + xref->count += count; + xrefsection = NULL; + ref = (ppref *)ppheap_take(&pdf->heap, count * sizeof(ppref)); + for (refindex = 0; refindex < count; ++refindex, ++ref) + { + ref->xref = xref; + ref->number = first + refindex; + if (iof_read(I, buffer, bufferbytes) != bufferbytes) + goto xref_stream_error; + b = buffer; + if (w1 == 0) + f1 = 1; // default type is 1 + else + for (f1 = 0, w = 0; w < w1; f1 = (f1 << 8)|(*b), ++w, ++b); + for (f2 = 0, w = 0; w < w2; f2 = (f2 << 8)|(*b), ++w, ++b); + for (f3 = 0, w = 0; w < w3; f3 = (f3 << 8)|(*b), ++w, ++b); + switch (f1) + { + case 0: + //--ref; + xrefsection = NULL; + --xref->count; + break; + case 1: + if (xrefsection == NULL) + { + xrefsection = ppxref_push_section(xref, &pdf->heap); + xrefsection->first = ref->number; + xrefsection->refs = ref; + } + xrefsection->last = ref->number; + ref->offset = f2; + ref->version = f3; + ref->object.type = PPNONE; + ref->object.any = NULL; + ref->length = 0; + break; + case 2: + if (xrefsection == NULL) + { + xrefsection = ppxref_push_section(xref, &pdf->heap); + xrefsection->first = ref->number; + xrefsection->refs = ref; + } + xrefsection->last = ref->number; + ref->offset = 0; // f2 is parent objstm, f3 is index in parent, both useless + ref->version = 0; // compressed objects has implicit version == 0 + ref->object.type = PPNONE; + ref->object.any = NULL; + ref->length = 0; + break; + default: + goto xref_stream_error; + } + } + } + /* sort sections */ + if (!ppxref_sort(xref)) + ; // case of xref->size == 0 handled by ppxref_load_chain() + /* close the stream _before_ loading prev xref */ + ppstream_done(xrefstream); + /* load prev and return */ + return ppxref_load_chain(pdf, xref); +xref_stream_error: + ppstream_done(xrefstream); + return NULL; +} + +/* +The following procedure loads xref /Prev, links xref->prev and typically returns xref. +Some docs contain empty xref (one section with zero objects) that is actually a proxy +to xref stream referred as /XRefStm (genuine concept of xrefs old/new style xrefs in +the same doc). In case of 0-length xref we ignore the proxy and return the target xref +(otherwise we would need annoying sanity check for xref->size > 0 on every ref search). +*/ + +static ppxref * ppxref_load_chain (ppdoc *pdf, ppxref *xref) +{ + ppdict *trailer; + ppuint xrefoffset; + ppxref *prevxref, *nextxref; + + trailer = ppxref_trailer(xref); + if (!ppdict_get_uint(trailer, "Prev", &xrefoffset)) // XRefStm is useless + return xref; // missing /Prev is obviously ok + for (nextxref = pdf->xref; nextxref != NULL; nextxref = nextxref->prev) + if (nextxref->offset == xrefoffset) // insane + return NULL; + if ((prevxref = ppxref_load(pdf, (size_t)xrefoffset)) == NULL) + return NULL; + if (xref->size > 0) + { + xref->prev = prevxref; + return xref; + } + if (pdf->xref == xref) + pdf->xref = prevxref; + return prevxref; +} + +static ppxref * ppxref_load (ppdoc *pdf, size_t xrefoffset) +{ + iof *I; + if ((I = ppdoc_reader(pdf, xrefoffset, PP_LENGTH_UNKNOWN)) == NULL) + return NULL; + ppscan_find(I); + if (ppscan_key(I, "xref")) + return ppxref_load_table(I, pdf, xrefoffset); + return ppxref_load_stream(I, pdf, xrefoffset); + // iof_close(I) does nothing here +} + +static void ppoffmap_sort (ppref **left, ppref **right) +{ + ppref **l, **r, *t; + ppuint pivot; + l = left, r = right; + pivot = (*(l + ((r - l) / 2)))->offset; + do + { // don't read from pointer! + while ((*l)->offset < pivot) ++l; + while ((*r)->offset > pivot) --r; + if (l <= r) + { + t = *l; + *l = *r; + *r = t; + ++l, --r; + } + } while (l <= r); + if (left < r) + ppoffmap_sort(left, r); + if (l < right) + ppoffmap_sort(l, right); +} + + +static void fix_trailer_references (ppdoc *pdf) +{ + ppxref *xref; + ppdict *trailer; + ppname *pkey; + ppobj *obj; + ppref *ref; + for (xref = pdf->xref; xref != NULL; xref = xref->prev) + { + if ((trailer = ppxref_trailer(xref)) == NULL) + continue; + for (ppdict_first(trailer, pkey, obj); *pkey != NULL; ppdict_next(pkey, obj)) + { // no need to go deeper in structs, all items in trailer except info and root must be direct refs + if (obj->type != PPREF) + continue; + ref = obj->ref; + if (ref->offset == 0) // unresolved? + if ((ref = ppxref_find(xref, ref->number)) != NULL) + obj->ref = ref; // at this moment the reference still points nothing, but should be the one with the proper offset + } + } +} + +/* +Here comes a procedure that loads all entries from all document bodies. We resolve references while +parsing objects and to make resolving correct, we need a complete chain of xref maps, and a knowledge +about possible linearized dict (first offset). So loading refs sorted by offsets makes sense (not sure +if it matters nowadays but we also avoid fseek() by large offsets). + +Here is the proc: + + - create a list of all refs in all bodies + - sort the list by offsets + - for every ref from the sorted list: + - estimate object length to avoid fread-ing more than necessary (not perfect but enough) + - fseek() to the proper offset, fread() entry data or its part + - parse the object with ppscan_obj(I, pdf, xref), where xref is not necessarily top pdf->xref + - save the actual ref->length (not sure if we need that?) + - make a stream if a dict is followed by "stream" keyword, also save the stream offset + - free the list +*/ + +static int ppdoc_load_objstm (ppstream *stream, ppdoc *pdf, ppxref *xref); + +static void ppdoc_load_entries (ppdoc *pdf) +{ + size_t objects, sectionindex, refnumber, offindex; + ppnum linearized; + ppref **offmap, **pref, *ref; + ppxref *xref; + ppxsec *xsec; + ppobj *obj; + ppname type; + int redundant_indirection = 0; + ppcrypt *crypt; + ppstream *stream; + + if ((objects = (size_t)ppdoc_objects(pdf)) == 0) // can't happen + return; + pref = offmap = (ppref **)pp_malloc(objects * sizeof(ppref *)); + objects = 0; // recount refs with offset > 0 + for (xref = pdf->xref; xref != NULL; xref = xref->prev) + for (sectionindex = 0, xsec = xref->sects; sectionindex < xref->size; ++sectionindex, ++xsec) + for (refnumber = xsec->first, ref = xsec->refs; refnumber <= xsec->last; ++refnumber, ++ref) + if (ref->offset > 0) // 0 means compressed or insane + *pref++ = ref, ++objects; + ppoffmap_sort(offmap, offmap + objects - 1); + + crypt = pdf->crypt; + for (offindex = 0, pref = offmap; offindex < objects; ) + { + ref = *pref; + ++pref; + ++offindex; + if (ref->object.type != PPNONE) // might be preloaded already (/Encrypt dict, stream filter dicts, stream /Length..) + continue; + if (offindex < objects) + ref->length = (*pref)->offset - ref->offset; + else + ref->length = pdf->filesize > ref->offset ? pdf->filesize - ref->offset : 0; + if (crypt != NULL) + { + ppcrypt_start_ref(crypt, ref); + obj = ppdoc_load_entry(pdf, ref); + ppcrypt_end_ref(crypt); + } + else + { + obj = ppdoc_load_entry(pdf, ref); + } + switch (obj->type) + { + case PPDICT: /* Check if the object at first offset is linearized dict. We need that to resolve all references properly. */ + if (offindex == 1 && ppdict_get_num(obj->dict, "Linearized", &linearized)) // /Linearized value is a version number, default 1.0 + pdf->flags |= PPDOC_LINEARIZED; + break; + case PPREF: + redundant_indirection = 1; + break; + default: + break; + } + // if pdf->crypt crypt->ref = NULL + } + + /* refs pointngs refs? cut. */ + if (redundant_indirection) + { + for (offindex = 0, pref = offmap; offindex < objects; ++offindex) + { + ref = *pref++; + if (ref->object.type == PPREF) + ref->object = ref->object.ref->object; // doing for all effectively cuts all insane chains + } + } + + /* now handle streams; update stream info (eg. /Length), load pdf 1.5 object streams + we could do earlier but then we would need to struggle with indirects */ + for (offindex = 0, pref = offmap; offindex < objects; ++offindex) + { + ref = *pref++; + obj = &ref->object; + if (obj->type != PPSTREAM) + continue; + stream = obj->stream; + if (crypt != NULL) + { + ppcrypt_start_ref(crypt, ref); + ppstream_info(stream, pdf); + ppcrypt_end_ref(crypt); + } + else + { + ppstream_info(stream, pdf); + } + if (ref->xref->trailer.type == PPSTREAM && (type = ppdict_get_name(stream->dict, "Type")) != NULL && ppname_is(type, "ObjStm")) // somewhat dummy.. + if (!ppdoc_load_objstm(stream, pdf, ref->xref)) + loggerf("invalid objects stream %s at offset " PPSIZEF, ppref_str(ref->number, ref->version), ref->offset); + } + pp_free(offmap); +} + +ppobj * ppdoc_load_entry (ppdoc *pdf, ppref *ref) +{ + iof *I; + size_t length; + ppxref *xref; + ppobj *obj; + ppstack *stack; + size_t streamoffset; + ppref *refref; + ppuint refnumber, refversion; + + length = ref->length > 0 ? ref->length : PP_LENGTH_UNKNOWN; // estimated or unknown + if ((I = ppdoc_reader(pdf, ref->offset, length)) == NULL || !ppscan_start_entry(I, ref)) + { + loggerf("invalid %s offset " PPSIZEF, ppref_str(ref->number, ref->version), ref->offset); + return &ref->object; // PPNONE + } + stack = &pdf->stack; + xref = ref->xref; // to resolve indirects properly + if ((obj = ppscan_obj(I, pdf, xref)) == NULL) + { + loggerf("invalid %s object at offset " PPSIZEF, ppref_str(ref->number, ref->version), ref->offset); + return &ref->object; // PPNONE + } + ref->object = *obj; + ppstack_pop(stack, 1); + obj = &ref->object; + ref->length = ppdoc_reader_tell(pdf, I) - ref->offset; + if (obj->type == PPDICT) + { + if (ppscan_start_stream(I, pdf, &streamoffset)) + { + obj->type = PPSTREAM; + obj->stream = ppstream_create(pdf, obj->dict, streamoffset); + } + } + else if (obj->type == PPINT) + { + ppscan_find(I); + if (ppscan_uint(I, &refversion) && ppscan_find(I) == 'R') + { + refnumber = (ppuint)obj->integer; + if ((refref = ppxref_find(xref, refnumber)) != NULL) + { + obj->type = PPREF; + obj->ref = refref; + } + else + { + obj->type = PPNONE; // as ppref_unresolved() + obj->any = NULL; + } + } + } + return obj; +} + +/* Loading entries from object stream + + /N is the number of contained entries + /First is the offset of the first item + +The stream consists of N pairs of numbers <objnum> <offset> <objnum> <offset> ... +Offsets are ascending (relative to the first), but ref numbers order is arbitrary. +PDF spec says there might be some additional data between objects, so we should obey offsets. +Which means we should basically load the stream at once (may be needed anyway to grab the stream [...]). +*/ + +static int ppdoc_load_objstm (ppstream *stream, ppdoc *pdf, ppxref *xref) +{ + ppdict *dict; // stream dict, actually still on stack + ppref *ref; + ppobj *obj; + ppuint items, firstoffset, offset, objnum, i, invalid = 0; + iof *I; + uint8_t *firstdata, *indexdata; + ppstack *stack; + + dict = stream->dict; + if (!ppdict_rget_uint(dict, "N", &items) || !ppdict_rget_uint(dict, "First", &firstoffset)) + return 0; + if ((I = ppstream_read(stream, 1, 1)) == NULL) + return 0; + firstdata = I->pos + firstoffset; + if (firstdata >= I->end) + goto invalid_objstm; + stack = &pdf->stack; + //if (pdf->crypt != NULL) + // ppcrypt_end_ref(pdf->crypt); // objects are not encrypted, pdf->crypt->ref ensured NULL + for (i = 0; i < items; ++i) + { + ppscan_find(I); + if (!ppscan_uint(I, &objnum)) + goto invalid_objstm; + ppscan_find(I); + if (!ppscan_uint(I, &offset)) + goto invalid_objstm; + if ((ref = ppxref_find_local(xref, objnum)) == NULL || ref->object.type != PPNONE) + { + loggerf("invalid compressed object number " PPUINTF " at position " PPUINTF, objnum, i); + ++invalid; + continue; + } + if (firstdata + offset >= I->end) + { + loggerf("invalid compressed object offset " PPUINTF " at position " PPUINTF, offset, i); + ++invalid; + continue; + } + indexdata = I->pos; // save position + I->pos = firstdata + offset; // go to the object + ppscan_find(I); + if ((obj = ppscan_obj(I, pdf, xref)) != NULL) + { + ref->object = *obj; + ppstack_pop(stack, 1); + // nothing more needed, as obj can never be indirect ref or stream + } + else + { + ++invalid; + loggerf("invalid compressed object %s at stream offset " PPUINTF, ppref_str(objnum, 0), offset); + } + I->pos = indexdata; // restore position and read next from index + } + ppstream_done(stream); + return invalid == 0; +invalid_objstm: + ppstream_done(stream); + return 0; +} + +/* main PDF loader proc */ + +ppcrypt_status ppdoc_crypt_pass (ppdoc *pdf, const void *userpass, size_t userpasslength, const void *ownerpass, size_t ownerpasslength) +{ + switch (pdf->cryptstatus) + { + case PPCRYPT_NONE: + case PPCRYPT_DONE: + case PPCRYPT_FAIL: + break; + case PPCRYPT_PASS: // initial status or really needs password + pdf->cryptstatus = ppdoc_crypt_init(pdf, userpass, userpasslength, ownerpass, ownerpasslength); + switch (pdf->cryptstatus) + { + case PPCRYPT_NONE: + case PPCRYPT_DONE: + ppdoc_load_entries(pdf); + break; + case PPCRYPT_PASS: // user needs to check ppdoc_crypt_status() and recall ppdoc_crypt_pass() with the proper password + case PPCRYPT_FAIL: // hopeless.. + break; + } + break; + } + return pdf->cryptstatus; +} + +static ppdoc * ppdoc_read (ppdoc *pdf, iof_file *input) +{ + uint8_t header[PPDOC_HEADER]; + size_t xrefoffset; + + input = &pdf->input; + if (iof_file_read(header, 1, PPDOC_HEADER, input) != PPDOC_HEADER || !ppdoc_header(pdf, header)) + return NULL; + if (!ppdoc_tail(pdf, input, &xrefoffset)) + return NULL; + if (ppxref_load(pdf, xrefoffset) == NULL) + return NULL; + fix_trailer_references(pdf); // after loading xrefs but before accessing trailer refs (/Encrypt might be a reference) + // check encryption, if any, try empty password + switch (ppdoc_crypt_pass(pdf, "", 0, NULL, 0)) + { + case PPCRYPT_NONE: // no encryption + case PPCRYPT_DONE: // encryption with an empty password + case PPCRYPT_PASS: // the user needs to check ppdoc_crypt_status() and call ppdoc_crypt_pass() + break; + case PPCRYPT_FAIL: // hopeless + //loggerf("decryption failed"); + //return NULL; + break; + } + return pdf; +} + +static void ppdoc_pages_init (ppdoc *pdf); + +static ppdoc * ppdoc_create (iof_file *input) +{ + ppdoc *pdf; + ppheap *heap; + + heap = ppheap_new(); + pdf = (ppdoc *)ppheap_take(&heap, sizeof(ppdoc)); + pdf->flags = 0; + pdf->heap = heap; + pdf->xref = NULL; + pdf->version[0] = '\0'; + pdf->crypt = NULL; + pdf->cryptstatus = PPCRYPT_PASS; // force encryption check on ppdoc_read() -> ppdoc_crypt_pass() + ppstack_init(&pdf->stack, &pdf->heap); + ppdoc_reader_init(pdf, input); + ppdoc_pages_init(pdf); + if (ppdoc_read(pdf, &pdf->input) != NULL) + return pdf; + ppdoc_free(pdf); + return NULL; +} + +ppdoc * ppdoc_load (const char *filename) +{ + FILE *file; + iof_file input; + if ((file = fopen(filename, "rb")) == NULL) + return NULL; + iof_file_init(&input, file); + input.flags |= IOF_CLOSE_FILE; + return ppdoc_create(&input); +} + +ppdoc * ppdoc_mem (const void *data, size_t size) +{ + iof_file input; + iof_file_rdata_init(&input, data, size); + input.flags |= IOF_BUFFER_ALLOC; // todo: 3 modes: borrow, take over, copy? + return ppdoc_create(&input); +} + +void ppdoc_free (ppdoc *pdf) +{ + //iof_file_free(&pdf->input); + iof_file_decref(&pdf->input); + ppstack_free_buffer(&pdf->stack); + ppheap_free(pdf->heap); // last! +} + +ppcrypt_status ppdoc_crypt_status (ppdoc *pdf) +{ + return pdf->cryptstatus; +} + +ppint ppdoc_permissions (ppdoc *pdf) +{ + return pdf->crypt != NULL ? pdf->crypt->permissions : (ppint)0xFFFFFFFFFFFFFFFF; +} + +/* pages access */ + +static pparray * pppage_node (ppdict *dict, ppuint *count, ppname *type) +{ + ppname *pkey, key; + ppobj *obj; + pparray *kids = NULL; + *count = 0; + *type = NULL; + for (ppdict_first(dict, pkey, obj); (key = *pkey) != NULL; ppdict_next(pkey, obj)) + { + switch (key[0]) + { + case 'T': + if (ppname_is(key, "Type")) + *type = ppobj_get_name(obj); + break; + case 'C': + if (ppname_is(key, "Count")) + ppobj_get_uint(obj, *count); + break; + case 'K': + if (ppname_is(key, "Kids")) + kids = ppobj_rget_array(obj); + break; + } + } + return kids; +} + +#define ppname_is_page(type) (type != NULL && ppname_is(type, "Page")) + +ppuint ppdoc_page_count (ppdoc *pdf) +{ + ppref *ref; + ppname type; + ppuint count; + if ((ref = ppxref_pages(pdf->xref)) == NULL) + return 0; + if (pppage_node(ref->object.dict, &count, &type) == NULL) + return ppname_is_page(type) ? 1 : 0; // acrobat and ghostscript accept documents with root /Pages entry being a reference to a sole /Page object + return count; +} + +ppref * ppdoc_page (ppdoc *pdf, ppuint index) +{ + ppdict *dict; + ppuint count; + pparray *kids; + size_t size, i; + ppobj *r, *o; + ppref *ref; + ppname type; + + + if ((ref = ppxref_pages(pdf->xref)) == NULL) + return NULL; + dict = ref->object.dict; + if ((kids = pppage_node(dict, &count, &type)) != NULL) + { + if (index < 1 || index > count) + return NULL; + } + else + { + return index == 1 && ppname_is_page(type) ? ref : NULL; + } +scan_array: + if (index <= count / 2) + { // probably shorter way from the beginning + for (i = 0, size = kids->size, r = pparray_at(kids, 0); i < size; ++i, ++r) + { + if (r->type != PPREF) + return NULL; + o = &r->ref->object; + if (o->type != PPDICT) + return NULL; + dict = o->dict; + if ((kids = pppage_node(dict, &count, &type)) != NULL) + { + if (index <= count) + goto scan_array; + index -= count; + continue; + } + if (index == 1 && ppname_is_page(type)) + return r->ref; + --index; + } + } + else if ((size = kids->size) > 0) // for safe (size-1) + { // probably shorter way from the end + index = count - index + 1; + for (i = 0, r = pparray_at(kids, size - 1); i < size; ++i, --r) + { + if (r->type != PPREF) + return NULL; + o = &r->ref->object; + if (o->type != PPDICT) + return NULL; + dict = o->dict; + if ((kids = pppage_node(dict, &count, &type)) != NULL) + { + if (index <= count) + goto scan_array; + index -= count; + continue; + } + if (index == 1 && ppname_is_page(type)) + return r->ref; + --index; + } + } + return NULL; +} + +/* +Through pages iterator. Iterating over pages tree just on the base of /Kids and /Parent keys +is ineffective, as to get next pageref we need to take parent, find the pageref in /Kids, +take next (or go upper).. Annoying. We use a dedicated stack for pages iterator. This could +actually be done with pdf->stack, but some operations may clear it, so safer to keep it independent +Besides, its depth is constant (set on first use), so no need for allocs. +*/ + +static void ppdoc_pages_init (ppdoc *pdf) +{ + pppages *pages; + pages = &pdf->pages; + pages->root = pages->parent = pages->buffer; + pages->depth = 0; + pages->space = PPPAGES_STACK_DEPTH; +} + +static ppkids * pppages_push (ppdoc *pdf, pparray *kids) +{ + ppkids *newroot, *bounds; + pppages *pages; + pages = &pdf->pages; + if (pages->depth == pages->space) + { + pages->space <<= 1; + newroot = (ppkids *)ppheap_take(&pdf->heap, pages->space * sizeof(ppkids)); + memcpy(newroot, pages->root, pages->depth * sizeof(ppkids)); + pages->root = newroot; + } + bounds = pages->parent = &pages->root[pages->depth++]; + bounds->current = pparray_at(kids, 0); + bounds->sentinel = pparray_at(kids, kids->size); + return bounds; +} + +#define pppages_pop(pages) (--((pages)->parent), --((pages)->depth)) + +static ppref * ppdoc_pages_group_first (ppdoc *pdf, ppref *ref) +{ + ppdict *dict; + pparray *kids; + ppuint count; + ppname type; + + dict = ref->object.dict; // typecheck made by callers + while ((kids = pppage_node(dict, &count, &type)) != NULL) + { + if ((ref = pparray_get_ref(kids, 0)) == NULL || ref->object.type != PPDICT) + return NULL; + pppages_push(pdf, kids); + dict = ref->object.dict; + } + return ppname_is_page(type) ? ref : NULL; +} + +ppref * ppdoc_first_page (ppdoc *pdf) +{ + ppref *ref; + pppages *pages; + if ((ref = ppdoc_pages(pdf)) == NULL) + return NULL; + pages = &pdf->pages; + pages->parent = pages->root; + pages->depth = 0; + return ppdoc_pages_group_first(pdf, ref); +} + +ppref * ppdoc_next_page (ppdoc *pdf) +{ + pppages *pages; + ppkids *bounds; + ppref *ref; + ppobj *obj; + pages = &pdf->pages; + while (pages->depth > 0) + { + bounds = pages->parent; + obj = ++bounds->current; + if (obj < bounds->sentinel) + { + if (obj->type != PPREF) + return NULL; + ref = obj->ref; + if (ref->object.type != PPDICT) + return NULL; + return ppdoc_pages_group_first(pdf, ref); + } + else + { // no next node, go upper + pppages_pop(pages); + } + } + return NULL; +} + +/* context */ + +ppcontext * ppcontext_new (void) +{ + ppheap *heap; + ppcontext *context; + heap = ppheap_new(); + context = (ppcontext *)pp_malloc(sizeof(ppcontext)); // not from priv heap, as we delete it on renew + context->heap = heap; + ppstack_init(&context->stack, &context->heap); + return context; +} + +void ppcontext_done (ppcontext *context) +{ + ppheap_renew(context->heap); + ppstack_clear(&context->stack); +} + +void ppcontext_free (ppcontext *context) +{ + ppstack_free_buffer(&context->stack); + ppheap_free(context->heap); + pp_free(context); +} + +/* page contents streams */ + +//#define ppcontents_first_stream(array) pparray_rget_stream(array, 0) + +static ppstream * ppcontents_first_stream (pparray *array) +{ + size_t i; + ppobj *obj; + ppref *ref; + for (pparray_first(array, i, obj); i < array->size; pparray_next(i, obj)) + if ((ref = ppobj_get_ref(obj)) != NULL && ref->object.type == PPSTREAM) + return ref->object.stream; + return NULL; +} + +static ppstream * ppcontents_next_stream (pparray *array, ppstream *stream) +{ + size_t i; + ppobj *obj; + ppref *ref; + for (pparray_first(array, i, obj); i < array->size; pparray_next(i, obj)) + if ((ref = ppobj_get_ref(obj)) != NULL && ref->object.type == PPSTREAM && ref->object.stream == stream) + if (++i < array->size && (ref = ppobj_get_ref(obj + 1)) != NULL && ref->object.type == PPSTREAM) + return ref->object.stream; + return NULL; +} + +ppstream * ppcontents_first (ppdict *dict) +{ + ppobj *contentsobj; + if ((contentsobj = ppdict_rget_obj(dict, "Contents")) == NULL) + return NULL; + switch (contentsobj->type) + { + case PPARRAY: + return ppcontents_first_stream(contentsobj->array); + case PPSTREAM: + return contentsobj->stream; + default: + break; + } + return NULL; +} + +ppstream * ppcontents_next (ppdict *dict, ppstream *stream) +{ + ppobj *contentsobj; + if ((contentsobj = ppdict_rget_obj(dict, "Contents")) == NULL) + return NULL; + switch (contentsobj->type) + { + case PPARRAY: + return ppcontents_next_stream(contentsobj->array, stream); + case PPSTREAM: + break; + default: + break; + } + return NULL; +} + +static ppobj * ppcontents_op (iof *I, ppstack *stack, size_t *psize, ppname *pname) +{ + ppobj *obj; + ppstack_clear(stack); + do { + if (ppscan_find(I) < 0) + return NULL; + if ((obj = ppscan_psobj(I, stack)) == NULL) + return NULL; + } while (obj->type != PPNAME || !ppname_exec(obj->name)); + *pname = obj->name; + *psize = stack->size - 1; + return stack->buf; +} + +ppobj * ppcontents_first_op (ppcontext *context, ppstream *stream, size_t *psize, ppname *pname) +{ + iof *I; + if ((I = ppstream_read(stream, 1, 0)) == NULL) + return NULL; + return ppcontents_op(I, &context->stack, psize, pname); +} + +ppobj * ppcontents_next_op (ppcontext *context, ppstream *stream, size_t *psize, ppname *pname) +{ + return ppcontents_op(ppstream_iof(stream), &context->stack, psize, pname); +} + +ppobj * ppcontents_parse (ppcontext *context, ppstream *stream, size_t *psize) +{ + iof *I; + ppstack *stack; + ppobj *obj; + stack = &context->stack; + ppstack_clear(stack); + if ((I = ppstream_read(stream, 1, 0)) == NULL) + return NULL; + while (ppscan_find(I) >= 0) + if ((obj = ppscan_psobj(I, stack)) == NULL) + goto error; + *psize = stack->size; + ppstream_done(stream); + return stack->buf; +error: + ppstream_done(stream); + return NULL; +} + +/* boxes */ + +pprect * pparray_to_rect (pparray *array, pprect *rect) +{ + ppobj *obj; + if (array->size != 4) + return NULL; + obj = pparray_at(array, 0); + if (!ppobj_get_num(obj, rect->lx)) return NULL; + obj = pparray_at(array, 1); + if (!ppobj_get_num(obj, rect->ly)) return NULL; + obj = pparray_at(array, 2); + if (!ppobj_get_num(obj, rect->rx)) return NULL; + obj = pparray_at(array, 3); + if (!ppobj_get_num(obj, rect->ry)) return NULL; + return rect; +} + +pprect * ppdict_get_rect (ppdict *dict, const char *name, pprect *rect) +{ + pparray *array; + return (array = ppdict_rget_array(dict, name)) != NULL ? pparray_to_rect(array, rect) : NULL; +} + +pprect * ppdict_get_box (ppdict *dict, const char *name, pprect *rect) +{ + do { + if (ppdict_get_rect(dict, name, rect) != NULL) + return rect; + dict = ppdict_rget_dict(dict, "Parent"); + } while (dict != NULL); + return NULL; +} + +ppmatrix * pparray_to_matrix (pparray *array, ppmatrix *matrix) +{ + ppobj *obj; + if (array->size != 6) + return NULL; + obj = pparray_at(array, 0); + if (!ppobj_get_num(obj, matrix->xx)) return NULL; + obj = pparray_at(array, 1); + if (!ppobj_get_num(obj, matrix->xy)) return NULL; + obj = pparray_at(array, 2); + if (!ppobj_get_num(obj, matrix->yx)) return NULL; + obj = pparray_at(array, 3); + if (!ppobj_get_num(obj, matrix->yy)) return NULL; + obj = pparray_at(array, 4); + if (!ppobj_get_num(obj, matrix->x)) return NULL; + obj = pparray_at(array, 5); + if (!ppobj_get_num(obj, matrix->y)) return NULL; + return matrix; +} + +ppmatrix * ppdict_get_matrix (ppdict *dict, const char *name, ppmatrix *matrix) +{ + pparray *array; + return (array = ppdict_rget_array(dict, name)) != NULL ? pparray_to_matrix(array, matrix) : NULL; +} + +/* logger */ + +void pplog_callback (pplogger_callback logger, void *alien) +{ + logger_callback((logger_function)logger, alien); +} + +int pplog_prefix (const char *prefix) +{ + return logger_prefix(prefix); +} + +/* version */ + +const char * ppdoc_version_string (ppdoc *pdf) +{ + return pdf->version; +} + +int ppdoc_version_number (ppdoc *pdf, int *minor) +{ + *minor = pdf->version[2] - '0'; + return pdf->version[0] - '0'; +} + +/* doc info */ + +size_t ppdoc_file_size (ppdoc *pdf) +{ + return pdf->filesize; +} + +ppuint ppdoc_objects (ppdoc *pdf) +{ + ppuint count; + ppxref *xref; + for (count = 0, xref = pdf->xref; xref != NULL; xref = xref->prev) + count += xref->count; + return count; +} + +size_t ppdoc_memory (ppdoc *pdf, size_t *waste) +{ + size_t used; + ppheap *heap; + used = 0, *waste = 0; + for (heap = pdf->heap; heap != NULL; heap = heap->prev) + { + used += heap->space; + *waste += heap->size; + } + return used; +} diff --git a/Build/source/texk/web2c/luatexdir/luapplib/ppload.c.orig b/Build/source/texk/web2c/luatexdir/luapplib/ppload.c.orig new file mode 100644 index 00000000000..fda726656ab --- /dev/null +++ b/Build/source/texk/web2c/luatexdir/luapplib/ppload.c.orig @@ -0,0 +1,2555 @@ + +#include <utilbasexx.h> + +#include "pplib.h" + +const char * ppobj_kind[] = { "none", "null", "bool", "integer", "number", "name", "string", "array", "dict", "stream", "ref" }; + +#define ignored_char(c) (c == 0x20 || c == 0x0A || c == 0x0D || c == 0x09 || c == 0x00) +#define newline_char(c) (c == 0x0A || c == 0x0D) +#define IGNORED_CHAR_CASE 0x20: case 0x0A: case 0x0D: case 0x09: case 0x00 +#define NEWLINE_CHAR_CASE 0x0A: case 0x0D +#define DIGIT_CHAR_CASE '0': case '1': case '2': case '3': case '4': case '5': case '6': case '7': case '8': case '9' +#define OCTAL_CHAR_CASE '0': case '1': case '2': case '3': case '4': case '5': case '6': case '7' + +#define MAX_INT_DIGITS 32 + +#define PP_LENGTH_UNKNOWN ((size_t)-1) + +static const char * ppref_str (ppuint refnumber, ppuint refversion) +{ + static char buffer[MAX_INT_DIGITS + 1 + MAX_INT_DIGITS + 1 + 1 + 1]; +#if defined(MSVC64)|| defined(MINGW64) + sprintf(buffer, PPUINTF " " PPUINTF " R", refnumber, refversion); +#else + sprintf(buffer, PPUINTF " " PPUINTF " R", (unsigned long)(refnumber), (unsigned long)(refversion)); +#endif + return buffer; +} + +/* name */ + +// pdf name delimiters: 0..32, ()<>[]{}/% +// # treated specially +// .+- are valid part of name; keep in mind names such as -| | |- .notdef ABCDEF+Font etc. +const char ppname_byte_lookup[] = { + 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, + 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, + 0, 1, 1, '#', 1, 0, 1, 1, 0, 0, 1, 1, 1, 1, 1, 0, + 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 0, 1, 0, 1, + 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, + 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 0, 1, 0, 1, 1, + 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, + 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 0, 1, 0, 1, 1, + 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, + 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, + 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, + 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, + 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, + 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, + 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, + 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1 +}; + +#define PPNAME_INIT (8+1) + +#define ppname_flush(O, ghost, siz, flgs) \ + (iof_put(O, '\0'), \ + ghost = (_ppname *)ppheap_flush(O, &siz), \ + ghost->flags = flgs, \ + ghost->size = siz - sizeof(_ppname) - 1, \ + (ppname)(ghost + 1)) + +#define ppname_flush_with_ego(O, ghost, siz, flgs) \ + (iof_put(O, '\0'), \ + iof_ensure(O, sizeof(ppname *)), \ + O->pos += sizeof(ppname *), \ + ghost = (_ppname *)ppheap_flush(O, &siz), \ + ghost->flags = flgs, \ + ghost->size = siz - sizeof(_ppname) - 1 - sizeof(ppname *), \ + (ppname)(ghost + 1)) + +#define ppname_set_alter_ego(name, ghost, ego) (*((ppname *)(name + (ghost)->size + 1)) = ego) +#define ppname_get_alter_ego(name) (*((ppname *)(name + ppname_size(name) + 1))) + +static ppname ppscan_name (iof *I, ppheap **pheap) +{ + int c, decode; + iof *O; + _ppname *ghost1, *ghost2; + ppname encoded, decoded; + size_t size; + const char *p, *e; + uint8_t v1, v2; + O = ppheap_buffer(pheap, sizeof(_ppname), PPNAME_INIT); + for (decode = 0, c = iof_char(I); c >= 0 && ppname_byte_lookup[c]; c = iof_next(I)) + { + if (c == '#') + decode = 1; + iof_put(O, c); + } + if (!decode) + return ppname_flush(O, ghost1, size, 0); + encoded = ppname_flush_with_ego(O, ghost1, size, 0|PPNAME_ENCODED); + O = ppheap_buffer(pheap, sizeof(_ppname), PPNAME_INIT); + for (p = encoded, e = encoded + ghost1->size; p < e; ++p) + { + if (*p == '#' && p + 2 < e ){ + v1 = base16_value(p[1]); + v2 = base16_value(p[2]); + iof_put(O, ((v1<<4)+v2)); + }else + iof_put(O, *p); + } + decoded = ppname_flush_with_ego(O, ghost2, size, 0|PPNAME_DECODED); + ppname_set_alter_ego(encoded, ghost1, decoded); + ppname_set_alter_ego(decoded, ghost2, encoded); + return encoded; +} + +static ppname ppscan_exec (iof *I, ppheap **pheap, int first) +{ + int c, decode; + iof *O; + _ppname *ghost1, *ghost2; + ppname encoded, decoded; + size_t size; + const char *p, *e; + uint8_t v1, v2; + + O = ppheap_buffer(pheap, sizeof(_ppname), PPNAME_INIT); + iof_put(O, first); + for (decode = 0, c = iof_char(I); c >= 0 && ppname_byte_lookup[c]; c = iof_next(I)) + { + if (c == '#') + decode = 1; + iof_put(O, c); + } + if (!decode) + return ppname_flush(O, ghost1, size, PPNAME_EXEC); + encoded = ppname_flush_with_ego(O, ghost1, size, PPNAME_EXEC|PPNAME_ENCODED); + O = ppheap_buffer(pheap, sizeof(_ppname), PPNAME_INIT); + for (p = encoded, e = encoded + ghost1->size; p < e; ++p) + { + if (*p == '#' && p + 2 < e ){ + v1 = base16_value(p[1]); + v2 = base16_value(p[2]); + iof_put(O, ((v1<<4)+v2)); + }else + iof_put(O, *p); + } + decoded = ppname_flush_with_ego(O, ghost2, size, PPNAME_EXEC|PPNAME_DECODED); + ppname_set_alter_ego(encoded, ghost1, decoded); + ppname_set_alter_ego(decoded, ghost2, encoded); + return encoded; +} + +static ppname ppexec_internal (const void *data, size_t size, ppheap **pheap) +{ // used only for artificial 'EI' operator name + iof *O; + _ppname *ghost1; + + O = ppheap_buffer(pheap, sizeof(_ppname), size); + iof_write(O, data, size); + return ppname_flush(O, ghost1, size, PPNAME_EXEC); +} + +ppname ppname_decoded (ppname name) +{ + const _ppname *ghost; + ghost = _ppname_ghost(name); + return (ghost->flags & PPNAME_ENCODED) ? ppname_get_alter_ego(name) : name; +} + +ppname ppname_encoded (ppname name) +{ + const _ppname *ghost; + ghost = _ppname_ghost(name); + return (ghost->flags & PPNAME_DECODED) ? ppname_get_alter_ego(name) : name; +} + +/* string */ + +#define PPSTRING_INIT (16+1) + +#define ppstring_flush(O, ghost, siz, flgs) \ + (iof_put(O, '\0'), \ + ghost = (_ppstring *)ppheap_flush(O, &siz), \ + ghost->flags = flgs, \ + ghost->size = siz - sizeof(_ppstring) - 1, \ + (ppstring)(ghost + 1)) + +#define ppstring_flush_with_ego(O, ghost, siz, flgs) \ + (iof_put(O, '\0'), \ + iof_ensure(O, sizeof(ppstring *)), \ + O->pos += sizeof(ppstring *), \ + ghost = (_ppstring *)ppheap_flush(O, &siz), \ + ghost->flags = flgs, \ + ghost->size = siz - sizeof(_ppstring) - 1 - sizeof(ppstring *), \ + (ppstring)(ghost + 1)) + +#define ppstring_utf16be_bom(decoded) (decoded[0] == ((char)0xFE) && decoded[1] == ((char)0xFF) ) +#define ppstring_utf16le_bom(decoded) (decoded[0] == ((char)0xFF) && decoded[1] == ((char)0xFE)) + +#define ppstring_check_bom(decoded, ghost) ((void)\ + (ghost->size >= 2 ? (ppstring_utf16be_bom(decoded) ? (ghost->flags |= PPSTRING_UTF16BE) : \ + (ppstring_utf16le_bom(decoded) ? (ghost->flags |= PPSTRING_UTF16LE) : 0)) : 0)) + +#define ppstring_check_bom2(decoded, ghost1, ghost2) ((void)\ + (ghost2->size >= 2 ? (ppstring_utf16be_bom(decoded) ? ((ghost1->flags |= PPSTRING_UTF16BE), (ghost2->flags |= PPSTRING_UTF16BE)) : \ + (ppstring_utf16le_bom(decoded) ? ((ghost1->flags |= PPSTRING_UTF16LE), (ghost2->flags |= PPSTRING_UTF16LE)) : 0)) : 0)) + +#define ppstring_set_alter_ego(string, ghost, ego) (*((ppstring *)(string + (ghost)->size + 1)) = ego) +#define ppstring_get_alter_ego(string) (*((ppstring *)(string + ppstring_size(string) + 1))) + +static ppstring ppscan_string (iof *I, ppheap **pheap) +{ + int c, decode, balance; + iof *O; + _ppstring *ghost1, *ghost2; + uint8_t *p, *e; + ppstring encoded, decoded; + size_t size; + O = ppheap_buffer(pheap, sizeof(_ppstring), PPSTRING_INIT); + for (decode = 0, balance = 0, c = iof_char(I); c >= 0; ) + { + switch (c) + { + case '\\': + decode = 1; // unescaping later + iof_put(O, '\\'); + if ((c = iof_next(I)) >= 0) + { + iof_put(O, c); + c = iof_next(I); + } + break; + case '(': // may be unescaped if balanced + ++balance; + iof_put(O, '('); + c = iof_next(I); + break; + case ')': + if (balance == 0) + { + c = IOFEOF; + ++I->pos; + break; + } + --balance; + iof_put(O, ')'); + c = iof_next(I); + break; + default: + iof_put(O, c); + c = iof_next(I); + } + } + if (!decode) + { + encoded = ppstring_flush(O, ghost1, size, 0); + ppstring_check_bom(encoded, ghost1); // any bytes can be there + return encoded; + } + encoded = ppstring_flush_with_ego(O, ghost1, size, 0|PPSTRING_ENCODED); + O = ppheap_buffer(pheap, sizeof(_ppstring), PPSTRING_INIT); + for (p = (uint8_t *)encoded, e = (uint8_t *)encoded + ghost1->size; p < e; ++p) + { + if (*p == '\\') + { + if (++p >= e) + break; + switch (*p) + { + case OCTAL_CHAR_CASE: + c = *p - '0'; + if (++p < e && *p >= '0' && *p <= '7') + { + c = (c << 3) + *p - '0'; + if (++p < e && *p >= '0' && *p <= '7') + c = (c << 3) + *p - '0'; + } + iof_put(O, c); + break; + case 'n': + iof_put(O, '\n'); + break; + case 'r': + iof_put(O, '\r'); + break; + case 't': + iof_put(O, '\t'); + break; + case 'b': + iof_put(O, '\b'); + break; + case 'f': + iof_put(O, '\f'); + break; + case NEWLINE_CHAR_CASE: // not a part of the string, ignore (pdf spec page 55) + break; + case '(': case ')': case '\\': + default: // for enything else backslash is ignored (pdf spec page 54) + iof_put(O, *p); + break; + } + } + else + iof_put(O, *p); + } + decoded = ppstring_flush_with_ego(O, ghost2, size, 0|PPSTRING_DECODED); + ppstring_check_bom2(decoded, ghost1, ghost2); + ppstring_set_alter_ego(encoded, ghost1, decoded); + ppstring_set_alter_ego(decoded, ghost2, encoded); + return encoded; +} + +/* Hex string may contain white characters. If odd number of digits, the last assumed to be '0' */ + +static ppstring ppscan_base16 (iof *I, ppheap **pheap) +{ + int c, v1, v2; + iof *O; + _ppstring *ghost1, *ghost2; + size_t size; + ppstring encoded, decoded; + uint8_t *p, *e; + + O = ppheap_buffer(pheap, sizeof(_ppstring), PPSTRING_INIT); + for (c = iof_char(I); c >= 0 && (base16_digit(c) || ignored_char(c)); c = iof_next(I)) + iof_put(O, c); + if (c == '>') + ++I->pos; + encoded = ppstring_flush_with_ego(O, ghost1, size, PPSTRING_BASE16|PPSTRING_ENCODED); + O = ppheap_buffer(pheap, sizeof(_ppstring), (ghost1->size >> 1) + 1); + for (p = (uint8_t *)encoded, e = (uint8_t *)encoded + ghost1->size; p < e; ++p) + { + if ((v1 = base16_value(*p)) < 0) // ignored + continue; + for (v2 = 0, ++p; p < e && (v2 = base16_value(*p)) < 0; ++p); + iof_put(O, (v1<<4)|v2); + } + decoded = ppstring_flush_with_ego(O, ghost2, size, 0|PPSTRING_DECODED); + ppstring_check_bom2(decoded, ghost1, ghost2); + ppstring_set_alter_ego(encoded, ghost1, decoded); + ppstring_set_alter_ego(decoded, ghost2, encoded); + return encoded; +} + +/* internal use only; binary string */ + +static ppstring ppstring_buffer (iof *O, ppheap **pheap) +{ + _ppstring *ghost1, *ghost2; + ppstring encoded, decoded; + uint8_t *p, *e; + size_t size; + + decoded = ppstring_flush_with_ego(O, ghost2, size, 0|PPSTRING_DECODED); + O = ppheap_buffer(pheap, sizeof(_ppstring), (ghost2->size << 1) + 1); + for (p = (uint8_t *)decoded, e = (uint8_t *)decoded + ghost2->size; p < e; ++p) + iof_set2(O, base16_uc_alphabet[(*p)>>4], base16_uc_alphabet[(*p)&15]); + encoded = ppstring_flush_with_ego(O, ghost1, size, PPSTRING_BASE16|PPSTRING_ENCODED); + ppstring_set_alter_ego(encoded, ghost1, decoded); + ppstring_set_alter_ego(decoded, ghost2, encoded); + return encoded; +} + +ppstring ppstring_internal (const void *data, size_t size, ppheap **pheap) +{ // so far used only for crypt key + iof *O; + + O = ppheap_buffer(pheap, sizeof(_ppstring), size); + iof_write(O, data, size); + return ppstring_buffer(O, pheap); +} + +/* PDF spec says nothing about base85 strings, but streams might be (afair no leading '<~' but present trailing '~>') */ + +static ppstring ppscan_base85 (iof *I, ppheap **pheap) +{ // bawse85 alphabet is 33.117, adobe also hires 'z' and 'y' for compression + int c; + iof *O, B; + _ppstring *ghost1, *ghost2; + size_t size; + ppstring encoded, decoded; + O = ppheap_buffer(pheap, sizeof(_ppstring), PPSTRING_INIT); + for (c = iof_char(I); (c >= '!' && c <= 'u') || c == 'z' || c == 'y'; c = iof_next(I)) + iof_put(O, c); + if (c == '~') + if ((c = iof_next(I)) == '>') + ++I->pos; + encoded = ppstring_flush_with_ego(O, ghost1, size, PPSTRING_BASE85|PPSTRING_ENCODED); + iof_string_reader(&B, encoded, ghost1->size); + O = ppheap_buffer(pheap, sizeof(_ppstring), (ghost1->size * 5 / 4) + 1); + base85_decode(&B, O); + decoded = ppstring_flush_with_ego(O, ghost2, size, 0|PPSTRING_DECODED); + ppstring_check_bom2(decoded, ghost1, ghost2); + ppstring_set_alter_ego(encoded, ghost1, decoded); + ppstring_set_alter_ego(decoded, ghost2, encoded); + return encoded; +} + +/* +Encrypted strings. In case of encrypted strings, we first need to decode the string (saving original form hardly makes sense), +then decrypt the string, and encode it again. +*/ + +const char ppstring_byte_escape[] = { /* -1 escaped with octal, >0 escaped with \\, 0 left intact*/ + -1,-1,-1,-1,-1,-1,-1,-1,'b','t','n',-1,'f','r',-1,-1, + -1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1, + 0, 0, 0, 0, 0, 0, 0, 0,'(',')', 0, 0, 0, 0, 0, 0, + 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, + 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, + 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,'\\', 0, 0, 0, + 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, + 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, + -1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1, + -1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1, + -1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1, + -1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1, + -1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1, + -1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1, + -1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1, + -1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1 +}; + + +static ppstring ppscan_crypt_string (iof *I, ppcrypt *crypt, ppheap **pheap) +{ + int c, b, balance, encode; + iof *O; + _ppstring *ghost1, *ghost2; + ppstring encoded, decoded; + uint8_t *p, *e; + size_t size; + + O = ppheap_buffer(pheap, sizeof(_ppstring), PPSTRING_INIT); + for (balance = 0, encode = 0, c = iof_char(I); c >= 0; ) + { + switch (c) + { + case '\\': + if ((c = iof_next(I)) < 0) + break; + encode = 1; + switch (c) + { + case OCTAL_CHAR_CASE: + b = c - '0'; + if ((c = iof_next(I)) >= 0 && c >= '0' && c <= '7') + { + b = (b << 3) + c - '0'; + if ((c = iof_next(I)) >= 0 && c >= '0' && c <= '7') + { + b = (b << 3) + c - '0'; + c = iof_next(I); + } + } + iof_put(O, b); + // c is set to the next char + break; + case 'n': + iof_put(O, '\n'); + c = iof_next(I); + break; + case 'r': + iof_put(O, '\r'); + c = iof_next(I); + break; + case 't': + iof_put(O, '\t'); + c = iof_next(I); + break; + case 'b': + iof_put(O, '\b'); + c = iof_next(I); + break; + case 'f': + iof_put(O, '\f'); + c = iof_next(I); + break; + case NEWLINE_CHAR_CASE: // not a part of the string, ignore (pdf spec page 55) + c = iof_next(I); + break; + case '(': case ')': case '\\': + default: // for enything else backslash is ignored (pdf spec page 54) + iof_put(O, c); + c = iof_next(I); + break; + } + break; + case '(': + ++balance; + encode = 1; + iof_put(O, '('); + c = iof_next(I); + break; + case ')': + if (balance == 0) + { + c = IOFEOF; + ++I->pos; + } + else + { + --balance; + //encode = 1; + iof_put(O, ')'); + c = iof_next(I); + } + break; + default: + if (ppstring_byte_escape[c] != 0) + encode = 1; + iof_put(O, c); + c = iof_next(I); + } + } + /* decrypt the buffer in place, update size */ + if (ppstring_decrypt(crypt, O->buf, iof_size(O), O->buf, &size)) + O->pos = O->buf + size; + /* make encoded counterpart */ + if (!encode) + { + decoded = ppstring_flush(O, ghost2, size, 0); + ppstring_check_bom(decoded, ghost2); + return decoded; + } + decoded = ppstring_flush_with_ego(O, ghost2, size, PPSTRING_DECODED); + O = ppheap_buffer(pheap, sizeof(_ppstring), ghost2->size); + for (p = (uint8_t *)decoded, e = (uint8_t *)decoded + ghost2->size; p < e; ++p) + { + switch ((b = ppstring_byte_escape[*p])) + { + case 0: + iof_put(O, *p); + break; + case -1: + iof_put4(O, '\\', (c >> 6) + '0', ((c >> 3) & 7) + '0', (c & 7) + '0'); + break; + default: + iof_put2(O, '\\', b); + break; + } + } + encoded = ppstring_flush_with_ego(O, ghost1, size, PPSTRING_ENCODED); + ppstring_check_bom2(decoded, ghost1, ghost2); + ppstring_set_alter_ego(encoded, ghost1, decoded); + ppstring_set_alter_ego(decoded, ghost2, encoded); + return encoded; +} + + +static ppstring ppscan_crypt_base16 (iof *I, ppcrypt *crypt, ppheap **pheap) +{ + int c, v1, v2; + iof *O; + _ppstring *ghost1, *ghost2; + ppstring encoded, decoded; + uint8_t *p, *e; + size_t size; + + O = ppheap_buffer(pheap, sizeof(_ppstring), PPSTRING_INIT); + // base16_decode(I, O); // no info about the last char.. + for (c = iof_char(I); c != '>' && c >= 0; ) + { + if ((v1 = base16_value(c)) < 0) + { + if (ignored_char(c)) + { + c = iof_next(I); + continue; + } + break; + } + do { + c = iof_next(I); + if ((v2 = base16_value(c)) >= 0) + { + c = iof_next(I); + break; + } + if (!ignored_char(c)) // c == '>' || c < 0 or some crap + { + v2 = 0; + break; + } + } while (1); + iof_put(O, (v1 << 4)|v2); + } + if (c == '>') + ++I->pos; + /* decrypt the buffer in place, update size */ + if (ppstring_decrypt(crypt, O->buf, iof_size(O), O->buf, &size)) + O->pos = O->buf + size; + decoded = ppstring_flush_with_ego(O, ghost2, size, PPSTRING_DECODED); + /* recreate an encoded form */ + O = ppheap_buffer(pheap, sizeof(_ppstring), (ghost2->size << 1) + 1); + for (p = (uint8_t *)decoded, e = (uint8_t *)decoded + ghost2->size; p < e; ++p) + iof_set2(O, base16_uc_alphabet[(*p)>>4], base16_uc_alphabet[(*p)&15]); + encoded = ppstring_flush_with_ego(O, ghost1, size, PPSTRING_BASE16|PPSTRING_ENCODED); + ppstring_check_bom2(decoded, ghost1, ghost2); + ppstring_set_alter_ego(encoded, ghost1, decoded); + ppstring_set_alter_ego(decoded, ghost2, encoded); + return encoded; +} + +/* ppstring alter ego switcher */ + +ppstring ppstring_decoded (ppstring string) +{ + const _ppstring *ghost; + ghost = _ppstring_ghost(string); + return (ghost->flags & PPSTRING_ENCODED) ? ppstring_get_alter_ego(string) : string; +} + +ppstring ppstring_encoded (ppstring string) +{ + const _ppstring *ghost; + ghost = _ppstring_ghost(string); + return (ghost->flags & PPSTRING_DECODED) ? ppstring_get_alter_ego(string) : string; +} + +/* scanner stack */ + +#define PPSTACK_BUFFER 512 + +static void ppstack_init (ppstack *stack, ppheap **pheap) +{ + stack->buf = stack->pos = (ppobj *)pp_malloc(PPSTACK_BUFFER * sizeof(ppobj)); + stack->size = 0; + stack->space = PPSTACK_BUFFER; + stack->pheap = pheap; +} + +#define ppstack_free_buffer(stack) (pp_free((stack)->buf)) + +static void ppstack_resize (ppstack *stack) +{ + ppobj *newbuffer; + stack->space <<= 1; + newbuffer = (ppobj *)pp_malloc(stack->space * sizeof(ppobj)); + memcpy(newbuffer, stack->buf, stack->size * sizeof(ppobj)); + ppstack_free_buffer(stack); + stack->buf = newbuffer; + stack->pos = newbuffer + stack->size; +} + +#define ppstack_push(stack) ((void)((stack)->size < (stack)->space || (ppstack_resize(stack), 0)), ++(stack)->size, (stack)->pos++) +#define ppstack_pop(stack, n) ((stack)->size -= (n), (stack)->pos -= (n)) +#define ppstack_at(stack, i) ((stack)->buf + i) +#define ppstack_clear(stack) ((stack)->pos = (stack)->buf, (stack)->size = 0) + +/* scanner commons */ + +#define ppscan_uint(I, u) iof_get_uintlw(I, u) +#define ppread_uint(s, u) string_to_uintlw((const char *)(s), u) + +static ppobj * ppscan_numobj (iof *I, ppobj *obj, int negative) +{ + ppint integer; + ppnum number; + int exponent; + int c; + c = iof_char(I); + iof_scan_integer(I, c, integer); + switch(c) + { + case '.': + { + number = (ppnum)integer; + c = iof_next(I); + iof_scan_fraction(I, c, number, exponent); + double_negative_exp10(number, exponent); + obj->type = PPNUM, obj->number = negative ? -number : number; + break; + } + default: + obj->type = PPINT, obj->integer = negative ? -integer : integer; + break; + } + return obj; +} + +static ppobj * ppscan_numobj_frac (iof *I, ppobj *obj, int negative) +{ + ppnum number; + int c, exponent; + + number = 0.0; + c = iof_next(I); + iof_scan_fraction(I, c, number, exponent); + double_negative_exp10(number, exponent); + obj->type = PPNUM, obj->number = negative ? -number : number; + return obj; +} + +static int ppscan_find (iof *I) +{ // skips whitechars and comments + int c; + for (c = iof_char(I); ; c = iof_next(I)) + { + switch (c) + { + case IGNORED_CHAR_CASE: + break; + case '%': { + do { + if ((c = iof_next(I)) < 0) + return c; + } while (!newline_char(c)); + break; + } + default: + return c; + } + } + return c; // never reached +} + +static int ppscan_keyword (iof *I, const char *keyword, size_t size) +{ + size_t i; + int c; + if (iof_left(I) >= size) + { + if (memcmp(I->pos, keyword, size) != 0) + return 0; + I->pos += size; + return 1; + } + // sticky case, we can't go back + for (i = 0, c = iof_char(I); i < size; ++i, ++keyword, c = iof_next(I)) + if (i != c) + return 0; + return 1; +} + +#define ppscan_key(I, literal) ppscan_keyword(I, "" literal, sizeof(literal) - 1) + +/* objects parser */ + +static ppref * ppref_unresolved (ppheap **pheap, ppuint refnumber, ppuint refversion) +{ + ppref *ref = (ppref *)ppheap_take(pheap, sizeof(ppref)); + memset(ref, 0, sizeof(ppref)); + ref->object.type = PPNONE; + ref->number = refnumber; + ref->version = refversion; + return ref; +} + +#define PPMARK PPNONE + +static ppobj * ppscan_obj (iof *I, ppdoc *pdf, ppxref *xref) +{ + int c; + ppobj *obj; + size_t mark, size; + ppuint refnumber, refversion; + ppref *ref; + ppstack *stack; + ppcrypt *crypt; + + stack = &pdf->stack; + c = iof_char(I); + switch (c) + { + case DIGIT_CHAR_CASE: + return ppscan_numobj(I, ppstack_push(stack), 0); + case '.': + return ppscan_numobj_frac(I, ppstack_push(stack), 0); + case '+': + ++I->pos; + return ppscan_numobj(I, ppstack_push(stack), 0); + case '-': + ++I->pos; + return ppscan_numobj(I, ppstack_push(stack), 1); + case '/': + ++I->pos; + obj = ppstack_push(stack); + obj->type = PPNAME; + obj->name = ppscan_name(I, &pdf->heap); + return obj; + case '(': + ++I->pos; + obj = ppstack_push(stack); + obj->type = PPSTRING; + if (ppcrypt_ref(pdf, crypt)) + obj->string = ppscan_crypt_string(I, crypt, &pdf->heap); + else + obj->string = ppscan_string(I, &pdf->heap); + return obj; + case '[': + mark = stack->size; + obj = ppstack_push(stack); + obj->type = PPMARK; // ppscan_obj() checks types backward for 'R', so set the type immediatelly (reserved for PPARRAY) + obj->any = NULL; + ++I->pos; + for (c = ppscan_find(I); c != ']'; c = ppscan_find(I)) + { + if (ppscan_obj(I, pdf, xref) == NULL) + { // callers assume that NULL returns means nothing pushed + size = stack->size - mark; // pop items AND the obj reserved for array + ppstack_pop(stack, size); + return NULL; + } + } + ++I->pos; + size = stack->size - mark - 1; + obj = ppstack_at(stack, mark); // stack might have been realocated + obj->type = PPARRAY; + obj->array = pparray_create(ppstack_at(stack, mark + 1), size, &pdf->heap); + ppstack_pop(stack, size); // pop array items, leave the array on top + return obj; + case '<': + if ((c = iof_next(I)) == '<') + { + mark = stack->size; + obj = ppstack_push(stack); + obj->type = PPMARK; + obj->any = NULL; + ++I->pos; + for (c = ppscan_find(I); c != '>'; c = ppscan_find(I)) + { + if (ppscan_obj(I, pdf, xref) == NULL) + { + size = stack->size - mark; + ppstack_pop(stack, size); + return NULL; + } + } + if (iof_next(I) == '>') + ++I->pos; + size = stack->size - mark - 1; + obj = ppstack_at(stack, mark); + obj->type = PPDICT; + obj->dict = ppdict_create(ppstack_at(stack, mark + 1), size, &pdf->heap); + ppstack_pop(stack, size); + return obj; + } + obj = ppstack_push(stack); + obj->type = PPSTRING; + if (ppcrypt_ref(pdf, crypt)) + obj->string = ppscan_crypt_base16(I, crypt, &pdf->heap); + else + obj->string = ppscan_base16(I, &pdf->heap); + return obj; + case 'R': + if (stack->size >= 2 && stack->pos[-1].type == PPINT && stack->pos[-2].type == PPINT) + { + ++I->pos; + obj = &stack->pos[-2]; + refnumber = (ppuint)obj->integer; + ppstack_pop(stack, 1); // pop version number, retype obj to a reference + if (xref == NULL || (ref = ppxref_find(xref, refnumber)) == NULL) + { /* pdf spec page 64: unresolvable reference is not an error, should just be treated as a reference to null. + we also need this to read trailer, where refs can't be resolved yet */ + refversion = (obj + 1)->integer; + //if (xref != NULL) + // loggerf("unresolved reference %s", ppref_str(refnumber, refversion)); + ref = ppref_unresolved(stack->pheap, refnumber, refversion); + } + obj->type = PPREF; + obj->ref = ref; + return obj; + } + break; + case 't': + if (iof_next(I) == 'r' && iof_next(I) == 'u' && iof_next(I) == 'e') + { + ++I->pos; + obj = ppstack_push(stack); + obj->type = PPBOOL; + obj->integer = 1; + return obj; + } + break; + case 'f': + if (iof_next(I) == 'a' && iof_next(I) == 'l' && iof_next(I) == 's' && iof_next(I) == 'e') + { + ++I->pos; + obj = ppstack_push(stack); + obj->type = PPBOOL; + obj->integer = 0; + return obj; + } + break; + case 'n': + if (iof_next(I) == 'u' && iof_next(I) == 'l' && iof_next(I) == 'l') + { + ++I->pos; + obj = ppstack_push(stack); + obj->type = PPNULL; + obj->any = NULL; + return obj; + } + break; + } + return NULL; +} + +/* +A variant for contents streams (aka postscript); wise of operators, blind to references. +We are still PDF, so we don't care about postscript specific stuff such as radix numbers +and scientific numbers notation. It takes ppstack * as context (no ppdoc *) to be able +to run contents parser beyond the scope of ppdoc heap. +*/ + +static ppstring ppstring_inline (iof *I, ppdict *imagedict, ppheap **pheap); + +static ppobj * ppscan_psobj (iof *I, ppstack *stack) +{ + int c; + ppobj *obj, *op; + size_t size, mark; + ppname exec; + + c = iof_char(I); + switch (c) + { + case DIGIT_CHAR_CASE: + return ppscan_numobj(I, ppstack_push(stack), 0); + case '.': + return ppscan_numobj_frac(I, ppstack_push(stack), 0); + case '+': + c = iof_next(I); + if (base10_digit(c)) // '+.abc' is probably an executable name, but we are not in postscript + return ppscan_numobj(I, ppstack_push(stack), 0); + else if (c == '.') + return ppscan_numobj_frac(I, ppstack_push(stack), 0); + obj = ppstack_push(stack); + obj->type = PPNAME; + obj->name = ppscan_exec(I, stack->pheap, '+'); + return obj; + case '-': + c = iof_next(I); + if (base10_digit(c)) // ditto, we would handle type1 '-|' '|-' operators though + return ppscan_numobj(I, ppstack_push(stack), 1); + else if (c == '.') + return ppscan_numobj_frac(I, ppstack_push(stack), 1); + obj = ppstack_push(stack); + obj->type = PPNAME; + obj->name = ppscan_exec(I, stack->pheap, '-'); + return obj; + case '/': + ++I->pos; + obj = ppstack_push(stack); + obj->type = PPNAME; + obj->name = ppscan_name(I, stack->pheap); + return obj; + case '(': + ++I->pos; + obj = ppstack_push(stack); + obj->type = PPSTRING; + obj->string = ppscan_string(I, stack->pheap); + return obj; + case '[': + mark = stack->size; + obj = ppstack_push(stack); + obj->type = PPMARK; + obj->any = NULL; + ++I->pos; + for (c = ppscan_find(I); c != ']'; c = ppscan_find(I)) + { + if (ppscan_psobj(I, stack) == NULL) + { + size = stack->size - mark; + ppstack_pop(stack, size); + return NULL; + } + } + ++I->pos; + size = stack->size - mark - 1; + obj = ppstack_at(stack, mark); + obj->type = PPARRAY; + obj->array = pparray_create(ppstack_at(stack, mark + 1), size, stack->pheap); + ppstack_pop(stack, size); + return obj; + case '<': + if ((c = iof_next(I)) == '<') + { + mark = stack->size; + obj = ppstack_push(stack); + obj->type = PPMARK; + obj->any = NULL; + ++I->pos; + for (c = ppscan_find(I); c != '>'; c = ppscan_find(I)) + { + if (ppscan_psobj(I, stack) == NULL) + { + size = stack->size - mark; + ppstack_pop(stack, size); + return NULL; + } + } + if (iof_next(I) == '>') + ++I->pos; + size = stack->size - mark - 1; + obj = ppstack_at(stack, mark); + obj->type = PPDICT; + obj->dict = ppdict_create(ppstack_at(stack, mark + 1), size, stack->pheap); + ppstack_pop(stack, size); + return obj; + } + obj = ppstack_push(stack); + obj->type = PPSTRING; + if (c == '~') + ++I->pos, obj->string = ppscan_base85(I, stack->pheap); + else + obj->string = ppscan_base16(I, stack->pheap); + return obj; + default: + if (c < 0 || !ppname_byte_lookup[c]) + break; // forbid empty names; dead loop otherwise + ++I->pos; + /* true false null practically don't occur in streams so it makes sense to assume that we get an operator name here. + If it happen to be a keyword we could give back those several bytes to the heap but.. heap buffer is tricky enough. */ + exec = ppscan_exec(I, stack->pheap, c); + obj = ppstack_push(stack); + switch (exec[0]) + { + case 't': + if (exec[1] == 'r' && exec[2] == 'u' && exec[3] == 'e' && exec[4] == '\0') + { + obj->type = PPBOOL; + obj->integer = 1; + // todo: drop exec + return obj; + } + break; + case 'f': + if (exec[1] == 'a' && exec[2] == 'l' && exec[3] == 's' && exec[4] == 'e' && exec[5] == '\0') + { + obj->type = PPBOOL; + obj->integer = 0; + // todo: drop exec + return obj; + } + break; + case 'n': + if (exec[1] == 'u' && exec[2] == 'l' && exec[3] == 'l' && exec[4] == '\0') + { + obj->type = PPNULL; + obj->any = NULL; + // todo: drop exec + return obj; + } + break; + case 'B': + /* + Inline images break rules of operand/operator syntax, so 'BI/ID' operators need to be treated as special syntactic keywords. + + BI <keyval pairs> ID<whitechar?><imagedata><whitechar?>EI + + We treat the image as a single syntactic token; BI starts collecting a dict, ID is the beginning of the data. Effectively EI + operator obtains two operands - dict and string. It is ok to put three items onto the stack, callers dont't assume there is just one. + */ + if (exec[1] == 'I' && exec[2] == '\0') + { + ppdict *imagedict; + /* key val pairs -> dict */ + mark = stack->size - 1; + obj->type = PPMARK; + obj->any = NULL; + for (c = ppscan_find(I); ; c = ppscan_find(I)) + { + if ((op = ppscan_psobj(I, stack)) == NULL) + { + size = stack->size - mark; + ppstack_pop(stack, size); + return NULL; + } + if (op->type == PPNAME && ppname_exec(op->name)) + { + if (!ppname_is(op->name, "ID")) + { // weird + size = stack->size - mark; + ppstack_pop(stack, size); + return NULL; + } + break; + } + } + size = stack->size - mark - 1; + obj = ppstack_at(stack, mark); + obj->type = PPDICT; + obj->dict = imagedict = ppdict_create(ppstack_at(stack, mark + 1), size, stack->pheap); + ppstack_pop(stack, size); + /* put image data string */ + obj = ppstack_push(stack); + obj->type = PPSTRING; + obj->string = ppstring_inline(I, imagedict, stack->pheap);; + /* put EI operator name */ + obj = ppstack_push(stack); + obj->type = PPNAME; + obj->name = ppexec_internal("EI", 2, stack->pheap); + return obj; + } + } + obj->type = PPNAME; + obj->name = exec; + return obj; + } + return NULL; +} + +/* +We try to get the exact inline image length from its dict params. If cannot predict the length, we have to scan the input until 'EI'. +I've checked on may examples that it gives the same results but one can never be sure, as 'EI' might happen to be a part of the data. +Stripping white char is also very heuristic; \0 is a white char in PDF and very likely to be a data byte.. weak method. +*/ + +static size_t inline_image_length (ppdict *dict) +{ + ppuint w, h, bpc, colors; + ppname cs; + + if (ppdict_get_uint(dict, "W", &w) && ppdict_get_uint(dict, "H", &h) && ppdict_get_uint(dict, "BPC", &bpc) && (cs = ppdict_get_name(dict, "CS")) != NULL) + { + if (ppname_is(cs, "DeviceGray")) + colors = 1; + else if (ppname_is(cs, "DeviceRGB")) + colors = 3; + else if (ppname_is(cs, "DeviceCMYK")) + colors = 4; + else + return PP_LENGTH_UNKNOWN; + return (w * h * bpc * colors + 7) >> 3; + } + return PP_LENGTH_UNKNOWN; +} + +static ppstring ppstring_inline (iof *I, ppdict *imagedict, ppheap **pheap) +{ + iof *O; + int c, d, e; + size_t length, leftin, leftout, bytes; + + O = ppheap_buffer(pheap, sizeof(_ppstring), PPSTRING_INIT); + c = iof_char(I); + if (ignored_char(c)) + c = iof_next(I); + + length = inline_image_length(imagedict); + if (length != PP_LENGTH_UNKNOWN) + { + while (length > 0 && iof_readable(I) && iof_writable(O)) + { + leftin = iof_left(I); + leftout = iof_left(O); + bytes = length; + if (bytes > leftin) bytes = leftin; + if (bytes > leftout) bytes = leftout; + memcpy(O->pos, I->pos, bytes); + I->pos += bytes; + O->pos += bytes; + length -= bytes; + } + // gobble EI + if (ppscan_find(I) == 'E') + if (iof_next(I) == 'I') + ++I->pos; + } + else + { + while (c >= 0) + { + if (c == 'E') + { + d = iof_next(I); + if (d == 'I') + { + e = iof_next(I); + if (!ppname_byte_lookup[e]) + { /* strip one newline from the end and stop */ + if (O->pos - 2 >= O->buf) // sanity + { + c = *(O->pos - 1); + if (ignored_char(c)) + { + if (c == 0x0A && *(O->pos - 2) == 0x0D) + O->pos -= 2; + else + O->pos -= 1; + } + } + break; + } + iof_put2(O, c, d); + c = e; + } + else + { + iof_put(O, c); + c = d; + } + } + else + { + iof_put(O, c); + c = iof_next(I); + } + } + } + return ppstring_buffer(O, pheap); +} + +/* input reader */ + +/* +PDF input is a pseudo file that either keeps FILE * or data. Reader iof * is a proxy to input +that provides byte-by-byte interface. Our iof structure is capable to link iof_file *input, +but t avoid redundant checks on IOF_DATA flag, here we link iof *I directly to FILE * or mem buffer. +When reading from file we need an internal buffer, which should be kept rather small, as it is +only used to parse xrefs and objects (no streams). We allocate the buffer from a private heap +(not static) to avoid conflicts when processing >1 pdfs at once. Besides, the input buffer may be +needed after loading the document, eg. to access references raw data. +*/ + +#define PPDOC_BUFFER 0xFFF // keep that small, it is only used to parse body objects + +static void ppdoc_reader_init (ppdoc *pdf, iof_file *input) +{ + iof *I; + pdf->input = *input; + input = &pdf->input; + input->refcount = 1; + I = &pdf->reader; + if (input->flags & IOF_DATA) + { + pdf->buffer = NULL; // input iof_file is the buffer + iof_string_reader(I, NULL, 0); // gets IOF_DATA flag + } + else + { + pdf->buffer = (uint8_t *)ppheap_take(&pdf->heap, PPDOC_BUFFER); + iof_setup_file_handle_reader(I, NULL, 0, iof_file_get_fh(input)); // gets IOF_FILE_HANDLE flag and FILE * + I->space = PPDOC_BUFFER; // used on refill + } +} + +/* +Whenever we need to read the input file, we fseek the to the given offset and fread to to the private buffer. +The length we need is not always predictable, in which case PPDOC_BUFFER bytes are read (keep it small). +I->buf = I->pos is set to the beginning, I->end set to the end (end is the first byte one shouldn't read). +*/ + +static iof * ppdoc_reader (ppdoc *pdf, size_t offset, size_t length) +{ + iof_file *input; + iof *I; + input = &pdf->input; + I = &pdf->reader; + if (iof_file_seek(input, offset, SEEK_SET) != 0) + return NULL; + I->flags &= ~IOF_STOPPED; + if (input->flags & IOF_DATA) + { + I->buf = I->pos = input->pos; + I->end = (length == PP_LENGTH_UNKNOWN || I->pos + length >= input->end) ? input->end : (I->pos + length); + } + else + { + I->buf = I->pos = pdf->buffer; // ->buf is actually permanently equal pdf->buffer but we might need some tricks + if (length == PP_LENGTH_UNKNOWN || length > PPDOC_BUFFER) + length = PPDOC_BUFFER; + length = fread(I->buf, 1, length, I->file); + I->end = I->buf + length; + } + return I; +} + +/* The position from the beginning of input +- for data buffer: (pdf->input.pos - pdf->input.buf) + (I->pos - I->buf) + I->buf == pdf->input.pos, so this resolves to (I->pos - pdf->input.buf), independent from I->buf +- for file buffer: ftell(pdf->input.file) - (I->end - I->pos) +*/ + +#define ppdoc_reader_tell(pdf, I) ((size_t)(((pdf)->input.flags & IOF_DATA) ? ((I)->pos - (pdf)->input.buf) : (ftell(iof_file_get_fh(&(pdf)->input)) - ((I)->end - (I)->pos)))) + +/* pdf */ + +#define PPDOC_HEADER 10 // "%PDF-?.??\n" + +static int ppdoc_header (ppdoc *pdf, uint8_t header[PPDOC_HEADER]) +{ + size_t i; + if (memcmp(header, "%PDF-", 5) != 0) + return 0; + for (i = 5; i < PPDOC_HEADER - 1 && !ignored_char(header[i]); ++i) + pdf->version[i - 5] = header[i]; + pdf->version[i - 5] = '\0'; + return 1; +} + +static int ppdoc_tail (ppdoc *pdf, iof_file *input, size_t *pxrefoffset) +{ + int c; + uint8_t tail[4*10], *p, back, tailbytes; + + if (iof_file_seek(input, 0, SEEK_END) != 0) + return 0; + pdf->filesize = (size_t)iof_file_tell(input); + // simple heuristic to avoif fgetc() / fseek(-2) hiccup: keep seeking back by len(startxref) + 1 == 10 + // until a letter found (assuming liberal white characters and tail length) + for (back = 1, tailbytes = 0; ; ++back) + { + if (iof_file_seek(input, -10, SEEK_CUR) != 0) + return 0; + tailbytes += 10; + c = iof_file_getc(input); + tailbytes -= 1; + switch (c) + { + case IGNORED_CHAR_CASE: + case DIGIT_CHAR_CASE: + case '%': case 'E': case 'O': case 'F': + if (back > 4) // 2 should be enough + return 0; + continue; + case 's': case 't': case 'a': case 'r': case 'x': case 'e': case 'f': + if (iof_file_read(tail, 1, tailbytes, input) != tailbytes) + return 0; + tail[tailbytes] = '\0'; + for (p = &tail[0]; ; ++p) + { + if (*p == '\0') + return 0; + if ((c = base10_value(*p)) >= 0) + break; + } + ppread_uint(p, pxrefoffset); + return 1; + default: + return 0; + } + } + return 0; +} + +/* xref/body */ + +static int ppscan_start_entry (iof *I, ppref *ref) +{ + ppuint u; + ppscan_find(I); if (!ppscan_uint(I, &u) || u != ref->number) return 0; + ppscan_find(I); if (!ppscan_uint(I, &u) || u != ref->version) return 0; + ppscan_find(I); if (!ppscan_key(I, "obj")) return 0; + ppscan_find(I); + return 1; +} + +static int ppscan_skip_entry (iof *I) +{ + size_t u; + ppscan_find(I); if (!ppscan_uint(I, &u)) return 0; + ppscan_find(I); if (!ppscan_uint(I, &u)) return 0; + ppscan_find(I); if (!ppscan_key(I, "obj")) return 0; + ppscan_find(I); + return 1; +} + +static int ppscan_start_stream (iof *I, ppdoc *pdf, size_t *streamoffset) +{ + int c; + ppscan_find(I); + if (ppscan_key(I, "stream")) + { // skip 1 or 2 whites (here we shouldn't just gobble all blanks) + c = iof_char(I); + if (ignored_char(c)) + { + c = iof_next(I); + if (ignored_char(c)) + ++I->pos; + } + *streamoffset = ppdoc_reader_tell(pdf, I); + return 1; + } + return 0; +} + +static ppxref * ppxref_load (ppdoc *pdf, size_t xrefoffset); +static ppxref * ppxref_load_chain (ppdoc *pdf, ppxref *xref); + +/* Parsing xref table + + 1 10 // first ref number and refs count + 0000000000 00000 n // 10-digits offset, 5 digits version, type identifier + 0000000000 00000 n // n states for normal I guess + 0000000000 00000 f // f states for free (not used) + ... + +Free entries seem to be a relic of ancient times, completelly useless for us. To avoid parsing xref table twice, +we waste some space on free entries by allocating one plane of refs for each section. Later on we slice sections, +so that effectively free entries are not involved in map. + +Subsequent refs gets number, version and offset. Other fields initialized when parsing PDF body. + +Having xref table loaded, we sort sections for future binary search (xref with objects count == 0 is considered invalid). + +Then we have to deal with the trailer dict. In general, to load objects and resolve references we need a complete chain +of xrefs (not only the top). To load the previous xref, we need its offset, which is given in trailer. So we have to +parse the trailer ignoring references, which might be unresolvable at this point (objects parser makes a dummy check +for xref != NULL on refs resolving ppscan_obj(), which irritates me but I don't want a separate parser for trailer..). +The same applies to xref streams, in which we have parse the trailer not having xref map at all. So the procedure is: + + - load xref map, initialize references, make it ready to search + - parse trailer ignoring references + - get /Prev xref offset and load older xref (linked list via ->prev) + - sort all refs in all xrefs by offset + - parse refs in order resolving references in contained objects + - fix trailer references + +First created xref becomes a pdf->xref (top xref). We link that early to control offsets already read (insane loops?). +*/ + +// Every xref table item "0000000000 00000 n" is said to be terminated with 2-byte EOL but we don't like relying on whites. +#define xref_item_length (10+1+5+1+1) + +static ppxref * ppxref_load_table (iof *I, ppdoc *pdf, size_t xrefoffset) +{ + ppxref *xref; + ppxsec *xrefsection; + ppref *ref; + ppuint first, count, refindex; + uint8_t buffer[xref_item_length + 1]; + const char *p; + const ppobj *obj; + + buffer[xref_item_length] = '\0'; + xref = ppxref_create(pdf, 0, xrefoffset); + if (pdf->xref == NULL) pdf->xref = xref; + for (ppscan_find(I); ppscan_uint(I, &first); ppscan_find(I)) + { + ppscan_find(I); + if (!ppscan_uint(I, &count)) + return NULL; + if (count == 0) // weird + continue; + xref->count += count; + xrefsection = NULL; + ref = (ppref *)ppheap_take(&pdf->heap, count * sizeof(ppref)); + for (refindex = 0; refindex < count; ++refindex, ++ref) + { + ref->xref = xref; + ref->number = first + refindex; + ppscan_find(I); + iof_read(I, buffer, xref_item_length); + switch (buffer[xref_item_length - 1]) + { + case 'n': + if (xrefsection == NULL) + { + xrefsection = ppxref_push_section(xref, &pdf->heap); + xrefsection->first = ref->number; + xrefsection->refs = ref; + } + xrefsection->last = ref->number; + for (p = (const char *)buffer; *p == '0'; ++p); + p = ppread_uint(p, &ref->offset); + for ( ; *p == ' ' || *p == '0'; ++p); + p = ppread_uint(p, &ref->version); + ref->object.type = PPNONE; // init for sanity + ref->object.any = NULL; + ref->length = 0; + break; + case 'f': + default: + --ref; + xrefsection = NULL; + --xref->count; + } + } + } + /* sort section */ + if (!ppxref_sort(xref)) + ; // case of xref->size == 0 handled by ppxref_load_chain() + /* get trailer ignoring refs */ + if (!ppscan_key(I, "trailer")) + return NULL; + ppscan_find(I); + if ((obj = ppscan_obj(I, pdf, NULL)) == NULL) + return NULL; + ppstack_pop(&pdf->stack, 1); + if (obj->type != PPDICT) + return NULL; + xref->trailer = *obj; + return ppxref_load_chain(pdf, xref); +} + +/* Parsing xref stream +First we load the trailer, ignoring references. Dict defines sections and fields lengths: + + /Size % max ref number plus 1 + /Index [ first count first count ... ] % a pair of numbers for every section, defaults to [0 Size] + /W [w1 w2 w3] % fields lengths, 0 states for omitted field + +xref stream data is a continuous stream of binary number triplets. First number is a type: + + 0 - free entry (as 'f' in xref table) + 1 - normal entry, followed by offset an version (as 'n' in xref table) + 2 - compressed entry, followed by parent object stream number and entry index + +0 and 1 are handled as 'n' and 'f' entries in xref table. For type 2 we normally initialize +ref->number and ref->version (the later is implicitly 0). ref->offset is set to 0 (invalid offset), +which is recognized by objects loader. +*/ + +#define XREF_STREAM_MAX_FIELD 4 + +static ppxref * ppxref_load_stream (iof *I, ppdoc *pdf, size_t xrefoffset) +{ + ppxref *xref; + ppxsec *xrefsection; + ppref *ref; + ppobj *obj; + ppstream *xrefstream; + size_t streamoffset; + ppuint w1, w2, w3, w, bufferbytes; + uint8_t buffer[3 * XREF_STREAM_MAX_FIELD], *b; + ppuint first, count, f1, f2, f3; + pparray *fieldwidths, *sectionindices; + ppobj sectionmock[2], *sectionfirst, *sectioncount; + size_t sections, sectionindex, refindex; + + if (!ppscan_skip_entry(I)) + return NULL; + if ((obj = ppscan_obj(I, pdf, NULL)) == NULL) + return NULL; + ppstack_pop(&pdf->stack, 1); + if (obj->type != PPDICT || !ppscan_start_stream(I, pdf, &streamoffset)) + return NULL; + xrefstream = ppstream_create(pdf, obj->dict, streamoffset); + /* All normal streams go through ppstream_info(), but it makes no sense for trailer stream (no crypt allowed, no refs yet). + So we just record the length and informative flag. Here we have to expect that /Length and /Filter are not indirects. */ + if (!ppdict_get_uint(obj->dict, "Length", &xrefstream->length)) + return NULL; + if (ppdict_get_obj(obj->dict, "Filter") != NULL) + xrefstream->flags |= PPSTREAM_COMPRESSED; + if ((fieldwidths = ppdict_get_array(xrefstream->dict, "W")) != NULL) + { + if (!pparray_get_uint(fieldwidths, 0, &w1)) w1 = 0; + if (!pparray_get_uint(fieldwidths, 1, &w2)) w2 = 0; + if (!pparray_get_uint(fieldwidths, 2, &w3)) w3 = 0; + } + else + w1 = w2 = w3 = 0; + if (w1 > XREF_STREAM_MAX_FIELD || w2 > XREF_STREAM_MAX_FIELD || w3 > XREF_STREAM_MAX_FIELD) + return NULL; + bufferbytes = w1 + w2 + w3; + if ((sectionindices = ppdict_get_array(xrefstream->dict, "Index")) != NULL) + { + sections = sectionindices->size >> 1; + sectionfirst = sectionindices->data; + } + else + { + sections = 1; + sectionmock[0].type = PPINT; + sectionmock[0].integer = 0; + sectionmock[1].type = PPINT; + if (!ppdict_get_int(xrefstream->dict, "Size", §ionmock[1].integer)) + sectionmock[1].integer = 0; + sectionfirst = §ionmock[0]; + } + if ((I = ppstream_read(xrefstream, 1, 0)) == NULL) + return NULL; // we fseek() so original I is useless anyway + xref = ppxref_create(pdf, sections, xrefoffset); + if (pdf->xref == NULL) pdf->xref = xref; + xref->trailer.type = PPSTREAM; + xref->trailer.stream = xrefstream; + for (sectionindex = 0; sectionindex < sections; ++sectionindex, sectionfirst += 2) + { + sectioncount = sectionfirst + 1; + if (!ppobj_get_uint(sectionfirst, first) || !ppobj_get_uint(sectioncount, count)) + goto xref_stream_error; + if (count == 0) + continue; + xref->count += count; + xrefsection = NULL; + ref = (ppref *)ppheap_take(&pdf->heap, count * sizeof(ppref)); + for (refindex = 0; refindex < count; ++refindex, ++ref) + { + ref->xref = xref; + ref->number = first + refindex; + if (iof_read(I, buffer, bufferbytes) != bufferbytes) + goto xref_stream_error; + b = buffer; + if (w1 == 0) + f1 = 1; // default type is 1 + else + for (f1 = 0, w = 0; w < w1; f1 = (f1 << 8)|(*b), ++w, ++b); + for (f2 = 0, w = 0; w < w2; f2 = (f2 << 8)|(*b), ++w, ++b); + for (f3 = 0, w = 0; w < w3; f3 = (f3 << 8)|(*b), ++w, ++b); + switch (f1) + { + case 0: + //--ref; + xrefsection = NULL; + --xref->count; + break; + case 1: + if (xrefsection == NULL) + { + xrefsection = ppxref_push_section(xref, &pdf->heap); + xrefsection->first = ref->number; + xrefsection->refs = ref; + } + xrefsection->last = ref->number; + ref->offset = f2; + ref->version = f3; + ref->object.type = PPNONE; + ref->object.any = NULL; + ref->length = 0; + break; + case 2: + if (xrefsection == NULL) + { + xrefsection = ppxref_push_section(xref, &pdf->heap); + xrefsection->first = ref->number; + xrefsection->refs = ref; + } + xrefsection->last = ref->number; + ref->offset = 0; // f2 is parent objstm, f3 is index in parent, both useless + ref->version = 0; // compressed objects has implicit version == 0 + ref->object.type = PPNONE; + ref->object.any = NULL; + ref->length = 0; + break; + default: + goto xref_stream_error; + } + } + } + /* sort sections */ + if (!ppxref_sort(xref)) + ; // case of xref->size == 0 handled by ppxref_load_chain() + /* close the stream _before_ loading prev xref */ + ppstream_done(xrefstream); + /* load prev and return */ + return ppxref_load_chain(pdf, xref); +xref_stream_error: + ppstream_done(xrefstream); + return NULL; +} + +/* +The following procedure loads xref /Prev, links xref->prev and typically returns xref. +Some docs contain empty xref (one section with zero objects) that is actually a proxy +to xref stream referred as /XRefStm (genuine concept of xrefs old/new style xrefs in +the same doc). In case of 0-length xref we ignore the proxy and return the target xref +(otherwise we would need annoying sanity check for xref->size > 0 on every ref search). +*/ + +static ppxref * ppxref_load_chain (ppdoc *pdf, ppxref *xref) +{ + ppdict *trailer; + ppuint xrefoffset; + ppxref *prevxref, *nextxref; + + trailer = ppxref_trailer(xref); + if (!ppdict_get_uint(trailer, "Prev", &xrefoffset)) // XRefStm is useless + return xref; // missing /Prev is obviously ok + for (nextxref = pdf->xref; nextxref != NULL; nextxref = nextxref->prev) + if (nextxref->offset == xrefoffset) // insane + return NULL; + if ((prevxref = ppxref_load(pdf, (size_t)xrefoffset)) == NULL) + return NULL; + if (xref->size > 0) + { + xref->prev = prevxref; + return xref; + } + if (pdf->xref == xref) + pdf->xref = prevxref; + return prevxref; +} + +static ppxref * ppxref_load (ppdoc *pdf, size_t xrefoffset) +{ + iof *I; + if ((I = ppdoc_reader(pdf, xrefoffset, PP_LENGTH_UNKNOWN)) == NULL) + return NULL; + ppscan_find(I); + if (ppscan_key(I, "xref")) + return ppxref_load_table(I, pdf, xrefoffset); + return ppxref_load_stream(I, pdf, xrefoffset); + // iof_close(I) does nothing here +} + +static void ppoffmap_sort (ppref **left, ppref **right) +{ + ppref **l, **r, *t; + ppuint pivot; + l = left, r = right; + pivot = (*(l + ((r - l) / 2)))->offset; + do + { // don't read from pointer! + while ((*l)->offset < pivot) ++l; + while ((*r)->offset > pivot) --r; + if (l <= r) + { + t = *l; + *l = *r; + *r = t; + ++l, --r; + } + } while (l <= r); + if (left < r) + ppoffmap_sort(left, r); + if (l < right) + ppoffmap_sort(l, right); +} + + +static void fix_trailer_references (ppdoc *pdf) +{ + ppxref *xref; + ppdict *trailer; + ppname *pkey; + ppobj *obj; + ppref *ref; + for (xref = pdf->xref; xref != NULL; xref = xref->prev) + { + if ((trailer = ppxref_trailer(xref)) == NULL) + continue; + for (ppdict_first(trailer, pkey, obj); *pkey != NULL; ppdict_next(pkey, obj)) + { // no need to go deeper in structs, all items in trailer except info and root must be direct refs + if (obj->type != PPREF) + continue; + ref = obj->ref; + if (ref->offset == 0) // unresolved? + if ((ref = ppxref_find(xref, ref->number)) != NULL) + obj->ref = ref; // at this moment the reference still points nothing, but should be the one with the proper offset + } + } +} + +/* +Here comes a procedure that loads all entries from all document bodies. We resolve references while +parsing objects and to make resolving correct, we need a complete chain of xref maps, and a knowledge +about possible linearized dict (first offset). So loading refs sorted by offsets makes sense (not sure +if it matters nowadays but we also avoid fseek() by large offsets). + +Here is the proc: + + - create a list of all refs in all bodies + - sort the list by offsets + - for every ref from the sorted list: + - estimate object length to avoid fread-ing more than necessary (not perfect but enough) + - fseek() to the proper offset, fread() entry data or its part + - parse the object with ppscan_obj(I, pdf, xref), where xref is not necessarily top pdf->xref + - save the actual ref->length (not sure if we need that?) + - make a stream if a dict is followed by "stream" keyword, also save the stream offset + - free the list +*/ + +static int ppdoc_load_objstm (ppstream *stream, ppdoc *pdf, ppxref *xref); + +static void ppdoc_load_entries (ppdoc *pdf) +{ + size_t objects, sectionindex, refnumber, offindex; + ppnum linearized; + ppref **offmap, **pref, *ref; + ppxref *xref; + ppxsec *xsec; + ppobj *obj; + ppname type; + int redundant_indirection = 0; + ppcrypt *crypt; + ppstream *stream; + + if ((objects = (size_t)ppdoc_objects(pdf)) == 0) // can't happen + return; + pref = offmap = (ppref **)pp_malloc(objects * sizeof(ppref *)); + objects = 0; // recount refs with offset > 0 + for (xref = pdf->xref; xref != NULL; xref = xref->prev) + for (sectionindex = 0, xsec = xref->sects; sectionindex < xref->size; ++sectionindex, ++xsec) + for (refnumber = xsec->first, ref = xsec->refs; refnumber <= xsec->last; ++refnumber, ++ref) + if (ref->offset > 0) // 0 means compressed or insane + *pref++ = ref, ++objects; + ppoffmap_sort(offmap, offmap + objects - 1); + + crypt = pdf->crypt; + for (offindex = 0, pref = offmap; offindex < objects; ) + { + ref = *pref; + ++pref; + ++offindex; + if (ref->object.type != PPNONE) // might be preloaded already (/Encrypt dict, stream filter dicts, stream /Length..) + continue; + if (offindex < objects) + ref->length = (*pref)->offset - ref->offset; + else + ref->length = pdf->filesize > ref->offset ? pdf->filesize - ref->offset : 0; + if (crypt != NULL) + { + ppcrypt_start_ref(crypt, ref); + obj = ppdoc_load_entry(pdf, ref); + ppcrypt_end_ref(crypt); + } + else + { + obj = ppdoc_load_entry(pdf, ref); + } + switch (obj->type) + { + case PPDICT: /* Check if the object at first offset is linearized dict. We need that to resolve all references properly. */ + if (offindex == 1 && ppdict_get_num(obj->dict, "Linearized", &linearized)) // /Linearized value is a version number, default 1.0 + pdf->flags |= PPDOC_LINEARIZED; + break; + case PPREF: + redundant_indirection = 1; + break; + default: + break; + } + // if pdf->crypt crypt->ref = NULL + } + + /* refs pointngs refs? cut. */ + if (redundant_indirection) + { + for (offindex = 0, pref = offmap; offindex < objects; ++offindex) + { + ref = *pref++; + if (ref->object.type == PPREF) + ref->object = ref->object.ref->object; // doing for all effectively cuts all insane chains + } + } + + /* now handle streams; update stream info (eg. /Length), load pdf 1.5 object streams + we could do earlier but then we would need to struggle with indirects */ + for (offindex = 0, pref = offmap; offindex < objects; ++offindex) + { + ref = *pref++; + obj = &ref->object; + if (obj->type != PPSTREAM) + continue; + stream = obj->stream; + if (crypt != NULL) + { + ppcrypt_start_ref(crypt, ref); + ppstream_info(stream, pdf); + ppcrypt_end_ref(crypt); + } + else + { + ppstream_info(stream, pdf); + } + if (ref->xref->trailer.type == PPSTREAM && (type = ppdict_get_name(stream->dict, "Type")) != NULL && ppname_is(type, "ObjStm")) // somewhat dummy.. + if (!ppdoc_load_objstm(stream, pdf, ref->xref)) + loggerf("invalid objects stream %s at offset " PPSIZEF, ppref_str(ref->number, ref->version), ref->offset); + } + pp_free(offmap); +} + +ppobj * ppdoc_load_entry (ppdoc *pdf, ppref *ref) +{ + iof *I; + size_t length; + ppxref *xref; + ppobj *obj; + ppstack *stack; + size_t streamoffset; + ppref *refref; + ppuint refnumber, refversion; + + length = ref->length > 0 ? ref->length : PP_LENGTH_UNKNOWN; // estimated or unknown + if ((I = ppdoc_reader(pdf, ref->offset, length)) == NULL || !ppscan_start_entry(I, ref)) + { + loggerf("invalid %s offset " PPSIZEF, ppref_str(ref->number, ref->version), ref->offset); + return &ref->object; // PPNONE + } + stack = &pdf->stack; + xref = ref->xref; // to resolve indirects properly + if ((obj = ppscan_obj(I, pdf, xref)) == NULL) + { + loggerf("invalid %s object at offset " PPSIZEF, ppref_str(ref->number, ref->version), ref->offset); + return &ref->object; // PPNONE + } + ref->object = *obj; + ppstack_pop(stack, 1); + obj = &ref->object; + ref->length = ppdoc_reader_tell(pdf, I) - ref->offset; + if (obj->type == PPDICT) + { + if (ppscan_start_stream(I, pdf, &streamoffset)) + { + obj->type = PPSTREAM; + obj->stream = ppstream_create(pdf, obj->dict, streamoffset); + } + } + else if (obj->type == PPINT) + { + ppscan_find(I); + if (ppscan_uint(I, &refversion) && ppscan_find(I) == 'R') + { + refnumber = (ppuint)obj->integer; + if ((refref = ppxref_find(xref, refnumber)) != NULL) + { + obj->type = PPREF; + obj->ref = refref; + } + else + { + obj->type = PPNONE; // as ppref_unresolved() + obj->any = NULL; + } + } + } + return obj; +} + +/* Loading entries from object stream + + /N is the number of contained entries + /First is the offset of the first item + +The stream consists of N pairs of numbers <objnum> <offset> <objnum> <offset> ... +Offsets are ascending (relative to the first), but ref numbers order is arbitrary. +PDF spec says there might be some additional data between objects, so we should obey offsets. +Which means we should basically load the stream at once (may be needed anyway to grab the stream [...]). +*/ + +static int ppdoc_load_objstm (ppstream *stream, ppdoc *pdf, ppxref *xref) +{ + ppdict *dict; // stream dict, actually still on stack + ppref *ref; + ppobj *obj; + ppuint items, firstoffset, offset, objnum, i, invalid = 0; + iof *I; + uint8_t *firstdata, *indexdata; + ppstack *stack; + + dict = stream->dict; + if (!ppdict_rget_uint(dict, "N", &items) || !ppdict_rget_uint(dict, "First", &firstoffset)) + return 0; + if ((I = ppstream_read(stream, 1, 1)) == NULL) + return 0; + firstdata = I->pos + firstoffset; + if (firstdata >= I->end) + goto invalid_objstm; + stack = &pdf->stack; + //if (pdf->crypt != NULL) + // ppcrypt_end_ref(pdf->crypt); // objects are not encrypted, pdf->crypt->ref ensured NULL + for (i = 0; i < items; ++i) + { + ppscan_find(I); + if (!ppscan_uint(I, &objnum)) + goto invalid_objstm; + ppscan_find(I); + if (!ppscan_uint(I, &offset)) + goto invalid_objstm; + if ((ref = ppxref_find_local(xref, objnum)) == NULL || ref->object.type != PPNONE) + { + loggerf("invalid compressed object number " PPUINTF " at position " PPUINTF, objnum, i); + ++invalid; + continue; + } + if (firstdata + offset >= I->end) + { + loggerf("invalid compressed object offset " PPUINTF " at position " PPUINTF, offset, i); + ++invalid; + continue; + } + indexdata = I->pos; // save position + I->pos = firstdata + offset; // go to the object + ppscan_find(I); + if ((obj = ppscan_obj(I, pdf, xref)) != NULL) + { + ref->object = *obj; + ppstack_pop(stack, 1); + // nothing more needed, as obj can never be indirect ref or stream + } + else + { + ++invalid; + loggerf("invalid compressed object %s at stream offset " PPUINTF, ppref_str(objnum, 0), offset); + } + I->pos = indexdata; // restore position and read next from index + } + ppstream_done(stream); + return invalid == 0; +invalid_objstm: + ppstream_done(stream); + return 0; +} + +/* main PDF loader proc */ + +ppcrypt_status ppdoc_crypt_pass (ppdoc *pdf, const void *userpass, size_t userpasslength, const void *ownerpass, size_t ownerpasslength) +{ + switch (pdf->cryptstatus) + { + case PPCRYPT_NONE: + case PPCRYPT_DONE: + case PPCRYPT_FAIL: + break; + case PPCRYPT_PASS: // initial status or really needs password + pdf->cryptstatus = ppdoc_crypt_init(pdf, userpass, userpasslength, ownerpass, ownerpasslength); + switch (pdf->cryptstatus) + { + case PPCRYPT_NONE: + case PPCRYPT_DONE: + ppdoc_load_entries(pdf); + break; + case PPCRYPT_PASS: // user needs to check ppdoc_crypt_status() and recall ppdoc_crypt_pass() with the proper password + case PPCRYPT_FAIL: // hopeless.. + break; + } + break; + } + return pdf->cryptstatus; +} + +static ppdoc * ppdoc_read (ppdoc *pdf, iof_file *input) +{ + uint8_t header[PPDOC_HEADER]; + size_t xrefoffset; + + input = &pdf->input; + if (iof_file_read(header, 1, PPDOC_HEADER, input) != PPDOC_HEADER || !ppdoc_header(pdf, header)) + return NULL; + if (!ppdoc_tail(pdf, input, &xrefoffset)) + return NULL; + if (ppxref_load(pdf, xrefoffset) == NULL) + return NULL; + fix_trailer_references(pdf); // after loading xrefs but before accessing trailer refs (/Encrypt might be a reference) + // check encryption, if any, try empty password + switch (ppdoc_crypt_pass(pdf, "", 0, NULL, 0)) + { + case PPCRYPT_NONE: // no encryption + case PPCRYPT_DONE: // encryption with an empty password + case PPCRYPT_PASS: // the user needs to check ppdoc_crypt_status() and call ppdoc_crypt_pass() + break; + case PPCRYPT_FAIL: // hopeless + //loggerf("decryption failed"); + //return NULL; + break; + } + return pdf; +} + +static void ppdoc_pages_init (ppdoc *pdf); + +static ppdoc * ppdoc_create (iof_file *input) +{ + ppdoc *pdf; + ppheap *heap; + + heap = ppheap_new(); + pdf = (ppdoc *)ppheap_take(&heap, sizeof(ppdoc)); + pdf->flags = 0; + pdf->heap = heap; + pdf->xref = NULL; + pdf->version[0] = '\0'; + pdf->crypt = NULL; + pdf->cryptstatus = PPCRYPT_PASS; // force encryption check on ppdoc_read() -> ppdoc_crypt_pass() + ppstack_init(&pdf->stack, &pdf->heap); + ppdoc_reader_init(pdf, input); + ppdoc_pages_init(pdf); + if (ppdoc_read(pdf, &pdf->input) != NULL) + return pdf; + ppdoc_free(pdf); + return NULL; +} + +ppdoc * ppdoc_load (const char *filename) +{ + FILE *file; + iof_file input; + if ((file = fopen(filename, "rb")) == NULL) + return NULL; + iof_file_init(&input, file); + input.flags |= IOF_CLOSE_FILE; + return ppdoc_create(&input); +} + +ppdoc * ppdoc_mem (const void *data, size_t size) +{ + iof_file input; + iof_file_rdata_init(&input, data, size); + input.flags |= IOF_BUFFER_ALLOC; // todo: 3 modes: borrow, take over, copy? + return ppdoc_create(&input); +} + +void ppdoc_free (ppdoc *pdf) +{ + //iof_file_free(&pdf->input); + iof_file_decref(&pdf->input); + ppstack_free_buffer(&pdf->stack); + ppheap_free(pdf->heap); // last! +} + +ppcrypt_status ppdoc_crypt_status (ppdoc *pdf) +{ + return pdf->cryptstatus; +} + +ppint ppdoc_permissions (ppdoc *pdf) +{ + return pdf->crypt != NULL ? pdf->crypt->permissions : (ppint)0xFFFFFFFFFFFFFFFF; +} + +/* pages access */ + +static pparray * pppage_node (ppdict *dict, ppuint *count, ppname *type) +{ + ppname *pkey, key; + ppobj *obj; + pparray *kids = NULL; + *count = 0; + *type = NULL; + for (ppdict_first(dict, pkey, obj); (key = *pkey) != NULL; ppdict_next(pkey, obj)) + { + switch (key[0]) + { + case 'T': + if (ppname_is(key, "Type")) + *type = ppobj_get_name(obj); + break; + case 'C': + if (ppname_is(key, "Count")) + ppobj_get_uint(obj, *count); + break; + case 'K': + if (ppname_is(key, "Kids")) + kids = ppobj_rget_array(obj); + break; + } + } + return kids; +} + +#define ppname_is_page(type) (type != NULL && ppname_is(type, "Page")) + +ppuint ppdoc_page_count (ppdoc *pdf) +{ + ppref *ref; + ppname type; + ppuint count; + if ((ref = ppxref_pages(pdf->xref)) == NULL) + return 0; + if (pppage_node(ref->object.dict, &count, &type) == NULL) + return ppname_is_page(type) ? 1 : 0; // acrobat and ghostscript accept documents with root /Pages entry being a reference to a sole /Page object + return count; +} + +ppref * ppdoc_page (ppdoc *pdf, ppuint index) +{ + ppdict *dict; + ppuint count; + pparray *kids; + size_t size, i; + ppobj *r, *o; + ppref *ref; + ppname type; + + + if ((ref = ppxref_pages(pdf->xref)) == NULL) + return NULL; + dict = ref->object.dict; + if ((kids = pppage_node(dict, &count, &type)) != NULL) + { + if (index < 1 || index > count) + return NULL; + } + else + { + return index == 1 && ppname_is_page(type) ? ref : NULL; + } +scan_array: + if (index <= count / 2) + { // probably shorter way from the beginning + for (i = 0, size = kids->size, r = pparray_at(kids, 0); i < size; ++i, ++r) + { + if (r->type != PPREF) + return NULL; + o = &r->ref->object; + if (o->type != PPDICT) + return NULL; + dict = o->dict; + if ((kids = pppage_node(dict, &count, &type)) != NULL) + { + if (index <= count) + goto scan_array; + index -= count; + continue; + } + if (index == 1 && ppname_is_page(type)) + return r->ref; + --index; + } + } + else if ((size = kids->size) > 0) // for safe (size-1) + { // probably shorter way from the end + index = count - index + 1; + for (i = 0, r = pparray_at(kids, size - 1); i < size; ++i, --r) + { + if (r->type != PPREF) + return NULL; + o = &r->ref->object; + if (o->type != PPDICT) + return NULL; + dict = o->dict; + if ((kids = pppage_node(dict, &count, &type)) != NULL) + { + if (index <= count) + goto scan_array; + index -= count; + continue; + } + if (index == 1 && ppname_is_page(type)) + return r->ref; + --index; + } + } + return NULL; +} + +/* +Through pages iterator. Iterating over pages tree just on the base of /Kids and /Parent keys +is ineffective, as to get next pageref we need to take parent, find the pageref in /Kids, +take next (or go upper).. Annoying. We use a dedicated stack for pages iterator. This could +actually be done with pdf->stack, but some operations may clear it, so safer to keep it independent +Besides, its depth is constant (set on first use), so no need for allocs. +*/ + +static void ppdoc_pages_init (ppdoc *pdf) +{ + pppages *pages; + pages = &pdf->pages; + pages->root = pages->parent = pages->buffer; + pages->depth = 0; + pages->space = PPPAGES_STACK_DEPTH; +} + +static ppkids * pppages_push (ppdoc *pdf, pparray *kids) +{ + ppkids *newroot, *bounds; + pppages *pages; + pages = &pdf->pages; + if (pages->depth == pages->space) + { + pages->space <<= 1; + newroot = (ppkids *)ppheap_take(&pdf->heap, pages->space * sizeof(ppkids)); + memcpy(newroot, pages->root, pages->depth * sizeof(ppkids)); + pages->root = newroot; + } + bounds = pages->parent = &pages->root[pages->depth++]; + bounds->current = pparray_at(kids, 0); + bounds->sentinel = pparray_at(kids, kids->size); + return bounds; +} + +#define pppages_pop(pages) (--((pages)->parent), --((pages)->depth)) + +static ppref * ppdoc_pages_group_first (ppdoc *pdf, ppref *ref) +{ + ppdict *dict; + pparray *kids; + ppuint count; + ppname type; + + dict = ref->object.dict; // typecheck made by callers + while ((kids = pppage_node(dict, &count, &type)) != NULL) + { + if ((ref = pparray_get_ref(kids, 0)) == NULL || ref->object.type != PPDICT) + return NULL; + pppages_push(pdf, kids); + dict = ref->object.dict; + } + return ppname_is_page(type) ? ref : NULL; +} + +ppref * ppdoc_first_page (ppdoc *pdf) +{ + ppref *ref; + pppages *pages; + if ((ref = ppdoc_pages(pdf)) == NULL) + return NULL; + pages = &pdf->pages; + pages->parent = pages->root; + pages->depth = 0; + return ppdoc_pages_group_first(pdf, ref); +} + +ppref * ppdoc_next_page (ppdoc *pdf) +{ + pppages *pages; + ppkids *bounds; + ppref *ref; + ppobj *obj; + pages = &pdf->pages; + while (pages->depth > 0) + { + bounds = pages->parent; + obj = ++bounds->current; + if (obj < bounds->sentinel) + { + if (obj->type != PPREF) + return NULL; + ref = obj->ref; + if (ref->object.type != PPDICT) + return NULL; + return ppdoc_pages_group_first(pdf, ref); + } + else + { // no next node, go upper + pppages_pop(pages); + } + } + return NULL; +} + +/* context */ + +ppcontext * ppcontext_new (void) +{ + ppheap *heap; + ppcontext *context; + heap = ppheap_new(); + context = (ppcontext *)pp_malloc(sizeof(ppcontext)); // not from priv heap, as we delete it on renew + context->heap = heap; + ppstack_init(&context->stack, &context->heap); + return context; +} + +void ppcontext_done (ppcontext *context) +{ + ppheap_renew(context->heap); + ppstack_clear(&context->stack); +} + +void ppcontext_free (ppcontext *context) +{ + ppstack_free_buffer(&context->stack); + ppheap_free(context->heap); + pp_free(context); +} + +/* page contents streams */ + +//#define ppcontents_first_stream(array) pparray_rget_stream(array, 0) + +static ppstream * ppcontents_first_stream (pparray *array) +{ + size_t i; + ppobj *obj; + ppref *ref; + for (pparray_first(array, i, obj); i < array->size; pparray_next(i, obj)) + if ((ref = ppobj_get_ref(obj)) != NULL && ref->object.type == PPSTREAM) + return ref->object.stream; + return NULL; +} + +static ppstream * ppcontents_next_stream (pparray *array, ppstream *stream) +{ + size_t i; + ppobj *obj; + ppref *ref; + for (pparray_first(array, i, obj); i < array->size; pparray_next(i, obj)) + if ((ref = ppobj_get_ref(obj)) != NULL && ref->object.type == PPSTREAM && ref->object.stream == stream) + if (++i < array->size && (ref = ppobj_get_ref(obj + 1)) != NULL && ref->object.type == PPSTREAM) + return ref->object.stream; + return NULL; +} + +ppstream * ppcontents_first (ppdict *dict) +{ + ppobj *contentsobj; + if ((contentsobj = ppdict_rget_obj(dict, "Contents")) == NULL) + return NULL; + switch (contentsobj->type) + { + case PPARRAY: + return ppcontents_first_stream(contentsobj->array); + case PPSTREAM: + return contentsobj->stream; + default: + break; + } + return NULL; +} + +ppstream * ppcontents_next (ppdict *dict, ppstream *stream) +{ + ppobj *contentsobj; + if ((contentsobj = ppdict_rget_obj(dict, "Contents")) == NULL) + return NULL; + switch (contentsobj->type) + { + case PPARRAY: + return ppcontents_next_stream(contentsobj->array, stream); + case PPSTREAM: + break; + default: + break; + } + return NULL; +} + +static ppobj * ppcontents_op (iof *I, ppstack *stack, size_t *psize, ppname *pname) +{ + ppobj *obj; + ppstack_clear(stack); + do { + if (ppscan_find(I) < 0) + return NULL; + if ((obj = ppscan_psobj(I, stack)) == NULL) + return NULL; + } while (obj->type != PPNAME || !ppname_exec(obj->name)); + *pname = obj->name; + *psize = stack->size - 1; + return stack->buf; +} + +ppobj * ppcontents_first_op (ppcontext *context, ppstream *stream, size_t *psize, ppname *pname) +{ + iof *I; + if ((I = ppstream_read(stream, 1, 0)) == NULL) + return NULL; + return ppcontents_op(I, &context->stack, psize, pname); +} + +ppobj * ppcontents_next_op (ppcontext *context, ppstream *stream, size_t *psize, ppname *pname) +{ + return ppcontents_op(ppstream_iof(stream), &context->stack, psize, pname); +} + +ppobj * ppcontents_parse (ppcontext *context, ppstream *stream, size_t *psize) +{ + iof *I; + ppstack *stack; + ppobj *obj; + stack = &context->stack; + ppstack_clear(stack); + if ((I = ppstream_read(stream, 1, 0)) == NULL) + return NULL; + while (ppscan_find(I) >= 0) + if ((obj = ppscan_psobj(I, stack)) == NULL) + goto error; + *psize = stack->size; + ppstream_done(stream); + return stack->buf; +error: + ppstream_done(stream); + return NULL; +} + +/* boxes */ + +pprect * pparray_to_rect (pparray *array, pprect *rect) +{ + ppobj *obj; + if (array->size != 4) + return NULL; + obj = pparray_at(array, 0); + if (!ppobj_get_num(obj, rect->lx)) return NULL; + obj = pparray_at(array, 1); + if (!ppobj_get_num(obj, rect->ly)) return NULL; + obj = pparray_at(array, 2); + if (!ppobj_get_num(obj, rect->rx)) return NULL; + obj = pparray_at(array, 3); + if (!ppobj_get_num(obj, rect->ry)) return NULL; + return rect; +} + +pprect * ppdict_get_rect (ppdict *dict, const char *name, pprect *rect) +{ + pparray *array; + return (array = ppdict_rget_array(dict, name)) != NULL ? pparray_to_rect(array, rect) : NULL; +} + +pprect * ppdict_get_box (ppdict *dict, const char *name, pprect *rect) +{ + do { + if (ppdict_get_rect(dict, name, rect) != NULL) + return rect; + dict = ppdict_rget_dict(dict, "Parent"); + } while (dict != NULL); + return NULL; +} + +ppmatrix * pparray_to_matrix (pparray *array, ppmatrix *matrix) +{ + ppobj *obj; + if (array->size != 6) + return NULL; + obj = pparray_at(array, 0); + if (!ppobj_get_num(obj, matrix->xx)) return NULL; + obj = pparray_at(array, 1); + if (!ppobj_get_num(obj, matrix->xy)) return NULL; + obj = pparray_at(array, 2); + if (!ppobj_get_num(obj, matrix->yx)) return NULL; + obj = pparray_at(array, 3); + if (!ppobj_get_num(obj, matrix->yy)) return NULL; + obj = pparray_at(array, 4); + if (!ppobj_get_num(obj, matrix->x)) return NULL; + obj = pparray_at(array, 5); + if (!ppobj_get_num(obj, matrix->y)) return NULL; + return matrix; +} + +ppmatrix * ppdict_get_matrix (ppdict *dict, const char *name, ppmatrix *matrix) +{ + pparray *array; + return (array = ppdict_rget_array(dict, name)) != NULL ? pparray_to_matrix(array, matrix) : NULL; +} + +/* logger */ + +void pplog_callback (pplogger_callback logger, void *alien) +{ + logger_callback((logger_function)logger, alien); +} + +int pplog_prefix (const char *prefix) +{ + return logger_prefix(prefix); +} + +/* version */ + +const char * ppdoc_version_string (ppdoc *pdf) +{ + return pdf->version; +} + +int ppdoc_version_number (ppdoc *pdf, int *minor) +{ + *minor = pdf->version[2] - '0'; + return pdf->version[0] - '0'; +} + +/* doc info */ + +size_t ppdoc_file_size (ppdoc *pdf) +{ + return pdf->filesize; +} + +ppuint ppdoc_objects (ppdoc *pdf) +{ + ppuint count; + ppxref *xref; + for (count = 0, xref = pdf->xref; xref != NULL; xref = xref->prev) + count += xref->count; + return count; +} + +size_t ppdoc_memory (ppdoc *pdf, size_t *waste) +{ + size_t used; + ppheap *heap; + used = 0, *waste = 0; + for (heap = pdf->heap; heap != NULL; heap = heap->prev) + { + used += heap->space; + *waste += heap->size; + } + return used; +} diff --git a/Build/source/texk/web2c/luatexdir/luapplib/ppload.h b/Build/source/texk/web2c/luatexdir/luapplib/ppload.h new file mode 100644 index 00000000000..16392839893 --- /dev/null +++ b/Build/source/texk/web2c/luatexdir/luapplib/ppload.h @@ -0,0 +1,54 @@ + +#ifndef PP_LOAD_H +#define PP_LOAD_H + +typedef struct { + ppobj *buf; // ppobjects buffer (allocated, not from our heap) + ppobj *pos; // current ppobj * + size_t size; // stack size + size_t space; // available space + ppheap **pheap; // allocator (parent pdf->stack->pheap or own) +} ppstack; + +typedef struct { + ppobj *current; + ppobj *sentinel; +} ppkids; + +#define PPPAGES_STACK_DEPTH 4 + +typedef struct { + ppkids buffer[PPPAGES_STACK_DEPTH]; + ppkids *root; + ppkids *parent; + ppuint depth; + ppuint space; +} pppages; + +struct ppdoc { + char version[5]; + iof_file input; + iof reader; + uint8_t *buffer; + size_t filesize; + ppxref *xref; + ppheap *heap; + ppstack stack; + pppages pages; + int flags; + ppcrypt *crypt; + ppcrypt_status cryptstatus; +}; + +#define PPDOC_LINEARIZED (1 << 0) + +ppobj * ppdoc_load_entry (ppdoc *pdf, ppref *ref); +#define ppobj_preloaded(pdf, obj) ((obj)->type != PPREF ? (obj) : ((obj)->ref->object.type == PPNONE ? ppdoc_load_entry(pdf, (obj)->ref) : &(obj)->ref->object)) +ppstring ppstring_internal (const void *data, size_t size, ppheap **pheap); + +struct ppcontext { + ppheap *heap; + ppstack stack; +}; + +#endif
\ No newline at end of file diff --git a/Build/source/texk/web2c/luatexdir/luapplib/ppstream.c b/Build/source/texk/web2c/luatexdir/luapplib/ppstream.c new file mode 100644 index 00000000000..0e4be7b8e28 --- /dev/null +++ b/Build/source/texk/web2c/luatexdir/luapplib/ppstream.c @@ -0,0 +1,310 @@ + +#include "ppfilter.h" +#include "pplib.h" + +ppstream * ppstream_create (ppdoc *pdf, ppdict *dict, size_t offset) +{ + ppstream *stream; + stream = (ppstream *)ppheap_take(&pdf->heap, sizeof(ppstream)); + stream->dict = dict; + stream->offset = offset; + //if (!ppdict_rget_uint(dict, "Length", &stream->length)) // may be indirect pointing PPNONE at this moment + // stream->length = 0; + stream->length = 0; + stream->input = &pdf->input; + stream->I = NULL; + stream->cryptkey = NULL; + stream->flags = 0; + return stream; +} + +/* codecs */ + +enum { + PPSTREAM_UNKNOWN = -1, + PPSTREAM_BASE16 = 0, + PPSTREAM_BASE85, + PPSTREAM_RUNLENGTH, + PPSTREAM_FLATE, + PPSTREAM_LZW, + PPSTREAM_CCITT, + PPSTREAM_DCT, + PPSTREAM_JBIG2, + PPSTREAM_JPX, + PPSTREAM_CRYPT +}; + +static int ppstream_codec_type (ppname name) +{ // one of those places where some hash wuld be nice.. + switch (name[0]) + { + case 'A': + if (ppname_is(name, "ASCIIHexDecode")) return PPSTREAM_BASE16; + if (ppname_is(name, "ASCII85Decode")) return PPSTREAM_BASE85; + break; + case 'R': + if (ppname_is(name, "RunLengthDecode")) return PPSTREAM_RUNLENGTH; + break; + case 'F': + if (ppname_is(name, "FlateDecode")) return PPSTREAM_FLATE; + break; + case 'L': + if (ppname_is(name, "LZWDecode")) return PPSTREAM_LZW; + break; + case 'D': + if (ppname_is(name, "DCTDecode")) return PPSTREAM_DCT; + break; + case 'C': + if (ppname_is(name, "CCITTFaxDecode")) return PPSTREAM_CCITT; + if (ppname_is(name, "Crypt")) return PPSTREAM_CRYPT; + break; + case 'J': + if (ppname_is(name, "JPXDecode")) return PPSTREAM_JPX; + if (ppname_is(name, "JBIG2Decode")) return PPSTREAM_JBIG2; + break; + } + return PPSTREAM_UNKNOWN; +} + +static iof * ppstream_predictor (ppdict *params, iof *N) +{ + ppint predictor, rowsamples, components, samplebits; + + if (!ppdict_get_int(params, "Predictor", &predictor) || predictor <= 1) + return N; + if (!ppdict_get_int(params, "Columns", &rowsamples) || rowsamples == 0) // sanity, filter probably expects >0 + rowsamples = 1;; + if (!ppdict_get_int(params, "Colors", &components) || components == 0) // ditto + components = 1; + if (!ppdict_get_int(params, "BitsPerComponent", &samplebits) || samplebits == 0) + samplebits = 8; + return iof_filter_predictor_decoder(N, (int)predictor, (int)rowsamples, (int)components, (int)samplebits); +} + +static iof * ppstream_decoder (ppstream *stream, int codectype, ppdict *params, iof *N) +{ + int flags; + iof *F, *P; + ppint earlychange; + ppstring cryptkey; + + switch (codectype) + { + case PPSTREAM_BASE16: + return iof_filter_base16_decoder(N); + case PPSTREAM_BASE85: + return iof_filter_base85_decoder(N); + case PPSTREAM_RUNLENGTH: + return iof_filter_runlength_decoder(N); + case PPSTREAM_FLATE: + if ((F = iof_filter_flate_decoder(N)) != NULL) + { + if (params != NULL) + { + if ((P = ppstream_predictor(params, F)) != NULL) + return P; + iof_close(F); + break; + } + return F; + } + break; + case PPSTREAM_LZW: + flags = LZW_DECODER_DEFAULTS; + if (params != NULL && ppdict_get_int(params, "EarlyChange", &earlychange) && earlychange == 0) // integer, not boolean + flags &= ~LZW_EARLY_INDEX; + if ((F = iof_filter_lzw_decoder(N, flags)) != NULL) + { + if (params != NULL) + { + if ((P = ppstream_predictor(params, F)) != NULL) + return P; + iof_close(F); + break; + } + return F; + } + break; + case PPSTREAM_CRYPT: + if ((cryptkey = stream->cryptkey) == NULL) + return N; // /Identity crypt + if (stream->flags & PPSTREAM_ENCRYPTED_AES) + return iof_filter_aes_decoder(N, cryptkey, ppstring_size(cryptkey)); + if (stream->flags & PPSTREAM_ENCRYPTED_RC4) + return iof_filter_rc4_decoder(N, cryptkey, ppstring_size(cryptkey)); + return NULL; // if neither AES or RC4 but cryptkey present, something went wrong; see ppstream_info() + case PPSTREAM_CCITT: + case PPSTREAM_DCT: + case PPSTREAM_JBIG2: + case PPSTREAM_JPX: + case PPSTREAM_UNKNOWN: + break; + } + return NULL; +} + +#define ppstream_image(type) (type == PPSTREAM_DCT || type == PPSTREAM_JBIG2 || PPSTREAM_JPX) + +#define ppstream_source(stream) iof_filter_stream_coreader((iof_file *)((stream)->input), (size_t)((stream)->offset), (size_t)((stream)->length)) +#define ppstream_auxsource(filename) iof_filter_file_reader(filename) + +static ppname ppstream_filter_name (ppobj *filterobj, size_t index) +{ + if (filterobj->type == PPNAME) + return index == 0 ? filterobj->name : NULL; + if (filterobj->type == PPARRAY) + return pparray_get_name(filterobj->array, index); + return NULL; +} + +static ppdict * ppstream_filter_params (ppobj *paramsobj, size_t index) +{ + if (paramsobj->type == PPDICT) + return index == 0 ? paramsobj->dict : NULL; + if (paramsobj->type == PPARRAY) + return pparray_rget_dict(paramsobj->array, index); + return NULL; +} + +static const char * ppstream_aux_filename (ppobj *filespec) +{ // mockup, here we should decode the string + if (filespec->type == PPSTRING) + { + return (const char *)(filespec->string); + } + // else might be a dict - todo + return NULL; +} + +iof * ppstream_read (ppstream *stream, int decode, int all) +{ + ppdict *dict; + iof *I, *F; + int codectype, external, owncrypt; + ppobj *filterobj, *paramsobj, *filespecobj; + ppname filter; + ppdict *params; + size_t index; + const char *filename; + + if (ppstream_iof(stream) != NULL) + return NULL; // usage error + + dict = stream->dict; + if ((filespecobj = ppdict_rget_obj(dict, "F")) != NULL) + { + filename = ppstream_aux_filename(filespecobj); + I = filename != NULL ? ppstream_auxsource(filename) : NULL; + external = 1; + } + else + { + I = ppstream_source(stream); + external = 0; + } + if (I == NULL) + return NULL; + /* If the stream is encrypted, decipher is the first to be applied */ + owncrypt = (stream->flags & PPSTREAM_ENCRYPTED_OWN) != 0; + if (!owncrypt) + { + if (stream->cryptkey != NULL) + { /* implied global crypt */ + if ((F = ppstream_decoder(stream, PPSTREAM_CRYPT, NULL, I)) == NULL) + goto stream_error; + I = F; + } /* otherwise no crypt at all or /Identity */ + } + if (decode || owncrypt) + { + filterobj = ppdict_rget_obj(dict, external ? "FFilter" : "Filter"); + if (filterobj != NULL) + { + paramsobj = ppdict_rget_obj(dict, external ? "FDecodeParms" : "DecodeParms"); + for (index = 0, filter = ppstream_filter_name(filterobj, 0); filter != NULL; filter = ppstream_filter_name(filterobj, ++index)) + { + params = paramsobj != NULL ? ppstream_filter_params(paramsobj, index) : NULL; + codectype = ppstream_codec_type(filter); + if ((F = ppstream_decoder(stream, codectype, params, I)) != NULL) + { + I = F; + if (owncrypt && !decode && codectype == PPSTREAM_CRYPT) + break; // /Crypt filter should always be first, so in practise we return decrypted but compressed + continue; + } + if (!ppstream_image(codectype)) // something unexpected + goto stream_error; + else // just treat image data (jpeg/jbig) as the target data + break; + } + } + } + if (all) + iof_load(I); + else + iof_input(I); + stream->I = I; + return I; +stream_error: + iof_close(I); + return NULL; +} + +uint8_t * ppstream_first (ppstream *stream, size_t *size, int decode) +{ + iof *I; + if ((I = ppstream_read(stream, decode, 0)) != NULL) + { + *size = (size_t)iof_left(I); + return I->pos; + } + *size = 0; + return NULL; +} + +uint8_t * ppstream_next (ppstream *stream, size_t *size) +{ + iof *I; + if ((I = ppstream_iof(stream)) != NULL) + { + I->pos = I->end; + if ((*size = iof_input(I)) > 0) + return I->pos; + } + *size = 0; + return NULL; +} + +uint8_t * ppstream_all (ppstream *stream, size_t *size, int decode) +{ + iof *I; + if ((I = ppstream_read(stream, decode, 1)) != NULL) + { + *size = (size_t)iof_left(I); + return I->pos; + } + *size = 0; + return NULL; +} + +void ppstream_done (ppstream *stream) +{ + iof *I; + if ((I = ppstream_iof(stream)) != NULL) + { + iof_close(I); + stream->I = NULL; + } +} + +/* */ + +void ppstream_init_buffers (void) +{ + iof_filters_init(); +} + +void ppstream_free_buffers (void) +{ + iof_filters_free(); +} diff --git a/Build/source/texk/web2c/luatexdir/luapplib/ppstream.h b/Build/source/texk/web2c/luatexdir/luapplib/ppstream.h new file mode 100644 index 00000000000..ae5c19cd9f2 --- /dev/null +++ b/Build/source/texk/web2c/luatexdir/luapplib/ppstream.h @@ -0,0 +1,9 @@ + +#ifndef PP_STREAM_H +#define PP_STREAM_H + +ppstream * ppstream_create (ppdoc *pdf, ppdict *dict, size_t offset); +iof * ppstream_read (ppstream *stream, int decode, int all); +#define ppstream_iof(stream) ((iof *)((stream)->I)) + +#endif
\ No newline at end of file diff --git a/Build/source/texk/web2c/luatexdir/luapplib/pptest1.c b/Build/source/texk/web2c/luatexdir/luapplib/pptest1.c new file mode 100644 index 00000000000..4671849f551 --- /dev/null +++ b/Build/source/texk/web2c/luatexdir/luapplib/pptest1.c @@ -0,0 +1,86 @@ + +#include <stdio.h> +#include "ppapi.h" + +static const char * sizenum (size_t s) +{ + static char buffer[32]; + if (s < 1000) + sprintf(buffer, "%uB", (unsigned)s); + else if (s < 1000000) + sprintf(buffer, "%.2fkB", (double)(s) / 1000); + else + sprintf(buffer, "%.2fMB", (double)(s) / 1000000); + return buffer; +} + +static const char * crypt_info (ppdoc *pdf) +{ + switch (ppdoc_crypt_status(pdf)) + { + case PPCRYPT_NONE: + return "none"; + case PPCRYPT_DONE: + return "empty password"; + case PPCRYPT_PASS: + return "nonempty password"; + default: + break; + } + return "this shouldn't happen"; +} + +static void print_info (ppdoc *pdf) +{ + ppdict *info; + ppstring creator, producer; + size_t memused, memwaste; + + if ((info = ppdoc_info(pdf)) != NULL) + { + if ((creator = ppdict_rget_string(info, "Creator")) != NULL) + printf(" creator: %s\n", ppstring_decoded(creator)); + if ((producer = ppdict_rget_string(info, "Producer")) != NULL) + printf(" producer: %s\n", ppstring_decoded(producer)); + } + printf(" version: %s\n", ppdoc_version_string(pdf)); + printf(" protection: %s\n", crypt_info(pdf)); + printf(" filesize: %s\n", sizenum(ppdoc_file_size(pdf))); + printf(" objects: " PPUINTF "\n", ppdoc_objects(pdf)); + printf(" pagecount: " PPUINTF "\n", ppdoc_page_count(pdf)); + memused = ppdoc_memory(pdf, &memwaste); + printf(" memused: %s\n", sizenum(memused)); + printf(" memwaste: %s\n", sizenum(memwaste)); +} + +static int usage (const char *argv0) +{ + printf("pplib " pplib_version ", " pplib_author "\n"); + printf("usage: %s file1.pdf file2.pdf ...\n", argv0); + return 0; +} + +int main (int argc, const char **argv) +{ + const char *filepath; + int a; + ppdoc *pdf; + + if (argc < 2) + return usage(argv[0]); + for (a = 1; a < argc; ++a) + { + filepath = argv[a]; + printf("loading %s... ", filepath); + pdf = ppdoc_load(filepath); + if (pdf == NULL) + { + printf("failed\n"); + continue; + } + printf("done.\n"); + print_info(pdf); + ppdoc_free(pdf); + } + return 0; +} diff --git a/Build/source/texk/web2c/luatexdir/luapplib/pptest2.c b/Build/source/texk/web2c/luatexdir/luapplib/pptest2.c new file mode 100644 index 00000000000..4b7f66a0112 --- /dev/null +++ b/Build/source/texk/web2c/luatexdir/luapplib/pptest2.c @@ -0,0 +1,139 @@ + +#include <stdio.h> +#include <assert.h> +#include "ppapi.h" + +static const char * get_file_name (const char *path) +{ + const char *fn, *p; + for (fn = p = path; *p != '\0'; ++p) + if (*p == '\\' || *p == '/') + fn = p + 1; + return fn; +} + +static void box_info (ppdict *pagedict, FILE *fh) +{ + const char *boxes[] = {"MediaBox", "CropBox", "BleedBox", "TrimBox", "ArtBox"}; + pprect rect; + size_t i; + for (i = 0; i < sizeof(boxes) / sizeof(const char *); ++i) + if (ppdict_get_box(pagedict, boxes[i], &rect)) + fprintf(fh, "%%%% %s [%f %f %f %f]\n", boxes[i], rect.lx, rect.ly, rect.rx, rect.ry); +} + +static int usage (const char *argv0) +{ + printf("pplib " pplib_version ", " pplib_author "\n"); + printf("usage: %s file1.pdf file2.pdf ...\n", argv0); + return 0; +} + +#define OUTDIR "." + +static void log_callback (const char *message, void *alien) +{ + fprintf((FILE *)alien, "\nooops: %s\n", message); +} + +int main (int argc, const char **argv) +{ + const char *filepath, *filename; + int a; + ppdoc *pdf; + ppref *pageref; + ppdict *pagedict; + int pageno; + char outname[1024]; + FILE *fh; + ppstream *stream; + uint8_t *data; + size_t size; + ppcontext *context; + ppobj *obj; + ppname op; + size_t operators; + + if (argc < 2) + return usage(argv[0]); + ppstream_init_buffers(); + pplog_callback(log_callback, stderr); + context = ppcontext_new(); + for (a = 1; a < argc; ++a) + { + filepath = argv[a]; + printf("loading %s... ", filepath); + pdf = ppdoc_load(filepath); + if (pdf == NULL) + { + printf("failed\n"); + continue; + } + printf("done.\n"); + switch (ppdoc_crypt_status(pdf)) + { + case PPCRYPT_NONE: + case PPCRYPT_DONE: + break; + case PPCRYPT_PASS: + if (ppdoc_crypt_pass(pdf, "dummy", 5, NULL, 0) == PPCRYPT_DONE || ppdoc_crypt_pass(pdf, NULL, 0, "dummy", 5) == PPCRYPT_DONE) + break; + printf("sorry, password needed\n"); + ppdoc_free(pdf); + continue; + case PPCRYPT_FAIL: + printf("sorry, encryption failed\n"); + ppdoc_free(pdf); + continue; + } + filename = get_file_name(filepath); + sprintf(outname, OUTDIR "/%s.out", filename); + fh = fopen(outname, "wb"); + if (fh == NULL) + { + printf("can't open %s for writing\n", outname); + continue; + } + for (pageref = ppdoc_first_page(pdf), pageno = 1; + pageref != NULL; + pageref = ppdoc_next_page(pdf), ++pageno) + { + pagedict = pageref->object.dict; + /* decompress contents data */ + fprintf(fh, "%%%% PAGE %d\n", pageno); + box_info(pagedict, fh); + for (stream = ppcontents_first(pagedict); + stream != NULL; + stream = ppcontents_next(pagedict, stream)) + { + for (data = ppstream_first(stream, &size, 1); + data != NULL; + data = ppstream_next(stream, &size)) + fwrite(data, size, 1, fh); + ppstream_done(stream); + } + /* now parse contents */ + for (stream = ppcontents_first(pagedict); + stream != NULL; + stream = ppcontents_next(pagedict, stream)) + { + operators = 0; + for (obj = ppcontents_first_op(context, stream, &size, &op); + obj != NULL; + obj = ppcontents_next_op(context, stream, &size, &op)) + ++operators; + fprintf(fh, "%%%% OPERATORS count " PPSIZEF "\n", operators); + ppstream_done(stream); + //obj = ppcontents_parse(context, stream, &size); + //fprintf(fh, "%%%% items count " PPSIZEF "\n", size); + fprintf(fh, "\n"); + } + ppcontext_done(context); + } + fclose(fh); + ppdoc_free(pdf); + } + ppcontext_free(context); + ppstream_free_buffers(); + return 0; +} diff --git a/Build/source/texk/web2c/luatexdir/luapplib/ppxref.c b/Build/source/texk/web2c/luatexdir/luapplib/ppxref.c new file mode 100644 index 00000000000..984f6cbe327 --- /dev/null +++ b/Build/source/texk/web2c/luatexdir/luapplib/ppxref.c @@ -0,0 +1,162 @@ + +#include "pplib.h" + +#define PPXREF_MAP_INIT 16 // number of xref sections + +ppxref * ppxref_create (ppdoc *pdf, size_t initsize, size_t xrefoffset) +{ + ppxref *xref; + + if (initsize == 0) // unknown + initsize = PPXREF_MAP_INIT; + xref = (ppxref *)ppheap_take(&pdf->heap, sizeof(ppxref) + initsize * sizeof(ppxsec)); + xref->sects = (ppxsec *)(xref + 1); + xref->size = 0; + xref->space = initsize; + xref->count = 0; + xref->trailer.type = PPNONE; + xref->trailer.dict = NULL; + xref->prev = NULL; + xref->pdf = pdf; + xref->offset = xrefoffset; + //xref->crypt = NULL; + return xref; +} + +ppxsec * ppxref_push_section (ppxref *xref, ppheap **pheap) +{ + ppxsec *sects; + if (xref->size < xref->space) + return &xref->sects[xref->size++]; + xref->space <<= 1; + sects = xref->sects; + xref->sects = (ppxsec *)ppheap_take(pheap, xref->space * sizeof(ppxsec)); // waste but rare + memcpy(xref->sects, sects, xref->size * sizeof(ppxsec)); + return &xref->sects[xref->size++]; +} + +static void ppxref_sort_sects (ppxsec *left, ppxsec *right) +{ + ppxsec *l, *r, *m, t; + ppuint first, last; + l = left, r = right, m = l + ((r - l) / 2); + first = m->first, last = m->last; + do + { // don't take first/last from pointer + while (l->first < first) ++l; + while (r->first > last) --r; + if (l <= r) + { + t = *l; + *l = *r; + *r = t; + ++l, --r; + } + } while (l <= r); + if (l < right) + ppxref_sort_sects(l, right); + if (r > left) + ppxref_sort_sects(left, r); +} + +int ppxref_sort (ppxref *xref) +{ + if (xref->size == 0) + return 0; + ppxref_sort_sects(xref->sects, xref->sects + xref->size - 1); + return 1; +} + +ppref * ppxref_find_local (ppxref *xref, ppuint refnumber) +{ + ppxsec *left, *right, *mid; + //if (xref->size == 0) // we don't allow that + // return NULL; + left = xref->sects; + right = xref->sects + xref->size - 1; + do + { + mid = left + ((right - left) / 2); + if (refnumber > mid->last) + left = mid + 1; + else if (refnumber < mid->first) + right = mid - 1; + else + return &mid->refs[refnumber - mid->first]; + } while (left <= right); + return NULL; +} + +ppref * ppxref_find (ppxref *xref, ppuint refnumber) +{ + ppref *ref; + ppxref *other; + + if ((ref = ppxref_find_local(xref, refnumber)) != NULL) + return ref; + if (xref->pdf->flags & PPDOC_LINEARIZED) + { + for (other = xref->pdf->xref; other != NULL; other = other->prev) + if (other != xref && (ref = ppxref_find_local(other, refnumber)) != NULL) + return ref; + } + else + { + for (other = xref->prev; other != NULL; other = other->prev) + if ((ref = ppxref_find_local(other, refnumber)) != NULL) + return ref; + /* This shouldn't happen, but I've met documents that have no linearized dict, + but their xrefs are prepared as for linearized; with "older" xrefs referring + to "newer". */ + for (other = xref->pdf->xref; other != NULL && other != xref; other = other->prev) + if ((ref = ppxref_find_local(other, refnumber)) != NULL) + return ref; + } + return NULL; +} + +ppdict * ppxref_trailer (ppxref *xref) +{ + switch (xref->trailer.type) + { + case PPDICT: + return xref->trailer.dict; + case PPSTREAM: + return xref->trailer.stream->dict; + default: + break; + } + return NULL; +} + +ppxref * ppdoc_xref (ppdoc *pdf) +{ + return pdf->xref; +} + +ppxref * ppxref_prev (ppxref *xref) +{ + return xref->prev; +} + +ppdict * ppxref_catalog (ppxref *xref) +{ + ppdict *trailer; + return (trailer = ppxref_trailer(xref)) != NULL ? ppdict_rget_dict(trailer, "Root") : NULL; +} + +ppdict * ppxref_info (ppxref *xref) +{ + ppdict *trailer; + return (trailer = ppxref_trailer(xref)) != NULL ? ppdict_rget_dict(trailer, "Info") : NULL; +} + +ppref * ppxref_pages (ppxref *xref) +{ + ppdict *dict; + ppref *ref; + + if ((dict = ppxref_catalog(xref)) == NULL || (ref = ppdict_get_ref(dict, "Pages")) == NULL) + return NULL; + return ref->object.type == PPDICT ? ref : NULL; +} diff --git a/Build/source/texk/web2c/luatexdir/luapplib/ppxref.h b/Build/source/texk/web2c/luatexdir/luapplib/ppxref.h new file mode 100644 index 00000000000..86ab98a512a --- /dev/null +++ b/Build/source/texk/web2c/luatexdir/luapplib/ppxref.h @@ -0,0 +1,35 @@ + +#ifndef PP_XREF_H +#define PP_XREF_H + +/* +What we call xref is actually "xref section" in PDF spec and what we call section is "xref subsection". +Our ppxref is a list of sections, sorted by xrefsection->first and xrefsection->last bounds. Every section +keeps a list of ppref *refs, enumerated from xrefsection->first to xrefsection->last. To find a reference +by number we make a binary search over sections bounds, then jump to the proper ppref *ref. +*/ + +typedef struct { + ppuint first; // first reference number in section + ppuint last; // last reference number in section + ppref *refs; // references list +} ppxsec; + +struct ppxref { + ppxsec *sects; // subsections list + size_t size; // actual sections size + size_t space; // available sections space + ppobj trailer; // trailer dict or stream + ppuint count; // count of references in all sections + ppxref *prev; // previous xref + ppdoc *pdf; // parent pdf to access entries in linearized docs + size_t offset; // file offset of xref + //ppcrypt *crypt; // per xref encryption state? +}; + +ppxref * ppxref_create (ppdoc *pdf, size_t initsize, size_t xrefoffset); +ppxsec * ppxref_push_section (ppxref *xref, ppheap **pheap); +int ppxref_sort (ppxref *xref); +ppref * ppxref_find_local (ppxref *xref, ppuint refnumber); + +#endif
\ No newline at end of file diff --git a/Build/source/texk/web2c/luatexdir/luapplib/util/utilbasexx.c b/Build/source/texk/web2c/luatexdir/luapplib/util/utilbasexx.c new file mode 100644 index 00000000000..6fdd6fb58c5 --- /dev/null +++ b/Build/source/texk/web2c/luatexdir/luapplib/util/utilbasexx.c @@ -0,0 +1,2108 @@ + +#include "utilnumber.h" +#include "utilmem.h" +#include "utilbasexx.h" + +/* filters state structs */ + +struct basexx_state { + size_t line, maxline; + size_t left; + int tail[5]; + int flush; +}; +
+struct runlength_state { + int run; + int flush; + int c1, c2; + uint8_t *pos; +}; + +struct eexec_state { + int key; + int flush; + int binary; + int c1; + size_t line, maxline; /* ascii encoder only */ + const char *initbytes; +}; + +/* config */ + +#if defined(BASEXX_PDF) +# define ignored(c) (c == 0x20 || c == 0x0A || c == 0x0C || c == 0x0D || c == 0x09 || c == 0x00) +# define base16_eof(c) (c == '>' || c < 0) +# define base85_eof(c) (c == '~' || c < 0) +#else +# define ignored(c) (c == 0x20 || c == 0x0A || c == 0x0D || c == 0x09) +# define base16_eof(c) (c < 0) +# define base85_eof(c) (c < 0) +#endif + +#define base64_eof(c) (c == '=' || c < 0) + +#define basexx_nl '\x0A' +#define put_nl(O, line, maxline, n) ((void)((line += n) > maxline && ((line = n), iof_set(O, basexx_nl)))) + +/* tail macros */ + +#define set_tail1(state, c1) (state->left = 1, state->tail[0] = c1) +#define set_tail2(state, c1, c2) (state->left = 2, state->tail[0] = c1, state->tail[1] = c2) +#define set_tail3(state, c1, c2, c3) (state->left = 3, state->tail[0] = c1, state->tail[1] = c2, state->tail[2] = c3) +#define set_tail4(state, c1, c2, c3, c4) (state->left = 4, state->tail[0] = c1, state->tail[1] = c2, state->tail[2] = c3, state->tail[3] = c4) +#define set_tail5(state, c1, c2, c3, c4, c5) \ + (state->left = 5, state->tail[0] = c1, state->tail[1] = c2, state->tail[2] = c3, state->tail[3] = c4, state->tail[4] = c5) + +#define get_tail1(state, c1) (state->left = 0, c1 = state->tail[0]) +#define get_tail2(state, c1, c2) (state->left = 0, c1 = state->tail[0], c2 = state->tail[1]) +#define get_tail3(state, c1, c2, c3) (state->left = 0, c1 = state->tail[0], c2 = state->tail[1], c3 = state->tail[2]) +#define get_tail4(state, c1, c2, c3, c4) (state->left = 0, c1 = state->tail[0], c2 = state->tail[1], c3 = state->tail[2], c4 = state->tail[3]) + +/* basexx state initialization */ + +void basexx_state_init_ln (basexx_state *state, size_t line, size_t maxline) +{ + state->line = line; + state->maxline = maxline; + state->left = 0; + state->flush = 0; +} + +/* base 16; xxxx|xxxx */ + +iof_status base16_encoded_uc (const void *data, size_t size, iof *O) +{ + const uint8_t *s, *e; + for (s = (const uint8_t *)data, e = s + size; s < e; ++s) + { + if (!iof_ensure(O, 2)) + return IOFFULL; + iof_set_uc_hex(O, *s); + } + return IOFEOF; +} + +iof_status base16_encoded_lc (const void *data, size_t size, iof *O) +{ + const uint8_t *s, *e; + for (s = (const uint8_t *)data, e = s + size; s < e; ++s) + { + if (!iof_ensure(O, 2)) + return IOFFULL; + iof_set_lc_hex(O, *s); + } + return IOFEOF; +} + +iof_status base16_encoded_uc_ln (const void *data, size_t size, iof *O, size_t line, size_t maxline) +{ + const uint8_t *s, *e; + for (s = (const uint8_t *)data, e = s + size; s < e; ++s) + { + if (!iof_ensure(O, 3)) + return IOFFULL; + put_nl(O, line, maxline, 2); + iof_set_uc_hex(O, *s); + } + return IOFFULL; +} + +iof_status base16_encoded_lc_ln (const void *data, size_t size, iof *O, size_t line, size_t maxline) +{ + const uint8_t *s, *e; + for (s = (const uint8_t *)data, e = s + size; s < e; ++s) + { + if (!iof_ensure(O, 3)) + return IOFFULL; + put_nl(O, line, maxline, 2); + iof_set_lc_hex(O, *s); + } + return IOFFULL; +} + +iof_status base16_encode_uc (iof *I, iof *O) +{ + register int c; + while (iof_ensure(O, 2)) + { + if ((c = iof_get(I)) < 0) + return IOFEOF; + iof_set_uc_hex(O, c); + } + return IOFFULL; +} + +iof_status base16_encode_state_uc (iof *I, iof *O, basexx_state *state) +{ + register int c; + while (iof_ensure(O, 2)) + { + if ((c = iof_get(I)) < 0) + return (state->flush ? IOFEOF : IOFEMPTY); + iof_set_uc_hex(O, c); + } + return IOFFULL; +} + +iof_status base16_encode_lc (iof *I, iof *O) +{ + register int c; + while (iof_ensure(O, 2)) + { + if ((c = iof_get(I)) < 0) + return IOFEOF; + iof_set_lc_hex(O, c); + } + return IOFFULL; +} + +iof_status base16_encode_state_lc (iof *I, iof *O, basexx_state *state) +{ + register int c; + while (iof_ensure(O, 2)) + { + if ((c = iof_get(I)) < 0) + return (state->flush ? IOFEOF : IOFEMPTY); + iof_set_lc_hex(O, c); + } + return IOFFULL; +} + +iof_status base16_encode_uc_ln (iof *I, iof *O, size_t line, size_t maxline) +{ + register int c; + while (iof_ensure(O, 3)) + { + if ((c = iof_get(I)) < 0) + return IOFEOF; + put_nl(O, line, maxline, 2); + iof_set_uc_hex(O, c); + } + return IOFFULL; +} + +iof_status base16_encode_state_uc_ln (iof *I, iof *O, basexx_state *state) +{ + register int c; + while (iof_ensure(O, 3)) + { + if ((c = iof_get(I)) < 0) + return (state->flush ? IOFEOF : IOFEMPTY); + put_nl(O, state->line, state->maxline, 2); + iof_set_uc_hex(O, c); + } + return IOFFULL; +} + +iof_status base16_encode_lc_ln (iof *I, iof *O, size_t line, size_t maxline) +{ + register int c; + while (iof_ensure(O, 3)) + { + if ((c = iof_get(I)) < 0) + return IOFEOF; + put_nl(O, line, maxline, 2); + iof_set_lc_hex(O, c); + } + return IOFFULL; +} + +iof_status base16_encode_state_lc_ln (iof *I, iof *O, basexx_state *state) +{ + register int c; + while (iof_ensure(O, 3)) + { + if ((c = iof_get(I)) < 0) + return (state->flush ? IOFEOF : IOFEMPTY); + put_nl(O, state->line, state->maxline, 2); + iof_set_lc_hex(O, c); + } + return IOFFULL; +} + +int base16_getc (iof *I) +{ + register int c1, c2; + do { c1 = iof_get(I); } while (ignored(c1)); + if (base16_eof(c1)) + return IOFEOF; + do { c2 = iof_get(I); } while (ignored(c2)); + if (base16_eof(c2)) + { + if ((c1 = base16_lookup[c1]) == -1) + return IOFERR; + return c1<<4; + } + if ((c1 = base16_lookup[c1]) == -1 || (c2 = base16_lookup[c2]) == -1) + return IOFERR; + return (c1<<4)|c2; +} + +int base16_lc_putc (iof *O, int c) +{ + if (iof_ensure(O, 2)) + iof_set_lc_hex(O, c); + return IOFFULL; +} + +int base16_uc_putc (iof *O, int c) +{ + if (iof_ensure(O, 2)) + iof_set_uc_hex(O, c); + return IOFFULL; +} + + +iof_status base16_decode (iof *I, iof *O) +{ + register int c1, c2; + while (iof_ensure(O, 1)) + { + do { c1 = iof_get(I); } while (ignored(c1)); + if (base16_eof(c1)) + return IOFEOF; + do { c2 = iof_get(I); } while (ignored(c2)); + if (base16_eof(c2)) + { + if ((c1 = base16_lookup[c1]) == -1) + return IOFERR; + iof_set(O, c1<<4); // c2 := '0' + return IOFEOF; + } + if ((c1 = base16_lookup[c1]) == -1 || (c2 = base16_lookup[c2]) == -1) + return IOFERR; + iof_set(O, (c1<<4)|c2); + } + return IOFFULL; +} + +iof_status base16_decode_state (iof *I, iof *O, basexx_state *state) +{ + register int c1, c2, d1, d2; + if (!(iof_ensure(O, 1))) + return IOFFULL; + switch(state->left) + { + case 0: goto byte0; + case 1: get_tail1(state, c1); goto byte1; + } + while (iof_ensure(O, 1)) + { + byte0: + do { c1 = iof_get(I); } while (ignored(c1)); + if (base16_eof(c1)) + return (state->flush ? IOFEOF : IOFEMPTY); + byte1: + do { c2 = iof_get(I); } while (ignored(c2)); + if (base16_eof(c2)) + { + set_tail1(state, c1); /* set tail to let the caller display invalid chars */ + if (state->flush) + { + if ((c1 = base16_lookup[c1]) == -1) + return IOFERR; + iof_set(O, c1<<4); // c2 := '0' + return IOFEOF; + } + return IOFEMPTY; + } + if ((d1 = base16_lookup[c1]) == -1 || (d2 = base16_lookup[c2]) == -1) + { + set_tail2(state, c1, c2); + return IOFERR; + } + iof_set(O, (d1<<4)|d2); + } + return IOFFULL; +} + +/* base 64; xxxxxx|xx xxxx|xxxx xx|xxxxxx */ + +const char base64_alphabet[] = "ABCDEFGHIJKLMNOPQRSTUVWXYZabcdefghijklmnopqrstuvwxyz0123456789+/"; + +const int base64_lookup[] = { + -1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1, + -1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1, + -1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,62,-1,-1,-1,63, + 52,53,54,55,56,57,58,59,60,61,-1,-1,-1,-1,-1,-1, + -1, 0, 1, 2, 3, 4, 5, 6, 7, 8,9 ,10,11,12,13,14, + 15,16,17,18,19,20,21,22,23,24,25,-1,-1,-1,-1,-1, + -1,26,27,28,29,30,31,32,33,34,35,36,37,38,39,40, + 41,42,43,44,45,46,47,48,49,50,51,-1,-1,-1,-1,-1, + -1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1, + -1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1, + -1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1, + -1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1, + -1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1, + -1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1, + -1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1, + -1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1 +}; + +#define base64_digit1(c1) base64_alphabet[c1>>2] +#define base64_digit2(c1, c2) base64_alphabet[((c1&3)<<4)|(c2>>4)] +#define base64_digit3(c2, c3) base64_alphabet[((c2&15)<<2)|(c3>>6)] +#define base64_digit4(c3) base64_alphabet[c3&63] + +#define base64_encode_word(O, c1, c2, c3) \ + iof_set4(O, base64_digit1(c1), base64_digit2(c1, c2), base64_digit3(c2, c3), base64_digit4(c3)) + +#define base64_encode_tail2(O, c1, c2) \ + iof_set3(O, base64_digit1(c1), base64_digit2(c1, c2), base64_digit3(c2, 0)) + +#define base64_encode_tail1(O, c1) \ + iof_set2(O, base64_digit1(c1), base64_digit2(c1, 0)) + +iof_status base64_encoded (const void *data, size_t size, iof *O) +{ + const uint8_t *s, *e; + uint8_t c1, c2, c3; + for (s = (const uint8_t *)data, e = s + size; s + 2 < e; ) + { + if (!iof_ensure(O, 4)) + return IOFFULL; + c1 = *s++; + c2 = *s++; + c3 = *s++; + base64_encode_word(O, c1, c2, c3); + } + switch (e - s) + { + case 0: + break; + case 1: + if (!iof_ensure(O, 2)) + return IOFFULL; + c1 = *s; + base64_encode_tail1(O, c1); + break; + case 2: + if (!iof_ensure(O, 3)) + return IOFFULL; + c1 = *s++; + c2 = *s; + base64_encode_tail2(O, c1, c2); + break; + } + return IOFEOF; +} + +iof_status base64_encoded_ln (const void *data, size_t size, iof *O, size_t line, size_t maxline) +{ + const uint8_t *s, *e; + uint8_t c1, c2, c3; + for (s = (const uint8_t *)data, e = s + size; s + 2 < e; ) + { + if (!iof_ensure(O, 5)) + return IOFFULL; + c1 = *s++; + c2 = *s++; + c3 = *s++; + put_nl(O, line, maxline, 4); + base64_encode_word(O, c1, c2, c3); + } + switch (e - s) + { + case 0: + break; + case 1: + if (!iof_ensure(O, 3)) + return IOFFULL; + c1 = *s; + put_nl(O, line, maxline, 2); + base64_encode_tail1(O, c1); + break; + case 2: + if (!iof_ensure(O, 4)) + return IOFFULL; + c1 = *s++; + c2 = *s; + put_nl(O, line, maxline, 3); + base64_encode_tail2(O, c1, c2); + break; + } + return IOFEOF; +} + +iof_status base64_encode (iof *I, iof *O) +{ + register int c1, c2, c3; + while(iof_ensure(O, 4)) + { + if ((c1 = iof_get(I)) < 0) + return IOFEOF; + if ((c2 = iof_get(I)) < 0) + { + base64_encode_tail1(O, c1); + return IOFEOF; + } + if ((c3 = iof_get(I)) < 0) + { + base64_encode_tail2(O, c1, c2); + return IOFEOF; + } + base64_encode_word(O, c1, c2, c3); + } + return IOFFULL; +} + +iof_status base64_encode_state (iof *I, iof *O, basexx_state *state) +{ + register int c1, c2, c3; + if (!(iof_ensure(O, 4))) + return IOFFULL; + switch(state->left) + { + case 0: goto byte0; + case 1: get_tail1(state, c1); goto byte1; + case 2: get_tail2(state, c1, c2); goto byte2; + } + while(iof_ensure(O, 4)) + { + byte0: + if ((c1 = iof_get(I)) < 0) + return (state->flush ? IOFEOF : IOFEMPTY); + byte1: + if ((c2 = iof_get(I)) < 0) + return (state->flush ? (base64_encode_tail1(O, c1), IOFEOF) : (set_tail1(state, c1), IOFEMPTY)); + byte2: + if ((c3 = iof_get(I)) < 0) + return (state->flush ? (base64_encode_tail2(O, c1, c2), IOFEOF) : (set_tail2(state, c1, c2), IOFEMPTY)); + base64_encode_word(O, c1, c2, c3); + } + return IOFFULL; +} + +iof_status base64_encode_ln (iof *I, iof *O, size_t line, size_t maxline) +{ + register int c1, c2, c3; + while(iof_ensure(O, 5)) + { + if ((c1 = iof_get(I)) < 0) + return IOFEOF; + if ((c2 = iof_get(I)) < 0) + { + put_nl(O, line, maxline, 2); + base64_encode_tail1(O, c1); + return IOFEOF; + } + if ((c3 = iof_get(I)) < 0) + { + put_nl(O, line, maxline, 3); + base64_encode_tail2(O, c1, c2); + return IOFEOF; + } + put_nl(O, line, maxline, 4); + base64_encode_word(O, c1, c2, c3); + } + return IOFFULL; +} + +iof_status base64_encode_state_ln (iof *I, iof *O, basexx_state *state) +{ + register int c1, c2, c3; + if (!(iof_ensure(O, 5))) + return IOFFULL; + switch(state->left) + { + case 0: goto byte0; + case 1: get_tail1(state, c1); goto byte1; + case 2: get_tail2(state, c1, c2); goto byte2; + } + while(iof_ensure(O, 5)) + { + byte0: + if ((c1 = iof_get(I)) < 0) + return (state->flush ? IOFEOF : IOFEMPTY); + byte1: + if ((c2 = iof_get(I)) < 0) + return (state->flush ? (put_nl(O, state->line, state->maxline, 2), base64_encode_tail1(O, c1), IOFEOF) : (set_tail1(state, c1), IOFEMPTY)); + byte2: + if ((c3 = iof_get(I)) < 0) + return (state->flush ? (put_nl(O, state->line, state->maxline, 3), base64_encode_tail2(O, c1, c2), IOFEOF) : (set_tail2(state, c1, c2), IOFEMPTY)); + put_nl(O, state->line, state->maxline, 4); + base64_encode_word(O, c1, c2, c3); + } + return IOFFULL; +} + +// #define base64_code(c1, c2, c3, c4) ((c1<<18)|(c2<<12)|(c3<<6)|c4) + +#define base64_decode_word(O, c1, c2, c3, c4) \ + iof_set3(O, (c1<<2)|(c2>>4), ((c2&15)<<4)|(c3>>2), ((c3&3)<<6)|c4) + +#define base64_decode_tail3(O, c1, c2, c3) \ + iof_set2(O, (c1<<2)|(c2>>4), ((c2&15)<<4)|(c3>>2)) + +#define base64_decode_tail2(O, c1, c2) \ + iof_set(O, (c1<<2)|(c2>>4)) + +iof_status base64_decode (iof *I, iof *O) +{ + register int c1, c2, c3, c4; + while(iof_ensure(O, 3)) + { + do { c1 = iof_get(I); } while (ignored(c1)); + if (base64_eof(c1)) + return IOFEOF; + do { c2 = iof_get(I); } while (ignored(c2)); + if (base64_eof(c2)) + return IOFERR; + do { c3 = iof_get(I); } while (ignored(c3)); + if (base64_eof(c3)) + { + if ((c1 = base64_lookup[c1]) == -1 || (c2 = base64_lookup[c2]) == -1) + return IOFERR; + base64_decode_tail2(O, c1, c2); + return IOFEOF; + } + do { c4 = iof_get(I); } while (ignored(c4)); + if (base64_eof(c4)) + { + if ((c1 = base64_lookup[c1]) == -1 || (c2 = base64_lookup[c2]) == -1 || (c3 = base64_lookup[c3]) == -1) + return IOFERR; + base64_decode_tail3(O, c1, c2, c3); + return IOFEOF; + } + if ((c1 = base64_lookup[c1]) == -1 || (c2 = base64_lookup[c2]) == -1 || + (c3 = base64_lookup[c3]) == -1 || (c4 = base64_lookup[c4]) == -1) + return IOFERR; + base64_decode_word(O, c1, c2, c3, c4); + } + return IOFFULL; +} + +iof_status base64_decode_state (iof *I, iof *O, basexx_state *state) +{ + register int c1, c2, c3, c4; + register int d1, d2, d3, d4; + switch(state->left) + { + case 0: goto byte0; + case 1: get_tail1(state, c1); goto byte1; + case 2: get_tail2(state, c1, c2); goto byte2; + case 3: get_tail3(state, c1, c2, c3); goto byte3; + } + while(iof_ensure(O, 3)) + { + byte0: + do { c1 = iof_get(I); } while (ignored(c1)); + if (base64_eof(c1)) + return (state->flush ? IOFEOF : IOFEMPTY); + byte1: + do { c2 = iof_get(I); } while (ignored(c2)); + if (base64_eof(c2)) + { + set_tail1(state, c1); /* set tail to let the caller make padding or display invalid char in case of error */ + return (state->flush ? IOFERR : IOFEMPTY); /* if state->flush then error; tail must have at least two bytes */ + } + byte2: + do { c3 = iof_get(I); } while (ignored(c3)); + if (base64_eof(c3)) + { + set_tail2(state, c1, c2); + if (state->flush) + { + if ((c1 = base64_lookup[c1]) == -1 || (c2 = base64_lookup[c2]) == -1) + return IOFERR; + base64_decode_tail2(O, c1, c2); + return IOFEOF; + } + else + return IOFEMPTY; + } + byte3: + do { c4 = iof_get(I); } while (ignored(c4)); + if (base64_eof(c4)) + { + set_tail3(state, c1, c2, c3); + if (state->flush) + { + if ((c1 = base64_lookup[c1]) == -1 || (c2 = base64_lookup[c2]) == -1 || (c3 = base64_lookup[c3]) == -1) + return IOFERR; + base64_decode_tail3(O, c1, c2, c3); + return IOFEOF; + } + else + return IOFEMPTY; + } + if ((d1 = base64_lookup[c1]) == -1 || (d2 = base64_lookup[c2]) == -1 || + (d3 = base64_lookup[c3]) == -1 || (d4 = base64_lookup[c4]) == -1) + { + set_tail4(state, c1, c2, c3, c4); + return IOFERR; + } + base64_decode_word(O, d1, d2, d3, d4); + } + return IOFFULL; +} + +/* base85 */ + +const char base85_alphabet[] = "!\"#$%&'()*+,-./0123456789:;<=>?@ABCDEFGHIJKLMNOPQRSTUVWXYZ[\\]^_`abcdefghijklmnopqrstu"; /* for completness, not used below */ + +const int base85_lookup[] = { + -1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1, + -1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1, + -1, 0, 1, 2, 3, 4, 5, 6, 7, 8, 9,10,11,12,13,14, + 15,16,17,18,19,20,21,22,23,24,25,26,27,28,29,30, + 31,32,33,34,35,36,37,38,39,40,41,42,43,44,45,46, + 47,48,49,50,51,52,53,54,55,56,57,58,59,60,61,62, + 63,64,65,66,67,68,69,70,71,72,73,74,75,76,77,78, + 79,80,81,82,83,84,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1, + -1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1, + -1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1, + -1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1, + -1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1, + -1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1, + -1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1, + -1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1, + -1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1 +}; + +#define base85_encode_word(O, code) \ + (*(O->pos+4) = '!' + code%85, code /= 85, *(O->pos+3) = '!' + code%85, code /= 85, \ + *(O->pos+2) = '!' + code%85, code /= 85, *(O->pos+1) = '!' + code%85, code /= 85, \ + *(O->pos) = '!' + code, \ + O->pos += 5) + +#define base85_encode_tail3(O, code) \ + (*(O->pos+3) = '!' + code%85, code /= 85, *(O->pos+2) = '!' + code%85, code /= 85, \ + *(O->pos+1) = '!' + code%85, code /= 85, *(O->pos) = '!' + code, \ + O->pos += 4) + +#define base85_encode_tail2(O, code) \ + (*(O->pos+2) = '!' + code%85, code /= 85, *(O->pos+1) = '!' + code%85, code /= 85, \ + *(O->pos) = '!' + code, \ + O->pos += 3) + +#define base85_encode_tail1(O, code) \ + (*(O->pos+1) = '!' + code%85, code /= 85, *(O->pos) = '!' + code, \ + O->pos += 2) + +iof_status base85_encoded (const void *data, size_t size, iof *O) +{ + unsigned int code; + const uint8_t *s, *e; + uint8_t c1, c2, c3, c4; + for (s = (const uint8_t *)data, e = s + size; s + 3 < e; ) + { + if (!iof_ensure(O, 5)) + return IOFFULL; + c1 = *s++; + c2 = *s++; + c3 = *s++; + c4 = *s++; + code = (c1<<24)|(c2<<16)|(c3<<8)|c4; + if (code == 0) + { + iof_set(O, 'z'); + continue; + } + base85_encode_word(O, code); + } + switch (e - s) + { + case 0: + break; + case 1: + if (!iof_ensure(O, 2)) + return IOFFULL; + c1 = *s; + code = (c1<<24)/85/85/85; + base85_encode_tail1(O, code); + break; + case 2: + if (!iof_ensure(O, 3)) + return IOFFULL; + c1 = *s++; + c2 = *s; + code = ((c1<<24)|(c2<<16))/85/85; + base85_encode_tail2(O, code); + break; + case 3: + if (!iof_ensure(O, 4)) + return IOFFULL; + c1 = *s++; + c2 = *s++; + c3 = *s; + code = ((c1<<24)|(c2<<16)|(c3<<8))/85; + base85_encode_tail3(O, code); + break; + } + return IOFEOF; +} + +iof_status base85_encoded_ln (const void *data, size_t size, iof *O, size_t line, size_t maxline) +{ + unsigned int code; + const uint8_t *s, *e; + uint8_t c1, c2, c3, c4; + for (s = (const uint8_t *)data, e = s + size; s + 3 < e; ) + { + if (!iof_ensure(O, 6)) + return IOFFULL; + c1 = *s++; + c2 = *s++; + c3 = *s++; + c4 = *s++; + code = (c1<<24)|(c2<<16)|(c3<<8)|c4; + if (code == 0) + { + put_nl(O, line, maxline, 1); + iof_set(O, 'z'); + continue; + } + put_nl(O, line, maxline, 5); + base85_encode_word(O, code); + } + switch (e - s) + { + case 0: + break; + case 1: + if (!iof_ensure(O, 3)) + return IOFFULL; + c1 = *s; + code = (c1<<24)/85/85/85; + put_nl(O, line, maxline, 2); + base85_encode_tail1(O, code); + break; + case 2: + if (!iof_ensure(O, 4)) + return IOFFULL; + c1 = *s++; + c2 = *s; + code = ((c1<<24)|(c2<<16))/85/85; + put_nl(O, line, maxline, 3); + base85_encode_tail2(O, code); + break; + case 3: + if (!iof_ensure(O, 5)) + return IOFFULL; + c1 = *s++; + c2 = *s++; + c3 = *s; + code = ((c1<<24)|(c2<<16)|(c3<<8))/85; + put_nl(O, line, maxline, 4); + base85_encode_tail3(O, code); + break; + } + return IOFEOF; +} + +iof_status base85_encode (iof *I, iof *O) +{ + register int c1, c2, c3, c4; + register unsigned int code; + while(iof_ensure(O, 5)) + { + if ((c1 = iof_get(I)) < 0) + return IOFEOF; + if ((c2 = iof_get(I)) < 0) + { + code = (c1<<24)/85/85/85; + base85_encode_tail1(O, code); + return IOFEOF; + } + if ((c3 = iof_get(I)) < 0) + { + code = ((c1<<24)|(c2<<16))/85/85; + base85_encode_tail2(O, code); + return IOFEOF; + } + if ((c4 = iof_get(I)) < 0) + { + code = ((c1<<24)|(c2<<16)|(c3<<8))/85; + base85_encode_tail3(O, code); + return IOFEOF; + } + code = (c1<<24)|(c2<<16)|(c3<<8)|c4; + if (code == 0) + { + iof_set(O, 'z'); + continue; + } + /* in btoa 'y' character stays for 0x20202020, but pdf does not support this */ + /* if (code == 0x20202020) + { + iof_set(O, 'y'); + continue; + } */ + base85_encode_word(O, code); + } + return IOFFULL; +} + +iof_status base85_encode_state (iof *I, iof *O, basexx_state *state) +{ + register int c1, c2, c3, c4; + register unsigned int code; + if (!(iof_ensure(O, 5))) + return IOFFULL; + switch(state->left) + { + case 0: goto byte0; + case 1: get_tail1(state, c1); goto byte1; + case 2: get_tail2(state, c1, c2); goto byte2; + case 3: get_tail3(state, c1, c2, c3); goto byte3; + } + while(iof_ensure(O, 5)) + { + byte0: + if ((c1 = iof_get(I)) < 0) + return (state->flush ? IOFEOF : IOFEMPTY); + byte1: + if ((c2 = iof_get(I)) < 0) + { + set_tail1(state, c1); + if (state->flush) + { + code = (c1<<24)/85/85/85; + base85_encode_tail1(O, code); + return IOFEOF; + } + return IOFEMPTY; + } + byte2: + if ((c3 = iof_get(I)) < 0) + { + set_tail2(state, c1, c2); + if (state->flush) + { + code = ((c1<<24)|(c2<<16))/85/85; + base85_encode_tail2(O, code); + return IOFEOF; + } + return IOFEMPTY; + } + byte3: + if ((c4 = iof_get(I)) < 0) + { + set_tail3(state, c1, c2, c3); + if (state->flush) + { + code = ((c1<<24)|(c2<<16)|(c3<<8))/85; + base85_encode_tail3(O, code); + return IOFEOF; + } + return IOFEMPTY; + } + code = (c1<<24)|(c2<<16)|(c3<<8)|c4; + if (code == 0) + { + iof_set(O, 'z'); + continue; + } + base85_encode_word(O, code); + } + return IOFFULL; +} + +iof_status base85_encode_ln (iof *I, iof *O, size_t line, size_t maxline) +{ + register int c1, c2, c3, c4; + register unsigned int code; + while(iof_ensure(O, 6)) + { + if ((c1 = iof_get(I)) < 0) + return IOFEOF; + if ((c2 = iof_get(I)) < 0) + { + code = (c1<<24)/85/85/85; + put_nl(O, line, maxline, 2); + base85_encode_tail1(O, code); + return IOFEOF; + } + if ((c3 = iof_get(I)) < 0) + { + code = ((c1<<24)|(c2<<16))/85/85; + put_nl(O, line, maxline, 3); + base85_encode_tail2(O, code); + return IOFEOF; + } + if ((c4 = iof_get(I)) < 0) + { + code = ((c1<<24)|(c2<<16)|(c3<<8))/85; + put_nl(O, line, maxline, 4); + base85_encode_tail3(O, code); + return IOFEOF; + } + code = (c1<<24)|(c2<<16)|(c3<<8)|c4; + if (code == 0) + { + put_nl(O, line, maxline, 1); + iof_set(O, 'z'); + continue; + } + put_nl(O, line, maxline, 5); + base85_encode_word(O, code); + } + return IOFFULL; +} + +iof_status base85_encode_state_ln (iof *I, iof *O, basexx_state *state) +{ + register int c1, c2, c3, c4; + register unsigned int code; + if (!(iof_ensure(O, 6))) + return IOFFULL; + switch(state->left) + { + case 0: goto byte0; + case 1: get_tail1(state, c1); goto byte1; + case 2: get_tail2(state, c1, c2); goto byte2; + case 3: get_tail3(state, c1, c2, c3); goto byte3; + } + while(iof_ensure(O, 6)) + { + byte0: + if ((c1 = iof_get(I)) < 0) + return (state->flush ? IOFEOF : IOFEMPTY); + byte1: + if ((c2 = iof_get(I)) < 0) + { + set_tail1(state, c1); + if (state->flush) + { + code = (c1<<24)/85/85/85; + put_nl(O, state->line, state->maxline, 2); + base85_encode_tail1(O, code); + return IOFEOF; + } + return IOFEMPTY; + } + byte2: + if ((c3 = iof_get(I)) < 0) + { + set_tail2(state, c1, c2); + if (state->flush) + { + code = ((c1<<24)|(c2<<16))/85/85; + put_nl(O, state->line, state->maxline, 3); + base85_encode_tail2(O, code); + return IOFEOF; + } + return IOFEMPTY; + } + byte3: + if ((c4 = iof_get(I)) < 0) + { + set_tail3(state, c1, c2, c3); + if (state->flush) + { + code = ((c1<<24)|(c2<<16)|(c3<<8))/85; + put_nl(O, state->line, state->maxline, 4); + base85_encode_tail3(O, code); + return IOFEOF; + } + return IOFEMPTY; + } + code = (c1<<24)|(c2<<16)|(c3<<8)|c4; + if (code == 0) + { + put_nl(O, state->line, state->maxline, 1); + iof_set(O, 'z'); + continue; + } + put_nl(O, state->line, state->maxline, 5); + base85_encode_word(O, code); + } + return IOFFULL; +} + +#define base85_code(c1, c2, c3, c4, c5) ((((c1*85+c2)*85+c3)*85+c4)*85+c5) + +iof_status base85_decode (iof *I, iof *O) +{ + register int c1, c2, c3, c4, c5; + register unsigned int code; + while (iof_ensure(O, 4)) + { + do { c1 = iof_get(I); } while (ignored(c1)); + if (base85_eof(c1)) + return IOFEOF; + switch (c1) + { + case 'z': + iof_set4(O, '\0', '\0', '\0', '\0'); + continue; + case 'y': + iof_set4(O, ' ', ' ', ' ', ' '); + continue; + } + do { c2 = iof_get(I); } while (ignored(c2)); + if (base85_eof(c2)) + return IOFERR; + do { c3 = iof_get(I); } while (ignored(c3)); + if (base85_eof(c3)) + { + if ((c1 = base85_lookup[c1]) == -1 || (c2 = base85_lookup[c2]) == -1) + return IOFERR; + code = base85_code(c1, c2, 84, 84, 84); /* padding with 'u' (117); 117-33 = 84 */ + iof_set(O, code>>24); + return IOFEOF; + } + do { c4 = iof_get(I); } while (ignored(c4)); + if (base85_eof(c4)) + { + if ((c1 = base85_lookup[c1]) == -1 || (c2 = base85_lookup[c2]) == -1 || (c3 = base85_lookup[c3]) == -1) + return IOFERR; + code = base85_code(c1, c2, c3, 84, 84); + iof_set2(O, code>>24, (code>>16)&255); + return IOFEOF; + } + do { c5 = iof_get(I); } while (ignored(c5)); + if (base85_eof(c5)) + { + if ((c1 = base85_lookup[c1]) == -1 || (c2 = base85_lookup[c2]) == -1 || + (c3 = base85_lookup[c3]) == -1 || (c4 = base85_lookup[c4]) == -1) + return IOFERR; + code = base85_code(c1, c2, c3, c4, 84); + iof_set3(O, code>>24, (code>>16)&255, (code>>8)&255); + return IOFEOF; + } + if ((c1 = base85_lookup[c1]) == -1 || (c2 = base85_lookup[c2]) == -1 || (c3 = base85_lookup[c3]) == -1 || + (c4 = base85_lookup[c4]) == -1 || (c5 = base85_lookup[c5]) == -1) + return IOFERR; + code = base85_code(c1, c2, c3, c4, c5); + iof_set4(O, code>>24, (code>>16)&255, (code>>8)&255, code&255); + } + return IOFFULL; +} + +iof_status base85_decode_state (iof *I, iof *O, basexx_state *state) +{ + register int c1, c2, c3, c4, c5; + register int d1, d2, d3, d4, d5; + register unsigned int code; + if (!(iof_ensure(O, 4))) + return IOFFULL; + switch(state->left) + { + case 0: goto byte0; + case 1: get_tail1(state, c1); goto byte1; + case 2: get_tail2(state, c1, c2); goto byte2; + case 3: get_tail3(state, c1, c2, c3); goto byte3; + case 4: get_tail4(state, c1, c2, c3, c4); goto byte4; + } + while (iof_ensure(O, 4)) + { + byte0: + do { c1 = iof_get(I); } while (ignored(c1)); + if (base85_eof(c1)) + return (state->flush ? IOFEOF : IOFEMPTY); + switch (c1) + { + case 'z': + iof_set4(O, '\0', '\0', '\0', '\0'); + continue; + case 'y': + iof_set4(O, ' ', ' ', ' ', ' '); + continue; + } + byte1: + do { c2 = iof_get(I); } while (ignored(c2)); + if (base85_eof(c2)) + { + set_tail1(state, c1); + return (state->flush ? IOFERR : IOFEMPTY); /* if state->flush then error; tail must have at least two bytes */ + } + byte2: + do { c3 = iof_get(I); } while (ignored(c3)); + if (base85_eof(c3)) + { + set_tail2(state, c1, c2); + if (state->flush) + { + if ((c1 = base85_lookup[c1]) == -1 || (c2 = base85_lookup[c2]) == -1) + return IOFERR; + code = base85_code(c1, c2, 84, 84, 84); + iof_set(O, code>>24); + return IOFEOF; + } + return IOFEMPTY; + } + byte3: + do { c4 = iof_get(I); } while (ignored(c4)); + if (base85_eof(c4)) + { + set_tail3(state, c1, c2, c3); + if (state->flush) + { + if ((c1 = base85_lookup[c1]) == -1 || (c2 = base85_lookup[c2]) == -1 || (c3 = base85_lookup[c3]) == -1) + return IOFERR; + code = base85_code(c1, c2, c3, 84, 84); + iof_set2(O, code>>24, (code>>16)&255); + return IOFEOF; + } + return IOFEMPTY; + } + byte4: + do { c5 = iof_get(I); } while (ignored(c5)); + if (base85_eof(c5)) + { + set_tail4(state, c1, c2, c3, c4); + if (state->flush) + { + if ((c1 = base85_lookup[c1]) == -1 || (c2 = base85_lookup[c2]) == -1 || + (c3 = base85_lookup[c3]) == -1 || (c4 = base85_lookup[c4]) == -1) + return IOFERR; + code = base85_code(c1, c2, c3, c4, 84); + iof_set3(O, code>>24, (code>>16)&255, (code>>8)&255); + return IOFEOF; + } + return IOFEMPTY; + } + if ((d1 = base85_lookup[c1]) == -1 || (d2 = base85_lookup[c2]) == -1 || (d3 = base85_lookup[c3]) == -1 || + (d4 = base85_lookup[c4]) == -1 || (d5 = base85_lookup[c5]) == -1) + { + set_tail5(state, c1, c2, c3, c4, c5); + return IOFERR; + } + code = base85_code(d1, d2, d3, d4, d5); + iof_set4(O, code>>24, (code>>16)&255, (code>>8)&255, code&255); + } + return IOFFULL; +} + +/* postscript run length */ + +void runlength_state_init (runlength_state *state) +{ + state->run = -1; + state->flush = 0; + state->c1 = 0; + state->c2 = 0; + state->pos = NULL; +} + +iof_status runlength_encode (iof *I, iof *O) +{ + register int c1, c2, run = -1; + uint8_t *pos; + c1 = 0; /* avoid warning */ + while (iof_ensure(O, 1+128+1)) + { /* ensured space for single length byte, up to 128 bytes to be copied, possible eod marker */ + pos = O->pos++; + switch (run) + { + case -1: /* initial state; get first byte */ + if ((c1 = iof_get(I)) < 0) + return (*pos = 128, IOFEOF); + run = 0; + // fall through + case 0: /* `repeat' state; get another byte and compare */ + if ((c2 = iof_get(I)) < 0) + return (*pos = 0, iof_set2(O, c1, 128), IOFEOF); + run = (c1 == c2 ? 257-2 : 0); + break; + } + if (run < 128) + { /* single length byte, up to 128 bytes to be copied, possible eod marker */ + iof_set(O, c1); + for (c1 = c2, c2 = iof_char(I); c1 != c2 && run < 127; c1 = c2, c2 = iof_next(I)) + { + if (c2 < 0) /* O->pos must not change until next call to calling encoder!!! */ + return (*pos = (uint8_t)run+1, iof_set2(O, c1, 128), IOFEOF); + iof_set(O, c1); + ++run; + } + } + else // if run > 128 + { + for (c2 = iof_get(I); c1 == c2 && run > 129; c2 = iof_get(I)) + --run; + if (c2 < 0) + return (*pos = (uint8_t)run, iof_set2(O, c1, 128), IOFEOF); + iof_set(O, c1); + } + *pos = (uint8_t)run; + c1 = c2; + run = 0; + } + return IOFFULL; +} + +iof_status runlength_encode_state (iof *I, iof *O, runlength_state *state) +{ + while (iof_ensure(O, 3)) /* single length byte, the byte to be repeated and eod */ + { + state->pos = O->pos++; + switch (state->run) + { + case -1: /* initial state; get first byte */ + if ((state->c1 = iof_get(I)) < 0) + return (state->flush ? (*state->pos = 128, IOFEOF) : IOFEMPTY); + state->run = 0; + // fall through + case 0: /* `repeat' state; get another byte and compare */ + if ((state->c2 = iof_get(I)) < 0) + return (state->flush ? (*state->pos = 0, iof_set2(O, state->c1, 128), IOFEOF) : IOFEMPTY); + state->run = (state->c1 == state->c2 ? 257-2 : 0); + break; + } + if (state->run < 128) + { /* ensure space for single length byte, up to 128 bytes to be copied, plus possible eod marker, minus those already copied */ + if (!iof_ensure(O, 1+128+1-state->run)) + return IOFFULL; + iof_set(O, state->c1); + for (state->c1 = state->c2, state->c2 = iof_char(I); + state->c1 != state->c2 && state->run < 127; + state->c1 = state->c2, state->c2 = iof_next(I)) + { + if (state->c2 < 0) /* O->pos must not change until next call to calling encoder!!! */ + return (state->flush ? (*state->pos = (uint8_t)state->run+1, iof_set2(O, state->c1, 128), IOFEOF) : IOFEMPTY); + iof_set(O, state->c1); + ++state->run; + } + } + else // if run > 128 + { + for (state->c2 = iof_get(I); state->c1 == state->c2 && state->run > 129; state->c2 = iof_get(I)) + --state->run; + if (state->c2 < 0) + return (state->flush ? (*state->pos = (uint8_t)state->run, iof_set2(O, state->c1, 128), IOFEOF) : IOFEMPTY); + iof_set(O, state->c1); + } + *state->pos = (uint8_t)state->run; + state->c1 = state->c2; + state->run = 0; + } + return IOFFULL; +} + +iof_status runlength_decode (iof *I, iof *O) +{ + register int c, run = -1; + while (1) + { + if (run == -1) /* initial state */ + { + if ((run = iof_get(I)) < 0) + { + run = -1; /* don't assume IOFEOF == -1 */ + return IOFEOF; + } + } + if (run < 128) + { /* copy (run + 1) following bytes */ + while (run > -1) + { + if (iof_ensure(O, 1)) + { + if ((c = iof_get(I)) < 0) + return IOFERR; + iof_set(O, c); + --run; + continue; + } + return IOFFULL; + } + } + else if (run > 128) + { /* replicate the following byte (257 - run) times */ + if ((c = iof_get(I)) < 0) /* cf. state-wise version; don't change input position until we got this byte */ + return IOFERR; + while (run < 257) + { + if (iof_ensure(O, 1)) + { + iof_set(O, c); + ++run; + continue; + } + return IOFFULL; + } + run = -1; + } + else // c == 128 + return IOFEOF; + } + // return IOFFULL; +} + +iof_status runlength_decode_state (iof *I, iof *O, runlength_state *state) +{ + register int c; + while (1) + { + if (state->run == -1) /* initial state */ + { + if ((state->run = iof_char(I)) < 0) + { + state->run = -1; /* don't assume IOFEOF == -1 */ + return (state->flush ? IOFEOF : IOFEMPTY); + } + ++I->pos; + } + if (state->run < 128) + { /* copy (state->run + 1) following bytes */ + while (state->run > -1) + { + if (iof_ensure(O, 1)) + { + if ((c = iof_char(I)) < 0) + return (state->flush ? IOFERR : IOFEMPTY); + ++I->pos; + iof_set(O, c); + --state->run; + continue; + } + return IOFFULL; + } + } + else if (state->run > 128) + { /* replicate the following byte (257 - state->run) times */ + if ((c = iof_char(I)) < 0) + return (state->flush ? IOFERR : IOFEMPTY); + ++I->pos; + while (state->run < 257) + { + if (iof_ensure(O, 1)) + { + iof_set(O, c); + ++state->run; + continue; + } + return IOFFULL; + } + state->run = -1; + } + else // c == 128 + return IOFEOF; + } + // return IOFFULL; +} + +/* eexec stream filter, type1 fonts spec page 63 */ + +void eexec_state_init_ln (eexec_state *state, size_t line, size_t maxline, const char *initbytes) +{ + state->key = -1; + state->flush = 0; + state->binary = maxline > 0; + state->c1 = -1; + state->line = line; + state->maxline = maxline; + state->initbytes = initbytes; +} + +#define eexec_getc(I, c1) if ((c1 = iof_get(I)) < 0) return c1 +#define eexec_key(key, c1) (key = ((((c1) + key)*52845 + 22719) & 65535)) +#define eexec_decipher(key, c1, c) (c = ((c1)^(key>>8)), eexec_key(key, c1)) +#define eexec_encipher(key, c1, c) (c = ((c1)^(key>>8)), eexec_key(key, c)) + +#ifndef lps_ignored_char +# define lps_ignored_char(c) (c == 0x20 || c == 0x0A || c == 0x0C || c == 0x0D || c == 0x09 || c == 0x00) +#endif + +#define eexec_getx_(I, c) \ + do { c = iof_get(I); } while (lps_ignored_char(c)); \ + if (c < 0) return IOFEOF; \ + if ((c = base16_lookup[c]) < 0) return IOFERR + +#define eexec_getx(I, c1, c2) eexec_getx_(I, c1); \ + do { c2 = iof_get(I); } while (lps_ignored_char(c2)); \ + if (c2 < 0) c2 = 0; else \ + if ((c2 = base16_lookup[c2]) < 0) return IOFERR + +static int eexec_decode_init (iof *I, int *key, int *binary) +{ + int c1, c2, c3, c4; + *key = 55665; + eexec_getc(I, c1); + eexec_getc(I, c2); + eexec_getc(I, c3); + eexec_getc(I, c4); /* four head bytes */ + + /* eexec data has no explicit data termination. The caller of eexec_decode() should ensure that either + the input iof allows to read no more then necessary, or the output has no more space then 512 bytes, + to land safely somewhere in 512 bytes that probably follows eexec */ + + *binary = (base16_lookup[c1] < 0 || base16_lookup[c2] < 0 || + base16_lookup[c3] < 0 || base16_lookup[c4] < 0); + if (*binary) + { /* gobble 4 random bytes keeping decipher key up-to-date */ + eexec_key(*key, c1); eexec_key(*key, c2); + eexec_key(*key, c3); eexec_key(*key, c4); + } + else /* pfa/postscript only, pdf requires binary eexec form */ + { /* gobble 4 random bytes (8 hex digits) keeping decipher key up-to-date */ + c1 = base16_lookup[c1], c2 = base16_lookup[c2], c3 = base16_lookup[c3], c4 = base16_lookup[c4]; + eexec_key(*key, (c1<<4)|c2); eexec_key(*key, (c3<<4)|c4); /* dummy bytes 1, 2 */ + eexec_getx(I, c1, c2); eexec_getx(I, c3, c4); + eexec_key(*key, (c1<<4)|c2); eexec_key(*key, (c3<<4)|c4); /* dummy bytes 3, 4 */ + } + return 0; +} + +iof_status eexec_decode (iof *I, iof *O) +{ + int c, c1, c2, key, binary, status; + if ((status = eexec_decode_init(I, &key, &binary)) < 0) + return status; + if (binary) + { + while (iof_ensure(O, 1)) + { + eexec_getc(I, c1); + eexec_decipher(key, c1, c); + iof_set(O, c); + } + } + else + { + while (iof_ensure(O, 1)) + { + eexec_getx(I, c1, c2); + eexec_decipher(key, (c1<<4)|c2, c); + iof_set(O, c); + } + } + return IOFFULL; +} + +iof_status eexec_decode_state (iof *I, iof *O, eexec_state *state) +{ + register int c, c1, c2, status; + if (state->key == -1) /* initial state */ + if ((status = eexec_decode_init(I, &state->key, &state->binary)) < 0) + return status; + if (state->binary) + { + while (iof_ensure(O, 1)) + { + if ((c1 = iof_get(I)) < 0) + return (state->flush ? IOFEOF : IOFEMPTY); + else if (c1 < 0) + return c1; + eexec_decipher(state->key, c1, c); + iof_set(O, c); + } + } + else + { + if (state->c1 > -1) + { + c1 = state->c1; + state->c1 = -1; + goto byte2; + } + while (iof_ensure(O, 1)) + { + eexec_getx_(I, c1); + byte2: + do { c2 = iof_get(I); } while (lps_ignored_char(c2)); + if (c2 < 0) // odd number of hex bytes? hmm... assume c2 is zero + { + if (state->flush) + { + eexec_decipher(state->key, (c1<<4), c); + iof_set(O, c); + return IOFEOF; + } + state->c1 = c1; + return IOFEMPTY; + } + else if (c2 < 0) + return c2; + if ((c2 = base16_lookup[c2]) < 0) + return IOFERR; + eexec_decipher(state->key, (c1<<4)|c2, c); + iof_set(O, c); + } + } + return IOFFULL; +} + +#define EEXEC_INIT_BYTES "" + +iof_status eexec_encode (iof *I, iof *O, size_t line, size_t maxline) +{ + int c1, c, key; + const char *p; + + key = 55665; + p = EEXEC_INIT_BYTES; + if (maxline == 0) + { + if (!iof_ensure(O, 4)) + return IOFFULL; + eexec_encipher(key, p[0], c); iof_set(O, c); + eexec_encipher(key, p[1], c); iof_set(O, c); + eexec_encipher(key, p[2], c); iof_set(O, c); + eexec_encipher(key, p[3], c); iof_set(O, c); + while (iof_ensure(O, 1)) + { + eexec_getc(I, c1); + eexec_encipher(key, c1, c); + iof_set(O, c); + } + } + else + { + if (!iof_ensure(O, 8 + 1)) + return IOFFULL; + eexec_encipher(key, p[0], c); put_nl(O, line, maxline, 2); iof_set_uc_hex(O, c); + eexec_encipher(key, p[1], c); put_nl(O, line, maxline, 2); iof_set_uc_hex(O, c); + eexec_encipher(key, p[2], c); put_nl(O, line, maxline, 2); iof_set_uc_hex(O, c); + eexec_encipher(key, p[3], c); put_nl(O, line, maxline, 2); iof_set_uc_hex(O, c); + while (iof_ensure(O, 2 + 1)) + { + eexec_getc(I, c1); + eexec_encipher(key, c1, c); + put_nl(O, line, maxline, 2); + iof_set_uc_hex(O, c); + } + } + return IOFFULL; +} + +iof_status eexec_encode_state (iof *I, iof *O, eexec_state *state) +{ + int c, c1; + const char *p; + + if (state->key == -1) + { + p = state->initbytes != NULL ? state->initbytes : EEXEC_INIT_BYTES; + if (state->binary) + { + if (!iof_ensure(O, 4)) + return IOFFULL; + state->key = 55665; + eexec_encipher(state->key, p[0], c); iof_set(O, c); + eexec_encipher(state->key, p[1], c); iof_set(O, c); + eexec_encipher(state->key, p[2], c); iof_set(O, c); + eexec_encipher(state->key, p[3], c); iof_set(O, c); + } + else + { + if (!iof_ensure(O, 8 + 1)) + return IOFFULL; + state->key = 55665; + eexec_encipher(state->key, p[0], c); put_nl(O, state->line, state->maxline, 2); iof_set_uc_hex(O, c); + eexec_encipher(state->key, p[1], c); put_nl(O, state->line, state->maxline, 2); iof_set_uc_hex(O, c); + eexec_encipher(state->key, p[2], c); put_nl(O, state->line, state->maxline, 2); iof_set_uc_hex(O, c); + eexec_encipher(state->key, p[3], c); put_nl(O, state->line, state->maxline, 2); iof_set_uc_hex(O, c); + } + } + if (state->binary) + { + while (iof_ensure(O, 1)) + { + if ((c1 = iof_get(I)) < 0) + return c1 < 0 ? (state->flush ? IOFEOF : IOFEMPTY) : c1; + eexec_encipher(state->key, c1, c); + iof_set(O, c); + } + } + else + { + while (iof_ensure(O, 2 + 1)) + { + if ((c1 = iof_get(I)) < 0) + return c1 < 0 ? (state->flush ? IOFEOF : IOFEMPTY) : c1; + eexec_encipher(state->key, c1, c); + put_nl(O, state->line, state->maxline, 2); + iof_set_uc_hex(O, c); + } + } + return IOFFULL; +} + +/* +Type1 charstrings are encoded/decoded in the same way, except that: +- initial key is 4330 +- initbytes might be other then 4; /lenIV key in Private dict +Type1 spec page 63. In practise we don't need iof interface here, +we always apply the codec in place. +*/ + +#define type1chstr_key(key, c1) (key = ((((c1) + key)*52845 + 22719) & 65535)) +#define type1chstr_decipher(key, c1, c) (c = ((c1)^(key>>8)), type1chstr_key(key, c1)) +#define type1chstr_encipher(key, c1, c) (c = ((c1)^(key>>8)), type1chstr_key(key, c)) + +int type1_charstring_decode (void *data, size_t size, void *outdata, uint8_t leniv) +{ /* data and outdata may be the same, output size is always size - leniv */ + uint8_t *input = (uint8_t *)data, *output = (uint8_t *)outdata; + size_t i; + int c, c1, key; + + if (size < 4) + return 0; + key = 4330; + for (i = 0; i < leniv; ++i) + { + c1 = input[i]; + type1chstr_key(key, c1); + } + for ( ; i < size; ++i) + { + c1 = input[i]; + type1chstr_decipher(key, c1, c); + output[i - leniv] = c; + } + return 1; +} + +#define TYPE1_CHSTR_INIT_BYTES EEXEC_INIT_BYTES + +int type1_charstring_encode (void *data, size_t size, void *outdata, uint8_t leniv) +{ /* outdata may be data - leniv, output size is always size + leniv */ + uint8_t *input = (uint8_t *)data, *output = (uint8_t *)outdata; + size_t i, j; + int c, c1, key; + + key = 4330; + for (i = 0, j = 0; i < leniv; ++i) + { + c1 = TYPE1_CHSTR_INIT_BYTES[j]; + //type1chstr_key(key, c1); + type1chstr_encipher(key, c1, c); + if (++j == 4) + j = 0; + output[i] = c; + } + for (i = 0; i < size; ++i) + { + c1 = input[i]; + type1chstr_encipher(key, c1, c); + output[i + leniv] = c; + } + return 1; +} + +/* filters */ + +// base16 decoder function + +static size_t base16_decoder (iof *F, iof_mode mode) +{ + basexx_state *state; + iof_status status; + size_t tail; + + switch(mode) + { + case IOFLOAD: + case IOFREAD: + if (F->flags & IOF_STOPPED) + return 0; + tail = iof_tail(F); + F->pos = F->buf + tail; + F->end = F->buf + F->space; + state = iof_filter_state(basexx_state *, F); + do { + status = base16_decode_state(F->next, F, state); + } while (mode == IOFLOAD && status == IOFFULL && iof_resize_buffer(F)); + return iof_decoder_retval(F, "base16", status); + case IOFCLOSE: + iof_free(F); + return 0; + default: + break; + } + return 0; +} + +// base16 encoder function + +static size_t base16_encoder (iof *F, iof_mode mode) +{ + basexx_state *state; + iof_status status; + + state = iof_filter_state(basexx_state *, F); + switch (mode) + { + case IOFFLUSH: + state->flush = 1; + // fall through + case IOFWRITE: + F->end = F->pos; + F->pos = F->buf; + status = base16_encode_state_ln(F, F->next, state); + return iof_encoder_retval(F, "base16", status); + case IOFCLOSE: + if (!state->flush) + base16_encoder(F, IOFFLUSH); + iof_free(F); + return 0; + default: + break; + } + return 0; +} + +// base64 decoder function + +static size_t base64_decoder (iof *F, iof_mode mode) +{ + basexx_state *state; + iof_status status; + size_t tail; + + switch(mode) + { + case IOFLOAD: + case IOFREAD: + if (F->flags & IOF_STOPPED) + return 0; + tail = iof_tail(F); + F->pos = F->buf + tail; + F->end = F->buf + F->space; + state = iof_filter_state(basexx_state *, F); + do { + status = base64_decode_state(F->next, F, state); + } while (mode == IOFLOAD && status == IOFFULL && iof_resize_buffer(F)); + return iof_decoder_retval(F, "base64", status); + case IOFCLOSE: + iof_free(F); + return 0; + default: + break; + } + return 0; +} + +// base64 encoder function + +static size_t base64_encoder (iof *F, iof_mode mode) +{ + basexx_state *state; + iof_status status; + + state = iof_filter_state(basexx_state *, F); + switch (mode) + { + case IOFFLUSH: + state->flush = 1; + // fall through + case IOFWRITE: + F->end = F->pos; + F->pos = F->buf; + status = base64_encode_state_ln(F, F->next, state); + return iof_encoder_retval(F, "base64", status); + case IOFCLOSE: + if (!state->flush) + base64_encoder(F, IOFFLUSH); + iof_free(F); + return 0; + default: + break; + } + return 0; +} + +// base85 decoder function + +static size_t base85_decoder (iof *F, iof_mode mode) +{ + basexx_state *state; + iof_status status; + size_t tail; + + switch(mode) + { + case IOFLOAD: + case IOFREAD: + if (F->flags & IOF_STOPPED) + return 0; + tail = iof_tail(F); + F->pos = F->buf + tail; + F->end = F->buf + F->space; + state = iof_filter_state(basexx_state *, F); + do { + status = base85_decode_state(F->next, F, state); + } while (mode == IOFLOAD && status == IOFFULL && iof_resize_buffer(F)); + return iof_decoder_retval(F, "base85", status); + case IOFCLOSE: + iof_free(F); + return 0; + default: + break; + } + return 0; +} + +// base85 encoder function + +static size_t base85_encoder (iof *F, iof_mode mode) +{ + basexx_state *state; + iof_status status; + + state = iof_filter_state(basexx_state *, F); + switch (mode) + { + case IOFFLUSH: + state->flush = 1; + // fall through + case IOFWRITE: + F->end = F->pos; + F->pos = F->buf; + status = base85_encode_state_ln(F, F->next, state); + return iof_encoder_retval(F, "base85", status); + case IOFCLOSE: + if (!state->flush) + base85_encoder(F, IOFFLUSH); + iof_free(F); + return 0; + default: + break; + } + return 0; +} + +// runlength decoder function + +static size_t runlength_decoder (iof *F, iof_mode mode) +{ + runlength_state *state; + iof_status status; + size_t tail; + + switch(mode) + { + case IOFLOAD: + case IOFREAD: + if (F->flags & IOF_STOPPED) + return 0; + tail = iof_tail(F); + F->pos = F->buf + tail; + F->end = F->buf + F->space; + state = iof_filter_state(runlength_state *, F); + do { + status = runlength_decode_state(F->next, F, state); + } while (mode == IOFLOAD && status == IOFFULL && iof_resize_buffer(F)); + return iof_decoder_retval(F, "runlength", status); + case IOFCLOSE: + iof_free(F); + return 0; + default: + break; + } + return 0; +} + +// runlength encoder function + +static size_t runlength_encoder (iof *F, iof_mode mode) +{ + runlength_state *state; + iof_status status; + + state = iof_filter_state(runlength_state *, F); + switch (mode) + { + case IOFFLUSH: + state->flush = 1; + // fall through + case IOFWRITE: + F->end = F->pos; + F->pos = F->buf; + status = runlength_encode_state(F, F->next, state); + return iof_encoder_retval(F, "runlength", status); + case IOFCLOSE: + if (!state->flush) + runlength_encoder(F, IOFFLUSH); + iof_free(F); + return 0; + default: + break; + } + return 0; +} + +// eexec decoder function + +static size_t eexec_decoder (iof *F, iof_mode mode) +{ + eexec_state *state; + iof_status status; + size_t tail; + + switch(mode) + { + case IOFLOAD: + case IOFREAD: + if (F->flags & IOF_STOPPED) + return 0; + tail = iof_tail(F); + F->pos = F->buf + tail; + F->end = F->buf + F->space; + state = iof_filter_state(eexec_state *, F); + do { + status = eexec_decode_state(F->next, F, state); + } while (mode == IOFLOAD && status == IOFFULL && iof_resize_buffer(F)); + return iof_decoder_retval(F, "eexec", status); + case IOFCLOSE: + iof_free(F); + return 0; + default: + break; + } + return 0; +} + +// eexec encoder function + +static size_t eexec_encoder (iof *F, iof_mode mode) +{ + eexec_state *state; + iof_status status; + + state = iof_filter_state(eexec_state *, F); + switch (mode) + { + case IOFFLUSH: + state->flush = 1; + // fall through + case IOFWRITE: + F->end = F->pos; + F->pos = F->buf; + status = eexec_encode_state(F, F->next, state); + return iof_encoder_retval(F, "eexec", status); + case IOFCLOSE: + if (!state->flush) + eexec_encoder(F, IOFFLUSH); + iof_free(F); + return 0; + default: + break; + } + return 0; +} + +// + +int iof_filter_basexx_encoder_ln (iof *F, size_t line, size_t maxline) +{ + basexx_state *state; + if (maxline > 8 && line < maxline) + { + state = iof_filter_state(basexx_state *, F); + state->line = line; + state->maxline = maxline; + return 1; + } + return 0; +} + +/* base 16 */ + +iof * iof_filter_base16_decoder (iof *N) +{ + iof *I; + basexx_state *state; + I = iof_filter_reader(base16_decoder, sizeof(basexx_state), &state); + iof_setup_next(I, N); + basexx_state_init(state); + state->flush = 1; // means N is supposed to be continuous input + return I; +} + +iof * iof_filter_base16_encoder (iof *N) +{ + iof *O; + basexx_state *state; + O = iof_filter_writer(base16_encoder, sizeof(basexx_state), &state); + iof_setup_next(O, N); + basexx_state_init(state); + return O; +} + +/* base 64 */ + +iof * iof_filter_base64_decoder (iof *N) +{ + iof *I; + basexx_state *state; + I = iof_filter_reader(base64_decoder, sizeof(basexx_state), &state); + iof_setup_next(I, N); + basexx_state_init(state); + state->flush = 1; + return I; +} + +iof * iof_filter_base64_encoder (iof *N) +{ + iof *O; + basexx_state *state; + O = iof_filter_writer(base64_encoder, sizeof(basexx_state), &state); + iof_setup_next(O, N); + basexx_state_init(state); + return O; +} + +/* base 85 */ + +iof * iof_filter_base85_decoder (iof *N) +{ + iof *I; + basexx_state *state; + I = iof_filter_reader(base85_decoder, sizeof(basexx_state), &state); + iof_setup_next(I, N); + basexx_state_init(state); + state->flush = 1; + return I; +} + +iof * iof_filter_base85_encoder (iof *N) +{ + iof *O; + basexx_state *state; + O = iof_filter_writer(base85_encoder, sizeof(basexx_state), &state); + iof_setup_next(O, N); + basexx_state_init(state); + return O; +} + +/* runlength stream filter */ + +iof * iof_filter_runlength_decoder (iof *N) +{ + iof *I; + runlength_state *state; + I = iof_filter_reader(runlength_decoder, sizeof(runlength_state), &state); + iof_setup_next(I, N); + runlength_state_init(state); + state->flush = 1; + return I; +} + +iof * iof_filter_runlength_encoder (iof *N) +{ + iof *O; + runlength_state *state; + O = iof_filter_writer(runlength_encoder, sizeof(runlength_state), &state); + iof_setup_next(O, N); + runlength_state_init(state); + return O; +} + +/* eexec stream filter, type1 fonts spec p. 63 */ + +iof * iof_filter_eexec_decoder (iof *N) +{ + iof *I; + eexec_state *state; + I = iof_filter_reader(eexec_decoder, sizeof(eexec_state), &state); + iof_setup_next(I, N); + eexec_state_init(state); + state->flush = 1; + return I; +} + +iof * iof_filter_eexec_encoder (iof *N) +{ + iof *O; + eexec_state *state; + O = iof_filter_writer(eexec_encoder, sizeof(eexec_state), &state); + iof_setup_next(O, N); + eexec_state_init(state); + return O; +} diff --git a/Build/source/texk/web2c/luatexdir/luapplib/util/utilbasexx.h b/Build/source/texk/web2c/luatexdir/luapplib/util/utilbasexx.h new file mode 100644 index 00000000000..52aac3772ca --- /dev/null +++ b/Build/source/texk/web2c/luatexdir/luapplib/util/utilbasexx.h @@ -0,0 +1,133 @@ + +/* base encodings */ + +#ifndef UTIL_BASEXX_H +#define UTIL_BASEXX_H + +#include "utiliof.h" + +/* base codecs state */ + +typedef struct basexx_state basexx_state; + +#define BASEXX_MAXLINE 80 +#define BASEXX_PDF + +void basexx_state_init_ln (basexx_state *state, size_t line, size_t maxline); +#define basexx_state_init(state) basexx_state_init_ln(state, 0, BASEXX_MAXLINE) + +/* base16 */ + +int base16_getc (iof *I); +int base16_uc_putc (iof *I, int c); +int base16_lc_putc (iof *I, int c); +#define base16_putc base16_uc_putc + +iof_status base16_encoded_uc (const void *data, size_t size, iof *O); +iof_status base16_encoded_lc (const void *data, size_t size, iof *O); +iof_status base16_encoded_uc_ln (const void *data, size_t size, iof *O, size_t line, size_t maxline); +iof_status base16_encoded_lc_ln (const void *data, size_t size, iof *O, size_t line, size_t maxline); + +iof_status base16_encode_uc (iof *I, iof *O); +iof_status base16_encode_lc (iof *I, iof *O); +iof_status base16_encode_uc_ln (iof *I, iof *O, size_t line, size_t maxline); +iof_status base16_encode_lc_ln (iof *I, iof *O, size_t line, size_t maxline); +iof_status base16_decode (iof *I, iof *O); + +#define base16_encoded base16_encoded_uc +#define base16_encoded_ln base16_encoded_uc_ln +#define base16_encode base16_encode_uc +#define base16_encode_ln base16_encode_uc_ln + +iof_status base16_encode_state_uc (iof *I, iof *O, basexx_state *state); +iof_status base16_encode_state_lc (iof *I, iof *O, basexx_state *state); +iof_status base16_encode_state_uc_ln (iof *I, iof *O, basexx_state *state); +iof_status base16_encode_state_lc_ln (iof *I, iof *O, basexx_state *state); +iof_status base16_decode_state (iof *I, iof *O, basexx_state *state); + +#define base16_encode_state base16_encode_state_uc +#define base16_encode_state_ln base16_encode_state_uc_ln + +/* base64 */ + +extern const char base64_alphabet[]; +extern const int base64_lookup[]; + +iof_status base64_encoded (const void *data, size_t size, iof *O); +iof_status base64_encoded_ln (const void *data, size_t size, iof *O, size_t line, size_t maxline); + +iof_status base64_encode (iof *I, iof *O); +iof_status base64_encode_ln (iof *I, iof *O, size_t line, size_t maxline); +iof_status base64_decode (iof *I, iof *O); + +iof_status base64_encode_state (iof *I, iof *O, basexx_state *state); +iof_status base64_encode_state_ln (iof *I, iof *O, basexx_state *state); +iof_status base64_decode_state (iof *I, iof *O, basexx_state *state); + +/* base85 */ + +extern const char base85_alphabet[]; +extern const int base85_lookup[]; + +iof_status base85_encoded (const void *data, size_t size, iof *O); +iof_status base85_encoded_ln (const void *data, size_t size, iof *O, size_t line, size_t maxline); + +iof_status base85_encode (iof *I, iof *O); +iof_status base85_encode_ln (iof *I, iof *O, size_t line, size_t maxline); +iof_status base85_decode (iof *I, iof *O); + +iof_status base85_encode_state (iof *I, iof *O, basexx_state *state); +iof_status base85_encode_state_ln (iof *I, iof *O, basexx_state *state); +iof_status base85_decode_state (iof *I, iof *O, basexx_state *state); + +/* run length */ + +typedef struct runlength_state runlength_state; + +void runlength_state_init (runlength_state *state); + +iof_status runlength_encode (iof *I, iof *O); +iof_status runlength_encode_state (iof *I, iof *O, runlength_state *state); + +iof_status runlength_decode (iof *I, iof *O); +iof_status runlength_decode_state (iof *I, iof *O, runlength_state *state); + +/* eexec */ + +typedef struct eexec_state eexec_state; + +#define EEXEC_MAXLINE 80 + +void eexec_state_init_ln (eexec_state *state, size_t line, size_t maxline, const char *initbytes); +#define eexec_state_init(state) eexec_state_init_ln(state, 0, EEXEC_MAXLINE, NULL) + +iof_status eexec_decode (iof *I, iof *O); +iof_status eexec_decode_state (iof *I, iof *O, eexec_state *state); + +iof_status eexec_encode (iof *I, iof *O, size_t line, size_t maxline); +iof_status eexec_encode_state (iof *I, iof *O, eexec_state *state); + +iof_status type1_charstring_decode (void *data, size_t size, void *outdata, uint8_t leniv); +iof_status type1_charstring_encode (void *data, size_t size, void *outdata, uint8_t leniv); + +/* filters */ + +int iof_filter_basexx_encoder_ln (iof *N, size_t line, size_t maxline); + +iof * iof_filter_base16_decoder (iof *N); +iof * iof_filter_base16_encoder (iof *N); + +iof * iof_filter_base64_decoder (iof *N); +iof * iof_filter_base64_encoder (iof *N); + +iof * iof_filter_base85_decoder (iof *N); +iof * iof_filter_base85_encoder (iof *N); + +iof * iof_filter_runlength_decoder (iof *N); +iof * iof_filter_runlength_encoder (iof *N); + +iof * iof_filter_eexec_decoder (iof *N); +iof * iof_filter_eexec_encoder (iof *N); + + +#endif diff --git a/Build/source/texk/web2c/luatexdir/luapplib/util/utilcrypt.c b/Build/source/texk/web2c/luatexdir/luapplib/util/utilcrypt.c new file mode 100644 index 00000000000..1750d01bd86 --- /dev/null +++ b/Build/source/texk/web2c/luatexdir/luapplib/util/utilcrypt.c @@ -0,0 +1,1191 @@ + +#include "utilmem.h" +#include "utilcrypt.h" +#include "utilcryptdef.h" +#include "utilmd5.h" + +/* rc4 */ + +/* +Initializer arguments: +- state - crypt state +- map - a space for rc4 bytes map; may be left NULL in which case will be allocated +- vkey - crypt key; may be left NULL iff map is provided and properly initialized +- keylength - the length of crypt key (from 5 to 16 bytes) +*/ + +rc4_state * rc4_state_initialize (rc4_state *state, rc4_map *map, const void *vkey, size_t keylength) +{ + int i, j; + uint8_t tmp; + const uint8_t *key; + key = (const uint8_t *)vkey; + if (keylength == 0 || keylength > 256) + return NULL; + state->flags = 0; + if (map != NULL) + { + state->map = map; + } + else + { + state->map = (rc4_map *)util_malloc(sizeof(rc4_map)); + state->flags |= RC4_STATE_ALLOC; + } + + if (key != NULL) + { + for (i = 0; i < 256; ++i) + state->smap[i] = (uint8_t)i; + for (i = 0, j = 0; i < 256; ++i) + { + j = (j + state->smap[i] + key[i % keylength]) & 255; + tmp = state->smap[i]; + state->smap[i] = state->smap[j]; + state->smap[j] = tmp; + } + } + state->i = 0; + state->j = 0; + state->flush = 0; /* caller is responsible to override if necessary */ + return state; +} + +void rc4_map_save (rc4_state *state, rc4_map *map) +{ + memcpy(map, state->map, sizeof(rc4_map)); +} + +void rc4_map_restore (rc4_state *state, rc4_map *map) +{ + memcpy(state->map, map, sizeof(rc4_map)); + //state->flags = 0; + //state->flush = 0; + state->i = 0; + state->j = 0; +} + +static uint8_t rc4_next_random_byte (rc4_state *state) +{ + uint8_t tmp; + state->i = (state->i + 1) & 255; + state->j = (state->j + state->smap[state->i]) & 255; + tmp = state->smap[state->i]; + state->smap[state->i] = state->smap[state->j]; + state->smap[state->j] = tmp; + return state->smap[(state->smap[state->i] + state->smap[state->j]) & 255]; +} + +iof_status rc4_crypt_state (iof *I, iof *O, rc4_state *state) +{ + uint8_t r; + int c; + while (iof_ensure(O, 1)) + { + if ((c = iof_get(I)) < 0) + return c == IOFERR ? IOFERR : (state->flush ? IOFEOF : IOFEMPTY); + r = rc4_next_random_byte(state); + //r = r ^ ((uint8_t)c); + //iof_set(O, r); + iof_set(O, r ^ ((uint8_t)c)); + } + return IOFFULL; +} + +iof_status rc4_crypt (iof *I, iof *O, const void *key, size_t keylength) +{ + int ret; + rc4_state state; + rc4_map map; + if (rc4_state_initialize(&state, &map, key, keylength) == NULL) + return IOFERR; + state.flush = 1; + ret = rc4_crypt_state(I, O, &state); + rc4_state_close(&state); + return ret; +} + +/* +Variants that operates on c-strings can worn inplace, so output and input can be the same address. +Variant that takes rc4_state pointer expects the state properly initialized. Keep in mind +the crypt procedure modifies rc4 bytes map. All returns the size of encrypted/decrypted +data, which is the same as input data length for rc4. +*/ + +size_t rc4_crypt_data (const void *input, size_t length, void *output, const void *key, size_t keylength) +{ + rc4_state state; + rc4_map map; + if (rc4_state_initialize(&state, &map, key, keylength) == NULL) + return 0; + return rc4_crypt_state_data(&state, input, length, output); + // no need to call rc4_state_close() +} + +size_t rc4_crypt_state_data (rc4_state *state, const void *input, size_t length, void *output) +{ /* state assumed to be initialized and with the proper state of smap */ + const uint8_t *inp; + uint8_t r, *out; + size_t size; + inp = (const uint8_t *)input; + out = (uint8_t *)output; + for (size = 0; size < length; ++size, ++inp, ++out) + { + r = rc4_next_random_byte(state); + *out = r ^ *inp; + } + return length; +} + +void rc4_state_close (rc4_state *state) +{ + if (state->smap != NULL && (state->flags & RC4_STATE_ALLOC)) + { + util_free(state->smap); + state->smap = NULL; + } +} + +/* aes; parts of code excerpted from https://github.com/kokke/tiny-AES128-C */ + +static const uint8_t sbox[256] = { + 0x63, 0x7c, 0x77, 0x7b, 0xf2, 0x6b, 0x6f, 0xc5, 0x30, 0x01, 0x67, 0x2b, 0xfe, 0xd7, 0xab, 0x76, + 0xca, 0x82, 0xc9, 0x7d, 0xfa, 0x59, 0x47, 0xf0, 0xad, 0xd4, 0xa2, 0xaf, 0x9c, 0xa4, 0x72, 0xc0, + 0xb7, 0xfd, 0x93, 0x26, 0x36, 0x3f, 0xf7, 0xcc, 0x34, 0xa5, 0xe5, 0xf1, 0x71, 0xd8, 0x31, 0x15, + 0x04, 0xc7, 0x23, 0xc3, 0x18, 0x96, 0x05, 0x9a, 0x07, 0x12, 0x80, 0xe2, 0xeb, 0x27, 0xb2, 0x75, + 0x09, 0x83, 0x2c, 0x1a, 0x1b, 0x6e, 0x5a, 0xa0, 0x52, 0x3b, 0xd6, 0xb3, 0x29, 0xe3, 0x2f, 0x84, + 0x53, 0xd1, 0x00, 0xed, 0x20, 0xfc, 0xb1, 0x5b, 0x6a, 0xcb, 0xbe, 0x39, 0x4a, 0x4c, 0x58, 0xcf, + 0xd0, 0xef, 0xaa, 0xfb, 0x43, 0x4d, 0x33, 0x85, 0x45, 0xf9, 0x02, 0x7f, 0x50, 0x3c, 0x9f, 0xa8, + 0x51, 0xa3, 0x40, 0x8f, 0x92, 0x9d, 0x38, 0xf5, 0xbc, 0xb6, 0xda, 0x21, 0x10, 0xff, 0xf3, 0xd2, + 0xcd, 0x0c, 0x13, 0xec, 0x5f, 0x97, 0x44, 0x17, 0xc4, 0xa7, 0x7e, 0x3d, 0x64, 0x5d, 0x19, 0x73, + 0x60, 0x81, 0x4f, 0xdc, 0x22, 0x2a, 0x90, 0x88, 0x46, 0xee, 0xb8, 0x14, 0xde, 0x5e, 0x0b, 0xdb, + 0xe0, 0x32, 0x3a, 0x0a, 0x49, 0x06, 0x24, 0x5c, 0xc2, 0xd3, 0xac, 0x62, 0x91, 0x95, 0xe4, 0x79, + 0xe7, 0xc8, 0x37, 0x6d, 0x8d, 0xd5, 0x4e, 0xa9, 0x6c, 0x56, 0xf4, 0xea, 0x65, 0x7a, 0xae, 0x08, + 0xba, 0x78, 0x25, 0x2e, 0x1c, 0xa6, 0xb4, 0xc6, 0xe8, 0xdd, 0x74, 0x1f, 0x4b, 0xbd, 0x8b, 0x8a, + 0x70, 0x3e, 0xb5, 0x66, 0x48, 0x03, 0xf6, 0x0e, 0x61, 0x35, 0x57, 0xb9, 0x86, 0xc1, 0x1d, 0x9e, + 0xe1, 0xf8, 0x98, 0x11, 0x69, 0xd9, 0x8e, 0x94, 0x9b, 0x1e, 0x87, 0xe9, 0xce, 0x55, 0x28, 0xdf, + 0x8c, 0xa1, 0x89, 0x0d, 0xbf, 0xe6, 0x42, 0x68, 0x41, 0x99, 0x2d, 0x0f, 0xb0, 0x54, 0xbb, 0x16 }; + +static const uint8_t rsbox[256] = +{ 0x52, 0x09, 0x6a, 0xd5, 0x30, 0x36, 0xa5, 0x38, 0xbf, 0x40, 0xa3, 0x9e, 0x81, 0xf3, 0xd7, 0xfb, + 0x7c, 0xe3, 0x39, 0x82, 0x9b, 0x2f, 0xff, 0x87, 0x34, 0x8e, 0x43, 0x44, 0xc4, 0xde, 0xe9, 0xcb, + 0x54, 0x7b, 0x94, 0x32, 0xa6, 0xc2, 0x23, 0x3d, 0xee, 0x4c, 0x95, 0x0b, 0x42, 0xfa, 0xc3, 0x4e, + 0x08, 0x2e, 0xa1, 0x66, 0x28, 0xd9, 0x24, 0xb2, 0x76, 0x5b, 0xa2, 0x49, 0x6d, 0x8b, 0xd1, 0x25, + 0x72, 0xf8, 0xf6, 0x64, 0x86, 0x68, 0x98, 0x16, 0xd4, 0xa4, 0x5c, 0xcc, 0x5d, 0x65, 0xb6, 0x92, + 0x6c, 0x70, 0x48, 0x50, 0xfd, 0xed, 0xb9, 0xda, 0x5e, 0x15, 0x46, 0x57, 0xa7, 0x8d, 0x9d, 0x84, + 0x90, 0xd8, 0xab, 0x00, 0x8c, 0xbc, 0xd3, 0x0a, 0xf7, 0xe4, 0x58, 0x05, 0xb8, 0xb3, 0x45, 0x06, + 0xd0, 0x2c, 0x1e, 0x8f, 0xca, 0x3f, 0x0f, 0x02, 0xc1, 0xaf, 0xbd, 0x03, 0x01, 0x13, 0x8a, 0x6b, + 0x3a, 0x91, 0x11, 0x41, 0x4f, 0x67, 0xdc, 0xea, 0x97, 0xf2, 0xcf, 0xce, 0xf0, 0xb4, 0xe6, 0x73, + 0x96, 0xac, 0x74, 0x22, 0xe7, 0xad, 0x35, 0x85, 0xe2, 0xf9, 0x37, 0xe8, 0x1c, 0x75, 0xdf, 0x6e, + 0x47, 0xf1, 0x1a, 0x71, 0x1d, 0x29, 0xc5, 0x89, 0x6f, 0xb7, 0x62, 0x0e, 0xaa, 0x18, 0xbe, 0x1b, + 0xfc, 0x56, 0x3e, 0x4b, 0xc6, 0xd2, 0x79, 0x20, 0x9a, 0xdb, 0xc0, 0xfe, 0x78, 0xcd, 0x5a, 0xf4, + 0x1f, 0xdd, 0xa8, 0x33, 0x88, 0x07, 0xc7, 0x31, 0xb1, 0x12, 0x10, 0x59, 0x27, 0x80, 0xec, 0x5f, + 0x60, 0x51, 0x7f, 0xa9, 0x19, 0xb5, 0x4a, 0x0d, 0x2d, 0xe5, 0x7a, 0x9f, 0x93, 0xc9, 0x9c, 0xef, + 0xa0, 0xe0, 0x3b, 0x4d, 0xae, 0x2a, 0xf5, 0xb0, 0xc8, 0xeb, 0xbb, 0x3c, 0x83, 0x53, 0x99, 0x61, + 0x17, 0x2b, 0x04, 0x7e, 0xba, 0x77, 0xd6, 0x26, 0xe1, 0x69, 0x14, 0x63, 0x55, 0x21, 0x0c, 0x7d }; + +/* +The round constant word array, rcon[i], contains the values given by +x to th e power (i-1) being powers of x (x is denoted as {02}) in the field GF(2^8) +Note that i starts at 1, not 0). +*/ + +static const uint8_t rcon[255] = { + 0x8d, 0x01, 0x02, 0x04, 0x08, 0x10, 0x20, 0x40, 0x80, 0x1b, 0x36, 0x6c, 0xd8, 0xab, 0x4d, 0x9a, + 0x2f, 0x5e, 0xbc, 0x63, 0xc6, 0x97, 0x35, 0x6a, 0xd4, 0xb3, 0x7d, 0xfa, 0xef, 0xc5, 0x91, 0x39, + 0x72, 0xe4, 0xd3, 0xbd, 0x61, 0xc2, 0x9f, 0x25, 0x4a, 0x94, 0x33, 0x66, 0xcc, 0x83, 0x1d, 0x3a, + 0x74, 0xe8, 0xcb, 0x8d, 0x01, 0x02, 0x04, 0x08, 0x10, 0x20, 0x40, 0x80, 0x1b, 0x36, 0x6c, 0xd8, + 0xab, 0x4d, 0x9a, 0x2f, 0x5e, 0xbc, 0x63, 0xc6, 0x97, 0x35, 0x6a, 0xd4, 0xb3, 0x7d, 0xfa, 0xef, + 0xc5, 0x91, 0x39, 0x72, 0xe4, 0xd3, 0xbd, 0x61, 0xc2, 0x9f, 0x25, 0x4a, 0x94, 0x33, 0x66, 0xcc, + 0x83, 0x1d, 0x3a, 0x74, 0xe8, 0xcb, 0x8d, 0x01, 0x02, 0x04, 0x08, 0x10, 0x20, 0x40, 0x80, 0x1b, + 0x36, 0x6c, 0xd8, 0xab, 0x4d, 0x9a, 0x2f, 0x5e, 0xbc, 0x63, 0xc6, 0x97, 0x35, 0x6a, 0xd4, 0xb3, + 0x7d, 0xfa, 0xef, 0xc5, 0x91, 0x39, 0x72, 0xe4, 0xd3, 0xbd, 0x61, 0xc2, 0x9f, 0x25, 0x4a, 0x94, + 0x33, 0x66, 0xcc, 0x83, 0x1d, 0x3a, 0x74, 0xe8, 0xcb, 0x8d, 0x01, 0x02, 0x04, 0x08, 0x10, 0x20, + 0x40, 0x80, 0x1b, 0x36, 0x6c, 0xd8, 0xab, 0x4d, 0x9a, 0x2f, 0x5e, 0xbc, 0x63, 0xc6, 0x97, 0x35, + 0x6a, 0xd4, 0xb3, 0x7d, 0xfa, 0xef, 0xc5, 0x91, 0x39, 0x72, 0xe4, 0xd3, 0xbd, 0x61, 0xc2, 0x9f, + 0x25, 0x4a, 0x94, 0x33, 0x66, 0xcc, 0x83, 0x1d, 0x3a, 0x74, 0xe8, 0xcb, 0x8d, 0x01, 0x02, 0x04, + 0x08, 0x10, 0x20, 0x40, 0x80, 0x1b, 0x36, 0x6c, 0xd8, 0xab, 0x4d, 0x9a, 0x2f, 0x5e, 0xbc, 0x63, + 0xc6, 0x97, 0x35, 0x6a, 0xd4, 0xb3, 0x7d, 0xfa, 0xef, 0xc5, 0x91, 0x39, 0x72, 0xe4, 0xd3, 0xbd, + 0x61, 0xc2, 0x9f, 0x25, 0x4a, 0x94, 0x33, 0x66, 0xcc, 0x83, 0x1d, 0x3a, 0x74, 0xe8, 0xcb }; + +/* block copying */ + +#define aes_copy_block(output, input) memcpy(output, input, 16) + +static void aes_copy_cbc (uint8_t *data, const uint8_t *input) +{ + uint8_t i; + for (i = 0; i < 16; ++i) + data[i] ^= input[i]; +} + +static void aes_copy_xor (uint8_t *data, const uint8_t *input, const uint8_t *iv) +{ + uint8_t i; + for (i = 0; i < 16; ++i) + data[i] = input[i] ^ iv[i]; +} + +/* key expansion */ + +#define AES_COLUMNS 4 // constant in aes + +static void key_expansion (aes_state *state, const uint8_t *key) +{ + uint32_t i, j, k; + uint8_t t[4]; + uint8_t *keydata, keywords, columns; + + keywords = (uint8_t)(state->keylength >> 2); + keydata = (uint8_t *)state->keyblock; + + /* the first round key is the key itself */ + for(i = 0; i < keywords; ++i) + { + keydata[(i * 4) + 0] = key[(i * 4) + 0]; + keydata[(i * 4) + 1] = key[(i * 4) + 1]; + keydata[(i * 4) + 2] = key[(i * 4) + 2]; + keydata[(i * 4) + 3] = key[(i * 4) + 3]; + } + + /* others derived from the first */ + + for(columns = AES_COLUMNS * (state->rounds + 1); i < columns; ++i) + { + for(j = 0; j < 4; ++j) + t[j] = keydata[(i - 1) * 4 + j]; + if (i % keywords == 0) + { + /* rotate the 4 bytes in a word to the left once; [a0,a1,a2,a3] becomes [a1,a2,a3,a0] */ + k = t[0]; + t[0] = t[1]; + t[1] = t[2]; + t[2] = t[3]; + t[3] = k; + + /* take a four-byte input word and apply the S-box to each of the four bytes to produce an output word */ + t[0] = sbox[t[0]]; + t[1] = sbox[t[1]]; + t[2] = sbox[t[2]]; + t[3] = sbox[t[3]]; + + t[0] = t[0] ^ rcon[i / keywords]; + } + else if (keywords > 6 && i % keywords == 4) + { + t[0] = sbox[t[0]]; + t[1] = sbox[t[1]]; + t[2] = sbox[t[2]]; + t[3] = sbox[t[3]]; + } + keydata[i * 4 + 0] = keydata[(i - keywords) * 4 + 0] ^ t[0]; + keydata[i * 4 + 1] = keydata[(i - keywords) * 4 + 1] ^ t[1]; + keydata[i * 4 + 2] = keydata[(i - keywords) * 4 + 2] ^ t[2]; + keydata[i * 4 + 3] = keydata[(i - keywords) * 4 + 3] ^ t[3]; + } + +} + +/* +An original implementation uses no private buffers except a keyblock. We need private buffers to +keep a CBC vector between calls and to be able to read input data not necessarily in 16-bytes blocks. +Encrypter would actually require only one such buffer, as CBC vector is applied on input data before +the actual cipher procedure. And CBC for the next chunk is simply the output from the previous. +Decrypter, however, applies the cipher first, then applies CBC to the output with a buffered init +vector, and the vector for the next call is the row input before cipher. Hence we need two 16-bytes +buffers for decrypter. +*/ + +/* +aes_state * aes_state_initialize_ecb (aes_state *State, uint8_t *keyblock, const uint8_t *key) +{ + state->flags = 0; + + state->flags |= AES_ECB_MODE; + + if (keyblock == NULL) + { + keyblock = util_malloc(sizeof(aes_keyblock)); + state->flags |= AES_STATE_ALLOC; + } + state->keyblock = keyblock; + key_expansion(state, key); + state->flush = 0; + return state; +} +*/ + +void aes_pdf_mode (aes_state *state) +{ + state->flags |= AES_INLINE_IV; + state->flags &= ~AES_NULL_PADDING; +} + +/* +Initialize arguments: +- state - crypt state +- keyblock - a space for aes key expansion; can be left NULL in which case will be allocated +- key - crypt key; can be left NULL iff keyblock is given and properly initialized +- keylength - the length of the key (16 or 32 bytes) +- iv - 16-bytes CBC initialization vector; + - if left NULL for encoder, one is generated and stored as state->iv + - can also be left NULL for decorer, but then AES_INLINE_IV must be set, as this informs decoder to take + an initialization vector from the beginning of the encrypted stream + +At the first approach, an initialization vector was copied to state block during initialization and encoders +assumed that the state block is the current initialization vector. This simplifies encrypting procedure, +as the output from every 16-bytes chunk encryption is an initialization vector for the next chunk. However, +it makes api usage cumbersome, as the user has to know that iv may need to be copied to state block +before each call. +*/ + +static int aes_key_length (aes_state *state, size_t keylength) +{ + state->keylength = keylength; + switch (keylength) + { + case 16: + state->rounds = 10; + break; + case 24: + state->rounds = 12; + break; + case 32: + state->rounds = 14; + break; + default: + return 0; + } + return 1; +} + +aes_state * aes_encode_initialize (aes_state *state, aes_keyblock *keyblock, const void *key, size_t keylength, const void *iv) +{ + state->flags = 0; + if (!aes_key_length(state, keylength)) + return NULL; + if (iv != NULL) + aes_copy_block(state->iv, iv); + else + aes_generate_iv(state->iv); + state->flags |= AES_HAS_IV; + + if (keyblock == NULL) + { + keyblock = (aes_keyblock *)util_malloc(sizeof(aes_keyblock)); + state->flags |= AES_STATE_ALLOC; + } + state->keyblock = keyblock; + if (key != NULL) /* if NULL we assume keyblock is given and already expanded */ + key_expansion(state, (const uint8_t *)key); + state->flush = 0; + return state; +} + +aes_state * aes_decode_initialize (aes_state *state, aes_keyblock *keyblock, const void *key, size_t keylength, const void *iv) +{ + state->flags = 0; + if (!aes_key_length(state, keylength)) + return NULL; + if (iv != NULL) + { + aes_copy_block(state->iv, iv); + state->flags |= AES_HAS_IV; + } + /* else if AES_INLINE_IV flag is set will be read from input */ + + if (keyblock == NULL) + { + keyblock = (aes_keyblock *)util_malloc(sizeof(aes_keyblock)); + state->flags |= AES_STATE_ALLOC; + } + state->keyblock = keyblock; + if (key != NULL) /* otherwise keyblock is assumed present and properly initialized */ + key_expansion(state, (const uint8_t *)key); + state->flush = 0; + return state; +} + +void aes_state_close (aes_state *state) +{ + if (state->keyblock != NULL && (state->flags & AES_STATE_ALLOC)) + util_free(state->keyblock); +} + +/* add round key */ + +static void aes_round_key (aes_block block, aes_block keyblock) +{ + uint8_t i, j; + for(i = 0; i < 4; ++i) + for(j = 0; j < 4; ++j) + block[i][j] ^= keyblock[i][j]; +} + +#define aes_add_key(block, keyblock, round) aes_round_key(block, (*keyblock)[round]) + +/* substitution */ + +static void aes_encode_sub (aes_block block) +{ + uint8_t i, j, v; + for(i = 0; i < 4; ++i) + for(j = 0; j < 4; ++j) + v = block[i][j], block[i][j] = sbox[v]; +} + +/* rows shift; the row index is the shift offset, the first order is not shifted */ + +static void aes_encode_shift (aes_block block) +{ + uint8_t tmp; + + /* 1st row rotated once */ + tmp = block[0][1]; + block[0][1] = block[1][1]; + block[1][1] = block[2][1]; + block[2][1] = block[3][1]; + block[3][1] = tmp; + + /* 2nd row rotated twice */ + tmp = block[0][2]; + block[0][2] = block[2][2]; + block[2][2] = tmp; + tmp = block[1][2]; + block[1][2] = block[3][2]; + block[3][2] = tmp; + + /* 3rd row rotated 3 times */ + tmp = block[0][3]; + block[0][3] = block[3][3]; + block[3][3] = block[2][3]; + block[2][3] = block[1][3]; + block[1][3] = tmp; +} + +static uint8_t xtime (uint8_t x) +{ + return ((x << 1) ^ (((x >> 7) & 1) * 0x1b)); +} + +/* mix columns */ + +static void aes_encode_mix (aes_block block) +{ + uint8_t i, tmp, tm, t; + + for(i = 0; i < 4; ++i) + { + t = block[i][0]; + tmp = block[i][0] ^ block[i][1] ^ block[i][2] ^ block[i][3] ; + tm = block[i][0] ^ block[i][1]; tm = xtime(tm); block[i][0] ^= tm ^ tmp; + tm = block[i][1] ^ block[i][2]; tm = xtime(tm); block[i][1] ^= tm ^ tmp; + tm = block[i][2] ^ block[i][3]; tm = xtime(tm); block[i][2] ^= tm ^ tmp; + tm = block[i][3] ^ t ; tm = xtime(tm); block[i][3] ^= tm ^ tmp; + } +} + +/* multiply is used to multiply numbers in the field GF(2^8) */ + +#define multiply(x, y) \ + ( ((y & 1) * x) ^ \ + ((y>>1 & 1) * xtime(x)) ^ \ + ((y>>2 & 1) * xtime(xtime(x))) ^ \ + ((y>>3 & 1) * xtime(xtime(xtime(x)))) ^ \ + ((y>>4 & 1) * xtime(xtime(xtime(xtime(x)))))) \ + +/* mix columns */ + +static void aes_decode_mix (aes_block block) +{ + int i; + uint8_t a, b, c, d; + + for(i = 0; i < 4; ++i) + { + a = block[i][0]; + b = block[i][1]; + c = block[i][2]; + d = block[i][3]; + block[i][0] = multiply(a, 0x0e) ^ multiply(b, 0x0b) ^ multiply(c, 0x0d) ^ multiply(d, 0x09); + block[i][1] = multiply(a, 0x09) ^ multiply(b, 0x0e) ^ multiply(c, 0x0b) ^ multiply(d, 0x0d); + block[i][2] = multiply(a, 0x0d) ^ multiply(b, 0x09) ^ multiply(c, 0x0e) ^ multiply(d, 0x0b); + block[i][3] = multiply(a, 0x0b) ^ multiply(b, 0x0d) ^ multiply(c, 0x09) ^ multiply(d, 0x0e); + } +} + +/* inverse substitution */ + +static void aes_decode_sub (aes_block block) +{ + uint8_t i, j, v; + for(i = 0; i < 4; ++i) + for(j = 0; j < 4; ++j) + v = block[i][j], block[i][j] = rsbox[v]; +} + +/* inverse shift rows */ + +static void aes_decode_shift (aes_block block) +{ + uint8_t tmp; + + /* 1st row rotated once right */ + tmp = block[3][1]; + block[3][1] = block[2][1]; + block[2][1] = block[1][1]; + block[1][1] = block[0][1]; + block[0][1] = tmp; + + /* 2st row rotated twice right */ + tmp = block[0][2]; + block[0][2] = block[2][2]; + block[2][2] = tmp; + tmp = block[1][2]; + block[1][2] = block[3][2]; + block[3][2] = tmp; + + /* 3rd row rotated 3 times right */ + tmp = block[0][3]; + block[0][3] = block[1][3]; + block[1][3] = block[2][3]; + block[2][3] = block[3][3]; + block[3][3] = tmp; +} + +/* aes block encoder */ + +static void aes_encode_cipher (aes_state *state) +{ + uint8_t round; + aes_add_key(state->block, state->keyblock, 0); + for (round = 1; round < state->rounds; ++round) + { + aes_encode_sub(state->block); + aes_encode_shift(state->block); + aes_encode_mix(state->block); + aes_add_key(state->block, state->keyblock, round); + } + aes_encode_sub(state->block); + aes_encode_shift(state->block); + aes_add_key(state->block, state->keyblock, state->rounds); +} + +/* aes block decoder */ + +static void aes_decode_cipher (aes_state *state) +{ + uint8_t round; + aes_add_key(state->block, state->keyblock, state->rounds); + for(round = state->rounds - 1; round > 0; --round) + { + aes_decode_shift(state->block); + aes_decode_sub(state->block); + aes_add_key(state->block, state->keyblock, round); + aes_decode_mix(state->block); + } + aes_decode_shift(state->block); + aes_decode_sub(state->block); + aes_add_key(state->block, state->keyblock, 0); +} + +/* tail block padding; RFC 2898, PKCS #5: Password-Based Cryptography Specification Version 2.0; pdf spec p. 119 */ + +#define aes_padding(state) ((state->flags & AES_NULL_PADDING) == 0) + +static void aes_put_padding (aes_state *state, uint8_t length) +{ + uint8_t pad; + pad = (aes_padding(state)) ? 16 - length : 0; + for (; length < 16; ++length) + state->data[length] = state->iv[length] ^ pad; +} + +static int aes_remove_padding (aes_state *state, uint8_t *data, uint8_t *length) +{ + uint8_t pad; + *length = 16; /* block length 16 means leave intact */ + if (aes_padding(state)) + { + pad = data[16 - 1]; + if (pad > 16) + return IOFERR; + for ( ; *length > 16 - pad; --(*length)) + if (data[*length - 1] != pad) + return IOFERR; + } + else + { + for ( ; *length > 0; --(*length)) + if (data[*length - 1] != '\0') + break; + } + return IOFEOF; +} + +/* aes codec */ + +/* make the cipher on input xor-ed with iv, save the output as a new iv, write the output */ +#define aes_encode_output(state, output) \ + (aes_encode_cipher(state), aes_copy_block(state->iv, state->data), aes_copy_block(output, state->data), output += 16) + +iof_status aes_encode_state (iof *I, iof *O, aes_state *state) +{ + int c; + + if (!(state->flags & AES_HAS_IV)) // weird + return IOFERR; + if ((state->flags & AES_INLINE_IV) && !(state->flags & AES_CONTINUE)) + { /* write iv at the beginning of encrypted data */ + if (!iof_ensure(O, 16)) + return IOFFULL; + aes_copy_block(O->pos, state->iv); + O->pos += 16; + state->flags |= AES_CONTINUE; + } + while (iof_ensure(O, 16)) + { + while (state->buffered < 16) + { + if ((c = iof_get(I)) != IOFEOF) + { /* get input byte XORed with iv */ + state->data[state->buffered] = state->iv[state->buffered] ^ ((uint8_t)c); + ++state->buffered; + } + else + { + if (state->flush) + { + if (state->buffered > 0 || aes_padding(state)) + { /* pad the last input chunk; for input divisable by 16, add 16 bytes 0x0f */ + aes_put_padding(state, state->buffered); + state->buffered = 16; + aes_encode_output(state, O->pos); + } + return IOFEOF; + } + else + return IOFEMPTY; + } + } + aes_encode_output(state, O->pos); + state->buffered = 0; + } + return IOFFULL; +} + +/* write iv to the output, save the raw input just buffered as iv for the next chunk, make the cipher, write out xoring with iv */ +#define aes_decode_output(state, output) \ + (aes_copy_block(output, state->iv), aes_copy_block(state->iv, state->data), aes_decode_cipher(state), aes_copy_cbc(output, state->data), output += 16) + +iof_status aes_decode_state (iof *I, iof *O, aes_state *state) +{ + int c, ret; + uint8_t lastlength; + + if ((state->flags & AES_INLINE_IV) && !(state->flags & AES_CONTINUE)) + { + while (state->buffered < 16) + { + if ((c = iof_get(I)) != IOFEOF) + state->iv[state->buffered++] = (uint8_t)c; + else + return state->flush ? IOFERR : IOFEMPTY; + } + state->flags |= AES_CONTINUE|AES_HAS_IV; + state->buffered = 0; + } + while (iof_ensure(O, 16)) + { + while (state->buffered < 16) + { + if ((c = iof_get(I)) != IOFEOF) + state->data[state->buffered++] = (uint8_t)c; + else + return state->flush ? IOFERR : IOFEMPTY; + } + aes_decode_output(state, O->pos); + if (state->flush) + { /* we have to check for EOF here, to remove eventual padding */ + if ((c = iof_get(I)) < 0) + { /* end of input at 16-bytes boundary; remove padding and quit */ + ret = aes_remove_padding(state, O->pos - 16, &lastlength); + O->pos -= 16 - lastlength; + return ret; + } + else + { /* beginning of the next block */ + state->buffered = 1; + state->data[0] = (uint8_t)c; + } + } + else + state->buffered = 0; + } + return IOFFULL; +} + +/* variants that works on c-strings; can work inplace (output==input) except encoder in pdf flavour */ + +/* +Codecs operating on c-string can generally work inplace (output==input), except encoder with AES_INLINE_IV flag set, +which outputs 16 bytes of initialization vector at the beginning of encrypted data. All return the size of encrypted/decrypted +data. Encoders output is the original length padded to a complete 16 bytes (plus eventual 16 bytes of initialization +vector, if AES_INLINE_IV is used). Default padding is unambiguously removed during decryption. AES_NULL_PADDING flag +forces using (ambiguous) NULL-byte padding, only if input length module 16 is greater then zero. + +An input data is supposed to be a complete data to be encrypted or decrypted. It is possible, however, to use those +codecs for scaterred data chunks by manipulating AES_INLINE_IV, AES_NULL_PADDING, AES_CONTINUE flags and data length. +Caller may assume that c-string codecs do not modify state flags. + +Encoder could actually be optimized by writing an initialization vector to a state block once. After every chunk encryption, +the output is the initialization vector for the next chunk. Since we use c-string codec variants on short strings, +the gain is neglectable in comparison with the weight of the aes crypt procedure. +*/ + +size_t aes_encode_data (const void *input, size_t length, void *output, const void *key, size_t keylength, const void *iv, int flags) +{ + aes_state state; + aes_keyblock keyblock; + + if (aes_encode_initialize(&state, &keyblock, key, keylength, iv) == NULL) + return 0; + state.flags |= flags; + return aes_encode_state_data(&state, input, length, output); + // aes_state_close(&state); +} + +size_t aes_encode_state_data (aes_state *state, const void *input, size_t length, void *output) +{ + const uint8_t *inp; + uint8_t *out, tail, t; + size_t size; + + inp = (const uint8_t *)input; + out = (uint8_t *)output; + + if (!(state->flags & AES_HAS_IV)) + return 0; + if ((state->flags & AES_INLINE_IV) && !(state->flags & AES_CONTINUE)) + { + aes_copy_block(out, state->iv); + out += 16; + } + // state->flags |= AES_CONTINUE; // do not modify state flags + + for (size = 0; size + 16 <= length; size += 16) + { + aes_copy_xor(state->data, inp, state->iv); + aes_encode_output(state, out); + inp += 16; + } + + if ((tail = (length % 16)) > 0 || aes_padding(state)) + { + for (t = 0; t < tail; ++t) + state->data[t] = inp[t] ^ state->iv[t]; + aes_put_padding(state, tail); + aes_encode_output(state, out); + size += 16; + } + if (state->flags & AES_INLINE_IV) + size += 16; /* iv written at the beginning of encoded data */ + + return size; +} + +size_t aes_decode_data (const void *input, size_t length, void *output, const void *key, size_t keylength, const void *iv, int flags) +{ + aes_state state; + aes_keyblock keyblock; + + if (aes_decode_initialize(&state, &keyblock, key, keylength, iv) == NULL) + return 0; + state.flags |= flags; + return aes_decode_state_data(&state, input, length, output); + // aes_state_close(&state); +} + +size_t aes_decode_state_data (aes_state *state, const void *input, size_t length, void *output) +{ + const uint8_t *inp; + uint8_t *out, lastlength; + size_t size; + + inp = (const uint8_t *)input; + out = (uint8_t *)output; + + if ((state->flags & AES_INLINE_IV) && !(state->flags & AES_CONTINUE)) + { + aes_copy_block(state->iv, inp); + // state->flags |= AES_HAS_IV; // do not modify state flags + inp += 16; + length = length >= 16 ? length - 16 : 0; + } + else if (!(state->flags & AES_HAS_IV)) + return 0; + // state->flags |= AES_CONTINUE; // do not modify state flags + for (size = 0; size + 16 <= length; size += 16) + { + aes_copy_block(state->data, inp); + aes_decode_output(state, out); + inp += 16; + } + + if (size >= 16) + { + aes_remove_padding(state, out - 16, &lastlength); + size = size - 16 + lastlength; + } + + return size; +} + +/* +pseudo-random bytes chain exceprted from eexec; not expected to have strong cryptographic properties +we only expect that it is (reasonably) unique and different for each call (not only function call, but also +a program call). A current trick with mangling pointer value gives satisfactory results, generally different +for every function call and a programm call. Note that the pseudo-input bytes starts from some inner address +bits, as they vary better; without that, the first byte tends to be "lazy". +*/ + +void random_bytes (uint8_t *output, size_t size) +{ + size_t i; + uint8_t p; + static uint16_t k = 55665; + for (i = 0; i < size; ++i) + { + p = ((uint8_t *)(&output))[(i + 2) % sizeof(uint8_t *)] ^ (uint8_t)size; // pseudo input byte ;) + k = (((p + k) * 52845 + 22719) & 65535); // xor-ed with pseudo-random sequence (kept between calls) + output[i] = p ^ (k >> 8); + } +} + +void aes_generate_iv (uint8_t output[16]) +{ + random_bytes(output, 16); +} + +/* filters */ + +// rc4 decoder function + +static size_t rc4_decoder (iof *F, iof_mode mode) +{ + rc4_state *state; + iof_status status; + size_t tail; + + state = iof_filter_state(rc4_state *, F); + switch(mode) + { + case IOFLOAD: + case IOFREAD: + if (F->flags & IOF_STOPPED) + return 0; + tail = iof_tail(F); + F->pos = F->buf + tail; + F->end = F->buf + F->space; + do { + status = rc4_decode_state(F->next, F, state); + } while (mode == IOFLOAD && status == IOFFULL && iof_resize_buffer(F)); + return iof_decoder_retval(F, "rc4", status); + case IOFCLOSE: + rc4_state_close(state); + iof_free(F); + return 0; + default: + break; + } + return 0; +} + +// rc4 encoder function + +static size_t rc4_encoder (iof *F, iof_mode mode) +{ + rc4_state *state; + iof_status status; + + state = iof_filter_state(rc4_state *, F); + switch (mode) + { + case IOFFLUSH: + state->flush = 1; + // fall through + case IOFWRITE: + F->end = F->pos; + F->pos = F->buf; + status = rc4_encode_state(F, F->next, state); + return iof_encoder_retval(F, "rc4", status); + case IOFCLOSE: + if (!state->flush) + rc4_encoder(F, IOFFLUSH); + rc4_state_close(state); + iof_free(F); + return 0; + default: + break; + } + return 0; +} + +// aes decoder function + +static size_t aes_decoder (iof *F, iof_mode mode) +{ + aes_state *state; + iof_status status; + size_t tail; + + state = iof_filter_state(aes_state *, F); + switch(mode) + { + case IOFLOAD: + case IOFREAD: + if (F->flags & IOF_STOPPED) + return 0; + tail = iof_tail(F); + F->pos = F->buf + tail; + F->end = F->buf + F->space; + do { + status = aes_decode_state(F->next, F, state); + } while (mode == IOFLOAD && status == IOFFULL && iof_resize_buffer(F)); + return iof_decoder_retval(F, "aes", status); + case IOFCLOSE: + aes_state_close(state); + iof_free(F); + return 0; + default: + break; + } + return 0; +} + +// aes encoder function + +static size_t aes_encoder (iof *F, iof_mode mode) +{ + aes_state *state; + iof_status status; + + state = iof_filter_state(aes_state *, F); + switch (mode) + { + case IOFFLUSH: + state->flush = 1; + // fall through + case IOFWRITE: + F->end = F->pos; + F->pos = F->buf; + status = aes_encode_state(F, F->next, state); + return iof_encoder_retval(F, "aes", status); + case IOFCLOSE: + if (!state->flush) + aes_encoder(F, IOFFLUSH); + aes_state_close(state); + iof_free(F); + return 0; + default: + break; + } + return 0; +} + +iof * iof_filter_rc4_decoder (iof *N, const void *key, size_t keylength) +{ + iof *I; + rc4_state *state; + + I = iof_filter_reader(rc4_decoder, sizeof(rc4_state), &state); + iof_setup_next(I, N); + if (rc4_state_init(state, key, keylength) == NULL) + { + iof_discard(I); + return NULL; + } + state->flush = 1; + return I; +} + +iof * iof_filter_rc4_encoder (iof *N, const void *key, size_t keylength) +{ + iof *O; + rc4_state *state; + + O = iof_filter_writer(rc4_encoder, sizeof(rc4_state), &state); + iof_setup_next(O, N); + if (rc4_state_init(state, key, keylength) == NULL) + { + iof_discard(O); + return NULL; + } + // state->flush = 1; + return O; +} + +/* aes crypt filters */ + +iof * iof_filter_aes_decoder (iof *N, const void *key, size_t keylength) +{ + iof *I; + aes_state *state; + + I = iof_filter_reader(aes_decoder, sizeof(aes_state), &state); + iof_setup_next(I, N); + if (aes_decode_init(state, key, keylength) == NULL) + { + iof_discard(I); + return NULL; + } + aes_pdf_mode(state); + state->flush = 1; + return I; +} + +iof * iof_filter_aes_encoder (iof *N, const void *key, size_t keylength) +{ + iof *O; + aes_state *state; + + O = iof_filter_writer(aes_encoder, sizeof(aes_state), &state); + iof_setup_next(O, N); + if (aes_encode_init(state, key, keylength) == NULL) + { + iof_discard(O); + return NULL; + } + aes_pdf_mode(state); + // state->flush = 1; + return O; +} + +/* test */ + +/* +static void show (void *p, size_t size, uint8_t round, uint8_t sym) +{ + uint8_t i; + printf("%c%c:", round, sym); + for (i = 0; i < size; ++i) + printf("%02x", ((uint8_t *)p)[i]); + printf("\n"); +} + +void aes_test (void) +{ + const uint8_t key[] = { 0x2b, 0x7e, 0x15, 0x16, 0x28, 0xae, 0xd2, 0xa6, 0xab, 0xf7, 0x15, 0x88, 0x09, 0xcf, 0x4f, 0x3c }; + const uint8_t iv[] = { 0x00, 0x01, 0x02, 0x03, 0x04, 0x05, 0x06, 0x07, 0x08, 0x09, 0x0a, 0x0b, 0x0c, 0x0d, 0x0e, 0x0f }; + const uint8_t inp[] = { + 0x6b, 0xc1, 0xbe, 0xe2, 0x2e, 0x40, 0x9f, 0x96, 0xe9, 0x3d, 0x7e, 0x11, 0x73, 0x93, 0x17, 0x2a, + 0xae, 0x2d, 0x8a, 0x57, 0x1e, 0x03, 0xac, 0x9c, 0x9e, 0xb7, 0x6f, 0xac, 0x45, 0xaf, 0x8e, 0x51, + 0x30, 0xc8, 0x1c, 0x46, 0xa3, 0x5c, 0xe4, 0x11, 0xe5, 0xfb, 0xc1, 0x19, 0x1a, 0x0a, 0x52, 0xef, + 0xf6, 0x9f, 0x24, 0x45, 0xdf, 0x4f, 0x9b, 0x17, 0xad, 0x2b, 0x41, 0x7b, 0xe6, 0x6c, 0x37, 0x10 }; + const uint8_t out[] = { + 0x76, 0x49, 0xab, 0xac, 0x81, 0x19, 0xb2, 0x46, 0xce, 0xe9, 0x8e, 0x9b, 0x12, 0xe9, 0x19, 0x7d, + 0x50, 0x86, 0xcb, 0x9b, 0x50, 0x72, 0x19, 0xee, 0x95, 0xdb, 0x11, 0x3a, 0x91, 0x76, 0x78, 0xb2, + 0x73, 0xbe, 0xd6, 0xb8, 0xe3, 0xc1, 0x74, 0x3b, 0x71, 0x16, 0xe6, 0x9e, 0x22, 0x22, 0x95, 0x16, + 0x3f, 0xf1, 0xca, 0xa1, 0x68, 0x1f, 0xac, 0x09, 0x12, 0x0e, 0xca, 0x30, 0x75, 0x86, 0xe1, 0xa7 }; + + uint8_t input[64], output[64]; + size_t inpsize, outsize; + int flags = AES_NULL_PADDING; + + //////////////////////////////////////////////////////////////////////////// + +//#define ENCODETO output +#define ENCODETO input // inplace + + inpsize = 64; + memcpy(input, inp, inpsize); + show(input, inpsize, '>', '>'); + outsize = aes_encode_data(input, inpsize, ENCODETO, key, 16, iv, flags); + show(ENCODETO, outsize, '<', '<'); + if (outsize == inpsize && memcmp(ENCODETO, out, outsize) == 0) + printf("ENCODER SUCCESS\n"); + else + printf("ENCODER FAILURE\n"); + + //////////////////////////////////////////////////////////////////////////// + +//#define DECODETO input +#define DECODETO output // in place + + outsize = 64; + memcpy(output, out, outsize); + show(output, outsize, '<', '<'); + inpsize = aes_decode_data(output, outsize, DECODETO, key, 16, iv, flags); + show(DECODETO, inpsize, '>', '>'); + if (inpsize == outsize && memcmp(DECODETO, inp, inpsize) == 0) + printf("DECODER SUCCESS\n"); + else + printf("DECODER FAILURE\n"); +} +*/ + +/* +Some example vectors + +================================ AES ECB 128-bit encryption mode ================================ + +Encryption key: 2b7e151628aed2a6abf7158809cf4f3c + +Test vector Cipher text +6bc1bee22e409f96e93d7e117393172a 3ad77bb40d7a3660a89ecaf32466ef97 +ae2d8a571e03ac9c9eb76fac45af8e51 f5d3d58503b9699de785895a96fdbaaf +30c81c46a35ce411e5fbc1191a0a52ef 43b1cd7f598ece23881b00e3ed030688 +f69f2445df4f9b17ad2b417be66c3710 7b0c785e27e8ad3f8223207104725dd4 + + +================================ AES ECB 192-bit encryption mode ================================ + +Encryption key: 8e73b0f7da0e6452c810f32b809079e562f8ead2522c6b7b + +Test vector Cipher text +6bc1bee22e409f96e93d7e117393172a bd334f1d6e45f25ff712a214571fa5cc +ae2d8a571e03ac9c9eb76fac45af8e51 974104846d0ad3ad7734ecb3ecee4eef +30c81c46a35ce411e5fbc1191a0a52ef ef7afd2270e2e60adce0ba2face6444e +f69f2445df4f9b17ad2b417be66c3710 9a4b41ba738d6c72fb16691603c18e0e + + +================================ AES ECB 256-bit encryption mode ================================ + +Encryption key: 603deb1015ca71be2b73aef0857d77811f352c073b6108d72d9810a30914dff4 + +Test vector Cipher text +6bc1bee22e409f96e93d7e117393172a f3eed1bdb5d2a03c064b5a7e3db181f8 +ae2d8a571e03ac9c9eb76fac45af8e51 591ccb10d410ed26dc5ba74a31362870 +30c81c46a35ce411e5fbc1191a0a52ef b6ed21b99ca6f4f9f153e7b1beafed1d +f69f2445df4f9b17ad2b417be66c3710 23304b7a39f9f3ff067d8d8f9e24ecc7 + +================================ AES CBC 128-bit encryption mode ================================ + +Encryption key: 2b7e151628aed2a6abf7158809cf4f3c + +Initialization vector Test vector Cipher text +000102030405060708090A0B0C0D0E0F 6bc1bee22e409f96e93d7e117393172a 7649abac8119b246cee98e9b12e9197d +7649ABAC8119B246CEE98E9B12E9197D ae2d8a571e03ac9c9eb76fac45af8e51 5086cb9b507219ee95db113a917678b2 +5086CB9B507219EE95DB113A917678B2 30c81c46a35ce411e5fbc1191a0a52ef 73bed6b8e3c1743b7116e69e22229516 +73BED6B8E3C1743B7116E69E22229516 f69f2445df4f9b17ad2b417be66c3710 3ff1caa1681fac09120eca307586e1a7 + +================================ AES CBC 192-bit encryption mode ================================ + +Encryption key: 8e73b0f7da0e6452c810f32b809079e562f8ead2522c6b7b + +Initialization vector Test vector Cipher text +000102030405060708090A0B0C0D0E0F 6bc1bee22e409f96e93d7e117393172a 4f021db243bc633d7178183a9fa071e8 +4F021DB243BC633D7178183A9FA071E8 ae2d8a571e03ac9c9eb76fac45af8e51 b4d9ada9ad7dedf4e5e738763f69145a +B4D9ADA9AD7DEDF4E5E738763F69145A 30c81c46a35ce411e5fbc1191a0a52ef 571b242012fb7ae07fa9baac3df102e0 +571B242012FB7AE07FA9BAAC3DF102E0 f69f2445df4f9b17ad2b417be66c3710 08b0e27988598881d920a9e64f5615cd + +================================ AES CBC 256-bit encryption mode ================================ + +Encryption key: 603deb1015ca71be2b73aef0857d77811f352c073b6108d72d9810a30914dff4 + +Initialization vector Test vector Cipher text +000102030405060708090A0B0C0D0E0F 6bc1bee22e409f96e93d7e117393172a f58c4c04d6e5f1ba779eabfb5f7bfbd6 +F58C4C04D6E5F1BA779EABFB5F7BFBD6 ae2d8a571e03ac9c9eb76fac45af8e51 9cfc4e967edb808d679f777bc6702c7d +9CFC4E967EDB808D679F777BC6702C7D 30c81c46a35ce411e5fbc1191a0a52ef 39f23369a9d9bacfa530e26304231461 +39F23369A9D9BACFA530E26304231461 f69f2445df4f9b17ad2b417be66c3710 b2eb05e2c39be9fcda6c19078c6a9d1b +*/
\ No newline at end of file diff --git a/Build/source/texk/web2c/luatexdir/luapplib/util/utilcrypt.h b/Build/source/texk/web2c/luatexdir/luapplib/util/utilcrypt.h new file mode 100644 index 00000000000..e5bf53cc5ce --- /dev/null +++ b/Build/source/texk/web2c/luatexdir/luapplib/util/utilcrypt.h @@ -0,0 +1,90 @@ +#ifndef UTIL_CRYPT_H +#define UTIL_CRYPT_H + +#include <stdint.h> +#include <stddef.h> +#include "utiliof.h" + +#ifndef UTIL_CRYPT_TIME +# define UTIL_CRYPT_TIME 0 +#endif + +/* RC4 */ + +typedef uint8_t rc4_map[256]; + +typedef struct rc4_state rc4_state; + +#define RC4_STATE_ALLOC (1<<0) + +UTILAPI rc4_state * rc4_state_initialize (rc4_state *state, rc4_map *map, const void *vkey, size_t keylength); +#define rc4_state_init(state, vkey, keylength) rc4_state_initialize(state, NULL, vkey, keylength) +UTILAPI void rc4_map_save (rc4_state *state, rc4_map *map); +UTILAPI void rc4_map_restore (rc4_state *state, rc4_map *map); + +/* Codecs operating on iof */ + +UTILAPI iof_status rc4_crypt_state (iof *I, iof *O, rc4_state *state); +#define rc4_encode_state(I, O, state) rc4_crypt_state(I, O, state) +#define rc4_decode_state(I, O, state) rc4_crypt_state(I, O, state) + +UTILAPI iof_status rc4_crypt (iof *I, iof *O, const void *key, size_t length); +#define rc4_encode(I, O) rc4_crypt(I, O, key, length) +#define rc4_decode(I, O) rc4_crypt(I, O, key, length) + +UTILAPI size_t rc4_crypt_data (const void *input, size_t length, void *output, const void *key, size_t keylength); +UTILAPI size_t rc4_crypt_state_data (rc4_state *state, const void *input, size_t length, void *output); +#define rc4_encode_data(input, length, output, key, keylength) rc4_crypt_data(input, length, output, key, keylength) +#define rc4_decode_data(input, length, output, key, keylength) rc4_crypt_data(input, length, output, key, keylength) +#define rc4_encode_state_data(state, input, length, output) rc4_crypt_state_data(state, input, length, output) +#define rc4_decode_state_data(state, input, length, output) rc4_crypt_state_data(state, input, length, output) + +UTILAPI void rc4_state_close (rc4_state *state); + +/* AES */ + +typedef uint8_t aes_block[4][4]; +typedef aes_block aes_keyblock[15]; // aes128 - 10+1, aes192 - 12+1, aes256 - 14+1 + +typedef struct aes_state aes_state; + +#define AES_STATE_ALLOC (1<<0) +//#define AES_ECB_MODE (1<<2) +#define AES_HAS_IV (1<<3) +#define AES_INLINE_IV (1<<4) +#define AES_CONTINUE (1<<5) +#define AES_NULL_PADDING (1<<6) + +UTILAPI void aes_pdf_mode (aes_state *state); +//UTILAPI aes_state * aes_state_initialize_ecb (aes_state *State, uint8_t *roundkey, const uint8_t *key); +UTILAPI aes_state * aes_encode_initialize (aes_state *state, aes_keyblock *keyblock, const void *key, size_t keylength, const void *iv); +UTILAPI aes_state * aes_decode_initialize (aes_state *state, aes_keyblock *keyblock, const void *key, size_t keylength, const void *iv); +#define aes_encode_init(state, key, keylength) aes_encode_initialize(state, NULL, key, keylength, NULL) +#define aes_decode_init(state, key, keylength) aes_decode_initialize(state, NULL, key, keylength, NULL) + +UTILAPI void aes_state_close (aes_state *state); + +/* Codecs operating on iof */ + +UTILAPI iof_status aes_encode_state (iof *I, iof *O, aes_state *state); +UTILAPI iof_status aes_decode_state (iof *I, iof *O, aes_state *state); + +UTILAPI size_t aes_encode_data (const void *input, size_t length, void *output, const void *key, size_t keylength, const void *iv, int flags); +UTILAPI size_t aes_encode_state_data (aes_state *state, const void *input, size_t length, void *output); +UTILAPI size_t aes_decode_data (const void *input, size_t length, void *output, const void *key, size_t keylength, const void *iv, int flags); +UTILAPI size_t aes_decode_state_data (aes_state *state, const void *input, size_t length, void *output); + +/* random bytes generator */ + +UTILAPI void random_bytes (uint8_t *output, size_t size); +UTILAPI void aes_generate_iv (uint8_t output[16]); + +/* filters */ + +iof * iof_filter_rc4_decoder (iof *N, const void *key, size_t keylength); +iof * iof_filter_rc4_encoder (iof *N, const void *key, size_t keylength); + +iof * iof_filter_aes_decoder (iof *N, const void *key, size_t keylength); +iof * iof_filter_aes_encoder (iof *N, const void *key, size_t keylength); + +#endif
\ No newline at end of file diff --git a/Build/source/texk/web2c/luatexdir/luapplib/util/utilcryptdef.h b/Build/source/texk/web2c/luatexdir/luapplib/util/utilcryptdef.h new file mode 100644 index 00000000000..02d70aa5ef6 --- /dev/null +++ b/Build/source/texk/web2c/luatexdir/luapplib/util/utilcryptdef.h @@ -0,0 +1,30 @@ + +#ifndef UTIL_CRYPTDEF_H +#define UTIL_CRYPTDEF_H + +struct rc4_state { + union { + rc4_map *map; + uint8_t *smap; + }; + int i, j; + int flush; + int flags; +}; + +struct aes_state { + size_t keylength; + int rounds; + //int keywords; + union { + aes_block block; + uint8_t data[16]; + }; + aes_keyblock *keyblock; + uint8_t iv[16]; + uint8_t buffered; + int flush; + int flags; +}; + +#endif
\ No newline at end of file diff --git a/Build/source/texk/web2c/luatexdir/luapplib/util/utildecl.h b/Build/source/texk/web2c/luatexdir/luapplib/util/utildecl.h new file mode 100644 index 00000000000..94ef73413eb --- /dev/null +++ b/Build/source/texk/web2c/luatexdir/luapplib/util/utildecl.h @@ -0,0 +1,28 @@ + +#ifndef UTIL_DECL_H +#define UTIL_DECL_H + +/* +UTILDLL - when building .dll +UTILEXE - when building .exe to import symbols from .dll +*/ + +#if defined (_WIN32) || defined(_WIN64) +# ifdef UTILDLL +# define UTILAPI __declspec(dllexport) +# define UTILDEF __declspec(dllexport) +# else +# ifdef UTILEXE +# define UTILAPI __declspec(dllimport) +# define UTILDEF +# else +# define UTILAPI +# define UTILDEF +# endif +# endif +#else +# define UTILAPI +# define UTILDEF +#endif + +#endif
\ No newline at end of file diff --git a/Build/source/texk/web2c/luatexdir/luapplib/util/utilflate.c b/Build/source/texk/web2c/luatexdir/luapplib/util/utilflate.c new file mode 100644 index 00000000000..8ecc240c9dc --- /dev/null +++ b/Build/source/texk/web2c/luatexdir/luapplib/util/utilflate.c @@ -0,0 +1,319 @@ + +#include "utilmem.h" +#include "utillog.h" +#include "utilflate.h" +#include <zlib.h> + +/* flate codec */ + +/* +Flate codec example provided at http://www.zlib.net/zpipe.c (http://www.zlib.net/zlib_how.html) uses the following scheme: +- provide input data buffer +- keep providing output until codec function uses it + +For encoder: + + z->zalloc = z->zfree = z->zopaque = NULL; + deflateInit(z, compression_level); + do { + z->next_in = <input buffer> + z->avail_in = <input buffer bytes> + do { + z->next_out = <output buffer> + z->avail_out = <output buffer bytes> + deflate(z, flush); + // write obtained output from deflate + } while (z->avail_out == 0); + assert(z->avail_in == 0); + } while (flush != Z_FINISH); + deflateEnd(z); + +'z' is an internal codec state of type z_stream, 'flush' is either Z_NO_FLUSH or Z_FINISH at the end of data. +deflate() ensures to consume the entire input if there are no obstackles to write an output. The inner loop +provides an output space as long as it is used by deflate(). When deflate() wrote everything it could, +it leaves z->avail_out > 0, which breaks the inner loop. At this point z->avail_in should also be zero. +The example documentation claims that the return codes from deflate() doesn't really need to be checked, +as checking z->avail_out for zero is enough. + +The scheme for decoder is pretty similar, but with substantial differences: +- the end of stream is automatically found by decoder, so using Z_FINISH flag to indicate an end of stream + is not necessary, but if provided, it MUST be given only if the EOF marker actually occurs in the input chunk, + and subsequent calls to inflate() must consequently use Z_FINISH +- calling inflate() as long as it uses the output buffer provided still works for decoder, but inflate() + does not ensure to consume the entire input, as it will read until end of stream marker +- the return code from inflate() must be checked to ensure the proper reaction on invalid data stream and + end of stream signals +- initialization must set an input buffer to NULL or to some existing chunk (the later helps zlib to perform + better on inflate(), but inflate() does the research on the first call anyway) + + z->zalloc = z->zfree = z->zopaque = NULL; + z->next_in = NULL, z->avail_in = 0; + inflateInit(z); + do { + z->next_in = <input buffer> + z->avail_in = <input buffer bytes> + do { + z->next_out = <output buffer> + z->avail_out = <output buffer bytes> + status = inflate(z, flush); + // check return status + // write obtained output from inflate + } while (z->avail_out == 0); + } while (status != Z_STREAM_END); + inflateEnd(z); + +Our wrapper generally follows "prepare input, keep pomping output" scheme, but we need to support handler function +breaks on IOFEMPTY and IOFFULL. For a consistent come back from those on subsequent calls to the handler function, +we use 3 states: +- FLATE_IN - get input, when got something then goto FALTE_OUT +- FLATE_OUT - set z_stream buffers and keep writing output until enything to write, then goto FLATE_IN or FLATE_DONE +- FLATE_DONE - we are done, no return from that state +Distinction of FLATE_IN and FLATE_OUT states guarantees that we will not get more input until zlib consumes the stuff +from the previous feed, possibly interrupted by IOFFULL return on filling the output buffer. This distinction is not +critical, but makes the filter running according to the scheme described above. Note that we set zlib input buffer +(z->next_in, z->avail_in) at the beginning of FLATE_OUT state. Also note that we always update our buffers according +to updated avail_in / avail_out values, just after a call to inflate() / deflate(). So no matter what have happens +between handler calls, zlib input buffer is in sync with ours. +*/ + +struct flate_state { + z_stream z; + int flush; + int status; + int level; /* encoder compression level -1..9 */ +}; + +enum { + FLATE_IN, + FLATE_OUT, + FLATE_DONE +}; + +flate_state * flate_decoder_init (flate_state *state) +{ /* initialize zlib */ + z_stream *z = &state->z; + z->zalloc = Z_NULL; + z->zfree = Z_NULL; + z->opaque = Z_NULL; + z->avail_in = 0; /* must be initialized before inflateInit() */ + z->next_in = Z_NULL; /* ditto */ + if (inflateInit(z) != Z_OK) + return NULL; + state->status = FLATE_IN; + return state; +} + +flate_state * flate_encoder_init (flate_state *state) +{ + z_stream *z = &state->z; + z->zalloc = Z_NULL; + z->zfree = Z_NULL; + z->opaque = Z_NULL; + z->avail_in = 0; + z->next_in = Z_NULL; + state->level = Z_DEFAULT_COMPRESSION; // will probably be moved upward + if (deflateInit(z, state->level) != Z_OK) + return NULL; + state->status = FLATE_IN; + return state; +} + +static const char * zmess (int zstatus) +{ + switch (zstatus) + { + case Z_OK: return "ok"; + case Z_STREAM_END: return "end of stream"; + case Z_BUF_ERROR: return "buffer error"; + case Z_STREAM_ERROR: return "stream error"; + case Z_NEED_DICT: return "need dict"; + case Z_DATA_ERROR: return "data error"; + case Z_MEM_ERROR: return "memory error"; + case Z_VERSION_ERROR: return "version error"; + case Z_ERRNO: return "io error"; + default: + break; + } + return "unknown error"; +} + +iof_status flate_decode_state (iof *I, iof *O, flate_state *state) +{ + z_stream *z; + int zstatus = Z_OK; + z = &state->z; + while (state->status != FLATE_DONE) + { + if (state->status == FLATE_IN) + { + if (!iof_readable(I)) + return state->flush ? IOFERR : IOFEMPTY; + state->status = FLATE_OUT; + } + z->next_in = (Bytef *)I->pos; + z->avail_in = (uInt)iof_left(I); + do { + if (!iof_writable(O)) + return IOFFULL; + z->next_out = (Bytef *)O->pos; + z->avail_out = (uInt)iof_left(O); + zstatus = inflate(z, Z_NO_FLUSH); + I->pos += iof_left(I) - z->avail_in; + O->pos += iof_left(O) - z->avail_out; + switch (zstatus) + { + case Z_OK: + case Z_STREAM_END: + break; + default: + loggerf("flate decoder %s (%d)", zmess(zstatus), zstatus); + return IOFERR; + } + } while (z->avail_out == 0); + state->status = zstatus == Z_STREAM_END ? FLATE_DONE : FLATE_IN; + } + return IOFEOF; +} + +iof_status flate_encode_state (iof *I, iof *O, flate_state *state) +{ + z_stream *z; + int zstatus; + z = &state->z; + while (state->status != FLATE_DONE) + { + if (state->status == FLATE_IN) + { + if (!iof_readable(I)) + if (!state->flush) + return IOFEMPTY; + state->status = FLATE_OUT; + } + z->next_in = (Bytef *)I->pos; + z->avail_in = (uInt)iof_left(I); + do { + if (!iof_writable(O)) + return IOFFULL; + z->next_out = (Bytef *)O->pos; + z->avail_out = (uInt)iof_left(O); + zstatus = deflate(z, state->flush ? Z_FINISH : Z_NO_FLUSH); + I->pos += iof_left(I) - z->avail_in; + O->pos += iof_left(O) - z->avail_out; + switch (zstatus) + { + case Z_OK: + case Z_STREAM_END: + break; + default: + loggerf("flate encoder %s (%d)", zmess(zstatus), zstatus); + return IOFERR; + } + } while (z->avail_out == 0); + state->status = state->flush ? FLATE_DONE : FLATE_IN; + } + return IOFEOF; +} + + +void flate_decoder_close (flate_state *state) +{ + inflateEnd(&state->z); +} + +void flate_encoder_close (flate_state *state) +{ + deflateEnd(&state->z); +} + +/* filter */ + +// flate decoder function + +static size_t flate_decoder (iof *F, iof_mode mode) +{ + flate_state *state; + iof_status status; + size_t tail; + + state = iof_filter_state(flate_state *, F); + switch(mode) + { + case IOFLOAD: + case IOFREAD: + if (F->flags & IOF_STOPPED) + return 0; + tail = iof_tail(F); + F->pos = F->buf + tail; + F->end = F->buf + F->space; + do { + status = flate_decode_state(F->next, F, state); + } while (mode == IOFLOAD && status == IOFFULL && iof_resize_buffer(F)); + return iof_decoder_retval(F, "flate", status); + case IOFCLOSE: + flate_decoder_close(state); + iof_free(F); + return 0; + default: + break; + } + return 0; +} + +// flate encoder function + +static size_t flate_encoder (iof *F, iof_mode mode) +{ + flate_state *state; + iof_status status; + + state = iof_filter_state(flate_state *, F); + switch (mode) + { + case IOFFLUSH: + state->flush = 1; + // fall through + case IOFWRITE: + F->end = F->pos; + F->pos = F->buf; + status = flate_encode_state(F, F->next, state); + return iof_encoder_retval(F, "flate", status); + case IOFCLOSE: + if (!state->flush) + flate_encoder(F, IOFFLUSH); + flate_encoder_close(state); + iof_free(F); + return 0; + default: + break; + } + return 0; +} + +iof * iof_filter_flate_decoder (iof *N) +{ + iof *I; + flate_state *state; + I = iof_filter_reader(flate_decoder, sizeof(flate_state), &state); + iof_setup_next(I, N); + if (flate_decoder_init(state) == NULL) + { + iof_discard(I); + return NULL; + } + state->flush = 1; + return I; +} + +iof * iof_filter_flate_encoder (iof *N) +{ + iof *O; + flate_state *state; + O = iof_filter_writer(flate_encoder, sizeof(flate_state), &state); + iof_setup_next(O, N); + if (flate_encoder_init(state) == NULL) + { + iof_discard(O); + return NULL; + } + return O; +} diff --git a/Build/source/texk/web2c/luatexdir/luapplib/util/utilflate.h b/Build/source/texk/web2c/luatexdir/luapplib/util/utilflate.h new file mode 100644 index 00000000000..09bdd666112 --- /dev/null +++ b/Build/source/texk/web2c/luatexdir/luapplib/util/utilflate.h @@ -0,0 +1,21 @@ +#ifndef UTIL_FLATE_H +#define UTIL_FLATE_H + +#include "utiliof.h" + +typedef struct flate_state flate_state; + +flate_state * flate_decoder_init (flate_state *state); +flate_state * flate_encoder_init (flate_state *state); + +iof_status flate_decode_state (iof *I, iof *O, flate_state *state); +iof_status flate_encode_state (iof *I, iof *O, flate_state *state); + +void flate_decoder_close (flate_state *state); +void flate_encoder_close (flate_state *state); + +iof * iof_filter_flate_decoder (iof *N); +iof * iof_filter_flate_encoder (iof *N); + + +#endif
\ No newline at end of file diff --git a/Build/source/texk/web2c/luatexdir/luapplib/util/utilfpred.c b/Build/source/texk/web2c/luatexdir/luapplib/util/utilfpred.c new file mode 100644 index 00000000000..ce9c129e249 --- /dev/null +++ b/Build/source/texk/web2c/luatexdir/luapplib/util/utilfpred.c @@ -0,0 +1,783 @@ +/* predictor filters; common for flate and lzw */ + +#include "utilmem.h" +#include "utillog.h" +#include "utilfpred.h" + +/* +Here we implement predictor filters used with flate and lzw compressions in PDF streams. The main idea of data prediction +is to compute and output the differences between data records instead of this records. Adjacent pixels in images are usually +similar, so differences between pixel values tends to be zero. And both Flate and LZW performs better when the input +is rather smooth. Although a preliminary use of predictors is related to bitmap data, The actual need for predictor filter +came from the fact that xref streams may also be predicted (usually with PNG up-predictor). + +PDF specification allows to use several predictor algorithms, specified by /Predictor key in /DecodeParms dictionary: + + 1 - no predictor (default) + 2 - TIFF horizontal predictor + 10 - PNG none predictor + 11 - PNG sub predictor + 12 - PNG up predictor + 13 - PNG average predictor + 14 - PNG paeth predictor + +All PNG predictors works on bytes, regardless the image color-depth. While encoding, every input data byte is decreased +by the appropriate byte of the previous pixel. Even if the pixel does not fit a full byte, PNG predictors use an artificial +pixel size rounded up to a full byte. PNG predictors utilizes previous (left) pixel, pixel above and previous to above +pixel. In case of PNG, the type of the predictor is written on a dedicated byte on the beginning of every scanline. It +means all predictor functions must maintain and information about left, above and left-above pixels. + +Despite the same differencing idea, TIFF predictors are different. The prediction process bases on pixel components, +which are not necessarily bytes (component of a pixel is added/substracted from a relevant component of a previous +pixel). In TIFF predictor 2, only the previous (the left) pixel is taken into account, there is no need to keep +an information about other surrounding pixels. Also there is no expicit algorithm marker in data; the same prediction +method is applied to all input rows. + +Not surprisingly, predictor encoders and decoders are pretty similar. Encoders take some input value and the previous +input value (or 0 at the beginning of the scanline) and output a difference between them. Decoders takes an input value, +previously decoded value (or zero) and outputs their sum. When encoding, the result is cast to the proper unsigned integer, +when decoding, modulo 256 (or appropriate) is used, which makes encoding and decoding looseless. + +Some extra bits trickery is involved in TIFF predictor function, when components doesn't fit bytes boundary. In that case, +an input is treated as an bits stream. Every input byte is "buffered" in a larger integer, as its lower bits (from right). +Every output value is taken from its higher (left) bits. In a special case of bits-per-component equal 1, we buffer all +pixel bits and use XOR to compute bits difference between pixels. I've excerpted that trick from poppler, but I'm not +really sure if it works any better, especially when the number of components per pixel is 1. In that case we do a hard +bit-by-bit work anyway. + +Predictor codecs state keeps a notion of surrounding pixels. PNG predictors uses left, up and upleft +pixel data, while TIFF predictor (2) only needs the previous (left) pixel. Important to note that PNG +predictors always work on bytes, no matter of color-depth (bits per component), while TIFF predictor +works on pixel components, which not necessarily fits into a complete byte. However, for PNG predictor +the size of a pixel still matters, because 'left' and 'upleft' refers to a corresponding pixel byte, +not necessarily previous byte. + +In PNG prediction, we record every pixel byte (in decoded form) in state->rowsave. At the end of a scanline +we copy state->rowsave to state->rowup, so that in the next scanline we can access up-pixel byte. +Left pixel byte is accessed as state->rowsave (the byte recently stored or virtual left edge byte \0). +Up-left pixel byte is accessed via state->rowup, but with state->pixelsize offset (same as left byte, possibly \0 +at the left edge of the row). Both state->rowup and state->rowsave has a safe span of pixelsize bytes on the left, +that are permanently \0. +*/ + +#define predictor_component_t unsigned short +#define predictor_pixel1b_t unsigned int + +typedef struct predictor_state { + int default_predictor; /* default predictor indicator */ + int current_predictor; /* current predictor, possibly taken from algorithm marker in PNG data */ + int rowsamples; /* number of pixels in a scanline (/DecodeParms << /Columns ... >>) */ + int compbits; /* number of bits per component (/DecodeParms << /BitsPerComponent ... >>) */ + int components; /* number of components (/DecodeParms << /Colors ... >>) */ + uint8_t *buffer; /* temporary private buffer area */ + uint8_t *rowin; /* an input row buffer position */ + int rowsize; /* size of a current scanline in bytes (rounded up) */ + int rowend; /* an input buffer end position */ + int rowindex; /* an output buffer position */ + union { + struct { /* used by PNG predictor codecs */ + uint8_t *rowup, *rowsave; /* previous scanline buffers */ + int predictorbyte; /* flag indicating that algorithm byte is read/written */ + int pixelsize; /* number of bytes per pixel (rounded up) */ + }; + struct { /* used by TIFF predictor codecs */ + union { + predictor_component_t *prevcomp; /* an array of left pixel components */ + predictor_pixel1b_t *prevpixel; /* left pixel value stored on a single integer (for 1bit color-depth) */ + }; + int compin, compout; /* bit stream buffers */ + int bitsin, bitsout; /* bit stream counters */ + int sampleindex; /* pixel counter */ + int compindex; /* component counter */ + int pixbufsize; /* size of pixel buffer in bytes */ + }; + }; + int flush; + int status; +} predictor_state; + +enum { + STATUS_LAST = 0, + STATUS_CONTINUE = 1 // any value different then IOFEOF, IOFERR, ... +}; + +predictor_state * predictor_decoder_init (predictor_state *state, int predictor, int rowsamples, int components, int compbits) +{ + int rowsize, pixelsize; +#define storage_pos(b, p, size) ((b = p), (p += size)) + uint8_t *buffer, *p; + size_t buffersize; + + pixelsize = (components * compbits + 7) >> 3; // to bytes, rounded up + rowsize = (rowsamples * components * compbits + 7) >> 3; + + state->default_predictor = state->current_predictor = predictor; + state->rowsamples = rowsamples; + state->components = components; + state->compbits = compbits; + + if (predictor == 2) + { /* tiff predictor */ + size_t compbuf, pixbuf; + compbuf = state->components * sizeof(predictor_component_t); + pixbuf = 1 * sizeof(predictor_pixel1b_t); + state->pixbufsize = (int)(compbuf > pixbuf ? compbuf : pixbuf); + buffersize = rowsize + state->pixbufsize; + buffer = (uint8_t *)util_calloc(buffersize, 1); +#if defined __arm__ || defined __ARM__ || defined ARM || defined __ARM || defined __arm || defined __ARM_ARCH ||defined __aarch64__ + { /*memory leak */ + predictor_component_t *c; + if (state->pixbufsize%(sizeof(predictor_component_t))) { + c = malloc(state->pixbufsize - state->pixbufsize%(sizeof(predictor_component_t)) + sizeof(predictor_component_t) ); + } else { + c = malloc(state->pixbufsize); + } + memcpy(c,(state->rowin + rowsize),state->pixbufsize); + if (state->prevcomp){ + free(state->prevcomp); + } + state->prevcomp = c; + } +#else + state->prevcomp = (predictor_component_t *)(state->rowin + rowsize); +#endif + state->sampleindex = state->compindex = 0; + state->bitsin = state->bitsout = 0; + state->compin = state->compout = 0; + } + else + { /* png predictors */ + buffersize = (3 * rowsize + 2 * pixelsize + 1) * sizeof(uint8_t); + p = buffer = (uint8_t *)util_calloc(buffersize, 1); + storage_pos(state->rowin, p, 1 + rowsize); // one extra byte for prediction algorithm tag + p += pixelsize; // pixelsize extra bytes for virtual left pixel at the edge, eg. rowup[-1] (permanently \0) + storage_pos(state->rowup, p, rowsize); // actual row byte + p += pixelsize; // ditto + storage_pos(state->rowsave, p, rowsize); + state->pixelsize = pixelsize; + state->predictorbyte = 0; + } + state->buffer = buffer; + state->rowsize = rowsize; + state->rowindex = 0; + state->rowend = 0; + state->status = STATUS_CONTINUE; + return state; +} + +predictor_state * predictor_encoder_init (predictor_state *state, int predictor, int rowsamples, int components, int compbits) +{ + return predictor_decoder_init(state, predictor, rowsamples, components, compbits); +} + +void predictor_decoder_close (predictor_state *state) +{ + util_free(state->buffer); +} + +void predictor_encoder_close (predictor_state *state) +{ + util_free(state->buffer); +} + +/* +Predictor type identifiers (pdf spec 76). lpdf doesn't hire the codec if predictor is 1. Predictor 15 indicates +that the type of PNG prediction algorithm may change in subsequent lines. We always check algorithm marker anyway. +*/ + +enum predictor_code { + NONE_PREDICTOR = 1, + TIFF_PREDICTOR = 2, + PNG_NONE_PREDICTOR = 10, + PNG_SUB_PREDICTOR = 11, + PNG_UP_PREDICTOR = 12, + PNG_AVERAGE_PREDICTOR = 13, + PNG_PAETH_PREDICTOR = 14, + PNG_OPTIMUM_PREDICTOR = 15 +}; + +/* +All predoctor codecs first read the entire data row into a buffer. This is not crucial for the process, +but allows to separate read/write states. In particular, there is one place in which codec functions +may return on EOD. +*/ + +#define start_row(state) (state->rowindex = 0, state->rowin = state->buffer) + +static int read_scanline (predictor_state *state, iof *I, int size) +{ + int rowtail, left; + while ((rowtail = size - state->rowend) > 0) + { + left = (int)iof_left(I); + if (left >= rowtail) + { + memcpy(state->buffer + state->rowend, I->pos, (size_t)rowtail); + state->rowend += rowtail; + I->pos += rowtail; + start_row(state); + break; + } + else + { + if ((rowtail = left) > 0) + { + memcpy(state->buffer + state->rowend, I->pos, (size_t)rowtail); + state->rowend += rowtail; + I->pos += rowtail; + } + if (iof_input(I) == 0) + { + if (state->rowend == 0) // no scanline to process, no more input + return state->flush ? IOFEOF : IOFEMPTY; + /* If we are here, there is an incomplete scanline in buffer: + - if there is a chance for more (state->flush == 0), than wait for more + - otherwise encode/decode the last incomplete line? + pdf spec p. 76 says that "A row occupies a whole number of bytes", + so this situation should be considered abnormal (not found so far). + */ + if (!state->flush) + return IOFEMPTY; + loggerf("incomplete scanline in predictor filter"); + //return IOFERR; + state->status = STATUS_LAST; + state->rowsize -= size - state->rowend; + start_row(state); + break; + } + } + } + return STATUS_CONTINUE; +} + +#define read_row(state, I, size, status) if ((status = read_scanline(state, I, size)) != STATUS_CONTINUE) return status + +#define ensure_output_bytes(O, n) if (!iof_ensure(O, n)) return IOFFULL + +#define tobyte(c) ((unsigned char)(c)) +#define tocomp(c) ((unsigned short)(c)) + +#define row_byte(state) (state->rowin[state->rowindex]) + +#define up_pixel_byte(state) (state->rowup[state->rowindex]) +#define upleft_pixel_byte(state) (state->rowup[state->rowindex - state->pixelsize]) +#define left_pixel_byte(state) (state->rowsave[state->rowindex - state->pixelsize]) + +#define save_pixel_byte(state, c) (state->rowsave[state->rowindex] = c) + +#define left_pixel_component(state) (state->prevcomp[state->compindex]) // tiff predictor with 2, 4, 8, 16 components +#define left_pixel_value(state) (state->prevpixel[0]) // tiff predictor with 1bit components + +#define save_pixel_component(state, c) ((void)\ + ((state->prevcomp[state->compindex] = c), \ + (++state->compindex < state->components || (state->compindex = 0)))) + +#define save_pixel_value(state, c) (state->prevpixel[0] = c) + +/* Once the codec function is done with the scanline, we set imaginary left pixel data to zero, and reset row counters to +zero in order to allow buffering another input scanline. */ + +#define reset_row(state) state->rowend = 0 + +#define reset_png_row(state) (memcpy(state->rowup, state->rowsave, state->rowsize), state->predictorbyte = 0, reset_row(state)) + +#define reset_tiff_row(state) \ + memset(state->prevcomp, 0, state->pixbufsize), \ + state->bitsin = state->bitsout = 0, \ + state->compin = state->compout = 0, \ + reset_row(state), \ + state->sampleindex = state->compindex = 0 + +/* PNG paeth predictor function; http://www.libpng.org/pub/png/book/chapter09.html +Compute the base value p := left + up - upleft, then choose that byte the closest +(of the smallest absolute difference) to the base value. Left byte has a precedence. */ + + +static int paeth (predictor_state *state) +{ + int p, p1, p2, p3; + p = left_pixel_byte(state) + up_pixel_byte(state) - upleft_pixel_byte(state); + p1 = p >= left_pixel_byte(state) ? (p - left_pixel_byte(state)) : (left_pixel_byte(state) - p); + p2 = p >= up_pixel_byte(state) ? (p - up_pixel_byte(state)) : (up_pixel_byte(state) - p); + p3 = p >= upleft_pixel_byte(state) ? (p - upleft_pixel_byte(state)) : (upleft_pixel_byte(state) - p); + return (p1 <= p2 && p1 <= p3) ? left_pixel_byte(state) : (p2 <= p3 ? up_pixel_byte(state) : upleft_pixel_byte(state)); +} + +/* predictor decoder */ + +iof_status predictor_decode_state (iof *I, iof *O, predictor_state *state) +{ + int status, c, d, outbytes; + while (state->status == STATUS_CONTINUE) + { + if (state->default_predictor >= 10) // PNG predictor? + { + read_row(state, I, state->rowsize + 1, status); + if (state->predictorbyte == 0) + { // we could actually check state->rowin <> state->buffer, but we need this flag for encoder anyway + state->current_predictor = row_byte(state) + 10; + state->predictorbyte = 1; + ++state->rowin; + } + } + else + { + read_row(state, I, state->rowsize, status); + } + switch (state->current_predictor) + { + case NONE_PREDICTOR: + for ( ; state->rowindex < state->rowsize; ++state->rowindex) + { + ensure_output_bytes(O, 1); + c = row_byte(state); + iof_set(O, c); + } + reset_row(state); + break; + case TIFF_PREDICTOR: + switch (state->compbits) + { + case 1: + outbytes = (state->components + 7) >> 3; + for ( ; state->sampleindex < state->rowsamples; ++state->sampleindex) + { + ensure_output_bytes(O, outbytes); + while (state->bitsin < state->components) + { + state->compin = (state->compin << 8) | row_byte(state); + state->bitsin += 8; + ++state->rowindex; + } + state->bitsin -= state->components; + d = state->compin >> state->bitsin; + state->compin &= (1 << state->bitsin) - 1; + c = d ^ left_pixel_value(state); + save_pixel_value(state, c); + state->compout = (state->compout << state->components) | c; + state->bitsout += state->components; + while (state->bitsout >= 8) + { + state->bitsout -= 8; + iof_set(O, state->compout >> state->bitsout); + state->compout &= (1 << state->bitsout) - 1; + } + } + if (state->bitsout > 0) + { + ensure_output_bytes(O, 1); + iof_set(O, state->compin << (8 - state->bitsout)); + } + break; + case 2: case 4: + for ( ; state->sampleindex < state->rowsamples; ++state->sampleindex) + { + for ( ; state->compindex < state->components; ) // state->compindex is ++ed by save_pixel_component() + { + ensure_output_bytes(O, 1); + if (state->bitsin < state->compbits) + { + state->compin = (state->compin << 8) | row_byte(state); + state->bitsin += 8; + ++state->rowindex; + } + state->bitsin -= state->compbits; + d = state->compin >> state->bitsin; + state->compin &= (1 << state->bitsin) - 1; + c = (d + left_pixel_component(state)) & 0xff; + save_pixel_component(state, c); + state->compout = (state->compout << state->compbits) | c; + state->bitsout += state->compbits; + if (state->bitsout >= 8) + { + state->bitsout -= 8; + iof_set(O, state->compout >> state->bitsout); + state->compout &= (1 << state->bitsout) - 1; + } + } + } + if (state->bitsout > 0) + { + ensure_output_bytes(O, 1); + iof_set(O, state->compin << (8 - state->bitsout)); + } + break; + case 8: + for ( ; state->rowindex < state->rowsize; ++state->rowindex) + { + ensure_output_bytes(O, 1); + c = (row_byte(state) + left_pixel_component(state)) & 0xff; + save_pixel_component(state, c); + iof_set(O, c); + } + break; + case 16: + for ( ; state->rowindex < state->rowsize - 1; ++state->rowindex) + { + ensure_output_bytes(O, 2); + d = row_byte(state) << 8; + ++state->rowindex; + d |= row_byte(state); + c = (d + left_pixel_component(state)) & 0xff; + save_pixel_component(state, c); + iof_set2(O, c >> 8, c & 0xff); + } + break; + default: + return IOFERR; + } + reset_tiff_row(state); + break; + case PNG_NONE_PREDICTOR: + for ( ; state->rowindex < state->rowsize; ++state->rowindex) + { + ensure_output_bytes(O, 1); + c = row_byte(state); + save_pixel_byte(state, c); // next row may need it + iof_set(O, c); + } + reset_png_row(state); + break; + case PNG_SUB_PREDICTOR: + for ( ; state->rowindex < state->rowsize; ++state->rowindex) + { + ensure_output_bytes(O, 1); + c = (row_byte(state) + left_pixel_byte(state)) & 0xff; + save_pixel_byte(state, c); + iof_set(O, c); + } + reset_png_row(state); + break; + case PNG_UP_PREDICTOR: + for ( ; state->rowindex < state->rowsize; ++state->rowindex) + { + ensure_output_bytes(O, 1); + c = (row_byte(state) + up_pixel_byte(state)) & 0xff; + save_pixel_byte(state, c); + iof_set(O, c); + } + reset_png_row(state); + break; + case PNG_AVERAGE_PREDICTOR: + for ( ; state->rowindex < state->rowsize; ++state->rowindex) + { + ensure_output_bytes(O, 1); + c = (row_byte(state) + ((up_pixel_byte(state) + left_pixel_byte(state)) / 2)) & 0xff; + save_pixel_byte(state, c); + iof_set(O, c); + } + reset_png_row(state); + break; + case PNG_PAETH_PREDICTOR: + for ( ; state->rowindex < state->rowsize; ++state->rowindex) + { + ensure_output_bytes(O, 1); + c = (row_byte(state) + paeth(state)) & 0xff; + save_pixel_byte(state, c); + iof_set(O, c); + } + reset_png_row(state); + break; + //case PNG_OPTIMUM_PREDICTOR: // valid as default_redictor, but not as algorithm identifier byte + default: + return IOFERR; + } + } + return state->status == STATUS_LAST ? IOFERR : IOFEOF; +} + +/* predictor encoder */ + +iof_status predictor_encode_state (iof *I, iof *O, predictor_state *state) +{ + int status, c, d, outbytes; + while (state->status == STATUS_CONTINUE) + { + read_row(state, I, state->rowsize, status); + if (state->current_predictor >= 10 && state->predictorbyte == 0) + { + ensure_output_bytes(O, 1); + iof_set(O, state->current_predictor - 10); + state->predictorbyte = 1; + } + switch (state->current_predictor) + { + case NONE_PREDICTOR: + for ( ; state->rowindex < state->rowsize; ++state->rowindex) + { + ensure_output_bytes(O, 1); + c = row_byte(state); + iof_set(O, c); + } + reset_row(state); + break; + case TIFF_PREDICTOR: + switch (state->compbits) + { + case 1: + outbytes = (state->components + 7) >> 3; + for ( ; state->sampleindex < state->rowsamples; ++state->sampleindex) + { + ensure_output_bytes(O, outbytes); + while (state->bitsin < state->components) + { + state->compin = (state->compin << 8) | row_byte(state); + state->bitsin += 8; + ++state->rowindex; + } + state->bitsin -= state->components; + c = state->compin >> state->bitsin; + state->compin &= (1 << state->bitsin) - 1; + d = c ^ left_pixel_value(state); + save_pixel_value(state, c); + state->compout = (state->compout << state->components) | d; + state->bitsout += state->components; + while (state->bitsout >= 8) + { + state->bitsout -= 8; + iof_set(O, state->compout >> state->bitsout); + state->compout &= (1 << state->bitsout) - 1; + } + } + if (state->bitsout > 0) + { + ensure_output_bytes(O, 1); + iof_set(O, state->compin << (8 - state->bitsout)); + } + break; + case 2: case 4: + for ( ; state->sampleindex < state->rowsamples; ++state->sampleindex) + { + for ( ; state->compindex < state->components; ) + { + ensure_output_bytes(O, 1); + if (state->bitsin < state->compbits) + { + state->compin = (state->compin << 8) | row_byte(state); + state->bitsin += 8; + ++state->rowindex; + } + state->bitsin -= state->compbits; + c = state->compin >> state->bitsin; + state->compin &= (1 << state->bitsin) - 1; + d = tocomp(c - left_pixel_component(state)); + save_pixel_component(state, c); + state->compout = (state->compout << state->compbits) | d; + state->bitsout += state->compbits; + if (state->bitsout >= 8) + { + state->bitsout -= 8; + iof_set(O, state->compout >> state->bitsout); + state->compout &= (1 << state->bitsout) - 1; + } + } + } + if (state->bitsout > 0) + { + ensure_output_bytes(O, 1); + iof_set(O, state->compin << (8 - state->bitsout)); + } + break; + case 8: + for ( ; state->rowindex < state->rowsize; ++state->rowindex) + { + ensure_output_bytes(O, 1); + c = row_byte(state); + d = tobyte(c - left_pixel_component(state)); + save_pixel_component(state, c); + iof_set(O, d); + } + break; + case 16: + for ( ; state->rowindex < state->rowsize - 1; ++state->rowindex) + { + ensure_output_bytes(O, 2); + c = row_byte(state) << 8; + ++state->rowindex; + c |= row_byte(state); + d = tocomp(c - left_pixel_component(state)); + save_pixel_component(state, c); + iof_set2(O, d >> 8, d & 0xff); + } + break; + default: + return IOFERR; + } + reset_tiff_row(state); + break; + case PNG_NONE_PREDICTOR: + for ( ; state->rowindex < state->rowsize; ++state->rowindex) + { + ensure_output_bytes(O, 1); + c = row_byte(state); + save_pixel_byte(state, c); // next row may need it + iof_set(O, c); + } + reset_png_row(state); + break; + case PNG_SUB_PREDICTOR: + for ( ; state->rowindex < state->rowsize; ++state->rowindex) + { + ensure_output_bytes(O, 1); + c = row_byte(state); + d = tobyte(c - left_pixel_byte(state)); + save_pixel_byte(state, c); + iof_set(O, d); + } + reset_png_row(state); + break; + case PNG_OPTIMUM_PREDICTOR: // not worthy to perform optimization + case PNG_UP_PREDICTOR: + for ( ; state->rowindex < state->rowsize; ++state->rowindex) + { + ensure_output_bytes(O, 1); + c = row_byte(state); + d = tobyte(c - up_pixel_byte(state)); + save_pixel_byte(state, c); + iof_set(O, d); + } + reset_png_row(state); + break; + case PNG_AVERAGE_PREDICTOR: + for ( ; state->rowindex < state->rowsize; ++state->rowindex) + { + ensure_output_bytes(O, 1); + c = row_byte(state); + d = tobyte(c - ((up_pixel_byte(state) + left_pixel_byte(state)) >> 1)); + save_pixel_byte(state, c); + iof_set(O, d); + } + reset_png_row(state); + break; + case PNG_PAETH_PREDICTOR: + for ( ; state->rowindex < state->rowsize; ++state->rowindex) + { + ensure_output_bytes(O, 1); + c = row_byte(state); + d = tobyte(c - paeth(state)); + save_pixel_byte(state, c); + iof_set(O, d); + } + reset_png_row(state); + break; + default: + return IOFERR; + } + } + return state->status == STATUS_LAST ? IOFERR : IOFEOF; +} + +iof_status predictor_decode (iof *I, iof *O, int predictor, int rowsamples, int components, int compbits) +{ + predictor_state state; + int ret; + predictor_decoder_init(&state, predictor, rowsamples, components, compbits); + state.flush = 1; + ret = predictor_decode_state(I, O, &state); + predictor_decoder_close(&state); + return ret; +} + +iof_status predictor_encode (iof *I, iof *O, int predictor, int rowsamples, int components, int compbits) +{ + predictor_state state; + int ret; + predictor_encoder_init(&state, predictor, rowsamples, components, compbits); + state.flush = 1; + ret = predictor_encode_state(I, O, &state); + predictor_encoder_close(&state); + return ret; +} + +/* filters */ + +// predictor decoder function + +static size_t predictor_decoder (iof *F, iof_mode mode) +{ + predictor_state *state; + iof_status status; + size_t tail; + + state = iof_filter_state(predictor_state *, F); + switch(mode) + { + case IOFLOAD: + case IOFREAD: + if (F->flags & IOF_STOPPED) + return 0; + tail = iof_tail(F); + F->pos = F->buf + tail; + F->end = F->buf + F->space; + do { + status = predictor_decode_state(F->next, F, state); + } while (mode == IOFLOAD && status == IOFFULL && iof_resize_buffer(F)); + return iof_decoder_retval(F, "predictor", status); + case IOFCLOSE: + predictor_decoder_close(state); + iof_free(F); + return 0; + default: + break; + } + return 0; +} + +// predictor encoder function + +static size_t predictor_encoder (iof *F, iof_mode mode) +{ + predictor_state *state; + iof_status status; + + state = iof_filter_state(predictor_state *, F); + switch (mode) + { + case IOFFLUSH: + state->flush = 1; + // fall through + case IOFWRITE: + F->end = F->pos; + F->pos = F->buf; + status = predictor_encode_state(F, F->next, state); + return iof_encoder_retval(F, "predictor", status); + case IOFCLOSE: + if (!state->flush) + predictor_encoder(F, IOFFLUSH); + predictor_encoder_close(state); + iof_free(F); + return 0; + default: + break; + } + return 0; +} + +iof * iof_filter_predictor_decoder (iof *N, int predictor, int rowsamples, int components, int compbits) +{ + iof *I; + predictor_state *state; + I = iof_filter_reader(predictor_decoder, sizeof(predictor_state), &state); + iof_setup_next(I, N); + if (predictor_decoder_init(state, predictor, rowsamples, components, compbits) == NULL) + { + iof_discard(I); + return NULL; + } + state->flush = 1; + return I; +} + +iof * iof_filter_predictor_encoder (iof *N, int predictor, int rowsamples, int components, int compbits) +{ + iof *O; + predictor_state *state; + O = iof_filter_writer(predictor_encoder, sizeof(predictor_state), &state); + iof_setup_next(O, N); + if (predictor_encoder_init(state, predictor, rowsamples, components, compbits) == NULL) + { + iof_discard(O); + return NULL; + } + return O; +} diff --git a/Build/source/texk/web2c/luatexdir/luapplib/util/utilfpred.c-OK b/Build/source/texk/web2c/luatexdir/luapplib/util/utilfpred.c-OK new file mode 100644 index 00000000000..8e45ed6c252 --- /dev/null +++ b/Build/source/texk/web2c/luatexdir/luapplib/util/utilfpred.c-OK @@ -0,0 +1,767 @@ +/* predictor filters; common for flate and lzw */ + +#include "utilmem.h" +#include "utillog.h" +#include "utilfpred.h" + +/* +Here we implement predictor filters used with flate and lzw compressions in PDF streams. The main idea of data prediction +is to compute and output the differences between data records instead of this records. Adjacent pixels in images are usually +similar, so differences between pixel values tends to be zero. And both Flate and LZW performs better when the input +is rather smooth. Although a preliminary use of predictors is related to bitmap data, The actual need for predictor filter +came from the fact that xref streams may also be predicted (usually with PNG up-predictor). + +PDF specification allows to use several predictor algorithms, specified by /Predictor key in /DecodeParms dictionary: + + 1 - no predictor (default) + 2 - TIFF horizontal predictor + 10 - PNG none predictor + 11 - PNG sub predictor + 12 - PNG up predictor + 13 - PNG average predictor + 14 - PNG paeth predictor + +All PNG predictors works on bytes, regardless the image color-depth. While encoding, every input data byte is decreased +by the appropriate byte of the previous pixel. Even if the pixel does not fit a full byte, PNG predictors use an artificial +pixel size rounded up to a full byte. PNG predictors utilizes previous (left) pixel, pixel above and previous to above +pixel. In case of PNG, the type of the predictor is written on a dedicated byte on the beginning of every scanline. It +means all predictor functions must maintain and information about left, above and left-above pixels. + +Despite the same differencing idea, TIFF predictors are different. The prediction process bases on pixel components, +which are not necessarily bytes (component of a pixel is added/substracted from a relevant component of a previous +pixel). In TIFF predictor 2, only the previous (the left) pixel is taken into account, there is no need to keep +an information about other surrounding pixels. Also there is no expicit algorithm marker in data; the same prediction +method is applied to all input rows. + +Not surprisingly, predictor encoders and decoders are pretty similar. Encoders take some input value and the previous +input value (or 0 at the beginning of the scanline) and output a difference between them. Decoders takes an input value, +previously decoded value (or zero) and outputs their sum. When encoding, the result is cast to the proper unsigned integer, +when decoding, modulo 256 (or appropriate) is used, which makes encoding and decoding looseless. + +Some extra bits trickery is involved in TIFF predictor function, when components doesn't fit bytes boundary. In that case, +an input is treated as an bits stream. Every input byte is "buffered" in a larger integer, as its lower bits (from right). +Every output value is taken from its higher (left) bits. In a special case of bits-per-component equal 1, we buffer all +pixel bits and use XOR to compute bits difference between pixels. I've excerpted that trick from poppler, but I'm not +really sure if it works any better, especially when the number of components per pixel is 1. In that case we do a hard +bit-by-bit work anyway. + +Predictor codecs state keeps a notion of surrounding pixels. PNG predictors uses left, up and upleft +pixel data, while TIFF predictor (2) only needs the previous (left) pixel. Important to note that PNG +predictors always work on bytes, no matter of color-depth (bits per component), while TIFF predictor +works on pixel components, which not necessarily fits into a complete byte. However, for PNG predictor +the size of a pixel still matters, because 'left' and 'upleft' refers to a corresponding pixel byte, +not necessarily previous byte. + +In PNG prediction, we record every pixel byte (in decoded form) in state->rowsave. At the end of a scanline +we copy state->rowsave to state->rowup, so that in the next scanline we can access up-pixel byte. +Left pixel byte is accessed as state->rowsave (the byte recently stored or virtual left edge byte \0). +Up-left pixel byte is accessed via state->rowup, but with state->pixelsize offset (same as left byte, possibly \0 +at the left edge of the row). Both state->rowup and state->rowsave has a safe span of pixelsize bytes on the left, +that are permanently \0. +*/ + +#define predictor_component_t unsigned short +#define predictor_pixel1b_t unsigned int + +typedef struct predictor_state { + int default_predictor; /* default predictor indicator */ + int current_predictor; /* current predictor, possibly taken from algorithm marker in PNG data */ + int rowsamples; /* number of pixels in a scanline (/DecodeParms << /Columns ... >>) */ + int compbits; /* number of bits per component (/DecodeParms << /BitsPerComponent ... >>) */ + int components; /* number of components (/DecodeParms << /Colors ... >>) */ + uint8_t *buffer; /* temporary private buffer area */ + uint8_t *rowin; /* an input row buffer position */ + int rowsize; /* size of a current scanline in bytes (rounded up) */ + int rowend; /* an input buffer end position */ + int rowindex; /* an output buffer position */ + union { + struct { /* used by PNG predictor codecs */ + uint8_t *rowup, *rowsave; /* previous scanline buffers */ + int predictorbyte; /* flag indicating that algorithm byte is read/written */ + int pixelsize; /* number of bytes per pixel (rounded up) */ + }; + struct { /* used by TIFF predictor codecs */ + union { + predictor_component_t *prevcomp; /* an array of left pixel components */ + predictor_pixel1b_t *prevpixel; /* left pixel value stored on a single integer (for 1bit color-depth) */ + }; + int compin, compout; /* bit stream buffers */ + int bitsin, bitsout; /* bit stream counters */ + int sampleindex; /* pixel counter */ + int compindex; /* component counter */ + int pixbufsize; /* size of pixel buffer in bytes */ + }; + }; + int flush; + int status; +} predictor_state; + +enum { + STATUS_LAST = 0, + STATUS_CONTINUE = 1 // any value different then IOFEOF, IOFERR, ... +}; + +predictor_state * predictor_decoder_init (predictor_state *state, int predictor, int rowsamples, int components, int compbits) +{ + int rowsize, pixelsize; +#define storage_pos(b, p, size) ((b = p), (p += size)) + uint8_t *buffer, *p; + size_t buffersize; + + pixelsize = (components * compbits + 7) >> 3; // to bytes, rounded up + rowsize = (rowsamples * components * compbits + 7) >> 3; + + state->default_predictor = state->current_predictor = predictor; + state->rowsamples = rowsamples; + state->components = components; + state->compbits = compbits; + + if (predictor == 2) + { /* tiff predictor */ + size_t compbuf, pixbuf; + compbuf = state->components * sizeof(predictor_component_t); + pixbuf = 1 * sizeof(predictor_pixel1b_t); + state->pixbufsize = (int)(compbuf > pixbuf ? compbuf : pixbuf); + buffersize = rowsize + state->pixbufsize; + buffer = (uint8_t *)util_calloc(buffersize, 1); + state->prevcomp = (predictor_component_t *)(state->rowin + rowsize); + state->sampleindex = state->compindex = 0; + state->bitsin = state->bitsout = 0; + state->compin = state->compout = 0; + } + else + { /* png predictors */ + buffersize = (3 * rowsize + 2 * pixelsize + 1) * sizeof(uint8_t); + p = buffer = (uint8_t *)util_calloc(buffersize, 1); + storage_pos(state->rowin, p, 1 + rowsize); // one extra byte for prediction algorithm tag + p += pixelsize; // pixelsize extra bytes for virtual left pixel at the edge, eg. rowup[-1] (permanently \0) + storage_pos(state->rowup, p, rowsize); // actual row byte + p += pixelsize; // ditto + storage_pos(state->rowsave, p, rowsize); + state->pixelsize = pixelsize; + state->predictorbyte = 0; + } + state->buffer = buffer; + state->rowsize = rowsize; + state->rowindex = 0; + state->rowend = 0; + state->status = STATUS_CONTINUE; + return state; +} + +predictor_state * predictor_encoder_init (predictor_state *state, int predictor, int rowsamples, int components, int compbits) +{ + return predictor_decoder_init(state, predictor, rowsamples, components, compbits); +} + +void predictor_decoder_close (predictor_state *state) +{ + util_free(state->buffer); +} + +void predictor_encoder_close (predictor_state *state) +{ + util_free(state->buffer); +} + +/* +Predictor type identifiers (pdf spec 76). lpdf doesn't hire the codec if predictor is 1. Predictor 15 indicates +that the type of PNG prediction algorithm may change in subsequent lines. We always check algorithm marker anyway. +*/ + +enum predictor_code { + NONE_PREDICTOR = 1, + TIFF_PREDICTOR = 2, + PNG_NONE_PREDICTOR = 10, + PNG_SUB_PREDICTOR = 11, + PNG_UP_PREDICTOR = 12, + PNG_AVERAGE_PREDICTOR = 13, + PNG_PAETH_PREDICTOR = 14, + PNG_OPTIMUM_PREDICTOR = 15 +}; + +/* +All predoctor codecs first read the entire data row into a buffer. This is not crucial for the process, +but allows to separate read/write states. In particular, there is one place in which codec functions +may return on EOD. +*/ + +#define start_row(state) (state->rowindex = 0, state->rowin = state->buffer) + +static int read_scanline (predictor_state *state, iof *I, int size) +{ + int rowtail, left; + while ((rowtail = size - state->rowend) > 0) + { + left = (int)iof_left(I); + if (left >= rowtail) + { + memcpy(state->buffer + state->rowend, I->pos, (size_t)rowtail); + state->rowend += rowtail; + I->pos += rowtail; + start_row(state); + break; + } + else + { + if ((rowtail = left) > 0) + { + memcpy(state->buffer + state->rowend, I->pos, (size_t)rowtail); + state->rowend += rowtail; + I->pos += rowtail; + } + if (iof_input(I) == 0) + { + if (state->rowend == 0) // no scanline to process, no more input + return state->flush ? IOFEOF : IOFEMPTY; + /* If we are here, there is an incomplete scanline in buffer: + - if there is a chance for more (state->flush == 0), than wait for more + - otherwise encode/decode the last incomplete line? + pdf spec p. 76 says that "A row occupies a whole number of bytes", + so this situation should be considered abnormal (not found so far). + */ + if (!state->flush) + return IOFEMPTY; + loggerf("incomplete scanline in predictor filter"); + //return IOFERR; + state->status = STATUS_LAST; + state->rowsize -= size - state->rowend; + start_row(state); + break; + } + } + } + return STATUS_CONTINUE; +} + +#define read_row(state, I, size, status) if ((status = read_scanline(state, I, size)) != STATUS_CONTINUE) return status + +#define ensure_output_bytes(O, n) if (!iof_ensure(O, n)) return IOFFULL + +#define tobyte(c) ((unsigned char)(c)) +#define tocomp(c) ((unsigned short)(c)) + +#define row_byte(state) (state->rowin[state->rowindex]) + +#define up_pixel_byte(state) (state->rowup[state->rowindex]) +#define upleft_pixel_byte(state) (state->rowup[state->rowindex - state->pixelsize]) +#define left_pixel_byte(state) (state->rowsave[state->rowindex - state->pixelsize]) + +#define save_pixel_byte(state, c) (state->rowsave[state->rowindex] = c) + +#define left_pixel_component(state) (state->prevcomp[state->compindex]) // tiff predictor with 2, 4, 8, 16 components +#define left_pixel_value(state) (state->prevpixel[0]) // tiff predictor with 1bit components + +#define save_pixel_component(state, c) ((void)\ + ((state->prevcomp[state->compindex] = c), \ + (++state->compindex < state->components || (state->compindex = 0)))) + +#define save_pixel_value(state, c) (state->prevpixel[0] = c) + +/* Once the codec function is done with the scanline, we set imaginary left pixel data to zero, and reset row counters to +zero in order to allow buffering another input scanline. */ + +#define reset_row(state) state->rowend = 0 + +#define reset_png_row(state) (memcpy(state->rowup, state->rowsave, state->rowsize), state->predictorbyte = 0, reset_row(state)) + +#define reset_tiff_row(state) \ + memset(state->prevcomp, 0, state->pixbufsize), \ + state->bitsin = state->bitsout = 0, \ + state->compin = state->compout = 0, \ + reset_row(state), \ + state->sampleindex = state->compindex = 0 + +/* PNG paeth predictor function; http://www.libpng.org/pub/png/book/chapter09.html +Compute the base value p := left + up - upleft, then choose that byte the closest +(of the smallest absolute difference) to the base value. Left byte has a precedence. */ + + +static int paeth (predictor_state *state) +{ + int p, p1, p2, p3; + p = left_pixel_byte(state) + up_pixel_byte(state) - upleft_pixel_byte(state); + p1 = p >= left_pixel_byte(state) ? (p - left_pixel_byte(state)) : (left_pixel_byte(state) - p); + p2 = p >= up_pixel_byte(state) ? (p - up_pixel_byte(state)) : (up_pixel_byte(state) - p); + p3 = p >= upleft_pixel_byte(state) ? (p - upleft_pixel_byte(state)) : (upleft_pixel_byte(state) - p); + return (p1 <= p2 && p1 <= p3) ? left_pixel_byte(state) : (p2 <= p3 ? up_pixel_byte(state) : upleft_pixel_byte(state)); +} + +/* predictor decoder */ + +iof_status predictor_decode_state (iof *I, iof *O, predictor_state *state) +{ + int status, c, d, outbytes; + while (state->status == STATUS_CONTINUE) + { + if (state->default_predictor >= 10) // PNG predictor? + { + read_row(state, I, state->rowsize + 1, status); + if (state->predictorbyte == 0) + { // we could actually check state->rowin <> state->buffer, but we need this flag for encoder anyway + state->current_predictor = row_byte(state) + 10; + state->predictorbyte = 1; + ++state->rowin; + } + } + else + { + read_row(state, I, state->rowsize, status); + } + switch (state->current_predictor) + { + case NONE_PREDICTOR: + for ( ; state->rowindex < state->rowsize; ++state->rowindex) + { + ensure_output_bytes(O, 1); + c = row_byte(state); + iof_set(O, c); + } + reset_row(state); + break; + case TIFF_PREDICTOR: + switch (state->compbits) + { + case 1: + outbytes = (state->components + 7) >> 3; + for ( ; state->sampleindex < state->rowsamples; ++state->sampleindex) + { + ensure_output_bytes(O, outbytes); + while (state->bitsin < state->components) + { + state->compin = (state->compin << 8) | row_byte(state); + state->bitsin += 8; + ++state->rowindex; + } + state->bitsin -= state->components; + d = state->compin >> state->bitsin; + state->compin &= (1 << state->bitsin) - 1; + c = d ^ left_pixel_value(state); + save_pixel_value(state, c); + state->compout = (state->compout << state->components) | c; + state->bitsout += state->components; + while (state->bitsout >= 8) + { + state->bitsout -= 8; + iof_set(O, state->compout >> state->bitsout); + state->compout &= (1 << state->bitsout) - 1; + } + } + if (state->bitsout > 0) + { + ensure_output_bytes(O, 1); + iof_set(O, state->compin << (8 - state->bitsout)); + } + break; + case 2: case 4: + for ( ; state->sampleindex < state->rowsamples; ++state->sampleindex) + { + for ( ; state->compindex < state->components; ) // state->compindex is ++ed by save_pixel_component() + { + ensure_output_bytes(O, 1); + if (state->bitsin < state->compbits) + { + state->compin = (state->compin << 8) | row_byte(state); + state->bitsin += 8; + ++state->rowindex; + } + state->bitsin -= state->compbits; + d = state->compin >> state->bitsin; + state->compin &= (1 << state->bitsin) - 1; + c = (d + left_pixel_component(state)) & 0xff; + save_pixel_component(state, c); + state->compout = (state->compout << state->compbits) | c; + state->bitsout += state->compbits; + if (state->bitsout >= 8) + { + state->bitsout -= 8; + iof_set(O, state->compout >> state->bitsout); + state->compout &= (1 << state->bitsout) - 1; + } + } + } + if (state->bitsout > 0) + { + ensure_output_bytes(O, 1); + iof_set(O, state->compin << (8 - state->bitsout)); + } + break; + case 8: + for ( ; state->rowindex < state->rowsize; ++state->rowindex) + { + ensure_output_bytes(O, 1); + c = (row_byte(state) + left_pixel_component(state)) & 0xff; + save_pixel_component(state, c); + iof_set(O, c); + } + break; + case 16: + for ( ; state->rowindex < state->rowsize - 1; ++state->rowindex) + { + ensure_output_bytes(O, 2); + d = row_byte(state) << 8; + ++state->rowindex; + d |= row_byte(state); + c = (d + left_pixel_component(state)) & 0xff; + save_pixel_component(state, c); + iof_set2(O, c >> 8, c & 0xff); + } + break; + default: + return IOFERR; + } + reset_tiff_row(state); + break; + case PNG_NONE_PREDICTOR: + for ( ; state->rowindex < state->rowsize; ++state->rowindex) + { + ensure_output_bytes(O, 1); + c = row_byte(state); + save_pixel_byte(state, c); // next row may need it + iof_set(O, c); + } + reset_png_row(state); + break; + case PNG_SUB_PREDICTOR: + for ( ; state->rowindex < state->rowsize; ++state->rowindex) + { + ensure_output_bytes(O, 1); + c = (row_byte(state) + left_pixel_byte(state)) & 0xff; + save_pixel_byte(state, c); + iof_set(O, c); + } + reset_png_row(state); + break; + case PNG_UP_PREDICTOR: + for ( ; state->rowindex < state->rowsize; ++state->rowindex) + { + ensure_output_bytes(O, 1); + c = (row_byte(state) + up_pixel_byte(state)) & 0xff; + save_pixel_byte(state, c); + iof_set(O, c); + } + reset_png_row(state); + break; + case PNG_AVERAGE_PREDICTOR: + for ( ; state->rowindex < state->rowsize; ++state->rowindex) + { + ensure_output_bytes(O, 1); + c = (row_byte(state) + ((up_pixel_byte(state) + left_pixel_byte(state)) / 2)) & 0xff; + save_pixel_byte(state, c); + iof_set(O, c); + } + reset_png_row(state); + break; + case PNG_PAETH_PREDICTOR: + for ( ; state->rowindex < state->rowsize; ++state->rowindex) + { + ensure_output_bytes(O, 1); + c = (row_byte(state) + paeth(state)) & 0xff; + save_pixel_byte(state, c); + iof_set(O, c); + } + reset_png_row(state); + break; + //case PNG_OPTIMUM_PREDICTOR: // valid as default_redictor, but not as algorithm identifier byte + default: + return IOFERR; + } + } + return state->status == STATUS_LAST ? IOFERR : IOFEOF; +} + +/* predictor encoder */ + +iof_status predictor_encode_state (iof *I, iof *O, predictor_state *state) +{ + int status, c, d, outbytes; + while (state->status == STATUS_CONTINUE) + { + read_row(state, I, state->rowsize, status); + if (state->current_predictor >= 10 && state->predictorbyte == 0) + { + ensure_output_bytes(O, 1); + iof_set(O, state->current_predictor - 10); + state->predictorbyte = 1; + } + switch (state->current_predictor) + { + case NONE_PREDICTOR: + for ( ; state->rowindex < state->rowsize; ++state->rowindex) + { + ensure_output_bytes(O, 1); + c = row_byte(state); + iof_set(O, c); + } + reset_row(state); + break; + case TIFF_PREDICTOR: + switch (state->compbits) + { + case 1: + outbytes = (state->components + 7) >> 3; + for ( ; state->sampleindex < state->rowsamples; ++state->sampleindex) + { + ensure_output_bytes(O, outbytes); + while (state->bitsin < state->components) + { + state->compin = (state->compin << 8) | row_byte(state); + state->bitsin += 8; + ++state->rowindex; + } + state->bitsin -= state->components; + c = state->compin >> state->bitsin; + state->compin &= (1 << state->bitsin) - 1; + d = c ^ left_pixel_value(state); + save_pixel_value(state, c); + state->compout = (state->compout << state->components) | d; + state->bitsout += state->components; + while (state->bitsout >= 8) + { + state->bitsout -= 8; + iof_set(O, state->compout >> state->bitsout); + state->compout &= (1 << state->bitsout) - 1; + } + } + if (state->bitsout > 0) + { + ensure_output_bytes(O, 1); + iof_set(O, state->compin << (8 - state->bitsout)); + } + break; + case 2: case 4: + for ( ; state->sampleindex < state->rowsamples; ++state->sampleindex) + { + for ( ; state->compindex < state->components; ) + { + ensure_output_bytes(O, 1); + if (state->bitsin < state->compbits) + { + state->compin = (state->compin << 8) | row_byte(state); + state->bitsin += 8; + ++state->rowindex; + } + state->bitsin -= state->compbits; + c = state->compin >> state->bitsin; + state->compin &= (1 << state->bitsin) - 1; + d = tocomp(c - left_pixel_component(state)); + save_pixel_component(state, c); + state->compout = (state->compout << state->compbits) | d; + state->bitsout += state->compbits; + if (state->bitsout >= 8) + { + state->bitsout -= 8; + iof_set(O, state->compout >> state->bitsout); + state->compout &= (1 << state->bitsout) - 1; + } + } + } + if (state->bitsout > 0) + { + ensure_output_bytes(O, 1); + iof_set(O, state->compin << (8 - state->bitsout)); + } + break; + case 8: + for ( ; state->rowindex < state->rowsize; ++state->rowindex) + { + ensure_output_bytes(O, 1); + c = row_byte(state); + d = tobyte(c - left_pixel_component(state)); + save_pixel_component(state, c); + iof_set(O, d); + } + break; + case 16: + for ( ; state->rowindex < state->rowsize - 1; ++state->rowindex) + { + ensure_output_bytes(O, 2); + c = row_byte(state) << 8; + ++state->rowindex; + c |= row_byte(state); + d = tocomp(c - left_pixel_component(state)); + save_pixel_component(state, c); + iof_set2(O, d >> 8, d & 0xff); + } + break; + default: + return IOFERR; + } + reset_tiff_row(state); + break; + case PNG_NONE_PREDICTOR: + for ( ; state->rowindex < state->rowsize; ++state->rowindex) + { + ensure_output_bytes(O, 1); + c = row_byte(state); + save_pixel_byte(state, c); // next row may need it + iof_set(O, c); + } + reset_png_row(state); + break; + case PNG_SUB_PREDICTOR: + for ( ; state->rowindex < state->rowsize; ++state->rowindex) + { + ensure_output_bytes(O, 1); + c = row_byte(state); + d = tobyte(c - left_pixel_byte(state)); + save_pixel_byte(state, c); + iof_set(O, d); + } + reset_png_row(state); + break; + case PNG_OPTIMUM_PREDICTOR: // not worthy to perform optimization + case PNG_UP_PREDICTOR: + for ( ; state->rowindex < state->rowsize; ++state->rowindex) + { + ensure_output_bytes(O, 1); + c = row_byte(state); + d = tobyte(c - up_pixel_byte(state)); + save_pixel_byte(state, c); + iof_set(O, d); + } + reset_png_row(state); + break; + case PNG_AVERAGE_PREDICTOR: + for ( ; state->rowindex < state->rowsize; ++state->rowindex) + { + ensure_output_bytes(O, 1); + c = row_byte(state); + d = tobyte(c - ((up_pixel_byte(state) + left_pixel_byte(state)) >> 1)); + save_pixel_byte(state, c); + iof_set(O, d); + } + reset_png_row(state); + break; + case PNG_PAETH_PREDICTOR: + for ( ; state->rowindex < state->rowsize; ++state->rowindex) + { + ensure_output_bytes(O, 1); + c = row_byte(state); + d = tobyte(c - paeth(state)); + save_pixel_byte(state, c); + iof_set(O, d); + } + reset_png_row(state); + break; + default: + return IOFERR; + } + } + return state->status == STATUS_LAST ? IOFERR : IOFEOF; +} + +iof_status predictor_decode (iof *I, iof *O, int predictor, int rowsamples, int components, int compbits) +{ + predictor_state state; + int ret; + predictor_decoder_init(&state, predictor, rowsamples, components, compbits); + state.flush = 1; + ret = predictor_decode_state(I, O, &state); + predictor_decoder_close(&state); + return ret; +} + +iof_status predictor_encode (iof *I, iof *O, int predictor, int rowsamples, int components, int compbits) +{ + predictor_state state; + int ret; + predictor_encoder_init(&state, predictor, rowsamples, components, compbits); + state.flush = 1; + ret = predictor_encode_state(I, O, &state); + predictor_encoder_close(&state); + return ret; +} + +/* filters */ + +// predictor decoder function + +static size_t predictor_decoder (iof *F, iof_mode mode) +{ + predictor_state *state; + iof_status status; + size_t tail; + + state = iof_filter_state(predictor_state *, F); + switch(mode) + { + case IOFLOAD: + case IOFREAD: + if (F->flags & IOF_STOPPED) + return 0; + tail = iof_tail(F); + F->pos = F->buf + tail; + F->end = F->buf + F->space; + do { + status = predictor_decode_state(F->next, F, state); + } while (mode == IOFLOAD && status == IOFFULL && iof_resize_buffer(F)); + return iof_decoder_retval(F, "predictor", status); + case IOFCLOSE: + predictor_decoder_close(state); + iof_free(F); + return 0; + default: + break; + } + return 0; +} + +// predictor encoder function + +static size_t predictor_encoder (iof *F, iof_mode mode) +{ + predictor_state *state; + iof_status status; + + state = iof_filter_state(predictor_state *, F); + switch (mode) + { + case IOFFLUSH: + state->flush = 1; + // fall through + case IOFWRITE: + F->end = F->pos; + F->pos = F->buf; + status = predictor_encode_state(F, F->next, state); + return iof_encoder_retval(F, "predictor", status); + case IOFCLOSE: + if (!state->flush) + predictor_encoder(F, IOFFLUSH); + predictor_encoder_close(state); + iof_free(F); + return 0; + default: + break; + } + return 0; +} + +iof * iof_filter_predictor_decoder (iof *N, int predictor, int rowsamples, int components, int compbits) +{ + iof *I; + predictor_state *state; + I = iof_filter_reader(predictor_decoder, sizeof(predictor_state), &state); + iof_setup_next(I, N); + if (predictor_decoder_init(state, predictor, rowsamples, components, compbits) == NULL) + { + iof_discard(I); + return NULL; + } + state->flush = 1; + return I; +} + +iof * iof_filter_predictor_encoder (iof *N, int predictor, int rowsamples, int components, int compbits) +{ + iof *O; + predictor_state *state; + O = iof_filter_writer(predictor_encoder, sizeof(predictor_state), &state); + iof_setup_next(O, N); + if (predictor_encoder_init(state, predictor, rowsamples, components, compbits) == NULL) + { + iof_discard(O); + return NULL; + } + return O; +} diff --git a/Build/source/texk/web2c/luatexdir/luapplib/util/utilfpred.h b/Build/source/texk/web2c/luatexdir/luapplib/util/utilfpred.h new file mode 100644 index 00000000000..6ae2f893586 --- /dev/null +++ b/Build/source/texk/web2c/luatexdir/luapplib/util/utilfpred.h @@ -0,0 +1,23 @@ +#ifndef UTIL_FILTER_PREDICTOR_H +#define UTIL_FILTER_PREDICTOR_H + +#include "utiliof.h" + +typedef struct predictor_state predictor_state; + +predictor_state * predictor_decoder_init (predictor_state *state, int predictor, int rowsamples, int components, int compbits); +predictor_state * predictor_encoder_init (predictor_state *state, int predictor, int rowsamples, int components, int compbits); + +void predictor_decoder_close (predictor_state *state); +void predictor_encoder_close (predictor_state *state); + +iof_status predictor_decode_state (iof *I, iof *O, predictor_state *state); +iof_status predictor_encode_state (iof *I, iof *O, predictor_state *state); + +iof_status predictor_decode (iof *I, iof *O, int predictor, int rowsamples, int components, int compbits); +iof_status predictor_encode (iof *I, iof *O, int predictor, int rowsamples, int components, int compbits); + +iof * iof_filter_predictor_decoder (iof *N, int predictor, int rowsamples, int components, int compbits); +iof * iof_filter_predictor_encoder (iof *N, int predictor, int rowsamples, int components, int compbits); + +#endif
\ No newline at end of file diff --git a/Build/source/texk/web2c/luatexdir/luapplib/util/utiliof.c b/Build/source/texk/web2c/luatexdir/luapplib/util/utiliof.c new file mode 100644 index 00000000000..7d07cf12d81 --- /dev/null +++ b/Build/source/texk/web2c/luatexdir/luapplib/util/utiliof.c @@ -0,0 +1,2949 @@ +/* input/iutput stream */ + +#include <stdlib.h> +#include <string.h> +#include <stdarg.h> + +#include "utilmem.h" +#include "utillog.h" +#include "utiliof.h" + +/* commons */ + +void * iof_copy_data (const void *data, size_t size) +{ + return memcpy(util_malloc(size), data, size); +} + +uint8_t * iof_copy_file_data (const char *filename, size_t *psize) +{ + FILE *file; + size_t size; + uint8_t *data; + if ((file = fopen(filename, "rb")) == NULL) + return NULL; + fseek(file, 0, SEEK_END); + size = (size_t)ftell(file); + data = (uint8_t *)util_malloc(size); + fseek(file, 0, SEEK_SET); + if ((*psize = fread(data, 1, size, file)) != size) + { + util_free(data); + data = NULL; + } + fclose(file); + return data; +} + +uint8_t * iof_copy_file_handle_data (FILE *file, size_t *psize) +{ + size_t size; + uint8_t *data; + //long offset = ftell(file); // keep offset intact? + fseek(file, 0, SEEK_END); + size = (size_t)ftell(file); + data = (uint8_t *)util_malloc(size); + fseek(file, 0, SEEK_SET); + if ((*psize = fread(data, 1, size, file)) != size) + { + util_free(data); + data = NULL; + } + //fseek(file, offset, SEEK_SET) + return data; +} + +FILE * iof_get_file (iof *F) +{ + if (F->flags & IOF_FILE) + return iof_file_get_file(F->iofile); + if (F->flags & IOF_FILE_HANDLE) + return F->file; + return NULL; +} + +const char * iof_status_kind (iof_status status) +{ + switch (status) + { + case IOFEOF: + return "IOFEOF"; + case IOFERR: + return "IOFERR"; + case IOFEMPTY: + return "IOFEMPTY"; + case IOFFULL: + return "IOFFULL"; + default: + break; + } + return "(unknown)"; +} + +/* shared pseudofile */ + +#define IOF_FILE_DEFAULTS 0 + +iof_file * iof_file_new (FILE *file) +{ + iof_file *iofile = (iof_file *)util_malloc(sizeof(iof_file)); + iof_file_set_fh(iofile, file); + iofile->offset = NULL; + iofile->size = 0; + iofile->name = NULL; + iofile->refcount = 0; + iofile->flags = IOF_FILE_DEFAULTS|IOF_ALLOC; + return iofile; +} + +iof_file * iof_file_init (iof_file *iofile, FILE *file) +{ + iof_file_set_fh(iofile, file); + iofile->offset = NULL; + iofile->size = 0; + iofile->name = NULL; + iofile->refcount = 0; + iofile->flags = IOF_FILE_DEFAULTS; + return iofile; +} + +iof_file * iof_file_rdata (const void *data, size_t size) +{ + iof_file *iofile = (iof_file *)util_malloc(sizeof(iof_file)); + iofile->rbuf = iofile->rpos = (const uint8_t *)data; + iofile->rend = iofile->rbuf + size; + iofile->offset = NULL; + iofile->size = 0; + iofile->name = NULL; + iofile->refcount = 0; + iofile->flags = IOF_FILE_DEFAULTS|IOF_ALLOC|IOF_DATA; + return iofile; +} + +iof_file * iof_file_rdata_init (iof_file *iofile, const void *data, size_t size) +{ + iofile->rbuf = iofile->rpos = (const uint8_t *)data; + iofile->rend = iofile->rbuf + size; + iofile->offset = NULL; + iofile->size = 0; // letse keep it consequently set to zero (only for user disposal) + iofile->name = NULL; + iofile->refcount = 0; + iofile->flags = IOF_FILE_DEFAULTS|IOF_DATA; + return iofile; +} + +iof_file * iof_file_wdata (void *data, size_t size) +{ + return iof_file_rdata((const void *)data, size); +} + +iof_file * iof_file_wdata_init (iof_file *iofile, void *data, size_t size) +{ + return iof_file_rdata_init(iofile, (const void *)data, size); +} + +/* typical uses so far */ + +iof_file * iof_file_reader_from_file_handle (iof_file *iofile, const char *filename, FILE *file, int preload, int closefile) +{ + uint8_t *data; + size_t size; + + if (preload) + { + if ((data = iof_copy_file_handle_data(file, &size)) == NULL) + { + if (closefile) + fclose(file); + return NULL; + } + if (iofile == NULL) + iofile = iof_file_rdata(data, size); + else + iof_file_rdata_init(iofile, data, size); + iofile->flags |= IOF_BUFFER_ALLOC; + if (closefile) + fclose(file); + } + else + { + if (iofile == NULL) + iofile = iof_file_new(file); + else + iof_file_init(iofile, file); + if (closefile) + iofile->flags |= IOF_CLOSE_FILE; + } + if (filename != NULL) + iof_file_set_name(iofile, filename); + return iofile; +} + +iof_file * iof_file_reader_from_file (iof_file *iofile, const char *filename, int preload) +{ + FILE *file; + if ((file = fopen(filename, "rb")) == NULL) + return NULL; + return iof_file_reader_from_file_handle(iofile, filename, file, preload, 1); +} + +iof_file * iof_file_reader_from_data (iof_file *iofile, const void *data, size_t size, int preload, int freedata) +{ + void *newdata; + if (data == NULL) + return NULL; + if (preload) + { + newdata = iof_copy_data(data, size); + if (iofile == NULL) + iofile = iof_file_rdata(newdata, size); + else + iof_file_rdata_init(iofile, newdata, size); + iofile->flags |= IOF_BUFFER_ALLOC; + //if (freedata) // hardly makes sense... we can't free const void * + // util_free((void *)data); + } + else + { + if (iofile == NULL) + iofile = iof_file_rdata(data, size); + else + iof_file_rdata_init(iofile, data, size); + if (freedata) + iofile->flags |= IOF_BUFFER_ALLOC; + } + return iofile; +} + +/* +iof_file * iof_file_writer_from_file (iof_file *iofile, const char *filename) +{ + FILE *file; + if ((file = fopen(filename, "wb")) == NULL) + return NULL; + if (iofile == NULL) + iofile = iof_file_new(file); + else + iof_file_init(iofile, file); + iofile->flags |= IOF_CLOSE_FILE; + iof_file_set_name(iofile, filename); + return iofile; +} +*/ + +/* +Because of limited number of FILE* handles available, we may need to close contained handle +between accessing it. In applications so far (fonts, images) we typically need the source +to parse the file on creation and to rewrite or reload the data on dump. All iof_file api +functions assume that iofile has FILE* opened. Reopening it on every access (ftell, fseek, +read/write) makes no sense, as we would effectively loose control. If the caller invalidates +iofile by closing and nulling its file handle, it is also responsible to reopen when necessary. +*/ + +int iof_file_close_input (iof_file *iofile) +{ + FILE *file; + if (iofile->flags & IOF_DATA) + return 0; + if ((file = iof_file_get_fh(iofile)) == NULL) + return 0; + fclose(file); + iof_file_set_fh(iofile, NULL); + iofile->flags &= ~IOF_RECLOSE_FILE; + iofile->flags |= IOF_REOPEN_FILE; + return 1; +} + +int iof_file_reopen_input (iof_file *iofile) +{ // returns true if iofile readable + FILE *file; + const char *filename; + if (iofile->flags & IOF_DATA) + return 1; + if ((file = iof_file_get_fh(iofile)) != NULL) + return 1; // if present, assumed readable + if ((filename = iofile->name) == NULL || (file = fopen(filename, "rb")) == NULL) + return 0; + iof_file_set_fh(iofile, file); + iofile->flags &= ~IOF_REOPEN_FILE; + iofile->flags |= IOF_RECLOSE_FILE; + return 1; +} + +/* freeing iof_file */ + +void iof_file_free (iof_file *iofile) +{ + FILE *file; + if (iofile->flags & IOF_DATA) + { + if (iofile->flags & IOF_BUFFER_ALLOC) + { + iofile->flags &= ~IOF_BUFFER_ALLOC; + if (iofile->buf != NULL) + { + util_free(iofile->buf); + iofile->buf = iofile->pos = iofile->end = NULL; + } + } + } + else if ((file = iof_file_get_fh(iofile)) != NULL) + { + if (iofile->flags & IOF_CLOSE_FILE) + fclose(file); + iof_file_set_fh(iofile, NULL); + } + iof_file_set_name(iofile, NULL); + if (iofile->flags & IOF_ALLOC) + util_free(iofile); +} + +/* set filename for reopen */ + +void iof_file_set_name (iof_file *iofile, const char *name) +{ + if (iofile->name != NULL) + util_free(iofile->name); + if (name != NULL) + iofile->name = iof_copy_data(name, strlen(name) + 1); + else + iofile->name = NULL; +} + +/* seek */ + +int iof_file_seek (iof_file *iofile, long offset, int whence) +{ + if (iofile->flags & IOF_DATA) + { + switch (whence) + { + case SEEK_SET: + if (offset >= 0 && iofile->buf + offset <= iofile->end) + { + iofile->pos = iofile->buf + offset; + return 0; + } + return -1; + case SEEK_CUR: + if ((offset >= 0 && iofile->pos + offset <= iofile->end) || (offset < 0 && iofile->pos + offset >= iofile->buf)) + { + iofile->pos += offset; + return 0; + } + return -1; + case SEEK_END: + if (offset <= 0 && iofile->end + offset >= iofile->buf) + { + iofile->pos = iofile->end + offset; + return 0; + } + return -1; + } + return -1; + } + return fseek(iof_file_get_fh(iofile), offset, whence); +} + +/* */ + +long iof_file_tell (iof_file *iofile) +{ + return (iofile->flags & IOF_DATA) ? (long)(iofile->pos - iofile->buf) : ftell(iof_file_get_fh(iofile)); +} + +size_t iof_file_size (iof_file *iofile) +{ + long pos, size; + FILE *file; + if (iofile->flags & IOF_DATA) + return (size_t)iof_space(iofile); + file = iof_file_get_fh(iofile); + pos = ftell(file); + fseek(file, 0, SEEK_END); + size = ftell(file); + fseek(file, pos, SEEK_SET); + return size; +} + +int iof_file_eof (iof_file *iofile) +{ + if (iofile->flags & IOF_DATA) + return iofile->pos == iofile->end ? -1 : 0; + return feof(iof_file_get_fh(iofile)); +} + +int iof_file_flush (iof_file *iofile) +{ + if (iofile->flags & IOF_DATA) + return 0; + return fflush(iof_file_get_fh(iofile)); +} + +size_t iof_file_read (void *ptr, size_t size, size_t items, iof_file *iofile) +{ + if (iofile->flags & IOF_DATA) + { + size_t bytes = size * items; + if (bytes > (size_t)iof_left(iofile)) + bytes = (size_t)iof_left(iofile); + memcpy(ptr, iofile->pos, bytes); + iofile->pos += bytes; + return bytes / size; // number of elements read + } + return fread(ptr, size, items, iof_file_get_fh(iofile)); +} + +static size_t iof_file_data_resizeto (iof_file *iofile, size_t space) +{ + uint8_t *newbuf; + size_t size; + size = iof_size(iofile); + if (iofile->flags & IOF_BUFFER_ALLOC) + { + newbuf = (uint8_t *)util_realloc(iofile->buf, space); + } + else + { + newbuf = (uint8_t *)util_malloc(space); + if (size > 0) + memcpy(newbuf, iofile->buf, size); + iofile->flags |= IOF_BUFFER_ALLOC; + } + iofile->buf = newbuf; + iofile->pos = newbuf + size; + iofile->end = newbuf + space; + return space - size; +} + +#define iof_file_data_resize(iofile) iof_file_data_resizeto(iofile, iof_space(iofile) << 1) + +size_t iof_file_write (const void *ptr, size_t size, size_t items, iof_file *iofile) +{ + if (iofile->flags & IOF_DATA) + { + size_t space, sizesofar, bytes; + bytes = size * items; + if (bytes > (size_t)iof_left(iofile)) + { + if ((space = iof_space(iofile)) == 0) // allow iofile->buf/end initially NULL + space = BUFSIZ; + for (sizesofar = iof_size(iofile), space <<= 1; sizesofar + bytes > space; space <<= 1) + ; + if (iof_file_data_resizeto(iofile, space) == 0) + return 0; + } + memcpy(iofile->pos, ptr, bytes); + iofile->pos += bytes; + return bytes / size; + } + return fwrite(ptr, size, items, iof_file_get_fh(iofile)); +} + +size_t iof_file_ensure (iof_file *iofile, size_t bytes) +{ + if (iofile->flags & IOF_DATA) + { + size_t space, sizesofar, left; + left = (size_t)iof_left(iofile); + if (bytes > left) + { + if ((space = iof_space(iofile)) == 0) // allow iofile->buf/end initially NULL + space = BUFSIZ; + for (sizesofar = iof_size(iofile), space <<= 1; sizesofar + bytes > space; space <<= 1); + return iof_file_data_resizeto(iofile, space); + } + return left; + } + return 0; +} + +int iof_file_getc (iof_file *iofile) +{ + if (iofile->flags & IOF_DATA) + return iofile->pos < iofile->end ? *iofile->pos++ : IOFEOF; + return fgetc(iof_file_get_fh(iofile)); +} + +int iof_file_putc (iof_file *iofile, int c) +{ + if (iofile->flags & IOF_DATA) + { + if (iofile->pos >= iofile->end) + if (iof_file_data_resize(iofile) == 0) + return IOFEOF; + *iofile->pos++ = (uint8_t)c; + return c; + } + return fputc(c, iof_file_get_fh(iofile)); +} + +static int iof_file_sync (iof_file *iofile, size_t *offset) +{ + if (iofile->offset != offset) + { + if (iofile->offset != NULL) + *iofile->offset = iof_file_tell(iofile); + iofile->offset = offset; + if (offset) // let offset be NULL + return iof_file_seek(iofile, (long)*offset, SEEK_SET); + } + return 0; +} + +//#define iof_file_unsync(iofile, poffset) (void)((iofile)->offset == poffset && (((iofile)->offset = NULL), 0)) +#define iof_file_unsync(iofile, poffset) ((void)poffset, (iofile)->offset = NULL) + +/* iof seek */ + +#define iof_reader_reset(I) ((I)->pos = (I)->end = (I)->buf) +#define iof_reader_reseek_file(I, offset, whence) (fseek((I)->file, offset, whence) == 0 ? (iof_reader_reset(I), 0) : -1) +#define iof_reader_reseek_iofile(I, offset, whence) (iof_file_seek((I)->iofile, offset, whence) == 0 ? (iof_reader_reset(I), 0) : -1) + +#define iof_writer_reset(O) ((O)->pos = (O)->buf) +#define iof_writer_reseek_file(O, offset, whence) (iof_flush(O), (fseek((O)->file, offset, whence) == 0 ? (iof_writer_reset(O), 0) : -1)) +#define iof_writer_reseek_iofile(O, offset, whence) (iof_flush(O), (iof_file_seek((O)->iofile, offset, whence) == 0 ? (iof_writer_reset(O), 0) : -1)) + +static int iof_reader_seek_data (iof *I, long offset, int whence) +{ + switch (whence) + { + case SEEK_SET: + if (offset >= 0 && I->buf + offset <= I->end) + { + I->pos = I->buf + offset; + return 0; + } + return -1; + case SEEK_CUR: + if ((offset >= 0 && I->pos + offset <= I->end) || (offset < 0 && I->pos + offset >= I->buf)) + { + I->pos += offset; + return 0; + } + return -1; + case SEEK_END: + if (offset <= 0 && I->end + offset >= I->buf) + { + I->pos = I->end + offset; + return 0; + } + return -1; + } + return -1; +} + +static int iof_reader_seek_iofile (iof *I, long offset, int whence) +{ + long fileoffset; + switch (whence) + { + case SEEK_SET: + fileoffset = iof_file_tell(I->iofile); + if (offset <= fileoffset && offset >= fileoffset - iof_space(I)) + { + I->pos = I->end - (fileoffset - offset); + return 0; + } + return iof_reader_reseek_iofile(I, offset, SEEK_SET); + case SEEK_CUR: + if ((offset >= 0 && I->pos + offset <= I->end) || (offset < 0 && I->pos + offset >= I->buf)) + { + I->pos += offset; + return 0; + } + return iof_reader_reseek_iofile(I, offset, SEEK_CUR); + case SEEK_END: + return iof_reader_reseek_iofile(I, offset, SEEK_END); // can we do better? + } + return -1; +} + +static int iof_reader_seek_file (iof *I, long offset, int whence) +{ + long fileoffset; + switch (whence) + { + case SEEK_SET: + fileoffset = ftell(I->file); + if (offset <= fileoffset && offset >= fileoffset - iof_space(I)) + { + I->pos = I->end - (fileoffset - offset); + return 0; + } + return iof_reader_reseek_file(I, offset, SEEK_SET); + case SEEK_CUR: + if ((offset >= 0 && I->pos + offset <= I->end) || (offset < 0 && I->pos + offset >= I->buf)) + { + I->pos += offset; + return 0; + } + return iof_reader_reseek_file(I, offset, SEEK_CUR); + case SEEK_END: + return iof_reader_reseek_file(I, offset, SEEK_END); // can we do better? + } + return -1; +} + +int iof_reader_seek (iof *I, long offset, int whence) +{ + I->flags &= ~IOF_STOPPED; + if (I->flags & IOF_FILE) + return iof_reader_seek_iofile(I, offset, whence); + if (I->flags & IOF_FILE_HANDLE) + return iof_reader_seek_file(I, offset, whence); + if (I->flags & IOF_DATA) + return iof_reader_seek_data(I, offset, whence); + return -1; +} + +int iof_reader_reseek (iof *I, long offset, int whence) +{ + I->flags &= ~IOF_STOPPED; + if (I->flags & IOF_FILE) + return iof_reader_reseek_iofile(I, offset, whence); + if (I->flags & IOF_FILE_HANDLE) + return iof_reader_reseek_file(I, offset, whence); + if (I->flags & IOF_DATA) + return iof_reader_seek_data(I, offset, whence); + return -1; +} + +static int iof_writer_seek_data (iof *O, long offset, int whence) +{ + /* + fseek() allows to seek after the end of file. Seeking does not increase the output file. + No byte is written before fwirte(). It seems to fill the gap with zeros. Until we really need that, + no seeking out of bounds for writers. + */ + O->flags &= ~IOF_STOPPED; + return iof_reader_seek_data(O, offset, whence); +} + +static int iof_writer_seek_iofile (iof *O, long offset, int whence) +{ + long fileoffset; + switch (whence) + { + case SEEK_SET: + fileoffset = iof_file_tell(O->iofile); + if (offset >= fileoffset && offset <= fileoffset + iof_space(O)) + { + O->pos = O->buf + (offset - fileoffset); + return 0; + } + return iof_writer_reseek_iofile(O, offset, SEEK_SET); + case SEEK_CUR: + if ((offset >=0 && O->pos + offset <= O->end) || (offset < 0 && O->pos + offset >= O->buf)) + { + O->pos += offset; + return 0; + } + return iof_writer_reseek_iofile(O, offset, SEEK_CUR); + case SEEK_END: + return iof_writer_reseek_iofile(O, offset, SEEK_END); + } + return -1; +} + +static int iof_writer_seek_file (iof *O, long offset, int whence) +{ + long fileoffset; + switch (whence) + { + case SEEK_SET: + fileoffset = ftell(O->file); + if (offset >= fileoffset && offset <= fileoffset + iof_space(O)) + { + O->pos = O->buf + (offset - fileoffset); + return 0; + } + return iof_writer_reseek_file(O, offset, SEEK_SET); + case SEEK_CUR: + if ((offset >=0 && O->pos + offset <= O->end) || (offset < 0 && O->pos + offset >= O->buf)) + { + O->pos += offset; + return 0; + } + return iof_writer_reseek_file(O, offset, SEEK_CUR); + case SEEK_END: + return iof_writer_reseek_file(O, offset, SEEK_END); + } + return -1; +} + +int iof_writer_seek (iof *I, long offset, int whence) +{ + I->flags &= ~IOF_STOPPED; + if (I->flags & IOF_FILE) + return iof_writer_seek_iofile(I, offset, whence); + if (I->flags & IOF_FILE_HANDLE) + return iof_writer_seek_file(I, offset, whence); + if (I->flags & IOF_DATA) + return iof_writer_seek_data(I, offset, whence); + return -1; +} + +int iof_writer_reseek (iof *I, long offset, int whence) +{ + I->flags &= ~IOF_STOPPED; + if (I->flags & IOF_FILE) + return iof_writer_reseek_iofile(I, offset, whence); + if (I->flags & IOF_FILE_HANDLE) + return iof_writer_reseek_file(I, offset, whence); + if (I->flags & IOF_DATA) + return iof_writer_seek_data(I, offset, whence); + return -1; +} + +int iof_seek (iof *F, long offset, int whence) +{ + return (F->flags & IOF_WRITER) ? iof_writer_seek(F, offset, whence) : iof_reader_seek(F, offset, whence); +} + +int iof_reseek (iof *F, long offset, int whence) +{ + return (F->flags & IOF_WRITER) ? iof_writer_reseek(F, offset, whence) : iof_reader_reseek(F, offset, whence); +} + +/* tell */ + +long iof_reader_tell (iof *I) +{ + if (I->flags & IOF_FILE) + return iof_file_tell(I->iofile) - (long)iof_left(I); + if (I->flags & IOF_FILE_HANDLE) + return ftell(I->file) - (long)iof_left(I); + //if (I->flags & IOF_DATA) + return (long)iof_size(I); +} + +long iof_writer_tell (iof *O) +{ + if (O->flags & IOF_FILE) + return iof_file_tell(O->iofile) + (long)iof_size(O); + if (O->flags & IOF_FILE_HANDLE) + return ftell(O->file) + (long)iof_size(O); + //if (I->flags & IOF_DATA) + return (long)iof_size(O); +} + +long iof_tell (iof *I) +{ + return (I->flags & IOF_WRITER) ? iof_writer_tell(I) : iof_reader_tell(I); +} + +size_t iof_fsize (iof *I) +{ + size_t pos, size; + if (I->flags & IOF_FILE) + return iof_file_size(I->iofile); + if (I->flags & IOF_FILE_HANDLE) + { + pos = (size_t)ftell(I->file); + fseek(I->file, 0, SEEK_END); + size = (size_t)ftell(I->file); + fseek(I->file, (long)pos, SEEK_SET); + return size; + } + //if (I->flags & IOF_DATA) + return (size_t)iof_space(I); +} + +/* save reader tail */ + +size_t iof_save_tail (iof *I) +{ + size_t size, left; + size = iof_size(I); + left = iof_left(I); + if (size >= left) + memcpy(I->buf, I->pos, left); + else + memmove(I->buf, I->pos, left); + return left; +} + +size_t iof_input_save_tail (iof *I, size_t back) +{ + size_t size; + I->flags |= IOF_TAIL; + I->pos -= back; + size = iof_input(I); + I->pos += back; + I->flags &= ~IOF_TAIL; + return size; // + back - back +} + +/* read from file */ + +/* iof free*/ + +static size_t file_read (iof *I); +static size_t file_load (iof *I); + +static size_t file_reader (iof *I, iof_mode mode) +{ + switch (mode) + { + case IOFREAD: + return file_read(I); + case IOFLOAD: + return file_load(I); + case IOFCLOSE: + iof_free(I); + return 0; + default: + return 0; + } +} + +iof * iof_setup_file_handle_reader (iof *I, void *buffer, size_t space, FILE *f) +{ + if (I == NULL) + iof_setup_reader(I, buffer, space); + else + iof_reader_buffer(I, buffer, space); + iof_setup_file(I, f); + I->more = file_reader; + return I; +} + +iof * iof_setup_file_reader (iof *I, void *buffer, size_t space, const char *filename) +{ + FILE *f; + if ((f = fopen(filename, "rb")) == NULL) + return NULL; + if (I == NULL) + iof_setup_reader(I, buffer, space); + else + iof_reader_buffer(I, buffer, space); + iof_setup_file(I, f); + I->flags |= IOF_CLOSE_FILE; + I->more = file_reader; + return I; +} + +/* write to file */ + +static size_t file_write (iof *O, int flush); + +static size_t file_writer (iof *O, iof_mode mode) +{ + switch (mode) + { + case IOFWRITE: + return file_write(O, 0); + case IOFFLUSH: + return file_write(O, 1); + case IOFCLOSE: + file_write(O, 1); + iof_free(O); + return 0; + default: + return 0; + } +} + +iof * iof_setup_file_handle_writer (iof *O, void *buffer, size_t space, FILE *f) +{ + if (O == NULL) + iof_setup_writer(O, buffer, space); + else + iof_writer_buffer(O, buffer, space); + iof_setup_file(O, f); + O->more = file_writer; + return O; +} + +iof * iof_setup_file_writer (iof *O, void *buffer, size_t space, const char *filename) +{ + FILE *f; + if ((f = fopen(filename, "wb")) == NULL) + return NULL; + if (O == NULL) + iof_setup_writer(O, buffer, space); + else + iof_writer_buffer(O, buffer, space); + iof_setup_file(O, f); + O->flags |= IOF_CLOSE_FILE; + O->more = file_writer; + return O; +} + +/* a dedicated handler for stdout/stderr */ + +static size_t stdout_writer (iof *O, iof_mode mode) +{ + switch(mode) + { + case IOFWRITE: + { + fwrite(O->buf, sizeof(uint8_t), iof_size(O), stdout); + O->pos = O->buf; + return O->space; + } + case IOFCLOSE: + case IOFFLUSH: + { + fwrite(O->buf, sizeof(uint8_t), iof_size(O), stdout); + fflush(stdout); + O->pos = O->buf; + return 0; + } + default: + break; + } + return 0; +} + +static size_t stderr_writer (iof *O, iof_mode mode) +{ + switch(mode) + { + case IOFWRITE: + { + fwrite(O->buf, sizeof(uint8_t), iof_size(O), stderr); + O->pos = O->buf; + return O->space; + } + case IOFCLOSE: + case IOFFLUSH: + { + fwrite(O->buf, sizeof(uint8_t), iof_size(O), stderr); + fflush(stderr); + O->pos = O->buf; + return 0; + } + default: + break; + } + return 0; +} + +static uint8_t iof_stdout_buffer[BUFSIZ]; +iof iof_stdout = IOF_WRITER_STRUCT(stdout_writer, NULL, iof_stdout_buffer, BUFSIZ, 0); + +static uint8_t iof_stderr_buffer[BUFSIZ]; +iof iof_stderr = IOF_WRITER_STRUCT(stderr_writer, NULL, iof_stderr_buffer, BUFSIZ, 0); + +/* read from somewhere */ + +iof * iof_reader (iof *I, void *link, iof_handler reader, const void *m, size_t bytes) +{ + I->space = 0; + I->link = link; + I->more = reader; + I->flags = 0; + I->refcount = 0; + if (m != NULL) + { + I->rbuf = I->rpos = (const uint8_t *)m; + I->rend = (const uint8_t *)m + bytes; + return I; + } + return NULL; +} + +iof * iof_string_reader (iof *I, const void *s, size_t bytes) +{ + I->space = 0; + I->link = NULL; + I->more = NULL; + I->flags = 0; // iof_string() sets IOF_DATA + I->refcount = 0; + if (s != NULL) + return iof_string(I, s, bytes); + return NULL; +} + +/* write somewhere */ + +iof * iof_writer (iof *O, void *link, iof_handler writer, void *m, size_t bytes) +{ + O->space = 0; + O->link = link; + O->more = writer; + O->flags = 0; + O->refcount = 0; + if (m != NULL && bytes > 0) + { + O->buf = O->pos = (uint8_t *)m; + O->end = (uint8_t *)m + bytes; + return O; + } + // return iof_null(O); + return NULL; +} + +/* write to growing bytes buffer */ + +static size_t iof_mem_handler (iof *O, iof_mode mode) +{ + switch(mode) + { + case IOFWRITE: + return iof_resize_buffer(O); + case IOFCLOSE: + iof_free(O); + return 0; + default: + return 0; + } +} + +iof * iof_setup_buffer (iof *O, void *buffer, size_t space) +{ + if (O == NULL) + iof_setup_writer(O, buffer, space); + else + iof_writer_buffer(O, buffer, space); + O->link = NULL; + O->flags |= IOF_DATA; + O->more = iof_mem_handler; + return O; +} + +iof * iof_setup_buffermin (iof *O, void *buffer, size_t space, size_t min) +{ + if ((O = iof_setup_buffer(O, buffer, space)) != NULL && space < min) // just allocate min now to avoid further rewriting + { + O->buf = O->pos = (uint8_t *)util_malloc(min); + O->flags |= IOF_BUFFER_ALLOC; + O->end = O->buf + min; + } + return O; +} + +iof * iof_buffer_create (size_t space) +{ + uint8_t *buffer; + iof *O; + space += sizeof(iof); + buffer = util_malloc(space); + if ((O = iof_setup_buffer(NULL, buffer, space)) != NULL) + O->flags |= IOF_ALLOC; + return O; +} + +/* set/get */ + +int iof_getc (iof *I) +{ + if (iof_readable(I)) + return *I->pos++; + return IOFEOF; +} + +int iof_putc (iof *O, int u) +{ + if (iof_writable(O)) + { + iof_set(O, u); + return (uint8_t)u; + } + return IOFFULL; +} + +size_t iof_skip (iof *I, size_t bytes) +{ + while (bytes) + { + if (iof_readable(I)) + ++I->pos; + else + break; + --bytes; + } + return bytes; +} + +/* from iof to iof */ + +iof_status iof_pass (iof *I, iof *O) +{ + size_t leftin, leftout; + if ((leftin = iof_left(I)) == 0) + leftin = iof_input(I); + while (leftin) + { + if ((leftout = iof_left(O)) == 0) + if ((leftout = iof_output(O)) == 0) + return IOFFULL; + while (leftin > leftout) + { + memcpy(O->pos, I->pos, leftout); + I->pos += leftout; + O->pos = O->end; /* eq. += leftout */ + leftin -= leftout; + if ((leftout = iof_output(O)) == 0) + return IOFFULL; + } + if (leftin) + { + memcpy(O->pos, I->pos, leftin); + I->pos = I->end; /* eq. += leftin */ + O->pos += leftin; + } + leftin = iof_input(I); + } + return IOFEOF; +} + +/* read n-bytes */ + +size_t iof_read (iof *I, void *to, size_t size) +{ + size_t leftin, done = 0; + char *s = (char *)to; + + if ((leftin = iof_left(I)) == 0) + if ((leftin = iof_input(I)) == 0) + return done; + while (size > leftin) + { + memcpy(s, I->pos, leftin * sizeof(uint8_t)); + size -= leftin; + done += leftin; + s += leftin; + I->pos = I->end; + if ((leftin = iof_input(I)) == 0) + return done; + } + if (size) + { + memcpy(s, I->pos, size * sizeof(uint8_t)); + I->pos += size; + done += size; + } + return done; +} + +/* rewrite FILE content (use fseek if needed) */ + +size_t iof_write_file_handle (iof *O, FILE *file) +{ + size_t leftout, size, readout; + if ((leftout = iof_left(O)) == 0) + if ((leftout = iof_output(O)) == 0) + return 0; + size = 0; + do { + readout = fread(O->pos, 1, leftout, file); + O->pos += readout; + size += readout; + } while(readout == leftout && (leftout = iof_output(O)) > 0); + return size; +} + +size_t iof_write_file (iof *O, const char *filename) +{ + FILE *file; + size_t size; + if ((file = fopen(filename, "rb")) == NULL) + return 0; + size = iof_write_file_handle(O, file); + fclose(file); + return size; +} + +size_t iof_write_iofile (iof *O, iof_file *iofile, int savepos) +{ + long offset; + size_t size; + FILE *file; + if (iofile->flags & IOF_DATA) + return iof_write(O, iofile->pos, (size_t)(iofile->end - iofile->pos)); + file = iof_file_get_fh(iofile); + if (savepos) + { + offset = ftell(file); + size = iof_write_file_handle(O, file); + fseek(file, offset, SEEK_SET); + return size; + } + return iof_write_file_handle(O, file); +} + +/* write n-bytes */ + +size_t iof_write (iof *O, const void *data, size_t size) +{ + size_t leftout, done = 0; + const char *s = (const char *)data; + if ((leftout = iof_left(O)) == 0) + if ((leftout = iof_output(O)) == 0) + return done; + while (size > leftout) + { + memcpy(O->pos, s, leftout * sizeof(uint8_t)); + size -= leftout; + done += leftout; + s += leftout; + O->pos = O->end; + if ((leftout = iof_output(O)) == 0) + return done; + } + if (size) + { + memcpy(O->pos, s, size * sizeof(uint8_t)); + O->pos += size; + done += size; + } + return done; +} + +/* write '\0'-terminated string */ + +iof_status iof_puts (iof *O, const void *data) +{ + const char *s = (const char *)data; + while (*s) + { + if (iof_writable(O)) + iof_set(O, *s++); + else + return IOFFULL; + } + return IOFEOF; // ? +} + +size_t iof_put_string (iof *O, const void *data) +{ + const char *p, *s = (const char *)data; + for (p = s; *p != '\0' && iof_writable(O); iof_set(O, *p++)); + return p - s; +} + +/* write byte n-times */ + +/* +iof_status iof_repc (iof *O, char c, size_t bytes) +{ + while (bytes) + { + if (iof_writable(O)) + iof_set(O, c); + else + return IOFFULL; + --bytes; + } + return IOFEOF; // ? +} +*/ + +size_t iof_repc (iof *O, char c, size_t bytes) +{ + size_t leftout, todo = bytes; + if ((leftout = iof_left(O)) == 0) + if ((leftout = iof_output(O)) == 0) + return 0; + while (bytes > leftout) + { + memset(O->pos, c, leftout); + bytes -= leftout; + O->pos = O->end; + if ((leftout = iof_output(O)) == 0) + return todo - bytes; + } + if (bytes) + { + memset(O->pos, c, bytes); + O->pos += bytes; + } + return todo; +} + +/* putfs */ + +#define IOF_FMT_SIZE 1024 + +size_t iof_putfs (iof *O, const char *format, ...) +{ + static char buffer[IOF_FMT_SIZE]; + va_list args; + va_start(args, format); + if (vsnprintf(buffer, IOF_FMT_SIZE, format, args) > 0) + { + va_end(args); + return iof_put_string(O, buffer); + } + else + { + va_end(args); + return iof_write(O, buffer, IOF_FMT_SIZE); + } +} + +/* integer from iof; return 1 on success, 0 otherwise */ + +int iof_get_int32 (iof *I, int32_t *number) +{ + int sign, c = iof_char(I); + iof_scan_sign(I, c, sign); + if (!base10_digit(c)) return 0; + iof_read_integer(I, c, *number); + if (sign) *number = -*number; + return 1; +} + +int iof_get_intlw (iof *I, intlw_t *number) +{ + int sign, c = iof_char(I); + iof_scan_sign(I, c, sign); + if (!base10_digit(c)) return 0; + iof_read_integer(I, c, *number); + if (sign) *number = -*number; + return 1; +} + +int iof_get_int64 (iof *I, int64_t *number) +{ + int sign, c = iof_char(I); + iof_scan_sign(I, c, sign); + if (!base10_digit(c)) return 0; + iof_read_integer(I, c, *number); + if (sign) *number = -*number; + return 1; +} + +int iof_get_uint32 (iof *I, uint32_t *number) +{ + int c = iof_char(I); + if (!base10_digit(c)) return 0; + iof_read_integer(I, c, *number); + return 1; +} + +int iof_get_uintlw (iof *I, uintlw_t *number) +{ + int c = iof_char(I); + if (!base10_digit(c)) return 0; + iof_read_integer(I, c, *number); + return 1; +} + +int iof_get_uint64 (iof *I, uint64_t *number) +{ + int c = iof_char(I); + if (!base10_digit(c)) return 0; + iof_read_integer(I, c, *number); + return 1; +} + +int iof_get_int32_radix (iof *I, int32_t *number, int radix) +{ + int sign, c = iof_char(I); + iof_scan_sign(I, c, sign); + if (!base10_digit(c)) return 0; + iof_read_radix(I, c, *number, radix); + if (sign) *number = -*number; + return 1; + +} + +int iof_get_intlw_radix (iof *I, intlw_t *number, int radix) +{ + int sign, c = iof_char(I); + iof_scan_sign(I, c, sign); + if (!base10_digit(c)) return 0; + iof_read_radix(I, c, *number, radix); + if (sign) *number = -*number; + return 1; +} + +int iof_get_int64_radix (iof *I, int64_t *number, int radix) +{ + int sign, c = iof_char(I); + iof_scan_sign(I, c, sign); + if (!base10_digit(c)) return 0; + iof_read_radix(I, c, *number, radix); + if (sign) *number = -*number; + return 1; +} + +int iof_get_uint32_radix (iof *I, uint32_t *number, int radix) +{ + int c = iof_char(I); + if (!base10_digit(c)) return 0; + iof_read_radix(I, c, *number, radix); + return 1; +} + +int iof_get_uintlw_radix (iof *I, uintlw_t *number, int radix) +{ + int c = iof_char(I); + if (!base10_digit(c)) return 0; + iof_read_radix(I, c, *number, radix); + return 1; +} + +int iof_get_uint64_radix (iof *I, uint64_t *number, int radix) +{ + int c = iof_char(I); + if (!base10_digit(c)) return 0; + iof_read_radix(I, c, *number, radix); + return 1; +} + +/* get roman to uint16_t, cf. roman_to_uint16() from utilnumber.c*/ + +/* todo: some trick in place of this macro horror? */ + +#define roman1000(c) (c == 'M' || c == 'm') +#define roman500(c) (c == 'D' || c == 'd') +#define roman100(c) (c == 'C' || c == 'c') +#define roman50(c) (c == 'L' || c == 'l') +#define roman10(c) (c == 'X' || c == 'x') +#define roman5(c) (c == 'V' || c == 'v') +#define roman1(c) (c == 'I' || c == 'i') + +#define roman100s(I, c) \ + (roman100(c) ? (100 + ((c = iof_next(I), roman100(c)) ? (100 + ((c = iof_next(I), roman100(c)) ? (c = iof_next(I), 100) : 0)) : 0)) : 0) +#define roman10s(I, c) \ + (roman10(c) ? (10 + ((c = iof_next(I), roman10(c)) ? (10 + ((c = iof_next(I), roman10(c)) ? (c = iof_next(I), 10) : 0)) : 0)) : 0) +#define roman1s(I, c) \ + (roman1(c) ? (1 + ((c = iof_next(I), roman1(c)) ? (1 + ((c = iof_next(I), roman1(c)) ? (c = iof_next(I), 1) : 0)) : 0)) : 0) + +int iof_get_roman (iof *I, uint16_t *number) +{ + int c; + /* M */ + for (*number = 0, c = iof_char(I); roman1000(c); *number += 1000, c = iof_next(I)); + /* D C */ + if (roman500(c)) + { + c = iof_next(I); + *number += 500 + roman100s(I, c); + } + else if (roman100(c)) + { + c = iof_next(I); + if (roman1000(c)) + { + c = iof_next(I); + *number += 900; + } + else if (roman500(c)) + { + c = iof_next(I); + *number += 400; + } + else + *number += 100 + roman100s(I, c); + } + /* L X */ + if (roman50(c)) + { + c = iof_next(I); + *number += 50 + roman10s(I, c); + } + else if (roman10(c)) + { + c = iof_next(I); + if (roman100(c)) + { + c = iof_next(I); + *number += 90; + } + else if (roman50(c)) + { + c = iof_next(I); + *number += 40; + } + else + *number += 10 + roman10s(I, c); + } + /* V I */ + if (roman5(c)) + { + c = iof_next(I); + *number += 5 + roman1s(I, c); + } + else if (roman1(c)) + { + c = iof_next(I); + if (roman10(c)) + { + c = iof_next(I); + *number += 9; + } + else if (roman5(c)) + { + c = iof_next(I); + *number += 4; + } + else + *number += 1 + roman1s(I, c); + } + return 1; +} + +/* double from iof; return 1 on success */ + +int iof_get_double (iof *I, double *number) // cf. string_to_double() +{ + int sign, exponent10, c = iof_char(I); + iof_scan_sign(I, c, sign); + iof_scan_decimal(I, c, *number); + if (c == '.') + { + c = iof_next(I); + iof_scan_fraction(I, c, *number, exponent10); + } + else + exponent10 = 0; + if (c == 'e' || c == 'E') + { + c = iof_next(I); + iof_scan_exponent10(I, c, exponent10); + } + double_exp10(*number, exponent10); + if (sign) *number = -*number; + return 1; +} + +int iof_get_float (iof *I, float *number) // cf. string_to_float() in utilnumber.c +{ + int sign, exponent10, c = iof_char(I); + iof_scan_sign(I, c, sign); + iof_scan_decimal(I, c, *number); + if (c == '.') + { + c = iof_next(I); + iof_scan_fraction(I, c, *number, exponent10); + } + else + exponent10 = 0; + if (c == 'e' || c == 'E') + { + c = iof_next(I); + iof_scan_exponent10(I, c, exponent10); + } + float_exp10(*number, exponent10); + if (sign) *number = -*number; + return 1; +} + +int iof_conv_double (iof *I, double *number) // cf. convert_to_double() in utilnumber.c +{ + int sign, exponent10, c = iof_char(I); + iof_scan_sign(I, c, sign); + iof_scan_decimal(I, c, *number); + if (c == '.' || c == ',') + { + c = iof_next(I); + iof_scan_fraction(I, c, *number, exponent10); + if (exponent10 < 0) + double_negative_exp10(*number, exponent10); + } + if (sign) *number = -*number; + return 1; +} + +int iof_conv_float (iof *I, float *number) // cf. convert_to_float() +{ + int sign, exponent10, c = iof_char(I); + iof_scan_sign(I, c, sign); + iof_scan_decimal(I, c, *number); + if (c == '.' || c == ',') + { + c = iof_next(I); + iof_scan_fraction(I, c, *number, exponent10); + if (exponent10 < 0) + float_negative_exp10(*number, exponent10); + } + if (sign) *number = -*number; + return 1; +} + +/* integer to iof; return a number of written bytes */ + +#define iof_copy_number_buffer(O, s, p) for (p = s; *p && iof_writable(O); iof_set(O, *p), ++p) + +size_t iof_put_int32 (iof *O, int32_t number) +{ + const char *s, *p; + s = int32_to_string(number); + iof_copy_number_buffer(O, s, p); + return p - s; +} + +size_t iof_put_intlw (iof *O, intlw_t number) +{ + const char *s, *p; + s = intlw_to_string(number); + iof_copy_number_buffer(O, s, p); + return p - s; +} + +size_t iof_put_int64 (iof *O, int64_t number) +{ + const char *s, *p; + s = int64_to_string(number); + iof_copy_number_buffer(O, s, p); + return p - s; +} + +size_t iof_put_uint32 (iof *O, uint32_t number) +{ + const char *s, *p; + s = uint32_to_string(number); + iof_copy_number_buffer(O, s, p); + return p - s; +} + +size_t iof_put_uintlw (iof *O, uintlw_t number) +{ + const char *s, *p; + s = uintlw_to_string(number); + iof_copy_number_buffer(O, s, p); + return p - s; +} + +size_t iof_put_uint64 (iof *O, uint64_t number) +{ + const char *s, *p; + s = uint64_to_string(number); + iof_copy_number_buffer(O, s, p); + return p - s; +} + +size_t iof_put_int32_radix (iof *O, int32_t number, int radix) +{ + const char *s, *p; + s = int32_to_radix(number, radix); + iof_copy_number_buffer(O, s, p); + return p - s; +} + +size_t iof_put_intlw_radix (iof *O, intlw_t number, int radix) +{ + const char *s, *p; + s = intlw_to_radix(number, radix); + iof_copy_number_buffer(O, s, p); + return p - s; +} + +size_t iof_put_int64_radix (iof *O, int64_t number, int radix) +{ + const char *s, *p; + s = int64_to_radix(number, radix); + iof_copy_number_buffer(O, s, p); + return p - s; +} + +size_t iof_put_uint32_radix (iof *O, uint32_t number, int radix) +{ + const char *s, *p; + s = uint32_to_radix(number, radix); + iof_copy_number_buffer(O, s, p); + return p - s; +} + +size_t iof_put_uintlw_radix (iof *O, uintlw_t number, int radix) +{ + const char *s, *p; + s = uintlw_to_radix(number, radix); + iof_copy_number_buffer(O, s, p); + return p - s; +} + +size_t iof_put_uint64_radix (iof *O, uint64_t number, int radix) +{ + const char *s, *p; + s = uint64_to_radix(number, radix); + iof_copy_number_buffer(O, s, p); + return p - s; +} + +/* roman numerals */ + +size_t iof_put_roman_uc (iof *O, uint16_t number) +{ + const char *s, *p; + s = uint16_to_roman_uc(number); + iof_copy_number_buffer(O, s, p); + return p - s; +} + +size_t iof_put_roman_lc (iof *O, uint16_t number) +{ + const char *s, *p; + s = uint16_to_roman_lc(number); + iof_copy_number_buffer(O, s, p); + return p - s; +} + +/* double/float to iof; return the number of written bytes */ + +size_t iof_put_double (iof *O, double number, int digits) +{ + const char *s, *p; + s = double_to_string(number, digits); + iof_copy_number_buffer(O, s, p); + return p - s; +} + +size_t iof_put_float (iof *O, float number, int digits) +{ + const char *s, *p; + s = float_to_string(number, digits); + iof_copy_number_buffer(O, s, p); + return p - s; +} + +/* iof to binary integer; pretty common */ + +int iof_get_be_uint2 (iof *I, uint32_t *pnumber) +{ + int c1, c2; + if ((c1 = iof_get(I)) < 0 || (c2 = iof_get(I)) < 0) + return 0; + *pnumber = (c1<<8)|c2; + return 1; +} + +int iof_get_be_uint3 (iof *I, uint32_t *pnumber) +{ + int c1, c2, c3; + if ((c1 = iof_get(I)) < 0 || (c2 = iof_get(I)) < 0 || (c3 = iof_get(I)) < 0) + return 0; + *pnumber = (c1<<16)|(c2<<8)|c3; + return 1; +} + +int iof_get_be_uint4 (iof *I, uint32_t *pnumber) +{ + int c1, c2, c3, c4; + if ((c1 = iof_get(I)) < 0 || (c2 = iof_get(I)) < 0 || (c3 = iof_get(I)) < 0 || (c4 = iof_get(I)) < 0) + return 0; + *pnumber = (c1<<24)|(c2<<16)|(c3<<8)|c4; + return 1; +} + +int iof_get_le_uint2 (iof *I, uint32_t *pnumber) +{ + int c1, c2; + if ((c1 = iof_get(I)) < 0 || (c2 = iof_get(I)) < 0) + return 0; + *pnumber = (c2<<8)|c1; + return 1; +} + +int iof_get_le_uint3 (iof *I, uint32_t *pnumber) +{ + int c1, c2, c3; + if ((c1 = iof_get(I)) < 0 || (c2 = iof_get(I)) < 0 || (c3 = iof_get(I)) < 0) + return 0; + *pnumber = (c3<<16)|(c2<<8)|c1; + return 1; +} + +int iof_get_le_uint4 (iof *I, uint32_t *pnumber) +{ + int c1, c2, c3, c4; + if ((c1 = iof_get(I)) < 0 || (c2 = iof_get(I)) < 0 || (c3 = iof_get(I)) < 0 || (c4 = iof_get(I)) < 0) + return 0; + *pnumber = (c4<<24)|(c3<<16)|(c2<<8)|c1; + return 1; +} + +/* iof input data */ + +uint8_t * iof_file_input_data (iof_file *iofile, size_t *psize, int *isnew) +{ + uint8_t *data; + if (iofile->flags & IOF_DATA) + { + data = iofile->buf; + *psize = iofile->end - iofile->buf; + *isnew = 0; + return data; + } + if (iof_file_reopen(iofile)) + { + data = iof_copy_file_handle_data(iof_file_get_fh(iofile), psize); + *isnew = 1; + iof_file_reclose(iofile); + return data; + } + return NULL; +} + +/* +uint8_t * iof_file_reader_data (iof_file *iofile, size_t *size) +{ + uint8_t *data; + if (!(iofile->flags & IOF_DATA) || iofile->pos == NULL || (*size = (size_t)iof_left(iofile)) == 0) + return NULL; + if (iofile->flags & IOF_BUFFER_ALLOC) + { + data = iofile->buf; // iofile->pos; // returned must be freeable, makes sense when ->buf == ->pos + iofile->flags &= ~IOF_BUFFER_ALLOC; + iofile->buf = iofile->pos = iofile->end = NULL; + return data; + } + data = (uint8_t *)util_malloc(*size); + memcpy(data, iofile->buf, *size); + return data; +} + +uint8_t * iof_file_writer_data (iof_file *iofile, size_t *size) +{ + uint8_t *data; + if (!(iofile->flags & IOF_DATA) || iofile->buf == NULL || (*size = (size_t)iof_size(iofile)) == 0) + return NULL; + if (iofile->flags & IOF_BUFFER_ALLOC) + { + iofile->flags &= ~IOF_BUFFER_ALLOC; + data = iofile->buf; + iofile->buf = iofile->pos = iofile->end = NULL; + return data; + } + data = (uint8_t *)util_malloc(*size); + memcpy(data, iofile->buf, *size); + return data; +} +*/ + +uint8_t * iof_reader_data (iof *I, size_t *psize) +{ + uint8_t *data; + *psize = (size_t)iof_left(I); + if (I->flags & IOF_BUFFER_ALLOC) + { + data = I->buf; // actually I->pos, but we have to return something freeable + I->flags &= ~IOF_BUFFER_ALLOC; + I->buf = NULL; + } + else + { + data = util_malloc(*psize); + memcpy(data, I->pos, *psize); + } + iof_close(I); + return data; +} + + +uint8_t * iof_writer_data (iof *O, size_t *psize) +{ + uint8_t *data; + *psize = (size_t)iof_size(O); + if (O->flags & IOF_BUFFER_ALLOC) + { + data = O->buf; + O->flags &= ~IOF_BUFFER_ALLOC; + O->buf = NULL; + } + else + { + data = util_malloc(*psize); + memcpy(data, O->buf, *psize); + } + iof_close(O); + return data; +} + +size_t iof_reader_to_file_handle (iof *I, FILE *file) +{ + size_t size; + for (size = 0; iof_readable(I); I->pos = I->end) + size += fwrite(I->buf, sizeof(uint8_t), iof_left(I), file); + return size; +} + +size_t iof_reader_to_file (iof *I, const char *filename) +{ + FILE *file; + size_t size; + if ((file = fopen(filename, "wb")) == NULL) + return 0; + for (size = 0; iof_readable(I); I->pos = I->end) + size += fwrite(I->buf, sizeof(uint8_t), iof_left(I), file); + fclose(file); + return size; +} + +/* debug */ + +size_t iof_data_to_file (const void *data, size_t size, const char *filename) +{ + FILE *fh; + if ((fh = fopen(filename, "wb")) == NULL) + return 0; + // size = fwrite(data, size, sizeof(uint8_t), fh); // WRONG, this always returns 1, as fwrite returns the number of elements successfully written out + size = fwrite(data, sizeof(uint8_t), size, fh); + fclose(fh); + return size; +} + +size_t iof_result_to_file_handle (iof *F, FILE *file) +{ + const void *data; + size_t size; + data = iof_result(F, size); + return iof_data_to_file_handle(data, size, file); +} + +size_t iof_result_to_file (iof *F, const char *filename) +{ + const void *data; + size_t size; + data = iof_result(F, size); + return iof_data_to_file(data, size, filename); +} + +void iof_debug (iof *I, const char *filename) +{ + FILE *file = fopen(filename, "wb"); + if (file != NULL) + { + fprintf(file, ">>> buf %p <<<\n", I->buf); + fwrite(I->buf, sizeof(uint8_t), iof_size(I), file); + fprintf(file, "\n>>> pos %p (%ld) <<<\n", I->pos, (long)iof_size(I)); + fwrite(I->pos, sizeof(uint8_t), iof_left(I), file); + fprintf(file, "\n>>> end %p (%ld) <<<\n", I->end, (long)iof_left(I)); + fwrite(I->end, sizeof(uint8_t), I->space - iof_space(I), file); + fprintf(file, "\n>>> end of buffer %p (%ld) <<<\n", I->buf + I->space, (long)(I->buf + I->space - I->end)); + fclose(file); + } +} + +/* common filters api */ + +/* sizes of filter states on x64 +size of iof_filter: 640 (no longer used; sizeof(iof) + sizeof larger state) +size of file_state: 16 +size of stream_state: 16 +size of flate_state: 104 +size of lzw_state: 56 +size of predictor_state: 104 +size of basexx_state: 48 +size of basexx_state: 48 +size of basexx_state: 48 +size of eexec_state: 40 +size of runlength_state: 24 +size of rc4_state: 24 +size of aes_state: 72 +size of img_state: 576 +size of img: 496 +*/ + +typedef struct iof_heap iof_heap; + +struct iof_heap { + uint8_t *data, *pos; + size_t size, space; + iof_heap *next, *prev; + int refcount; +}; + +typedef struct { + iof_heap *heap; +} iof_heap_ghost; + +static iof_heap * iof_buffers_heap = NULL; +static iof_heap * iof_filters_heap = NULL; + +#define IOF_HEAP_FILTERS_COUNT 4 +#define IOF_BUFFER_SIZE 262144 // (1<<18) +#define IOF_FILTER_SIZE 1024 +// sizeof(iof_filter) on x64 is now 640, img_state 576, img 496, others 16-104 +#define IOF_BUFFER_HEAP_SIZE (IOF_HEAP_FILTERS_COUNT * (IOF_BUFFER_SIZE + sizeof(iof_heap_ghost))) +#define IOF_FILTER_HEAP_SIZE (IOF_HEAP_FILTERS_COUNT * (IOF_FILTER_SIZE + sizeof(iof_heap_ghost))) + +static iof_heap * iof_heap_new (size_t space) +{ + iof_heap *iofheap; + iofheap = (iof_heap *)util_malloc(sizeof(iof_heap) + space); + iofheap->data = iofheap->pos = (uint8_t *)(iofheap + 1); + iofheap->size = iofheap->space = space; + iofheap->next = NULL; + iofheap->prev = NULL; + iofheap->refcount = 0; + return iofheap; +} + +#define iof_heap_free(iofheap) util_free(iofheap) + +void iof_filters_init (void) +{ + if (iof_buffers_heap == NULL) + iof_buffers_heap = iof_heap_new(IOF_BUFFER_HEAP_SIZE); + if (iof_filters_heap == NULL) + iof_filters_heap = iof_heap_new(IOF_FILTER_HEAP_SIZE); +} + +void iof_filters_free (void) +{ + iof_heap *heap, *next; + for (heap = iof_buffers_heap; heap != NULL; heap = next) + { + next = heap->next; + if (heap->refcount != 0) + loggerf("not closed iof filters left (%d)", heap->refcount); + if (next != NULL) + loggerf("iof filters heap left"); + iof_heap_free(heap); + } + iof_buffers_heap = NULL; + for (heap = iof_filters_heap; heap != NULL; heap = next) + { + next = heap->next; + if (heap->refcount != 0) + loggerf("not closed iof buffers left (%d)", heap->refcount); + if (next != NULL) + loggerf("iof buffers heap left"); + iof_heap_free(heap); + } + iof_filters_heap = NULL; +} + +#if defined __arm__ || defined __ARM__ || defined ARM || defined __ARM || defined __arm || defined __ARM_ARCH ||defined __aarch64__ +#define iof_heap_get(hp, ghost, data, siz) \ + (ghost = (iof_heap_ghost *)((void*)((hp)->pos)), \ + ghost->heap = hp, \ + data = (uint8_t *)(ghost + 1), \ + (hp)->pos += siz, \ + (hp)->size -= siz, \ + ++(hp)->refcount) +#else +#define iof_heap_get(hp, ghost, data, siz) \ + (ghost = (iof_heap_ghost *)((hp)->pos), \ + ghost->heap = hp, \ + data = (uint8_t *)(ghost + 1), \ + (hp)->pos += siz, \ + (hp)->size -= siz, \ + ++(hp)->refcount) + +#endif + + +static void * iof_heap_take (iof_heap **pheap, size_t size) +{ + uint8_t *data; + iof_heap_ghost *ghost; + iof_heap *heap, *newheap, *next; + + heap = *pheap; + size += sizeof(iof_heap_ghost); + if (heap->size >= size) + { /* take cheap mem from main heap */ + iof_heap_get(heap, ghost, data, size); + return data; + } + if (size <= heap->space >> 1) + { /* make new cheap heap, make it front */ + *pheap = newheap = iof_heap_new(heap->space); + newheap->next = heap; + heap->prev = newheap; + iof_heap_get(newheap, ghost, data, size); + return data; + } + /* size much larger than expected? should not happen. + make a single-item heap, keep the front heap intact. */ + newheap = iof_heap_new(size); + if ((next = heap->next) != NULL) + { + newheap->next = next; + next->prev = newheap; + } + heap->next = newheap; + newheap->prev = heap; + iof_heap_get(newheap, ghost, data, size); + return data; +} + +void iof_heap_back (void *data) +{ + iof_heap_ghost *ghost; + iof_heap *heap, *next, *prev; + + ghost = ((iof_heap_ghost *)data) - 1; + heap = ghost->heap; + if (heap->refcount == 0) + loggerf("invalid use of iof heap, refcount < 0"); + if (--heap->refcount <= 0) + { + if ((prev = heap->prev) != NULL) + { /* free the heap */ + if ((next = heap->next) != NULL) + prev->next = next, next->prev = prev; + else + prev->next = NULL; + iof_heap_free(heap); + } + else + { /* this is the front heap, just reset */ + heap->pos = heap->data; + heap->size = heap->space; + } + } +} + +void * iof_filter_new (size_t size) +{ // to be removed + void *data; + + iof_filters_init(); + data = iof_heap_take(&iof_filters_heap, size); + return memset(data, 0, size); +} + +static uint8_t * iof_filter_buffer_new (size_t *psize) +{ + iof_filters_init(); + *psize = IOF_BUFFER_SIZE; + return iof_heap_take(&iof_buffers_heap, IOF_BUFFER_SIZE); +} + +iof * iof_filter_reader_new (iof_handler handler, size_t statesize, void **pstate) +{ + iof *F; + void *filter; + uint8_t *buffer; + size_t buffersize; + + iof_filters_init(); + filter = iof_heap_take(&iof_filters_heap, sizeof(iof) + statesize); + F = (iof *)memset(filter, 0, sizeof(iof) + statesize); + buffer = iof_filter_buffer_new(&buffersize); + iof_reader_buffer(F, buffer, buffersize); + F->flags |= IOF_HEAP|IOF_BUFFER_HEAP; + F->more = handler; + *pstate = (F + 1); + return F; +} + +iof * iof_filter_reader_with_buffer_new (iof_handler handler, size_t statesize, void **pstate, void *buffer, size_t buffersize) +{ // for filters that has own buffer (string, some image filters) + iof *F; + void *filter; + iof_filters_init(); + filter = iof_heap_take(&iof_filters_heap, sizeof(iof) + statesize); + F = (iof *)memset(filter, 0, sizeof(iof) + statesize); + iof_reader_buffer(F, buffer, buffersize); + F->flags |= IOF_HEAP; + F->more = handler; + *pstate = (F + 1); + return F; +} + +iof * iof_filter_writer_new (iof_handler handler, size_t statesize, void **pstate) +{ + iof *F; + void *filter; + uint8_t *buffer; + size_t buffersize; + + iof_filters_init(); + filter = iof_heap_take(&iof_filters_heap, sizeof(iof) + statesize); + F = (iof *)memset(filter, 0, sizeof(iof) + statesize); + buffer = iof_filter_buffer_new(&buffersize); + iof_writer_buffer(F, buffer, buffersize); + F->flags |= IOF_HEAP|IOF_BUFFER_HEAP; + F->more = handler; + *pstate = (F + 1); + return F; +} + +iof * iof_filter_writer_with_buffer_new (iof_handler handler, size_t statesize, void **pstate, void *buffer, size_t size) +{ + iof *F; + void *filter; + size_t buffersize; + + iof_filters_init(); + filter = iof_heap_take(&iof_filters_heap, sizeof(iof) + statesize); + F = (iof *)memset(filter, 0, sizeof(iof) + statesize); + buffer = iof_filter_buffer_new(&buffersize); + iof_writer_buffer(F, buffer, buffersize); + F->flags |= IOF_HEAP; + F->more = handler; + *pstate = (F + 1); + return F; +} + +/* close */ + +#define iof_close_next(F) ((void)(iof_decref((F)->next), (F)->next = NULL, 0)) +/* when filter creation fails, we should take care to destroy the filter but leave ->next intact */ +#define iof_clear_next(F) ((void)(iof_unref((F)->next), (F)->next = NULL, 0)) + +#define iof_close_buffer(F) ((void)\ + ((F)->buf != NULL ? \ + ((F->flags & IOF_BUFFER_ALLOC) ? (util_free((F)->buf), (F)->buf = NULL, 0) : \ + ((F->flags & IOF_BUFFER_HEAP) ? (iof_filter_buffer_free((F)->buf), (F)->buf = NULL, 0) : ((F)->buf = NULL, 0))) : 0)) + +/* closing underlying file handle */ + +static void iof_close_file (iof *F) +{ + FILE *file; + //if (F->flags & IOF_FILE_HANDLE) + //{ + if ((file = F->file) != NULL) + { + if (F->flags & IOF_CLOSE_FILE) + fclose(F->file); + F->file = NULL; + } + //} +} + +/* a very special variant for reader filters initiated with iof_file_reopen(). It also calls + iof_file_reclose(), which takes an effect only if previously reopened, but better to keep + all this thin ice separated. Used in filters: iofile_reader, iofile_stream_reader, image + decoders. */ + +static void iof_close_iofile (iof *F) +{ + iof_file *iofile; + //if (F->flags & IOF_FILE) + //{ + if ((iofile = F->iofile) != NULL) + { + iof_file_unsync(iofile, NULL); + iof_file_reclose(iofile); // takes an effect iff prevoiusly reopened + iof_file_decref(iofile); + F->iofile = NULL; + } + //} +} + +void iof_free (iof *F) +{ + if (F->flags & IOF_FILE_HANDLE) + iof_close_file(F); + else if (F->flags & IOF_FILE) + iof_close_iofile(F); + else if (F->flags & IOF_NEXT) + iof_close_next(F); + iof_close_buffer(F); + if (F->flags & IOF_HEAP) + iof_filter_free(F); + else if (F->flags & IOF_ALLOC) + util_free(F); +} + +void iof_discard (iof *F) +{ // so far used only on failed filters creation; as iof_free() but don't dare to release ->next + if (F->flags & IOF_FILE_HANDLE) + iof_close_file(F); + else if (F->flags & IOF_FILE) + iof_close_iofile(F); + else if (F->flags & IOF_NEXT) + iof_close_next(F); + iof_close_buffer(F); + if (F->flags & IOF_HEAP) + iof_filter_free(F); + else if (F->flags & IOF_ALLOC) + util_free(F); +} + +/* resizing buffer */ + +size_t iof_resize_buffer_to (iof *O, size_t space) +{ + uint8_t *buf; + + if (O->flags & IOF_BUFFER_ALLOC) + { + buf = (uint8_t *)util_realloc(O->buf, space); + } + else + { + buf = (uint8_t *)util_malloc(space); + memcpy(buf, O->buf, iof_size(O)); + if (O->flags & IOF_BUFFER_HEAP) + { + iof_filter_buffer_free(O->buf); + O->flags &= ~IOF_BUFFER_HEAP; + } + O->flags |= IOF_BUFFER_ALLOC; + + } + O->pos = buf + iof_size(O); + O->end = buf + space; + O->buf = buf; + O->space = space; + return iof_left(O); +} + +/* */ + +size_t iof_decoder_retval (iof *I, const char *type, iof_status status) +{ + switch (status) + { + case IOFERR: + case IOFEMPTY: // should never happen as we set state.flush = 1 on decoders init + loggerf("%s decoder error (%d, %s)", type, status, iof_status_kind(status)); + I->flags |= IOF_STOPPED; + return 0; + case IOFEOF: // this is the last chunk, + I->flags |= IOF_STOPPED; // so stop it and fall + case IOFFULL: // prepare pointers to read from I->buf + I->end = I->pos; + I->pos = I->buf; + return I->end - I->buf; + } + loggerf("%s decoder bug, invalid retval %d", type, status); + return 0; +} + +size_t iof_encoder_retval (iof *O, const char *type, iof_status status) +{ + switch (status) + { + case IOFERR: + case IOFFULL: + loggerf("%s encoder error (%d, %s)", type, status, iof_status_kind(status)); + return 0; + case IOFEMPTY: + O->pos = O->buf; + O->end = O->buf + O->space; + return O->space; + case IOFEOF: + return 0; + } + loggerf("%s encoder bug, invalid retval %d", type, status); + return 0; +} + +/* file/stream state */ + +typedef struct { + size_t length; + size_t offset; +} file_state; + + +#define file_state_init(state, off, len) ((state)->offset = off, (state)->length = len) + +typedef struct { + size_t length; + size_t offset; +} stream_state; + +#define stream_state_init(state, off, len) ((state)->offset = off, (state)->length = len) + +static size_t file_read (iof *I) +{ + size_t bytes, tail; + if (I->flags & IOF_STOPPED) + return 0; + tail = iof_tail(I); + if ((bytes = tail + fread(I->buf + tail, sizeof(uint8_t), I->space - tail, I->file)) < I->space) + I->flags |= IOF_STOPPED; + I->pos = I->buf; + I->end = I->buf + bytes; + return bytes; +} + +static size_t iofile_read (iof *I, size_t *poffset) +{ + size_t bytes, tail; + if (I->flags & IOF_STOPPED) + return 0; + iof_file_sync(I->iofile, poffset); + tail = iof_tail(I); + if ((bytes = tail + iof_file_read(I->buf + tail, sizeof(uint8_t), I->space - tail, I->iofile)) < I->space) + { + I->flags |= IOF_STOPPED; + iof_file_unsync(I->iofile, poffset); + } + I->pos = I->buf; + I->end = I->buf + bytes; + return bytes; +} + +static size_t file_load (iof *I) +{ + size_t bytes, left, tail; + if (I->flags & IOF_STOPPED) + return 0; + tail = iof_tail(I); + I->pos = I->buf + tail; + I->end = I->buf + I->space; /* don't assume its done when initializing the filter */ + left = I->space - tail; + do { + bytes = fread(I->pos, sizeof(uint8_t), left, I->file); + I->pos += bytes; + } while (bytes == left && (left = iof_resize_buffer(I)) > 0); + I->flags |= IOF_STOPPED; + return iof_loaded(I); +} + +static size_t iofile_load (iof *I, size_t *poffset) +{ + size_t bytes, left, tail; + if (I->flags & IOF_STOPPED) + return 0; + tail = iof_tail(I); + I->pos = I->buf + tail; + I->end = I->buf + I->space; /* don't assume its done when initializing the filter */ + left = I->space - tail; + iof_file_sync(I->iofile, poffset); + do { + bytes = iof_file_read(I->pos, sizeof(uint8_t), left, I->iofile); + I->pos += bytes; + } while (bytes == left && (left = iof_resize_buffer(I)) > 0); + I->flags |= IOF_STOPPED; + iof_file_unsync(I->iofile, poffset); + return iof_loaded(I); +} + +static size_t filter_file_reader (iof *I, iof_mode mode) +{ + switch (mode) + { + case IOFREAD: + return file_read(I); + case IOFLOAD: + return file_load(I); + case IOFCLOSE: + iof_free(I); + return 0; + default: + return 0; + } +} + +static size_t filter_iofile_reader (iof *I, iof_mode mode) +{ + file_state *state; + state = iof_filter_state(file_state *, I); + switch (mode) + { + case IOFREAD: + return iofile_read(I, &state->offset); + case IOFLOAD: + return iofile_load(I, &state->offset); + case IOFCLOSE: + iof_free(I); + return 0; + default: + return 0; + } +} + +static size_t file_write (iof *O, int flush) +{ + size_t bytes; + if ((bytes = iof_size(O)) > 0) + if (bytes != fwrite(O->buf, sizeof(uint8_t), bytes, O->file)) + return 0; + if (flush) + fflush(O->file); + O->end = O->buf + O->space; // remains intact actually + O->pos = O->buf; + return O->space; +} + +static size_t iofile_write (iof *O, size_t *poffset, int flush) +{ + size_t bytes; + iof_file_sync(O->iofile, poffset); + if ((bytes = iof_size(O)) > 0) + { + if (bytes != iof_file_write(O->buf, sizeof(uint8_t), bytes, O->iofile)) + { + iof_file_unsync(O->iofile, poffset); + return 0; + } + } + if (flush) + iof_file_flush(O->iofile); + O->end = O->buf + O->space; // remains intact actually + O->pos = O->buf; + return O->space; +} + +static size_t filter_file_writer (iof *O, iof_mode mode) +{ + switch (mode) + { + case IOFWRITE: + return file_write(O, 0); + case IOFFLUSH: + return file_write(O, 1); + case IOFCLOSE: + file_write(O, 1); + iof_free(O); + return 0; + default: + return 0; + } +} + +static size_t filter_iofile_writer (iof *O, iof_mode mode) +{ + file_state *state; + state = iof_filter_state(file_state *, O); + switch (mode) + { + case IOFWRITE: + return iofile_write(O, &state->offset, 0); + case IOFFLUSH: + return iofile_write(O, &state->offset, 1); + case IOFCLOSE: + iofile_write(O, &state->offset, 1); + iof_free(O); + return 0; + default: + return 0; + } +} + +/* filter from FILE* */ + +iof * iof_filter_file_handle_reader (FILE *file) +{ + iof *I; + file_state *state; + if (file == NULL) + return NULL; + I = iof_filter_reader(filter_file_reader, sizeof(file_state), &state); + iof_setup_file(I, file); + file_state_init(state, 0, 0); + return I; +} + +iof * iof_filter_file_handle_writer (FILE *file) +{ + iof *O; + file_state *state; + if (file == NULL) + return NULL; + O = iof_filter_writer(filter_file_writer, sizeof(file_state), &state); + iof_setup_file(O, file); + file_state_init(state, 0, 0); + return O; +} + +/* filter from iof_file * */ + +iof * iof_filter_iofile_reader (iof_file *iofile, size_t offset) +{ + iof *I; + file_state *state; + if (!iof_file_reopen(iofile)) + return NULL; + I = iof_filter_reader(filter_iofile_reader, sizeof(file_state), &state); + iof_setup_iofile(I, iofile); + file_state_init(state, offset, 0); + return I; +} + +iof * iof_filter_iofile_writer (iof_file *iofile, size_t offset) +{ + iof *O; + file_state *state; + O = iof_filter_writer(filter_iofile_writer, sizeof(file_state), &state); + iof_setup_iofile(O, iofile); + file_state_init(state, offset, 0); + return O; +} + +/* filter from filename */ + +iof * iof_filter_file_reader (const char *filename) +{ + iof *I; + file_state *state; + FILE *file; + if ((file = fopen(filename, "rb")) == NULL) + return NULL; + I = iof_filter_reader(filter_file_reader, sizeof(file_state), &state); + iof_setup_file(I, file); + file_state_init(state, 0, 0); + I->flags |= IOF_CLOSE_FILE; + return I; +} + +iof * iof_filter_file_writer (const char *filename) +{ + iof *O; + file_state *state; + FILE *file; + if ((file = fopen(filename, "wb")) == NULL) + return NULL; + O = iof_filter_writer(filter_file_writer, sizeof(file_state), &state); + iof_setup_file(O, file); + file_state_init(state, 0, 0); + O->flags |= IOF_CLOSE_FILE; + return O; +} + +/* from string */ + +static size_t dummy_handler (iof *I, iof_mode mode) +{ + switch (mode) + { + case IOFCLOSE: + iof_free(I); + return 0; + default: + return 0; + } +} + +iof * iof_filter_string_reader (const void *s, size_t length) +{ + iof *I; + void *dummy; + I = iof_filter_reader_with_buffer(dummy_handler, 0, &dummy, NULL, 0); + I->rbuf = I->rpos = (const uint8_t *)s; + I->rend = (const uint8_t *)s + length; + // I->space = length; + return I; +} + +iof * iof_filter_string_writer (const void *s, size_t length) +{ + iof *O; + void *dummy; + O = iof_filter_reader_with_buffer(dummy_handler, 0, &dummy, NULL, 0); + O->rbuf = O->rpos = (const uint8_t *)s; + O->rend = (const uint8_t *)s + length; + // O->space = length; + return O; +} + +iof * iof_filter_buffer_writer (size_t size) +{ // filter alternative of iof_buffer_create() + iof *O; + void *dummy; + uint8_t *buffer; + if (size > IOF_BUFFER_SIZE) + { + buffer = (uint8_t *)util_malloc(size); + O = iof_filter_writer_with_buffer(iof_mem_handler, 0, &dummy, buffer, size); + O->flags |= IOF_BUFFER_ALLOC; + return O; + } + return iof_filter_writer(iof_mem_handler, 0, &dummy); +} + +/* stream */ + +static size_t file_stream_read (iof *I, size_t *plength) +{ + size_t bytes, tail; + if (I->flags & IOF_STOPPED || *plength == 0) + return 0; + tail = iof_tail(I); + if (I->space - tail >= *plength) + { + bytes = tail + fread(I->buf + tail, sizeof(uint8_t), *plength, I->file); + I->flags |= IOF_STOPPED; + *plength = 0; + } + else + { + bytes = tail + fread(I->buf + tail, sizeof(uint8_t), I->space - tail, I->file); + *plength -= bytes - tail; + } + I->pos = I->buf; + I->end = I->buf + bytes; + return bytes; +} + +static size_t iofile_stream_read (iof *I, size_t *plength, size_t *poffset) +{ + size_t bytes, tail; + if (I->flags & IOF_STOPPED || *plength == 0) + return 0; + tail = iof_tail(I); + iof_file_sync(I->iofile, poffset); + if (I->space - tail >= *plength) + { + bytes = tail + iof_file_read(I->buf + tail, sizeof(uint8_t), *plength, I->iofile); + iof_file_unsync(I->iofile, poffset); + I->flags |= IOF_STOPPED; + *plength = 0; + } + else + { + bytes = tail + iof_file_read(I->buf + tail, sizeof(uint8_t), I->space - tail, I->iofile); + *plength -= bytes - tail; + } + I->pos = I->buf; + I->end = I->buf + bytes; + return bytes; +} + +static size_t file_stream_load (iof *I, size_t *plength) +{ + size_t bytes, tail; + if (I->flags & IOF_STOPPED || *plength == 0) + return 0; + tail = iof_tail(I); + if (I->space - tail < *plength) + if (iof_resize_buffer_to(I, tail + *plength) == 0) + return 0; + bytes = tail + fread(I->buf + tail, sizeof(uint8_t), *plength, I->file); + I->flags |= IOF_STOPPED; + *plength = 0; + I->pos = I->buf; + I->end = I->buf + bytes; + return bytes; +} + +static size_t iofile_stream_load (iof *I, size_t *plength, size_t *poffset) +{ + size_t bytes, tail; + if (I->flags & IOF_STOPPED || *plength == 0) + return 0; + iof_file_sync(I->iofile, poffset); + tail = iof_tail(I); + if (I->space - tail < *plength) + if (iof_resize_buffer_to(I, tail + *plength) == 0) + return 0; + bytes = tail + iof_file_read(I->buf + tail, sizeof(uint8_t), *plength, I->iofile); + iof_file_unsync(I->iofile, poffset); + I->flags |= IOF_STOPPED; + *plength = 0; + I->pos = I->buf; + I->end = I->buf + bytes; + return bytes; +} + +static size_t filter_file_stream_reader (iof *I, iof_mode mode) +{ + stream_state *state; + state = iof_filter_state(stream_state *, I); + switch(mode) + { + case IOFREAD: + return file_stream_read(I, &state->length); + case IOFLOAD: + return file_stream_load(I, &state->length); + case IOFCLOSE: + iof_free(I); + return 0; + default: + return 0; + } +} + +static size_t filter_iofile_stream_reader (iof *I, iof_mode mode) +{ + stream_state *state; + state = iof_filter_state(stream_state *, I); + switch(mode) + { + case IOFREAD: + return iofile_stream_read(I, &state->length, &state->offset); + case IOFLOAD: + return iofile_stream_load(I, &state->length, &state->offset); + case IOFCLOSE: + iof_free(I); + return 0; + default: + return 0; + } +} + +iof * iof_filter_stream_reader (FILE *file, size_t offset, size_t length) +{ + iof *I; + stream_state *state; + I = iof_filter_reader(filter_file_stream_reader, sizeof(stream_state), &state); + iof_setup_file(I, file); + stream_state_init(state, offset, length); + fseek(file, (long)offset, SEEK_SET); // or perhaps it should be call in file_stream_read(), like iof_file_sync()? + return I; +} + +iof * iof_filter_stream_coreader (iof_file *iofile, size_t offset, size_t length) +{ + iof *I; + stream_state *state; + if (!iof_file_reopen(iofile)) + return NULL; + I = iof_filter_reader(filter_iofile_stream_reader, sizeof(stream_state), &state); + iof_setup_iofile(I, iofile); + stream_state_init(state, offset, length); + return I; +} + +static size_t file_stream_write (iof *O, size_t *plength, int flush) +{ + size_t bytes; + if ((bytes = iof_size(O)) > 0) + { + if (bytes != fwrite(O->buf, sizeof(uint8_t), bytes, O->file)) + { + *plength += bytes; + return 0; + } + } + if (flush) + fflush(O->file); + *plength += bytes; + O->end = O->buf + O->space; // remains intact + O->pos = O->buf; + return O->space; +} + +static size_t iofile_stream_write (iof *O, size_t *plength, size_t *poffset, int flush) +{ + size_t bytes; + if ((bytes = iof_size(O)) > 0) + { + iof_file_sync(O->iofile, poffset); + if (bytes != iof_file_write(O->buf, sizeof(uint8_t), bytes, O->iofile)) + { + *plength += bytes; + iof_file_unsync(O->iofile, poffset); + return 0; + } + } + if (flush) + iof_file_flush(O->iofile); + *plength += bytes; + O->end = O->buf + O->space; // remains intact + O->pos = O->buf; + return O->space; +} + +static size_t filter_file_stream_writer (iof *O, iof_mode mode) +{ + stream_state *state; + state = iof_filter_state(stream_state *, O); + switch (mode) + { + case IOFWRITE: + return file_stream_write(O, &state->length, 0); + case IOFFLUSH: + return file_stream_write(O, &state->length, 1); + case IOFCLOSE: + file_stream_write(O, &state->length, 1); + iof_free(O); + return 0; + default: + return 0; + } +} + +static size_t filter_iofile_stream_writer (iof *O, iof_mode mode) +{ + stream_state *state; + state = iof_filter_state(stream_state *, O); + switch (mode) + { + case IOFWRITE: + return iofile_stream_write(O, &state->length, &state->offset, 0); + case IOFFLUSH: + return iofile_stream_write(O, &state->length, &state->offset, 1); + case IOFCLOSE: + iofile_stream_write(O, &state->length, &state->offset, 1); + iof_free(O); + return 0; + default: + return 0; + } +} + +iof * iof_filter_stream_writer (FILE *file) +{ + iof *O; + stream_state *state; + O = iof_filter_writer(filter_file_stream_writer, sizeof(stream_state), &state); + iof_setup_file(O, file); + stream_state_init(state, 0, 0); + return O; +} + +iof * iof_filter_stream_cowriter (iof_file *iofile, size_t offset) +{ + iof *O; + stream_state *state; + O = iof_filter_writer(filter_iofile_stream_writer, sizeof(stream_state), &state); + iof_setup_iofile(O, iofile); + stream_state_init(state, offset, 0); + return O; +} + +/* very specific for images; get input from already created strem filter, exchange the filter but keep the buffer */ + +FILE * iof_filter_file_reader_source (iof *I, size_t *poffset, size_t *plength) +{ + stream_state *sstate; + file_state *fstate; + if (I->more == filter_file_stream_reader) // I is the result of iof_filter_stream_reader() + { + sstate = iof_filter_state(stream_state *, I); + *poffset = sstate->offset; + *plength = sstate->length; // might be 0 but it is ok for file readers + return I->file; + } + if (I->more == filter_file_reader) + { + fstate = iof_filter_state(file_state *, I); + *poffset = fstate->offset; + *plength = fstate->length; // might be 0 but it is ok for file readers + return I->file; + } + return NULL; +} + +iof_file * iof_filter_file_coreader_source (iof *I, size_t *poffset, size_t *plength) +{ + stream_state *sstate; + file_state *fstate; + if (I->more == filter_iofile_stream_reader) // I is the result of iof_filter_stream_coreader() + { + sstate = iof_filter_state(stream_state *, I); + *poffset = sstate->offset; + *plength = sstate->length; + return I->iofile; + } + if (I->more == filter_iofile_reader) + { + fstate = iof_filter_state(file_state *, I); + *poffset = fstate->offset; + *plength = fstate->length; + return I->iofile; + } + return NULL; +} + +iof * iof_filter_reader_replacement (iof *P, iof_handler handler, size_t statesize, void **pstate) +{ // called after iof_filter_file_reader_source(), no need to check if F is filter from iof heap and if has buffer from iof heap + iof *F; + F = iof_filter_reader_with_buffer(handler, statesize, pstate, P->buf, P->space); + F->flags |= IOF_BUFFER_HEAP; + //iof_reader_buffer(P, NULL, 0); + //P->flags &= ~IOF_BUFFER_HEAP; + iof_filter_free(P); + return F; +} + + + + + + + + + + + + + + + + + + + + + + + diff --git a/Build/source/texk/web2c/luatexdir/luapplib/util/utiliof.c-OK b/Build/source/texk/web2c/luatexdir/luapplib/util/utiliof.c-OK new file mode 100644 index 00000000000..8815c8d5a14 --- /dev/null +++ b/Build/source/texk/web2c/luatexdir/luapplib/util/utiliof.c-OK @@ -0,0 +1,2938 @@ +/* input/iutput stream */ + +#include <stdlib.h> +#include <string.h> +#include <stdarg.h> + +#include "utilmem.h" +#include "utillog.h" +#include "utiliof.h" + +/* commons */ + +void * iof_copy_data (const void *data, size_t size) +{ + return memcpy(util_malloc(size), data, size); +} + +uint8_t * iof_copy_file_data (const char *filename, size_t *psize) +{ + FILE *file; + size_t size; + uint8_t *data; + if ((file = fopen(filename, "rb")) == NULL) + return NULL; + fseek(file, 0, SEEK_END); + size = (size_t)ftell(file); + data = (uint8_t *)util_malloc(size); + fseek(file, 0, SEEK_SET); + if ((*psize = fread(data, 1, size, file)) != size) + { + util_free(data); + data = NULL; + } + fclose(file); + return data; +} + +uint8_t * iof_copy_file_handle_data (FILE *file, size_t *psize) +{ + size_t size; + uint8_t *data; + //long offset = ftell(file); // keep offset intact? + fseek(file, 0, SEEK_END); + size = (size_t)ftell(file); + data = (uint8_t *)util_malloc(size); + fseek(file, 0, SEEK_SET); + if ((*psize = fread(data, 1, size, file)) != size) + { + util_free(data); + data = NULL; + } + //fseek(file, offset, SEEK_SET) + return data; +} + +FILE * iof_get_file (iof *F) +{ + if (F->flags & IOF_FILE) + return iof_file_get_file(F->iofile); + if (F->flags & IOF_FILE_HANDLE) + return F->file; + return NULL; +} + +const char * iof_status_kind (iof_status status) +{ + switch (status) + { + case IOFEOF: + return "IOFEOF"; + case IOFERR: + return "IOFERR"; + case IOFEMPTY: + return "IOFEMPTY"; + case IOFFULL: + return "IOFFULL"; + default: + break; + } + return "(unknown)"; +} + +/* shared pseudofile */ + +#define IOF_FILE_DEFAULTS 0 + +iof_file * iof_file_new (FILE *file) +{ + iof_file *iofile = (iof_file *)util_malloc(sizeof(iof_file)); + iof_file_set_fh(iofile, file); + iofile->offset = NULL; + iofile->size = 0; + iofile->name = NULL; + iofile->refcount = 0; + iofile->flags = IOF_FILE_DEFAULTS|IOF_ALLOC; + return iofile; +} + +iof_file * iof_file_init (iof_file *iofile, FILE *file) +{ + iof_file_set_fh(iofile, file); + iofile->offset = NULL; + iofile->size = 0; + iofile->name = NULL; + iofile->refcount = 0; + iofile->flags = IOF_FILE_DEFAULTS; + return iofile; +} + +iof_file * iof_file_rdata (const void *data, size_t size) +{ + iof_file *iofile = (iof_file *)util_malloc(sizeof(iof_file)); + iofile->rbuf = iofile->rpos = (const uint8_t *)data; + iofile->rend = iofile->rbuf + size; + iofile->offset = NULL; + iofile->size = 0; + iofile->name = NULL; + iofile->refcount = 0; + iofile->flags = IOF_FILE_DEFAULTS|IOF_ALLOC|IOF_DATA; + return iofile; +} + +iof_file * iof_file_rdata_init (iof_file *iofile, const void *data, size_t size) +{ + iofile->rbuf = iofile->rpos = (const uint8_t *)data; + iofile->rend = iofile->rbuf + size; + iofile->offset = NULL; + iofile->size = 0; // letse keep it consequently set to zero (only for user disposal) + iofile->name = NULL; + iofile->refcount = 0; + iofile->flags = IOF_FILE_DEFAULTS|IOF_DATA; + return iofile; +} + +iof_file * iof_file_wdata (void *data, size_t size) +{ + return iof_file_rdata((const void *)data, size); +} + +iof_file * iof_file_wdata_init (iof_file *iofile, void *data, size_t size) +{ + return iof_file_rdata_init(iofile, (const void *)data, size); +} + +/* typical uses so far */ + +iof_file * iof_file_reader_from_file_handle (iof_file *iofile, const char *filename, FILE *file, int preload, int closefile) +{ + uint8_t *data; + size_t size; + + if (preload) + { + if ((data = iof_copy_file_handle_data(file, &size)) == NULL) + { + if (closefile) + fclose(file); + return NULL; + } + if (iofile == NULL) + iofile = iof_file_rdata(data, size); + else + iof_file_rdata_init(iofile, data, size); + iofile->flags |= IOF_BUFFER_ALLOC; + if (closefile) + fclose(file); + } + else + { + if (iofile == NULL) + iofile = iof_file_new(file); + else + iof_file_init(iofile, file); + if (closefile) + iofile->flags |= IOF_CLOSE_FILE; + } + if (filename != NULL) + iof_file_set_name(iofile, filename); + return iofile; +} + +iof_file * iof_file_reader_from_file (iof_file *iofile, const char *filename, int preload) +{ + FILE *file; + if ((file = fopen(filename, "rb")) == NULL) + return NULL; + return iof_file_reader_from_file_handle(iofile, filename, file, preload, 1); +} + +iof_file * iof_file_reader_from_data (iof_file *iofile, const void *data, size_t size, int preload, int freedata) +{ + void *newdata; + if (data == NULL) + return NULL; + if (preload) + { + newdata = iof_copy_data(data, size); + if (iofile == NULL) + iofile = iof_file_rdata(newdata, size); + else + iof_file_rdata_init(iofile, newdata, size); + iofile->flags |= IOF_BUFFER_ALLOC; + //if (freedata) // hardly makes sense... we can't free const void * + // util_free((void *)data); + } + else + { + if (iofile == NULL) + iofile = iof_file_rdata(data, size); + else + iof_file_rdata_init(iofile, data, size); + if (freedata) + iofile->flags |= IOF_BUFFER_ALLOC; + } + return iofile; +} + +/* +iof_file * iof_file_writer_from_file (iof_file *iofile, const char *filename) +{ + FILE *file; + if ((file = fopen(filename, "wb")) == NULL) + return NULL; + if (iofile == NULL) + iofile = iof_file_new(file); + else + iof_file_init(iofile, file); + iofile->flags |= IOF_CLOSE_FILE; + iof_file_set_name(iofile, filename); + return iofile; +} +*/ + +/* +Because of limited number of FILE* handles available, we may need to close contained handle +between accessing it. In applications so far (fonts, images) we typically need the source +to parse the file on creation and to rewrite or reload the data on dump. All iof_file api +functions assume that iofile has FILE* opened. Reopening it on every access (ftell, fseek, +read/write) makes no sense, as we would effectively loose control. If the caller invalidates +iofile by closing and nulling its file handle, it is also responsible to reopen when necessary. +*/ + +int iof_file_close_input (iof_file *iofile) +{ + FILE *file; + if (iofile->flags & IOF_DATA) + return 0; + if ((file = iof_file_get_fh(iofile)) == NULL) + return 0; + fclose(file); + iof_file_set_fh(iofile, NULL); + iofile->flags &= ~IOF_RECLOSE_FILE; + iofile->flags |= IOF_REOPEN_FILE; + return 1; +} + +int iof_file_reopen_input (iof_file *iofile) +{ // returns true if iofile readable + FILE *file; + const char *filename; + if (iofile->flags & IOF_DATA) + return 1; + if ((file = iof_file_get_fh(iofile)) != NULL) + return 1; // if present, assumed readable + if ((filename = iofile->name) == NULL || (file = fopen(filename, "rb")) == NULL) + return 0; + iof_file_set_fh(iofile, file); + iofile->flags &= ~IOF_REOPEN_FILE; + iofile->flags |= IOF_RECLOSE_FILE; + return 1; +} + +/* freeing iof_file */ + +void iof_file_free (iof_file *iofile) +{ + FILE *file; + if (iofile->flags & IOF_DATA) + { + if (iofile->flags & IOF_BUFFER_ALLOC) + { + iofile->flags &= ~IOF_BUFFER_ALLOC; + if (iofile->buf != NULL) + { + util_free(iofile->buf); + iofile->buf = iofile->pos = iofile->end = NULL; + } + } + } + else if ((file = iof_file_get_fh(iofile)) != NULL) + { + if (iofile->flags & IOF_CLOSE_FILE) + fclose(file); + iof_file_set_fh(iofile, NULL); + } + iof_file_set_name(iofile, NULL); + if (iofile->flags & IOF_ALLOC) + util_free(iofile); +} + +/* set filename for reopen */ + +void iof_file_set_name (iof_file *iofile, const char *name) +{ + if (iofile->name != NULL) + util_free(iofile->name); + if (name != NULL) + iofile->name = iof_copy_data(name, strlen(name) + 1); + else + iofile->name = NULL; +} + +/* seek */ + +int iof_file_seek (iof_file *iofile, long offset, int whence) +{ + if (iofile->flags & IOF_DATA) + { + switch (whence) + { + case SEEK_SET: + if (offset >= 0 && iofile->buf + offset <= iofile->end) + { + iofile->pos = iofile->buf + offset; + return 0; + } + return -1; + case SEEK_CUR: + if ((offset >= 0 && iofile->pos + offset <= iofile->end) || (offset < 0 && iofile->pos + offset >= iofile->buf)) + { + iofile->pos += offset; + return 0; + } + return -1; + case SEEK_END: + if (offset <= 0 && iofile->end + offset >= iofile->buf) + { + iofile->pos = iofile->end + offset; + return 0; + } + return -1; + } + return -1; + } + return fseek(iof_file_get_fh(iofile), offset, whence); +} + +/* */ + +long iof_file_tell (iof_file *iofile) +{ + return (iofile->flags & IOF_DATA) ? (long)(iofile->pos - iofile->buf) : ftell(iof_file_get_fh(iofile)); +} + +size_t iof_file_size (iof_file *iofile) +{ + long pos, size; + FILE *file; + if (iofile->flags & IOF_DATA) + return (size_t)iof_space(iofile); + file = iof_file_get_fh(iofile); + pos = ftell(file); + fseek(file, 0, SEEK_END); + size = ftell(file); + fseek(file, pos, SEEK_SET); + return size; +} + +int iof_file_eof (iof_file *iofile) +{ + if (iofile->flags & IOF_DATA) + return iofile->pos == iofile->end ? -1 : 0; + return feof(iof_file_get_fh(iofile)); +} + +int iof_file_flush (iof_file *iofile) +{ + if (iofile->flags & IOF_DATA) + return 0; + return fflush(iof_file_get_fh(iofile)); +} + +size_t iof_file_read (void *ptr, size_t size, size_t items, iof_file *iofile) +{ + if (iofile->flags & IOF_DATA) + { + size_t bytes = size * items; + if (bytes > (size_t)iof_left(iofile)) + bytes = (size_t)iof_left(iofile); + memcpy(ptr, iofile->pos, bytes); + iofile->pos += bytes; + return bytes / size; // number of elements read + } + return fread(ptr, size, items, iof_file_get_fh(iofile)); +} + +static size_t iof_file_data_resizeto (iof_file *iofile, size_t space) +{ + uint8_t *newbuf; + size_t size; + size = iof_size(iofile); + if (iofile->flags & IOF_BUFFER_ALLOC) + { + newbuf = (uint8_t *)util_realloc(iofile->buf, space); + } + else + { + newbuf = (uint8_t *)util_malloc(space); + if (size > 0) + memcpy(newbuf, iofile->buf, size); + iofile->flags |= IOF_BUFFER_ALLOC; + } + iofile->buf = newbuf; + iofile->pos = newbuf + size; + iofile->end = newbuf + space; + return space - size; +} + +#define iof_file_data_resize(iofile) iof_file_data_resizeto(iofile, iof_space(iofile) << 1) + +size_t iof_file_write (const void *ptr, size_t size, size_t items, iof_file *iofile) +{ + if (iofile->flags & IOF_DATA) + { + size_t space, sizesofar, bytes; + bytes = size * items; + if (bytes > (size_t)iof_left(iofile)) + { + if ((space = iof_space(iofile)) == 0) // allow iofile->buf/end initially NULL + space = BUFSIZ; + for (sizesofar = iof_size(iofile), space <<= 1; sizesofar + bytes > space; space <<= 1) + ; + if (iof_file_data_resizeto(iofile, space) == 0) + return 0; + } + memcpy(iofile->pos, ptr, bytes); + iofile->pos += bytes; + return bytes / size; + } + return fwrite(ptr, size, items, iof_file_get_fh(iofile)); +} + +size_t iof_file_ensure (iof_file *iofile, size_t bytes) +{ + if (iofile->flags & IOF_DATA) + { + size_t space, sizesofar, left; + left = (size_t)iof_left(iofile); + if (bytes > left) + { + if ((space = iof_space(iofile)) == 0) // allow iofile->buf/end initially NULL + space = BUFSIZ; + for (sizesofar = iof_size(iofile), space <<= 1; sizesofar + bytes > space; space <<= 1); + return iof_file_data_resizeto(iofile, space); + } + return left; + } + return 0; +} + +int iof_file_getc (iof_file *iofile) +{ + if (iofile->flags & IOF_DATA) + return iofile->pos < iofile->end ? *iofile->pos++ : IOFEOF; + return fgetc(iof_file_get_fh(iofile)); +} + +int iof_file_putc (iof_file *iofile, int c) +{ + if (iofile->flags & IOF_DATA) + { + if (iofile->pos >= iofile->end) + if (iof_file_data_resize(iofile) == 0) + return IOFEOF; + *iofile->pos++ = (uint8_t)c; + return c; + } + return fputc(c, iof_file_get_fh(iofile)); +} + +static int iof_file_sync (iof_file *iofile, size_t *offset) +{ + if (iofile->offset != offset) + { + if (iofile->offset != NULL) + *iofile->offset = iof_file_tell(iofile); + iofile->offset = offset; + if (offset) // let offset be NULL + return iof_file_seek(iofile, (long)*offset, SEEK_SET); + } + return 0; +} + +//#define iof_file_unsync(iofile, poffset) (void)((iofile)->offset == poffset && (((iofile)->offset = NULL), 0)) +#define iof_file_unsync(iofile, poffset) ((void)poffset, (iofile)->offset = NULL) + +/* iof seek */ + +#define iof_reader_reset(I) ((I)->pos = (I)->end = (I)->buf) +#define iof_reader_reseek_file(I, offset, whence) (fseek((I)->file, offset, whence) == 0 ? (iof_reader_reset(I), 0) : -1) +#define iof_reader_reseek_iofile(I, offset, whence) (iof_file_seek((I)->iofile, offset, whence) == 0 ? (iof_reader_reset(I), 0) : -1) + +#define iof_writer_reset(O) ((O)->pos = (O)->buf) +#define iof_writer_reseek_file(O, offset, whence) (iof_flush(O), (fseek((O)->file, offset, whence) == 0 ? (iof_writer_reset(O), 0) : -1)) +#define iof_writer_reseek_iofile(O, offset, whence) (iof_flush(O), (iof_file_seek((O)->iofile, offset, whence) == 0 ? (iof_writer_reset(O), 0) : -1)) + +static int iof_reader_seek_data (iof *I, long offset, int whence) +{ + switch (whence) + { + case SEEK_SET: + if (offset >= 0 && I->buf + offset <= I->end) + { + I->pos = I->buf + offset; + return 0; + } + return -1; + case SEEK_CUR: + if ((offset >= 0 && I->pos + offset <= I->end) || (offset < 0 && I->pos + offset >= I->buf)) + { + I->pos += offset; + return 0; + } + return -1; + case SEEK_END: + if (offset <= 0 && I->end + offset >= I->buf) + { + I->pos = I->end + offset; + return 0; + } + return -1; + } + return -1; +} + +static int iof_reader_seek_iofile (iof *I, long offset, int whence) +{ + long fileoffset; + switch (whence) + { + case SEEK_SET: + fileoffset = iof_file_tell(I->iofile); + if (offset <= fileoffset && offset >= fileoffset - iof_space(I)) + { + I->pos = I->end - (fileoffset - offset); + return 0; + } + return iof_reader_reseek_iofile(I, offset, SEEK_SET); + case SEEK_CUR: + if ((offset >= 0 && I->pos + offset <= I->end) || (offset < 0 && I->pos + offset >= I->buf)) + { + I->pos += offset; + return 0; + } + return iof_reader_reseek_iofile(I, offset, SEEK_CUR); + case SEEK_END: + return iof_reader_reseek_iofile(I, offset, SEEK_END); // can we do better? + } + return -1; +} + +static int iof_reader_seek_file (iof *I, long offset, int whence) +{ + long fileoffset; + switch (whence) + { + case SEEK_SET: + fileoffset = ftell(I->file); + if (offset <= fileoffset && offset >= fileoffset - iof_space(I)) + { + I->pos = I->end - (fileoffset - offset); + return 0; + } + return iof_reader_reseek_file(I, offset, SEEK_SET); + case SEEK_CUR: + if ((offset >= 0 && I->pos + offset <= I->end) || (offset < 0 && I->pos + offset >= I->buf)) + { + I->pos += offset; + return 0; + } + return iof_reader_reseek_file(I, offset, SEEK_CUR); + case SEEK_END: + return iof_reader_reseek_file(I, offset, SEEK_END); // can we do better? + } + return -1; +} + +int iof_reader_seek (iof *I, long offset, int whence) +{ + I->flags &= ~IOF_STOPPED; + if (I->flags & IOF_FILE) + return iof_reader_seek_iofile(I, offset, whence); + if (I->flags & IOF_FILE_HANDLE) + return iof_reader_seek_file(I, offset, whence); + if (I->flags & IOF_DATA) + return iof_reader_seek_data(I, offset, whence); + return -1; +} + +int iof_reader_reseek (iof *I, long offset, int whence) +{ + I->flags &= ~IOF_STOPPED; + if (I->flags & IOF_FILE) + return iof_reader_reseek_iofile(I, offset, whence); + if (I->flags & IOF_FILE_HANDLE) + return iof_reader_reseek_file(I, offset, whence); + if (I->flags & IOF_DATA) + return iof_reader_seek_data(I, offset, whence); + return -1; +} + +static int iof_writer_seek_data (iof *O, long offset, int whence) +{ + /* + fseek() allows to seek after the end of file. Seeking does not increase the output file. + No byte is written before fwirte(). It seems to fill the gap with zeros. Until we really need that, + no seeking out of bounds for writers. + */ + O->flags &= ~IOF_STOPPED; + return iof_reader_seek_data(O, offset, whence); +} + +static int iof_writer_seek_iofile (iof *O, long offset, int whence) +{ + long fileoffset; + switch (whence) + { + case SEEK_SET: + fileoffset = iof_file_tell(O->iofile); + if (offset >= fileoffset && offset <= fileoffset + iof_space(O)) + { + O->pos = O->buf + (offset - fileoffset); + return 0; + } + return iof_writer_reseek_iofile(O, offset, SEEK_SET); + case SEEK_CUR: + if ((offset >=0 && O->pos + offset <= O->end) || (offset < 0 && O->pos + offset >= O->buf)) + { + O->pos += offset; + return 0; + } + return iof_writer_reseek_iofile(O, offset, SEEK_CUR); + case SEEK_END: + return iof_writer_reseek_iofile(O, offset, SEEK_END); + } + return -1; +} + +static int iof_writer_seek_file (iof *O, long offset, int whence) +{ + long fileoffset; + switch (whence) + { + case SEEK_SET: + fileoffset = ftell(O->file); + if (offset >= fileoffset && offset <= fileoffset + iof_space(O)) + { + O->pos = O->buf + (offset - fileoffset); + return 0; + } + return iof_writer_reseek_file(O, offset, SEEK_SET); + case SEEK_CUR: + if ((offset >=0 && O->pos + offset <= O->end) || (offset < 0 && O->pos + offset >= O->buf)) + { + O->pos += offset; + return 0; + } + return iof_writer_reseek_file(O, offset, SEEK_CUR); + case SEEK_END: + return iof_writer_reseek_file(O, offset, SEEK_END); + } + return -1; +} + +int iof_writer_seek (iof *I, long offset, int whence) +{ + I->flags &= ~IOF_STOPPED; + if (I->flags & IOF_FILE) + return iof_writer_seek_iofile(I, offset, whence); + if (I->flags & IOF_FILE_HANDLE) + return iof_writer_seek_file(I, offset, whence); + if (I->flags & IOF_DATA) + return iof_writer_seek_data(I, offset, whence); + return -1; +} + +int iof_writer_reseek (iof *I, long offset, int whence) +{ + I->flags &= ~IOF_STOPPED; + if (I->flags & IOF_FILE) + return iof_writer_reseek_iofile(I, offset, whence); + if (I->flags & IOF_FILE_HANDLE) + return iof_writer_reseek_file(I, offset, whence); + if (I->flags & IOF_DATA) + return iof_writer_seek_data(I, offset, whence); + return -1; +} + +int iof_seek (iof *F, long offset, int whence) +{ + return (F->flags & IOF_WRITER) ? iof_writer_seek(F, offset, whence) : iof_reader_seek(F, offset, whence); +} + +int iof_reseek (iof *F, long offset, int whence) +{ + return (F->flags & IOF_WRITER) ? iof_writer_reseek(F, offset, whence) : iof_reader_reseek(F, offset, whence); +} + +/* tell */ + +long iof_reader_tell (iof *I) +{ + if (I->flags & IOF_FILE) + return iof_file_tell(I->iofile) - (long)iof_left(I); + if (I->flags & IOF_FILE_HANDLE) + return ftell(I->file) - (long)iof_left(I); + //if (I->flags & IOF_DATA) + return (long)iof_size(I); +} + +long iof_writer_tell (iof *O) +{ + if (O->flags & IOF_FILE) + return iof_file_tell(O->iofile) + (long)iof_size(O); + if (O->flags & IOF_FILE_HANDLE) + return ftell(O->file) + (long)iof_size(O); + //if (I->flags & IOF_DATA) + return (long)iof_size(O); +} + +long iof_tell (iof *I) +{ + return (I->flags & IOF_WRITER) ? iof_writer_tell(I) : iof_reader_tell(I); +} + +size_t iof_fsize (iof *I) +{ + size_t pos, size; + if (I->flags & IOF_FILE) + return iof_file_size(I->iofile); + if (I->flags & IOF_FILE_HANDLE) + { + pos = (size_t)ftell(I->file); + fseek(I->file, 0, SEEK_END); + size = (size_t)ftell(I->file); + fseek(I->file, (long)pos, SEEK_SET); + return size; + } + //if (I->flags & IOF_DATA) + return (size_t)iof_space(I); +} + +/* save reader tail */ + +size_t iof_save_tail (iof *I) +{ + size_t size, left; + size = iof_size(I); + left = iof_left(I); + if (size >= left) + memcpy(I->buf, I->pos, left); + else + memmove(I->buf, I->pos, left); + return left; +} + +size_t iof_input_save_tail (iof *I, size_t back) +{ + size_t size; + I->flags |= IOF_TAIL; + I->pos -= back; + size = iof_input(I); + I->pos += back; + I->flags &= ~IOF_TAIL; + return size; // + back - back +} + +/* read from file */ + +/* iof free*/ + +static size_t file_read (iof *I); +static size_t file_load (iof *I); + +static size_t file_reader (iof *I, iof_mode mode) +{ + switch (mode) + { + case IOFREAD: + return file_read(I); + case IOFLOAD: + return file_load(I); + case IOFCLOSE: + iof_free(I); + return 0; + default: + return 0; + } +} + +iof * iof_setup_file_handle_reader (iof *I, void *buffer, size_t space, FILE *f) +{ + if (I == NULL) + iof_setup_reader(I, buffer, space); + else + iof_reader_buffer(I, buffer, space); + iof_setup_file(I, f); + I->more = file_reader; + return I; +} + +iof * iof_setup_file_reader (iof *I, void *buffer, size_t space, const char *filename) +{ + FILE *f; + if ((f = fopen(filename, "rb")) == NULL) + return NULL; + if (I == NULL) + iof_setup_reader(I, buffer, space); + else + iof_reader_buffer(I, buffer, space); + iof_setup_file(I, f); + I->flags |= IOF_CLOSE_FILE; + I->more = file_reader; + return I; +} + +/* write to file */ + +static size_t file_write (iof *O, int flush); + +static size_t file_writer (iof *O, iof_mode mode) +{ + switch (mode) + { + case IOFWRITE: + return file_write(O, 0); + case IOFFLUSH: + return file_write(O, 1); + case IOFCLOSE: + file_write(O, 1); + iof_free(O); + return 0; + default: + return 0; + } +} + +iof * iof_setup_file_handle_writer (iof *O, void *buffer, size_t space, FILE *f) +{ + if (O == NULL) + iof_setup_writer(O, buffer, space); + else + iof_writer_buffer(O, buffer, space); + iof_setup_file(O, f); + O->more = file_writer; + return O; +} + +iof * iof_setup_file_writer (iof *O, void *buffer, size_t space, const char *filename) +{ + FILE *f; + if ((f = fopen(filename, "wb")) == NULL) + return NULL; + if (O == NULL) + iof_setup_writer(O, buffer, space); + else + iof_writer_buffer(O, buffer, space); + iof_setup_file(O, f); + O->flags |= IOF_CLOSE_FILE; + O->more = file_writer; + return O; +} + +/* a dedicated handler for stdout/stderr */ + +static size_t stdout_writer (iof *O, iof_mode mode) +{ + switch(mode) + { + case IOFWRITE: + { + fwrite(O->buf, sizeof(uint8_t), iof_size(O), stdout); + O->pos = O->buf; + return O->space; + } + case IOFCLOSE: + case IOFFLUSH: + { + fwrite(O->buf, sizeof(uint8_t), iof_size(O), stdout); + fflush(stdout); + O->pos = O->buf; + return 0; + } + default: + break; + } + return 0; +} + +static size_t stderr_writer (iof *O, iof_mode mode) +{ + switch(mode) + { + case IOFWRITE: + { + fwrite(O->buf, sizeof(uint8_t), iof_size(O), stderr); + O->pos = O->buf; + return O->space; + } + case IOFCLOSE: + case IOFFLUSH: + { + fwrite(O->buf, sizeof(uint8_t), iof_size(O), stderr); + fflush(stderr); + O->pos = O->buf; + return 0; + } + default: + break; + } + return 0; +} + +static uint8_t iof_stdout_buffer[BUFSIZ]; +iof iof_stdout = IOF_WRITER_STRUCT(stdout_writer, NULL, iof_stdout_buffer, BUFSIZ, 0); + +static uint8_t iof_stderr_buffer[BUFSIZ]; +iof iof_stderr = IOF_WRITER_STRUCT(stderr_writer, NULL, iof_stderr_buffer, BUFSIZ, 0); + +/* read from somewhere */ + +iof * iof_reader (iof *I, void *link, iof_handler reader, const void *m, size_t bytes) +{ + I->space = 0; + I->link = link; + I->more = reader; + I->flags = 0; + I->refcount = 0; + if (m != NULL) + { + I->rbuf = I->rpos = (const uint8_t *)m; + I->rend = (const uint8_t *)m + bytes; + return I; + } + return NULL; +} + +iof * iof_string_reader (iof *I, const void *s, size_t bytes) +{ + I->space = 0; + I->link = NULL; + I->more = NULL; + I->flags = 0; // iof_string() sets IOF_DATA + I->refcount = 0; + if (s != NULL) + return iof_string(I, s, bytes); + return NULL; +} + +/* write somewhere */ + +iof * iof_writer (iof *O, void *link, iof_handler writer, void *m, size_t bytes) +{ + O->space = 0; + O->link = link; + O->more = writer; + O->flags = 0; + O->refcount = 0; + if (m != NULL && bytes > 0) + { + O->buf = O->pos = (uint8_t *)m; + O->end = (uint8_t *)m + bytes; + return O; + } + // return iof_null(O); + return NULL; +} + +/* write to growing bytes buffer */ + +static size_t iof_mem_handler (iof *O, iof_mode mode) +{ + switch(mode) + { + case IOFWRITE: + return iof_resize_buffer(O); + case IOFCLOSE: + iof_free(O); + return 0; + default: + return 0; + } +} + +iof * iof_setup_buffer (iof *O, void *buffer, size_t space) +{ + if (O == NULL) + iof_setup_writer(O, buffer, space); + else + iof_writer_buffer(O, buffer, space); + O->link = NULL; + O->flags |= IOF_DATA; + O->more = iof_mem_handler; + return O; +} + +iof * iof_setup_buffermin (iof *O, void *buffer, size_t space, size_t min) +{ + if ((O = iof_setup_buffer(O, buffer, space)) != NULL && space < min) // just allocate min now to avoid further rewriting + { + O->buf = O->pos = (uint8_t *)util_malloc(min); + O->flags |= IOF_BUFFER_ALLOC; + O->end = O->buf + min; + } + return O; +} + +iof * iof_buffer_create (size_t space) +{ + uint8_t *buffer; + iof *O; + space += sizeof(iof); + buffer = util_malloc(space); + if ((O = iof_setup_buffer(NULL, buffer, space)) != NULL) + O->flags |= IOF_ALLOC; + return O; +} + +/* set/get */ + +int iof_getc (iof *I) +{ + if (iof_readable(I)) + return *I->pos++; + return IOFEOF; +} + +int iof_putc (iof *O, int u) +{ + if (iof_writable(O)) + { + iof_set(O, u); + return (uint8_t)u; + } + return IOFFULL; +} + +size_t iof_skip (iof *I, size_t bytes) +{ + while (bytes) + { + if (iof_readable(I)) + ++I->pos; + else + break; + --bytes; + } + return bytes; +} + +/* from iof to iof */ + +iof_status iof_pass (iof *I, iof *O) +{ + size_t leftin, leftout; + if ((leftin = iof_left(I)) == 0) + leftin = iof_input(I); + while (leftin) + { + if ((leftout = iof_left(O)) == 0) + if ((leftout = iof_output(O)) == 0) + return IOFFULL; + while (leftin > leftout) + { + memcpy(O->pos, I->pos, leftout); + I->pos += leftout; + O->pos = O->end; /* eq. += leftout */ + leftin -= leftout; + if ((leftout = iof_output(O)) == 0) + return IOFFULL; + } + if (leftin) + { + memcpy(O->pos, I->pos, leftin); + I->pos = I->end; /* eq. += leftin */ + O->pos += leftin; + } + leftin = iof_input(I); + } + return IOFEOF; +} + +/* read n-bytes */ + +size_t iof_read (iof *I, void *to, size_t size) +{ + size_t leftin, done = 0; + char *s = (char *)to; + + if ((leftin = iof_left(I)) == 0) + if ((leftin = iof_input(I)) == 0) + return done; + while (size > leftin) + { + memcpy(s, I->pos, leftin * sizeof(uint8_t)); + size -= leftin; + done += leftin; + s += leftin; + I->pos = I->end; + if ((leftin = iof_input(I)) == 0) + return done; + } + if (size) + { + memcpy(s, I->pos, size * sizeof(uint8_t)); + I->pos += size; + done += size; + } + return done; +} + +/* rewrite FILE content (use fseek if needed) */ + +size_t iof_write_file_handle (iof *O, FILE *file) +{ + size_t leftout, size, readout; + if ((leftout = iof_left(O)) == 0) + if ((leftout = iof_output(O)) == 0) + return 0; + size = 0; + do { + readout = fread(O->pos, 1, leftout, file); + O->pos += readout; + size += readout; + } while(readout == leftout && (leftout = iof_output(O)) > 0); + return size; +} + +size_t iof_write_file (iof *O, const char *filename) +{ + FILE *file; + size_t size; + if ((file = fopen(filename, "rb")) == NULL) + return 0; + size = iof_write_file_handle(O, file); + fclose(file); + return size; +} + +size_t iof_write_iofile (iof *O, iof_file *iofile, int savepos) +{ + long offset; + size_t size; + FILE *file; + if (iofile->flags & IOF_DATA) + return iof_write(O, iofile->pos, (size_t)(iofile->end - iofile->pos)); + file = iof_file_get_fh(iofile); + if (savepos) + { + offset = ftell(file); + size = iof_write_file_handle(O, file); + fseek(file, offset, SEEK_SET); + return size; + } + return iof_write_file_handle(O, file); +} + +/* write n-bytes */ + +size_t iof_write (iof *O, const void *data, size_t size) +{ + size_t leftout, done = 0; + const char *s = (const char *)data; + if ((leftout = iof_left(O)) == 0) + if ((leftout = iof_output(O)) == 0) + return done; + while (size > leftout) + { + memcpy(O->pos, s, leftout * sizeof(uint8_t)); + size -= leftout; + done += leftout; + s += leftout; + O->pos = O->end; + if ((leftout = iof_output(O)) == 0) + return done; + } + if (size) + { + memcpy(O->pos, s, size * sizeof(uint8_t)); + O->pos += size; + done += size; + } + return done; +} + +/* write '\0'-terminated string */ + +iof_status iof_puts (iof *O, const void *data) +{ + const char *s = (const char *)data; + while (*s) + { + if (iof_writable(O)) + iof_set(O, *s++); + else + return IOFFULL; + } + return IOFEOF; // ? +} + +size_t iof_put_string (iof *O, const void *data) +{ + const char *p, *s = (const char *)data; + for (p = s; *p != '\0' && iof_writable(O); iof_set(O, *p++)); + return p - s; +} + +/* write byte n-times */ + +/* +iof_status iof_repc (iof *O, char c, size_t bytes) +{ + while (bytes) + { + if (iof_writable(O)) + iof_set(O, c); + else + return IOFFULL; + --bytes; + } + return IOFEOF; // ? +} +*/ + +size_t iof_repc (iof *O, char c, size_t bytes) +{ + size_t leftout, todo = bytes; + if ((leftout = iof_left(O)) == 0) + if ((leftout = iof_output(O)) == 0) + return 0; + while (bytes > leftout) + { + memset(O->pos, c, leftout); + bytes -= leftout; + O->pos = O->end; + if ((leftout = iof_output(O)) == 0) + return todo - bytes; + } + if (bytes) + { + memset(O->pos, c, bytes); + O->pos += bytes; + } + return todo; +} + +/* putfs */ + +#define IOF_FMT_SIZE 1024 + +size_t iof_putfs (iof *O, const char *format, ...) +{ + static char buffer[IOF_FMT_SIZE]; + va_list args; + va_start(args, format); + if (vsnprintf(buffer, IOF_FMT_SIZE, format, args) > 0) + { + va_end(args); + return iof_put_string(O, buffer); + } + else + { + va_end(args); + return iof_write(O, buffer, IOF_FMT_SIZE); + } +} + +/* integer from iof; return 1 on success, 0 otherwise */ + +int iof_get_int32 (iof *I, int32_t *number) +{ + int sign, c = iof_char(I); + iof_scan_sign(I, c, sign); + if (!base10_digit(c)) return 0; + iof_read_integer(I, c, *number); + if (sign) *number = -*number; + return 1; +} + +int iof_get_intlw (iof *I, intlw_t *number) +{ + int sign, c = iof_char(I); + iof_scan_sign(I, c, sign); + if (!base10_digit(c)) return 0; + iof_read_integer(I, c, *number); + if (sign) *number = -*number; + return 1; +} + +int iof_get_int64 (iof *I, int64_t *number) +{ + int sign, c = iof_char(I); + iof_scan_sign(I, c, sign); + if (!base10_digit(c)) return 0; + iof_read_integer(I, c, *number); + if (sign) *number = -*number; + return 1; +} + +int iof_get_uint32 (iof *I, uint32_t *number) +{ + int c = iof_char(I); + if (!base10_digit(c)) return 0; + iof_read_integer(I, c, *number); + return 1; +} + +int iof_get_uintlw (iof *I, uintlw_t *number) +{ + int c = iof_char(I); + if (!base10_digit(c)) return 0; + iof_read_integer(I, c, *number); + return 1; +} + +int iof_get_uint64 (iof *I, uint64_t *number) +{ + int c = iof_char(I); + if (!base10_digit(c)) return 0; + iof_read_integer(I, c, *number); + return 1; +} + +int iof_get_int32_radix (iof *I, int32_t *number, int radix) +{ + int sign, c = iof_char(I); + iof_scan_sign(I, c, sign); + if (!base10_digit(c)) return 0; + iof_read_radix(I, c, *number, radix); + if (sign) *number = -*number; + return 1; + +} + +int iof_get_intlw_radix (iof *I, intlw_t *number, int radix) +{ + int sign, c = iof_char(I); + iof_scan_sign(I, c, sign); + if (!base10_digit(c)) return 0; + iof_read_radix(I, c, *number, radix); + if (sign) *number = -*number; + return 1; +} + +int iof_get_int64_radix (iof *I, int64_t *number, int radix) +{ + int sign, c = iof_char(I); + iof_scan_sign(I, c, sign); + if (!base10_digit(c)) return 0; + iof_read_radix(I, c, *number, radix); + if (sign) *number = -*number; + return 1; +} + +int iof_get_uint32_radix (iof *I, uint32_t *number, int radix) +{ + int c = iof_char(I); + if (!base10_digit(c)) return 0; + iof_read_radix(I, c, *number, radix); + return 1; +} + +int iof_get_uintlw_radix (iof *I, uintlw_t *number, int radix) +{ + int c = iof_char(I); + if (!base10_digit(c)) return 0; + iof_read_radix(I, c, *number, radix); + return 1; +} + +int iof_get_uint64_radix (iof *I, uint64_t *number, int radix) +{ + int c = iof_char(I); + if (!base10_digit(c)) return 0; + iof_read_radix(I, c, *number, radix); + return 1; +} + +/* get roman to uint16_t, cf. roman_to_uint16() from utilnumber.c*/ + +/* todo: some trick in place of this macro horror? */ + +#define roman1000(c) (c == 'M' || c == 'm') +#define roman500(c) (c == 'D' || c == 'd') +#define roman100(c) (c == 'C' || c == 'c') +#define roman50(c) (c == 'L' || c == 'l') +#define roman10(c) (c == 'X' || c == 'x') +#define roman5(c) (c == 'V' || c == 'v') +#define roman1(c) (c == 'I' || c == 'i') + +#define roman100s(I, c) \ + (roman100(c) ? (100 + ((c = iof_next(I), roman100(c)) ? (100 + ((c = iof_next(I), roman100(c)) ? (c = iof_next(I), 100) : 0)) : 0)) : 0) +#define roman10s(I, c) \ + (roman10(c) ? (10 + ((c = iof_next(I), roman10(c)) ? (10 + ((c = iof_next(I), roman10(c)) ? (c = iof_next(I), 10) : 0)) : 0)) : 0) +#define roman1s(I, c) \ + (roman1(c) ? (1 + ((c = iof_next(I), roman1(c)) ? (1 + ((c = iof_next(I), roman1(c)) ? (c = iof_next(I), 1) : 0)) : 0)) : 0) + +int iof_get_roman (iof *I, uint16_t *number) +{ + int c; + /* M */ + for (*number = 0, c = iof_char(I); roman1000(c); *number += 1000, c = iof_next(I)); + /* D C */ + if (roman500(c)) + { + c = iof_next(I); + *number += 500 + roman100s(I, c); + } + else if (roman100(c)) + { + c = iof_next(I); + if (roman1000(c)) + { + c = iof_next(I); + *number += 900; + } + else if (roman500(c)) + { + c = iof_next(I); + *number += 400; + } + else + *number += 100 + roman100s(I, c); + } + /* L X */ + if (roman50(c)) + { + c = iof_next(I); + *number += 50 + roman10s(I, c); + } + else if (roman10(c)) + { + c = iof_next(I); + if (roman100(c)) + { + c = iof_next(I); + *number += 90; + } + else if (roman50(c)) + { + c = iof_next(I); + *number += 40; + } + else + *number += 10 + roman10s(I, c); + } + /* V I */ + if (roman5(c)) + { + c = iof_next(I); + *number += 5 + roman1s(I, c); + } + else if (roman1(c)) + { + c = iof_next(I); + if (roman10(c)) + { + c = iof_next(I); + *number += 9; + } + else if (roman5(c)) + { + c = iof_next(I); + *number += 4; + } + else + *number += 1 + roman1s(I, c); + } + return 1; +} + +/* double from iof; return 1 on success */ + +int iof_get_double (iof *I, double *number) // cf. string_to_double() +{ + int sign, exponent10, c = iof_char(I); + iof_scan_sign(I, c, sign); + iof_scan_decimal(I, c, *number); + if (c == '.') + { + c = iof_next(I); + iof_scan_fraction(I, c, *number, exponent10); + } + else + exponent10 = 0; + if (c == 'e' || c == 'E') + { + c = iof_next(I); + iof_scan_exponent10(I, c, exponent10); + } + double_exp10(*number, exponent10); + if (sign) *number = -*number; + return 1; +} + +int iof_get_float (iof *I, float *number) // cf. string_to_float() in utilnumber.c +{ + int sign, exponent10, c = iof_char(I); + iof_scan_sign(I, c, sign); + iof_scan_decimal(I, c, *number); + if (c == '.') + { + c = iof_next(I); + iof_scan_fraction(I, c, *number, exponent10); + } + else + exponent10 = 0; + if (c == 'e' || c == 'E') + { + c = iof_next(I); + iof_scan_exponent10(I, c, exponent10); + } + float_exp10(*number, exponent10); + if (sign) *number = -*number; + return 1; +} + +int iof_conv_double (iof *I, double *number) // cf. convert_to_double() in utilnumber.c +{ + int sign, exponent10, c = iof_char(I); + iof_scan_sign(I, c, sign); + iof_scan_decimal(I, c, *number); + if (c == '.' || c == ',') + { + c = iof_next(I); + iof_scan_fraction(I, c, *number, exponent10); + if (exponent10 < 0) + double_negative_exp10(*number, exponent10); + } + if (sign) *number = -*number; + return 1; +} + +int iof_conv_float (iof *I, float *number) // cf. convert_to_float() +{ + int sign, exponent10, c = iof_char(I); + iof_scan_sign(I, c, sign); + iof_scan_decimal(I, c, *number); + if (c == '.' || c == ',') + { + c = iof_next(I); + iof_scan_fraction(I, c, *number, exponent10); + if (exponent10 < 0) + float_negative_exp10(*number, exponent10); + } + if (sign) *number = -*number; + return 1; +} + +/* integer to iof; return a number of written bytes */ + +#define iof_copy_number_buffer(O, s, p) for (p = s; *p && iof_writable(O); iof_set(O, *p), ++p) + +size_t iof_put_int32 (iof *O, int32_t number) +{ + const char *s, *p; + s = int32_to_string(number); + iof_copy_number_buffer(O, s, p); + return p - s; +} + +size_t iof_put_intlw (iof *O, intlw_t number) +{ + const char *s, *p; + s = intlw_to_string(number); + iof_copy_number_buffer(O, s, p); + return p - s; +} + +size_t iof_put_int64 (iof *O, int64_t number) +{ + const char *s, *p; + s = int64_to_string(number); + iof_copy_number_buffer(O, s, p); + return p - s; +} + +size_t iof_put_uint32 (iof *O, uint32_t number) +{ + const char *s, *p; + s = uint32_to_string(number); + iof_copy_number_buffer(O, s, p); + return p - s; +} + +size_t iof_put_uintlw (iof *O, uintlw_t number) +{ + const char *s, *p; + s = uintlw_to_string(number); + iof_copy_number_buffer(O, s, p); + return p - s; +} + +size_t iof_put_uint64 (iof *O, uint64_t number) +{ + const char *s, *p; + s = uint64_to_string(number); + iof_copy_number_buffer(O, s, p); + return p - s; +} + +size_t iof_put_int32_radix (iof *O, int32_t number, int radix) +{ + const char *s, *p; + s = int32_to_radix(number, radix); + iof_copy_number_buffer(O, s, p); + return p - s; +} + +size_t iof_put_intlw_radix (iof *O, intlw_t number, int radix) +{ + const char *s, *p; + s = intlw_to_radix(number, radix); + iof_copy_number_buffer(O, s, p); + return p - s; +} + +size_t iof_put_int64_radix (iof *O, int64_t number, int radix) +{ + const char *s, *p; + s = int64_to_radix(number, radix); + iof_copy_number_buffer(O, s, p); + return p - s; +} + +size_t iof_put_uint32_radix (iof *O, uint32_t number, int radix) +{ + const char *s, *p; + s = uint32_to_radix(number, radix); + iof_copy_number_buffer(O, s, p); + return p - s; +} + +size_t iof_put_uintlw_radix (iof *O, uintlw_t number, int radix) +{ + const char *s, *p; + s = uintlw_to_radix(number, radix); + iof_copy_number_buffer(O, s, p); + return p - s; +} + +size_t iof_put_uint64_radix (iof *O, uint64_t number, int radix) +{ + const char *s, *p; + s = uint64_to_radix(number, radix); + iof_copy_number_buffer(O, s, p); + return p - s; +} + +/* roman numerals */ + +size_t iof_put_roman_uc (iof *O, uint16_t number) +{ + const char *s, *p; + s = uint16_to_roman_uc(number); + iof_copy_number_buffer(O, s, p); + return p - s; +} + +size_t iof_put_roman_lc (iof *O, uint16_t number) +{ + const char *s, *p; + s = uint16_to_roman_lc(number); + iof_copy_number_buffer(O, s, p); + return p - s; +} + +/* double/float to iof; return the number of written bytes */ + +size_t iof_put_double (iof *O, double number, int digits) +{ + const char *s, *p; + s = double_to_string(number, digits); + iof_copy_number_buffer(O, s, p); + return p - s; +} + +size_t iof_put_float (iof *O, float number, int digits) +{ + const char *s, *p; + s = float_to_string(number, digits); + iof_copy_number_buffer(O, s, p); + return p - s; +} + +/* iof to binary integer; pretty common */ + +int iof_get_be_uint2 (iof *I, uint32_t *pnumber) +{ + int c1, c2; + if ((c1 = iof_get(I)) < 0 || (c2 = iof_get(I)) < 0) + return 0; + *pnumber = (c1<<8)|c2; + return 1; +} + +int iof_get_be_uint3 (iof *I, uint32_t *pnumber) +{ + int c1, c2, c3; + if ((c1 = iof_get(I)) < 0 || (c2 = iof_get(I)) < 0 || (c3 = iof_get(I)) < 0) + return 0; + *pnumber = (c1<<16)|(c2<<8)|c3; + return 1; +} + +int iof_get_be_uint4 (iof *I, uint32_t *pnumber) +{ + int c1, c2, c3, c4; + if ((c1 = iof_get(I)) < 0 || (c2 = iof_get(I)) < 0 || (c3 = iof_get(I)) < 0 || (c4 = iof_get(I)) < 0) + return 0; + *pnumber = (c1<<24)|(c2<<16)|(c3<<8)|c4; + return 1; +} + +int iof_get_le_uint2 (iof *I, uint32_t *pnumber) +{ + int c1, c2; + if ((c1 = iof_get(I)) < 0 || (c2 = iof_get(I)) < 0) + return 0; + *pnumber = (c2<<8)|c1; + return 1; +} + +int iof_get_le_uint3 (iof *I, uint32_t *pnumber) +{ + int c1, c2, c3; + if ((c1 = iof_get(I)) < 0 || (c2 = iof_get(I)) < 0 || (c3 = iof_get(I)) < 0) + return 0; + *pnumber = (c3<<16)|(c2<<8)|c1; + return 1; +} + +int iof_get_le_uint4 (iof *I, uint32_t *pnumber) +{ + int c1, c2, c3, c4; + if ((c1 = iof_get(I)) < 0 || (c2 = iof_get(I)) < 0 || (c3 = iof_get(I)) < 0 || (c4 = iof_get(I)) < 0) + return 0; + *pnumber = (c4<<24)|(c3<<16)|(c2<<8)|c1; + return 1; +} + +/* iof input data */ + +uint8_t * iof_file_input_data (iof_file *iofile, size_t *psize, int *isnew) +{ + uint8_t *data; + if (iofile->flags & IOF_DATA) + { + data = iofile->buf; + *psize = iofile->end - iofile->buf; + *isnew = 0; + return data; + } + if (iof_file_reopen(iofile)) + { + data = iof_copy_file_handle_data(iof_file_get_fh(iofile), psize); + *isnew = 1; + iof_file_reclose(iofile); + return data; + } + return NULL; +} + +/* +uint8_t * iof_file_reader_data (iof_file *iofile, size_t *size) +{ + uint8_t *data; + if (!(iofile->flags & IOF_DATA) || iofile->pos == NULL || (*size = (size_t)iof_left(iofile)) == 0) + return NULL; + if (iofile->flags & IOF_BUFFER_ALLOC) + { + data = iofile->buf; // iofile->pos; // returned must be freeable, makes sense when ->buf == ->pos + iofile->flags &= ~IOF_BUFFER_ALLOC; + iofile->buf = iofile->pos = iofile->end = NULL; + return data; + } + data = (uint8_t *)util_malloc(*size); + memcpy(data, iofile->buf, *size); + return data; +} + +uint8_t * iof_file_writer_data (iof_file *iofile, size_t *size) +{ + uint8_t *data; + if (!(iofile->flags & IOF_DATA) || iofile->buf == NULL || (*size = (size_t)iof_size(iofile)) == 0) + return NULL; + if (iofile->flags & IOF_BUFFER_ALLOC) + { + iofile->flags &= ~IOF_BUFFER_ALLOC; + data = iofile->buf; + iofile->buf = iofile->pos = iofile->end = NULL; + return data; + } + data = (uint8_t *)util_malloc(*size); + memcpy(data, iofile->buf, *size); + return data; +} +*/ + +uint8_t * iof_reader_data (iof *I, size_t *psize) +{ + uint8_t *data; + *psize = (size_t)iof_left(I); + if (I->flags & IOF_BUFFER_ALLOC) + { + data = I->buf; // actually I->pos, but we have to return something freeable + I->flags &= ~IOF_BUFFER_ALLOC; + I->buf = NULL; + } + else + { + data = util_malloc(*psize); + memcpy(data, I->pos, *psize); + } + iof_close(I); + return data; +} + + +uint8_t * iof_writer_data (iof *O, size_t *psize) +{ + uint8_t *data; + *psize = (size_t)iof_size(O); + if (O->flags & IOF_BUFFER_ALLOC) + { + data = O->buf; + O->flags &= ~IOF_BUFFER_ALLOC; + O->buf = NULL; + } + else + { + data = util_malloc(*psize); + memcpy(data, O->buf, *psize); + } + iof_close(O); + return data; +} + +size_t iof_reader_to_file_handle (iof *I, FILE *file) +{ + size_t size; + for (size = 0; iof_readable(I); I->pos = I->end) + size += fwrite(I->buf, sizeof(uint8_t), iof_left(I), file); + return size; +} + +size_t iof_reader_to_file (iof *I, const char *filename) +{ + FILE *file; + size_t size; + if ((file = fopen(filename, "wb")) == NULL) + return 0; + for (size = 0; iof_readable(I); I->pos = I->end) + size += fwrite(I->buf, sizeof(uint8_t), iof_left(I), file); + fclose(file); + return size; +} + +/* debug */ + +size_t iof_data_to_file (const void *data, size_t size, const char *filename) +{ + FILE *fh; + if ((fh = fopen(filename, "wb")) == NULL) + return 0; + // size = fwrite(data, size, sizeof(uint8_t), fh); // WRONG, this always returns 1, as fwrite returns the number of elements successfully written out + size = fwrite(data, sizeof(uint8_t), size, fh); + fclose(fh); + return size; +} + +size_t iof_result_to_file_handle (iof *F, FILE *file) +{ + const void *data; + size_t size; + data = iof_result(F, size); + return iof_data_to_file_handle(data, size, file); +} + +size_t iof_result_to_file (iof *F, const char *filename) +{ + const void *data; + size_t size; + data = iof_result(F, size); + return iof_data_to_file(data, size, filename); +} + +void iof_debug (iof *I, const char *filename) +{ + FILE *file = fopen(filename, "wb"); + if (file != NULL) + { + fprintf(file, ">>> buf %p <<<\n", I->buf); + fwrite(I->buf, sizeof(uint8_t), iof_size(I), file); + fprintf(file, "\n>>> pos %p (%ld) <<<\n", I->pos, (long)iof_size(I)); + fwrite(I->pos, sizeof(uint8_t), iof_left(I), file); + fprintf(file, "\n>>> end %p (%ld) <<<\n", I->end, (long)iof_left(I)); + fwrite(I->end, sizeof(uint8_t), I->space - iof_space(I), file); + fprintf(file, "\n>>> end of buffer %p (%ld) <<<\n", I->buf + I->space, (long)(I->buf + I->space - I->end)); + fclose(file); + } +} + +/* common filters api */ + +/* sizes of filter states on x64 +size of iof_filter: 640 (no longer used; sizeof(iof) + sizeof larger state) +size of file_state: 16 +size of stream_state: 16 +size of flate_state: 104 +size of lzw_state: 56 +size of predictor_state: 104 +size of basexx_state: 48 +size of basexx_state: 48 +size of basexx_state: 48 +size of eexec_state: 40 +size of runlength_state: 24 +size of rc4_state: 24 +size of aes_state: 72 +size of img_state: 576 +size of img: 496 +*/ + +typedef struct iof_heap iof_heap; + +struct iof_heap { + uint8_t *data, *pos; + size_t size, space; + iof_heap *next, *prev; + int refcount; +}; + +typedef struct { + iof_heap *heap; +} iof_heap_ghost; + +static iof_heap * iof_buffers_heap = NULL; +static iof_heap * iof_filters_heap = NULL; + +#define IOF_HEAP_FILTERS_COUNT 4 +#define IOF_BUFFER_SIZE 262144 // (1<<18) +#define IOF_FILTER_SIZE 1024 +// sizeof(iof_filter) on x64 is now 640, img_state 576, img 496, others 16-104 +#define IOF_BUFFER_HEAP_SIZE (IOF_HEAP_FILTERS_COUNT * (IOF_BUFFER_SIZE + sizeof(iof_heap_ghost))) +#define IOF_FILTER_HEAP_SIZE (IOF_HEAP_FILTERS_COUNT * (IOF_FILTER_SIZE + sizeof(iof_heap_ghost))) + +static iof_heap * iof_heap_new (size_t space) +{ + iof_heap *iofheap; + iofheap = (iof_heap *)util_malloc(sizeof(iof_heap) + space); + iofheap->data = iofheap->pos = (uint8_t *)(iofheap + 1); + iofheap->size = iofheap->space = space; + iofheap->next = NULL; + iofheap->prev = NULL; + iofheap->refcount = 0; + return iofheap; +} + +#define iof_heap_free(iofheap) util_free(iofheap) + +void iof_filters_init (void) +{ + if (iof_buffers_heap == NULL) + iof_buffers_heap = iof_heap_new(IOF_BUFFER_HEAP_SIZE); + if (iof_filters_heap == NULL) + iof_filters_heap = iof_heap_new(IOF_FILTER_HEAP_SIZE); +} + +void iof_filters_free (void) +{ + iof_heap *heap, *next; + for (heap = iof_buffers_heap; heap != NULL; heap = next) + { + next = heap->next; + if (heap->refcount != 0) + loggerf("not closed iof filters left (%d)", heap->refcount); + if (next != NULL) + loggerf("iof filters heap left"); + iof_heap_free(heap); + } + iof_buffers_heap = NULL; + for (heap = iof_filters_heap; heap != NULL; heap = next) + { + next = heap->next; + if (heap->refcount != 0) + loggerf("not closed iof buffers left (%d)", heap->refcount); + if (next != NULL) + loggerf("iof buffers heap left"); + iof_heap_free(heap); + } + iof_filters_heap = NULL; +} + +#define iof_heap_get(hp, ghost, data, siz) \ + (ghost = (iof_heap_ghost *)((hp)->pos), \ + ghost->heap = hp, \ + data = (uint8_t *)(ghost + 1), \ + (hp)->pos += siz, \ + (hp)->size -= siz, \ + ++(hp)->refcount) + + +static void * iof_heap_take (iof_heap **pheap, size_t size) +{ + uint8_t *data; + iof_heap_ghost *ghost; + iof_heap *heap, *newheap, *next; + + heap = *pheap; + size += sizeof(iof_heap_ghost); + if (heap->size >= size) + { /* take cheap mem from main heap */ + iof_heap_get(heap, ghost, data, size); + return data; + } + if (size <= heap->space >> 1) + { /* make new cheap heap, make it front */ + *pheap = newheap = iof_heap_new(heap->space); + newheap->next = heap; + heap->prev = newheap; + iof_heap_get(newheap, ghost, data, size); + return data; + } + /* size much larger than expected? should not happen. + make a single-item heap, keep the front heap intact. */ + newheap = iof_heap_new(size); + if ((next = heap->next) != NULL) + { + newheap->next = next; + next->prev = newheap; + } + heap->next = newheap; + newheap->prev = heap; + iof_heap_get(newheap, ghost, data, size); + return data; +} + +void iof_heap_back (void *data) +{ + iof_heap_ghost *ghost; + iof_heap *heap, *next, *prev; + + ghost = ((iof_heap_ghost *)data) - 1; + heap = ghost->heap; + if (heap->refcount == 0) + loggerf("invalid use of iof heap, refcount < 0"); + if (--heap->refcount <= 0) + { + if ((prev = heap->prev) != NULL) + { /* free the heap */ + if ((next = heap->next) != NULL) + prev->next = next, next->prev = prev; + else + prev->next = NULL; + iof_heap_free(heap); + } + else + { /* this is the front heap, just reset */ + heap->pos = heap->data; + heap->size = heap->space; + } + } +} + +void * iof_filter_new (size_t size) +{ // to be removed + void *data; + + iof_filters_init(); + data = iof_heap_take(&iof_filters_heap, size); + return memset(data, 0, size); +} + +static uint8_t * iof_filter_buffer_new (size_t *psize) +{ + iof_filters_init(); + *psize = IOF_BUFFER_SIZE; + return iof_heap_take(&iof_buffers_heap, IOF_BUFFER_SIZE); +} + +iof * iof_filter_reader_new (iof_handler handler, size_t statesize, void **pstate) +{ + iof *F; + void *filter; + uint8_t *buffer; + size_t buffersize; + + iof_filters_init(); + filter = iof_heap_take(&iof_filters_heap, sizeof(iof) + statesize); + F = (iof *)memset(filter, 0, sizeof(iof) + statesize); + buffer = iof_filter_buffer_new(&buffersize); + iof_reader_buffer(F, buffer, buffersize); + F->flags |= IOF_HEAP|IOF_BUFFER_HEAP; + F->more = handler; + *pstate = (F + 1); + return F; +} + +iof * iof_filter_reader_with_buffer_new (iof_handler handler, size_t statesize, void **pstate, void *buffer, size_t buffersize) +{ // for filters that has own buffer (string, some image filters) + iof *F; + void *filter; + iof_filters_init(); + filter = iof_heap_take(&iof_filters_heap, sizeof(iof) + statesize); + F = (iof *)memset(filter, 0, sizeof(iof) + statesize); + iof_reader_buffer(F, buffer, buffersize); + F->flags |= IOF_HEAP; + F->more = handler; + *pstate = (F + 1); + return F; +} + +iof * iof_filter_writer_new (iof_handler handler, size_t statesize, void **pstate) +{ + iof *F; + void *filter; + uint8_t *buffer; + size_t buffersize; + + iof_filters_init(); + filter = iof_heap_take(&iof_filters_heap, sizeof(iof) + statesize); + F = (iof *)memset(filter, 0, sizeof(iof) + statesize); + buffer = iof_filter_buffer_new(&buffersize); + iof_writer_buffer(F, buffer, buffersize); + F->flags |= IOF_HEAP|IOF_BUFFER_HEAP; + F->more = handler; + *pstate = (F + 1); + return F; +} + +iof * iof_filter_writer_with_buffer_new (iof_handler handler, size_t statesize, void **pstate, void *buffer, size_t size) +{ + iof *F; + void *filter; + size_t buffersize; + + iof_filters_init(); + filter = iof_heap_take(&iof_filters_heap, sizeof(iof) + statesize); + F = (iof *)memset(filter, 0, sizeof(iof) + statesize); + buffer = iof_filter_buffer_new(&buffersize); + iof_writer_buffer(F, buffer, buffersize); + F->flags |= IOF_HEAP; + F->more = handler; + *pstate = (F + 1); + return F; +} + +/* close */ + +#define iof_close_next(F) ((void)(iof_decref((F)->next), (F)->next = NULL, 0)) +/* when filter creation fails, we should take care to destroy the filter but leave ->next intact */ +#define iof_clear_next(F) ((void)(iof_unref((F)->next), (F)->next = NULL, 0)) + +#define iof_close_buffer(F) ((void)\ + ((F)->buf != NULL ? \ + ((F->flags & IOF_BUFFER_ALLOC) ? (util_free((F)->buf), (F)->buf = NULL, 0) : \ + ((F->flags & IOF_BUFFER_HEAP) ? (iof_filter_buffer_free((F)->buf), (F)->buf = NULL, 0) : ((F)->buf = NULL, 0))) : 0)) + +/* closing underlying file handle */ + +static void iof_close_file (iof *F) +{ + FILE *file; + //if (F->flags & IOF_FILE_HANDLE) + //{ + if ((file = F->file) != NULL) + { + if (F->flags & IOF_CLOSE_FILE) + fclose(F->file); + F->file = NULL; + } + //} +} + +/* a very special variant for reader filters initiated with iof_file_reopen(). It also calls + iof_file_reclose(), which takes an effect only if previously reopened, but better to keep + all this thin ice separated. Used in filters: iofile_reader, iofile_stream_reader, image + decoders. */ + +static void iof_close_iofile (iof *F) +{ + iof_file *iofile; + //if (F->flags & IOF_FILE) + //{ + if ((iofile = F->iofile) != NULL) + { + iof_file_unsync(iofile, NULL); + iof_file_reclose(iofile); // takes an effect iff prevoiusly reopened + iof_file_decref(iofile); + F->iofile = NULL; + } + //} +} + +void iof_free (iof *F) +{ + if (F->flags & IOF_FILE_HANDLE) + iof_close_file(F); + else if (F->flags & IOF_FILE) + iof_close_iofile(F); + else if (F->flags & IOF_NEXT) + iof_close_next(F); + iof_close_buffer(F); + if (F->flags & IOF_HEAP) + iof_filter_free(F); + else if (F->flags & IOF_ALLOC) + util_free(F); +} + +void iof_discard (iof *F) +{ // so far used only on failed filters creation; as iof_free() but don't dare to release ->next + if (F->flags & IOF_FILE_HANDLE) + iof_close_file(F); + else if (F->flags & IOF_FILE) + iof_close_iofile(F); + else if (F->flags & IOF_NEXT) + iof_close_next(F); + iof_close_buffer(F); + if (F->flags & IOF_HEAP) + iof_filter_free(F); + else if (F->flags & IOF_ALLOC) + util_free(F); +} + +/* resizing buffer */ + +size_t iof_resize_buffer_to (iof *O, size_t space) +{ + uint8_t *buf; + + if (O->flags & IOF_BUFFER_ALLOC) + { + buf = (uint8_t *)util_realloc(O->buf, space); + } + else + { + buf = (uint8_t *)util_malloc(space); + memcpy(buf, O->buf, iof_size(O)); + if (O->flags & IOF_BUFFER_HEAP) + { + iof_filter_buffer_free(O->buf); + O->flags &= ~IOF_BUFFER_HEAP; + } + O->flags |= IOF_BUFFER_ALLOC; + + } + O->pos = buf + iof_size(O); + O->end = buf + space; + O->buf = buf; + O->space = space; + return iof_left(O); +} + +/* */ + +size_t iof_decoder_retval (iof *I, const char *type, iof_status status) +{ + switch (status) + { + case IOFERR: + case IOFEMPTY: // should never happen as we set state.flush = 1 on decoders init + loggerf("%s decoder error (%d, %s)", type, status, iof_status_kind(status)); + I->flags |= IOF_STOPPED; + return 0; + case IOFEOF: // this is the last chunk, + I->flags |= IOF_STOPPED; // so stop it and fall + case IOFFULL: // prepare pointers to read from I->buf + I->end = I->pos; + I->pos = I->buf; + return I->end - I->buf; + } + loggerf("%s decoder bug, invalid retval %d", type, status); + return 0; +} + +size_t iof_encoder_retval (iof *O, const char *type, iof_status status) +{ + switch (status) + { + case IOFERR: + case IOFFULL: + loggerf("%s encoder error (%d, %s)", type, status, iof_status_kind(status)); + return 0; + case IOFEMPTY: + O->pos = O->buf; + O->end = O->buf + O->space; + return O->space; + case IOFEOF: + return 0; + } + loggerf("%s encoder bug, invalid retval %d", type, status); + return 0; +} + +/* file/stream state */ + +typedef struct { + size_t length; + size_t offset; +} file_state; + + +#define file_state_init(state, off, len) ((state)->offset = off, (state)->length = len) + +typedef struct { + size_t length; + size_t offset; +} stream_state; + +#define stream_state_init(state, off, len) ((state)->offset = off, (state)->length = len) + +static size_t file_read (iof *I) +{ + size_t bytes, tail; + if (I->flags & IOF_STOPPED) + return 0; + tail = iof_tail(I); + if ((bytes = tail + fread(I->buf + tail, sizeof(uint8_t), I->space - tail, I->file)) < I->space) + I->flags |= IOF_STOPPED; + I->pos = I->buf; + I->end = I->buf + bytes; + return bytes; +} + +static size_t iofile_read (iof *I, size_t *poffset) +{ + size_t bytes, tail; + if (I->flags & IOF_STOPPED) + return 0; + iof_file_sync(I->iofile, poffset); + tail = iof_tail(I); + if ((bytes = tail + iof_file_read(I->buf + tail, sizeof(uint8_t), I->space - tail, I->iofile)) < I->space) + { + I->flags |= IOF_STOPPED; + iof_file_unsync(I->iofile, poffset); + } + I->pos = I->buf; + I->end = I->buf + bytes; + return bytes; +} + +static size_t file_load (iof *I) +{ + size_t bytes, left, tail; + if (I->flags & IOF_STOPPED) + return 0; + tail = iof_tail(I); + I->pos = I->buf + tail; + I->end = I->buf + I->space; /* don't assume its done when initializing the filter */ + left = I->space - tail; + do { + bytes = fread(I->pos, sizeof(uint8_t), left, I->file); + I->pos += bytes; + } while (bytes == left && (left = iof_resize_buffer(I)) > 0); + I->flags |= IOF_STOPPED; + return iof_loaded(I); +} + +static size_t iofile_load (iof *I, size_t *poffset) +{ + size_t bytes, left, tail; + if (I->flags & IOF_STOPPED) + return 0; + tail = iof_tail(I); + I->pos = I->buf + tail; + I->end = I->buf + I->space; /* don't assume its done when initializing the filter */ + left = I->space - tail; + iof_file_sync(I->iofile, poffset); + do { + bytes = iof_file_read(I->pos, sizeof(uint8_t), left, I->iofile); + I->pos += bytes; + } while (bytes == left && (left = iof_resize_buffer(I)) > 0); + I->flags |= IOF_STOPPED; + iof_file_unsync(I->iofile, poffset); + return iof_loaded(I); +} + +static size_t filter_file_reader (iof *I, iof_mode mode) +{ + switch (mode) + { + case IOFREAD: + return file_read(I); + case IOFLOAD: + return file_load(I); + case IOFCLOSE: + iof_free(I); + return 0; + default: + return 0; + } +} + +static size_t filter_iofile_reader (iof *I, iof_mode mode) +{ + file_state *state; + state = iof_filter_state(file_state *, I); + switch (mode) + { + case IOFREAD: + return iofile_read(I, &state->offset); + case IOFLOAD: + return iofile_load(I, &state->offset); + case IOFCLOSE: + iof_free(I); + return 0; + default: + return 0; + } +} + +static size_t file_write (iof *O, int flush) +{ + size_t bytes; + if ((bytes = iof_size(O)) > 0) + if (bytes != fwrite(O->buf, sizeof(uint8_t), bytes, O->file)) + return 0; + if (flush) + fflush(O->file); + O->end = O->buf + O->space; // remains intact actually + O->pos = O->buf; + return O->space; +} + +static size_t iofile_write (iof *O, size_t *poffset, int flush) +{ + size_t bytes; + iof_file_sync(O->iofile, poffset); + if ((bytes = iof_size(O)) > 0) + { + if (bytes != iof_file_write(O->buf, sizeof(uint8_t), bytes, O->iofile)) + { + iof_file_unsync(O->iofile, poffset); + return 0; + } + } + if (flush) + iof_file_flush(O->iofile); + O->end = O->buf + O->space; // remains intact actually + O->pos = O->buf; + return O->space; +} + +static size_t filter_file_writer (iof *O, iof_mode mode) +{ + switch (mode) + { + case IOFWRITE: + return file_write(O, 0); + case IOFFLUSH: + return file_write(O, 1); + case IOFCLOSE: + file_write(O, 1); + iof_free(O); + return 0; + default: + return 0; + } +} + +static size_t filter_iofile_writer (iof *O, iof_mode mode) +{ + file_state *state; + state = iof_filter_state(file_state *, O); + switch (mode) + { + case IOFWRITE: + return iofile_write(O, &state->offset, 0); + case IOFFLUSH: + return iofile_write(O, &state->offset, 1); + case IOFCLOSE: + iofile_write(O, &state->offset, 1); + iof_free(O); + return 0; + default: + return 0; + } +} + +/* filter from FILE* */ + +iof * iof_filter_file_handle_reader (FILE *file) +{ + iof *I; + file_state *state; + if (file == NULL) + return NULL; + I = iof_filter_reader(filter_file_reader, sizeof(file_state), &state); + iof_setup_file(I, file); + file_state_init(state, 0, 0); + return I; +} + +iof * iof_filter_file_handle_writer (FILE *file) +{ + iof *O; + file_state *state; + if (file == NULL) + return NULL; + O = iof_filter_writer(filter_file_writer, sizeof(file_state), &state); + iof_setup_file(O, file); + file_state_init(state, 0, 0); + return O; +} + +/* filter from iof_file * */ + +iof * iof_filter_iofile_reader (iof_file *iofile, size_t offset) +{ + iof *I; + file_state *state; + if (!iof_file_reopen(iofile)) + return NULL; + I = iof_filter_reader(filter_iofile_reader, sizeof(file_state), &state); + iof_setup_iofile(I, iofile); + file_state_init(state, offset, 0); + return I; +} + +iof * iof_filter_iofile_writer (iof_file *iofile, size_t offset) +{ + iof *O; + file_state *state; + O = iof_filter_writer(filter_iofile_writer, sizeof(file_state), &state); + iof_setup_iofile(O, iofile); + file_state_init(state, offset, 0); + return O; +} + +/* filter from filename */ + +iof * iof_filter_file_reader (const char *filename) +{ + iof *I; + file_state *state; + FILE *file; + if ((file = fopen(filename, "rb")) == NULL) + return NULL; + I = iof_filter_reader(filter_file_reader, sizeof(file_state), &state); + iof_setup_file(I, file); + file_state_init(state, 0, 0); + I->flags |= IOF_CLOSE_FILE; + return I; +} + +iof * iof_filter_file_writer (const char *filename) +{ + iof *O; + file_state *state; + FILE *file; + if ((file = fopen(filename, "wb")) == NULL) + return NULL; + O = iof_filter_writer(filter_file_writer, sizeof(file_state), &state); + iof_setup_file(O, file); + file_state_init(state, 0, 0); + O->flags |= IOF_CLOSE_FILE; + return O; +} + +/* from string */ + +static size_t dummy_handler (iof *I, iof_mode mode) +{ + switch (mode) + { + case IOFCLOSE: + iof_free(I); + return 0; + default: + return 0; + } +} + +iof * iof_filter_string_reader (const void *s, size_t length) +{ + iof *I; + void *dummy; + I = iof_filter_reader_with_buffer(dummy_handler, 0, &dummy, NULL, 0); + I->rbuf = I->rpos = (const uint8_t *)s; + I->rend = (const uint8_t *)s + length; + // I->space = length; + return I; +} + +iof * iof_filter_string_writer (const void *s, size_t length) +{ + iof *O; + void *dummy; + O = iof_filter_reader_with_buffer(dummy_handler, 0, &dummy, NULL, 0); + O->rbuf = O->rpos = (const uint8_t *)s; + O->rend = (const uint8_t *)s + length; + // O->space = length; + return O; +} + +iof * iof_filter_buffer_writer (size_t size) +{ // filter alternative of iof_buffer_create() + iof *O; + void *dummy; + uint8_t *buffer; + if (size > IOF_BUFFER_SIZE) + { + buffer = (uint8_t *)util_malloc(size); + O = iof_filter_writer_with_buffer(iof_mem_handler, 0, &dummy, buffer, size); + O->flags |= IOF_BUFFER_ALLOC; + return O; + } + return iof_filter_writer(iof_mem_handler, 0, &dummy); +} + +/* stream */ + +static size_t file_stream_read (iof *I, size_t *plength) +{ + size_t bytes, tail; + if (I->flags & IOF_STOPPED || *plength == 0) + return 0; + tail = iof_tail(I); + if (I->space - tail >= *plength) + { + bytes = tail + fread(I->buf + tail, sizeof(uint8_t), *plength, I->file); + I->flags |= IOF_STOPPED; + *plength = 0; + } + else + { + bytes = tail + fread(I->buf + tail, sizeof(uint8_t), I->space - tail, I->file); + *plength -= bytes - tail; + } + I->pos = I->buf; + I->end = I->buf + bytes; + return bytes; +} + +static size_t iofile_stream_read (iof *I, size_t *plength, size_t *poffset) +{ + size_t bytes, tail; + if (I->flags & IOF_STOPPED || *plength == 0) + return 0; + tail = iof_tail(I); + iof_file_sync(I->iofile, poffset); + if (I->space - tail >= *plength) + { + bytes = tail + iof_file_read(I->buf + tail, sizeof(uint8_t), *plength, I->iofile); + iof_file_unsync(I->iofile, poffset); + I->flags |= IOF_STOPPED; + *plength = 0; + } + else + { + bytes = tail + iof_file_read(I->buf + tail, sizeof(uint8_t), I->space - tail, I->iofile); + *plength -= bytes - tail; + } + I->pos = I->buf; + I->end = I->buf + bytes; + return bytes; +} + +static size_t file_stream_load (iof *I, size_t *plength) +{ + size_t bytes, tail; + if (I->flags & IOF_STOPPED || *plength == 0) + return 0; + tail = iof_tail(I); + if (I->space - tail < *plength) + if (iof_resize_buffer_to(I, tail + *plength) == 0) + return 0; + bytes = tail + fread(I->buf + tail, sizeof(uint8_t), *plength, I->file); + I->flags |= IOF_STOPPED; + *plength = 0; + I->pos = I->buf; + I->end = I->buf + bytes; + return bytes; +} + +static size_t iofile_stream_load (iof *I, size_t *plength, size_t *poffset) +{ + size_t bytes, tail; + if (I->flags & IOF_STOPPED || *plength == 0) + return 0; + iof_file_sync(I->iofile, poffset); + tail = iof_tail(I); + if (I->space - tail < *plength) + if (iof_resize_buffer_to(I, tail + *plength) == 0) + return 0; + bytes = tail + iof_file_read(I->buf + tail, sizeof(uint8_t), *plength, I->iofile); + iof_file_unsync(I->iofile, poffset); + I->flags |= IOF_STOPPED; + *plength = 0; + I->pos = I->buf; + I->end = I->buf + bytes; + return bytes; +} + +static size_t filter_file_stream_reader (iof *I, iof_mode mode) +{ + stream_state *state; + state = iof_filter_state(stream_state *, I); + switch(mode) + { + case IOFREAD: + return file_stream_read(I, &state->length); + case IOFLOAD: + return file_stream_load(I, &state->length); + case IOFCLOSE: + iof_free(I); + return 0; + default: + return 0; + } +} + +static size_t filter_iofile_stream_reader (iof *I, iof_mode mode) +{ + stream_state *state; + state = iof_filter_state(stream_state *, I); + switch(mode) + { + case IOFREAD: + return iofile_stream_read(I, &state->length, &state->offset); + case IOFLOAD: + return iofile_stream_load(I, &state->length, &state->offset); + case IOFCLOSE: + iof_free(I); + return 0; + default: + return 0; + } +} + +iof * iof_filter_stream_reader (FILE *file, size_t offset, size_t length) +{ + iof *I; + stream_state *state; + I = iof_filter_reader(filter_file_stream_reader, sizeof(stream_state), &state); + iof_setup_file(I, file); + stream_state_init(state, offset, length); + fseek(file, (long)offset, SEEK_SET); // or perhaps it should be call in file_stream_read(), like iof_file_sync()? + return I; +} + +iof * iof_filter_stream_coreader (iof_file *iofile, size_t offset, size_t length) +{ + iof *I; + stream_state *state; + if (!iof_file_reopen(iofile)) + return NULL; + I = iof_filter_reader(filter_iofile_stream_reader, sizeof(stream_state), &state); + iof_setup_iofile(I, iofile); + stream_state_init(state, offset, length); + return I; +} + +static size_t file_stream_write (iof *O, size_t *plength, int flush) +{ + size_t bytes; + if ((bytes = iof_size(O)) > 0) + { + if (bytes != fwrite(O->buf, sizeof(uint8_t), bytes, O->file)) + { + *plength += bytes; + return 0; + } + } + if (flush) + fflush(O->file); + *plength += bytes; + O->end = O->buf + O->space; // remains intact + O->pos = O->buf; + return O->space; +} + +static size_t iofile_stream_write (iof *O, size_t *plength, size_t *poffset, int flush) +{ + size_t bytes; + if ((bytes = iof_size(O)) > 0) + { + iof_file_sync(O->iofile, poffset); + if (bytes != iof_file_write(O->buf, sizeof(uint8_t), bytes, O->iofile)) + { + *plength += bytes; + iof_file_unsync(O->iofile, poffset); + return 0; + } + } + if (flush) + iof_file_flush(O->iofile); + *plength += bytes; + O->end = O->buf + O->space; // remains intact + O->pos = O->buf; + return O->space; +} + +static size_t filter_file_stream_writer (iof *O, iof_mode mode) +{ + stream_state *state; + state = iof_filter_state(stream_state *, O); + switch (mode) + { + case IOFWRITE: + return file_stream_write(O, &state->length, 0); + case IOFFLUSH: + return file_stream_write(O, &state->length, 1); + case IOFCLOSE: + file_stream_write(O, &state->length, 1); + iof_free(O); + return 0; + default: + return 0; + } +} + +static size_t filter_iofile_stream_writer (iof *O, iof_mode mode) +{ + stream_state *state; + state = iof_filter_state(stream_state *, O); + switch (mode) + { + case IOFWRITE: + return iofile_stream_write(O, &state->length, &state->offset, 0); + case IOFFLUSH: + return iofile_stream_write(O, &state->length, &state->offset, 1); + case IOFCLOSE: + iofile_stream_write(O, &state->length, &state->offset, 1); + iof_free(O); + return 0; + default: + return 0; + } +} + +iof * iof_filter_stream_writer (FILE *file) +{ + iof *O; + stream_state *state; + O = iof_filter_writer(filter_file_stream_writer, sizeof(stream_state), &state); + iof_setup_file(O, file); + stream_state_init(state, 0, 0); + return O; +} + +iof * iof_filter_stream_cowriter (iof_file *iofile, size_t offset) +{ + iof *O; + stream_state *state; + O = iof_filter_writer(filter_iofile_stream_writer, sizeof(stream_state), &state); + iof_setup_iofile(O, iofile); + stream_state_init(state, offset, 0); + return O; +} + +/* very specific for images; get input from already created strem filter, exchange the filter but keep the buffer */ + +FILE * iof_filter_file_reader_source (iof *I, size_t *poffset, size_t *plength) +{ + stream_state *sstate; + file_state *fstate; + if (I->more == filter_file_stream_reader) // I is the result of iof_filter_stream_reader() + { + sstate = iof_filter_state(stream_state *, I); + *poffset = sstate->offset; + *plength = sstate->length; // might be 0 but it is ok for file readers + return I->file; + } + if (I->more == filter_file_reader) + { + fstate = iof_filter_state(file_state *, I); + *poffset = fstate->offset; + *plength = fstate->length; // might be 0 but it is ok for file readers + return I->file; + } + return NULL; +} + +iof_file * iof_filter_file_coreader_source (iof *I, size_t *poffset, size_t *plength) +{ + stream_state *sstate; + file_state *fstate; + if (I->more == filter_iofile_stream_reader) // I is the result of iof_filter_stream_coreader() + { + sstate = iof_filter_state(stream_state *, I); + *poffset = sstate->offset; + *plength = sstate->length; + return I->iofile; + } + if (I->more == filter_iofile_reader) + { + fstate = iof_filter_state(file_state *, I); + *poffset = fstate->offset; + *plength = fstate->length; + return I->iofile; + } + return NULL; +} + +iof * iof_filter_reader_replacement (iof *P, iof_handler handler, size_t statesize, void **pstate) +{ // called after iof_filter_file_reader_source(), no need to check if F is filter from iof heap and if has buffer from iof heap + iof *F; + F = iof_filter_reader_with_buffer(handler, statesize, pstate, P->buf, P->space); + F->flags |= IOF_BUFFER_HEAP; + //iof_reader_buffer(P, NULL, 0); + //P->flags &= ~IOF_BUFFER_HEAP; + iof_filter_free(P); + return F; +} + + + + + + + + + + + + + + + + + + + + + + + diff --git a/Build/source/texk/web2c/luatexdir/luapplib/util/utiliof.h b/Build/source/texk/web2c/luatexdir/luapplib/util/utiliof.h new file mode 100644 index 00000000000..853a9ae523f --- /dev/null +++ b/Build/source/texk/web2c/luatexdir/luapplib/util/utiliof.h @@ -0,0 +1,669 @@ + +#ifndef UTIL_IOF_H +#define UTIL_IOF_H + +#include <stdio.h> // for FILE * +#include <errno.h> // for errno +#include <string.h> // for strerror() +#include <stdint.h> // for uintN_t + +#include "utildecl.h" +#include "utilnumber.h" + +/* handler call modes */ + +typedef enum { + IOFREAD = 0, /* read to buffer */ + IOFLOAD = 1, /* read all to buffer */ + IOFWRITE = 2, /* write buffer to the output */ + IOFFLUSH = 3, /* flush buffer to the output */ + IOFCLOSE = 4 /* (flush and) close */ +} iof_mode; + +/* return statuses */ + +typedef enum { + IOFEOF = -1, /* end of input */ + IOFEMPTY = -2, /* end of input buffer*/ + IOFFULL = -3, /* end of output buffer */ + IOFERR = -4 /* error */ +} iof_status; + +const char * iof_status_kind (iof_status status); + +/* iof_file */ + +typedef struct iof_file { + union { + FILE *iofh; // access via iof_file_get_fh / iof_file_set_fh (below) + union { + struct { uint8_t *buf, *pos, *end; }; + struct { const uint8_t *rbuf, *rpos, *rend; }; // to trick compiler warnings about cast discarding const + }; + }; + size_t *offset; + char *name; + size_t size; + int refcount; + int flags; +} iof_file; + +/* iof handler function */ + +typedef struct iof iof; +typedef size_t (*iof_handler) (iof *I, iof_mode mode); + +/* iof structure */ + +#define IOF_MEMBERS \ + union { \ + struct { uint8_t *buf, *pos, *end; }; \ + struct { uint16_t *hbuf, *hpos, *hend; }; \ + struct { uint32_t *ibuf, *ipos, *iend; }; \ + struct { const uint8_t *rbuf, *rpos, *rend; }; \ + }; \ + size_t space; \ + iof_handler more; \ + union { iof *next; FILE *file; iof_file *iofile; void *link; }; \ + int flags; \ + int refcount + +/* + buf -- the beginning of buffer + pos -- the current position + end -- the end of buffer + space -- private space size, not always eq. (end - buf) + more -- handler function + next/file/iofile/link -- reader source or writer target + source -- source filter + flags -- private filter info + refcount -- refcount +*/ + +struct iof { + IOF_MEMBERS; +}; + +typedef void (*iof_dump_function) (const void *value, iof *O); + +/* flags */ + +#define IOF_ALLOC (1<<0) // iof is allocated +#define IOF_HEAP (1<<1) // iof taken from iof heap +#define IOF_BUFFER_ALLOC (1<<2) // buffer allocated +#define IOF_BUFFER_HEAP (1<<3) // buffer taken from iof heap + +#define IOF_SHORT (1<<4) // buffer uses 16bit integers +#define IOF_LONG (1<<5) // buffer uses 32bit integers + +#define IOF_TAIL (1<<6) // preserve reader tail +#define IOF_READER (1<<7) // is reader +#define IOF_WRITER (1<<8) // is writer + +#define IOF_DATA (1<<9) // binds some memory +#define IOF_FILE_HANDLE (1<<10) // links FILE * +#define IOF_FILE (1<<11) // links iof_file * +#define IOF_NEXT (1<<12) // links next iof * +#define IOF_CLOSE_FILE (1<<13) // close FILE * on free +#define IOF_REOPEN_FILE (1<<14) // close/reopen mode for iof_file +#define IOF_RECLOSE_FILE (1<<15) // ditto + +#define IOF_STOPPED (1<<16) // stopped + +// #define IOF_CUSTOM (1<<17) // first custom flag + +#define IOF_BUFSIZ (sizeof(iof) + BUFSIZ*sizeof(uint8_t)) + +/* +reading buffer -- all of buf, pos, end pointers are initialized to the beginning of the private buffer, + next call to a handler function moves the end pointer to bufer+space +writer -- buf and pos pointers initialized to the beginning of the buffer, end initialized to bufer+space + +Every call to handler returns size_t number of bytes +available (to write/read) or 0 if there is no more space. + +We usually align the data buffer just after the iof structure. +This is convenient, especially when a memory for the structure +and its buffer is to be allocated. In the case of growing output +buffers we used to check if the memory of the buffer is allocated +by the handler function using test (O->buf != (O+1)). We don't use +it any longer not to rely on little secrets. Now there is an explicit +IOF_BUFFER_ALLOC flag for that. IOF_ALLOC tells if the structure +itself is taken from malloc (not used so far). Assuming the buffer size +is way larger the sizeof(iof) +*/ + +/* initializers */ + +#define IOF_READER_STRUCT(handler, file, buffer, size, flags) \ + { {{ (uint8_t *)(buffer), (uint8_t *)(buffer), (uint8_t *)(buffer) }}, size, handler, { file }, flags|IOF_READER, 0 } + +#define IOF_WRITER_STRUCT(handler, file, buffer, size, flags) \ + { {{ (uint8_t *)(buffer), (uint8_t *)(buffer), (uint8_t *)(buffer) + size }}, size, handler, { file }, flags|IOF_WRITER, 0 } + +#define IOF_STRING_STRUCT(buffer, size) \ + { {{ (uint8_t *)(buffer), (uint8_t *)(buffer), (uint8_t *)(buffer) + size }}, size, NULL, { NULL }, 0|IOF_READER|IOF_DATA, 0 } + +#define IOF_STRING() IOF_STRING_STRUCT(0, 0) + +/* refcount */ + +#define iof_incref(I) (++(I)->refcount) +#define iof_decref(I) ((void)(--(I)->refcount <= 0 && iof_close(I))) +#define iof_unref(I) (--(I)->refcount) + +/* setting up iof and buffer from mem buffer of a given size */ + +#define iof_setup_reader(I, buffer, size) \ + ((I) = (iof *)(buffer), iof_reader_buffer(I, (I)+1, size - sizeof(iof))) + +#define iof_setup_writer(O, buffer, size) \ + ((O) = (iof *)buffer, iof_writer_buffer(O, (O)+1, size - sizeof(iof))) + +/* binding buffer of a given size */ + +#define iof_reader_buffer(I, buffer, size) \ + ((I)->buf = (I)->pos = (I)->end = (uint8_t *)(buffer), \ + (I)->space = size, (I)->flags = 0|IOF_READER, (I)->refcount = 0) + +#define iof_writer_buffer(O, buffer, size) \ + ((O)->buf = (O)->pos = (uint8_t *)(buffer), \ + (O)->end = (uint8_t *)(buffer) + size, \ + (O)->space = size, (O)->flags = 0|IOF_WRITER, (O)->refcount = 0) + +/* basics */ + +#define iof_space(I) ((I)->end - (I)->buf) +#define iof_left(I) ((I)->end - (I)->pos) +#define iof_size(I) ((I)->pos - (I)->buf) + +#define iof_input(I) ((I)->more ? (I)->more((I), IOFREAD) : 0lu) +#define iof_load(I) ((I)->more ? (I)->more((I), IOFLOAD) : 0lu) + +#define iof_output(O) ((O)->more ? (O)->more((O), IOFWRITE) : 0lu) +//#define iof_flush(O) ((O)->pos > (O)->buf && (O)->more ? (O)->more(O, IOFFLUSH) : 0lu) +// flush should be unconditional, because encoders emits EOD markers only on flush +#define iof_flush(O) ((O)->more ? (O)->more(O, IOFFLUSH) : 0lu) +#define iof_close(O) ((O)->more ? (O)->more(O, IOFCLOSE) : 0lu) + +#define iof_stop(F) ((void)(F->pos = F->end = F->buf, F->flags |= IOF_STOPPED)) + +/* +Rewriting reader tail to the beginning of new data portion; readers reacting on IOFREAD +mode must be aware of some not yet read data, but treat it necessary only if IOF_TAIL flag is set. +Parsers using iof input may protect not yet read data when there may be a need to put bytes +back to the stream. This is trivial when I->pos > I->buf, as we can make a move by --I->pos. +But when there is a need to put back more then one byte, we can protect the data tail, so that +realoder will rewrite it to the beginning of new data chunk. + + iof_tail(I) - internal, used by iof handlers at IOFREAD mode + iof_protect_tail(I) - used by parsers to ensure some bytes chunk in one piece + +*/ + +size_t iof_save_tail (iof *I); +#define iof_tail(I) (((I)->flags & IOF_TAIL) && (I)->pos < (I)->end ? iof_save_tail(I) : 0) + +size_t iof_input_save_tail (iof *I, size_t back); +#define iof_protect_tail(I, back, length) ((iof_left(I) >= (length) - (back)) ? 1 : (iof_input_save_tail(I, back) >= length - back)) + +//uint8_t * iof_tail_data (iof *I, size_t *ptail); +//#define iof_tail_free(data) util_free(data) + +/* panic */ + +// #define iof_panic(mess) return 0 +#ifndef iof_panic + #define iof_panic(mess) (fputs(mess, stderr), abort()) +#endif +//#define iof_memory_error() iof_panic(strerror(errno)) +#define iof_fwrite_error() iof_panic(strerror(errno)) + +/* generic helpers */ + +UTILAPI uint8_t * iof_copy_file_data (const char *filename, size_t *psize); +UTILAPI uint8_t * iof_copy_file_handle_data (FILE *file, size_t *psize); + +/* In the future we may need releasing file handle and restoring it from iofile->name, so access file handle via macros */ + +#define iof_file_get_fh(iofile) ((iofile)->iofh) +#define iof_file_set_fh(iofile, fh) ((iofile)->iofh = fh) +#define iof_file_get_file(iofile) (((iofile)->flags & IOF_DATA) ? NULL : iof_file_get_fh(iofile)) +FILE * iof_get_file (iof *F); + +/* basic iof_file interface */ + +iof_file * iof_file_new (FILE *file); +iof_file * iof_file_init (iof_file *iofile, FILE *file); + +iof_file * iof_file_rdata (const void *data, size_t size); +iof_file * iof_file_wdata (void *data, size_t size); + +iof_file * iof_file_rdata_init (iof_file *iofile, const void *data, size_t size); +iof_file * iof_file_wdata_init (iof_file *iofile, void *data, size_t size); + +iof_file * iof_file_reader_from_file_handle (iof_file *iofile, const char *filename, FILE *file, int preload, int closefile); +iof_file * iof_file_reader_from_file (iof_file *iofile, const char *filename, int preload); +iof_file * iof_file_reader_from_data (iof_file *iofile, const void *data, size_t size, int preload, int freedata); +//iof_file * iof_file_writer_from_file (iof_file *iofile, const char *filename); + +void * iof_copy_data (const void *data, size_t size); +#define iof_data_free(data) util_free(data) +#define iof_file_wdata_copy(data, size) iof_file_wdata(iof_copy_data(data, size), size) +#define iof_file_rdata_copy(data, size) iof_file_rdata(iof_copy_data(data, size), size) + +void iof_file_free (iof_file *iofile); + +#define iof_file_get_name(iofile) ((iofile)->name) +void iof_file_set_name (iof_file *iofile, const char *name); + +#define iof_file_incref(iofile) (++(iofile)->refcount) +#define iof_file_decref(iofile) ((void)(--(iofile)->refcount <= 0 && (iof_file_free(iofile), 0))) + +int iof_file_seek (iof_file *iofile, long offset, int whence); +long iof_file_tell (iof_file *iofile); +size_t iof_file_size (iof_file *iofile); +int iof_file_eof (iof_file *iofile); + +size_t iof_file_read (void *ptr, size_t size, size_t items, iof_file *iofile); +size_t iof_file_write (const void *ptr, size_t size, size_t items, iof_file *iofile); +size_t iof_file_ensure (iof_file *iofile, size_t bytes); +int iof_file_flush (iof_file *iofile); + +int iof_file_getc (iof_file *iofile); +int iof_file_putc (iof_file *iofile, int c); + +int iof_file_close_input (iof_file *iofile); +int iof_file_reopen_input (iof_file *iofile); + +#define iof_file_reopen(iofile) (((iofile)->flags & IOF_REOPEN_FILE) ? iof_file_reopen_input(iofile) : 1) +#define iof_file_reclose(iofile) (void)(((iofile)->flags & IOF_RECLOSE_FILE) ? iof_file_close_input(iofile) : 0) + +/* wrappers of basic operations for iof */ + +int iof_reader_seek (iof *I, long offset, int whence); +int iof_reader_reseek (iof *I, long offset, int whence); +int iof_writer_seek (iof *I, long offset, int whence); +int iof_writer_reseek (iof *I, long offset, int whence); + +int iof_seek (iof *I, long offset, int whence); +int iof_reseek (iof *I, long offset, int whence); + +long iof_reader_tell (iof *I); +long iof_writer_tell (iof *I); +long iof_tell (iof *I); +size_t iof_fsize (iof *I); + +#define iof_setup_iofile(I, f) (iof_file_incref(f), (I)->iofile = f, (I)->flags |= IOF_FILE) +#define iof_setup_file(I, fh) ((I)->file = fh, (I)->flags |= IOF_FILE_HANDLE) +#define iof_setup_next(I, N) ((I)->next = N, iof_incref(N), (I)->flags |= IOF_NEXT) + +/* file handler reader and writer */ + +UTILAPI iof * iof_setup_file_handle_reader (iof *I, void *buffer, size_t space, FILE *f); +UTILAPI iof * iof_setup_file_handle_writer (iof *O, void *buffer, size_t space, FILE *f); + +#define iof_get_file_handle_reader(buffer, space, fh) iof_setup_file_handle_reader(NULL, buffer, space, fh) +#define iof_get_file_handle_writer(buffer, space, fh) iof_setup_file_handle_writer(NULL, buffer, space, fh) + +/* file reader and writer */ + +UTILAPI iof * iof_setup_file_reader (iof *I, void *buffer, size_t space, const char *filename); +UTILAPI iof * iof_setup_file_writer (iof *O, void *buffer, size_t space, const char *filename); + +#define iof_get_file_reader(buffer, space, filename) iof_setup_file_reader(NULL, buffer, space, filename) +#define iof_get_file_writer(buffer, space, filename) iof_setup_file_writer(NULL, buffer, space, filename) + +/* mem writer */ + +UTILAPI iof * iof_setup_buffer (iof *O, void *buffer, size_t space); +UTILAPI iof * iof_setup_buffermin (iof *O, void *buffer, size_t space, size_t min); + +#define iof_buffer(buffer, space) iof_setup_buffer(NULL, buffer, space) +#define iof_buffermin(buffer, space, min) iof_setup_buffermin(NULL, buffer, space, min) + +UTILAPI iof * iof_buffer_create (size_t space); +#define iof_buffer_new() iof_buffer_create(BUFSIZ) + +/* custom handler */ + +UTILAPI iof * iof_reader (iof *I, void *link, iof_handler reader, const void *s, size_t bytes); +UTILAPI iof * iof_writer (iof *O, void *link, iof_handler writer, void *s, size_t bytes); + +/* stdout wrapper */ + +extern UTILAPI iof iof_stdout; +extern UTILAPI iof iof_stderr; + +/* simple string reader */ + +UTILAPI iof * iof_string_reader (iof *I, const void *s, size_t bytes); + +#define iof_string(I, s, bytes) \ + (((I)->rbuf = (I)->rpos = (const uint8_t *)s), ((I)->rend = (I)->rbuf + (bytes)), ((I)->flags |= IOF_DATA), (I)) + +/* dummies */ + +UTILAPI iof * iof_dummy (void *buffer, size_t space); +UTILAPI iof * iof_null (void *buffer, size_t space); + +/* checking available space */ + +#define iof_loadable(I) ((I)->pos < (I)->end || iof_load(I)) +#define iof_readable(I) ((I)->pos < (I)->end || iof_input(I)) +#define iof_writable(O) ((O)->pos < (O)->end || iof_output(O)) + +#define iof_hloadable iof_loadable +#define iof_iloadable iof_loadable + +#define iof_hreadable iof_readable +#define iof_ireadable iof_readable + +#define iof_hwritable iof_writable +#define iof_iwritable iof_writable + +/* ensure space to write several bytes (several means less then I->space) */ + +#define iof_ensure(O, n) ((O)->pos+(n)-1 < (O)->end || iof_output(O)) // iof_ensure(O, 1) eq iof_writable(O) +#define iof_hensure(O, n) ((O)->hpos+(n)-1 < (O)->hend || iof_output(O)) +#define iof_iensure(O, n) ((O)->ipos+(n)-1 < (O)->iend || iof_output(O)) + +/* reading */ + +UTILAPI int iof_getc (iof *I); +UTILAPI int iof_hgetc (iof *I); +UTILAPI int iof_igetc (iof *I); + +// UTILAPI int iof_cmp (iof *I, const char *s); +// UTILAPI int iof_cmpn (iof *I, const char *s, size_t bytes); + +UTILAPI iof_status iof_pass (iof *I, iof *O); +#define iof_hpass iof_pass +#define iof_ipass iof_pass + +/* readers helpers */ + +UTILAPI size_t iof_read (iof *I, void *s, size_t bytes); +UTILAPI size_t iof_hread (iof *I, void *s, size_t bytes); +UTILAPI size_t iof_iread (iof *I, void *s, size_t bytes); + +UTILAPI size_t iof_skip (iof *I, size_t bytes); +UTILAPI size_t iof_hskip (iof *I, size_t bytes); +UTILAPI size_t iof_iskip (iof *I, size_t bytes); + +/* get */ + +#define iof_pos(I) (*(I)->pos++) +#define iof_hpos(I) (*(I)->hpos++) +#define iof_ipos(I) (*(I)->ipos++) + +#define iof_get(I) (iof_readable(I) ? (int)(*(I)->pos++) : IOFEOF) +#define iof_hget(I) (iof_hreadable(I) ? (int)(*(I)->hpos++) : IOFEOF) +#define iof_iget(I) (iof_ireadable(I) ? (int)(*(I)->ipos++) : IOFEOF) + +#define iof_char(I) (iof_readable(I) ? (int)(*(I)->pos) : IOFEOF) +#define iof_hcurr(I) (iof_hreadable(I) ? (int)(*(I)->hpos) : IOFEOF) +#define iof_icurr(I) (iof_ireadable(I) ? (int)(*(I)->ipos) : IOFEOF) + +#define iof_next(I) (++(I)->pos, iof_char(I)) +#define iof_hnext(I) (++(I)->hpos, iof_hcurr(I)) +#define iof_inext(I) (++(I)->ipos, iof_icurr(I)) + +/* unget */ + +/* +If possible, we just move the position backward. If it is not possible to +move backward, we call iof_backup(I, c) that sets all pointers to the end of +a private backup space, then moves buf AND pos pointers backward and set c at +pos (==buf). We can backup characters as long as there is a private space. If +several calls to iof_backup() are followed by iof_get(), pos pointer +increases in normal way and so the use of another iof_unget() works just fine +by moving the position. Once we swallow all backup characters (when +pos==end), backup handler restores the previous pointers. + +Obviously we assume that the character provided to iof_unget() is always the +character just obtained from iof_get(). We CAN'T just overwrite the character +at a given position as the space we read may not be writable. + +When backup is in use, we can only get bytes until automatically restored. +*/ + +/* backup */ + +/* +#define iof_uses_backup(I) ((I)->more == iof_unget_handler) + +#define iof_save(I, B) \ + ((B)->buf = (I)->buf, (B)->pos = (I)->pos, (B)->end = (I)->end, (B)->space = (I)->space, \ + (B)->link = I->link, (B)->more = (I)->more, (B)->flags = (I)->flags) +#define iof_restore(B, I) iof_save(I, B) + +#define iof_unget(I, c) \ + ((void)(c == (uint8_t)c ? ((I)->pos > (I)->buf ? --(I)->pos : iof_backup(I, c)) : 0) +int iof_backup (iof *I, int c); +*/ + +/* writing */ + +UTILAPI size_t iof_write_file_handle (iof *O, FILE *file); +UTILAPI size_t iof_write_file (iof *O, const char *filename); +UTILAPI size_t iof_write_iofile (iof *O, iof_file *iofile, int savepos); + +UTILAPI int iof_putc (iof *O, int u); +UTILAPI int iof_hputc (iof *O, int u); +UTILAPI int iof_iputc (iof *O, int u); + +UTILAPI size_t iof_write (iof *O, const void *data, size_t size); +UTILAPI size_t iof_hwrite (iof *O, const void *data, size_t size); +UTILAPI size_t iof_iwrite (iof *O, const void *data, size_t size); + +UTILAPI iof_status iof_puts (iof *O, const void *data); +UTILAPI size_t iof_put_string (iof *O, const void *data); +UTILAPI size_t iof_putfs (iof *O, const char *format, ...); +UTILAPI size_t iof_repc (iof *O, char c, size_t bytes); + +#define iof_putl(O, s) iof_write(O, "" s, sizeof(s)-1) +//#define iof_putl iof_puts + +#define iof_set(O, c) (*(O)->pos++ = (uint8_t)(c)) +#define iof_set2(O, c1, c2) (iof_set(O, c1), iof_set(O, c2)) +#define iof_set3(O, c1, c2, c3) (iof_set(O, c1), iof_set(O, c2), iof_set(O, c3)) +#define iof_set4(O, c1, c2, c3, c4) (iof_set(O, c1), iof_set(O, c2), iof_set(O, c3), iof_set(O, c4)) +#define iof_set5(O, c1, c2, c3, c4, c5) (iof_set(O, c1), iof_set(O, c2), iof_set(O, c3), iof_set(O, c4), iof_set(O, c5)) + +#define iof_hset(O, c) (*(O)->hpos++ = (uint16_t)(c)) +#define iof_iset(O, c) (*(O)->ipos++ = (uint32_t)(c)) + +#define iof_put(O, c) ((void)iof_ensure(O, 1), iof_set(O, c)) +#define iof_put2(O, c1, c2) ((void)iof_ensure(O, 2), iof_set2(O, c1, c2)) +#define iof_put3(O, c1, c2, c3) ((void)iof_ensure(O, 3), iof_set3(O, c1, c2, c3)) +#define iof_put4(O, c1, c2, c3, c4) ((void)iof_ensure(O, 4), iof_set4(O, c1, c2, c3, c4)) +#define iof_put5(O, c1, c2, c3, c4, c5) ((void)iof_ensure(O, 5), iof_set5(O, c1, c2, c3, c4, c5)) + +#define iof_hput(O, c) ((void)iof_hensure(O, 1), iof_hset(O, c)) +#define iof_iput(O, c) ((void)iof_iensure(O, 1), iof_iset(O, c)) + +#define iof_put_uc_hex(O, c) iof_put2(O, base16_uc_digit1(c), base16_uc_digit2(c)) +#define iof_put_lc_hex(O, c) iof_put2(O, base16_lc_digit1(c), base16_lc_digit2(c)) +#define iof_set_uc_hex(O, c) iof_set2(O, base16_uc_digit1(c), base16_uc_digit2(c)) +#define iof_set_lc_hex(O, c) iof_set2(O, base16_lc_digit1(c), base16_lc_digit2(c)) +#define iof_put_hex iof_put_uc_hex +#define iof_set_hex iof_set_uc_hex + +/* number from iof; return 1 on success, 0 otherwise */ + +#define iof_scan_sign(I, c, sign) _scan_sign(c, sign, iof_next(I)) +#define iof_scan_integer(I, c, number) _scan_integer(c, number, iof_next(I)) +#define iof_scan_radix(I, c, number, radix) _scan_radix(c, number, radix, iof_next(I)) +#define iof_read_integer(I, c, number) _read_integer(c, number, iof_next(I)) +#define iof_read_radix(I, c, number, radix) _read_radix(c, number, radix, iof_next(I)) + +#define iof_scan_decimal(I, c, number) _scan_decimal(c, number, iof_next(I)) +#define iof_scan_fraction(I, c, number, exponent10) _scan_fraction(c, number, exponent10, iof_next(I)) +#define iof_scan_exponent10(I, c, exponent10) _scan_exponent10(c, exponent10, iof_next(I)) + +UTILAPI int iof_get_int32 (iof *I, int32_t *number); +UTILAPI int iof_get_intlw (iof *I, intlw_t *number); +UTILAPI int iof_get_int64 (iof *I, int64_t *number); + +UTILAPI int iof_get_uint32 (iof *I, uint32_t *number); +UTILAPI int iof_get_uintlw (iof *I, uintlw_t *number); +UTILAPI int iof_get_uint64 (iof *I, uint64_t *number); + +UTILAPI int iof_get_int32_radix (iof *I, int32_t *number, int radix); +UTILAPI int iof_get_intlw_radix (iof *I, intlw_t *number, int radix); +UTILAPI int iof_get_int64_radix (iof *I, int64_t *number, int radix); + +UTILAPI int iof_get_uint32_radix (iof *I, uint32_t *number, int radix); +UTILAPI int iof_get_uintlw_radix (iof *I, uintlw_t *number, int radix); +UTILAPI int iof_get_uint64_radix (iof *I, uint64_t *number, int radix); + +UTILAPI int iof_get_roman (iof *I, unsigned short int *number); + +UTILAPI int iof_get_double (iof *I, double *number); +UTILAPI int iof_get_float (iof *I, float *number); + +UTILAPI int iof_conv_double (iof *I, double *number); +UTILAPI int iof_conv_float (iof *I, float *number); + +/* number to iof; return a number of written bytes */ + +UTILAPI size_t iof_put_int32 (iof *O, int32_t number); +UTILAPI size_t iof_put_intlw (iof *O, intlw_t number); +UTILAPI size_t iof_put_int64 (iof *O, int64_t number); + +UTILAPI size_t iof_put_uint32 (iof *O, uint32_t number); +UTILAPI size_t iof_put_uintlw (iof *O, uintlw_t number); +UTILAPI size_t iof_put_uint64 (iof *O, uint64_t number); + +UTILAPI size_t iof_put_int32_radix (iof *O, int32_t number, int radix); +UTILAPI size_t iof_put_intlw_radix (iof *O, intlw_t number, int radix); +UTILAPI size_t iof_put_int64_radix (iof *O, int64_t number, int radix); + +UTILAPI size_t iof_put_uint32_radix (iof *O, uint32_t number, int radix); +UTILAPI size_t iof_put_uintlw_radix (iof *O, uintlw_t number, int radix); +UTILAPI size_t iof_put_uint64_radix (iof *O, uint64_t number, int radix); + +UTILAPI size_t iof_put_roman_uc (iof *O, uint16_t number); +UTILAPI size_t iof_put_roman_lc (iof *O, uint16_t number); +#define iof_put_roman(I, number) iof_put_roman_uc(I, number) + +UTILAPI size_t iof_put_double(iof *O, double number, int digits); +UTILAPI size_t iof_put_float(iof *O, float number, int digits); + +/* common stuff for binary integers */ + +UTILAPI int iof_get_be_uint2 (iof *I, uint32_t *pnumber); +UTILAPI int iof_get_be_uint3 (iof *I, uint32_t *pnumber); +UTILAPI int iof_get_be_uint4 (iof *I, uint32_t *pnumber); + +UTILAPI int iof_get_le_uint2 (iof *I, uint32_t *pnumber); +UTILAPI int iof_get_le_uint3 (iof *I, uint32_t *pnumber); +UTILAPI int iof_get_le_uint4 (iof *I, uint32_t *pnumber); + +// iof_set() and iof_put() suite casts arguments to uint8_t, so we don't need &0xff mask + +#define iof_set_be_uint1(O, u) iof_set(O, u) +#define iof_set_be_uint2(O, u) iof_set2(O, (u)>>8, u) +#define iof_set_be_uint3(O, u) iof_set3(O, (u)>>16, (u)>>8, u) +#define iof_set_be_uint4(O, u) iof_set4(O, (u)>>24, (u)>>16, (u)>>8, u) + +#define iof_set_le_uint1(O, u) iof_set(O, u) +#define iof_set_le_uint2(O, u) iof_set2(O, u, (u)>>8) +#define iof_set_le_uint3(O, u) iof_set3(O, u, (u)>>8, (u)>>16) +#define iof_set_le_uint4(O, u) iof_set4(O, u, (u)>>8, (u)>>16, (u)>>24) + +#define iof_put_be_uint1(O, u) iof_put(O, u) +#define iof_put_be_uint2(O, u) iof_put2(O, (u)>>8, u) +#define iof_put_be_uint3(O, u) iof_put3(O, (u)>>16, (u)>>8, u) +#define iof_put_be_uint4(O, u) iof_put4(O, (u)>>24, (u)>>16, (u)>>8, u) + +#define iof_put_le_uint1(O, u) iof_put(O, u) +#define iof_put_le_uint2(O, u) iof_put2(O, u, (u)>>8) +#define iof_put_le_uint3(O, u) iof_put3(O, u, (u)>>8, (u)>>16) +#define iof_put_le_uint4(O, u) iof_put4(O, u, (u)>>8, (u)>>16, (u)>>24) + +/* buffer results */ + +#define iof_reader_result(I, size) ((size = iof_left(I)), (I)->pos) +#define iof_writer_result(I, size) ((size = iof_size(I)), (I)->buf) +#define iof_result(I, size) (((I)->flags & IOF_READER) ? iof_reader_result(I, size) : iof_writer_result(I, size)) + +uint8_t * iof_file_input_data (iof_file *iofile, size_t *psize, int *isnew); +//uint8_t * iof_file_reader_data (iof_file *iofile, size_t *size); +//uint8_t * iof_file_writer_data (iof_file *iofile, size_t *size); + +uint8_t * iof_reader_data (iof *I, size_t *psize); +uint8_t * iof_writer_data (iof *O, size_t *psize); +size_t iof_reader_to_file_handle (iof *I, FILE *file); +size_t iof_reader_to_file (iof *I, const char *filename); + +#define iof_loaded(I) ((I)->end = (I)->pos, (I)->pos = (I)->buf, iof_left(I)) + +#define iof_data_to_file_handle(data, size, file) fwrite(data, sizeof(uint8_t), size, file) +UTILAPI size_t iof_data_to_file (const void *data, size_t size, const char *filename); + +UTILAPI size_t iof_result_to_file_handle (iof *F, FILE *file); +UTILAPI size_t iof_result_to_file (iof *F, const char *filename); +UTILAPI void iof_debug (iof *I, const char *filename); + +/* common filters allocator */ + +void iof_filters_init (void); +void iof_filters_free (void); + +void * iof_filter_new (size_t size); +void iof_heap_back (void *data); +#define iof_filter_free(F) iof_heap_back(F) +#define iof_filter_buffer_free(data) iof_heap_back(data) + +// &((void *)pstate + +iof * iof_filter_reader_new (iof_handler handler, size_t statesize, void **pstate); +#define iof_filter_reader(handler, statesize, pstate) iof_filter_reader_new(handler, statesize, (void **)(pstate)) +iof * iof_filter_reader_with_buffer_new (iof_handler handler, size_t statesize, void **pstate, void *buffer, size_t buffersize); +#define iof_filter_reader_with_buffer(handler, statesize, pstate, buffer, buffersize) iof_filter_reader_with_buffer_new(handler, statesize, (void **)(pstate), buffer, buffersize) +iof * iof_filter_writer_new (iof_handler handler, size_t statesize, void **pstate); +#define iof_filter_writer(handler, statesize, pstate) iof_filter_writer_new(handler, statesize, (void **)(pstate)) +iof * iof_filter_writer_with_buffer_new (iof_handler handler, size_t statesize, void **pstate, void *buffer, size_t buffersize); +#define iof_filter_writer_with_buffer(handler, statesize, pstate, buffer, buffersize) iof_filter_writer_with_buffer_new(handler, statesize, (void **)(pstate), buffer, buffersize) + +#define iof_filter_state(statetype, F) (statetype)((F) + 1) + +void iof_free (iof *F); +void iof_discard (iof *F); + +size_t iof_resize_buffer_to (iof *O, size_t space); +#define iof_resize_buffer(O) iof_resize_buffer_to(O, (O)->space << 1) + +size_t iof_decoder_retval (iof *I, const char *type, iof_status status); +size_t iof_encoder_retval (iof *O, const char *type, iof_status status); + +/* filters */ + +iof * iof_filter_file_handle_reader (FILE *file); +iof * iof_filter_file_handle_writer (FILE *file); + +iof * iof_filter_iofile_reader (iof_file *iofile, size_t offset); +iof * iof_filter_iofile_writer (iof_file *iofile, size_t offset); + +iof * iof_filter_file_reader (const char *filename); +iof * iof_filter_file_writer (const char *filename); + +iof * iof_filter_string_reader (const void *s, size_t length); +iof * iof_filter_string_writer (const void *s, size_t length); + +iof * iof_filter_buffer_writer (size_t size); + +iof * iof_filter_stream_reader (FILE *file, size_t offset, size_t length); +iof * iof_filter_stream_coreader (iof_file *iofile, size_t offset, size_t length); + +iof * iof_filter_stream_writer (FILE *file); +iof * iof_filter_stream_cowriter (iof_file *iofile, size_t offset); + +FILE * iof_filter_file_reader_source (iof *I, size_t *poffset, size_t *plength); +iof_file * iof_filter_file_coreader_source (iof *I, size_t *poffset, size_t *plength); +iof * iof_filter_reader_replacement (iof *P, iof_handler handler, size_t statesize, void **pstate); +#define iof_filter_reader_replace(P, handler, statesize, pstate) iof_filter_reader_replacement(P, handler, statesize, (void **)(pstate)) + +#endif
\ No newline at end of file diff --git a/Build/source/texk/web2c/luatexdir/luapplib/util/utillog.c b/Build/source/texk/web2c/luatexdir/luapplib/util/utillog.c new file mode 100644 index 00000000000..1b38f74679c --- /dev/null +++ b/Build/source/texk/web2c/luatexdir/luapplib/util/utillog.c @@ -0,0 +1,60 @@ + +#include <stdio.h> +#include <string.h> // strlen +#include <stdarg.h> +#include "utillog.h" + +#define LOGGER_BUFFER_SIZE 256 +#define LOGGER_PREFIX_SIZE 32 + +typedef struct { + logger_function callback; + void *context; + size_t pfxlen; +} logger_struct; + +static logger_struct logger = { 0, NULL, 0 }; + +static char logger_buffer[LOGGER_BUFFER_SIZE+LOGGER_PREFIX_SIZE]; + +void loggerf (const char *format, ...) +{ + va_list args; + int length; + + va_start(args, format); + length = vsnprintf(logger_buffer + logger.pfxlen, LOGGER_BUFFER_SIZE, format, args); + if (length > 0) + { + if (length > LOGGER_BUFFER_SIZE) + length = LOGGER_BUFFER_SIZE; + } + else + { + loggerf("logger encoding error '%s'", format); + length = (int)strlen(logger_buffer); + } + length += (int)logger.pfxlen; + if (logger.callback) + logger.callback(logger_buffer, logger.context); + else + printf("\n%s\n", logger_buffer); + va_end(args); +} + +void logger_callback (logger_function callback, void *context) +{ + logger.callback = callback; + logger.context = context; +} + +int logger_prefix (const char *prefix) +{ + size_t pfxlen; + pfxlen = strlen(prefix); + if (pfxlen > LOGGER_PREFIX_SIZE) + return 0; + memcpy(logger_buffer, prefix, pfxlen); + logger.pfxlen = pfxlen; + return 1; +} diff --git a/Build/source/texk/web2c/luatexdir/luapplib/util/utillog.h b/Build/source/texk/web2c/luatexdir/luapplib/util/utillog.h new file mode 100644 index 00000000000..c30e0ff0f57 --- /dev/null +++ b/Build/source/texk/web2c/luatexdir/luapplib/util/utillog.h @@ -0,0 +1,10 @@ + +#ifndef UTIL_LOG_H +#define UTIL_LOG_H + +typedef void (*logger_function) (const char *message, void *alien); +void loggerf (const char *format, ...); +void logger_callback (logger_function callback, void *context); +int logger_prefix (const char *prefix); + +#endif
\ No newline at end of file diff --git a/Build/source/texk/web2c/luatexdir/luapplib/util/utillzw.c b/Build/source/texk/web2c/luatexdir/luapplib/util/utillzw.c new file mode 100644 index 00000000000..4cd22121449 --- /dev/null +++ b/Build/source/texk/web2c/luatexdir/luapplib/util/utillzw.c @@ -0,0 +1,701 @@ +/* lzw implementation for postscript/pdf filters +# Notes on LZW + +# Encoder + +Initially the table contains 256 entires for single bytes. Encoder consumes +input bytes trying to find the longest sequence stored so far in the table. +Once it finds a sequence that is not present in the table, it outputs the table +index of the longest sequence found (accumulated bytes except the last +consumed) and pushes the new sequence (accumulated bytes including the last +one) on the top of the table. The last taken byte is not yet written to the +output, it becomes the beginning of the new sequence to accumulate. Initially, +encoder outputs 9-bit codes. While the table grows, the number of bits for each +code increases up to 12. In example, after adding a table entry of index 511 it +is high time to switch to 10-bit bytes. /EarlyChange=true parameter in stream +dictionary (both postscript and pdf) informs to increase the number of bits one +code earlier then necessary. Looks pretty much like an early days bug that +became a specification :) I have never found a PDF having /EarlyChange key +specified anyway. + +Once the table becomes full (or when encoder decides it is worthy), +a clear-table marker (code 256) purges the table and restores codes length to +9. End-of-data marker (code 257) ends the stream. Conventionally, the beginning +of the stream starts with clear-table marker. + +Postscript allows to provide a /UnitLength which determines the bit length of +codes. The above description assumes UnitLength=8 (default). Allowed values are +from 3 to 8. Different UnitLength also affects markers; clear-table is then +2^UnitLength and end-of-data marker is 2^UnitLenth+1. + +Encoder outputs 9-12bit codes that are packed into bytes using high-bits-first +scheme (default) or low-bits-scheme. + +PDF spec p. 73 (PS spec p. 135 gives an mistaken output sequence and so +mistaken output bytes) + +Input character sequence (decimal) +45 45 45 45 45 65 45 45 45 66 + +Output 9bit codes (decimal) +256 45 258 258 65 259 66 257 + +Output 9bit codes (binary) +100000000 000101101 100000010 100000010 001000001 100000011 001000010 100000001 + +Output bytes (LowBitsFirst=false); eight high-order bits of code becomes +the first byte, remaining low-order bit of code becomes the high-order bit of the +next byte; +10000000 00001011 01100000 01010000 00100010 00001100 00001100 10000101 00000001 +-> 80 0B 60 50 22 0C 0C 85 01 + +Output bytes (binary, LowBitsFirst=true); eight low-order bits of code becomes +the first byte, remaining high-order bit of code becomes low-order bit of the +next byte; +00000000 01011011 00001000 00010100 00011000 01100100 10100000 10000000 10010000 +-> 00 5B 08 14 18 64 A0 80 90 + +# Decoder + +Decoder consumes input bytes transforming them to 9 to 12 bit codes. Initially +it starts with 9bit codes and the table of 258 fixed codes (same as encoder). +Basically, it interprets incoming codes as table indices (except 256 and 257 +markers) and it outputs byte sequences stored at given indices. It also +upbuilds the table and changes the number of bits of codes when necessary. The +key point on lzw is that both encoder and decoder builds the table +synchronously. + +However, decoder needs some "knowledge" about how encoder works to be able to +interpret a table index that it doesn't have so far. Look that the output from +encoder in the example above. The first output code is conventional clear-table +(256). Then comes a code 45. So far so good, decoder interprets code 45 as +a (fixed) entry of the table, emitting byte 45. The next code is 258, which is +should be interpreted as an index in the table. Oops, encoder doesn't have one +yet. If that occurs, it means that encoder was able to output the new entry +code just after adding it to a table. It means that + + sequence_before + next_byte == next_byte + sequence_after + +This may happen not only for sequences like 45 45 45, but also symmetric series +such as abcbabcba; abcb + a == a + bcba. Decoder must be aware of that and if +it gets a code one larger than the top table index, it should create one on-fly +by appending last entry sequence by the first by of the last entry. + +# UnitLength + +Postscript specification mentions about UnitLength parameter that can be used +in LZW decoder (not allowed in encoder), with possible values from 3 to 8. This +parameter determines the number of bits per code; form UnitLength + 1 to 12. It +also determines which codes are used for clear-table marker (2^UnitLength) and +end-of-data marker ((2^UnitLength)+1). Postscript specification says (page 134): + +"Initially, the code length is (UnitLength + 1) bits and the table contains only +entries for the (2^UnitLength + 2) fixed codes. As encoding proceeds, entries are +appended to the table, associating new codes with longer and longer input character +sequences. The encoding and decoding filters maintain identical copies of +this table." + +Later on page 136 Postscript specification says: + +"Data that has been LZW-encoded with a UnitLength less than 8 consists only of +codes in the range 0 to 2^UnitLength - 1; consequently, the LZWDecode filter produces +only codes in that range when read. UnitLength also affects the encoded +representation, as described above." + +UnitLength (Postscript only) and LowBitsFirst are used only by decoder. +EarlyChange should obviously be respected by both encoder and decoder. When +table index reaches current bit length boundary (511, 1023, ...) it must react +by increasing the number of bits of input code. But if the index reaches it +maximum value (when the table is full), decoder is NOT supposed to clear the +table. When the table is full, encoder must emit clear-table marker and it +emits this code using 12 bits and reinitialize code bits after that. It means +that, when the table is full, decoder should get one more 12-bit code (which +should be clear-table marker) and actually clear the table and reinitialize +code bits after that. + +# Clear-table vs last entry track (after tries and checks) + +It is also not quite clear what should actually happen when encoder gets a full +table and it is supposed to emit clear-table marker. When it gets full, it +means that it has just appended another entry to the table. And that happens +only the input sequence collected so far plus the last byte is not present in +the table. Encoder is supposed to output the table index of the present +sequence and set the recent byte as a starting index of the new sequence to be +collected. Even if it is time to clear the table, encoder is still supposed to +keep the track of the last table entry. Decoder, however, must drop the track of the +last code on clear-table. + +# Decoder table vs encoder table + +While decoding we need query lzw table by (subsequent) numeric codes and output +character sequences stored in the table. While encoding we need to query the +table on every input byte and fetch indices pointing to character sequences. +Note that we never need to query the entire table for the longest sequence +found so far. The encoder table do not need to access the longest character +sequence at one piece. It is enough to keep the track of the current table +index and the very next byte. We organize an encoder table into a search tree, +where every node contains its table index (value) and last byte (key). Except +initial tree content, every node is created on the base of the previous node +and it conceptually point the sequence represented by that nodo consists of the +previous node sequence plus the next byte. + +Every new node is a descendant of the node it has been derived from. Every node +has a map (a search subtree) indexed by suffix byte value, pointing to +descendants nodes. Every node also has binary tentackles (left/right fields) +necessary to search the map (except initials, every node lives in a map of some +ancestor node). The key point is that on every input byte we don't search the +entire tree, but only the map of the current node children. The map tree is +a simple binary tree with no balancing mechanism (not worthy to optimize an +ephemeric structure that may be upbuilt more often then queried). + +In our implementation, decoder table requires 4069 entries (topmost index 4095). +Encoder table, however, needs 4097 entries to handle the case when EarlyIndex +parameter is 0 (I have never a chance to test that in practise). The node of index +4096 might be added to a search tree, but its code is never emitted; the lookup +is purged just after adding that node. + +todo: +- support for LowBitsFirst encoding +*/ + +#include "utilmem.h" +#include "utillzw.h" + +/* filter state struct */ + +typedef struct lzw_entry { + union { + const char *rdata; // to be able to init with string literal + char *data; + }; + int size; +} lzw_entry; + +#define lzw_index short + +typedef struct lzw_node lzw_node; + +struct lzw_node { + lzw_index index; + unsigned char suffix; + lzw_node *left; + lzw_node *right; + lzw_node *map; +}; + +struct lzw_state { + union { + lzw_node *lookup; /* encoder table */ + lzw_entry *table; /* decoder table */ + }; + lzw_index index; /* table index */ + union { + lzw_node *lastnode; /* previous encoder table node */ + struct { + lzw_entry *lastentry; /* previous decoder table entry */ + int tailbytes; /* num of bytes of lastentry not yet written out */ + }; + }; + int basebits; /* /UnitLength parameter (8) */ + int codebits; /* current code bits */ + int lastbyte; /* previosly read byte */ + int tailbits; /* lastbyte bits not yet consumed */ + int flush; /* encoder */ + int flags; /* options */ +}; + +/* macros */ + +#define LZW_MIN_BITS 3 +#define LZW_MAX_BITS 12 +#define LZW_TABLE_SIZE (1 << LZW_MAX_BITS) +#define LZW_LOOKUP_SIZE (LZW_TABLE_SIZE + 1) + +#define lzw_bit_range(bits) (bits >= LZW_MIN_BITS && bits <= LZW_BASE_BITS) +#define lzw_base_bits(flags) (flags & ((1 << 4) - 1)) // 4 low bits of flags is basebits (UnitLength) + +#define lzw_initial_codes(state) (1 << state->basebits) +#define lzw_clear_code(state) lzw_initial_codes(state) +#define lzw_eod_code(state) (lzw_initial_codes(state) + 1) +#define lzw_initial_index(state) (lzw_initial_codes(state) + 2) + +#define lzw_max_index(state) ((1 << state->codebits) - ((state->flags & LZW_EARLY_INDEX) ? 1 : 0)) +#define lzw_check_bits(state) ((void)(state->index == lzw_max_index(state) && state->codebits < LZW_MAX_BITS && ++state->codebits)) + +#define lzw_malloc util_malloc +#define lzw_free util_free + +/* decoder */ + +static struct lzw_entry lzw_initial_table[] = { + {{"\x00"}, 1}, {{"\x01"}, 1}, {{"\x02"}, 1}, {{"\x03"}, 1}, {{"\x04"}, 1}, {{"\x05"}, 1}, {{"\x06"}, 1}, {{"\x07"}, 1}, {{"\x08"}, 1}, {{"\x09"}, 1}, {{"\x0A"}, 1}, {{"\x0B"}, 1}, {{"\x0C"}, 1}, {{"\x0D"}, 1}, {{"\x0E"}, 1}, {{"\x0F"}, 1}, + {{"\x10"}, 1}, {{"\x11"}, 1}, {{"\x12"}, 1}, {{"\x13"}, 1}, {{"\x14"}, 1}, {{"\x15"}, 1}, {{"\x16"}, 1}, {{"\x17"}, 1}, {{"\x18"}, 1}, {{"\x19"}, 1}, {{"\x1A"}, 1}, {{"\x1B"}, 1}, {{"\x1C"}, 1}, {{"\x1D"}, 1}, {{"\x1E"}, 1}, {{"\x1F"}, 1}, + {{"\x20"}, 1}, {{"\x21"}, 1}, {{"\x22"}, 1}, {{"\x23"}, 1}, {{"\x24"}, 1}, {{"\x25"}, 1}, {{"\x26"}, 1}, {{"\x27"}, 1}, {{"\x28"}, 1}, {{"\x29"}, 1}, {{"\x2A"}, 1}, {{"\x2B"}, 1}, {{"\x2C"}, 1}, {{"\x2D"}, 1}, {{"\x2E"}, 1}, {{"\x2F"}, 1}, + {{"\x30"}, 1}, {{"\x31"}, 1}, {{"\x32"}, 1}, {{"\x33"}, 1}, {{"\x34"}, 1}, {{"\x35"}, 1}, {{"\x36"}, 1}, {{"\x37"}, 1}, {{"\x38"}, 1}, {{"\x39"}, 1}, {{"\x3A"}, 1}, {{"\x3B"}, 1}, {{"\x3C"}, 1}, {{"\x3D"}, 1}, {{"\x3E"}, 1}, {{"\x3F"}, 1}, + {{"\x40"}, 1}, {{"\x41"}, 1}, {{"\x42"}, 1}, {{"\x43"}, 1}, {{"\x44"}, 1}, {{"\x45"}, 1}, {{"\x46"}, 1}, {{"\x47"}, 1}, {{"\x48"}, 1}, {{"\x49"}, 1}, {{"\x4A"}, 1}, {{"\x4B"}, 1}, {{"\x4C"}, 1}, {{"\x4D"}, 1}, {{"\x4E"}, 1}, {{"\x4F"}, 1}, + {{"\x50"}, 1}, {{"\x51"}, 1}, {{"\x52"}, 1}, {{"\x53"}, 1}, {{"\x54"}, 1}, {{"\x55"}, 1}, {{"\x56"}, 1}, {{"\x57"}, 1}, {{"\x58"}, 1}, {{"\x59"}, 1}, {{"\x5A"}, 1}, {{"\x5B"}, 1}, {{"\x5C"}, 1}, {{"\x5D"}, 1}, {{"\x5E"}, 1}, {{"\x5F"}, 1}, + {{"\x60"}, 1}, {{"\x61"}, 1}, {{"\x62"}, 1}, {{"\x63"}, 1}, {{"\x64"}, 1}, {{"\x65"}, 1}, {{"\x66"}, 1}, {{"\x67"}, 1}, {{"\x68"}, 1}, {{"\x69"}, 1}, {{"\x6A"}, 1}, {{"\x6B"}, 1}, {{"\x6C"}, 1}, {{"\x6D"}, 1}, {{"\x6E"}, 1}, {{"\x6F"}, 1}, + {{"\x70"}, 1}, {{"\x71"}, 1}, {{"\x72"}, 1}, {{"\x73"}, 1}, {{"\x74"}, 1}, {{"\x75"}, 1}, {{"\x76"}, 1}, {{"\x77"}, 1}, {{"\x78"}, 1}, {{"\x79"}, 1}, {{"\x7A"}, 1}, {{"\x7B"}, 1}, {{"\x7C"}, 1}, {{"\x7D"}, 1}, {{"\x7E"}, 1}, {{"\x7F"}, 1}, + {{"\x80"}, 1}, {{"\x81"}, 1}, {{"\x82"}, 1}, {{"\x83"}, 1}, {{"\x84"}, 1}, {{"\x85"}, 1}, {{"\x86"}, 1}, {{"\x87"}, 1}, {{"\x88"}, 1}, {{"\x89"}, 1}, {{"\x8A"}, 1}, {{"\x8B"}, 1}, {{"\x8C"}, 1}, {{"\x8D"}, 1}, {{"\x8E"}, 1}, {{"\x8F"}, 1}, + {{"\x90"}, 1}, {{"\x91"}, 1}, {{"\x92"}, 1}, {{"\x93"}, 1}, {{"\x94"}, 1}, {{"\x95"}, 1}, {{"\x96"}, 1}, {{"\x97"}, 1}, {{"\x98"}, 1}, {{"\x99"}, 1}, {{"\x9A"}, 1}, {{"\x9B"}, 1}, {{"\x9C"}, 1}, {{"\x9D"}, 1}, {{"\x9E"}, 1}, {{"\x9F"}, 1}, + {{"\xA0"}, 1}, {{"\xA1"}, 1}, {{"\xA2"}, 1}, {{"\xA3"}, 1}, {{"\xA4"}, 1}, {{"\xA5"}, 1}, {{"\xA6"}, 1}, {{"\xA7"}, 1}, {{"\xA8"}, 1}, {{"\xA9"}, 1}, {{"\xAA"}, 1}, {{"\xAB"}, 1}, {{"\xAC"}, 1}, {{"\xAD"}, 1}, {{"\xAE"}, 1}, {{"\xAF"}, 1}, + {{"\xB0"}, 1}, {{"\xB1"}, 1}, {{"\xB2"}, 1}, {{"\xB3"}, 1}, {{"\xB4"}, 1}, {{"\xB5"}, 1}, {{"\xB6"}, 1}, {{"\xB7"}, 1}, {{"\xB8"}, 1}, {{"\xB9"}, 1}, {{"\xBA"}, 1}, {{"\xBB"}, 1}, {{"\xBC"}, 1}, {{"\xBD"}, 1}, {{"\xBE"}, 1}, {{"\xBF"}, 1}, + {{"\xC0"}, 1}, {{"\xC1"}, 1}, {{"\xC2"}, 1}, {{"\xC3"}, 1}, {{"\xC4"}, 1}, {{"\xC5"}, 1}, {{"\xC6"}, 1}, {{"\xC7"}, 1}, {{"\xC8"}, 1}, {{"\xC9"}, 1}, {{"\xCA"}, 1}, {{"\xCB"}, 1}, {{"\xCC"}, 1}, {{"\xCD"}, 1}, {{"\xCE"}, 1}, {{"\xCF"}, 1}, + {{"\xD0"}, 1}, {{"\xD1"}, 1}, {{"\xD2"}, 1}, {{"\xD3"}, 1}, {{"\xD4"}, 1}, {{"\xD5"}, 1}, {{"\xD6"}, 1}, {{"\xD7"}, 1}, {{"\xD8"}, 1}, {{"\xD9"}, 1}, {{"\xDA"}, 1}, {{"\xDB"}, 1}, {{"\xDC"}, 1}, {{"\xDD"}, 1}, {{"\xDE"}, 1}, {{"\xDF"}, 1}, + {{"\xE0"}, 1}, {{"\xE1"}, 1}, {{"\xE2"}, 1}, {{"\xE3"}, 1}, {{"\xE4"}, 1}, {{"\xE5"}, 1}, {{"\xE6"}, 1}, {{"\xE7"}, 1}, {{"\xE8"}, 1}, {{"\xE9"}, 1}, {{"\xEA"}, 1}, {{"\xEB"}, 1}, {{"\xEC"}, 1}, {{"\xED"}, 1}, {{"\xEE"}, 1}, {{"\xEF"}, 1}, + {{"\xF0"}, 1}, {{"\xF1"}, 1}, {{"\xF2"}, 1}, {{"\xF3"}, 1}, {{"\xF4"}, 1}, {{"\xF5"}, 1}, {{"\xF6"}, 1}, {{"\xF7"}, 1}, {{"\xF8"}, 1}, {{"\xF9"}, 1}, {{"\xFA"}, 1}, {{"\xFB"}, 1}, {{"\xFC"}, 1}, {{"\xFD"}, 1}, {{"\xFE"}, 1}, {{"\xFF"}, 1} +}; + +#define lzw_entry_at(state, index) (&state->table[index]) + +static lzw_state * lzw_decoder_init_table (lzw_state *state, lzw_entry *table, int flags) +{ + state->basebits = lzw_base_bits(flags); // first four bits or flags + if (!lzw_bit_range(state->basebits)) + return NULL; + state->flags = flags; + if ((state->table = table) == NULL) + { + state->table = (lzw_entry *)lzw_malloc(LZW_TABLE_SIZE * sizeof(lzw_entry)); + state->flags |= LZW_TABLE_ALLOC; + } + memcpy(state->table, lzw_initial_table, (size_t)lzw_initial_codes(state)*sizeof(lzw_entry)); + // memset(&state->table[lzw_initial_codes(state)], 0, 2*sizeof(lzw_entry)); // eod and clear entries never accessed + state->codebits = state->basebits + 1; + state->index = lzw_initial_index(state); + state->lastentry = NULL; + state->tailbytes = 0; + state->lastbyte = 0; + state->tailbits = 0; + return state; +} + +lzw_state * lzw_decoder_init (lzw_state *state, int flags) +{ + return lzw_decoder_init_table(state, NULL, flags); +} + +static void lzw_decoder_clear (lzw_state *state) +{ + lzw_entry *entry; + lzw_index initindex = lzw_initial_index(state); + while (state->index > initindex) + { + entry = lzw_entry_at(state, --state->index); + lzw_free(entry->data); + // entry->data = NULL; + // entry->size = 0; + } + state->lastentry = NULL; + state->tailbytes = 0; + state->codebits = state->basebits + 1; +} + +void lzw_decoder_close (lzw_state *state) +{ + lzw_decoder_clear(state); + if (state->flags & LZW_TABLE_ALLOC) + lzw_free(state->table); +} + +static int lzw_next_entry (lzw_state *state, lzw_entry *nextentry) +{ + lzw_entry *lastentry, *newentry; + if ((lastentry = state->lastentry) == NULL) + return 1; /* its ok */ + if (state->index == LZW_TABLE_SIZE) + return 0; /* invalid input; eod marker expected earlier */ + /* put the new entry on the top of the table */ + newentry = lzw_entry_at(state, state->index++); + /* its size is the last entrtyy size plus 1 */ + newentry->size = lastentry->size + 1; + /* its content is the content of the last entry, */ + newentry->data = (char *)lzw_malloc((size_t)newentry->size); + memcpy(newentry->data, lastentry->data, lastentry->size); + /* plus the first byte of the new entry (usually fixed code entry) */ + newentry->data[newentry->size - 1] = nextentry->data[0]; + return 1; +} + +#define lzw_write_bytes(O, state) ((state->tailbytes -= (int)iof_write(O, state->lastentry->data, (size_t)state->tailbytes)) == 0) + +iof_status lzw_decode_state (iof *I, iof *O, lzw_state *state) +{ + const lzw_index clear = lzw_clear_code(state), eod = lzw_eod_code(state); + lzw_index code; + lzw_entry *entry; + if (state->lastentry != NULL) + { /* write out the tail from the last call */ + if (state->tailbytes > 0 && !lzw_write_bytes(O, state)) + return IOFFULL; + /* do what we normally do at the end of the loop body below */ + lzw_check_bits(state); + } + // if (state->flags & LZW_LOW_BITS_FIRST) + // return IOFERR; + while (1) + { + /* get input code of length state->codebits */ + code = (state->lastbyte & ((1 << state->tailbits) - 1)) << (state->codebits - state->tailbits); + for (state->tailbits -= state->codebits; state->tailbits < 0; ) + { + get_code: + if ((state->lastbyte = iof_get(I)) < 0) + return state->flush ? IOFEOF : state->lastbyte; + state->tailbits += 8; + if (state->tailbits < 0) + { + code |= (state->lastbyte << (-state->tailbits)); + goto get_code; + } + else + { + code |= (state->lastbyte >> state->tailbits); + break; + } + } + /* interpret the code */ + if (code < state->index) + { /* single byte code or special marker */ + if (code == clear) + { + lzw_decoder_clear(state); + continue; + } + if (code == eod) + return IOFEOF; + entry = lzw_entry_at(state, code); + if (!lzw_next_entry(state, entry)) + return IOFERR; + } + else if (code == state->index) + { /* apparently encoder has emitted the code of the key just created (see notes) */ + if (!lzw_next_entry(state, state->lastentry)) + return IOFERR; + entry = lzw_entry_at(state, state->index - 1); + } + else + { /* invalid input code */ + return IOFERR; + } + /* record the entry found */ + state->lastentry = entry; + /* emit the sequence pointed by that entry */ + state->tailbytes = entry->size; + if (!lzw_write_bytes(O, state)) + return IOFFULL; + /* check and update code bits */ + lzw_check_bits(state); + } + return state->lastbyte; // never reached +} + +/* encoder */ + +#define lzw_node_at(state, index) (&state->lookup[index]) + +#define lzw_node_init(node, i, c) (node->index = i, node->suffix = c, node->left = NULL, node->right = NULL, node->map = NULL) + +static lzw_state * lzw_encoder_init_table (lzw_state *state, lzw_node *lookup, int flags) +{ + lzw_index index; + lzw_node *node; + state->basebits = lzw_base_bits(flags); // first four bits of flags is base bits of code (default 8) + if (!lzw_bit_range(state->basebits)) + return NULL; + state->flags = flags; + if ((state->lookup = lookup) == NULL) + { + state->lookup = lzw_malloc(LZW_LOOKUP_SIZE*sizeof(lzw_node)); + state->flags |= LZW_TABLE_ALLOC; + } + state->index = lzw_initial_index(state); + for (index = 0; index < lzw_initial_codes(state); ++index) + { + node = lzw_node_at(state, index); + lzw_node_init(node, index, (unsigned char)index); + } + state->codebits = state->basebits + 1; + state->lastnode = NULL; + state->lastbyte = 0; + state->tailbits = 0; + return state; +} + +lzw_state * lzw_encoder_init (lzw_state *state, int flags) +{ + return lzw_encoder_init_table(state, NULL, flags); +} + +void lzw_encoder_close (lzw_state *state) +{ + if (state->flags & LZW_TABLE_ALLOC) + lzw_free(state->lookup); +} + +static void lzw_encoder_clear (lzw_state *state) +{ + lzw_node *node; + lzw_index index; + /* clear fixed nodes */ + for (index = 0; index < lzw_initial_codes(state); ++index) + { + node = lzw_node_at(state, index); + lzw_node_init(node, index, (unsigned char)index); + } + /* reset table index */ + state->index = lzw_initial_index(state); + /* reset code bits */ + state->codebits = state->basebits + 1; +} + +static void lzw_put_code (iof *O, lzw_state *state, lzw_index code, int todobits) +{ + int leftbits, rightbits; + do + { + leftbits = 8 - state->tailbits; + rightbits = todobits - leftbits; + if (rightbits >= 0) + { + state->lastbyte |= (code >> rightbits); + iof_put(O, state->lastbyte); + code = code & ((1 << rightbits) - 1); + todobits -= leftbits; + state->lastbyte = 0; + state->tailbits = 0; + } + else + { + state->lastbyte |= (code << (-rightbits)); + state->tailbits += todobits; + return; + } + } while (1); +} + +static iof_status lzw_encode_last (iof *O, lzw_state *state) +{ + if (state->flush) + { + /* put the last code if any */ + if (state->lastnode != NULL) + lzw_put_code(O, state, state->lastnode->index, state->codebits); + /* put eod marker, */ + lzw_put_code(O, state, lzw_eod_code(state), state->codebits); + /* with tail bits set to 0 */ + if (state->tailbits > 0) + lzw_put_code(O, state, 0, 8 - state->tailbits); + return IOFEOF; + } + return IOFEMPTY; +} + +static lzw_node * lzw_node_push (lzw_state *state, unsigned char suffix) +{ + lzw_node *node; + node = lzw_node_at(state, state->index); + lzw_node_init(node, state->index, suffix); + ++state->index; + return node; +} + +static int lzw_next_node (lzw_state *state, unsigned char suffix) +{ + lzw_node *node; + if ((node = state->lastnode->map) == NULL) + { + state->lastnode->map = lzw_node_push(state, suffix); + return 0; + } + while (1) + { + if (suffix < node->suffix) + { + if (node->left == NULL) + { + node->left = lzw_node_push(state, suffix); + return 0; + } + node = node->left; + } + else if (suffix > node->suffix) + { + if (node->right == NULL) + { + node->right = lzw_node_push(state, suffix); + return 0; + } + node = node->right; + } + else + { + state->lastnode = node; + return 1; + } + } + return 0; // never reached +} + +iof_status lzw_encode_state (iof *I, iof *O, lzw_state *state) +{ + int byte; + if (state->lastnode == NULL) + { /* first call only; following convention, put clear-table marker */ + if (!iof_ensure(O, 2)) + return IOFFULL; + lzw_put_code(O, state, lzw_clear_code(state), state->codebits); + /* get the first input byte and initialize the current table entry */ + if ((byte = iof_get(I)) < 0) + return lzw_encode_last(O, state); + state->lastnode = lzw_node_at(state, byte); + } + while (iof_ensure(O, 2)) + { /* we need to write at most 2 bytes on each iteration */ + if ((byte = iof_get(I)) < 0) + return lzw_encode_last(O, state); + if (lzw_next_node(state, (unsigned char)byte) == 0) + { /* means that the key hasn't been found and the new entry has just been created */ + /* output the code pointing the longest sequence so far */ + lzw_put_code(O, state, state->lastnode->index, state->codebits); + /* update code bits */ + if (state->index == lzw_max_index(state) + 1) + { + if (state->codebits < LZW_MAX_BITS) + ++state->codebits; + else + { + /* put clear-table marker */ + lzw_put_code(O, state, lzw_clear_code(state), state->codebits); + /* reset the table */ + lzw_encoder_clear(state); + } + } + /* in any case, recent byte becomes the current table code */ + state->lastnode = lzw_node_at(state, byte); + } + /* otherwise no new entry is appended and state->lastnode points the longer sequence just found */ + } + return IOFFULL; +} + +/* single call codecs */ + +iof_status lzw_decode (iof *I, iof *O, int flags) +{ + lzw_state state = { { 0 } }; // shut overactive warnings + lzw_entry table[LZW_TABLE_SIZE]; + int ret; + lzw_decoder_init_table(&state, table, flags); + state.flush = 1; + ret = lzw_decode_state(I, O, &state); + // iof_flush(O); // ? + lzw_decoder_close(&state); + return ret; +} + +iof_status lzw_encode (iof *I, iof *O, int flags) +{ + lzw_state state; + lzw_node lookup[LZW_LOOKUP_SIZE]; + int ret; + lzw_encoder_init_table(&state, lookup, flags); + state.flush = 1; + ret = lzw_encode_state(I, O, &state); + // iof_flush(O); // ? + lzw_encoder_close(&state); + return ret; +} + +/* filters */ + +// lzw decoder function + +static size_t lzw_decoder (iof *F, iof_mode mode) +{ + lzw_state *state; + iof_status status; + size_t tail; + + state = iof_filter_state(lzw_state *, F); + switch(mode) + { + case IOFLOAD: + case IOFREAD: + if (F->flags & IOF_STOPPED) + return 0; + tail = iof_tail(F); + F->pos = F->buf + tail; + F->end = F->buf + F->space; + do { + status = lzw_decode_state(F->next, F, state); + } while (mode == IOFLOAD && status == IOFFULL && iof_resize_buffer(F)); + return iof_decoder_retval(F, "lzw", status); + case IOFCLOSE: + lzw_decoder_close(state); + iof_free(F); + return 0; + default: + break; + } + return 0; +} + +// lzw encoder function + +static size_t lzw_encoder (iof *F, iof_mode mode) +{ + lzw_state *state; + iof_status status; + + state = iof_filter_state(lzw_state *, F); + switch (mode) + { + case IOFFLUSH: + state->flush = 1; + // fall through + case IOFWRITE: + F->end = F->pos; + F->pos = F->buf; + status = lzw_encode_state(F, F->next, state); + return iof_encoder_retval(F, "lzw", status); + case IOFCLOSE: + if (!state->flush) + lzw_encoder(F, IOFFLUSH); + lzw_encoder_close(state); + iof_free(F); + return 0; + default: + break; + } + return 0; +} + +iof * iof_filter_lzw_decoder (iof *N, int flags) +{ + iof *I; + lzw_state *state; + I = iof_filter_reader(lzw_decoder, sizeof(lzw_state), &state); + iof_setup_next(I, N); + if (lzw_decoder_init(state, flags) == NULL) + { + iof_discard(I); + return NULL; + } + state->flush = 1; + return I; +} + +iof * iof_filter_lzw_encoder (iof *N, int flags) +{ + iof *O; + lzw_state *state; + O = iof_filter_writer(lzw_encoder, sizeof(lzw_state), &state); + iof_setup_next(O, N); + if (lzw_encoder_init(state, flags) == NULL) + { + iof_discard(O); + return NULL; + } + return O; +} diff --git a/Build/source/texk/web2c/luatexdir/luapplib/util/utillzw.h b/Build/source/texk/web2c/luatexdir/luapplib/util/utillzw.h new file mode 100644 index 00000000000..9e3a085d43c --- /dev/null +++ b/Build/source/texk/web2c/luatexdir/luapplib/util/utillzw.h @@ -0,0 +1,30 @@ +#ifndef UTIL_LZW_H +#define UTIL_LZW_H + +#include "utiliof.h" + +typedef struct lzw_state lzw_state; + +#define LZW_BASE_BITS 8 +#define LZW_TABLE_ALLOC (1<<4) +#define LZW_EARLY_INDEX (1<<5) +//#define LZW_LOW_BITS_FIRST (1<<6) +#define LZW_DECODER_DEFAULTS (LZW_BASE_BITS|LZW_EARLY_INDEX|0) +#define LZW_ENCODER_DEFAULTS (LZW_BASE_BITS|LZW_EARLY_INDEX|0) + +lzw_state * lzw_decoder_init (lzw_state *state, int flags); +lzw_state * lzw_encoder_init (lzw_state *state, int flags); + +void lzw_decoder_close (lzw_state *state); +void lzw_encoder_close (lzw_state *state); + +iof_status lzw_encode_state (iof *I, iof *O, lzw_state *state); +iof_status lzw_decode_state (iof *I, iof *O, lzw_state *state); + +iof_status lzw_encode (iof *I, iof *O, int flags); +iof_status lzw_decode (iof *I, iof *O, int flags); + +iof * iof_filter_lzw_decoder (iof *N, int flags); +iof * iof_filter_lzw_encoder (iof *N, int flags); + +#endif
\ No newline at end of file diff --git a/Build/source/texk/web2c/luatexdir/luapplib/util/utilmd5.c b/Build/source/texk/web2c/luatexdir/luapplib/util/utilmd5.c new file mode 100644 index 00000000000..a2926b0aa8b --- /dev/null +++ b/Build/source/texk/web2c/luatexdir/luapplib/util/utilmd5.c @@ -0,0 +1,413 @@ +/* md5 implementation by Peter Deutsch (ghost@aladdin.com) with some convenience shorthands */ + +/* begin of md5.c */ + +/* + Copyright (C) 1999, 2000, 2002 Aladdin Enterprises. All rights reserved. + + This software is provided 'as-is', without any express or implied + warranty. In no event will the authors be held liable for any damages + arising from the use of this software. + + Permission is granted to anyone to use this software for any purpose, + including commercial applications, and to alter it and redistribute it + freely, subject to the following restrictions: + + 1. The origin of this software must not be misrepresented; you must not + claim that you wrote the original software. If you use this software + in a product, an acknowledgment in the product documentation would be + appreciated but is not required. + 2. Altered source versions must be plainly marked as such, and must not be + misrepresented as being the original software. + 3. This notice may not be removed or altered from any source distribution. + + L. Peter Deutsch + ghost@aladdin.com + + */ +/* $Id: md5.c,v 1.6 2002/04/13 19:20:28 lpd Exp $ */ +/* + Independent implementation of MD5 (RFC 1321). + + This code implements the MD5 Algorithm defined in RFC 1321, whose + text is available at + http://www.ietf.org/rfc/rfc1321.txt + The code is derived from the text of the RFC, including the test suite + (section A.5) but excluding the rest of Appendix A. It does not include + any code or documentation that is identified in the RFC as being + copyrighted. + + The original and principal author of md5.c is L. Peter Deutsch + <ghost@aladdin.com>. Other authors are noted in the change history + that follows (in reverse chronological order): + + 2002-04-13 lpd Clarified derivation from RFC 1321; now handles byte order + either statically or dynamically; added missing #include <string.h> + in library. + 2002-03-11 lpd Corrected argument list for main(), and added int return + type, in test program and T value program. + 2002-02-21 lpd Added missing #include <stdio.h> in test program. + 2000-07-03 lpd Patched to eliminate warnings about "constant is + unsigned in ANSI C, signed in traditional"; made test program + self-checking. + 1999-11-04 lpd Edited comments slightly for automatic TOC extraction. + 1999-10-18 lpd Fixed typo in header comment (ansi2knr rather than md5). + 1999-05-03 lpd Original version. + */ + +//#include "md5.h" +#include "utilmd5.h" +#include <string.h> + +#undef BYTE_ORDER /* 1 = big-endian, -1 = little-endian, 0 = unknown */ +#ifdef ARCH_IS_BIG_ENDIAN +# define BYTE_ORDER (ARCH_IS_BIG_ENDIAN ? 1 : -1) +#else +# define BYTE_ORDER 0 +#endif + +#define T_MASK ((uint32_t)~0) +#define T1 /* 0xd76aa478 */ (T_MASK ^ 0x28955b87) +#define T2 /* 0xe8c7b756 */ (T_MASK ^ 0x173848a9) +#define T3 0x242070db +#define T4 /* 0xc1bdceee */ (T_MASK ^ 0x3e423111) +#define T5 /* 0xf57c0faf */ (T_MASK ^ 0x0a83f050) +#define T6 0x4787c62a +#define T7 /* 0xa8304613 */ (T_MASK ^ 0x57cfb9ec) +#define T8 /* 0xfd469501 */ (T_MASK ^ 0x02b96afe) +#define T9 0x698098d8 +#define T10 /* 0x8b44f7af */ (T_MASK ^ 0x74bb0850) +#define T11 /* 0xffff5bb1 */ (T_MASK ^ 0x0000a44e) +#define T12 /* 0x895cd7be */ (T_MASK ^ 0x76a32841) +#define T13 0x6b901122 +#define T14 /* 0xfd987193 */ (T_MASK ^ 0x02678e6c) +#define T15 /* 0xa679438e */ (T_MASK ^ 0x5986bc71) +#define T16 0x49b40821 +#define T17 /* 0xf61e2562 */ (T_MASK ^ 0x09e1da9d) +#define T18 /* 0xc040b340 */ (T_MASK ^ 0x3fbf4cbf) +#define T19 0x265e5a51 +#define T20 /* 0xe9b6c7aa */ (T_MASK ^ 0x16493855) +#define T21 /* 0xd62f105d */ (T_MASK ^ 0x29d0efa2) +#define T22 0x02441453 +#define T23 /* 0xd8a1e681 */ (T_MASK ^ 0x275e197e) +#define T24 /* 0xe7d3fbc8 */ (T_MASK ^ 0x182c0437) +#define T25 0x21e1cde6 +#define T26 /* 0xc33707d6 */ (T_MASK ^ 0x3cc8f829) +#define T27 /* 0xf4d50d87 */ (T_MASK ^ 0x0b2af278) +#define T28 0x455a14ed +#define T29 /* 0xa9e3e905 */ (T_MASK ^ 0x561c16fa) +#define T30 /* 0xfcefa3f8 */ (T_MASK ^ 0x03105c07) +#define T31 0x676f02d9 +#define T32 /* 0x8d2a4c8a */ (T_MASK ^ 0x72d5b375) +#define T33 /* 0xfffa3942 */ (T_MASK ^ 0x0005c6bd) +#define T34 /* 0x8771f681 */ (T_MASK ^ 0x788e097e) +#define T35 0x6d9d6122 +#define T36 /* 0xfde5380c */ (T_MASK ^ 0x021ac7f3) +#define T37 /* 0xa4beea44 */ (T_MASK ^ 0x5b4115bb) +#define T38 0x4bdecfa9 +#define T39 /* 0xf6bb4b60 */ (T_MASK ^ 0x0944b49f) +#define T40 /* 0xbebfbc70 */ (T_MASK ^ 0x4140438f) +#define T41 0x289b7ec6 +#define T42 /* 0xeaa127fa */ (T_MASK ^ 0x155ed805) +#define T43 /* 0xd4ef3085 */ (T_MASK ^ 0x2b10cf7a) +#define T44 0x04881d05 +#define T45 /* 0xd9d4d039 */ (T_MASK ^ 0x262b2fc6) +#define T46 /* 0xe6db99e5 */ (T_MASK ^ 0x1924661a) +#define T47 0x1fa27cf8 +#define T48 /* 0xc4ac5665 */ (T_MASK ^ 0x3b53a99a) +#define T49 /* 0xf4292244 */ (T_MASK ^ 0x0bd6ddbb) +#define T50 0x432aff97 +#define T51 /* 0xab9423a7 */ (T_MASK ^ 0x546bdc58) +#define T52 /* 0xfc93a039 */ (T_MASK ^ 0x036c5fc6) +#define T53 0x655b59c3 +#define T54 /* 0x8f0ccc92 */ (T_MASK ^ 0x70f3336d) +#define T55 /* 0xffeff47d */ (T_MASK ^ 0x00100b82) +#define T56 /* 0x85845dd1 */ (T_MASK ^ 0x7a7ba22e) +#define T57 0x6fa87e4f +#define T58 /* 0xfe2ce6e0 */ (T_MASK ^ 0x01d3191f) +#define T59 /* 0xa3014314 */ (T_MASK ^ 0x5cfebceb) +#define T60 0x4e0811a1 +#define T61 /* 0xf7537e82 */ (T_MASK ^ 0x08ac817d) +#define T62 /* 0xbd3af235 */ (T_MASK ^ 0x42c50dca) +#define T63 0x2ad7d2bb +#define T64 /* 0xeb86d391 */ (T_MASK ^ 0x14792c6e) + + +static void +md5_process(md5_state_t *pms, const uint8_t *data /*[64]*/) +{ + uint32_t + a = pms->abcd[0], b = pms->abcd[1], + c = pms->abcd[2], d = pms->abcd[3]; + uint32_t t; +#if BYTE_ORDER > 0 + /* Define storage only for big-endian CPUs. */ + uint32_t X[16]; +#else + /* Define storage for little-endian or both types of CPUs. */ + uint32_t xbuf[16]; + const uint32_t *X; +#endif + + { +#if BYTE_ORDER == 0 + /* + * Determine dynamically whether this is a big-endian or + * little-endian machine, since we can use a more efficient + * algorithm on the latter. + */ + static const int w = 1; + + if (*((const uint8_t *)&w)) /* dynamic little-endian */ +#endif +#if BYTE_ORDER <= 0 /* little-endian */ + { + /* + * On little-endian machines, we can process properly aligned + * data without copying it. + */ + if (!((data - (const uint8_t *)0) & 3)) { + /* data are properly aligned */ + X = (const uint32_t *)data; + } else { + /* not aligned */ + memcpy(xbuf, data, 64); + X = xbuf; + } + } +#endif +#if BYTE_ORDER == 0 + else /* dynamic big-endian */ +#endif +#if BYTE_ORDER >= 0 /* big-endian */ + { + /* + * On big-endian machines, we must arrange the bytes in the + * right order. + */ + const uint8_t *xp = data; + int i; + +# if BYTE_ORDER == 0 + X = xbuf; /* (dynamic only) */ +# else +# define xbuf X /* (static only) */ +# endif + for (i = 0; i < 16; ++i, xp += 4) + xbuf[i] = xp[0] + (xp[1] << 8) + (xp[2] << 16) + (xp[3] << 24); + } +#endif + } + +#define ROTATE_LEFT(x, n) (((x) << (n)) | ((x) >> (32 - (n)))) + + /* Round 1. */ + /* Let [abcd k s i] denote the operation + a = b + ((a + F(b,c,d) + X[k] + T[i]) <<< s). */ +#define F(x, y, z) (((x) & (y)) | (~(x) & (z))) +#define SET(a, b, c, d, k, s, Ti)\ + t = a + F(b,c,d) + X[k] + Ti;\ + a = ROTATE_LEFT(t, s) + b + /* Do the following 16 operations. */ + SET(a, b, c, d, 0, 7, T1); + SET(d, a, b, c, 1, 12, T2); + SET(c, d, a, b, 2, 17, T3); + SET(b, c, d, a, 3, 22, T4); + SET(a, b, c, d, 4, 7, T5); + SET(d, a, b, c, 5, 12, T6); + SET(c, d, a, b, 6, 17, T7); + SET(b, c, d, a, 7, 22, T8); + SET(a, b, c, d, 8, 7, T9); + SET(d, a, b, c, 9, 12, T10); + SET(c, d, a, b, 10, 17, T11); + SET(b, c, d, a, 11, 22, T12); + SET(a, b, c, d, 12, 7, T13); + SET(d, a, b, c, 13, 12, T14); + SET(c, d, a, b, 14, 17, T15); + SET(b, c, d, a, 15, 22, T16); +#undef SET + + /* Round 2. */ + /* Let [abcd k s i] denote the operation + a = b + ((a + G(b,c,d) + X[k] + T[i]) <<< s). */ +#define G(x, y, z) (((x) & (z)) | ((y) & ~(z))) +#define SET(a, b, c, d, k, s, Ti)\ + t = a + G(b,c,d) + X[k] + Ti;\ + a = ROTATE_LEFT(t, s) + b + /* Do the following 16 operations. */ + SET(a, b, c, d, 1, 5, T17); + SET(d, a, b, c, 6, 9, T18); + SET(c, d, a, b, 11, 14, T19); + SET(b, c, d, a, 0, 20, T20); + SET(a, b, c, d, 5, 5, T21); + SET(d, a, b, c, 10, 9, T22); + SET(c, d, a, b, 15, 14, T23); + SET(b, c, d, a, 4, 20, T24); + SET(a, b, c, d, 9, 5, T25); + SET(d, a, b, c, 14, 9, T26); + SET(c, d, a, b, 3, 14, T27); + SET(b, c, d, a, 8, 20, T28); + SET(a, b, c, d, 13, 5, T29); + SET(d, a, b, c, 2, 9, T30); + SET(c, d, a, b, 7, 14, T31); + SET(b, c, d, a, 12, 20, T32); +#undef SET + + /* Round 3. */ + /* Let [abcd k s t] denote the operation + a = b + ((a + H(b,c,d) + X[k] + T[i]) <<< s). */ +#define H(x, y, z) ((x) ^ (y) ^ (z)) +#define SET(a, b, c, d, k, s, Ti)\ + t = a + H(b,c,d) + X[k] + Ti;\ + a = ROTATE_LEFT(t, s) + b + /* Do the following 16 operations. */ + SET(a, b, c, d, 5, 4, T33); + SET(d, a, b, c, 8, 11, T34); + SET(c, d, a, b, 11, 16, T35); + SET(b, c, d, a, 14, 23, T36); + SET(a, b, c, d, 1, 4, T37); + SET(d, a, b, c, 4, 11, T38); + SET(c, d, a, b, 7, 16, T39); + SET(b, c, d, a, 10, 23, T40); + SET(a, b, c, d, 13, 4, T41); + SET(d, a, b, c, 0, 11, T42); + SET(c, d, a, b, 3, 16, T43); + SET(b, c, d, a, 6, 23, T44); + SET(a, b, c, d, 9, 4, T45); + SET(d, a, b, c, 12, 11, T46); + SET(c, d, a, b, 15, 16, T47); + SET(b, c, d, a, 2, 23, T48); +#undef SET + + /* Round 4. */ + /* Let [abcd k s t] denote the operation + a = b + ((a + I(b,c,d) + X[k] + T[i]) <<< s). */ +#define I(x, y, z) ((y) ^ ((x) | ~(z))) +#define SET(a, b, c, d, k, s, Ti)\ + t = a + I(b,c,d) + X[k] + Ti;\ + a = ROTATE_LEFT(t, s) + b + /* Do the following 16 operations. */ + SET(a, b, c, d, 0, 6, T49); + SET(d, a, b, c, 7, 10, T50); + SET(c, d, a, b, 14, 15, T51); + SET(b, c, d, a, 5, 21, T52); + SET(a, b, c, d, 12, 6, T53); + SET(d, a, b, c, 3, 10, T54); + SET(c, d, a, b, 10, 15, T55); + SET(b, c, d, a, 1, 21, T56); + SET(a, b, c, d, 8, 6, T57); + SET(d, a, b, c, 15, 10, T58); + SET(c, d, a, b, 6, 15, T59); + SET(b, c, d, a, 13, 21, T60); + SET(a, b, c, d, 4, 6, T61); + SET(d, a, b, c, 11, 10, T62); + SET(c, d, a, b, 2, 15, T63); + SET(b, c, d, a, 9, 21, T64); +#undef SET + + /* Then perform the following additions. (That is increment each + of the four registers by the value it had before this block + was started.) */ + pms->abcd[0] += a; + pms->abcd[1] += b; + pms->abcd[2] += c; + pms->abcd[3] += d; +} + +void +pplib_md5_init(md5_state_t *pms) +{ + pms->count[0] = pms->count[1] = 0; + pms->abcd[0] = 0x67452301; + pms->abcd[1] = /*0xefcdab89*/ T_MASK ^ 0x10325476; + pms->abcd[2] = /*0x98badcfe*/ T_MASK ^ 0x67452301; + pms->abcd[3] = 0x10325476; +} + +void +md5_add(md5_state_t *pms, const void *input, size_t size) +{ + const uint8_t *p = (const uint8_t *)input; + int nbytes = (int)size; // PJ + int left = nbytes; + int offset = (pms->count[0] >> 3) & 63; + uint32_t nbits = (uint32_t)(nbytes << 3); + + if (nbytes <= 0) + return; + + /* Update the message length. */ + pms->count[1] += nbytes >> 29; + pms->count[0] += nbits; + if (pms->count[0] < nbits) + pms->count[1]++; + + /* Process an initial partial block. */ + if (offset) { + int copy = (offset + nbytes > 64 ? 64 - offset : nbytes); + + memcpy(pms->buf + offset, p, copy); + if (offset + copy < 64) + return; + p += copy; + left -= copy; + md5_process(pms, pms->buf); + } + + /* Process full blocks. */ + for (; left >= 64; p += 64, left -= 64) + md5_process(pms, p); + + /* Process a final partial block. */ + if (left) + memcpy(pms->buf, p, left); +} + +void +md5_put(md5_state_t *pms, uint8_t digest[16]) +{ + static const uint8_t pad[64] = { + 0x80, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, + 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, + 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, + 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0 + }; + uint8_t data[8]; + int i; + + /* Save the length before padding. */ + for (i = 0; i < 8; ++i) + data[i] = (uint8_t)(pms->count[i >> 2] >> ((i & 3) << 3)); + /* Pad to 56 bytes mod 64. */ + md5_add(pms, pad, ((55 - (pms->count[0] >> 3)) & 63) + 1); + /* Append the length. */ + md5_add(pms, data, 8); + for (i = 0; i < 16; ++i) + digest[i] = (uint8_t)(pms->abcd[i >> 2] >> ((i & 3) << 3)); +} + +/* end of md5.h */ + +static md5_state_t state; + +void pplib_md5 (const void *input, size_t length, uint8_t output[16]) +{ + pplib_md5_init(&state); + md5_add(&state, input, length); + md5_put(&state, output); +} + +void md5init (void) +{ + pplib_md5_init(&state); +} + +void md5add (const void *input, size_t length) +{ + md5_add(&state, input, length); +} + +void md5put (uint8_t output[16]) +{ + md5_put(&state, output); +} diff --git a/Build/source/texk/web2c/luatexdir/luapplib/util/utilmd5.c.orig b/Build/source/texk/web2c/luatexdir/luapplib/util/utilmd5.c.orig new file mode 100644 index 00000000000..06fb03c5bce --- /dev/null +++ b/Build/source/texk/web2c/luatexdir/luapplib/util/utilmd5.c.orig @@ -0,0 +1,413 @@ +/* md5 implementation by Peter Deutsch (ghost@aladdin.com) with some convenience shorthands */ + +/* begin of md5.c */ + +/* + Copyright (C) 1999, 2000, 2002 Aladdin Enterprises. All rights reserved. + + This software is provided 'as-is', without any express or implied + warranty. In no event will the authors be held liable for any damages + arising from the use of this software. + + Permission is granted to anyone to use this software for any purpose, + including commercial applications, and to alter it and redistribute it + freely, subject to the following restrictions: + + 1. The origin of this software must not be misrepresented; you must not + claim that you wrote the original software. If you use this software + in a product, an acknowledgment in the product documentation would be + appreciated but is not required. + 2. Altered source versions must be plainly marked as such, and must not be + misrepresented as being the original software. + 3. This notice may not be removed or altered from any source distribution. + + L. Peter Deutsch + ghost@aladdin.com + + */ +/* $Id: md5.c,v 1.6 2002/04/13 19:20:28 lpd Exp $ */ +/* + Independent implementation of MD5 (RFC 1321). + + This code implements the MD5 Algorithm defined in RFC 1321, whose + text is available at + http://www.ietf.org/rfc/rfc1321.txt + The code is derived from the text of the RFC, including the test suite + (section A.5) but excluding the rest of Appendix A. It does not include + any code or documentation that is identified in the RFC as being + copyrighted. + + The original and principal author of md5.c is L. Peter Deutsch + <ghost@aladdin.com>. Other authors are noted in the change history + that follows (in reverse chronological order): + + 2002-04-13 lpd Clarified derivation from RFC 1321; now handles byte order + either statically or dynamically; added missing #include <string.h> + in library. + 2002-03-11 lpd Corrected argument list for main(), and added int return + type, in test program and T value program. + 2002-02-21 lpd Added missing #include <stdio.h> in test program. + 2000-07-03 lpd Patched to eliminate warnings about "constant is + unsigned in ANSI C, signed in traditional"; made test program + self-checking. + 1999-11-04 lpd Edited comments slightly for automatic TOC extraction. + 1999-10-18 lpd Fixed typo in header comment (ansi2knr rather than md5). + 1999-05-03 lpd Original version. + */ + +//#include "md5.h" +#include "utilmd5.h" +#include <string.h> + +#undef BYTE_ORDER /* 1 = big-endian, -1 = little-endian, 0 = unknown */ +#ifdef ARCH_IS_BIG_ENDIAN +# define BYTE_ORDER (ARCH_IS_BIG_ENDIAN ? 1 : -1) +#else +# define BYTE_ORDER 0 +#endif + +#define T_MASK ((uint32_t)~0) +#define T1 /* 0xd76aa478 */ (T_MASK ^ 0x28955b87) +#define T2 /* 0xe8c7b756 */ (T_MASK ^ 0x173848a9) +#define T3 0x242070db +#define T4 /* 0xc1bdceee */ (T_MASK ^ 0x3e423111) +#define T5 /* 0xf57c0faf */ (T_MASK ^ 0x0a83f050) +#define T6 0x4787c62a +#define T7 /* 0xa8304613 */ (T_MASK ^ 0x57cfb9ec) +#define T8 /* 0xfd469501 */ (T_MASK ^ 0x02b96afe) +#define T9 0x698098d8 +#define T10 /* 0x8b44f7af */ (T_MASK ^ 0x74bb0850) +#define T11 /* 0xffff5bb1 */ (T_MASK ^ 0x0000a44e) +#define T12 /* 0x895cd7be */ (T_MASK ^ 0x76a32841) +#define T13 0x6b901122 +#define T14 /* 0xfd987193 */ (T_MASK ^ 0x02678e6c) +#define T15 /* 0xa679438e */ (T_MASK ^ 0x5986bc71) +#define T16 0x49b40821 +#define T17 /* 0xf61e2562 */ (T_MASK ^ 0x09e1da9d) +#define T18 /* 0xc040b340 */ (T_MASK ^ 0x3fbf4cbf) +#define T19 0x265e5a51 +#define T20 /* 0xe9b6c7aa */ (T_MASK ^ 0x16493855) +#define T21 /* 0xd62f105d */ (T_MASK ^ 0x29d0efa2) +#define T22 0x02441453 +#define T23 /* 0xd8a1e681 */ (T_MASK ^ 0x275e197e) +#define T24 /* 0xe7d3fbc8 */ (T_MASK ^ 0x182c0437) +#define T25 0x21e1cde6 +#define T26 /* 0xc33707d6 */ (T_MASK ^ 0x3cc8f829) +#define T27 /* 0xf4d50d87 */ (T_MASK ^ 0x0b2af278) +#define T28 0x455a14ed +#define T29 /* 0xa9e3e905 */ (T_MASK ^ 0x561c16fa) +#define T30 /* 0xfcefa3f8 */ (T_MASK ^ 0x03105c07) +#define T31 0x676f02d9 +#define T32 /* 0x8d2a4c8a */ (T_MASK ^ 0x72d5b375) +#define T33 /* 0xfffa3942 */ (T_MASK ^ 0x0005c6bd) +#define T34 /* 0x8771f681 */ (T_MASK ^ 0x788e097e) +#define T35 0x6d9d6122 +#define T36 /* 0xfde5380c */ (T_MASK ^ 0x021ac7f3) +#define T37 /* 0xa4beea44 */ (T_MASK ^ 0x5b4115bb) +#define T38 0x4bdecfa9 +#define T39 /* 0xf6bb4b60 */ (T_MASK ^ 0x0944b49f) +#define T40 /* 0xbebfbc70 */ (T_MASK ^ 0x4140438f) +#define T41 0x289b7ec6 +#define T42 /* 0xeaa127fa */ (T_MASK ^ 0x155ed805) +#define T43 /* 0xd4ef3085 */ (T_MASK ^ 0x2b10cf7a) +#define T44 0x04881d05 +#define T45 /* 0xd9d4d039 */ (T_MASK ^ 0x262b2fc6) +#define T46 /* 0xe6db99e5 */ (T_MASK ^ 0x1924661a) +#define T47 0x1fa27cf8 +#define T48 /* 0xc4ac5665 */ (T_MASK ^ 0x3b53a99a) +#define T49 /* 0xf4292244 */ (T_MASK ^ 0x0bd6ddbb) +#define T50 0x432aff97 +#define T51 /* 0xab9423a7 */ (T_MASK ^ 0x546bdc58) +#define T52 /* 0xfc93a039 */ (T_MASK ^ 0x036c5fc6) +#define T53 0x655b59c3 +#define T54 /* 0x8f0ccc92 */ (T_MASK ^ 0x70f3336d) +#define T55 /* 0xffeff47d */ (T_MASK ^ 0x00100b82) +#define T56 /* 0x85845dd1 */ (T_MASK ^ 0x7a7ba22e) +#define T57 0x6fa87e4f +#define T58 /* 0xfe2ce6e0 */ (T_MASK ^ 0x01d3191f) +#define T59 /* 0xa3014314 */ (T_MASK ^ 0x5cfebceb) +#define T60 0x4e0811a1 +#define T61 /* 0xf7537e82 */ (T_MASK ^ 0x08ac817d) +#define T62 /* 0xbd3af235 */ (T_MASK ^ 0x42c50dca) +#define T63 0x2ad7d2bb +#define T64 /* 0xeb86d391 */ (T_MASK ^ 0x14792c6e) + + +static void +md5_process(md5_state_t *pms, const uint8_t *data /*[64]*/) +{ + uint32_t + a = pms->abcd[0], b = pms->abcd[1], + c = pms->abcd[2], d = pms->abcd[3]; + uint32_t t; +#if BYTE_ORDER > 0 + /* Define storage only for big-endian CPUs. */ + uint32_t X[16]; +#else + /* Define storage for little-endian or both types of CPUs. */ + uint32_t xbuf[16]; + const uint32_t *X; +#endif + + { +#if BYTE_ORDER == 0 + /* + * Determine dynamically whether this is a big-endian or + * little-endian machine, since we can use a more efficient + * algorithm on the latter. + */ + static const int w = 1; + + if (*((const uint8_t *)&w)) /* dynamic little-endian */ +#endif +#if BYTE_ORDER <= 0 /* little-endian */ + { + /* + * On little-endian machines, we can process properly aligned + * data without copying it. + */ + if (!((data - (const uint8_t *)0) & 3)) { + /* data are properly aligned */ + X = (const uint32_t *)data; + } else { + /* not aligned */ + memcpy(xbuf, data, 64); + X = xbuf; + } + } +#endif +#if BYTE_ORDER == 0 + else /* dynamic big-endian */ +#endif +#if BYTE_ORDER >= 0 /* big-endian */ + { + /* + * On big-endian machines, we must arrange the bytes in the + * right order. + */ + const uint8_t *xp = data; + int i; + +# if BYTE_ORDER == 0 + X = xbuf; /* (dynamic only) */ +# else +# define xbuf X /* (static only) */ +# endif + for (i = 0; i < 16; ++i, xp += 4) + xbuf[i] = xp[0] + (xp[1] << 8) + (xp[2] << 16) + (xp[3] << 24); + } +#endif + } + +#define ROTATE_LEFT(x, n) (((x) << (n)) | ((x) >> (32 - (n)))) + + /* Round 1. */ + /* Let [abcd k s i] denote the operation + a = b + ((a + F(b,c,d) + X[k] + T[i]) <<< s). */ +#define F(x, y, z) (((x) & (y)) | (~(x) & (z))) +#define SET(a, b, c, d, k, s, Ti)\ + t = a + F(b,c,d) + X[k] + Ti;\ + a = ROTATE_LEFT(t, s) + b + /* Do the following 16 operations. */ + SET(a, b, c, d, 0, 7, T1); + SET(d, a, b, c, 1, 12, T2); + SET(c, d, a, b, 2, 17, T3); + SET(b, c, d, a, 3, 22, T4); + SET(a, b, c, d, 4, 7, T5); + SET(d, a, b, c, 5, 12, T6); + SET(c, d, a, b, 6, 17, T7); + SET(b, c, d, a, 7, 22, T8); + SET(a, b, c, d, 8, 7, T9); + SET(d, a, b, c, 9, 12, T10); + SET(c, d, a, b, 10, 17, T11); + SET(b, c, d, a, 11, 22, T12); + SET(a, b, c, d, 12, 7, T13); + SET(d, a, b, c, 13, 12, T14); + SET(c, d, a, b, 14, 17, T15); + SET(b, c, d, a, 15, 22, T16); +#undef SET + + /* Round 2. */ + /* Let [abcd k s i] denote the operation + a = b + ((a + G(b,c,d) + X[k] + T[i]) <<< s). */ +#define G(x, y, z) (((x) & (z)) | ((y) & ~(z))) +#define SET(a, b, c, d, k, s, Ti)\ + t = a + G(b,c,d) + X[k] + Ti;\ + a = ROTATE_LEFT(t, s) + b + /* Do the following 16 operations. */ + SET(a, b, c, d, 1, 5, T17); + SET(d, a, b, c, 6, 9, T18); + SET(c, d, a, b, 11, 14, T19); + SET(b, c, d, a, 0, 20, T20); + SET(a, b, c, d, 5, 5, T21); + SET(d, a, b, c, 10, 9, T22); + SET(c, d, a, b, 15, 14, T23); + SET(b, c, d, a, 4, 20, T24); + SET(a, b, c, d, 9, 5, T25); + SET(d, a, b, c, 14, 9, T26); + SET(c, d, a, b, 3, 14, T27); + SET(b, c, d, a, 8, 20, T28); + SET(a, b, c, d, 13, 5, T29); + SET(d, a, b, c, 2, 9, T30); + SET(c, d, a, b, 7, 14, T31); + SET(b, c, d, a, 12, 20, T32); +#undef SET + + /* Round 3. */ + /* Let [abcd k s t] denote the operation + a = b + ((a + H(b,c,d) + X[k] + T[i]) <<< s). */ +#define H(x, y, z) ((x) ^ (y) ^ (z)) +#define SET(a, b, c, d, k, s, Ti)\ + t = a + H(b,c,d) + X[k] + Ti;\ + a = ROTATE_LEFT(t, s) + b + /* Do the following 16 operations. */ + SET(a, b, c, d, 5, 4, T33); + SET(d, a, b, c, 8, 11, T34); + SET(c, d, a, b, 11, 16, T35); + SET(b, c, d, a, 14, 23, T36); + SET(a, b, c, d, 1, 4, T37); + SET(d, a, b, c, 4, 11, T38); + SET(c, d, a, b, 7, 16, T39); + SET(b, c, d, a, 10, 23, T40); + SET(a, b, c, d, 13, 4, T41); + SET(d, a, b, c, 0, 11, T42); + SET(c, d, a, b, 3, 16, T43); + SET(b, c, d, a, 6, 23, T44); + SET(a, b, c, d, 9, 4, T45); + SET(d, a, b, c, 12, 11, T46); + SET(c, d, a, b, 15, 16, T47); + SET(b, c, d, a, 2, 23, T48); +#undef SET + + /* Round 4. */ + /* Let [abcd k s t] denote the operation + a = b + ((a + I(b,c,d) + X[k] + T[i]) <<< s). */ +#define I(x, y, z) ((y) ^ ((x) | ~(z))) +#define SET(a, b, c, d, k, s, Ti)\ + t = a + I(b,c,d) + X[k] + Ti;\ + a = ROTATE_LEFT(t, s) + b + /* Do the following 16 operations. */ + SET(a, b, c, d, 0, 6, T49); + SET(d, a, b, c, 7, 10, T50); + SET(c, d, a, b, 14, 15, T51); + SET(b, c, d, a, 5, 21, T52); + SET(a, b, c, d, 12, 6, T53); + SET(d, a, b, c, 3, 10, T54); + SET(c, d, a, b, 10, 15, T55); + SET(b, c, d, a, 1, 21, T56); + SET(a, b, c, d, 8, 6, T57); + SET(d, a, b, c, 15, 10, T58); + SET(c, d, a, b, 6, 15, T59); + SET(b, c, d, a, 13, 21, T60); + SET(a, b, c, d, 4, 6, T61); + SET(d, a, b, c, 11, 10, T62); + SET(c, d, a, b, 2, 15, T63); + SET(b, c, d, a, 9, 21, T64); +#undef SET + + /* Then perform the following additions. (That is increment each + of the four registers by the value it had before this block + was started.) */ + pms->abcd[0] += a; + pms->abcd[1] += b; + pms->abcd[2] += c; + pms->abcd[3] += d; +} + +void +md5_init(md5_state_t *pms) +{ + pms->count[0] = pms->count[1] = 0; + pms->abcd[0] = 0x67452301; + pms->abcd[1] = /*0xefcdab89*/ T_MASK ^ 0x10325476; + pms->abcd[2] = /*0x98badcfe*/ T_MASK ^ 0x67452301; + pms->abcd[3] = 0x10325476; +} + +void +md5_add(md5_state_t *pms, const void *input, size_t size) +{ + const uint8_t *p = (const uint8_t *)input; + int nbytes = (int)size; // PJ + int left = nbytes; + int offset = (pms->count[0] >> 3) & 63; + uint32_t nbits = (uint32_t)(nbytes << 3); + + if (nbytes <= 0) + return; + + /* Update the message length. */ + pms->count[1] += nbytes >> 29; + pms->count[0] += nbits; + if (pms->count[0] < nbits) + pms->count[1]++; + + /* Process an initial partial block. */ + if (offset) { + int copy = (offset + nbytes > 64 ? 64 - offset : nbytes); + + memcpy(pms->buf + offset, p, copy); + if (offset + copy < 64) + return; + p += copy; + left -= copy; + md5_process(pms, pms->buf); + } + + /* Process full blocks. */ + for (; left >= 64; p += 64, left -= 64) + md5_process(pms, p); + + /* Process a final partial block. */ + if (left) + memcpy(pms->buf, p, left); +} + +void +md5_put(md5_state_t *pms, uint8_t digest[16]) +{ + static const uint8_t pad[64] = { + 0x80, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, + 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, + 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, + 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0 + }; + uint8_t data[8]; + int i; + + /* Save the length before padding. */ + for (i = 0; i < 8; ++i) + data[i] = (uint8_t)(pms->count[i >> 2] >> ((i & 3) << 3)); + /* Pad to 56 bytes mod 64. */ + md5_add(pms, pad, ((55 - (pms->count[0] >> 3)) & 63) + 1); + /* Append the length. */ + md5_add(pms, data, 8); + for (i = 0; i < 16; ++i) + digest[i] = (uint8_t)(pms->abcd[i >> 2] >> ((i & 3) << 3)); +} + +/* end of md5.h */ + +static md5_state_t state; + +void md5 (const void *input, size_t length, uint8_t output[16]) +{ + md5_init(&state); + md5_add(&state, input, length); + md5_put(&state, output); +} + +void md5init (void) +{ + md5_init(&state); +} + +void md5add (const void *input, size_t length) +{ + md5_add(&state, input, length); +} + +void md5put (uint8_t output[16]) +{ + md5_put(&state, output); +} diff --git a/Build/source/texk/web2c/luatexdir/luapplib/util/utilmd5.h b/Build/source/texk/web2c/luatexdir/luapplib/util/utilmd5.h new file mode 100644 index 00000000000..2fb4bc58de7 --- /dev/null +++ b/Build/source/texk/web2c/luatexdir/luapplib/util/utilmd5.h @@ -0,0 +1,115 @@ +#ifndef UTIL_MD5_H +#define UTIL_MD5_H + +#include <stdint.h> // for uintX8_t +#include <stddef.h> // for size_t +#include "utildecl.h" + +/* begin of md5.h */ + +/* + Copyright (C) 1999, 2002 Aladdin Enterprises. All rights reserved. + + This software is provided 'as-is', without any express or implied + warranty. In no event will the authors be held liable for any damages + arising from the use of this software. + + Permission is granted to anyone to use this software for any purpose, + including commercial applications, and to alter it and redistribute it + freely, subject to the following restrictions: + + 1. The origin of this software must not be misrepresented; you must not + claim that you wrote the original software. If you use this software + in a product, an acknowledgment in the product documentation would be + appreciated but is not required. + 2. Altered source versions must be plainly marked as such, and must not be + misrepresented as being the original software. + 3. This notice may not be removed or altered from any source distribution. + + L. Peter Deutsch + ghost@aladdin.com + + */ +/* $Id: md5.h,v 1.4 2002/04/13 19:20:28 lpd Exp $ */ +/* + Independent implementation of MD5 (RFC 1321). + + This code implements the MD5 Algorithm defined in RFC 1321, whose + text is available at + http://www.ietf.org/rfc/rfc1321.txt + The code is derived from the text of the RFC, including the test suite + (section A.5) but excluding the rest of Appendix A. It does not include + any code or documentation that is identified in the RFC as being + copyrighted. + + The original and principal author of md5.h is L. Peter Deutsch + <ghost@aladdin.com>. Other authors are noted in the change history + that follows (in reverse chronological order): + + 2002-04-13 lpd Removed support for non-ANSI compilers; removed + references to Ghostscript; clarified derivation from RFC 1321; + now handles byte order either statically or dynamically. + 1999-11-04 lpd Edited comments slightly for automatic TOC extraction. + 1999-10-18 lpd Fixed typo in header comment (ansi2knr rather than md5); + added conditionalization for C++ compilation from Martin + Purschke <purschke@bnl.gov>. + 1999-05-03 lpd Original version. + */ + +//#ifndef md5_INCLUDED +//# define md5_INCLUDED + +/* + * This package supports both compile-time and run-time determination of CPU + * byte order. If ARCH_IS_BIG_ENDIAN is defined as 0, the code will be + * compiled to run only on little-endian CPUs; if ARCH_IS_BIG_ENDIAN is + * defined as non-zero, the code will be compiled to run only on big-endian + * CPUs; if ARCH_IS_BIG_ENDIAN is not defined, the code will be compiled to + * run on either big- or little-endian CPUs, but will run slightly less + * efficiently on either one than if ARCH_IS_BIG_ENDIAN is defined. + */ + +// PJ: replaced with uint8_t and uint32_t +//typedef unsigned char md5_byte_t; /* 8-bit byte */ +//typedef unsigned int md5_word_t; /* 32-bit word */ + +/* Define the state of the MD5 Algorithm. */ +typedef struct md5_state_s { + uint32_t count[2]; /* message length in bits, lsw first */ + uint32_t abcd[4]; /* digest buffer */ + uint8_t buf[64]; /* accumulate block */ +} md5_state_t; + +#ifdef __cplusplus +extern "C" +{ +#endif + +/* Initialize the algorithm. */ +UTILAPI void pplib_md5_init (md5_state_t *pms); + +/* Append a string to the message. */ +//void md5_add(md5_state_t *pms, const uint8_t *data, int nbytes); // PJ +UTILAPI void md5_add (md5_state_t *pms, const void *input, size_t size); + +/* Finish the message and return the digest. */ +//void md5_finish(md5_state_t *pms, uint8_t digest[16]); // PJ +UTILAPI void md5_put (md5_state_t *pms, uint8_t digest[16]); + +#ifdef __cplusplus +} /* end extern "C" */ +#endif + +//#endif /* md5_INCLUDED */ + +/* end of md5.h */ + +#define md5_state md5_state_t + +UTILAPI void md5init (void); +UTILAPI void md5add (const void *input, size_t length); +UTILAPI void md5put (uint8_t output[16]); + +UTILAPI void pplib_md5 (const void *input, size_t length, uint8_t output[16]); + +#endif
\ No newline at end of file diff --git a/Build/source/texk/web2c/luatexdir/luapplib/util/utilmd5.h.orig b/Build/source/texk/web2c/luatexdir/luapplib/util/utilmd5.h.orig new file mode 100644 index 00000000000..c4eecff4f75 --- /dev/null +++ b/Build/source/texk/web2c/luatexdir/luapplib/util/utilmd5.h.orig @@ -0,0 +1,115 @@ +#ifndef UTIL_MD5_H +#define UTIL_MD5_H + +#include <stdint.h> // for uintX8_t +#include <stddef.h> // for size_t +#include "utildecl.h" + +/* begin of md5.h */ + +/* + Copyright (C) 1999, 2002 Aladdin Enterprises. All rights reserved. + + This software is provided 'as-is', without any express or implied + warranty. In no event will the authors be held liable for any damages + arising from the use of this software. + + Permission is granted to anyone to use this software for any purpose, + including commercial applications, and to alter it and redistribute it + freely, subject to the following restrictions: + + 1. The origin of this software must not be misrepresented; you must not + claim that you wrote the original software. If you use this software + in a product, an acknowledgment in the product documentation would be + appreciated but is not required. + 2. Altered source versions must be plainly marked as such, and must not be + misrepresented as being the original software. + 3. This notice may not be removed or altered from any source distribution. + + L. Peter Deutsch + ghost@aladdin.com + + */ +/* $Id: md5.h,v 1.4 2002/04/13 19:20:28 lpd Exp $ */ +/* + Independent implementation of MD5 (RFC 1321). + + This code implements the MD5 Algorithm defined in RFC 1321, whose + text is available at + http://www.ietf.org/rfc/rfc1321.txt + The code is derived from the text of the RFC, including the test suite + (section A.5) but excluding the rest of Appendix A. It does not include + any code or documentation that is identified in the RFC as being + copyrighted. + + The original and principal author of md5.h is L. Peter Deutsch + <ghost@aladdin.com>. Other authors are noted in the change history + that follows (in reverse chronological order): + + 2002-04-13 lpd Removed support for non-ANSI compilers; removed + references to Ghostscript; clarified derivation from RFC 1321; + now handles byte order either statically or dynamically. + 1999-11-04 lpd Edited comments slightly for automatic TOC extraction. + 1999-10-18 lpd Fixed typo in header comment (ansi2knr rather than md5); + added conditionalization for C++ compilation from Martin + Purschke <purschke@bnl.gov>. + 1999-05-03 lpd Original version. + */ + +//#ifndef md5_INCLUDED +//# define md5_INCLUDED + +/* + * This package supports both compile-time and run-time determination of CPU + * byte order. If ARCH_IS_BIG_ENDIAN is defined as 0, the code will be + * compiled to run only on little-endian CPUs; if ARCH_IS_BIG_ENDIAN is + * defined as non-zero, the code will be compiled to run only on big-endian + * CPUs; if ARCH_IS_BIG_ENDIAN is not defined, the code will be compiled to + * run on either big- or little-endian CPUs, but will run slightly less + * efficiently on either one than if ARCH_IS_BIG_ENDIAN is defined. + */ + +// PJ: replaced with uint8_t and uint32_t +//typedef unsigned char md5_byte_t; /* 8-bit byte */ +//typedef unsigned int md5_word_t; /* 32-bit word */ + +/* Define the state of the MD5 Algorithm. */ +typedef struct md5_state_s { + uint32_t count[2]; /* message length in bits, lsw first */ + uint32_t abcd[4]; /* digest buffer */ + uint8_t buf[64]; /* accumulate block */ +} md5_state_t; + +#ifdef __cplusplus +extern "C" +{ +#endif + +/* Initialize the algorithm. */ +UTILAPI void md5_init (md5_state_t *pms); + +/* Append a string to the message. */ +//void md5_add(md5_state_t *pms, const uint8_t *data, int nbytes); // PJ +UTILAPI void md5_add (md5_state_t *pms, const void *input, size_t size); + +/* Finish the message and return the digest. */ +//void md5_finish(md5_state_t *pms, uint8_t digest[16]); // PJ +UTILAPI void md5_put (md5_state_t *pms, uint8_t digest[16]); + +#ifdef __cplusplus +} /* end extern "C" */ +#endif + +//#endif /* md5_INCLUDED */ + +/* end of md5.h */ + +#define md5_state md5_state_t + +UTILAPI void md5init (void); +UTILAPI void md5add (const void *input, size_t length); +UTILAPI void md5put (uint8_t output[16]); + +UTILAPI void md5 (const void *input, size_t length, uint8_t output[16]); + +#endif
\ No newline at end of file diff --git a/Build/source/texk/web2c/luatexdir/luapplib/util/utilmem.c b/Build/source/texk/web2c/luatexdir/luapplib/util/utilmem.c new file mode 100644 index 00000000000..1fa00a0c10a --- /dev/null +++ b/Build/source/texk/web2c/luatexdir/luapplib/util/utilmem.c @@ -0,0 +1,68 @@ + +#include <string.h> // for memcpy + +#include "utilmem.h" +#include "utillog.h" + +#ifndef util_memerr +# if defined(_WIN64) || defined(__MINGW32__) +# define util_memerr(size) { loggerf("ooops, not enough memory (%I64u)", ((unsigned long long)(size))); abort(); } +# else +# define util_memerr(size) { loggerf("ooops, not enough memory (%llu)", ((unsigned long long)(size))); abort(); } +# endif +#endif + +void * util_malloc (size_t size) +{ + void *m; + if ((m = malloc(size)) == NULL) + util_memerr(size); + return m; +} + +void * util_calloc (size_t num, size_t size) +{ + void *m; + if ((m = calloc(num, size)) == NULL) + util_memerr(size); + return m; +} + +void * util_realloc (void *m, size_t size) +{ + if ((m = realloc(m, size)) == NULL) + util_memerr(size); + return m; +} + +/* common array resizer + +data -- the beginning of array +unit -- sizeof array element +size -- current array size +extra -- requested extra size +space -- pointer to available space +allocated -- flag indicating if *data has been allocated (with malloc) + +*/ + +void util_resize (void **data, size_t unit, size_t size, size_t extra, size_t *space, int allocated) +{ + if (*space == 0) + *space = 4; // ... better keep *space non-zero to avoid it + do { *space <<= 1; } + while (size + extra > *space); + + if (allocated) + { + *data = util_realloc(*data, *space * unit); + } + else + { + void *newdata = util_malloc(*space * unit); + if (*data != NULL) + memcpy(newdata, *data, size * unit); + *data = newdata; + } +} + diff --git a/Build/source/texk/web2c/luatexdir/luapplib/util/utilmem.h b/Build/source/texk/web2c/luatexdir/luapplib/util/utilmem.h new file mode 100644 index 00000000000..e2d1286187b --- /dev/null +++ b/Build/source/texk/web2c/luatexdir/luapplib/util/utilmem.h @@ -0,0 +1,59 @@ + +#ifndef UTIL_MEM_H +#define UTIL_MEM_H + +#include <stdlib.h> // for size_t and alloc functions +#include "utildecl.h" + +UTILAPI void * util_malloc (size_t size); +UTILAPI void * util_calloc (size_t num, size_t size); +UTILAPI void * util_realloc (void *m, size_t size); + +#define util_free free // not a call + +/* arrays */ + +#define array_header(element_type, integer_type) \ + element_type *data; integer_type size; integer_type space + +#define array_of(element_type) array_header(element_type, size_t) + +#define array_data_is_allocated(array) ((array)->data != (void *)((array) + 1)) + +#define array_create(array, array_type, element_type, init) \ + (array = (array_type *)util_malloc(sizeof(array_type) + ((init) > 0 ? ((init) * sizeof(element_type)) : 1)), \ + array_init(array, element_type, init)) + +#define array_init(array, element_type, init) \ + ((array)->data = (element_type *)((array) + 1), (array)->size = 0, (array)->space = init) + +#define array_init_data(array, element_type, init) \ + ((array)->data = (element_type *)util_malloc((init) * sizeof(element_type)), (array)->size = 0, (array)->space = init) + +#define array_free_data(array) \ + ((void)(array_data_is_allocated(array) && (util_free((array)->data), 0))) + +#define array_free(array) \ + ((void)(array_free_data(array), (util_free(array), 0))) + +void util_resize (void **data, size_t unit, size_t size, size_t extra, size_t *space, int allocated); + +#define array_ensure(array, unit, extra) \ + ((void)((array)->size + (extra) > (array)->space && \ + (util_resize((void **)(&(array)->data), unit, (array)->size, extra, &(array)->space, array_data_is_allocated(array)), 0))) + +#define array_ensure_alloc(array, unit, extra, isalloc) \ + ((void)((array)->size + (extra) > (array)->space && \ + (util_resize((void **)(&(array)->data), unit, (array)->size, extra, &(array)->space, isalloc), 0))) + +#define array_at(array, index) ((array)->data + index) +#define array_index(array, index) (index < (array)->size ? array_at(array, index) : NULL) +#define array_top(array) ((array)->data + (array)->size - 1) + +#define array_from_bottom(array, index) array_at(array, index - 1) +#define array_from_top(array, negindex) array_at(array, (array)->size + (negindex)) + +#define array_push(array) ((array)->data + ((array)->size)++) +#define array_pop(array) ((array)->data + --((array)->size)) + +#endif diff --git a/Build/source/texk/web2c/luatexdir/luapplib/util/utilnumber.c b/Build/source/texk/web2c/luatexdir/luapplib/util/utilnumber.c new file mode 100644 index 00000000000..c87261e15ab --- /dev/null +++ b/Build/source/texk/web2c/luatexdir/luapplib/util/utilnumber.c @@ -0,0 +1,1325 @@ + +#include <math.h> /* for log10() and floor() */ +#include <stdio.h> /* for printf() */ + +#include "utilnumber.h" + +const int base10_lookup[] = { + -1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1, + -1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1, + -1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1, + 0, 1, 2, 3, 4, 5, 6, 7, 8, 9,-1,-1,-1,-1,-1,-1, + -1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1, + -1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1, + -1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1, + -1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1, + -1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1, + -1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1, + -1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1, + -1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1, + -1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1, + -1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1, + -1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1, + -1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1 +}; + +const int base16_lookup[] = { + -1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1, + -1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1, + -1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1, + 0, 1, 2, 3, 4, 5, 6, 7, 8, 9,-1,-1,-1,-1,-1,-1, + -1,10,11,12,13,14,15,-1,-1,-1,-1,-1,-1,-1,-1,-1, + -1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1, + -1,10,11,12,13,14,15,-1,-1,-1,-1,-1,-1,-1,-1,-1, + -1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1, + -1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1, + -1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1, + -1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1, + -1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1, + -1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1, + -1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1, + -1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1, + -1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1 +}; + +const int base26_lookup[] = { + -1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1, + -1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1, + -1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1, + -1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1, + -1, 1, 2, 3, 4, 5, 6, 7, 8, 9,10,11,12,13,14,15, + 16,17,18,19,20,21,22,23,24,25,26,-1,-1,-1,-1,-1, + -1, 1, 2, 3, 4, 5, 6, 7, 8, 9,10,11,12,13,14,15, + 16,17,18,19,20,21,22,23,24,25,26,-1,-1,-1,-1,-1, + -1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1, + -1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1, + -1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1, + -1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1, + -1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1, + -1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1, + -1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1, + -1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1 +}; + +const int base36_lookup[] = { + -1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1, + -1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1, + -1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1, + 0, 1, 2, 3, 4, 5, 6, 7, 8, 9,-1,-1,-1,-1,-1,-1, + -1,10,11,12,13,14,15,16,17,18,19,20,21,22,23,24, + 25,26,27,28,29,30,31,32,33,34,35,-1,-1,-1,-1,-1, + -1,10,11,12,13,14,15,16,17,18,19,20,21,22,23,24, + 25,26,27,28,29,30,31,32,33,34,35,-1,-1,-1,-1,-1, + -1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1, + -1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1, + -1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1, + -1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1, + -1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1, + -1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1, + -1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1, + -1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1 +}; + +/* integer from string; return a pointer to character next to the last digit */ + +#define string_scan_sign(s, c, sign) _scan_sign(c, sign, *++s) +#define string_scan_integer(s, c, number) _scan_integer(c, number, *++s) +#define string_scan_radix(s, c, number, radix) _scan_radix(c, number, radix, *++s) +#define string_read_integer(s, c, number) _read_integer(c, number, *++s) +#define string_read_radix(s, c, number, radix) _read_radix(c, number, radix, *++s) + +const char * string_to_int32 (const char *s, int32_t *number) +{ + int sign, c = *s; + string_scan_sign(s, c, sign); + string_scan_integer(s, c, *number); + if (sign) *number = -*number; + return s; +} + +const char * string_to_intlw (const char *s, intlw_t *number) +{ + int sign, c = *s; + string_scan_sign(s, c, sign); + string_scan_integer(s, c, *number); + if (sign) *number = -*number; + return s; +} + +const char * string_to_int64 (const char *s, int64_t *number) +{ + int sign, c = *s; + string_scan_sign(s, c, sign); + string_scan_integer(s, c, *number); + if (sign) *number = -*number; + return s; +} + +const char * string_to_uint32 (const char *s, uint32_t *number) +{ + int c = *s; + string_scan_integer(s, c, *number); + return s; +} + +const char * string_to_uintlw (const char *s, uintlw_t *number) +{ + int c = *s; + string_scan_integer(s, c, *number); + return s; +} + +const char * string_to_uint64 (const char *s, uint64_t *number) +{ + int c = *s; + string_scan_integer(s, c, *number); + return s; +} + +const char * radix_to_int32 (const char *s, int32_t *number, int radix) +{ + int sign, c = *s; + string_scan_sign(s, c, sign); + string_scan_radix(s, c, *number, radix); + if (sign) *number = -*number; + return s; +} + +const char * radix_to_intlw (const char *s, intlw_t *number, int radix) +{ + int sign, c = *s; + string_scan_sign(s, c, sign); + string_scan_radix(s, c, *number, radix); + if (sign) *number = -*number; + return s; +} + +const char * radix_to_int64 (const char *s, int64_t *number, int radix) +{ + int sign, c = *s; + string_scan_sign(s, c, sign); + string_scan_radix(s, c, *number, radix); + if (sign) *number = -*number; + return s; +} + +const char * radix_to_uint32 (const char *s, uint32_t *number, int radix) +{ + int c = *s; + string_scan_radix(s, c, *number, radix); + return s; +} + +const char * radix_to_uintlw (const char *s, uintlw_t *number, int radix) +{ + int c = *s; + string_scan_radix(s, c, *number, radix); + return s; +} + +const char * radix_to_uint64 (const char *s, uint64_t *number, int radix) +{ + int c = *s; + string_scan_radix(s, c, *number, radix); + return s; +} + +/* roman to uint16_t */ + +#define roman1000(c) (c == 'M' || c == 'm') +#define roman500(c) (c == 'D' || c == 'd') +#define roman100(c) (c == 'C' || c == 'c') +#define roman50(c) (c == 'L' || c == 'l') +#define roman10(c) (c == 'X' || c == 'x') +#define roman5(c) (c == 'V' || c == 'v') +#define roman1(c) (c == 'I' || c == 'i') + +#define roman100s(p) (roman100(*p) ? (100 + ((++p, roman100(*p)) ? (100 + ((++p, roman100(*p)) ? (++p, 100) : 0)) : 0)) : 0) +#define roman10s(p) (roman10(*p) ? (10 + ((++p, roman10(*p)) ? (10 + ((++p, roman10(*p)) ? (++p, 10) : 0)) : 0)) : 0) +#define roman1s(p) (roman1(*p) ? (1 + ((++p, roman1(*p)) ? (1 + ((++p, roman1(*p)) ? (++p, 1) : 0)) : 0)) : 0) + +const char * roman_to_uint16 (const char *s, uint16_t *number) +{ + const char *p; + /* M */ + for (*number = 0, p = s; roman1000(*p); *number += 1000, ++p); + /* D C */ + if (roman500(*p)) + { + ++p; + *number += 500 + roman100s(p); + } + else if (roman100(*p)) + { + ++p; + if (roman1000(*p)) + { + ++p; + *number += 900; + } + else if (roman500(*p)) + { + ++p; + *number += 400; + } + else + *number += 100 + roman100s(p); + } + /* L X */ + if (roman50(*p)) + { + ++p; + *number += 50 + roman10s(p); + } + else if (roman10(*p)) + { + ++p; + if (roman100(*p)) + { + ++p; + *number += 90; + } + else if (roman50(*p)) + { + ++p; + *number += 40; + } + else + *number += 10 + roman10s(p); + } + /* V I */ + if (roman5(*p)) + { + ++p; + *number += 5 + roman1s(p); + } + else if (roman1(*p)) + { + ++p; + if (roman10(*p)) + { + ++p; + *number += 9; + } + else if (roman5(*p)) + { + ++p; + *number += 4; + } + else + *number += 1 + roman1s(p); + } + return p; +} + +/* integer to string; return a pointer to null-terminated static const string */ + +static char integer_buffer[MAX_INTEGER_DIGITS] = {'\0'}; +#define end_of_integer_buffer (integer_buffer + MAX_INTEGER_DIGITS - 1) + +/* writing integers */ + +#define number_printrev_signed(p, number, quotient) \ + do { \ + quotient = number; number /= 10; \ + *--p = base10_palindrome[9 + (quotient - number*10)]; \ + } while (number); \ + if (quotient < 0) *--p = '-' + +#define number_printrev_unsigned(p, number, quotient) \ + do { \ + quotient = number; number /= 10; \ + *--p = (char)(quotient - integer_multiplied10(number)) + '0'; \ + } while (number) + +char * int32_as_string (int32_t number, char **e) +{ + char *p; + int quotient; + p = end_of_integer_buffer; *p = '\0'; + if (e != NULL) *e = p; + number_printrev_signed(p, number, quotient); + return p; +} + +char * intlw_as_string (intlw_t number, char **e) +{ + char *p; + intlw_t quotient; + p = end_of_integer_buffer; *p = '\0'; + if (e != NULL) *e = p; + number_printrev_signed(p, number, quotient); + return p; +} + +char * int64_as_string (int64_t number, char **e) +{ + char *p; + int64_t quotient; + p = end_of_integer_buffer; *p = '\0'; + if (e != NULL) *e = p; + number_printrev_signed(p, number, quotient); + return p; +} + +char * uint32_as_string (uint32_t number, char **e) +{ + char *p; + uint32_t quotient; + p = end_of_integer_buffer; *p = '\0'; + if (e != NULL) *e = p; + number_printrev_unsigned(p, number, quotient); + return p; +} + +char * uintlw_as_string (uintlw_t number, char **e) +{ + char *p; + uintlw_t quotient; + p = end_of_integer_buffer; *p = '\0'; + if (e != NULL) *e = p; + number_printrev_unsigned(p, number, quotient); + return p; +} + +char * uint64_as_string (uint64_t number, char **e) +{ + char *p; + uint64_t quotient; + p = end_of_integer_buffer; *p = '\0'; + if (e != NULL) *e = p; + number_printrev_unsigned(p, number, quotient); + return p; +} + +/* radix variant */ + +#define number_printrev_signed_radix_uc(p, number, radix, quotient) \ + do { \ + quotient = number; number /= radix; \ + *--p = base36_uc_palindrome[MAX_RADIX - 1 + (quotient - number*radix)]; \ + } while (number) + +#define number_printrev_signed_radix_lc(p, number, radix, quotient) \ + do { \ + quotient = number; number /= radix; \ + *--p = base36_lc_palindrome[MAX_RADIX - 1 + (quotient - number*radix)]; \ + } while (number) + +#define number_printrev_signed_radix(p, number, radix, quotient) \ + do { \ + if (radix > 0) { number_printrev_signed_radix_uc(p, number, radix, quotient); } \ + else { radix = -radix; number_printrev_signed_radix_lc(p, number, radix, quotient); } \ + if (quotient < 0) *--p = '-'; \ + } while (0) + +#define number_printrev_unsigned_radix_uc(p, number, radix, quotient) \ + do { \ + quotient = number; number /= radix; \ + *--p = base36_uc_alphabet[quotient % radix]; \ + } while (number) + +#define number_printrev_unsigned_radix_lc(p, number, radix, quotient) \ + do { \ + quotient = number; number /= radix; \ + *--p = base36_lc_alphabet[quotient % radix]; \ + } while (number) + +#define number_printrev_unsigned_radix(p, number, radix, quotient) \ + do { \ + if (radix > 0) { number_printrev_unsigned_radix_uc(p, number, radix, quotient); } \ + else { radix = -radix; number_printrev_unsigned_radix_lc(p, number, radix, quotient); } \ + } while (0) + +char * int32_as_radix (int number, int radix, char **e) +{ + char *p; + int quotient; + p = end_of_integer_buffer; *p = '\0'; + if (e != NULL) *e = p; + number_printrev_signed_radix(p, number, radix, quotient); + return p; +} + +char * intlw_as_radix (intlw_t number, int radix, char **e) +{ + char *p; + intlw_t quotient; + p = end_of_integer_buffer; *p = '\0'; + if (e != NULL) *e = p; + number_printrev_signed_radix(p, number, radix, quotient); + return p; +} + +char * int64_as_radix (int64_t number, int radix, char **e) +{ + char *p; + int64_t quotient; + p = end_of_integer_buffer; *p = '\0'; + if (e != NULL) *e = p; + number_printrev_signed_radix(p, number, radix, quotient); + return p; +} + +char * uint32_as_radix (uint32_t number, int radix, char **e) +{ + char *p; + uint32_t quotient; + p = end_of_integer_buffer; *p = '\0'; + if (e != NULL) *e = p; + number_printrev_unsigned_radix(p, number, radix, quotient); + return p; +} + +char * uintlw_as_radix (uintlw_t number, int radix, char **e) +{ + char *p; + uintlw_t quotient; + p = end_of_integer_buffer; *p = '\0'; + if (e != NULL) *e = p; + number_printrev_unsigned_radix(p, number, radix, quotient); + return p; +} + +char * uint64_as_radix (uint64_t number, int radix, char **e) +{ + char *p; + uint64_t quotient; + p = end_of_integer_buffer; *p = '\0'; + if (e != NULL) *e = p; + number_printrev_unsigned_radix(p, number, radix, quotient); + return p; +} + +/* aaa, aab, aac, ...; unsigned only. 0 gives empty string */ + +#define string_scan_alpha(s, c, number, radix) \ + for (number = 0, c = *s; (c = base26_value(c)) > 0; number = number * radix + c, c = *++s) + +const char * alpha_to_uint32 (const char *s, uint32_t *number) +{ + int c; + string_scan_alpha(s, c, *number, 26); + return s; +} + +const char * alpha_to_uintlw (const char *s, uintlw_t *number) +{ + int c; + string_scan_alpha(s, c, *number, 26); + return s; +} + +const char * alpha_to_uint64 (const char *s, uint64_t *number) +{ + int c; + string_scan_alpha(s, c, *number, 26); + return s; +} + +#define number_printrev_unsigned_alpha_uc(p, number, radix, quotient) \ + while (number > 0) { \ + quotient = --number; number /= radix; \ + *--p = base26_uc_alphabet[quotient % radix]; \ + } + +#define number_printrev_unsigned_alpha_lc(p, number, radix, quotient) \ + while (number > 0) { \ + quotient = --number; number /= radix; \ + *--p = base26_lc_alphabet[quotient % radix]; \ + } + +char * uint32_as_alpha_uc (uint32_t number, char **e) +{ + char *p; + uint32_t quotient; + p = end_of_integer_buffer; *p = '\0'; + if (e != NULL) *e = p; + number_printrev_unsigned_alpha_uc(p, number, 26, quotient); + return p; +} + +char * uint32_as_alpha_lc (uint32_t number, char **e) +{ + char *p; + uint32_t quotient; + p = end_of_integer_buffer; *p = '\0'; + if (e != NULL) *e = p; + number_printrev_unsigned_alpha_lc(p, number, 26, quotient); + return p; +} + +char * uintlw_as_alpha_uc (uintlw_t number, char **e) +{ + char *p; + uintlw_t quotient; + p = end_of_integer_buffer; *p = '\0'; + if (e != NULL) *e = p; + number_printrev_unsigned_alpha_uc(p, number, 26, quotient); + return p; +} + +char * uintlw_as_alpha_lc (uintlw_t number, char **e) +{ + char *p; + uintlw_t quotient; + p = end_of_integer_buffer; *p = '\0'; + if (e != NULL) *e = p; + number_printrev_unsigned_alpha_lc(p, number, 26, quotient); + return p; +} + +char * uint64_as_alpha_uc (uint64_t number, char **e) +{ + char *p; + uint64_t quotient; + p = end_of_integer_buffer; *p = '\0'; + if (e != NULL) *e = p; + number_printrev_unsigned_alpha_uc(p, number, 26, quotient); + return p; +} + +char * uint64_as_alpha_lc (uint64_t number, char **e) +{ + char *p; + uint64_t quotient; + p = end_of_integer_buffer; *p = '\0'; + if (e != NULL) *e = p; + number_printrev_unsigned_alpha_lc(p, number, 26, quotient); + return p; +} + +/* a variant of alphabetic, a, b, c, ..., z, aa, bb, cc, ..., zz (eg. pdf page labelling) */ + +#define string_scan_alphan(s, c, number, radix) \ + do { \ + number = 0; \ + if ((c = base26_value(*s)) > 0) { \ + number = c; \ + while (c == base26_value(*++s)) number += radix; \ + } \ + } while (0) + +const char * alphan_to_uint32 (const char *s, uint32_t *number) +{ + int c; + string_scan_alphan(s, c, *number, 26); + return s; +} + +const char * alphan_to_uintlw (const char *s, uintlw_t *number) +{ + int c; + string_scan_alphan(s, c, *number, 26); + return s; +} + +const char * alphan_to_uint64 (const char *s, uintlw_t *number) +{ + int c; + string_scan_alphan(s, c, *number, 26); + return s; +} + +#define number_print_alphan_uc(s, c, number, radix) \ + if (number > 0) { \ + for (c = (--number) % radix, number -= c; ; number -= radix) { \ + *s++ = base26_uc_alphabet[c]; \ + if (number == 0 || p >= end_of_integer_buffer) break; \ + } \ + } + +#define number_print_alphan_lc(s, c, number, radix) \ + if (number > 0) { \ + for (c = (--number) % radix, number -= c; ; number -= radix) { \ + *s++ = base26_lc_alphabet[c]; \ + if (number == 0 || p >= end_of_integer_buffer) break; \ + } \ + } + +char * uint32_as_alphan_uc (uint32_t number, char **e) +{ + char *p; + uint8_t c; + p = integer_buffer; + number_print_alphan_uc(p, c, number, 26); + *p = '\0'; if (e != NULL) *e = p; + return integer_buffer; +} + +char * uint32_as_alphan_lc (uint32_t number, char **e) +{ + char *p; + uint8_t c; + p = integer_buffer; + number_print_alphan_lc(p, c, number, 26); + *p = '\0'; if (e != NULL) *e = p; + return integer_buffer; +} + +char * uintlw_as_alphan_uc (uintlw_t number, char **e) +{ + char *p; + uint8_t c; + p = integer_buffer; + number_print_alphan_uc(p, c, number, 26); + *p = '\0'; if (e != NULL) *e = p; + return integer_buffer; +} + +char * uintlw_as_alphan_lc (uintlw_t number, char **e) +{ + char *p; + uint8_t c; + p = integer_buffer; + number_print_alphan_lc(p, c, number, 26); + *p = '\0'; if (e != NULL) *e = p; + return integer_buffer; +} + +char * uint64_as_alphan_uc (uint64_t number, char **e) +{ + char *p; + uint8_t c; + p = integer_buffer; + number_print_alphan_uc(p, c, number, 26); + *p = '\0'; if (e != NULL) *e = p; + return integer_buffer; +} + +char * uint64_as_alphan_lc (uint64_t number, char **e) +{ + char *p; + uint8_t c; + p = integer_buffer; + number_print_alphan_lc(p, c, number, 26); + *p = '\0'; if (e != NULL) *e = p; + return integer_buffer; +} + +/* roman numeral */ + +/* todo: large roman numerals? http://mathforum.org/library/drmath/view/57569.html */ + +#define base_roman_uc_alphabet "MDCLXVI" +#define base_roman_lc_alphabet "mdclxvi" + +static const uint32_t base_roman_values[] = { 1000, 500, 100, 50, 10, 5, 1 }; + +#define integer_to_roman(p, number, alphabet) \ + { \ + uint32_t k, j, v, u; \ + for (j = 0, v = base_roman_values[0]; number > 0; ) \ + { \ + if (number >= v) \ + { \ + *p++ = alphabet[j]; \ + number -= v; \ + continue; \ + } \ + if (j & 1) \ + k = j + 1; \ + else \ + k = j + 2; \ + u = base_roman_values[k]; \ + if (number + u >= v) \ + { \ + *p++ = alphabet[k]; \ + number += u; \ + } \ + else \ + v = base_roman_values[++j]; \ + } \ + } + +char * uint16_as_roman_uc (uint16_t number, char **e) +{ + char *p = integer_buffer; + integer_to_roman(p, number, base_roman_uc_alphabet); + if (e != NULL) + *e = p; + *p = '\0'; + return integer_buffer; +} + +char * uint16_as_roman_lc (uint16_t number, char **e) +{ + char *p = integer_buffer; + integer_to_roman(p, number, base_roman_lc_alphabet); + if (e != NULL) + *e = p; + *p = '\0'; + return integer_buffer; +} + +/* IEEE-754 */ + +#define BINARY_MODF 1 + +#define NOT_A_NUMBER_STRING "NaN" +#define INFINITY_STRING "INF" +#define SIGNED_INFINITY 1 +#define SIGNED_ZERO 0 +#define SIGNED_NOT_A_NUMBER 0 +#define RADIX_CHAR '.' + +/* double/float to decimal */ + +typedef struct ieee_double { + union { + double number; + uint64_t bits; + }; + uint64_t fraction; + int exponent, sign; +} ieee_double; + +typedef struct ieee_float { + union { + float number; + uint32_t bits; + }; + uint32_t fraction; + int exponent, sign; +} ieee_float; + +#define IEEE_DOUBLE_BIAS 1023 +#define IEEE_DOUBLE_MIN_EXPONENT -1023 +#define IEEE_DOUBLE_MAX_EXPONENT (0x7ff - IEEE_DOUBLE_BIAS) + +#define IEEE_FLOAT_BIAS 127 +#define IEEE_FLOAT_MIN_EXPONENT -127 +#define IEEE_FLOAT_MAX_EXPONENT (0xff - IEEE_FLOAT_BIAS) + +#define ieee_double_fraction(i) (i & 0x000fffffffffffffull) +#define ieee_double_exponent(i) ((0x7ff & (i >> 52)) - IEEE_DOUBLE_BIAS) +#define ieee_double_sign(ieee_number) ((void)((ieee_number.sign = ieee_number.bits >> 63) && (ieee_number.number = -ieee_number.number))) +#define ieee_double_init(ieee_number, number) \ + ieee_number.number = number, \ + ieee_number.fraction = ieee_double_fraction(ieee_number.bits), \ + ieee_number.exponent = ieee_double_exponent(ieee_number.bits) + +#define ieee_float_fraction(i) (i & 0x007fffff) +#define ieee_float_exponent(i) ((0xff & (i >> 23)) - IEEE_FLOAT_BIAS) +#define ieee_float_sign(ieee_number) ((void)((ieee_number.sign = ieee_number.bits >> 31) && (ieee_number.number = -ieee_number.number))) +#define ieee_float_init(ieee_number, number) \ + ieee_number.number = number, \ + ieee_number.fraction = ieee_float_fraction(ieee_number.bits), \ + ieee_number.exponent = ieee_float_exponent(ieee_number.bits) + +/* special cases */ + +#define ieee_double_is_zero(ieee_number) (ieee_number.number == 0) // || ieee_double_too_small(ieee_number) ? +#define ieee_double_too_small(ieee_number) (ieee_number.exponent == 0 && ieee_number.fraction != 0) // denormalized, implicit fracion bit not set + +#define ieee_float_is_zero(ieee_number) (ieee_number.number == 0) // || ieee_float_too_small(ieee_number) ? +#define ieee_float_too_small(ieee_number) (ieee_number.exponent == 0 && ieee_number.fraction != 0) + +#define ieee_double_zero_string(ieee_number) (SIGNED_ZERO && ieee_number.sign ? "-0" : "0") +#define ieee_double_infinity_string(ieee_number) (SIGNED_INFINITY && ieee_number.sign ? "-" INFINITY_STRING : INFINITY_STRING) + +#define ieee_float_zero_string ieee_double_zero_string +#define ieee_float_infinity_string ieee_double_infinity_string + +#define ieee_double_special_case(ieee_number) (ieee_number.exponent == IEEE_DOUBLE_MAX_EXPONENT) +#define ieee_double_special_string(ieee_number) (ieee_number.fraction ? NOT_A_NUMBER_STRING : ieee_double_infinity_string(ieee_number)) + +#define ieee_float_special_case(ieee_number) (ieee_number.exponent == IEEE_FLOAT_MAX_EXPONENT) +#define ieee_float_special_string(ieee_number) (ieee_number.fraction ? NOT_A_NUMBER_STRING : ieee_float_infinity_string(ieee_number)) + +#if 0 + +const double double_binary_power10[] = +{ + 1.0e1, 1.0e2, 1.0e4, 1.0e8, 1.0e16, 1.0e32, 1.0e64, 1.0e128, 1.0e256 +}; + +const float float_binary_power10[] = +{ + 1.0e1, 1.0e2, 1.0e4, 1.0e8, 1.0e16, 1.0e32 +}; + +const double double_binary_negpower10[] = +{ + 1.0e-1, 1.0e-2, 1.0e-4, 1.0e-8, 1.0e-16, 1.0e-32 +}; + +const float float_binary_negpower10[] = +{ + 1.0e-1, 1.0e-2, 1.0e-4, 1.0e-8, 1.0e-16, 1.0e-32 +}; + +#else + +const double double_decimal_power10[] = { + 1.0e0, 1.0e1, 1.0e2, 1.0e3, 1.0e4, 1.0e5, 1.0e6, 1.0e7, 1.0e8, 1.0e9, + 1.0e10, 1.0e11, 1.0e12, 1.0e13, 1.0e14, 1.0e15, 1.0e16, 1.0e17, 1.0e18, 1.0e19, + 1.0e20, 1.0e21, 1.0e22, 1.0e23, 1.0e24, 1.0e25, 1.0e26, 1.0e27, 1.0e28, 1.0e29, + 1.0e30, 1.0e31, 1.0e32, 1.0e33, 1.0e34, 1.0e35, 1.0e36, 1.0e37, 1.0e38, 1.0e39, + 1.0e40, 1.0e41, 1.0e42, 1.0e43, 1.0e44, 1.0e45, 1.0e46, 1.0e47, 1.0e48, 1.0e49, + 1.0e50, 1.0e51, 1.0e52, 1.0e53, 1.0e54, 1.0e55, 1.0e56, 1.0e57, 1.0e58, 1.0e59, + 1.0e60, 1.0e61, 1.0e62, 1.0e63, 1.0e64, 1.0e65, 1.0e66, 1.0e67, 1.0e68, 1.0e69, + 1.0e70, 1.0e71, 1.0e72, 1.0e73, 1.0e74, 1.0e75, 1.0e76, 1.0e77, 1.0e78, 1.0e79, + 1.0e80, 1.0e81, 1.0e82, 1.0e83, 1.0e84, 1.0e85, 1.0e86, 1.0e87, 1.0e88, 1.0e89, + 1.0e90, 1.0e91, 1.0e92, 1.0e93, 1.0e94, 1.0e95, 1.0e96, 1.0e97, 1.0e98, 1.0e99, + 1.0e100, 1.0e101, 1.0e102, 1.0e103, 1.0e104, 1.0e105, 1.0e106, 1.0e107, 1.0e108, 1.0e109, + 1.0e110, 1.0e111, 1.0e112, 1.0e113, 1.0e114, 1.0e115, 1.0e116, 1.0e117, 1.0e118, 1.0e119, + 1.0e120, 1.0e121, 1.0e122, 1.0e123, 1.0e124, 1.0e125, 1.0e126, 1.0e127, 1.0e128, 1.0e129, + 1.0e130, 1.0e131, 1.0e132, 1.0e133, 1.0e134, 1.0e135, 1.0e136, 1.0e137, 1.0e138, 1.0e139, + 1.0e140, 1.0e141, 1.0e142, 1.0e143, 1.0e144, 1.0e145, 1.0e146, 1.0e147, 1.0e148, 1.0e149, + 1.0e150, 1.0e151, 1.0e152, 1.0e153, 1.0e154, 1.0e155, 1.0e156, 1.0e157, 1.0e158, 1.0e159, + 1.0e160, 1.0e161, 1.0e162, 1.0e163, 1.0e164, 1.0e165, 1.0e166, 1.0e167, 1.0e168, 1.0e169, + 1.0e170, 1.0e171, 1.0e172, 1.0e173, 1.0e174, 1.0e175, 1.0e176, 1.0e177, 1.0e178, 1.0e179, + 1.0e180, 1.0e181, 1.0e182, 1.0e183, 1.0e184, 1.0e185, 1.0e186, 1.0e187, 1.0e188, 1.0e189, + 1.0e190, 1.0e191, 1.0e192, 1.0e193, 1.0e194, 1.0e195, 1.0e196, 1.0e197, 1.0e198, 1.0e199, + 1.0e200, 1.0e201, 1.0e202, 1.0e203, 1.0e204, 1.0e205, 1.0e206, 1.0e207, 1.0e208, 1.0e209, + 1.0e210, 1.0e211, 1.0e212, 1.0e213, 1.0e214, 1.0e215, 1.0e216, 1.0e217, 1.0e218, 1.0e219, + 1.0e220, 1.0e221, 1.0e222, 1.0e223, 1.0e224, 1.0e225, 1.0e226, 1.0e227, 1.0e228, 1.0e229, + 1.0e230, 1.0e231, 1.0e232, 1.0e233, 1.0e234, 1.0e235, 1.0e236, 1.0e237, 1.0e238, 1.0e239, + 1.0e240, 1.0e241, 1.0e242, 1.0e243, 1.0e244, 1.0e245, 1.0e246, 1.0e247, 1.0e248, 1.0e249, + 1.0e250, 1.0e251, 1.0e252, 1.0e253, 1.0e254, 1.0e255, 1.0e256, 1.0e257, 1.0e258, 1.0e259, + 1.0e260, 1.0e261, 1.0e262, 1.0e263, 1.0e264, 1.0e265, 1.0e266, 1.0e267, 1.0e268, 1.0e269, + 1.0e270, 1.0e271, 1.0e272, 1.0e273, 1.0e274, 1.0e275, 1.0e276, 1.0e277, 1.0e278, 1.0e279, + 1.0e280, 1.0e281, 1.0e282, 1.0e283, 1.0e284, 1.0e285, 1.0e286, 1.0e287, 1.0e288, 1.0e289, + 1.0e290, 1.0e291, 1.0e292, 1.0e293, 1.0e294, 1.0e295, 1.0e296, 1.0e297, 1.0e298, 1.0e299, + 1.0e300, 1.0e301, 1.0e302, 1.0e303, 1.0e304, 1.0e305, 1.0e306, 1.0e307, 1.0e308 +}; + +const float float_decimal_power10[] = { + 1.0e0f, 1.0e1f, 1.0e2f, 1.0e3f, 1.0e4f, 1.0e5f, 1.0e6f, 1.0e7f, 1.0e8f, 1.0e9f, + 1.0e10f, 1.0e11f, 1.0e12f, 1.0e13f, 1.0e14f, 1.0e15f, 1.0e16f, 1.0e17f, 1.0e18f, 1.0e19f, + 1.0e20f, 1.0e21f, 1.0e22f, 1.0e23f, 1.0e24f, 1.0e25f, 1.0e26f, 1.0e27f, 1.0e28f, 1.0e29f, + 1.0e30f, 1.0e31f, 1.0e32f, 1.0e33f, 1.0e34f, 1.0e35f, 1.0e36f, 1.0e37f, 1.0e38f +}; + +const double double_decimal_negpower10[] = { + 1.0e0, 1.0e-1, 1.0e-2, 1.0e-3, 1.0e-4, 1.0e-5, 1.0e-6, 1.0e-7, 1.0e-8, 1.0e-9, + 1.0e-10, 1.0e-11, 1.0e-12, 1.0e-13, 1.0e-14, 1.0e-15, 1.0e-16, 1.0e-17, 1.0e-18, 1.0e-19, + 1.0e-20, 1.0e-21, 1.0e-22, 1.0e-23, 1.0e-24, 1.0e-25, 1.0e-26, 1.0e-27, 1.0e-28, 1.0e-29, + 1.0e-30, 1.0e-31, 1.0e-32, 1.0e-33, 1.0e-34, 1.0e-35, 1.0e-36, 1.0e-37, 1.0e-38, 1.0e-39, + 1.0e-40, 1.0e-41, 1.0e-42, 1.0e-43, 1.0e-44, 1.0e-45, 1.0e-46, 1.0e-47, 1.0e-48, 1.0e-49, + 1.0e-50, 1.0e-51, 1.0e-52, 1.0e-53, 1.0e-54, 1.0e-55, 1.0e-56, 1.0e-57, 1.0e-58, 1.0e-59, + 1.0e-60, 1.0e-61, 1.0e-62, 1.0e-63, 1.0e-64, 1.0e-65, 1.0e-66, 1.0e-67, 1.0e-68, 1.0e-69, + 1.0e-70, 1.0e-71, 1.0e-72, 1.0e-73, 1.0e-74, 1.0e-75, 1.0e-76, 1.0e-77, 1.0e-78, 1.0e-79, + 1.0e-80, 1.0e-81, 1.0e-82, 1.0e-83, 1.0e-84, 1.0e-85, 1.0e-86, 1.0e-87, 1.0e-88, 1.0e-89, + 1.0e-90, 1.0e-91, 1.0e-92, 1.0e-93, 1.0e-94, 1.0e-95, 1.0e-96, 1.0e-97, 1.0e-98, 1.0e-99, + 1.0e-100, 1.0e-101, 1.0e-102, 1.0e-103, 1.0e-104, 1.0e-105, 1.0e-106, 1.0e-107, 1.0e-108, 1.0e-109, + 1.0e-110, 1.0e-111, 1.0e-112, 1.0e-113, 1.0e-114, 1.0e-115, 1.0e-116, 1.0e-117, 1.0e-118, 1.0e-119, + 1.0e-120, 1.0e-121, 1.0e-122, 1.0e-123, 1.0e-124, 1.0e-125, 1.0e-126, 1.0e-127, 1.0e-128, 1.0e-129, + 1.0e-130, 1.0e-131, 1.0e-132, 1.0e-133, 1.0e-134, 1.0e-135, 1.0e-136, 1.0e-137, 1.0e-138, 1.0e-139, + 1.0e-140, 1.0e-141, 1.0e-142, 1.0e-143, 1.0e-144, 1.0e-145, 1.0e-146, 1.0e-147, 1.0e-148, 1.0e-149, + 1.0e-150, 1.0e-151, 1.0e-152, 1.0e-153, 1.0e-154, 1.0e-155, 1.0e-156, 1.0e-157, 1.0e-158, 1.0e-159, + 1.0e-160, 1.0e-161, 1.0e-162, 1.0e-163, 1.0e-164, 1.0e-165, 1.0e-166, 1.0e-167, 1.0e-168, 1.0e-169, + 1.0e-170, 1.0e-171, 1.0e-172, 1.0e-173, 1.0e-174, 1.0e-175, 1.0e-176, 1.0e-177, 1.0e-178, 1.0e-179, + 1.0e-180, 1.0e-181, 1.0e-182, 1.0e-183, 1.0e-184, 1.0e-185, 1.0e-186, 1.0e-187, 1.0e-188, 1.0e-189, + 1.0e-190, 1.0e-191, 1.0e-192, 1.0e-193, 1.0e-194, 1.0e-195, 1.0e-196, 1.0e-197, 1.0e-198, 1.0e-199, + 1.0e-200, 1.0e-201, 1.0e-202, 1.0e-203, 1.0e-204, 1.0e-205, 1.0e-206, 1.0e-207, 1.0e-208, 1.0e-209, + 1.0e-210, 1.0e-211, 1.0e-212, 1.0e-213, 1.0e-214, 1.0e-215, 1.0e-216, 1.0e-217, 1.0e-218, 1.0e-219, + 1.0e-220, 1.0e-221, 1.0e-222, 1.0e-223, 1.0e-224, 1.0e-225, 1.0e-226, 1.0e-227, 1.0e-228, 1.0e-229, + 1.0e-230, 1.0e-231, 1.0e-232, 1.0e-233, 1.0e-234, 1.0e-235, 1.0e-236, 1.0e-237, 1.0e-238, 1.0e-239, + 1.0e-240, 1.0e-241, 1.0e-242, 1.0e-243, 1.0e-244, 1.0e-245, 1.0e-246, 1.0e-247, 1.0e-248, 1.0e-249, + 1.0e-250, 1.0e-251, 1.0e-252, 1.0e-253, 1.0e-254, 1.0e-255, 1.0e-256, 1.0e-257, 1.0e-258, 1.0e-259, + 1.0e-260, 1.0e-261, 1.0e-262, 1.0e-263, 1.0e-264, 1.0e-265, 1.0e-266, 1.0e-267, 1.0e-268, 1.0e-269, + 1.0e-270, 1.0e-271, 1.0e-272, 1.0e-273, 1.0e-274, 1.0e-275, 1.0e-276, 1.0e-277, 1.0e-278, 1.0e-279, + 1.0e-280, 1.0e-281, 1.0e-282, 1.0e-283, 1.0e-284, 1.0e-285, 1.0e-286, 1.0e-287, 1.0e-288, 1.0e-289, + 1.0e-290, 1.0e-291, 1.0e-292, 1.0e-293, 1.0e-294, 1.0e-295, 1.0e-296, 1.0e-297, 1.0e-298, 1.0e-299, + 1.0e-300, 1.0e-301, 1.0e-302, 1.0e-303, 1.0e-304, 1.0e-305, 1.0e-306, 1.0e-307, 1.0e-308 +}; + +const float float_decimal_negpower10[] = { + 1.0e0f, 1.0e-1f, 1.0e-2f, 1.0e-3f, 1.0e-4f, 1.0e-5f, 1.0e-6f, 1.0e-7f, 1.0e-8f, 1.0e-9f, + 1.0e-10f, 1.0e-11f, 1.0e-12f, 1.0e-13f, 1.0e-14f, 1.0e-15f, 1.0e-16f, 1.0e-17f, 1.0e-18f, 1.0e-19f, + 1.0e-20f, 1.0e-21f, 1.0e-22f, 1.0e-23f, 1.0e-24f, 1.0e-25f, 1.0e-26f, 1.0e-27f, 1.0e-28f, 1.0e-29f, + 1.0e-30f, 1.0e-31f, 1.0e-32f, 1.0e-33f, 1.0e-34f, 1.0e-35f, 1.0e-36f, 1.0e-37f, 1.0e-38f +}; + +#endif + +/* scale number by floor(log10(number)) + 1 so that the result is in range [0.1, 1) */ + +#define ieee_double_exponent10(ieee_number) ((int)floor(log10(ieee_number.number)) + 1) +#define ieee_float_exponent10(ieee_number) ((int)floorf(log10f(ieee_number.number)) + 1) // floorf, log10f ? + +#define ieee_double_exp10(ieee_number, exponent10) \ + exponent10 = ieee_double_exponent10(ieee_number); \ + if (exponent10 > 0) { \ + double_negative_exp10(ieee_number.number, -exponent10); \ + ieee_number.fraction = ieee_double_fraction(ieee_number.bits); \ + ieee_number.exponent = ieee_double_exponent(ieee_number.bits); \ + } else if (exponent10 < 0) { \ + double_positive_exp10(ieee_number.number, -exponent10); \ + ieee_number.fraction = ieee_double_fraction(ieee_number.bits); \ + ieee_number.exponent = ieee_double_exponent(ieee_number.bits); \ + } + +#define ieee_float_exp10(ieee_number, exponent10) \ + exponent10 = ieee_float_exponent10(ieee_number); \ + if (exponent10 > 0) { \ + float_negative_exp10(ieee_number.number, -exponent10); \ + ieee_number.fraction = ieee_float_fraction(ieee_number.bits); \ + ieee_number.exponent = ieee_float_exponent(ieee_number.bits); \ + } else if (exponent10 < 0) { \ + float_positive_exp10(ieee_number.number, -exponent10); \ + ieee_number.fraction = ieee_float_fraction(ieee_number.bits); \ + ieee_number.exponent = ieee_float_exponent(ieee_number.bits); \ + } + +#if BINARY_MODF + +/* unhide implicit bit 53, produce 56-bit denormalised fraction (binary exponent already in range [-4, -1]) */ + +#define ieee_double_denormalize(ieee_number) \ + (ieee_number.exponent == IEEE_DOUBLE_MIN_EXPONENT ? (++ieee_number.exponent, 0) : (ieee_number.fraction |= (1ull<<52))), \ + ieee_number.fraction <<= (ieee_number.exponent + 4) + +/* unhide implicit bit 24, produce 27-bit denormalized fraction (binary exponent already in range [-4, -1]) */ + +#define ieee_float_denormalize(ieee_number) \ + (ieee_number.exponent == IEEE_FLOAT_MIN_EXPONENT ? (++ieee_number.exponent, 0) : (ieee_number.fraction |= (1<<23))), \ + ieee_number.fraction <<= (ieee_number.exponent + 4) + +/* turn off significant bits over 56 (integer part), multiply by 10, return new integer part (subsequent decimal digit) */ + +#define ieee_double_binary_fraction(ieee_number) \ + (ieee_number.fraction &= ((1ull<<56) - 1), \ + ieee_number.fraction = (ieee_number.fraction << 1) + (ieee_number.fraction << 3), \ + ieee_number.fraction >> 56) + +/* turn off significant bits over 27 (integer part), multiply by 10, return the integer part (subsequent decimal digit) */ + +#define ieee_float_binary_fraction(ieee_number) \ + (ieee_number.fraction &= ((1<<27) - 1), \ + ieee_number.fraction = (ieee_number.fraction << 1) + (ieee_number.fraction << 3), \ + ieee_number.fraction >> 27) + +#define ieee_double_decimal(ieee_number, exponent10, digits, p) \ + ieee_number_decimal(ieee_double_binary_fraction, ieee_number, exponent10, digits, p) +#define ieee_float_decimal(ieee_number, exponent10, digits, p) \ + ieee_number_decimal(ieee_float_binary_fraction, ieee_number, exponent10, digits, p) +#define ieee_double_decimal_dot(ieee_number, exponent10, digits, p, dot) \ + ieee_number_decimal_dot(ieee_double_binary_fraction, ieee_number, exponent10, digits, p, dot) +#define ieee_float_decimal_dot(ieee_number, exponent10, digits, p, dot) \ + ieee_number_decimal_dot(ieee_float_binary_fraction, ieee_number, exponent10, digits, p, dot) + +#else + +/* generic method */ + +#define ieee_double_decimal_fraction(ieee_number, i) (ieee_number.number = modf(10*ieee_number.number, &i), i) +#define ieee_float_decimal_fraction(ieee_number, i) (ieee_number.number = (float)modf(10*ieee_number.number, &i), i) // ??? + +#define ieee_double_decimal(ieee_number, exponent10, digits, p) \ + ieee_number_decimal(ieee_double_decimal_fraction, ieee_number, exponent10, digits, p) +#define ieee_float_decimal(ieee_number, exponent10, digits, p) \ + ieee_number_decimal(ieee_float_decimal_fraction, ieee_number, exponent10, digits, p) +#define ieee_double_decimal_dot(ieee_number, exponent10, digits, p, dot) \ + ieee_number_decimal_dot(ieee_double_decimal_fraction, ieee_number, exponent10, digits, p, dot) +#define ieee_float_decimal_dot(ieee_number, exponent10, digits, p, dot) \ + ieee_number_decimal_dot(ieee_float_decimal_fraction, ieee_number, exponent10, digits, p, dot) + +#endif + +#define ieee_number_decimal(method, ieee_number, exponent10, digits, p) \ + ieee_double_denormalize(ieee_number); \ + if (ieee_number.sign) *p++ = '-'; \ + if (exponent10 <= 0) \ + for (*p++ = '0', *p++ = RADIX_CHAR; exponent10 && digits; *p++ = '0', ++exponent10, --digits); \ + else \ + { \ + do { *p++ = '0' + (char)method(ieee_number); } while (--exponent10); \ + *p++ = RADIX_CHAR; \ + } \ + for ( ; digits && ieee_number.fraction; --digits) \ + *p++ = '0' + (char)method(ieee_number) + +#define ieee_number_decimal_dot(method, ieee_number, exponent10, digits, p, dot) \ + ieee_double_denormalize(ieee_number); \ + if (ieee_number.sign) *p++ = '-'; \ + if (exponent10 <= 0) \ + { \ + *p++ = '0'; \ + if (dot != NULL) *dot = p; \ + for (*p++ = RADIX_CHAR; exponent10 && digits; *p++ = '0', ++exponent10, --digits); \ + } \ + else \ + { \ + do { *p++ = '0' + (char)method(ieee_number); } while (--exponent10); \ + if (dot != NULL) *dot = p; \ + *p++ = RADIX_CHAR; \ + } \ + for ( ; digits && ieee_number.fraction; --digits) \ + *p++ = '0' + (char)method(ieee_number) + +/* rounding to nearest integer */ + +#if BINARY_MODF +/* check if the mantissa has the most significant bit set, means >= 0.5 */ +# define ieee_double_half(ieee_number) (ieee_number.fraction & (1ull<<55)) +# define ieee_float_half(ieee_number) (ieee_number.fraction & (1<<26)) +#else +# define ieee_double_half(ieee_number) (ieee_number.number >= 0.5) +# define ieee_float_half(ieee_number) (ieee_number.number >= 0.5) +#endif + +/* rounding to nearest integer */ + +#define buffer_ceil(s, p, sign) \ + { \ + while (*--p == '9'); \ + if (*p != RADIX_CHAR) ++*p++; \ + else { \ + char *q; \ + for (q = p - 1; ; --q) { \ + if (*q < '9') { ++*q; break; } \ + *q = '0'; \ + if (q == s) \ + *--s = '1'; \ + else if (sign && q - 1 == s) \ + *s = '1', *--s = '-'; \ + } \ + } \ + } + +#define buffer_remove_trailing_zeros(s, p, sign) \ + { \ + while (*--p == '0'); \ + if (*p != RADIX_CHAR) \ + ++p; \ + else if (!SIGNED_ZERO && sign && p - 2 == s && *(p - 1) == '0') \ + p -= 2, *p++ = '0'; \ + } + +// if digits parameter was initially less then exponent10, then exponent10 > 0 and ieee_double_half(ieee_number) is irrelevant +#define ieee_double_round(ieee_number, exponent10, s, p) \ + if (exponent10 == 0 && ieee_double_half(ieee_number)) \ + { buffer_ceil(s, p, ieee_number.sign); } \ + else \ + { buffer_remove_trailing_zeros(s, p, ieee_number.sign); } + +#define ieee_float_round(ieee_number, exponent10, s, p) \ + if (exponent10 == 0 && ieee_float_half(ieee_number)) \ + { buffer_ceil(s, p, ieee_number.sign); } \ + else \ + { buffer_remove_trailing_zeros(s, p, ieee_number.sign); } + +/* double to decimal */ + +static char number_buffer[512]; + +#define ieee_copy_special_string(special, p, _p) \ + for (p = (char *)number_buffer, _p = special; ; ++p, ++_p) { \ + if ((*p = *_p) == '\0') break; \ + } + +#define ieee_copy_special_string_re(special, p, _p, r, e) \ + for (p = (char *)number_buffer, _p = special; ; ++p, ++_p) { \ + if ((*p = *_p) == '\0') { \ + if (r != NULL) *r = NULL; \ + if (e != NULL) *e = p; \ + break; \ + } \ + } + +char * double_to_string (double number, int digits) +{ + ieee_double ieee_number; + int exponent10; + char *s, *p; const char *_p; + ieee_double_init(ieee_number, number); + ieee_double_sign(ieee_number); + if (ieee_double_is_zero(ieee_number)) // to avoid crash on log10(number) + { + ieee_copy_special_string(ieee_double_zero_string(ieee_number), p, _p); + return (char *)number_buffer; + } + if (ieee_double_special_case(ieee_number)) + { + ieee_copy_special_string(ieee_double_special_string(ieee_number), p, _p); + return (char *)number_buffer; + } + s = p = number_buffer + 1; + ieee_double_exp10(ieee_number, exponent10); + ieee_double_decimal(ieee_number, exponent10, digits, p); + ieee_double_round(ieee_number, exponent10, s, p); + *p = '\0'; + return s; +} + +char * double_as_string (double number, int digits, char **r, char **e) +{ + ieee_double ieee_number; + int exponent10; + char *s, *p; const char *_p; + s = p = number_buffer + 1; + ieee_double_init(ieee_number, number); + ieee_double_sign(ieee_number); + if (ieee_double_is_zero(ieee_number)) // to avoid crash on log10(number) + { + ieee_copy_special_string_re(ieee_double_zero_string(ieee_number), p, _p, r, e); + return (char *)number_buffer; + } + if (ieee_double_special_case(ieee_number)) + { + ieee_copy_special_string_re(ieee_double_special_string(ieee_number), p, _p, r, e); + return (char *)number_buffer; + } + ieee_double_exp10(ieee_number, exponent10); + ieee_double_decimal_dot(ieee_number, exponent10, digits, p, r); + ieee_double_round(ieee_number, exponent10, s, p); + if (e != NULL) *e = p; + *p = '\0'; + return s; +} + +/* float to decimal */ + +char * float_to_string (float number, int digits) +{ + ieee_float ieee_number; + int exponent10; + char *s, *p; const char *_p; + ieee_float_init(ieee_number, number); + ieee_float_sign(ieee_number); + if (ieee_float_is_zero(ieee_number)) + { + ieee_copy_special_string(ieee_float_zero_string(ieee_number), p, _p); + return (char *)number_buffer; + } + if (ieee_float_special_case(ieee_number)) + { + ieee_copy_special_string(ieee_float_special_string(ieee_number), p, _p); + return (char *)number_buffer; + } + s = p = number_buffer + 1; + ieee_float_exp10(ieee_number, exponent10); + ieee_float_decimal(ieee_number, exponent10, digits, p); + ieee_float_round(ieee_number, exponent10, s, p); + *p = '\0'; + return s; +} + +char * float_as_string (float number, int digits, char **r, char **e) +{ + ieee_float ieee_number; + int exponent10; + char *s, *p; const char *_p; + s = p = number_buffer + 1; + ieee_float_init(ieee_number, number); + ieee_float_sign(ieee_number); + if (ieee_float_is_zero(ieee_number)) + { + ieee_copy_special_string_re(ieee_float_zero_string(ieee_number), p, _p, r, e); + return (char *)number_buffer; + } + if (ieee_float_special_case(ieee_number)) + { + ieee_copy_special_string_re(ieee_float_special_string(ieee_number), p, _p, r, e); + return (char *)number_buffer; + } + ieee_float_exp10(ieee_number, exponent10); + ieee_float_decimal_dot(ieee_number, exponent10, digits, p, r); + ieee_float_round(ieee_number, exponent10, s, p); + if (e != NULL) *e = p; + *p = '\0'; + return s; +} + +/* decimal string to double/float */ + +#define string_scan_decimal(s, c, number) _scan_decimal(c, number, *++s) +#define string_scan_fraction(s, c, number, exponent10) _scan_fraction(c, number, exponent10, *++s) +#define string_scan_exponent10(s, c, exponent10) _scan_exponent10(c, exponent10, *++s) + +const char * string_to_double (const char *s, double *number) +{ + int sign, exponent10, c = *s; + string_scan_sign(s, c, sign); + string_scan_decimal(s, c, *number); + if (c == '.') + { + c = *++s; + string_scan_fraction(s, c, *number, exponent10); + } + else + exponent10 = 0; + if (c == 'e' || c == 'E') + { + c = *++s; + string_scan_exponent10(s, c, exponent10); + } + double_exp10(*number, exponent10); + if (sign) *number = -*number; + return s; +} + +const char * string_to_float (const char *s, float *number) +{ + int sign, exponent10, c = *s; + string_scan_sign(s, c, sign); + string_scan_decimal(s, c, *number); + if (c == '.') + { + c = *++s; + string_scan_fraction(s, c, *number, exponent10); + } + else + exponent10 = 0; + if (c == 'e' || c == 'E') + { + c = *++s; + string_scan_exponent10(s, c, exponent10); + } + float_exp10(*number, exponent10); + if (sign) *number = -*number; + return s; +} + +/* conventional form */ + +const char * convert_to_double (const char *s, double *number) +{ + int sign, c = *s; + string_scan_sign(s, c, sign); + string_scan_decimal(s, c, *number); + if (c == '.' || c == ',') + { + int exponent10; + c = *++s; + string_scan_fraction(s, c, *number, exponent10); + if (exponent10 < 0) + double_negative_exp10(*number, exponent10); + } + if (sign) *number = -*number; + return s; +} + +const char * convert_to_float (const char *s, float *number) +{ + int sign, c = *s; + string_scan_sign(s, c, sign); + string_scan_decimal(s, c, *number); + if (c == '.' || c == ',') + { + int exponent10; + c = *++s; + string_scan_fraction(s, c, *number, exponent10); + if (exponent10 < 0) + float_negative_exp10(*number, exponent10); + } + if (sign) *number = -*number; + return s; +} + +/* pretty common stuff */ + +size_t bytes_to_hex_lc (const void *input, size_t size, unsigned char *output) +{ + size_t i; + const unsigned char *p; + for (i = 0, p = (const unsigned char *)input; i < size; ++i, ++p) + { + *output++ = base16_lc_digit1(*p); + *output++ = base16_lc_digit2(*p); + } + *output = '\0'; + return 2*size + 1; +} + +size_t bytes_to_hex_uc (const void *input, size_t size, unsigned char *output) +{ + size_t i; + const unsigned char *p; + for (i = 0, p = (const unsigned char *)input; i < size; ++i, ++p) + { + *output++ = base16_uc_digit1(*p); + *output++ = base16_uc_digit2(*p); + } + *output = '\0'; + return 2*size + 1; +} + +size_t hex_to_bytes (const void *input, size_t size, unsigned char *output) +{ + size_t i; + int c1, c2; + const unsigned char *p; + for (i = 1, p = (const unsigned char *)input; i < size; i += 2) + { + c1 = base16_value(*p); + ++p; + c2 = base16_value(*p); + ++p; + if (c1 >= 0 && c2 >= 0) + *output++ = (unsigned char)((c1<<4)|c2); + else + break; + } + return i >> 1; +} + +void print_as_hex (const void *input, size_t bytes) +{ + const unsigned char *p; + for (p = (const unsigned char *)input; bytes > 0; --bytes, ++p) + printf("%02x", *p); +}
\ No newline at end of file diff --git a/Build/source/texk/web2c/luatexdir/luapplib/util/utilnumber.h b/Build/source/texk/web2c/luatexdir/luapplib/util/utilnumber.h new file mode 100644 index 00000000000..3c96571b5ce --- /dev/null +++ b/Build/source/texk/web2c/luatexdir/luapplib/util/utilnumber.h @@ -0,0 +1,354 @@ +#ifndef UTIL_NUMBER_H +#define UTIL_NUMBER_H + +#include <stddef.h> // for size_t + +#include "utilplat.h" +#include "utildecl.h" + +/* since 'long' isn't long for msvc64/mingw64, we need a type for machine word */ + +#if !defined(__cplusplus) || !defined(_MSC_VER) +# include <stdint.h> // int*_t types are in standard in msvc++ +#endif + +//#if defined(MSVC64) || defined(__MINGW64__) || defined(__x86_64__) || UINTPTR_MAX > 0xffffffff +//# define BIT64 +//#else +//# define BIT64 +//#endif + +#if defined(_WIN64) || defined(__MINGW32__) +# define INT64F "%I64d" +# define UINT64F "%I64u" +#else +# define INT64F "%lld" +# define UINT64F "%llu" +#endif + +#if defined(MSVC64) +# define intlw_t int64_t +# define uintlw_t uint64_t +# define INTLW(N) N##I64 +# define UINTLW(N) N##UI64 +# define INTLWF INT64F +# define UINTLWF UINT64F +#elif defined(__MINGW64__) +# define intlw_t int64_t +# define uintlw_t uint64_t +# define INTLW(N) N##LL +# define UINTLW(N) N##ULL +# define INTLWF INT64F +# define UINTLWF UINT64F +#else // 32bit or sane 64bit (LP64) +# define intlw_t long +# define uintlw_t size_t /*unsigned long*/ +# define INTLW(N) N##L +# define UINTLW(N) N##UL +# define INTLWF "%ld" +# define UINTLWF "%lu" +#endif + +/* basic constants */ + +#define MAX_RADIX 36 +// #define MAX_INTEGER_DIGITS 65 /* 64-bit number in binary form plus '\0' */ +#define MAX_INTEGER_DIGITS 128 // to handle romannumeral of short int + +#define base36_uc_alphabet "0123456789ABCDEFGHIJKLMNOPQRSTUVWXYZ" +#define base36_lc_alphabet "0123456789abcdefghijklmnopqrstuvwxyz" + +#define base26_uc_alphabet "ABCDEFGHIJKLMNOPQRSTUVWXYZ" +#define base26_lc_alphabet "abcdefghijklmnopqrstuvwxyz" +extern const int base26_lookup[]; + +#define base36_lc_palindrome "zyxwvutsrqponmlkjihgfedcba9876543210123456789abcdefghijklmnopqrstuvwxyz" +#define base36_uc_palindrome "ZYXWVUTSRQPONMLKJIHGFEDCBA9876543210123456789ABCDEFGHIJKLMNOPQRSTUVWXYZ" + +extern const int base36_lookup[]; + +#define base10_palindrome "9876543210123456789" +#define base10_alphabet "0123456789" +extern const int base10_lookup[]; + +#define base16_uc_alphabet "0123456789ABCDEF" +#define base16_lc_alphabet "0123456789abcdef" +extern const int base16_lookup[]; + +#define base16_uc_digit1(c) base16_uc_alphabet[(c)>>4] +#define base16_uc_digit2(c) base16_uc_alphabet[(c)&15] +#define base16_lc_digit1(c) base16_lc_alphabet[(c)>>4] +#define base16_lc_digit2(c) base16_lc_alphabet[(c)&15] + +#define base8_digit(c) ((unsigned)(c - '0') <= (unsigned)('7' - '0')) +#define base8_value(c) (base8_digit(c) ? (c) - '0' : -1) + +#define base10_digit(c) ((unsigned)(c - '0') <= (unsigned)('9' - '0')) +#define base10_value(c) (base10_lookup[(uint8_t)c]) + +#define base16_digit(c) (base16_lookup[(uint8_t)c] >= 0) +#define base16_value(c) (base16_lookup[(uint8_t)c]) + +#define base26_digit(c) (base26_lookup[(uint8_t)c] >= 0) +#define base26_value(c) (base26_lookup[(uint8_t)c]) + +#define base36_digit(c) (base36_lookup[(uint8_t)c] >= 0) +#define base36_value(c) (base36_lookup[(uint8_t)c]) + +//#define base_digit(c, radix) ((unsigned)(base36_lookup[c]) < (unsigned)(radix)) +//#define base_value(c, radix) (base_digit(c, radix) ? base36_lookup[c] : -1) + +/* integer from string; return a pointer to character next to the last digit */ + +UTILAPI const char * string_to_int32 (const char *s, int32_t *number); +UTILAPI const char * string_to_intlw (const char *s, intlw_t *number); +UTILAPI const char * string_to_int64 (const char *s, int64_t *number); + +UTILAPI const char * string_to_uint32 (const char *s, uint32_t *number); +UTILAPI const char * string_to_uintlw (const char *s, uintlw_t *number); +UTILAPI const char * string_to_uint64 (const char *s, uint64_t *number); + +UTILAPI const char * radix_to_int32 (const char *s, int32_t *number, int radix); +UTILAPI const char * radix_to_intlw (const char *s, intlw_t *number, int radix); +UTILAPI const char * radix_to_int64 (const char *s, int64_t *number, int radix); + +UTILAPI const char * radix_to_uint32 (const char *s, uint32_t *number, int radix); +UTILAPI const char * radix_to_uintlw (const char *s, uintlw_t *number, int radix); +UTILAPI const char * radix_to_uint64 (const char *s, uint64_t *number, int radix); + +UTILAPI const char * roman_to_uint16 (const char *s, uint16_t *number); + +UTILAPI const char * alpha_to_uint32 (const char *s, uint32_t *number); +UTILAPI const char * alpha_to_uintlw (const char *s, uintlw_t *number); +UTILAPI const char * alpha_to_uint64 (const char *s, uint64_t *number); + +UTILAPI const char * alphan_to_uint32 (const char *s, uint32_t *number); +UTILAPI const char * alphan_to_uintlw (const char *s, uintlw_t *number); +UTILAPI const char * alphan_to_uint64 (const char *s, uintlw_t *number); + +/* +integer to string; return a pointer to null-terminated static const string +same but also stores pointer to trailing null (to be used for firther formatting) +*/ + +UTILAPI char * int32_as_string (int32_t number, char **e); +UTILAPI char * intlw_as_string (intlw_t number, char **e); +UTILAPI char * int64_as_string (int64_t number, char **e); + +UTILAPI char * uint32_as_string (uint32_t number, char **e); +UTILAPI char * uintlw_as_string (uintlw_t number, char **e); +UTILAPI char * uint64_as_string (uint64_t number, char **e); + +#define int32_to_string(number) int32_as_string(number, NULL) +#define intlw_to_string(number) intlw_as_string(number, NULL) +#define int64_to_string(number) int64_as_string(number, NULL) + +#define uint32_to_string(number) uint32_as_string(number, NULL) +#define uintlw_to_string(number) uintlw_as_string(number, NULL) +#define uint64_to_string(number) uint64_as_string(number, NULL) + +UTILAPI char * int32_as_radix (int32_t number, int radix, char **e); +UTILAPI char * intlw_as_radix (intlw_t number, int radix, char **e); +UTILAPI char * int64_as_radix (int64_t number, int radix, char **e); + +UTILAPI char * uint32_as_radix (uint32_t number, int radix, char **e); +UTILAPI char * uintlw_as_radix (uintlw_t number, int radix, char **e); +UTILAPI char * uint64_as_radix (uint64_t number, int radix, char **e); + +#define int32_to_radix(number, radix) int32_as_radix(number, radix, NULL) +#define intlw_to_radix(number, radix) intlw_as_radix(number, radix, NULL) +#define int64_to_radix(number, radix) int64_as_radix(number, radix, NULL) + +#define uint32_to_radix(number, radix) uint32_as_radix(number, radix, NULL) +#define uintlw_to_radix(number, radix) uintlw_as_radix(number, radix, NULL) +#define uint64_to_radix(number, radix) uint64_as_radix(number, radix, NULL) + +UTILAPI char * uint32_as_alpha_uc (uint32_t number, char **e); +UTILAPI char * uint32_as_alpha_lc (uint32_t number, char **e); +UTILAPI char * uintlw_as_alpha_uc (uintlw_t number, char **e); +UTILAPI char * uintlw_as_alpha_lc (uintlw_t number, char **e); +UTILAPI char * uint64_as_alpha_uc (uint64_t number, char **e); +UTILAPI char * uint64_as_alpha_lc (uint64_t number, char **e); + +#define uint32_to_alpha_uc(number) uint32_as_alpha_uc(number, NULL) +#define uint32_to_alpha_lc(number) uint32_as_alpha_lc(number, NULL) +#define uintlw_to_alpha_uc(number) uintlw_as_alpha_uc(number, NULL) +#define uintlw_to_alpha_lc(number) uintlw_as_alpha_lc(number, NULL) +#define uint64_to_alpha_uc(number) uint64_as_alpha_uc(number, NULL) +#define uint64_to_alpha_lc(number) uint64_as_alpha_lc(number, NULL) + +UTILAPI char * uint32_as_alphan_uc (uint32_t number, char **e); +UTILAPI char * uint32_as_alphan_lc (uint32_t number, char **e); +UTILAPI char * uintlw_as_alphan_uc (uintlw_t number, char **e); +UTILAPI char * uintlw_as_alphan_lc (uintlw_t number, char **e); +UTILAPI char * uint64_as_alphan_uc (uint64_t number, char **e); +UTILAPI char * uint64_as_alphan_lc (uint64_t number, char **e); + +#define uint32_to_alphan_uc(number) uint32_as_alpha_uc(number, NULL) +#define uint32_to_alphan_lc(number) uint32_as_alpha_lc(number, NULL) +#define uintlw_to_alphan_uc(number) uintlw_as_alpha_uc(number, NULL) +#define uintlw_to_alphan_lc(number) uintlw_as_alpha_lc(number, NULL) +#define uint64_to_alphan_uc(number) uint64_as_alpha_uc(number, NULL) +#define uint64_to_alphan_lc(number) uint64_as_alpha_lc(number, NULL) + +/* roman numeral (limited to uint16_t) */ + +UTILAPI char * uint16_as_roman_uc (uint16_t number, char **e); +UTILAPI char * uint16_as_roman_lc (uint16_t number, char **e); + +#define uint16_to_roman_uc(number) uint16_as_roman_uc(number, NULL) +#define uint16_to_roman_lc(number) uint16_as_roman_lc(number, NULL) + +#define uint16_as_roman(number) uint16_as_roman_uc(number) +#define uint16_to_roman(number) uint16_to_roman_uc(number) + +/* double/float to string */ + +UTILAPI char * double_to_string (double number, int digits); +UTILAPI char * float_to_string (float number, int digits); + +UTILAPI char * double_as_string (double number, int digits, char **r, char **e); +UTILAPI char * float_as_string (float number, int digits, char **r, char **e); + +/* string to double/float */ + +UTILAPI const char * string_to_double (const char *s, double *number); +UTILAPI const char * string_to_float (const char *s, float *number); + +/* convenience form accepting comma among a dot, with not exp notation (eg. pdf) */ + +UTILAPI const char * convert_to_double (const char *s, double *number); +UTILAPI const char * convert_to_float (const char *s, float *number); + +/* binary data parsers helpers */ + +#define get_byte1(i) ((i)&255) +#define get_byte2(i) (((i)>>8)&255) +#define get_byte3(i) (((i)>>16)&255) +#define get_byte4(i) (((i)>>24)&255) + +#define read_uint16be_as(s, int_type) ((int_type)((s[0]<<8)|s[1])) +#define read_uint32be_as(s, int_type) ((int_type)((s[0]<<24)|(s[1]<<16)|(s[2]<<8)|s[3])) + +#define read_uint16le_as(s, int_type) ((int_type)((s[1]<<8)|s[0])) +#define read_uint32le_as(s, int_type) ((int_type)((s[3]<<24)|(s[2]<<16)|(s[1]<<8)|s[0])) + +#define read_uint16_native(s) (*((uint16_t *)(s))) +#define read_uint32_native(s) (*((uint32_t *)(s))) +#define read_int16_native(s) (*((int16_t *)(s))) +#define read_int32_native(s) (*((int32_t *)(s))) + +#define scan_uint16be_as(s, int_type) (s += 2, (int_type)((s[-2]<<8)|s[-1])) +#define scan_uint32be_as(s, int_type) (s += 4, (int_type)((s[-4]<<24)|(s[-3]<<16)|(s[-2]<<8)|s[-1])) + +#define scan_uint16le_as(s, int_type) (s += 2, (int_type)((s[-1]<<8)|s[-2])) +#define scan_uint32le_as(s, int_type) (s += 4, (int_type)((s[-1]<<24)|(s[-2]<<16)|(s[-3]<<8)|s[-4])) + +#define scan_uint16_native(s) (s += 2, read_uint16_native(s-2)) +#define scan_uint32_native(s) (s += 4, read_uint32_native(s-4)) +#define scan_int16_native(s) (s += 2, read_int16_native(s-2)) +#define scan_int32_native(s) (s += 4, read_int32_native(s-4)) + +#define read_fixed16_16_as(s, float_type) (((float_type)read_uint32be_as(s, signed int))/(1<<16)) +#define read_fixed2_14_as(s, float_type) (((float_type)read_uint16be_as(s, signed short))/(1<<14)) + +#define scan_fixed16_16_as(s, float_type) (((float_type)scan_uint32be_as(s, signed int))/(1<<16)) +#define scan_fixed2_14_as(s, float_type) (((float_type)scan_uint16be_as(s, signed short))/(1<<14)) + +/* internal procedures */ + +#define _scan_sign(c, sign, next) \ + do { if (c == '-') { sign = 1; c = next; } else if (c == '+') { sign = 0; c = next; } else sign = 0; } while (0) + +#define integer_multiplied10(number) (((number) << 1) + ((number) << 3)) + +#define _scan_integer(c, number, next) \ + for (number = 0; base10_digit(c); number = integer_multiplied10(number) + (c - '0'), c = next) +#define _scan_radix(c, number, radix, next) \ + for (number = 0; (c = base36_value(c)) >= 0 && c < radix; number = number * radix + c, c = next) + +#define _read_integer(c, number, next) \ + for (number = c - '0', c = next; base10_digit(c); number = integer_multiplied10(number) + (c - '0'), c = next) +#define _read_radix(c, number, radix, next) \ + for (number = c - '0', c = next; (c = base36_value(c)) >= 0 && c < radix; number = number * radix + c, c = next) + +/* rationals */ + +#define _scan_decimal(c, number, next) \ + for (number = 0; base10_digit(c); number = number*10 + (c - '0'), c = next) +#define _scan_fraction(c, number, exponent10, next) \ + for (exponent10 = 0; base10_digit(c); --exponent10, number = number*10 + (c - '0'), c = next) + +#define _scan_exponent10(c, exponent10, next) \ + do { \ + int eexponent10, eexpsign; \ + _scan_sign(c, eexpsign, next); \ + _scan_integer(c, eexponent10, next); \ + if (eexpsign) \ + exponent10 -= eexponent10; \ + else \ + exponent10 += eexponent10; \ + } while(0) + +#if 0 + +// kept just for sentiment ;) + +extern const double double_binary_power10[]; +extern const float float_binary_power10[]; +extern const double double_binary_negpower10[]; +extern const float float_binary_negpower10[]; + +#define double_negative_exp10(number, exponent) \ +{ const double *bp10; int e = ((exponent) < 511 ? 511 : -(exponent)); \ + for (bp10 = double_binary_negpower10; e > 0; e >>= 1, ++bp10) \ + if (e & 1) number *= *bp10; } + +#define float_negative_exp10(number, exponent) \ +{ const float *bp10; int e = ((exponent) < 64 ? 64 : -(exponent)); \ + for (bp10 = float_binary_negpower10; e > 0; e >>= 1, ++bp10) \ + if (e & 1) number *= *bp10; } + +#define double_positive_exp10(number, exponent) \ +{ const double *bp10; int e = ((exponent) > 511 ? 511 : (exponent)); \ + for (bp10 = double_binary_power10; e > 0; e >>= 1, ++bp10) \ + if (e & 1) number *= *bp10; } + +#define float_positive_exp10(number, exponent) \ +{ const float *bp10; int e = ((exponent) > 64 ? 64 : (exponent)); \ + for (bp10 = double_binary_power10; e > 0; e >>= 1, ++bp10) \ + if (e & 1) number *= *bp10; } + +#define double_exp10(number, exponent) \ + if ((exponent) < 0) double_negative_exp10(number, exponent) else if ((exponent) > 0) double_positive_exp10(number, exponent) + +#define float_exp10(number, exponent) \ + if ((exponent) < 0) float_negative_exp10(number, exponent) else if ((exponent) > 0) float_positive_exp10(number, exponent) + +#else + +extern const double double_decimal_power10[]; +extern const float float_decimal_power10[]; +extern const double double_decimal_negpower10[]; +extern const float float_decimal_negpower10[]; + +#define double_negative_exp10(number, exponent) ((number) *= double_decimal_negpower10[(exponent) < -308 ? 308 : -(exponent)]) +#define double_positive_exp10(number, exponent) ((number) *= double_decimal_power10[(exponent) > 308 ? 308 : (exponent)]) + +#define float_negative_exp10(number, exponent) ((number) *= float_decimal_negpower10[(exponent) < -38 ? 38 : -(exponent)]) +#define float_positive_exp10(number, exponent) ((number) *= float_decimal_power10[(exponent) > 38 ? 38 : (exponent)]) + +#define double_exp10(number, exponent) ((void)(((exponent) < 0 && double_negative_exp10(number, exponent)) || (((exponent) > 0 && double_positive_exp10(number, exponent))))) +#define float_exp10(number, exponent) ((void)(((exponent) < 0 && float_negative_exp10(number, exponent)) || (((exponent) > 0 && float_positive_exp10(number, exponent))))) + +#endif + +/* pretty common stuff */ + +#define bytes_to_hex(input, size, output) bytes_to_hex_lc(input, size, output) +UTILAPI size_t bytes_to_hex_lc (const void *input, size_t size, uint8_t *output); +UTILAPI size_t bytes_to_hex_uc (const void *input, size_t size, uint8_t *output); +UTILAPI size_t hex_to_bytes (const void *input, size_t size, uint8_t *output); +UTILAPI void print_as_hex (const void *input, size_t bytes); + +#endif
\ No newline at end of file diff --git a/Build/source/texk/web2c/luatexdir/luapplib/util/utilplat.h b/Build/source/texk/web2c/luatexdir/luapplib/util/utilplat.h new file mode 100644 index 00000000000..037a1e23199 --- /dev/null +++ b/Build/source/texk/web2c/luatexdir/luapplib/util/utilplat.h @@ -0,0 +1,23 @@ + +#ifndef UTIL_PLAT_H +#define UTIL_PLAT_H + +#if defined(_WIN32) || defined(WIN32) +# ifdef _MSC_VER +# if defined(_M_64) || defined(_WIN64) +# define MSVC64 +# else +# define MSVC32 +# endif +# else +# if defined(__MINGW64__) +# define MINGW64 +# else +# if defined(__MINGW32__) +# define MINGW32 +# endif +# endif +# endif +#endif + +#endif
\ No newline at end of file diff --git a/Build/source/texk/web2c/luatexdir/luapplib/util/utilsha.c b/Build/source/texk/web2c/luatexdir/luapplib/util/utilsha.c new file mode 100644 index 00000000000..66532073023 --- /dev/null +++ b/Build/source/texk/web2c/luatexdir/luapplib/util/utilsha.c @@ -0,0 +1,1290 @@ +/* sha2 implementation by Aaron D. Gifford (http://www.aarongifford.com) */ + +#if defined __arm__ || defined __ARM__ || defined ARM || defined __ARM || defined __arm || defined __ARM_ARCH ||defined __aarch64__ +#include <stdlib.h> +#endif + +#ifndef BYTE_ORDER +#define BYTE_ORDER LITTLE_ENDIAN +#endif +/* begin of sha2.c */ + +/* + * FILE: sha2.c + * AUTHOR: Aaron D. Gifford - http://www.aarongifford.com/ + * + * Copyright (c) 2000-2001, Aaron D. Gifford + * All rights reserved. + * + * Redistribution and use in source and binary forms, with or without + * modification, are permitted provided that the following conditions + * are met: + * 1. Redistributions of source code must retain the above copyright + * notice, this list of conditions and the following disclaimer. + * 2. Redistributions in binary form must reproduce the above copyright + * notice, this list of conditions and the following disclaimer in the + * documentation and/or other materials provided with the distribution. + * 3. Neither the name of the copyright holder nor the names of contributors + * may be used to endorse or promote products derived from this software + * without specific prior written permission. + * + * THIS SOFTWARE IS PROVIDED BY THE AUTHOR AND CONTRIBUTOR(S) ``AS IS'' AND + * ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE + * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE + * ARE DISCLAIMED. IN NO EVENT SHALL THE AUTHOR OR CONTRIBUTOR(S) BE LIABLE + * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL + * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS + * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) + * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT + * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY + * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF + * SUCH DAMAGE. + * + * $Id: sha2.c,v 1.1 2001/11/08 00:01:51 adg Exp adg $ + */ + +#include <string.h> /* memcpy()/memset() or bcopy()/bzero() */ +#if defined __arm__ || defined __ARM__ || defined ARM || defined __ARM || defined __arm || defined __ARM_ARCH ||defined __aarch64__ +#include <stdlib.h> +#endif +#include <assert.h> /* assert() */ +//#include "sha2.h" +#include "utilsha.h" + +/* + * ASSERT NOTE: + * Some sanity checking code is included using assert(). On my FreeBSD + * system, this additional code can be removed by compiling with NDEBUG + * defined. Check your own systems manpage on assert() to see how to + * compile WITHOUT the sanity checking code on your system. + * + * UNROLLED TRANSFORM LOOP NOTE: + * You can define SHA2_UNROLL_TRANSFORM to use the unrolled transform + * loop version for the hash transform rounds (defined using macros + * later in this file). Either define on the command line, for example: + * + * cc -DSHA2_UNROLL_TRANSFORM -o sha2 sha2.c sha2prog.c + * + * or define below: + * + * #define SHA2_UNROLL_TRANSFORM + * + */ + + +/*** SHA-256/384/512 Machine Architecture Definitions *****************/ +/* + * BYTE_ORDER NOTE: + * + * Please make sure that your system defines BYTE_ORDER. If your + * architecture is little-endian, make sure it also defines + * LITTLE_ENDIAN and that the two (BYTE_ORDER and LITTLE_ENDIAN) are + * equivilent. + * + * If your system does not define the above, then you can do so by + * hand like this: + * + * #define LITTLE_ENDIAN 1234 + * #define BIG_ENDIAN 4321 + * + * And for little-endian machines, add: + * + * #define BYTE_ORDER LITTLE_ENDIAN + * + * Or for big-endian machines: + * + * #define BYTE_ORDER BIG_ENDIAN + * + * The FreeBSD machine this was written on defines BYTE_ORDER + * appropriately by including <sys/types.h> (which in turn includes + * <machine/endian.h> where the appropriate definitions are actually + * made). + */ +#if !defined(BYTE_ORDER) || (BYTE_ORDER != LITTLE_ENDIAN && BYTE_ORDER != BIG_ENDIAN) +#error Define BYTE_ORDER to be equal to either LITTLE_ENDIAN or BIG_ENDIAN +#endif + +/* + * Define the followingsha2_* types to types of the correct length on + * the native archtecture. Most BSD systems and Linux define u_intXX_t + * types. Machines with very recent ANSI C headers, can use the + * uintXX_t definintions from inttypes.h by defining SHA2_USE_INTTYPES_H + * during compile or in the sha.h header file. + * + * Machines that support neither u_intXX_t nor inttypes.h's uintXX_t + * will need to define these three typedefs below (and the appropriate + * ones in sha.h too) by hand according to their system architecture. + * + * Thank you, Jun-ichiro itojun Hagino, for suggesting using u_intXX_t + * types and pointing out recent ANSI C support for uintXX_t in inttypes.h. + * + * PJ: replace by uintX_t + */ + +//typedef uint8_t sha2_byte; /* Exactly 1 byte */ +//typedef uint32_t sha2_word32; /* Exactly 4 bytes */ +//typedef uint64_t sha2_word64; /* Exactly 8 bytes */ + +/*** SHA-256/384/512 Various Length Definitions ***********************/ +/* NOTE: Most of these are in sha2.h */ +#define SHA256_SHORT_BLOCK_LENGTH (SHA256_BLOCK_LENGTH - 8) +#define SHA384_SHORT_BLOCK_LENGTH (SHA384_BLOCK_LENGTH - 16) +#define SHA512_SHORT_BLOCK_LENGTH (SHA512_BLOCK_LENGTH - 16) + + +/*** ENDIAN REVERSAL MACROS *******************************************/ +#if BYTE_ORDER == LITTLE_ENDIAN +#define REVERSE32(w,x) { \ + uint32_t tmp = (w); \ + tmp = (tmp >> 16) | (tmp << 16); \ + (x) = ((tmp & 0xff00ff00UL) >> 8) | ((tmp & 0x00ff00ffUL) << 8); \ +} +#define REVERSE64(w,x) { \ + uint64_t tmp = (w); \ + tmp = (tmp >> 32) | (tmp << 32); \ + tmp = ((tmp & 0xff00ff00ff00ff00ULL) >> 8) | \ + ((tmp & 0x00ff00ff00ff00ffULL) << 8); \ + (x) = ((tmp & 0xffff0000ffff0000ULL) >> 16) | \ + ((tmp & 0x0000ffff0000ffffULL) << 16); \ +} +#endif /* BYTE_ORDER == LITTLE_ENDIAN */ + +/* + * Macro for incrementally adding the unsigned 64-bit integer n to the + * unsigned 128-bit integer (represented using a two-element array of + * 64-bit words): + */ +#define ADDINC128(w,n) { \ + (w)[0] += (uint64_t)(n); \ + if ((w)[0] < (n)) { \ + (w)[1]++; \ + } \ +} + +/* + * Macros for copying blocks of memory and for zeroing out ranges + * of memory. Using these macros makes it easy to switch from + * using memset()/memcpy() and using bzero()/bcopy(). + * + * Please define either SHA2_USE_MEMSET_MEMCPY or define + * SHA2_USE_BZERO_BCOPY depending on which function set you + * choose to use: + */ +#if !defined(SHA2_USE_MEMSET_MEMCPY) && !defined(SHA2_USE_BZERO_BCOPY) +/* Default to memset()/memcpy() if no option is specified */ +#define SHA2_USE_MEMSET_MEMCPY 1 +#endif +#if defined(SHA2_USE_MEMSET_MEMCPY) && defined(SHA2_USE_BZERO_BCOPY) +/* Abort with an error if BOTH options are defined */ +#error Define either SHA2_USE_MEMSET_MEMCPY or SHA2_USE_BZERO_BCOPY, not both! +#endif + +#ifdef SHA2_USE_MEMSET_MEMCPY +#define MEMSET_BZERO(p,l) memset((p), 0, (l)) +#define MEMCPY_BCOPY(d,s,l) memcpy((d), (s), (l)) +#endif +#ifdef SHA2_USE_BZERO_BCOPY +#define MEMSET_BZERO(p,l) bzero((p), (l)) +#define MEMCPY_BCOPY(d,s,l) bcopy((s), (d), (l)) +#endif + + +/*** THE SIX LOGICAL FUNCTIONS ****************************************/ +/* + * Bit shifting and rotation (used by the six SHA-XYZ logical functions: + * + * NOTE: The naming of R and S appears backwards here (R is a SHIFT and + * S is a ROTATION) because the SHA-256/384/512 description document + * (see http://csrc.nist.gov/cryptval/shs/sha256-384-512.pdf) uses this + * same "backwards" definition. + */ +/* Shift-right (used in SHA-256, SHA-384, and SHA-512): */ +#define R(b,x) ((x) >> (b)) +/* 32-bit Rotate-right (used in SHA-256): */ +#define S32(b,x) (((x) >> (b)) | ((x) << (32 - (b)))) +/* 64-bit Rotate-right (used in SHA-384 and SHA-512): */ +#define S64(b,x) (((x) >> (b)) | ((x) << (64 - (b)))) + +/* Two of six logical functions used in SHA-256, SHA-384, and SHA-512: */ +#define Ch(x,y,z) (((x) & (y)) ^ ((~(x)) & (z))) +#define Maj(x,y,z) (((x) & (y)) ^ ((x) & (z)) ^ ((y) & (z))) + +/* Four of six logical functions used in SHA-256: */ +#define Sigma0_256(x) (S32(2, (x)) ^ S32(13, (x)) ^ S32(22, (x))) +#define Sigma1_256(x) (S32(6, (x)) ^ S32(11, (x)) ^ S32(25, (x))) +#define sigma0_256(x) (S32(7, (x)) ^ S32(18, (x)) ^ R(3 , (x))) +#define sigma1_256(x) (S32(17, (x)) ^ S32(19, (x)) ^ R(10, (x))) + +/* Four of six logical functions used in SHA-384 and SHA-512: */ +#define Sigma0_512(x) (S64(28, (x)) ^ S64(34, (x)) ^ S64(39, (x))) +#define Sigma1_512(x) (S64(14, (x)) ^ S64(18, (x)) ^ S64(41, (x))) +#define sigma0_512(x) (S64( 1, (x)) ^ S64( 8, (x)) ^ R( 7, (x))) +#define sigma1_512(x) (S64(19, (x)) ^ S64(61, (x)) ^ R( 6, (x))) + +/*** INTERNAL FUNCTION PROTOTYPES *************************************/ +/* NOTE: These should not be accessed directly from outside this + * library -- they are intended for private internal visibility/use + * only. + */ +static void SHA512_Last(SHA512_CTX*); +static void SHA256_Transform(SHA256_CTX*, const uint32_t*); +static void SHA512_Transform(SHA512_CTX*, const uint64_t*); + + +/*** SHA-XYZ INITIAL HASH VALUES AND CONSTANTS ************************/ +/* Hash constant words K for SHA-256: */ +static const uint32_t K256[64] = { + 0x428a2f98UL, 0x71374491UL, 0xb5c0fbcfUL, 0xe9b5dba5UL, + 0x3956c25bUL, 0x59f111f1UL, 0x923f82a4UL, 0xab1c5ed5UL, + 0xd807aa98UL, 0x12835b01UL, 0x243185beUL, 0x550c7dc3UL, + 0x72be5d74UL, 0x80deb1feUL, 0x9bdc06a7UL, 0xc19bf174UL, + 0xe49b69c1UL, 0xefbe4786UL, 0x0fc19dc6UL, 0x240ca1ccUL, + 0x2de92c6fUL, 0x4a7484aaUL, 0x5cb0a9dcUL, 0x76f988daUL, + 0x983e5152UL, 0xa831c66dUL, 0xb00327c8UL, 0xbf597fc7UL, + 0xc6e00bf3UL, 0xd5a79147UL, 0x06ca6351UL, 0x14292967UL, + 0x27b70a85UL, 0x2e1b2138UL, 0x4d2c6dfcUL, 0x53380d13UL, + 0x650a7354UL, 0x766a0abbUL, 0x81c2c92eUL, 0x92722c85UL, + 0xa2bfe8a1UL, 0xa81a664bUL, 0xc24b8b70UL, 0xc76c51a3UL, + 0xd192e819UL, 0xd6990624UL, 0xf40e3585UL, 0x106aa070UL, + 0x19a4c116UL, 0x1e376c08UL, 0x2748774cUL, 0x34b0bcb5UL, + 0x391c0cb3UL, 0x4ed8aa4aUL, 0x5b9cca4fUL, 0x682e6ff3UL, + 0x748f82eeUL, 0x78a5636fUL, 0x84c87814UL, 0x8cc70208UL, + 0x90befffaUL, 0xa4506cebUL, 0xbef9a3f7UL, 0xc67178f2UL +}; + +/* Initial hash value H for SHA-256: */ +static const uint32_t sha256_initial_hash_value[8] = { + 0x6a09e667UL, + 0xbb67ae85UL, + 0x3c6ef372UL, + 0xa54ff53aUL, + 0x510e527fUL, + 0x9b05688cUL, + 0x1f83d9abUL, + 0x5be0cd19UL +}; + +/* Hash constant words K for SHA-384 and SHA-512: */ +static const uint64_t K512[80] = { + 0x428a2f98d728ae22ULL, 0x7137449123ef65cdULL, + 0xb5c0fbcfec4d3b2fULL, 0xe9b5dba58189dbbcULL, + 0x3956c25bf348b538ULL, 0x59f111f1b605d019ULL, + 0x923f82a4af194f9bULL, 0xab1c5ed5da6d8118ULL, + 0xd807aa98a3030242ULL, 0x12835b0145706fbeULL, + 0x243185be4ee4b28cULL, 0x550c7dc3d5ffb4e2ULL, + 0x72be5d74f27b896fULL, 0x80deb1fe3b1696b1ULL, + 0x9bdc06a725c71235ULL, 0xc19bf174cf692694ULL, + 0xe49b69c19ef14ad2ULL, 0xefbe4786384f25e3ULL, + 0x0fc19dc68b8cd5b5ULL, 0x240ca1cc77ac9c65ULL, + 0x2de92c6f592b0275ULL, 0x4a7484aa6ea6e483ULL, + 0x5cb0a9dcbd41fbd4ULL, 0x76f988da831153b5ULL, + 0x983e5152ee66dfabULL, 0xa831c66d2db43210ULL, + 0xb00327c898fb213fULL, 0xbf597fc7beef0ee4ULL, + 0xc6e00bf33da88fc2ULL, 0xd5a79147930aa725ULL, + 0x06ca6351e003826fULL, 0x142929670a0e6e70ULL, + 0x27b70a8546d22ffcULL, 0x2e1b21385c26c926ULL, + 0x4d2c6dfc5ac42aedULL, 0x53380d139d95b3dfULL, + 0x650a73548baf63deULL, 0x766a0abb3c77b2a8ULL, + 0x81c2c92e47edaee6ULL, 0x92722c851482353bULL, + 0xa2bfe8a14cf10364ULL, 0xa81a664bbc423001ULL, + 0xc24b8b70d0f89791ULL, 0xc76c51a30654be30ULL, + 0xd192e819d6ef5218ULL, 0xd69906245565a910ULL, + 0xf40e35855771202aULL, 0x106aa07032bbd1b8ULL, + 0x19a4c116b8d2d0c8ULL, 0x1e376c085141ab53ULL, + 0x2748774cdf8eeb99ULL, 0x34b0bcb5e19b48a8ULL, + 0x391c0cb3c5c95a63ULL, 0x4ed8aa4ae3418acbULL, + 0x5b9cca4f7763e373ULL, 0x682e6ff3d6b2b8a3ULL, + 0x748f82ee5defb2fcULL, 0x78a5636f43172f60ULL, + 0x84c87814a1f0ab72ULL, 0x8cc702081a6439ecULL, + 0x90befffa23631e28ULL, 0xa4506cebde82bde9ULL, + 0xbef9a3f7b2c67915ULL, 0xc67178f2e372532bULL, + 0xca273eceea26619cULL, 0xd186b8c721c0c207ULL, + 0xeada7dd6cde0eb1eULL, 0xf57d4f7fee6ed178ULL, + 0x06f067aa72176fbaULL, 0x0a637dc5a2c898a6ULL, + 0x113f9804bef90daeULL, 0x1b710b35131c471bULL, + 0x28db77f523047d84ULL, 0x32caab7b40c72493ULL, + 0x3c9ebe0a15c9bebcULL, 0x431d67c49c100d4cULL, + 0x4cc5d4becb3e42b6ULL, 0x597f299cfc657e2aULL, + 0x5fcb6fab3ad6faecULL, 0x6c44198c4a475817ULL +}; + +/* Initial hash value H for SHA-384 */ +static const uint64_t sha384_initial_hash_value[8] = { + 0xcbbb9d5dc1059ed8ULL, + 0x629a292a367cd507ULL, + 0x9159015a3070dd17ULL, + 0x152fecd8f70e5939ULL, + 0x67332667ffc00b31ULL, + 0x8eb44a8768581511ULL, + 0xdb0c2e0d64f98fa7ULL, + 0x47b5481dbefa4fa4ULL +}; + +/* Initial hash value H for SHA-512 */ +static const uint64_t sha512_initial_hash_value[8] = { + 0x6a09e667f3bcc908ULL, + 0xbb67ae8584caa73bULL, + 0x3c6ef372fe94f82bULL, + 0xa54ff53a5f1d36f1ULL, + 0x510e527fade682d1ULL, + 0x9b05688c2b3e6c1fULL, + 0x1f83d9abfb41bd6bULL, + 0x5be0cd19137e2179ULL +}; + +/* + * Constant used by SHA256/384/512_End() functions for converting the + * digest to a readable hexadecimal character string: + */ + +//static const char *sha2_hex_digits = "0123456789abcdef"; // PJ + + +/*** SHA-256: *********************************************************/ +static void SHA256_Init(SHA256_CTX* context) { + if (context == (SHA256_CTX*)0) { + return; + } + MEMCPY_BCOPY(context->state, sha256_initial_hash_value, SHA256_DIGEST_LENGTH); + MEMSET_BZERO(context->buffer, SHA256_BLOCK_LENGTH); + context->bitcount = 0; +} + +#ifdef SHA2_UNROLL_TRANSFORM + +/* Unrolled SHA-256 round macros: */ + +#if BYTE_ORDER == LITTLE_ENDIAN + +#define ROUND256_0_TO_15(a,b,c,d,e,f,g,h) \ + REVERSE32(*data++, W256[j]); \ + T1 = (h) + Sigma1_256(e) + Ch((e), (f), (g)) + \ + K256[j] + W256[j]; \ + (d) += T1; \ + (h) = T1 + Sigma0_256(a) + Maj((a), (b), (c)); \ + j++ + + +#else /* BYTE_ORDER == LITTLE_ENDIAN */ + +#define ROUND256_0_TO_15(a,b,c,d,e,f,g,h) \ + T1 = (h) + Sigma1_256(e) + Ch((e), (f), (g)) + \ + K256[j] + (W256[j] = *data++); \ + (d) += T1; \ + (h) = T1 + Sigma0_256(a) + Maj((a), (b), (c)); \ + j++ + +#endif /* BYTE_ORDER == LITTLE_ENDIAN */ + +#define ROUND256(a,b,c,d,e,f,g,h) \ + s0 = W256[(j+1)&0x0f]; \ + s0 = sigma0_256(s0); \ + s1 = W256[(j+14)&0x0f]; \ + s1 = sigma1_256(s1); \ + T1 = (h) + Sigma1_256(e) + Ch((e), (f), (g)) + K256[j] + \ + (W256[j&0x0f] += s1 + W256[(j+9)&0x0f] + s0); \ + (d) += T1; \ + (h) = T1 + Sigma0_256(a) + Maj((a), (b), (c)); \ + j++ + +void SHA256_Transform(SHA256_CTX* context, const uint32_t* data) { + uint32_t a, b, c, d, e, f, g, h, s0, s1; + uint32_t T1, *W256; + int j; + + W256 = (uint32_t*)context->buffer; + + /* Initialize registers with the prev. intermediate value */ + a = context->state[0]; + b = context->state[1]; + c = context->state[2]; + d = context->state[3]; + e = context->state[4]; + f = context->state[5]; + g = context->state[6]; + h = context->state[7]; + + j = 0; + do { + /* Rounds 0 to 15 (unrolled): */ + ROUND256_0_TO_15(a,b,c,d,e,f,g,h); + ROUND256_0_TO_15(h,a,b,c,d,e,f,g); + ROUND256_0_TO_15(g,h,a,b,c,d,e,f); + ROUND256_0_TO_15(f,g,h,a,b,c,d,e); + ROUND256_0_TO_15(e,f,g,h,a,b,c,d); + ROUND256_0_TO_15(d,e,f,g,h,a,b,c); + ROUND256_0_TO_15(c,d,e,f,g,h,a,b); + ROUND256_0_TO_15(b,c,d,e,f,g,h,a); + } while (j < 16); + + /* Now for the remaining rounds to 64: */ + do { + ROUND256(a,b,c,d,e,f,g,h); + ROUND256(h,a,b,c,d,e,f,g); + ROUND256(g,h,a,b,c,d,e,f); + ROUND256(f,g,h,a,b,c,d,e); + ROUND256(e,f,g,h,a,b,c,d); + ROUND256(d,e,f,g,h,a,b,c); + ROUND256(c,d,e,f,g,h,a,b); + ROUND256(b,c,d,e,f,g,h,a); + } while (j < 64); + + /* Compute the current intermediate hash value */ + context->state[0] += a; + context->state[1] += b; + context->state[2] += c; + context->state[3] += d; + context->state[4] += e; + context->state[5] += f; + context->state[6] += g; + context->state[7] += h; + + /* Clean up */ + a = b = c = d = e = f = g = h = T1 = 0; +} + +#else /* SHA2_UNROLL_TRANSFORM */ + +static void SHA256_Transform(SHA256_CTX* context, const uint32_t* data) { + uint32_t a, b, c, d, e, f, g, h, s0, s1; + uint32_t T1, T2, *W256; + int j; + +#if defined __arm__ || defined __ARM__ || defined ARM || defined __ARM || defined __arm || defined __ARM_ARCH ||defined __aarch64__ + W256 = (uint32_t*)((void *)context->buffer); +#else + W256 = (uint32_t*)context->buffer; +#endif + + + /* Initialize registers with the prev. intermediate value */ + a = context->state[0]; + b = context->state[1]; + c = context->state[2]; + d = context->state[3]; + e = context->state[4]; + f = context->state[5]; + g = context->state[6]; + h = context->state[7]; + + j = 0; + do { +#if BYTE_ORDER == LITTLE_ENDIAN + /* Copy data while converting to host byte order */ + REVERSE32(*data++,W256[j]); + /* Apply the SHA-256 compression function to update a..h */ + T1 = h + Sigma1_256(e) + Ch(e, f, g) + K256[j] + W256[j]; +#else /* BYTE_ORDER == LITTLE_ENDIAN */ + /* Apply the SHA-256 compression function to update a..h with copy */ + T1 = h + Sigma1_256(e) + Ch(e, f, g) + K256[j] + (W256[j] = *data++); +#endif /* BYTE_ORDER == LITTLE_ENDIAN */ + T2 = Sigma0_256(a) + Maj(a, b, c); + h = g; + g = f; + f = e; + e = d + T1; + d = c; + c = b; + b = a; + a = T1 + T2; + + j++; + } while (j < 16); + + do { + /* Part of the message block expansion: */ + s0 = W256[(j+1)&0x0f]; + s0 = sigma0_256(s0); + s1 = W256[(j+14)&0x0f]; + s1 = sigma1_256(s1); + + /* Apply the SHA-256 compression function to update a..h */ + T1 = h + Sigma1_256(e) + Ch(e, f, g) + K256[j] + + (W256[j&0x0f] += s1 + W256[(j+9)&0x0f] + s0); + T2 = Sigma0_256(a) + Maj(a, b, c); + h = g; + g = f; + f = e; + e = d + T1; + d = c; + c = b; + b = a; + a = T1 + T2; + + j++; + } while (j < 64); + + /* Compute the current intermediate hash value */ + context->state[0] += a; + context->state[1] += b; + context->state[2] += c; + context->state[3] += d; + context->state[4] += e; + context->state[5] += f; + context->state[6] += g; + context->state[7] += h; + + /* Clean up */ + a = b = c = d = e = f = g = h = T1 = T2 = 0; +} + +#endif /* SHA2_UNROLL_TRANSFORM */ + +static void SHA256_Update(SHA256_CTX* context, const uint8_t *data, size_t len) { + unsigned int freespace, usedspace; + + if (len == 0) { + /* Calling with no data is valid - we do nothing */ + return; + } + + /* Sanity check: */ + assert(context != (SHA256_CTX*)0 && data != (uint8_t*)0); + + usedspace = (context->bitcount >> 3) % SHA256_BLOCK_LENGTH; + if (usedspace > 0) { + /* Calculate how much free space is available in the buffer */ + freespace = SHA256_BLOCK_LENGTH - usedspace; + + if (len >= freespace) { + /* Fill the buffer completely and process it */ + MEMCPY_BCOPY(&context->buffer[usedspace], data, freespace); + context->bitcount += freespace << 3; + len -= freespace; + data += freespace; +#if defined __arm__ || defined __ARM__ || defined ARM || defined __ARM || defined __arm || defined __ARM_ARCH ||defined __aarch64__ + SHA256_Transform(context, (uint32_t*)((void *)context->buffer)); +#else + SHA256_Transform(context, (uint32_t*)context->buffer); +#endif + + } else { + /* The buffer is not yet full */ + MEMCPY_BCOPY(&context->buffer[usedspace], data, len); + context->bitcount += len << 3; + /* Clean up: */ + usedspace = freespace = 0; + return; + } + } + while (len >= SHA256_BLOCK_LENGTH) { + /* Process as many complete blocks as we can */ +#if defined __arm__ || defined __ARM__ || defined ARM || defined __ARM || defined __arm || defined __ARM_ARCH ||defined __aarch64__ + SHA256_Transform(context, (const uint32_t*)(const void *)(data)); +#else + SHA256_Transform(context, (const uint32_t*)data); +#endif + + context->bitcount += SHA256_BLOCK_LENGTH << 3; + len -= SHA256_BLOCK_LENGTH; + data += SHA256_BLOCK_LENGTH; + } + if (len > 0) { + /* There's left-overs, so save 'em */ + MEMCPY_BCOPY(context->buffer, data, len); + context->bitcount += len << 3; + } + /* Clean up: */ + usedspace = freespace = 0; +} + +static void SHA256_Final(uint8_t digest[], SHA256_CTX* context) { +#if defined __arm__ || defined __ARM__ || defined ARM || defined __ARM || defined __arm || defined __ARM_ARCH ||defined __aarch64__ + uint32_t *d ; +#else + uint32_t *d = (uint32_t*)digest; + +#endif + unsigned int usedspace; + + /* Sanity check: */ + assert(context != (SHA256_CTX*)0); +#if defined __arm__ || defined __ARM__ || defined ARM || defined __ARM || defined __arm || defined __ARM_ARCH ||defined __aarch64__ + d = malloc(sizeof(uint32_t)*8); /* why 8 ? see below for loop */ + assert(d); +#endif + + /* If no digest buffer is passed, we don't bother doing this: */ + if (digest != (uint8_t*)0) { + usedspace = (context->bitcount >> 3) % SHA256_BLOCK_LENGTH; +#if BYTE_ORDER == LITTLE_ENDIAN + /* Convert FROM host byte order */ + REVERSE64(context->bitcount,context->bitcount); +#endif + if (usedspace > 0) { + /* Begin padding with a 1 bit: */ + context->buffer[usedspace++] = 0x80; + + if (usedspace <= SHA256_SHORT_BLOCK_LENGTH) { + /* Set-up for the last transform: */ + MEMSET_BZERO(&context->buffer[usedspace], SHA256_SHORT_BLOCK_LENGTH - usedspace); + } else { + if (usedspace < SHA256_BLOCK_LENGTH) { + MEMSET_BZERO(&context->buffer[usedspace], SHA256_BLOCK_LENGTH - usedspace); + } + /* Do second-to-last transform: */ +#if defined __arm__ || defined __ARM__ || defined ARM || defined __ARM || defined __arm || defined __ARM_ARCH ||defined __aarch64__ + SHA256_Transform(context, (uint32_t*)(void *)(context->buffer)); +#else + SHA256_Transform(context, (uint32_t*)context->buffer); +#endif + + + /* And set-up for the last transform: */ + MEMSET_BZERO(context->buffer, SHA256_SHORT_BLOCK_LENGTH); + } + } else { + /* Set-up for the last transform: */ + MEMSET_BZERO(context->buffer, SHA256_SHORT_BLOCK_LENGTH); + + /* Begin padding with a 1 bit: */ + *context->buffer = 0x80; + } + /* Set the bit count: */ + //*(uint64_t*)&context->buffer[SHA256_SHORT_BLOCK_LENGTH] = context->bitcount; // aliasing violation warning + context->buffer64[SHA256_SHORT_BLOCK_LENGTH / sizeof(uint64_t)] = context->bitcount; + + /* Final transform: */ +#if defined __arm__ || defined __ARM__ || defined ARM || defined __ARM || defined __arm || defined __ARM_ARCH ||defined __aarch64__ + SHA256_Transform(context, (uint32_t*)((void *)(context->buffer))); + +#else + SHA256_Transform(context, (uint32_t*)context->buffer); + +#endif + +#if BYTE_ORDER == LITTLE_ENDIAN + { + /* Convert TO host byte order */ + int j; + for (j = 0; j < 8; j++) { + REVERSE32(context->state[j],context->state[j]); + *d++ = context->state[j]; + } + } +#else + MEMCPY_BCOPY(d, context->state, SHA256_DIGEST_LENGTH); +#endif + } + +#if defined __arm__ || defined __ARM__ || defined ARM || defined __ARM || defined __arm || defined __ARM_ARCH ||defined __aarch64__ + memcpy(digest,d,SHA256_DIGEST_LENGTH); + free(d); +#endif + + /* Clean up state data: */ + MEMSET_BZERO(context, sizeof(*context)); + usedspace = 0; +} + +/* +static char *SHA256_End(SHA256_CTX* context, char buffer[]) { + uint8_t digest[SHA256_DIGEST_LENGTH], *d = digest; + int i; + + / * Sanity check: * / + assert(context != (SHA256_CTX*)0); + + if (buffer != (char*)0) { + SHA256_Final(digest, context); + + for (i = 0; i < SHA256_DIGEST_LENGTH; i++) { + *buffer++ = sha2_hex_digits[(*d & 0xf0) >> 4]; + *buffer++ = sha2_hex_digits[*d & 0x0f]; + d++; + } + *buffer = (char)0; + } else { + MEMSET_BZERO(context, sizeof(context)); + } + MEMSET_BZERO(digest, SHA256_DIGEST_LENGTH); + return buffer; +} + +static char* SHA256_Data(const uint8_t* data, size_t len, char digest[SHA256_DIGEST_STRING_LENGTH]) { + SHA256_CTX context; + + SHA256_Init(&context); + SHA256_Update(&context, data, len); + return SHA256_End(&context, digest); +} +*/ + + +/*** SHA-512: *********************************************************/ +static void SHA512_Init(SHA512_CTX* context) { + if (context == (SHA512_CTX*)0) { + return; + } + MEMCPY_BCOPY(context->state, sha512_initial_hash_value, SHA512_DIGEST_LENGTH); + MEMSET_BZERO(context->buffer, SHA512_BLOCK_LENGTH); + context->bitcount[0] = context->bitcount[1] = 0; +} + +#ifdef SHA2_UNROLL_TRANSFORM + +/* Unrolled SHA-512 round macros: */ +#if BYTE_ORDER == LITTLE_ENDIAN + +#define ROUND512_0_TO_15(a,b,c,d,e,f,g,h) \ + REVERSE64(*data++, W512[j]); \ + T1 = (h) + Sigma1_512(e) + Ch((e), (f), (g)) + \ + K512[j] + W512[j]; \ + (d) += T1, \ + (h) = T1 + Sigma0_512(a) + Maj((a), (b), (c)), \ + j++ + + +#else /* BYTE_ORDER == LITTLE_ENDIAN */ + +#define ROUND512_0_TO_15(a,b,c,d,e,f,g,h) \ + T1 = (h) + Sigma1_512(e) + Ch((e), (f), (g)) + \ + K512[j] + (W512[j] = *data++); \ + (d) += T1; \ + (h) = T1 + Sigma0_512(a) + Maj((a), (b), (c)); \ + j++ + +#endif /* BYTE_ORDER == LITTLE_ENDIAN */ + +#define ROUND512(a,b,c,d,e,f,g,h) \ + s0 = W512[(j+1)&0x0f]; \ + s0 = sigma0_512(s0); \ + s1 = W512[(j+14)&0x0f]; \ + s1 = sigma1_512(s1); \ + T1 = (h) + Sigma1_512(e) + Ch((e), (f), (g)) + K512[j] + \ + (W512[j&0x0f] += s1 + W512[(j+9)&0x0f] + s0); \ + (d) += T1; \ + (h) = T1 + Sigma0_512(a) + Maj((a), (b), (c)); \ + j++ + +static void SHA512_Transform(SHA512_CTX* context, const uint64_t* data) { + uint64_t a, b, c, d, e, f, g, h, s0, s1; + uint64_t T1, *W512 = (uint64_t*)context->buffer; + int j; + + /* Initialize registers with the prev. intermediate value */ + a = context->state[0]; + b = context->state[1]; + c = context->state[2]; + d = context->state[3]; + e = context->state[4]; + f = context->state[5]; + g = context->state[6]; + h = context->state[7]; + + j = 0; + do { + ROUND512_0_TO_15(a,b,c,d,e,f,g,h); + ROUND512_0_TO_15(h,a,b,c,d,e,f,g); + ROUND512_0_TO_15(g,h,a,b,c,d,e,f); + ROUND512_0_TO_15(f,g,h,a,b,c,d,e); + ROUND512_0_TO_15(e,f,g,h,a,b,c,d); + ROUND512_0_TO_15(d,e,f,g,h,a,b,c); + ROUND512_0_TO_15(c,d,e,f,g,h,a,b); + ROUND512_0_TO_15(b,c,d,e,f,g,h,a); + } while (j < 16); + + /* Now for the remaining rounds up to 79: */ + do { + ROUND512(a,b,c,d,e,f,g,h); + ROUND512(h,a,b,c,d,e,f,g); + ROUND512(g,h,a,b,c,d,e,f); + ROUND512(f,g,h,a,b,c,d,e); + ROUND512(e,f,g,h,a,b,c,d); + ROUND512(d,e,f,g,h,a,b,c); + ROUND512(c,d,e,f,g,h,a,b); + ROUND512(b,c,d,e,f,g,h,a); + } while (j < 80); + + /* Compute the current intermediate hash value */ + context->state[0] += a; + context->state[1] += b; + context->state[2] += c; + context->state[3] += d; + context->state[4] += e; + context->state[5] += f; + context->state[6] += g; + context->state[7] += h; + + /* Clean up */ + a = b = c = d = e = f = g = h = T1 = 0; +} + +#else /* SHA2_UNROLL_TRANSFORM */ + +static void SHA512_Transform(SHA512_CTX* context, const uint64_t* data) { + uint64_t a, b, c, d, e, f, g, h, s0, s1; + uint64_t T1, T2, *W512 = (uint64_t*)((void *)context->buffer); + int j; + + /* Initialize registers with the prev. intermediate value */ + a = context->state[0]; + b = context->state[1]; + c = context->state[2]; + d = context->state[3]; + e = context->state[4]; + f = context->state[5]; + g = context->state[6]; + h = context->state[7]; + + j = 0; + do { +#if BYTE_ORDER == LITTLE_ENDIAN + /* Convert TO host byte order */ + REVERSE64(*data++, W512[j]); + /* Apply the SHA-512 compression function to update a..h */ + T1 = h + Sigma1_512(e) + Ch(e, f, g) + K512[j] + W512[j]; +#else /* BYTE_ORDER == LITTLE_ENDIAN */ + /* Apply the SHA-512 compression function to update a..h with copy */ + T1 = h + Sigma1_512(e) + Ch(e, f, g) + K512[j] + (W512[j] = *data++); +#endif /* BYTE_ORDER == LITTLE_ENDIAN */ + T2 = Sigma0_512(a) + Maj(a, b, c); + h = g; + g = f; + f = e; + e = d + T1; + d = c; + c = b; + b = a; + a = T1 + T2; + + j++; + } while (j < 16); + + do { + /* Part of the message block expansion: */ + s0 = W512[(j+1)&0x0f]; + s0 = sigma0_512(s0); + s1 = W512[(j+14)&0x0f]; + s1 = sigma1_512(s1); + + /* Apply the SHA-512 compression function to update a..h */ + T1 = h + Sigma1_512(e) + Ch(e, f, g) + K512[j] + + (W512[j&0x0f] += s1 + W512[(j+9)&0x0f] + s0); + T2 = Sigma0_512(a) + Maj(a, b, c); + h = g; + g = f; + f = e; + e = d + T1; + d = c; + c = b; + b = a; + a = T1 + T2; + + j++; + } while (j < 80); + + /* Compute the current intermediate hash value */ + context->state[0] += a; + context->state[1] += b; + context->state[2] += c; + context->state[3] += d; + context->state[4] += e; + context->state[5] += f; + context->state[6] += g; + context->state[7] += h; + + /* Clean up */ + a = b = c = d = e = f = g = h = T1 = T2 = 0; +} + +#endif /* SHA2_UNROLL_TRANSFORM */ + +static void SHA512_Update(SHA512_CTX* context, const uint8_t *data, size_t len) { + unsigned int freespace, usedspace; + + if (len == 0) { + /* Calling with no data is valid - we do nothing */ + return; + } + + /* Sanity check: */ + assert(context != (SHA512_CTX*)0 && data != (uint8_t*)0); + + usedspace = (context->bitcount[0] >> 3) % SHA512_BLOCK_LENGTH; + if (usedspace > 0) { + /* Calculate how much free space is available in the buffer */ + freespace = SHA512_BLOCK_LENGTH - usedspace; + + if (len >= freespace) { + /* Fill the buffer completely and process it */ + MEMCPY_BCOPY(&context->buffer[usedspace], data, freespace); + ADDINC128(context->bitcount, freespace << 3); + len -= freespace; + data += freespace; +#if defined __arm__ || defined __ARM__ || defined ARM || defined __ARM || defined __arm || defined __ARM_ARCH ||defined __aarch64__ + SHA512_Transform(context, (uint64_t*)((void *)context->buffer)); +#else + SHA512_Transform(context, (uint64_t*)context->buffer); +#endif + } else { + /* The buffer is not yet full */ + MEMCPY_BCOPY(&context->buffer[usedspace], data, len); + ADDINC128(context->bitcount, len << 3); + /* Clean up: */ + usedspace = freespace = 0; + return; + } + } + while (len >= SHA512_BLOCK_LENGTH) { + /* Process as many complete blocks as we can */ +#if defined __arm__ || defined __ARM__ || defined ARM || defined __ARM || defined __arm || defined __ARM_ARCH ||defined __aarch64__ + SHA512_Transform(context, (uint64_t*)((void *)context->buffer)); +#else + SHA512_Transform(context, (uint64_t*)context->buffer); +#endif + ADDINC128(context->bitcount, SHA512_BLOCK_LENGTH << 3); + len -= SHA512_BLOCK_LENGTH; + data += SHA512_BLOCK_LENGTH; + } + if (len > 0) { + /* There's left-overs, so save 'em */ + MEMCPY_BCOPY(context->buffer, data, len); + ADDINC128(context->bitcount, len << 3); + } + /* Clean up: */ + usedspace = freespace = 0; +} + +static void SHA512_Last(SHA512_CTX* context) { + unsigned int usedspace; + + usedspace = (context->bitcount[0] >> 3) % SHA512_BLOCK_LENGTH; +#if BYTE_ORDER == LITTLE_ENDIAN + /* Convert FROM host byte order */ + REVERSE64(context->bitcount[0],context->bitcount[0]); + REVERSE64(context->bitcount[1],context->bitcount[1]); +#endif + if (usedspace > 0) { + /* Begin padding with a 1 bit: */ + context->buffer[usedspace++] = 0x80; + + if (usedspace <= SHA512_SHORT_BLOCK_LENGTH) { + /* Set-up for the last transform: */ + MEMSET_BZERO(&context->buffer[usedspace], SHA512_SHORT_BLOCK_LENGTH - usedspace); + } else { + if (usedspace < SHA512_BLOCK_LENGTH) { + MEMSET_BZERO(&context->buffer[usedspace], SHA512_BLOCK_LENGTH - usedspace); + } + /* Do second-to-last transform: */ +#if defined __arm__ || defined __ARM__ || defined ARM || defined __ARM || defined __arm || defined __ARM_ARCH ||defined __aarch64__ + SHA512_Transform(context, (uint64_t*)((void *)context->buffer)); +#else + SHA512_Transform(context, (uint64_t*)context->buffer); +#endif + + + /* And set-up for the last transform: */ + MEMSET_BZERO(context->buffer, SHA512_BLOCK_LENGTH - 2); + } + } else { + /* Prepare for final transform: */ + MEMSET_BZERO(context->buffer, SHA512_SHORT_BLOCK_LENGTH); + + /* Begin padding with a 1 bit: */ + *context->buffer = 0x80; + } + /* Store the length of input data (in bits): */ + //*(uint64_t*)&context->buffer[SHA512_SHORT_BLOCK_LENGTH] = context->bitcount[1]; // aliasing violation warning + //*(uint64_t*)&context->buffer[SHA512_SHORT_BLOCK_LENGTH+8] = context->bitcount[0]; + context->buffer64[SHA512_SHORT_BLOCK_LENGTH / sizeof(uint64_t)] = context->bitcount[1]; + context->buffer64[SHA512_SHORT_BLOCK_LENGTH / sizeof(uint64_t) + 1] = context->bitcount[0]; + + /* Final transform: */ +#if defined __arm__ || defined __ARM__ || defined ARM || defined __ARM || defined __arm || defined __ARM_ARCH ||defined __aarch64__ + SHA512_Transform(context, (uint64_t*)((void *)context->buffer)); +#else + SHA512_Transform(context, (uint64_t*)context->buffer); +#endif + +} + +static void SHA512_Final(uint8_t digest[], SHA512_CTX* context) { +#if defined __arm__ || defined __ARM__ || defined ARM || defined __ARM || defined __arm || defined __ARM_ARCH ||defined __aarch64__ + uint64_t *d ; +#else + uint64_t *d = (uint64_t*)digest; + +#endif + + /* Sanity check: */ + assert(context != (SHA512_CTX*)0); +#if defined __arm__ || defined __ARM__ || defined ARM || defined __ARM || defined __arm || defined __ARM_ARCH ||defined __aarch64__ + d = malloc(sizeof(uint64_t)*8); /* why 8 ? see below for loop */ + assert(d); +#endif + + /* If no digest buffer is passed, we don't bother doing this: */ + if (digest != (uint8_t*)0) { + SHA512_Last(context); + + /* Save the hash data for output: */ +#if BYTE_ORDER == LITTLE_ENDIAN + { + /* Convert TO host byte order */ + int j; + for (j = 0; j < 8; j++) { + REVERSE64(context->state[j],context->state[j]); + *d++ = context->state[j]; + } + } +#else + MEMCPY_BCOPY(d, context->state, SHA512_DIGEST_LENGTH); +#endif + } +#if defined __arm__ || defined __ARM__ || defined ARM || defined __ARM || defined __arm || defined __ARM_ARCH ||defined __aarch64__ + memcpy(digest,d,SHA512_DIGEST_LENGTH); + free(d); +#endif + + /* Zero out state data */ + MEMSET_BZERO(context, sizeof(*context)); +} + +/* +static char *SHA512_End(SHA512_CTX* context, char buffer[]) { + uint8_t digest[SHA512_DIGEST_LENGTH], *d = digest; + int i; + + / * Sanity check: * / + assert(context != (SHA512_CTX*)0); + + if (buffer != (char*)0) { + SHA512_Final(digest, context); + + for (i = 0; i < SHA512_DIGEST_LENGTH; i++) { + *buffer++ = sha2_hex_digits[(*d & 0xf0) >> 4]; + *buffer++ = sha2_hex_digits[*d & 0x0f]; + d++; + } + *buffer = (char)0; + } else { + MEMSET_BZERO(context, sizeof(context)); + } + MEMSET_BZERO(digest, SHA512_DIGEST_LENGTH); + return buffer; +} + +static char* SHA512_Data(const uint8_t* data, size_t len, char digest[SHA512_DIGEST_STRING_LENGTH]) { + SHA512_CTX context; + + SHA512_Init(&context); + SHA512_Update(&context, data, len); + return SHA512_End(&context, digest); +} +*/ + +/*** SHA-384: *********************************************************/ +static void SHA384_Init(SHA384_CTX* context) { + if (context == (SHA384_CTX*)0) { + return; + } + MEMCPY_BCOPY(context->state, sha384_initial_hash_value, SHA512_DIGEST_LENGTH); + MEMSET_BZERO(context->buffer, SHA384_BLOCK_LENGTH); + context->bitcount[0] = context->bitcount[1] = 0; +} + +static void SHA384_Update(SHA384_CTX* context, const uint8_t* data, size_t len) { + SHA512_Update((SHA512_CTX*)context, data, len); +} + +static void SHA384_Final(uint8_t digest[], SHA384_CTX* context) { +#if defined __arm__ || defined __ARM__ || defined ARM || defined __ARM || defined __arm || defined __ARM_ARCH ||defined __aarch64__ + uint64_t *d; +#else + uint64_t *d = (uint64_t*)digest; +#endif + + + + /* Sanity check: */ + assert(context != (SHA384_CTX*)0); +#if defined __arm__ || defined __ARM__ || defined ARM || defined __ARM || defined __arm || defined __ARM_ARCH ||defined __aarch64__ + d = malloc(sizeof(uint64_t)*6); /* why 6 ? see below for loop */ + assert(d); +#endif + + /* If no digest buffer is passed, we don't bother doing this: */ + if (digest != (uint8_t*)0) { + SHA512_Last((SHA512_CTX*)context); + + /* Save the hash data for output: */ +#if BYTE_ORDER == LITTLE_ENDIAN + { + /* Convert TO host byte order */ + int j; + for (j = 0; j < 6; j++) { + REVERSE64(context->state[j],context->state[j]); + *d++ = context->state[j]; + } + } +#else + MEMCPY_BCOPY(d, context->state, SHA384_DIGEST_LENGTH); +#endif + } +#if defined __arm__ || defined __ARM__ || defined ARM || defined __ARM || defined __arm || defined __ARM_ARCH ||defined __aarch64__ + memcpy(digest,d,SHA384_DIGEST_LENGTH); + free(d); +#endif + /* Zero out state data */ + MEMSET_BZERO(context, sizeof(*context)); +} + +/* +static char *SHA384_End(SHA384_CTX* context, char buffer[]) { + uint8_t digest[SHA384_DIGEST_LENGTH], *d = digest; + int i; + + / * Sanity check: * / + assert(context != (SHA384_CTX*)0); + + if (buffer != (char*)0) { + SHA384_Final(digest, context); + + for (i = 0; i < SHA384_DIGEST_LENGTH; i++) { + *buffer++ = sha2_hex_digits[(*d & 0xf0) >> 4]; + *buffer++ = sha2_hex_digits[*d & 0x0f]; + d++; + } + *buffer = (char)0; + } else { + MEMSET_BZERO(context, sizeof(context)); + } + MEMSET_BZERO(digest, SHA384_DIGEST_LENGTH); + return buffer; +} + +static char* SHA384_Data(const uint8_t* data, size_t len, char digest[SHA384_DIGEST_STRING_LENGTH]) { + SHA384_CTX context; + + SHA384_Init(&context); + SHA384_Update(&context, data, len); + return SHA384_End(&context, digest); +} +*/ + +/* end of sha2.c */ + +void sha256_init (sha256_state *state) +{ + SHA256_Init(state); +} + +void sha384_init (sha384_state *state) +{ + SHA384_Init(state); +} + +void sha512_init (sha512_state *state) +{ + SHA512_Init(state); +} + + +void sha256_add (sha256_state *state, const void *data, size_t size) +{ + SHA256_Update(state, (const uint8_t *)data, size); +} + +void sha384_add (sha384_state *state, const void *data, size_t size) +{ + SHA384_Update(state, (const uint8_t *)data, size); +} + +void sha512_add (sha512_state *state, const void *data, size_t size) +{ + SHA512_Update(state, (const uint8_t *)data, size); +} + + +void sha256_put (sha256_state *state, uint8_t digest[SHA256_DIGEST_LENGTH]) +{ + SHA256_Final(digest, state); +} + +void sha384_put (sha384_state *state, uint8_t digest[SHA384_DIGEST_LENGTH]) +{ + SHA384_Final(digest, state); +} + +void sha512_put (sha384_state *state, uint8_t digest[SHA512_DIGEST_LENGTH]) +{ + SHA512_Final(digest, state); +} + + +void sha256 (const void *data, size_t size, uint8_t digest[SHA256_DIGEST_LENGTH]) +{ + sha256_state state; + SHA256_Init(&state); + SHA256_Update(&state, (const uint8_t *)data, size); + SHA256_Final(digest, &state); +} + +void sha384 (const void *data, size_t size, uint8_t digest[SHA384_DIGEST_LENGTH]) +{ + sha384_state state; + SHA384_Init(&state); + SHA384_Update(&state, (const uint8_t *)data, size); + SHA384_Final(digest, &state); +} + +void sha512 (const void *data, size_t size, uint8_t digest[SHA512_DIGEST_LENGTH]) +{ + sha512_state state; + SHA512_Init(&state); + SHA512_Update(&state, (const uint8_t *)data, size); + SHA512_Final(digest, &state); +} + +static sha256_state sha256state; +static sha384_state sha384state; +static sha512_state sha512state; + + +void sha256init (void) +{ + SHA256_Init(&sha256state); +} + +void sha384init (void) +{ + SHA384_Init(&sha384state); +} + +void sha512init (void) +{ + SHA512_Init(&sha512state); +} + + +void sha256add (const void *data, size_t size) +{ + SHA256_Update(&sha256state, (const uint8_t *)data, size); +} + +void sha384add (const void *data, size_t size) +{ + SHA384_Update(&sha384state, (const uint8_t *)data, size); +} + +void sha512add (const void *data, size_t size) +{ + SHA512_Update(&sha512state, (const uint8_t *)data, size); +} + + +void sha256put (uint8_t digest[SHA256_DIGEST_LENGTH]) +{ + SHA256_Final(digest, &sha256state); +} + +void sha384put (uint8_t digest[SHA384_DIGEST_LENGTH]) +{ + SHA384_Final(digest, &sha384state); +} + +void sha512put (uint8_t digest[SHA512_DIGEST_LENGTH]) +{ + SHA512_Final(digest, &sha512state); +} diff --git a/Build/source/texk/web2c/luatexdir/luapplib/util/utilsha.c-OK b/Build/source/texk/web2c/luatexdir/luapplib/util/utilsha.c-OK new file mode 100644 index 00000000000..a4dc6865a82 --- /dev/null +++ b/Build/source/texk/web2c/luatexdir/luapplib/util/utilsha.c-OK @@ -0,0 +1,1198 @@ +/* sha2 implementation by Aaron D. Gifford (http://www.aarongifford.com) */ + +#define BYTE_ORDER LITTLE_ENDIAN + +/* begin of sha2.c */ + +/* + * FILE: sha2.c + * AUTHOR: Aaron D. Gifford - http://www.aarongifford.com/ + * + * Copyright (c) 2000-2001, Aaron D. Gifford + * All rights reserved. + * + * Redistribution and use in source and binary forms, with or without + * modification, are permitted provided that the following conditions + * are met: + * 1. Redistributions of source code must retain the above copyright + * notice, this list of conditions and the following disclaimer. + * 2. Redistributions in binary form must reproduce the above copyright + * notice, this list of conditions and the following disclaimer in the + * documentation and/or other materials provided with the distribution. + * 3. Neither the name of the copyright holder nor the names of contributors + * may be used to endorse or promote products derived from this software + * without specific prior written permission. + * + * THIS SOFTWARE IS PROVIDED BY THE AUTHOR AND CONTRIBUTOR(S) ``AS IS'' AND + * ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE + * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE + * ARE DISCLAIMED. IN NO EVENT SHALL THE AUTHOR OR CONTRIBUTOR(S) BE LIABLE + * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL + * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS + * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) + * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT + * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY + * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF + * SUCH DAMAGE. + * + * $Id: sha2.c,v 1.1 2001/11/08 00:01:51 adg Exp adg $ + */ + +#include <string.h> /* memcpy()/memset() or bcopy()/bzero() */ +#include <assert.h> /* assert() */ +//#include "sha2.h" +#include "utilsha.h" + +/* + * ASSERT NOTE: + * Some sanity checking code is included using assert(). On my FreeBSD + * system, this additional code can be removed by compiling with NDEBUG + * defined. Check your own systems manpage on assert() to see how to + * compile WITHOUT the sanity checking code on your system. + * + * UNROLLED TRANSFORM LOOP NOTE: + * You can define SHA2_UNROLL_TRANSFORM to use the unrolled transform + * loop version for the hash transform rounds (defined using macros + * later in this file). Either define on the command line, for example: + * + * cc -DSHA2_UNROLL_TRANSFORM -o sha2 sha2.c sha2prog.c + * + * or define below: + * + * #define SHA2_UNROLL_TRANSFORM + * + */ + + +/*** SHA-256/384/512 Machine Architecture Definitions *****************/ +/* + * BYTE_ORDER NOTE: + * + * Please make sure that your system defines BYTE_ORDER. If your + * architecture is little-endian, make sure it also defines + * LITTLE_ENDIAN and that the two (BYTE_ORDER and LITTLE_ENDIAN) are + * equivilent. + * + * If your system does not define the above, then you can do so by + * hand like this: + * + * #define LITTLE_ENDIAN 1234 + * #define BIG_ENDIAN 4321 + * + * And for little-endian machines, add: + * + * #define BYTE_ORDER LITTLE_ENDIAN + * + * Or for big-endian machines: + * + * #define BYTE_ORDER BIG_ENDIAN + * + * The FreeBSD machine this was written on defines BYTE_ORDER + * appropriately by including <sys/types.h> (which in turn includes + * <machine/endian.h> where the appropriate definitions are actually + * made). + */ +#if !defined(BYTE_ORDER) || (BYTE_ORDER != LITTLE_ENDIAN && BYTE_ORDER != BIG_ENDIAN) +#error Define BYTE_ORDER to be equal to either LITTLE_ENDIAN or BIG_ENDIAN +#endif + +/* + * Define the followingsha2_* types to types of the correct length on + * the native archtecture. Most BSD systems and Linux define u_intXX_t + * types. Machines with very recent ANSI C headers, can use the + * uintXX_t definintions from inttypes.h by defining SHA2_USE_INTTYPES_H + * during compile or in the sha.h header file. + * + * Machines that support neither u_intXX_t nor inttypes.h's uintXX_t + * will need to define these three typedefs below (and the appropriate + * ones in sha.h too) by hand according to their system architecture. + * + * Thank you, Jun-ichiro itojun Hagino, for suggesting using u_intXX_t + * types and pointing out recent ANSI C support for uintXX_t in inttypes.h. + * + * PJ: replace by uintX_t + */ + +//typedef uint8_t sha2_byte; /* Exactly 1 byte */ +//typedef uint32_t sha2_word32; /* Exactly 4 bytes */ +//typedef uint64_t sha2_word64; /* Exactly 8 bytes */ + +/*** SHA-256/384/512 Various Length Definitions ***********************/ +/* NOTE: Most of these are in sha2.h */ +#define SHA256_SHORT_BLOCK_LENGTH (SHA256_BLOCK_LENGTH - 8) +#define SHA384_SHORT_BLOCK_LENGTH (SHA384_BLOCK_LENGTH - 16) +#define SHA512_SHORT_BLOCK_LENGTH (SHA512_BLOCK_LENGTH - 16) + + +/*** ENDIAN REVERSAL MACROS *******************************************/ +#if BYTE_ORDER == LITTLE_ENDIAN +#define REVERSE32(w,x) { \ + uint32_t tmp = (w); \ + tmp = (tmp >> 16) | (tmp << 16); \ + (x) = ((tmp & 0xff00ff00UL) >> 8) | ((tmp & 0x00ff00ffUL) << 8); \ +} +#define REVERSE64(w,x) { \ + uint64_t tmp = (w); \ + tmp = (tmp >> 32) | (tmp << 32); \ + tmp = ((tmp & 0xff00ff00ff00ff00ULL) >> 8) | \ + ((tmp & 0x00ff00ff00ff00ffULL) << 8); \ + (x) = ((tmp & 0xffff0000ffff0000ULL) >> 16) | \ + ((tmp & 0x0000ffff0000ffffULL) << 16); \ +} +#endif /* BYTE_ORDER == LITTLE_ENDIAN */ + +/* + * Macro for incrementally adding the unsigned 64-bit integer n to the + * unsigned 128-bit integer (represented using a two-element array of + * 64-bit words): + */ +#define ADDINC128(w,n) { \ + (w)[0] += (uint64_t)(n); \ + if ((w)[0] < (n)) { \ + (w)[1]++; \ + } \ +} + +/* + * Macros for copying blocks of memory and for zeroing out ranges + * of memory. Using these macros makes it easy to switch from + * using memset()/memcpy() and using bzero()/bcopy(). + * + * Please define either SHA2_USE_MEMSET_MEMCPY or define + * SHA2_USE_BZERO_BCOPY depending on which function set you + * choose to use: + */ +#if !defined(SHA2_USE_MEMSET_MEMCPY) && !defined(SHA2_USE_BZERO_BCOPY) +/* Default to memset()/memcpy() if no option is specified */ +#define SHA2_USE_MEMSET_MEMCPY 1 +#endif +#if defined(SHA2_USE_MEMSET_MEMCPY) && defined(SHA2_USE_BZERO_BCOPY) +/* Abort with an error if BOTH options are defined */ +#error Define either SHA2_USE_MEMSET_MEMCPY or SHA2_USE_BZERO_BCOPY, not both! +#endif + +#ifdef SHA2_USE_MEMSET_MEMCPY +#define MEMSET_BZERO(p,l) memset((p), 0, (l)) +#define MEMCPY_BCOPY(d,s,l) memcpy((d), (s), (l)) +#endif +#ifdef SHA2_USE_BZERO_BCOPY +#define MEMSET_BZERO(p,l) bzero((p), (l)) +#define MEMCPY_BCOPY(d,s,l) bcopy((s), (d), (l)) +#endif + + +/*** THE SIX LOGICAL FUNCTIONS ****************************************/ +/* + * Bit shifting and rotation (used by the six SHA-XYZ logical functions: + * + * NOTE: The naming of R and S appears backwards here (R is a SHIFT and + * S is a ROTATION) because the SHA-256/384/512 description document + * (see http://csrc.nist.gov/cryptval/shs/sha256-384-512.pdf) uses this + * same "backwards" definition. + */ +/* Shift-right (used in SHA-256, SHA-384, and SHA-512): */ +#define R(b,x) ((x) >> (b)) +/* 32-bit Rotate-right (used in SHA-256): */ +#define S32(b,x) (((x) >> (b)) | ((x) << (32 - (b)))) +/* 64-bit Rotate-right (used in SHA-384 and SHA-512): */ +#define S64(b,x) (((x) >> (b)) | ((x) << (64 - (b)))) + +/* Two of six logical functions used in SHA-256, SHA-384, and SHA-512: */ +#define Ch(x,y,z) (((x) & (y)) ^ ((~(x)) & (z))) +#define Maj(x,y,z) (((x) & (y)) ^ ((x) & (z)) ^ ((y) & (z))) + +/* Four of six logical functions used in SHA-256: */ +#define Sigma0_256(x) (S32(2, (x)) ^ S32(13, (x)) ^ S32(22, (x))) +#define Sigma1_256(x) (S32(6, (x)) ^ S32(11, (x)) ^ S32(25, (x))) +#define sigma0_256(x) (S32(7, (x)) ^ S32(18, (x)) ^ R(3 , (x))) +#define sigma1_256(x) (S32(17, (x)) ^ S32(19, (x)) ^ R(10, (x))) + +/* Four of six logical functions used in SHA-384 and SHA-512: */ +#define Sigma0_512(x) (S64(28, (x)) ^ S64(34, (x)) ^ S64(39, (x))) +#define Sigma1_512(x) (S64(14, (x)) ^ S64(18, (x)) ^ S64(41, (x))) +#define sigma0_512(x) (S64( 1, (x)) ^ S64( 8, (x)) ^ R( 7, (x))) +#define sigma1_512(x) (S64(19, (x)) ^ S64(61, (x)) ^ R( 6, (x))) + +/*** INTERNAL FUNCTION PROTOTYPES *************************************/ +/* NOTE: These should not be accessed directly from outside this + * library -- they are intended for private internal visibility/use + * only. + */ +static void SHA512_Last(SHA512_CTX*); +static void SHA256_Transform(SHA256_CTX*, const uint32_t*); +static void SHA512_Transform(SHA512_CTX*, const uint64_t*); + + +/*** SHA-XYZ INITIAL HASH VALUES AND CONSTANTS ************************/ +/* Hash constant words K for SHA-256: */ +static const uint32_t K256[64] = { + 0x428a2f98UL, 0x71374491UL, 0xb5c0fbcfUL, 0xe9b5dba5UL, + 0x3956c25bUL, 0x59f111f1UL, 0x923f82a4UL, 0xab1c5ed5UL, + 0xd807aa98UL, 0x12835b01UL, 0x243185beUL, 0x550c7dc3UL, + 0x72be5d74UL, 0x80deb1feUL, 0x9bdc06a7UL, 0xc19bf174UL, + 0xe49b69c1UL, 0xefbe4786UL, 0x0fc19dc6UL, 0x240ca1ccUL, + 0x2de92c6fUL, 0x4a7484aaUL, 0x5cb0a9dcUL, 0x76f988daUL, + 0x983e5152UL, 0xa831c66dUL, 0xb00327c8UL, 0xbf597fc7UL, + 0xc6e00bf3UL, 0xd5a79147UL, 0x06ca6351UL, 0x14292967UL, + 0x27b70a85UL, 0x2e1b2138UL, 0x4d2c6dfcUL, 0x53380d13UL, + 0x650a7354UL, 0x766a0abbUL, 0x81c2c92eUL, 0x92722c85UL, + 0xa2bfe8a1UL, 0xa81a664bUL, 0xc24b8b70UL, 0xc76c51a3UL, + 0xd192e819UL, 0xd6990624UL, 0xf40e3585UL, 0x106aa070UL, + 0x19a4c116UL, 0x1e376c08UL, 0x2748774cUL, 0x34b0bcb5UL, + 0x391c0cb3UL, 0x4ed8aa4aUL, 0x5b9cca4fUL, 0x682e6ff3UL, + 0x748f82eeUL, 0x78a5636fUL, 0x84c87814UL, 0x8cc70208UL, + 0x90befffaUL, 0xa4506cebUL, 0xbef9a3f7UL, 0xc67178f2UL +}; + +/* Initial hash value H for SHA-256: */ +static const uint32_t sha256_initial_hash_value[8] = { + 0x6a09e667UL, + 0xbb67ae85UL, + 0x3c6ef372UL, + 0xa54ff53aUL, + 0x510e527fUL, + 0x9b05688cUL, + 0x1f83d9abUL, + 0x5be0cd19UL +}; + +/* Hash constant words K for SHA-384 and SHA-512: */ +static const uint64_t K512[80] = { + 0x428a2f98d728ae22ULL, 0x7137449123ef65cdULL, + 0xb5c0fbcfec4d3b2fULL, 0xe9b5dba58189dbbcULL, + 0x3956c25bf348b538ULL, 0x59f111f1b605d019ULL, + 0x923f82a4af194f9bULL, 0xab1c5ed5da6d8118ULL, + 0xd807aa98a3030242ULL, 0x12835b0145706fbeULL, + 0x243185be4ee4b28cULL, 0x550c7dc3d5ffb4e2ULL, + 0x72be5d74f27b896fULL, 0x80deb1fe3b1696b1ULL, + 0x9bdc06a725c71235ULL, 0xc19bf174cf692694ULL, + 0xe49b69c19ef14ad2ULL, 0xefbe4786384f25e3ULL, + 0x0fc19dc68b8cd5b5ULL, 0x240ca1cc77ac9c65ULL, + 0x2de92c6f592b0275ULL, 0x4a7484aa6ea6e483ULL, + 0x5cb0a9dcbd41fbd4ULL, 0x76f988da831153b5ULL, + 0x983e5152ee66dfabULL, 0xa831c66d2db43210ULL, + 0xb00327c898fb213fULL, 0xbf597fc7beef0ee4ULL, + 0xc6e00bf33da88fc2ULL, 0xd5a79147930aa725ULL, + 0x06ca6351e003826fULL, 0x142929670a0e6e70ULL, + 0x27b70a8546d22ffcULL, 0x2e1b21385c26c926ULL, + 0x4d2c6dfc5ac42aedULL, 0x53380d139d95b3dfULL, + 0x650a73548baf63deULL, 0x766a0abb3c77b2a8ULL, + 0x81c2c92e47edaee6ULL, 0x92722c851482353bULL, + 0xa2bfe8a14cf10364ULL, 0xa81a664bbc423001ULL, + 0xc24b8b70d0f89791ULL, 0xc76c51a30654be30ULL, + 0xd192e819d6ef5218ULL, 0xd69906245565a910ULL, + 0xf40e35855771202aULL, 0x106aa07032bbd1b8ULL, + 0x19a4c116b8d2d0c8ULL, 0x1e376c085141ab53ULL, + 0x2748774cdf8eeb99ULL, 0x34b0bcb5e19b48a8ULL, + 0x391c0cb3c5c95a63ULL, 0x4ed8aa4ae3418acbULL, + 0x5b9cca4f7763e373ULL, 0x682e6ff3d6b2b8a3ULL, + 0x748f82ee5defb2fcULL, 0x78a5636f43172f60ULL, + 0x84c87814a1f0ab72ULL, 0x8cc702081a6439ecULL, + 0x90befffa23631e28ULL, 0xa4506cebde82bde9ULL, + 0xbef9a3f7b2c67915ULL, 0xc67178f2e372532bULL, + 0xca273eceea26619cULL, 0xd186b8c721c0c207ULL, + 0xeada7dd6cde0eb1eULL, 0xf57d4f7fee6ed178ULL, + 0x06f067aa72176fbaULL, 0x0a637dc5a2c898a6ULL, + 0x113f9804bef90daeULL, 0x1b710b35131c471bULL, + 0x28db77f523047d84ULL, 0x32caab7b40c72493ULL, + 0x3c9ebe0a15c9bebcULL, 0x431d67c49c100d4cULL, + 0x4cc5d4becb3e42b6ULL, 0x597f299cfc657e2aULL, + 0x5fcb6fab3ad6faecULL, 0x6c44198c4a475817ULL +}; + +/* Initial hash value H for SHA-384 */ +static const uint64_t sha384_initial_hash_value[8] = { + 0xcbbb9d5dc1059ed8ULL, + 0x629a292a367cd507ULL, + 0x9159015a3070dd17ULL, + 0x152fecd8f70e5939ULL, + 0x67332667ffc00b31ULL, + 0x8eb44a8768581511ULL, + 0xdb0c2e0d64f98fa7ULL, + 0x47b5481dbefa4fa4ULL +}; + +/* Initial hash value H for SHA-512 */ +static const uint64_t sha512_initial_hash_value[8] = { + 0x6a09e667f3bcc908ULL, + 0xbb67ae8584caa73bULL, + 0x3c6ef372fe94f82bULL, + 0xa54ff53a5f1d36f1ULL, + 0x510e527fade682d1ULL, + 0x9b05688c2b3e6c1fULL, + 0x1f83d9abfb41bd6bULL, + 0x5be0cd19137e2179ULL +}; + +/* + * Constant used by SHA256/384/512_End() functions for converting the + * digest to a readable hexadecimal character string: + */ + +//static const char *sha2_hex_digits = "0123456789abcdef"; // PJ + + +/*** SHA-256: *********************************************************/ +static void SHA256_Init(SHA256_CTX* context) { + if (context == (SHA256_CTX*)0) { + return; + } + MEMCPY_BCOPY(context->state, sha256_initial_hash_value, SHA256_DIGEST_LENGTH); + MEMSET_BZERO(context->buffer, SHA256_BLOCK_LENGTH); + context->bitcount = 0; +} + +#ifdef SHA2_UNROLL_TRANSFORM + +/* Unrolled SHA-256 round macros: */ + +#if BYTE_ORDER == LITTLE_ENDIAN + +#define ROUND256_0_TO_15(a,b,c,d,e,f,g,h) \ + REVERSE32(*data++, W256[j]); \ + T1 = (h) + Sigma1_256(e) + Ch((e), (f), (g)) + \ + K256[j] + W256[j]; \ + (d) += T1; \ + (h) = T1 + Sigma0_256(a) + Maj((a), (b), (c)); \ + j++ + + +#else /* BYTE_ORDER == LITTLE_ENDIAN */ + +#define ROUND256_0_TO_15(a,b,c,d,e,f,g,h) \ + T1 = (h) + Sigma1_256(e) + Ch((e), (f), (g)) + \ + K256[j] + (W256[j] = *data++); \ + (d) += T1; \ + (h) = T1 + Sigma0_256(a) + Maj((a), (b), (c)); \ + j++ + +#endif /* BYTE_ORDER == LITTLE_ENDIAN */ + +#define ROUND256(a,b,c,d,e,f,g,h) \ + s0 = W256[(j+1)&0x0f]; \ + s0 = sigma0_256(s0); \ + s1 = W256[(j+14)&0x0f]; \ + s1 = sigma1_256(s1); \ + T1 = (h) + Sigma1_256(e) + Ch((e), (f), (g)) + K256[j] + \ + (W256[j&0x0f] += s1 + W256[(j+9)&0x0f] + s0); \ + (d) += T1; \ + (h) = T1 + Sigma0_256(a) + Maj((a), (b), (c)); \ + j++ + +void SHA256_Transform(SHA256_CTX* context, const uint32_t* data) { + uint32_t a, b, c, d, e, f, g, h, s0, s1; + uint32_t T1, *W256; + int j; + + W256 = (uint32_t*)context->buffer; + + /* Initialize registers with the prev. intermediate value */ + a = context->state[0]; + b = context->state[1]; + c = context->state[2]; + d = context->state[3]; + e = context->state[4]; + f = context->state[5]; + g = context->state[6]; + h = context->state[7]; + + j = 0; + do { + /* Rounds 0 to 15 (unrolled): */ + ROUND256_0_TO_15(a,b,c,d,e,f,g,h); + ROUND256_0_TO_15(h,a,b,c,d,e,f,g); + ROUND256_0_TO_15(g,h,a,b,c,d,e,f); + ROUND256_0_TO_15(f,g,h,a,b,c,d,e); + ROUND256_0_TO_15(e,f,g,h,a,b,c,d); + ROUND256_0_TO_15(d,e,f,g,h,a,b,c); + ROUND256_0_TO_15(c,d,e,f,g,h,a,b); + ROUND256_0_TO_15(b,c,d,e,f,g,h,a); + } while (j < 16); + + /* Now for the remaining rounds to 64: */ + do { + ROUND256(a,b,c,d,e,f,g,h); + ROUND256(h,a,b,c,d,e,f,g); + ROUND256(g,h,a,b,c,d,e,f); + ROUND256(f,g,h,a,b,c,d,e); + ROUND256(e,f,g,h,a,b,c,d); + ROUND256(d,e,f,g,h,a,b,c); + ROUND256(c,d,e,f,g,h,a,b); + ROUND256(b,c,d,e,f,g,h,a); + } while (j < 64); + + /* Compute the current intermediate hash value */ + context->state[0] += a; + context->state[1] += b; + context->state[2] += c; + context->state[3] += d; + context->state[4] += e; + context->state[5] += f; + context->state[6] += g; + context->state[7] += h; + + /* Clean up */ + a = b = c = d = e = f = g = h = T1 = 0; +} + +#else /* SHA2_UNROLL_TRANSFORM */ + +static void SHA256_Transform(SHA256_CTX* context, const uint32_t* data) { + uint32_t a, b, c, d, e, f, g, h, s0, s1; + uint32_t T1, T2, *W256; + int j; + + W256 = (uint32_t*)context->buffer; + + /* Initialize registers with the prev. intermediate value */ + a = context->state[0]; + b = context->state[1]; + c = context->state[2]; + d = context->state[3]; + e = context->state[4]; + f = context->state[5]; + g = context->state[6]; + h = context->state[7]; + + j = 0; + do { +#if BYTE_ORDER == LITTLE_ENDIAN + /* Copy data while converting to host byte order */ + REVERSE32(*data++,W256[j]); + /* Apply the SHA-256 compression function to update a..h */ + T1 = h + Sigma1_256(e) + Ch(e, f, g) + K256[j] + W256[j]; +#else /* BYTE_ORDER == LITTLE_ENDIAN */ + /* Apply the SHA-256 compression function to update a..h with copy */ + T1 = h + Sigma1_256(e) + Ch(e, f, g) + K256[j] + (W256[j] = *data++); +#endif /* BYTE_ORDER == LITTLE_ENDIAN */ + T2 = Sigma0_256(a) + Maj(a, b, c); + h = g; + g = f; + f = e; + e = d + T1; + d = c; + c = b; + b = a; + a = T1 + T2; + + j++; + } while (j < 16); + + do { + /* Part of the message block expansion: */ + s0 = W256[(j+1)&0x0f]; + s0 = sigma0_256(s0); + s1 = W256[(j+14)&0x0f]; + s1 = sigma1_256(s1); + + /* Apply the SHA-256 compression function to update a..h */ + T1 = h + Sigma1_256(e) + Ch(e, f, g) + K256[j] + + (W256[j&0x0f] += s1 + W256[(j+9)&0x0f] + s0); + T2 = Sigma0_256(a) + Maj(a, b, c); + h = g; + g = f; + f = e; + e = d + T1; + d = c; + c = b; + b = a; + a = T1 + T2; + + j++; + } while (j < 64); + + /* Compute the current intermediate hash value */ + context->state[0] += a; + context->state[1] += b; + context->state[2] += c; + context->state[3] += d; + context->state[4] += e; + context->state[5] += f; + context->state[6] += g; + context->state[7] += h; + + /* Clean up */ + a = b = c = d = e = f = g = h = T1 = T2 = 0; +} + +#endif /* SHA2_UNROLL_TRANSFORM */ + +static void SHA256_Update(SHA256_CTX* context, const uint8_t *data, size_t len) { + unsigned int freespace, usedspace; + + if (len == 0) { + /* Calling with no data is valid - we do nothing */ + return; + } + + /* Sanity check: */ + assert(context != (SHA256_CTX*)0 && data != (uint8_t*)0); + + usedspace = (context->bitcount >> 3) % SHA256_BLOCK_LENGTH; + if (usedspace > 0) { + /* Calculate how much free space is available in the buffer */ + freespace = SHA256_BLOCK_LENGTH - usedspace; + + if (len >= freespace) { + /* Fill the buffer completely and process it */ + MEMCPY_BCOPY(&context->buffer[usedspace], data, freespace); + context->bitcount += freespace << 3; + len -= freespace; + data += freespace; + SHA256_Transform(context, (uint32_t*)context->buffer); + } else { + /* The buffer is not yet full */ + MEMCPY_BCOPY(&context->buffer[usedspace], data, len); + context->bitcount += len << 3; + /* Clean up: */ + usedspace = freespace = 0; + return; + } + } + while (len >= SHA256_BLOCK_LENGTH) { + /* Process as many complete blocks as we can */ + SHA256_Transform(context, (const uint32_t*)data); + context->bitcount += SHA256_BLOCK_LENGTH << 3; + len -= SHA256_BLOCK_LENGTH; + data += SHA256_BLOCK_LENGTH; + } + if (len > 0) { + /* There's left-overs, so save 'em */ + MEMCPY_BCOPY(context->buffer, data, len); + context->bitcount += len << 3; + } + /* Clean up: */ + usedspace = freespace = 0; +} + +static void SHA256_Final(uint8_t digest[], SHA256_CTX* context) { + uint32_t *d = (uint32_t*)digest; + unsigned int usedspace; + + /* Sanity check: */ + assert(context != (SHA256_CTX*)0); + + /* If no digest buffer is passed, we don't bother doing this: */ + if (digest != (uint8_t*)0) { + usedspace = (context->bitcount >> 3) % SHA256_BLOCK_LENGTH; +#if BYTE_ORDER == LITTLE_ENDIAN + /* Convert FROM host byte order */ + REVERSE64(context->bitcount,context->bitcount); +#endif + if (usedspace > 0) { + /* Begin padding with a 1 bit: */ + context->buffer[usedspace++] = 0x80; + + if (usedspace <= SHA256_SHORT_BLOCK_LENGTH) { + /* Set-up for the last transform: */ + MEMSET_BZERO(&context->buffer[usedspace], SHA256_SHORT_BLOCK_LENGTH - usedspace); + } else { + if (usedspace < SHA256_BLOCK_LENGTH) { + MEMSET_BZERO(&context->buffer[usedspace], SHA256_BLOCK_LENGTH - usedspace); + } + /* Do second-to-last transform: */ + SHA256_Transform(context, (uint32_t*)context->buffer); + + /* And set-up for the last transform: */ + MEMSET_BZERO(context->buffer, SHA256_SHORT_BLOCK_LENGTH); + } + } else { + /* Set-up for the last transform: */ + MEMSET_BZERO(context->buffer, SHA256_SHORT_BLOCK_LENGTH); + + /* Begin padding with a 1 bit: */ + *context->buffer = 0x80; + } + /* Set the bit count: */ + //*(uint64_t*)&context->buffer[SHA256_SHORT_BLOCK_LENGTH] = context->bitcount; // aliasing violation warning + context->buffer64[SHA256_SHORT_BLOCK_LENGTH / sizeof(uint64_t)] = context->bitcount; + + /* Final transform: */ + SHA256_Transform(context, (uint32_t*)context->buffer); + +#if BYTE_ORDER == LITTLE_ENDIAN + { + /* Convert TO host byte order */ + int j; + for (j = 0; j < 8; j++) { + REVERSE32(context->state[j],context->state[j]); + *d++ = context->state[j]; + } + } +#else + MEMCPY_BCOPY(d, context->state, SHA256_DIGEST_LENGTH); +#endif + } + + /* Clean up state data: */ + MEMSET_BZERO(context, sizeof(*context)); + usedspace = 0; +} + +/* +static char *SHA256_End(SHA256_CTX* context, char buffer[]) { + uint8_t digest[SHA256_DIGEST_LENGTH], *d = digest; + int i; + + / * Sanity check: * / + assert(context != (SHA256_CTX*)0); + + if (buffer != (char*)0) { + SHA256_Final(digest, context); + + for (i = 0; i < SHA256_DIGEST_LENGTH; i++) { + *buffer++ = sha2_hex_digits[(*d & 0xf0) >> 4]; + *buffer++ = sha2_hex_digits[*d & 0x0f]; + d++; + } + *buffer = (char)0; + } else { + MEMSET_BZERO(context, sizeof(context)); + } + MEMSET_BZERO(digest, SHA256_DIGEST_LENGTH); + return buffer; +} + +static char* SHA256_Data(const uint8_t* data, size_t len, char digest[SHA256_DIGEST_STRING_LENGTH]) { + SHA256_CTX context; + + SHA256_Init(&context); + SHA256_Update(&context, data, len); + return SHA256_End(&context, digest); +} +*/ + + +/*** SHA-512: *********************************************************/ +static void SHA512_Init(SHA512_CTX* context) { + if (context == (SHA512_CTX*)0) { + return; + } + MEMCPY_BCOPY(context->state, sha512_initial_hash_value, SHA512_DIGEST_LENGTH); + MEMSET_BZERO(context->buffer, SHA512_BLOCK_LENGTH); + context->bitcount[0] = context->bitcount[1] = 0; +} + +#ifdef SHA2_UNROLL_TRANSFORM + +/* Unrolled SHA-512 round macros: */ +#if BYTE_ORDER == LITTLE_ENDIAN + +#define ROUND512_0_TO_15(a,b,c,d,e,f,g,h) \ + REVERSE64(*data++, W512[j]); \ + T1 = (h) + Sigma1_512(e) + Ch((e), (f), (g)) + \ + K512[j] + W512[j]; \ + (d) += T1, \ + (h) = T1 + Sigma0_512(a) + Maj((a), (b), (c)), \ + j++ + + +#else /* BYTE_ORDER == LITTLE_ENDIAN */ + +#define ROUND512_0_TO_15(a,b,c,d,e,f,g,h) \ + T1 = (h) + Sigma1_512(e) + Ch((e), (f), (g)) + \ + K512[j] + (W512[j] = *data++); \ + (d) += T1; \ + (h) = T1 + Sigma0_512(a) + Maj((a), (b), (c)); \ + j++ + +#endif /* BYTE_ORDER == LITTLE_ENDIAN */ + +#define ROUND512(a,b,c,d,e,f,g,h) \ + s0 = W512[(j+1)&0x0f]; \ + s0 = sigma0_512(s0); \ + s1 = W512[(j+14)&0x0f]; \ + s1 = sigma1_512(s1); \ + T1 = (h) + Sigma1_512(e) + Ch((e), (f), (g)) + K512[j] + \ + (W512[j&0x0f] += s1 + W512[(j+9)&0x0f] + s0); \ + (d) += T1; \ + (h) = T1 + Sigma0_512(a) + Maj((a), (b), (c)); \ + j++ + +static void SHA512_Transform(SHA512_CTX* context, const uint64_t* data) { + uint64_t a, b, c, d, e, f, g, h, s0, s1; + uint64_t T1, *W512 = (uint64_t*)context->buffer; + int j; + + /* Initialize registers with the prev. intermediate value */ + a = context->state[0]; + b = context->state[1]; + c = context->state[2]; + d = context->state[3]; + e = context->state[4]; + f = context->state[5]; + g = context->state[6]; + h = context->state[7]; + + j = 0; + do { + ROUND512_0_TO_15(a,b,c,d,e,f,g,h); + ROUND512_0_TO_15(h,a,b,c,d,e,f,g); + ROUND512_0_TO_15(g,h,a,b,c,d,e,f); + ROUND512_0_TO_15(f,g,h,a,b,c,d,e); + ROUND512_0_TO_15(e,f,g,h,a,b,c,d); + ROUND512_0_TO_15(d,e,f,g,h,a,b,c); + ROUND512_0_TO_15(c,d,e,f,g,h,a,b); + ROUND512_0_TO_15(b,c,d,e,f,g,h,a); + } while (j < 16); + + /* Now for the remaining rounds up to 79: */ + do { + ROUND512(a,b,c,d,e,f,g,h); + ROUND512(h,a,b,c,d,e,f,g); + ROUND512(g,h,a,b,c,d,e,f); + ROUND512(f,g,h,a,b,c,d,e); + ROUND512(e,f,g,h,a,b,c,d); + ROUND512(d,e,f,g,h,a,b,c); + ROUND512(c,d,e,f,g,h,a,b); + ROUND512(b,c,d,e,f,g,h,a); + } while (j < 80); + + /* Compute the current intermediate hash value */ + context->state[0] += a; + context->state[1] += b; + context->state[2] += c; + context->state[3] += d; + context->state[4] += e; + context->state[5] += f; + context->state[6] += g; + context->state[7] += h; + + /* Clean up */ + a = b = c = d = e = f = g = h = T1 = 0; +} + +#else /* SHA2_UNROLL_TRANSFORM */ + +static void SHA512_Transform(SHA512_CTX* context, const uint64_t* data) { + uint64_t a, b, c, d, e, f, g, h, s0, s1; + uint64_t T1, T2, *W512 = (uint64_t*)context->buffer; + int j; + + /* Initialize registers with the prev. intermediate value */ + a = context->state[0]; + b = context->state[1]; + c = context->state[2]; + d = context->state[3]; + e = context->state[4]; + f = context->state[5]; + g = context->state[6]; + h = context->state[7]; + + j = 0; + do { +#if BYTE_ORDER == LITTLE_ENDIAN + /* Convert TO host byte order */ + REVERSE64(*data++, W512[j]); + /* Apply the SHA-512 compression function to update a..h */ + T1 = h + Sigma1_512(e) + Ch(e, f, g) + K512[j] + W512[j]; +#else /* BYTE_ORDER == LITTLE_ENDIAN */ + /* Apply the SHA-512 compression function to update a..h with copy */ + T1 = h + Sigma1_512(e) + Ch(e, f, g) + K512[j] + (W512[j] = *data++); +#endif /* BYTE_ORDER == LITTLE_ENDIAN */ + T2 = Sigma0_512(a) + Maj(a, b, c); + h = g; + g = f; + f = e; + e = d + T1; + d = c; + c = b; + b = a; + a = T1 + T2; + + j++; + } while (j < 16); + + do { + /* Part of the message block expansion: */ + s0 = W512[(j+1)&0x0f]; + s0 = sigma0_512(s0); + s1 = W512[(j+14)&0x0f]; + s1 = sigma1_512(s1); + + /* Apply the SHA-512 compression function to update a..h */ + T1 = h + Sigma1_512(e) + Ch(e, f, g) + K512[j] + + (W512[j&0x0f] += s1 + W512[(j+9)&0x0f] + s0); + T2 = Sigma0_512(a) + Maj(a, b, c); + h = g; + g = f; + f = e; + e = d + T1; + d = c; + c = b; + b = a; + a = T1 + T2; + + j++; + } while (j < 80); + + /* Compute the current intermediate hash value */ + context->state[0] += a; + context->state[1] += b; + context->state[2] += c; + context->state[3] += d; + context->state[4] += e; + context->state[5] += f; + context->state[6] += g; + context->state[7] += h; + + /* Clean up */ + a = b = c = d = e = f = g = h = T1 = T2 = 0; +} + +#endif /* SHA2_UNROLL_TRANSFORM */ + +static void SHA512_Update(SHA512_CTX* context, const uint8_t *data, size_t len) { + unsigned int freespace, usedspace; + + if (len == 0) { + /* Calling with no data is valid - we do nothing */ + return; + } + + /* Sanity check: */ + assert(context != (SHA512_CTX*)0 && data != (uint8_t*)0); + + usedspace = (context->bitcount[0] >> 3) % SHA512_BLOCK_LENGTH; + if (usedspace > 0) { + /* Calculate how much free space is available in the buffer */ + freespace = SHA512_BLOCK_LENGTH - usedspace; + + if (len >= freespace) { + /* Fill the buffer completely and process it */ + MEMCPY_BCOPY(&context->buffer[usedspace], data, freespace); + ADDINC128(context->bitcount, freespace << 3); + len -= freespace; + data += freespace; + SHA512_Transform(context, (uint64_t*)context->buffer); + } else { + /* The buffer is not yet full */ + MEMCPY_BCOPY(&context->buffer[usedspace], data, len); + ADDINC128(context->bitcount, len << 3); + /* Clean up: */ + usedspace = freespace = 0; + return; + } + } + while (len >= SHA512_BLOCK_LENGTH) { + /* Process as many complete blocks as we can */ + SHA512_Transform(context, (const uint64_t*)data); + ADDINC128(context->bitcount, SHA512_BLOCK_LENGTH << 3); + len -= SHA512_BLOCK_LENGTH; + data += SHA512_BLOCK_LENGTH; + } + if (len > 0) { + /* There's left-overs, so save 'em */ + MEMCPY_BCOPY(context->buffer, data, len); + ADDINC128(context->bitcount, len << 3); + } + /* Clean up: */ + usedspace = freespace = 0; +} + +static void SHA512_Last(SHA512_CTX* context) { + unsigned int usedspace; + + usedspace = (context->bitcount[0] >> 3) % SHA512_BLOCK_LENGTH; +#if BYTE_ORDER == LITTLE_ENDIAN + /* Convert FROM host byte order */ + REVERSE64(context->bitcount[0],context->bitcount[0]); + REVERSE64(context->bitcount[1],context->bitcount[1]); +#endif + if (usedspace > 0) { + /* Begin padding with a 1 bit: */ + context->buffer[usedspace++] = 0x80; + + if (usedspace <= SHA512_SHORT_BLOCK_LENGTH) { + /* Set-up for the last transform: */ + MEMSET_BZERO(&context->buffer[usedspace], SHA512_SHORT_BLOCK_LENGTH - usedspace); + } else { + if (usedspace < SHA512_BLOCK_LENGTH) { + MEMSET_BZERO(&context->buffer[usedspace], SHA512_BLOCK_LENGTH - usedspace); + } + /* Do second-to-last transform: */ + SHA512_Transform(context, (uint64_t*)context->buffer); + + /* And set-up for the last transform: */ + MEMSET_BZERO(context->buffer, SHA512_BLOCK_LENGTH - 2); + } + } else { + /* Prepare for final transform: */ + MEMSET_BZERO(context->buffer, SHA512_SHORT_BLOCK_LENGTH); + + /* Begin padding with a 1 bit: */ + *context->buffer = 0x80; + } + /* Store the length of input data (in bits): */ + //*(uint64_t*)&context->buffer[SHA512_SHORT_BLOCK_LENGTH] = context->bitcount[1]; // aliasing violation warning + //*(uint64_t*)&context->buffer[SHA512_SHORT_BLOCK_LENGTH+8] = context->bitcount[0]; + context->buffer64[SHA512_SHORT_BLOCK_LENGTH / sizeof(uint64_t)] = context->bitcount[1]; + context->buffer64[SHA512_SHORT_BLOCK_LENGTH / sizeof(uint64_t) + 1] = context->bitcount[0]; + + /* Final transform: */ + SHA512_Transform(context, (uint64_t*)context->buffer); +} + +static void SHA512_Final(uint8_t digest[], SHA512_CTX* context) { + uint64_t *d = (uint64_t*)digest; + + /* Sanity check: */ + assert(context != (SHA512_CTX*)0); + + /* If no digest buffer is passed, we don't bother doing this: */ + if (digest != (uint8_t*)0) { + SHA512_Last(context); + + /* Save the hash data for output: */ +#if BYTE_ORDER == LITTLE_ENDIAN + { + /* Convert TO host byte order */ + int j; + for (j = 0; j < 8; j++) { + REVERSE64(context->state[j],context->state[j]); + *d++ = context->state[j]; + } + } +#else + MEMCPY_BCOPY(d, context->state, SHA512_DIGEST_LENGTH); +#endif + } + + /* Zero out state data */ + MEMSET_BZERO(context, sizeof(*context)); +} + +/* +static char *SHA512_End(SHA512_CTX* context, char buffer[]) { + uint8_t digest[SHA512_DIGEST_LENGTH], *d = digest; + int i; + + / * Sanity check: * / + assert(context != (SHA512_CTX*)0); + + if (buffer != (char*)0) { + SHA512_Final(digest, context); + + for (i = 0; i < SHA512_DIGEST_LENGTH; i++) { + *buffer++ = sha2_hex_digits[(*d & 0xf0) >> 4]; + *buffer++ = sha2_hex_digits[*d & 0x0f]; + d++; + } + *buffer = (char)0; + } else { + MEMSET_BZERO(context, sizeof(context)); + } + MEMSET_BZERO(digest, SHA512_DIGEST_LENGTH); + return buffer; +} + +static char* SHA512_Data(const uint8_t* data, size_t len, char digest[SHA512_DIGEST_STRING_LENGTH]) { + SHA512_CTX context; + + SHA512_Init(&context); + SHA512_Update(&context, data, len); + return SHA512_End(&context, digest); +} +*/ + +/*** SHA-384: *********************************************************/ +static void SHA384_Init(SHA384_CTX* context) { + if (context == (SHA384_CTX*)0) { + return; + } + MEMCPY_BCOPY(context->state, sha384_initial_hash_value, SHA512_DIGEST_LENGTH); + MEMSET_BZERO(context->buffer, SHA384_BLOCK_LENGTH); + context->bitcount[0] = context->bitcount[1] = 0; +} + +static void SHA384_Update(SHA384_CTX* context, const uint8_t* data, size_t len) { + SHA512_Update((SHA512_CTX*)context, data, len); +} + +static void SHA384_Final(uint8_t digest[], SHA384_CTX* context) { + uint64_t *d = (uint64_t*)digest; + + /* Sanity check: */ + assert(context != (SHA384_CTX*)0); + + /* If no digest buffer is passed, we don't bother doing this: */ + if (digest != (uint8_t*)0) { + SHA512_Last((SHA512_CTX*)context); + + /* Save the hash data for output: */ +#if BYTE_ORDER == LITTLE_ENDIAN + { + /* Convert TO host byte order */ + int j; + for (j = 0; j < 6; j++) { + REVERSE64(context->state[j],context->state[j]); + *d++ = context->state[j]; + } + } +#else + MEMCPY_BCOPY(d, context->state, SHA384_DIGEST_LENGTH); +#endif + } + + /* Zero out state data */ + MEMSET_BZERO(context, sizeof(*context)); +} + +/* +static char *SHA384_End(SHA384_CTX* context, char buffer[]) { + uint8_t digest[SHA384_DIGEST_LENGTH], *d = digest; + int i; + + / * Sanity check: * / + assert(context != (SHA384_CTX*)0); + + if (buffer != (char*)0) { + SHA384_Final(digest, context); + + for (i = 0; i < SHA384_DIGEST_LENGTH; i++) { + *buffer++ = sha2_hex_digits[(*d & 0xf0) >> 4]; + *buffer++ = sha2_hex_digits[*d & 0x0f]; + d++; + } + *buffer = (char)0; + } else { + MEMSET_BZERO(context, sizeof(context)); + } + MEMSET_BZERO(digest, SHA384_DIGEST_LENGTH); + return buffer; +} + +static char* SHA384_Data(const uint8_t* data, size_t len, char digest[SHA384_DIGEST_STRING_LENGTH]) { + SHA384_CTX context; + + SHA384_Init(&context); + SHA384_Update(&context, data, len); + return SHA384_End(&context, digest); +} +*/ + +/* end of sha2.c */ + +void sha256_init (sha256_state *state) +{ + SHA256_Init(state); +} + +void sha384_init (sha384_state *state) +{ + SHA384_Init(state); +} + +void sha512_init (sha512_state *state) +{ + SHA512_Init(state); +} + + +void sha256_add (sha256_state *state, const void *data, size_t size) +{ + SHA256_Update(state, (const uint8_t *)data, size); +} + +void sha384_add (sha384_state *state, const void *data, size_t size) +{ + SHA384_Update(state, (const uint8_t *)data, size); +} + +void sha512_add (sha512_state *state, const void *data, size_t size) +{ + SHA512_Update(state, (const uint8_t *)data, size); +} + + +void sha256_put (sha256_state *state, uint8_t digest[SHA256_DIGEST_LENGTH]) +{ + SHA256_Final(digest, state); +} + +void sha384_put (sha384_state *state, uint8_t digest[SHA384_DIGEST_LENGTH]) +{ + SHA384_Final(digest, state); +} + +void sha512_put (sha384_state *state, uint8_t digest[SHA512_DIGEST_LENGTH]) +{ + SHA512_Final(digest, state); +} + + +void sha256 (const void *data, size_t size, uint8_t digest[SHA256_DIGEST_LENGTH]) +{ + sha256_state state; + SHA256_Init(&state); + SHA256_Update(&state, (const uint8_t *)data, size); + SHA256_Final(digest, &state); +} + +void sha384 (const void *data, size_t size, uint8_t digest[SHA384_DIGEST_LENGTH]) +{ + sha384_state state; + SHA384_Init(&state); + SHA384_Update(&state, (const uint8_t *)data, size); + SHA384_Final(digest, &state); +} + +void sha512 (const void *data, size_t size, uint8_t digest[SHA512_DIGEST_LENGTH]) +{ + sha512_state state; + SHA512_Init(&state); + SHA512_Update(&state, (const uint8_t *)data, size); + SHA512_Final(digest, &state); +} + +static sha256_state sha256state; +static sha384_state sha384state; +static sha512_state sha512state; + + +void sha256init (void) +{ + SHA256_Init(&sha256state); +} + +void sha384init (void) +{ + SHA384_Init(&sha384state); +} + +void sha512init (void) +{ + SHA512_Init(&sha512state); +} + + +void sha256add (const void *data, size_t size) +{ + SHA256_Update(&sha256state, (const uint8_t *)data, size); +} + +void sha384add (const void *data, size_t size) +{ + SHA384_Update(&sha384state, (const uint8_t *)data, size); +} + +void sha512add (const void *data, size_t size) +{ + SHA512_Update(&sha512state, (const uint8_t *)data, size); +} + + +void sha256put (uint8_t digest[SHA256_DIGEST_LENGTH]) +{ + SHA256_Final(digest, &sha256state); +} + +void sha384put (uint8_t digest[SHA384_DIGEST_LENGTH]) +{ + SHA384_Final(digest, &sha384state); +} + +void sha512put (uint8_t digest[SHA512_DIGEST_LENGTH]) +{ + SHA512_Final(digest, &sha512state); +} diff --git a/Build/source/texk/web2c/luatexdir/luapplib/util/utilsha.h b/Build/source/texk/web2c/luatexdir/luapplib/util/utilsha.h new file mode 100644 index 00000000000..53f35458181 --- /dev/null +++ b/Build/source/texk/web2c/luatexdir/luapplib/util/utilsha.h @@ -0,0 +1,134 @@ +/* sha2 implementation by Aaron D. Gifford (http://www.aarongifford.com) */ + +#ifndef UTIL_SHA_H +#define UTIL_SHA_H + +#include <stddef.h> +#include <stdint.h> +#include "utildecl.h" + +/* begin of sha2.h */ + +/* + * FILE: sha2.h + * AUTHOR: Aaron D. Gifford - http://www.aarongifford.com/ + * + * Copyright (c) 2000-2001, Aaron D. Gifford + * All rights reserved. + * + * Redistribution and use in source and binary forms, with or without + * modification, are permitted provided that the following conditions + * are met: + * 1. Redistributions of source code must retain the above copyright + * notice, this list of conditions and the following disclaimer. + * 2. Redistributions in binary form must reproduce the above copyright + * notice, this list of conditions and the following disclaimer in the + * documentation and/or other materials provided with the distribution. + * 3. Neither the name of the copyright holder nor the names of contributors + * may be used to endorse or promote products derived from this software + * without specific prior written permission. + * + * THIS SOFTWARE IS PROVIDED BY THE AUTHOR AND CONTRIBUTOR(S) ``AS IS'' AND + * ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE + * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE + * ARE DISCLAIMED. IN NO EVENT SHALL THE AUTHOR OR CONTRIBUTOR(S) BE LIABLE + * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL + * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS + * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) + * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT + * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY + * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF + * SUCH DAMAGE. + * + * $Id: sha2.h,v 1.1 2001/11/08 00:02:01 adg Exp adg $ + */ + +/*** SHA-256/384/512 Various Length Definitions ***********************/ + +#define SHA256_BLOCK_LENGTH 64 +#define SHA256_DIGEST_LENGTH 32 +#define SHA256_DIGEST_STRING_LENGTH (SHA256_DIGEST_LENGTH * 2 + 1) +#define SHA384_BLOCK_LENGTH 128 +#define SHA384_DIGEST_LENGTH 48 +#define SHA384_DIGEST_STRING_LENGTH (SHA384_DIGEST_LENGTH * 2 + 1) +#define SHA512_BLOCK_LENGTH 128 +#define SHA512_DIGEST_LENGTH 64 +#define SHA512_DIGEST_STRING_LENGTH (SHA512_DIGEST_LENGTH * 2 + 1) + +typedef struct _SHA256_CTX { + uint32_t state[8]; + uint64_t bitcount; + union { + uint8_t buffer[SHA256_BLOCK_LENGTH]; + uint64_t buffer64[SHA256_BLOCK_LENGTH / sizeof(uint64_t)]; // to avoid warnings about violating aliasing rules + }; +} SHA256_CTX; + +typedef struct _SHA512_CTX { + uint64_t state[8]; + uint64_t bitcount[2]; + union { + uint8_t buffer[SHA512_BLOCK_LENGTH]; + uint64_t buffer64[SHA512_BLOCK_LENGTH / sizeof(uint64_t)]; + }; +} SHA512_CTX; + +typedef SHA512_CTX SHA384_CTX; + +/*** SHA-256/384/512 Function Prototypes ******************************/ + +/* +void SHA256_Init(SHA256_CTX *); +void SHA256_Update(SHA256_CTX*, const u_int8_t*, size_t); +void SHA256_Final(u_int8_t[SHA256_DIGEST_LENGTH], SHA256_CTX*); +char* SHA256_End(SHA256_CTX*, char[SHA256_DIGEST_STRING_LENGTH]); +char* SHA256_Data(const u_int8_t*, size_t, char[SHA256_DIGEST_STRING_LENGTH]); + +void SHA384_Init(SHA384_CTX*); +void SHA384_Update(SHA384_CTX*, const u_int8_t*, size_t); +void SHA384_Final(u_int8_t[SHA384_DIGEST_LENGTH], SHA384_CTX*); +char* SHA384_End(SHA384_CTX*, char[SHA384_DIGEST_STRING_LENGTH]); +char* SHA384_Data(const u_int8_t*, size_t, char[SHA384_DIGEST_STRING_LENGTH]); + +void SHA512_Init(SHA512_CTX*); +void SHA512_Update(SHA512_CTX*, const u_int8_t*, size_t); +void SHA512_Final(u_int8_t[SHA512_DIGEST_LENGTH], SHA512_CTX*); +char* SHA512_End(SHA512_CTX*, char[SHA512_DIGEST_STRING_LENGTH]); +char* SHA512_Data(const u_int8_t*, size_t, char[SHA512_DIGEST_STRING_LENGTH]); +*/ + +/* end of sha2.h */ + +#define sha256_state SHA256_CTX +#define sha384_state SHA384_CTX +#define sha512_state SHA512_CTX + +UTILAPI void sha256_init (sha256_state *state); +UTILAPI void sha384_init (sha384_state *state); +UTILAPI void sha512_init (sha512_state *state); + +UTILAPI void sha256_add (sha256_state *state, const void *data, size_t size); +UTILAPI void sha384_add (sha384_state *state, const void *data, size_t size); +UTILAPI void sha512_add (sha512_state *state, const void *data, size_t size); + +UTILAPI void sha256_put (sha256_state *state, uint8_t digest[SHA256_DIGEST_LENGTH]); +UTILAPI void sha384_put (sha384_state *state, uint8_t digest[SHA384_DIGEST_LENGTH]); +UTILAPI void sha512_put (sha512_state *state, uint8_t digest[SHA512_DIGEST_LENGTH]); + +UTILAPI void sha256 (const void *data, size_t size, uint8_t digest[SHA256_DIGEST_LENGTH]); +UTILAPI void sha384 (const void *data, size_t size, uint8_t digest[SHA384_DIGEST_LENGTH]); +UTILAPI void sha512 (const void *data, size_t size, uint8_t digest[SHA512_DIGEST_LENGTH]); + +UTILAPI void sha256init (void); +UTILAPI void sha384init (void); +UTILAPI void sha512init (void); + +UTILAPI void sha256add (const void *data, size_t size); +UTILAPI void sha384add (const void *data, size_t size); +UTILAPI void sha512add (const void *data, size_t size); + +UTILAPI void sha256put (uint8_t digest[SHA256_DIGEST_LENGTH]); +UTILAPI void sha384put (uint8_t digest[SHA384_DIGEST_LENGTH]); +UTILAPI void sha512put (uint8_t digest[SHA512_DIGEST_LENGTH]); + +#endif
\ No newline at end of file diff --git a/Build/source/texk/web2c/luatexdir/luapplib/zlib/zconf.h b/Build/source/texk/web2c/luatexdir/luapplib/zlib/zconf.h new file mode 100644 index 00000000000..02ce56c4313 --- /dev/null +++ b/Build/source/texk/web2c/luatexdir/luapplib/zlib/zconf.h @@ -0,0 +1,428 @@ +/* zconf.h -- configuration of the zlib compression library + * Copyright (C) 1995-2010 Jean-loup Gailly. + * For conditions of distribution and use, see copyright notice in zlib.h + */ + +/* @(#) $Id$ */ + +#ifndef ZCONF_H +#define ZCONF_H + +/* + * If you *really* need a unique prefix for all types and library functions, + * compile with -DZ_PREFIX. The "standard" zlib should be compiled without it. + * Even better than compiling with -DZ_PREFIX would be to use configure to set + * this permanently in zconf.h using "./configure --zprefix". + */ +#ifdef Z_PREFIX /* may be set to #if 1 by ./configure */ + +/* all linked symbols */ +# define _dist_code z__dist_code +# define _length_code z__length_code +# define _tr_align z__tr_align +# define _tr_flush_block z__tr_flush_block +# define _tr_init z__tr_init +# define _tr_stored_block z__tr_stored_block +# define _tr_tally z__tr_tally +# define adler32 z_adler32 +# define adler32_combine z_adler32_combine +# define adler32_combine64 z_adler32_combine64 +# define compress z_compress +# define compress2 z_compress2 +# define compressBound z_compressBound +# define crc32 z_crc32 +# define crc32_combine z_crc32_combine +# define crc32_combine64 z_crc32_combine64 +# define deflate z_deflate +# define deflateBound z_deflateBound +# define deflateCopy z_deflateCopy +# define deflateEnd z_deflateEnd +# define deflateInit2_ z_deflateInit2_ +# define deflateInit_ z_deflateInit_ +# define deflateParams z_deflateParams +# define deflatePrime z_deflatePrime +# define deflateReset z_deflateReset +# define deflateSetDictionary z_deflateSetDictionary +# define deflateSetHeader z_deflateSetHeader +# define deflateTune z_deflateTune +# define deflate_copyright z_deflate_copyright +# define get_crc_table z_get_crc_table +# define gz_error z_gz_error +# define gz_intmax z_gz_intmax +# define gz_strwinerror z_gz_strwinerror +# define gzbuffer z_gzbuffer +# define gzclearerr z_gzclearerr +# define gzclose z_gzclose +# define gzclose_r z_gzclose_r +# define gzclose_w z_gzclose_w +# define gzdirect z_gzdirect +# define gzdopen z_gzdopen +# define gzeof z_gzeof +# define gzerror z_gzerror +# define gzflush z_gzflush +# define gzgetc z_gzgetc +# define gzgets z_gzgets +# define gzoffset z_gzoffset +# define gzoffset64 z_gzoffset64 +# define gzopen z_gzopen +# define gzopen64 z_gzopen64 +# define gzprintf z_gzprintf +# define gzputc z_gzputc +# define gzputs z_gzputs +# define gzread z_gzread +# define gzrewind z_gzrewind +# define gzseek z_gzseek +# define gzseek64 z_gzseek64 +# define gzsetparams z_gzsetparams +# define gztell z_gztell +# define gztell64 z_gztell64 +# define gzungetc z_gzungetc +# define gzwrite z_gzwrite +# define inflate z_inflate +# define inflateBack z_inflateBack +# define inflateBackEnd z_inflateBackEnd +# define inflateBackInit_ z_inflateBackInit_ +# define inflateCopy z_inflateCopy +# define inflateEnd z_inflateEnd +# define inflateGetHeader z_inflateGetHeader +# define inflateInit2_ z_inflateInit2_ +# define inflateInit_ z_inflateInit_ +# define inflateMark z_inflateMark +# define inflatePrime z_inflatePrime +# define inflateReset z_inflateReset +# define inflateReset2 z_inflateReset2 +# define inflateSetDictionary z_inflateSetDictionary +# define inflateSync z_inflateSync +# define inflateSyncPoint z_inflateSyncPoint +# define inflateUndermine z_inflateUndermine +# define inflate_copyright z_inflate_copyright +# define inflate_fast z_inflate_fast +# define inflate_table z_inflate_table +# define uncompress z_uncompress +# define zError z_zError +# define zcalloc z_zcalloc +# define zcfree z_zcfree +# define zlibCompileFlags z_zlibCompileFlags +# define zlibVersion z_zlibVersion + +/* all zlib typedefs in zlib.h and zconf.h */ +# define Byte z_Byte +# define Bytef z_Bytef +# define alloc_func z_alloc_func +# define charf z_charf +# define free_func z_free_func +# define gzFile z_gzFile +# define gz_header z_gz_header +# define gz_headerp z_gz_headerp +# define in_func z_in_func +# define intf z_intf +# define out_func z_out_func +# define uInt z_uInt +# define uIntf z_uIntf +# define uLong z_uLong +# define uLongf z_uLongf +# define voidp z_voidp +# define voidpc z_voidpc +# define voidpf z_voidpf + +/* all zlib structs in zlib.h and zconf.h */ +# define gz_header_s z_gz_header_s +# define internal_state z_internal_state + +#endif + +#if defined(__MSDOS__) && !defined(MSDOS) +# define MSDOS +#endif +#if (defined(OS_2) || defined(__OS2__)) && !defined(OS2) +# define OS2 +#endif +#if defined(_WINDOWS) && !defined(WINDOWS) +# define WINDOWS +#endif +#if defined(_WIN32) || defined(_WIN32_WCE) || defined(__WIN32__) +# ifndef WIN32 +# define WIN32 +# endif +#endif +#if (defined(MSDOS) || defined(OS2) || defined(WINDOWS)) && !defined(WIN32) +# if !defined(__GNUC__) && !defined(__FLAT__) && !defined(__386__) +# ifndef SYS16BIT +# define SYS16BIT +# endif +# endif +#endif + +/* + * Compile with -DMAXSEG_64K if the alloc function cannot allocate more + * than 64k bytes at a time (needed on systems with 16-bit int). + */ +#ifdef SYS16BIT +# define MAXSEG_64K +#endif +#ifdef MSDOS +# define UNALIGNED_OK +#endif + +#ifdef __STDC_VERSION__ +# ifndef STDC +# define STDC +# endif +# if __STDC_VERSION__ >= 199901L +# ifndef STDC99 +# define STDC99 +# endif +# endif +#endif +#if !defined(STDC) && (defined(__STDC__) || defined(__cplusplus)) +# define STDC +#endif +#if !defined(STDC) && (defined(__GNUC__) || defined(__BORLANDC__)) +# define STDC +#endif +#if !defined(STDC) && (defined(MSDOS) || defined(WINDOWS) || defined(WIN32)) +# define STDC +#endif +#if !defined(STDC) && (defined(OS2) || defined(__HOS_AIX__)) +# define STDC +#endif + +#if defined(__OS400__) && !defined(STDC) /* iSeries (formerly AS/400). */ +# define STDC +#endif + +#ifndef STDC +# ifndef const /* cannot use !defined(STDC) && !defined(const) on Mac */ +# define const /* note: need a more gentle solution here */ +# endif +#endif + +/* Some Mac compilers merge all .h files incorrectly: */ +#if defined(__MWERKS__)||defined(applec)||defined(THINK_C)||defined(__SC__) +# define NO_DUMMY_DECL +#endif + +/* Maximum value for memLevel in deflateInit2 */ +#ifndef MAX_MEM_LEVEL +# ifdef MAXSEG_64K +# define MAX_MEM_LEVEL 8 +# else +# define MAX_MEM_LEVEL 9 +# endif +#endif + +/* Maximum value for windowBits in deflateInit2 and inflateInit2. + * WARNING: reducing MAX_WBITS makes minigzip unable to extract .gz files + * created by gzip. (Files created by minigzip can still be extracted by + * gzip.) + */ +#ifndef MAX_WBITS +# define MAX_WBITS 15 /* 32K LZ77 window */ +#endif + +/* The memory requirements for deflate are (in bytes): + (1 << (windowBits+2)) + (1 << (memLevel+9)) + that is: 128K for windowBits=15 + 128K for memLevel = 8 (default values) + plus a few kilobytes for small objects. For example, if you want to reduce + the default memory requirements from 256K to 128K, compile with + make CFLAGS="-O -DMAX_WBITS=14 -DMAX_MEM_LEVEL=7" + Of course this will generally degrade compression (there's no free lunch). + + The memory requirements for inflate are (in bytes) 1 << windowBits + that is, 32K for windowBits=15 (default value) plus a few kilobytes + for small objects. +*/ + + /* Type declarations */ + +#ifndef OF /* function prototypes */ +# ifdef STDC +# define OF(args) args +# else +# define OF(args) () +# endif +#endif + +/* The following definitions for FAR are needed only for MSDOS mixed + * model programming (small or medium model with some far allocations). + * This was tested only with MSC; for other MSDOS compilers you may have + * to define NO_MEMCPY in zutil.h. If you don't need the mixed model, + * just define FAR to be empty. + */ +#ifdef SYS16BIT +# if defined(M_I86SM) || defined(M_I86MM) + /* MSC small or medium model */ +# define SMALL_MEDIUM +# ifdef _MSC_VER +# define FAR _far +# else +# define FAR far +# endif +# endif +# if (defined(__SMALL__) || defined(__MEDIUM__)) + /* Turbo C small or medium model */ +# define SMALL_MEDIUM +# ifdef __BORLANDC__ +# define FAR _far +# else +# define FAR far +# endif +# endif +#endif + +#if defined(WINDOWS) || defined(WIN32) + /* If building or using zlib as a DLL, define ZLIB_DLL. + * This is not mandatory, but it offers a little performance increase. + */ +# ifdef ZLIB_DLL +# if defined(WIN32) && (!defined(__BORLANDC__) || (__BORLANDC__ >= 0x500)) +# ifdef ZLIB_INTERNAL +# define ZEXTERN extern __declspec(dllexport) +# else +# define ZEXTERN extern __declspec(dllimport) +# endif +# endif +# endif /* ZLIB_DLL */ + /* If building or using zlib with the WINAPI/WINAPIV calling convention, + * define ZLIB_WINAPI. + * Caution: the standard ZLIB1.DLL is NOT compiled using ZLIB_WINAPI. + */ +# ifdef ZLIB_WINAPI +# ifdef FAR +# undef FAR +# endif +# include <windows.h> + /* No need for _export, use ZLIB.DEF instead. */ + /* For complete Windows compatibility, use WINAPI, not __stdcall. */ +# define ZEXPORT WINAPI +# ifdef WIN32 +# define ZEXPORTVA WINAPIV +# else +# define ZEXPORTVA FAR CDECL +# endif +# endif +#endif + +#if defined (__BEOS__) +# ifdef ZLIB_DLL +# ifdef ZLIB_INTERNAL +# define ZEXPORT __declspec(dllexport) +# define ZEXPORTVA __declspec(dllexport) +# else +# define ZEXPORT __declspec(dllimport) +# define ZEXPORTVA __declspec(dllimport) +# endif +# endif +#endif + +#ifndef ZEXTERN +# define ZEXTERN extern +#endif +#ifndef ZEXPORT +# define ZEXPORT +#endif +#ifndef ZEXPORTVA +# define ZEXPORTVA +#endif + +#ifndef FAR +# define FAR +#endif + +#if !defined(__MACTYPES__) +typedef unsigned char Byte; /* 8 bits */ +#endif +typedef unsigned int uInt; /* 16 bits or more */ +typedef unsigned long uLong; /* 32 bits or more */ + +#ifdef SMALL_MEDIUM + /* Borland C/C++ and some old MSC versions ignore FAR inside typedef */ +# define Bytef Byte FAR +#else + typedef Byte FAR Bytef; +#endif +typedef char FAR charf; +typedef int FAR intf; +typedef uInt FAR uIntf; +typedef uLong FAR uLongf; + +#ifdef STDC + typedef void const *voidpc; + typedef void FAR *voidpf; + typedef void *voidp; +#else + typedef Byte const *voidpc; + typedef Byte FAR *voidpf; + typedef Byte *voidp; +#endif + +#ifdef HAVE_UNISTD_H /* may be set to #if 1 by ./configure */ +# define Z_HAVE_UNISTD_H +#endif + +#ifdef STDC +# include <sys/types.h> /* for off_t */ +#endif + +/* a little trick to accommodate both "#define _LARGEFILE64_SOURCE" and + * "#define _LARGEFILE64_SOURCE 1" as requesting 64-bit operations, (even + * though the former does not conform to the LFS document), but considering + * both "#undef _LARGEFILE64_SOURCE" and "#define _LARGEFILE64_SOURCE 0" as + * equivalently requesting no 64-bit operations + */ +#if -_LARGEFILE64_SOURCE - -1 == 1 +# undef _LARGEFILE64_SOURCE +#endif + +#if defined(Z_HAVE_UNISTD_H) || defined(_LARGEFILE64_SOURCE) +# include <unistd.h> /* for SEEK_* and off_t */ +# ifdef VMS +# include <unixio.h> /* for off_t */ +# endif +# ifndef z_off_t +# define z_off_t off_t +# endif +#endif + +#ifndef SEEK_SET +# define SEEK_SET 0 /* Seek from beginning of file. */ +# define SEEK_CUR 1 /* Seek from current position. */ +# define SEEK_END 2 /* Set file pointer to EOF plus "offset" */ +#endif + +#ifndef z_off_t +# define z_off_t long +#endif + +#if defined(_LARGEFILE64_SOURCE) && _LFS64_LARGEFILE-0 +# define z_off64_t off64_t +#else +# define z_off64_t z_off_t +#endif + +#if defined(__OS400__) +# define NO_vsnprintf +#endif + +#if defined(__MVS__) +# define NO_vsnprintf +#endif + +/* MVS linker does not support external names larger than 8 bytes */ +#if defined(__MVS__) + #pragma map(deflateInit_,"DEIN") + #pragma map(deflateInit2_,"DEIN2") + #pragma map(deflateEnd,"DEEND") + #pragma map(deflateBound,"DEBND") + #pragma map(inflateInit_,"ININ") + #pragma map(inflateInit2_,"ININ2") + #pragma map(inflateEnd,"INEND") + #pragma map(inflateSync,"INSY") + #pragma map(inflateSetDictionary,"INSEDI") + #pragma map(compressBound,"CMBND") + #pragma map(inflate_table,"INTABL") + #pragma map(inflate_fast,"INFA") + #pragma map(inflate_copyright,"INCOPY") +#endif + +#endif /* ZCONF_H */ diff --git a/Build/source/texk/web2c/luatexdir/luapplib/zlib/zlib.h b/Build/source/texk/web2c/luatexdir/luapplib/zlib/zlib.h new file mode 100644 index 00000000000..bfbba83e8ee --- /dev/null +++ b/Build/source/texk/web2c/luatexdir/luapplib/zlib/zlib.h @@ -0,0 +1,1613 @@ +/* zlib.h -- interface of the 'zlib' general purpose compression library + version 1.2.5, April 19th, 2010 + + Copyright (C) 1995-2010 Jean-loup Gailly and Mark Adler + + This software is provided 'as-is', without any express or implied + warranty. In no event will the authors be held liable for any damages + arising from the use of this software. + + Permission is granted to anyone to use this software for any purpose, + including commercial applications, and to alter it and redistribute it + freely, subject to the following restrictions: + + 1. The origin of this software must not be misrepresented; you must not + claim that you wrote the original software. If you use this software + in a product, an acknowledgment in the product documentation would be + appreciated but is not required. + 2. Altered source versions must be plainly marked as such, and must not be + misrepresented as being the original software. + 3. This notice may not be removed or altered from any source distribution. + + Jean-loup Gailly Mark Adler + jloup@gzip.org madler@alumni.caltech.edu + + + The data format used by the zlib library is described by RFCs (Request for + Comments) 1950 to 1952 in the files http://www.ietf.org/rfc/rfc1950.txt + (zlib format), rfc1951.txt (deflate format) and rfc1952.txt (gzip format). +*/ + +#ifndef ZLIB_H +#define ZLIB_H + +#include "zconf.h" + +#ifdef __cplusplus +extern "C" { +#endif + +#define ZLIB_VERSION "1.2.5" +#define ZLIB_VERNUM 0x1250 +#define ZLIB_VER_MAJOR 1 +#define ZLIB_VER_MINOR 2 +#define ZLIB_VER_REVISION 5 +#define ZLIB_VER_SUBREVISION 0 + +/* + The 'zlib' compression library provides in-memory compression and + decompression functions, including integrity checks of the uncompressed data. + This version of the library supports only one compression method (deflation) + but other algorithms will be added later and will have the same stream + interface. + + Compression can be done in a single step if the buffers are large enough, + or can be done by repeated calls of the compression function. In the latter + case, the application must provide more input and/or consume the output + (providing more output space) before each call. + + The compressed data format used by default by the in-memory functions is + the zlib format, which is a zlib wrapper documented in RFC 1950, wrapped + around a deflate stream, which is itself documented in RFC 1951. + + The library also supports reading and writing files in gzip (.gz) format + with an interface similar to that of stdio using the functions that start + with "gz". The gzip format is different from the zlib format. gzip is a + gzip wrapper, documented in RFC 1952, wrapped around a deflate stream. + + This library can optionally read and write gzip streams in memory as well. + + The zlib format was designed to be compact and fast for use in memory + and on communications channels. The gzip format was designed for single- + file compression on file systems, has a larger header than zlib to maintain + directory information, and uses a different, slower check method than zlib. + + The library does not install any signal handler. The decoder checks + the consistency of the compressed data, so the library should never crash + even in case of corrupted input. +*/ + +typedef voidpf (*alloc_func) OF((voidpf opaque, uInt items, uInt size)); +typedef void (*free_func) OF((voidpf opaque, voidpf address)); + +struct internal_state; + +typedef struct z_stream_s { + Bytef *next_in; /* next input byte */ + uInt avail_in; /* number of bytes available at next_in */ + uLong total_in; /* total nb of input bytes read so far */ + + Bytef *next_out; /* next output byte should be put there */ + uInt avail_out; /* remaining free space at next_out */ + uLong total_out; /* total nb of bytes output so far */ + + char *msg; /* last error message, NULL if no error */ + struct internal_state FAR *state; /* not visible by applications */ + + alloc_func zalloc; /* used to allocate the internal state */ + free_func zfree; /* used to free the internal state */ + voidpf opaque; /* private data object passed to zalloc and zfree */ + + int data_type; /* best guess about the data type: binary or text */ + uLong adler; /* adler32 value of the uncompressed data */ + uLong reserved; /* reserved for future use */ +} z_stream; + +typedef z_stream FAR *z_streamp; + +/* + gzip header information passed to and from zlib routines. See RFC 1952 + for more details on the meanings of these fields. +*/ +typedef struct gz_header_s { + int text; /* true if compressed data believed to be text */ + uLong time; /* modification time */ + int xflags; /* extra flags (not used when writing a gzip file) */ + int os; /* operating system */ + Bytef *extra; /* pointer to extra field or Z_NULL if none */ + uInt extra_len; /* extra field length (valid if extra != Z_NULL) */ + uInt extra_max; /* space at extra (only when reading header) */ + Bytef *name; /* pointer to zero-terminated file name or Z_NULL */ + uInt name_max; /* space at name (only when reading header) */ + Bytef *comment; /* pointer to zero-terminated comment or Z_NULL */ + uInt comm_max; /* space at comment (only when reading header) */ + int hcrc; /* true if there was or will be a header crc */ + int done; /* true when done reading gzip header (not used + when writing a gzip file) */ +} gz_header; + +typedef gz_header FAR *gz_headerp; + +/* + The application must update next_in and avail_in when avail_in has dropped + to zero. It must update next_out and avail_out when avail_out has dropped + to zero. The application must initialize zalloc, zfree and opaque before + calling the init function. All other fields are set by the compression + library and must not be updated by the application. + + The opaque value provided by the application will be passed as the first + parameter for calls of zalloc and zfree. This can be useful for custom + memory management. The compression library attaches no meaning to the + opaque value. + + zalloc must return Z_NULL if there is not enough memory for the object. + If zlib is used in a multi-threaded application, zalloc and zfree must be + thread safe. + + On 16-bit systems, the functions zalloc and zfree must be able to allocate + exactly 65536 bytes, but will not be required to allocate more than this if + the symbol MAXSEG_64K is defined (see zconf.h). WARNING: On MSDOS, pointers + returned by zalloc for objects of exactly 65536 bytes *must* have their + offset normalized to zero. The default allocation function provided by this + library ensures this (see zutil.c). To reduce memory requirements and avoid + any allocation of 64K objects, at the expense of compression ratio, compile + the library with -DMAX_WBITS=14 (see zconf.h). + + The fields total_in and total_out can be used for statistics or progress + reports. After compression, total_in holds the total size of the + uncompressed data and may be saved for use in the decompressor (particularly + if the decompressor wants to decompress everything in a single step). +*/ + + /* constants */ + +#define Z_NO_FLUSH 0 +#define Z_PARTIAL_FLUSH 1 +#define Z_SYNC_FLUSH 2 +#define Z_FULL_FLUSH 3 +#define Z_FINISH 4 +#define Z_BLOCK 5 +#define Z_TREES 6 +/* Allowed flush values; see deflate() and inflate() below for details */ + +#define Z_OK 0 +#define Z_STREAM_END 1 +#define Z_NEED_DICT 2 +#define Z_ERRNO (-1) +#define Z_STREAM_ERROR (-2) +#define Z_DATA_ERROR (-3) +#define Z_MEM_ERROR (-4) +#define Z_BUF_ERROR (-5) +#define Z_VERSION_ERROR (-6) +/* Return codes for the compression/decompression functions. Negative values + * are errors, positive values are used for special but normal events. + */ + +#define Z_NO_COMPRESSION 0 +#define Z_BEST_SPEED 1 +#define Z_BEST_COMPRESSION 9 +#define Z_DEFAULT_COMPRESSION (-1) +/* compression levels */ + +#define Z_FILTERED 1 +#define Z_HUFFMAN_ONLY 2 +#define Z_RLE 3 +#define Z_FIXED 4 +#define Z_DEFAULT_STRATEGY 0 +/* compression strategy; see deflateInit2() below for details */ + +#define Z_BINARY 0 +#define Z_TEXT 1 +#define Z_ASCII Z_TEXT /* for compatibility with 1.2.2 and earlier */ +#define Z_UNKNOWN 2 +/* Possible values of the data_type field (though see inflate()) */ + +#define Z_DEFLATED 8 +/* The deflate compression method (the only one supported in this version) */ + +#define Z_NULL 0 /* for initializing zalloc, zfree, opaque */ + +#define zlib_version zlibVersion() +/* for compatibility with versions < 1.0.2 */ + + + /* basic functions */ + +ZEXTERN const char * ZEXPORT zlibVersion OF((void)); +/* The application can compare zlibVersion and ZLIB_VERSION for consistency. + If the first character differs, the library code actually used is not + compatible with the zlib.h header file used by the application. This check + is automatically made by deflateInit and inflateInit. + */ + +/* +ZEXTERN int ZEXPORT deflateInit OF((z_streamp strm, int level)); + + Initializes the internal stream state for compression. The fields + zalloc, zfree and opaque must be initialized before by the caller. If + zalloc and zfree are set to Z_NULL, deflateInit updates them to use default + allocation functions. + + The compression level must be Z_DEFAULT_COMPRESSION, or between 0 and 9: + 1 gives best speed, 9 gives best compression, 0 gives no compression at all + (the input data is simply copied a block at a time). Z_DEFAULT_COMPRESSION + requests a default compromise between speed and compression (currently + equivalent to level 6). + + deflateInit returns Z_OK if success, Z_MEM_ERROR if there was not enough + memory, Z_STREAM_ERROR if level is not a valid compression level, or + Z_VERSION_ERROR if the zlib library version (zlib_version) is incompatible + with the version assumed by the caller (ZLIB_VERSION). msg is set to null + if there is no error message. deflateInit does not perform any compression: + this will be done by deflate(). +*/ + + +ZEXTERN int ZEXPORT deflate OF((z_streamp strm, int flush)); +/* + deflate compresses as much data as possible, and stops when the input + buffer becomes empty or the output buffer becomes full. It may introduce + some output latency (reading input without producing any output) except when + forced to flush. + + The detailed semantics are as follows. deflate performs one or both of the + following actions: + + - Compress more input starting at next_in and update next_in and avail_in + accordingly. If not all input can be processed (because there is not + enough room in the output buffer), next_in and avail_in are updated and + processing will resume at this point for the next call of deflate(). + + - Provide more output starting at next_out and update next_out and avail_out + accordingly. This action is forced if the parameter flush is non zero. + Forcing flush frequently degrades the compression ratio, so this parameter + should be set only when necessary (in interactive applications). Some + output may be provided even if flush is not set. + + Before the call of deflate(), the application should ensure that at least + one of the actions is possible, by providing more input and/or consuming more + output, and updating avail_in or avail_out accordingly; avail_out should + never be zero before the call. The application can consume the compressed + output when it wants, for example when the output buffer is full (avail_out + == 0), or after each call of deflate(). If deflate returns Z_OK and with + zero avail_out, it must be called again after making room in the output + buffer because there might be more output pending. + + Normally the parameter flush is set to Z_NO_FLUSH, which allows deflate to + decide how much data to accumulate before producing output, in order to + maximize compression. + + If the parameter flush is set to Z_SYNC_FLUSH, all pending output is + flushed to the output buffer and the output is aligned on a byte boundary, so + that the decompressor can get all input data available so far. (In + particular avail_in is zero after the call if enough output space has been + provided before the call.) Flushing may degrade compression for some + compression algorithms and so it should be used only when necessary. This + completes the current deflate block and follows it with an empty stored block + that is three bits plus filler bits to the next byte, followed by four bytes + (00 00 ff ff). + + If flush is set to Z_PARTIAL_FLUSH, all pending output is flushed to the + output buffer, but the output is not aligned to a byte boundary. All of the + input data so far will be available to the decompressor, as for Z_SYNC_FLUSH. + This completes the current deflate block and follows it with an empty fixed + codes block that is 10 bits long. This assures that enough bytes are output + in order for the decompressor to finish the block before the empty fixed code + block. + + If flush is set to Z_BLOCK, a deflate block is completed and emitted, as + for Z_SYNC_FLUSH, but the output is not aligned on a byte boundary, and up to + seven bits of the current block are held to be written as the next byte after + the next deflate block is completed. In this case, the decompressor may not + be provided enough bits at this point in order to complete decompression of + the data provided so far to the compressor. It may need to wait for the next + block to be emitted. This is for advanced applications that need to control + the emission of deflate blocks. + + If flush is set to Z_FULL_FLUSH, all output is flushed as with + Z_SYNC_FLUSH, and the compression state is reset so that decompression can + restart from this point if previous compressed data has been damaged or if + random access is desired. Using Z_FULL_FLUSH too often can seriously degrade + compression. + + If deflate returns with avail_out == 0, this function must be called again + with the same value of the flush parameter and more output space (updated + avail_out), until the flush is complete (deflate returns with non-zero + avail_out). In the case of a Z_FULL_FLUSH or Z_SYNC_FLUSH, make sure that + avail_out is greater than six to avoid repeated flush markers due to + avail_out == 0 on return. + + If the parameter flush is set to Z_FINISH, pending input is processed, + pending output is flushed and deflate returns with Z_STREAM_END if there was + enough output space; if deflate returns with Z_OK, this function must be + called again with Z_FINISH and more output space (updated avail_out) but no + more input data, until it returns with Z_STREAM_END or an error. After + deflate has returned Z_STREAM_END, the only possible operations on the stream + are deflateReset or deflateEnd. + + Z_FINISH can be used immediately after deflateInit if all the compression + is to be done in a single step. In this case, avail_out must be at least the + value returned by deflateBound (see below). If deflate does not return + Z_STREAM_END, then it must be called again as described above. + + deflate() sets strm->adler to the adler32 checksum of all input read + so far (that is, total_in bytes). + + deflate() may update strm->data_type if it can make a good guess about + the input data type (Z_BINARY or Z_TEXT). In doubt, the data is considered + binary. This field is only for information purposes and does not affect the + compression algorithm in any manner. + + deflate() returns Z_OK if some progress has been made (more input + processed or more output produced), Z_STREAM_END if all input has been + consumed and all output has been produced (only when flush is set to + Z_FINISH), Z_STREAM_ERROR if the stream state was inconsistent (for example + if next_in or next_out was Z_NULL), Z_BUF_ERROR if no progress is possible + (for example avail_in or avail_out was zero). Note that Z_BUF_ERROR is not + fatal, and deflate() can be called again with more input and more output + space to continue compressing. +*/ + + +ZEXTERN int ZEXPORT deflateEnd OF((z_streamp strm)); +/* + All dynamically allocated data structures for this stream are freed. + This function discards any unprocessed input and does not flush any pending + output. + + deflateEnd returns Z_OK if success, Z_STREAM_ERROR if the + stream state was inconsistent, Z_DATA_ERROR if the stream was freed + prematurely (some input or output was discarded). In the error case, msg + may be set but then points to a static string (which must not be + deallocated). +*/ + + +/* +ZEXTERN int ZEXPORT inflateInit OF((z_streamp strm)); + + Initializes the internal stream state for decompression. The fields + next_in, avail_in, zalloc, zfree and opaque must be initialized before by + the caller. If next_in is not Z_NULL and avail_in is large enough (the + exact value depends on the compression method), inflateInit determines the + compression method from the zlib header and allocates all data structures + accordingly; otherwise the allocation will be deferred to the first call of + inflate. If zalloc and zfree are set to Z_NULL, inflateInit updates them to + use default allocation functions. + + inflateInit returns Z_OK if success, Z_MEM_ERROR if there was not enough + memory, Z_VERSION_ERROR if the zlib library version is incompatible with the + version assumed by the caller, or Z_STREAM_ERROR if the parameters are + invalid, such as a null pointer to the structure. msg is set to null if + there is no error message. inflateInit does not perform any decompression + apart from possibly reading the zlib header if present: actual decompression + will be done by inflate(). (So next_in and avail_in may be modified, but + next_out and avail_out are unused and unchanged.) The current implementation + of inflateInit() does not process any header information -- that is deferred + until inflate() is called. +*/ + + +ZEXTERN int ZEXPORT inflate OF((z_streamp strm, int flush)); +/* + inflate decompresses as much data as possible, and stops when the input + buffer becomes empty or the output buffer becomes full. It may introduce + some output latency (reading input without producing any output) except when + forced to flush. + + The detailed semantics are as follows. inflate performs one or both of the + following actions: + + - Decompress more input starting at next_in and update next_in and avail_in + accordingly. If not all input can be processed (because there is not + enough room in the output buffer), next_in is updated and processing will + resume at this point for the next call of inflate(). + + - Provide more output starting at next_out and update next_out and avail_out + accordingly. inflate() provides as much output as possible, until there is + no more input data or no more space in the output buffer (see below about + the flush parameter). + + Before the call of inflate(), the application should ensure that at least + one of the actions is possible, by providing more input and/or consuming more + output, and updating the next_* and avail_* values accordingly. The + application can consume the uncompressed output when it wants, for example + when the output buffer is full (avail_out == 0), or after each call of + inflate(). If inflate returns Z_OK and with zero avail_out, it must be + called again after making room in the output buffer because there might be + more output pending. + + The flush parameter of inflate() can be Z_NO_FLUSH, Z_SYNC_FLUSH, Z_FINISH, + Z_BLOCK, or Z_TREES. Z_SYNC_FLUSH requests that inflate() flush as much + output as possible to the output buffer. Z_BLOCK requests that inflate() + stop if and when it gets to the next deflate block boundary. When decoding + the zlib or gzip format, this will cause inflate() to return immediately + after the header and before the first block. When doing a raw inflate, + inflate() will go ahead and process the first block, and will return when it + gets to the end of that block, or when it runs out of data. + + The Z_BLOCK option assists in appending to or combining deflate streams. + Also to assist in this, on return inflate() will set strm->data_type to the + number of unused bits in the last byte taken from strm->next_in, plus 64 if + inflate() is currently decoding the last block in the deflate stream, plus + 128 if inflate() returned immediately after decoding an end-of-block code or + decoding the complete header up to just before the first byte of the deflate + stream. The end-of-block will not be indicated until all of the uncompressed + data from that block has been written to strm->next_out. The number of + unused bits may in general be greater than seven, except when bit 7 of + data_type is set, in which case the number of unused bits will be less than + eight. data_type is set as noted here every time inflate() returns for all + flush options, and so can be used to determine the amount of currently + consumed input in bits. + + The Z_TREES option behaves as Z_BLOCK does, but it also returns when the + end of each deflate block header is reached, before any actual data in that + block is decoded. This allows the caller to determine the length of the + deflate block header for later use in random access within a deflate block. + 256 is added to the value of strm->data_type when inflate() returns + immediately after reaching the end of the deflate block header. + + inflate() should normally be called until it returns Z_STREAM_END or an + error. However if all decompression is to be performed in a single step (a + single call of inflate), the parameter flush should be set to Z_FINISH. In + this case all pending input is processed and all pending output is flushed; + avail_out must be large enough to hold all the uncompressed data. (The size + of the uncompressed data may have been saved by the compressor for this + purpose.) The next operation on this stream must be inflateEnd to deallocate + the decompression state. The use of Z_FINISH is never required, but can be + used to inform inflate that a faster approach may be used for the single + inflate() call. + + In this implementation, inflate() always flushes as much output as + possible to the output buffer, and always uses the faster approach on the + first call. So the only effect of the flush parameter in this implementation + is on the return value of inflate(), as noted below, or when it returns early + because Z_BLOCK or Z_TREES is used. + + If a preset dictionary is needed after this call (see inflateSetDictionary + below), inflate sets strm->adler to the adler32 checksum of the dictionary + chosen by the compressor and returns Z_NEED_DICT; otherwise it sets + strm->adler to the adler32 checksum of all output produced so far (that is, + total_out bytes) and returns Z_OK, Z_STREAM_END or an error code as described + below. At the end of the stream, inflate() checks that its computed adler32 + checksum is equal to that saved by the compressor and returns Z_STREAM_END + only if the checksum is correct. + + inflate() can decompress and check either zlib-wrapped or gzip-wrapped + deflate data. The header type is detected automatically, if requested when + initializing with inflateInit2(). Any information contained in the gzip + header is not retained, so applications that need that information should + instead use raw inflate, see inflateInit2() below, or inflateBack() and + perform their own processing of the gzip header and trailer. + + inflate() returns Z_OK if some progress has been made (more input processed + or more output produced), Z_STREAM_END if the end of the compressed data has + been reached and all uncompressed output has been produced, Z_NEED_DICT if a + preset dictionary is needed at this point, Z_DATA_ERROR if the input data was + corrupted (input stream not conforming to the zlib format or incorrect check + value), Z_STREAM_ERROR if the stream structure was inconsistent (for example + next_in or next_out was Z_NULL), Z_MEM_ERROR if there was not enough memory, + Z_BUF_ERROR if no progress is possible or if there was not enough room in the + output buffer when Z_FINISH is used. Note that Z_BUF_ERROR is not fatal, and + inflate() can be called again with more input and more output space to + continue decompressing. If Z_DATA_ERROR is returned, the application may + then call inflateSync() to look for a good compression block if a partial + recovery of the data is desired. +*/ + + +ZEXTERN int ZEXPORT inflateEnd OF((z_streamp strm)); +/* + All dynamically allocated data structures for this stream are freed. + This function discards any unprocessed input and does not flush any pending + output. + + inflateEnd returns Z_OK if success, Z_STREAM_ERROR if the stream state + was inconsistent. In the error case, msg may be set but then points to a + static string (which must not be deallocated). +*/ + + + /* Advanced functions */ + +/* + The following functions are needed only in some special applications. +*/ + +/* +ZEXTERN int ZEXPORT deflateInit2 OF((z_streamp strm, + int level, + int method, + int windowBits, + int memLevel, + int strategy)); + + This is another version of deflateInit with more compression options. The + fields next_in, zalloc, zfree and opaque must be initialized before by the + caller. + + The method parameter is the compression method. It must be Z_DEFLATED in + this version of the library. + + The windowBits parameter is the base two logarithm of the window size + (the size of the history buffer). It should be in the range 8..15 for this + version of the library. Larger values of this parameter result in better + compression at the expense of memory usage. The default value is 15 if + deflateInit is used instead. + + windowBits can also be -8..-15 for raw deflate. In this case, -windowBits + determines the window size. deflate() will then generate raw deflate data + with no zlib header or trailer, and will not compute an adler32 check value. + + windowBits can also be greater than 15 for optional gzip encoding. Add + 16 to windowBits to write a simple gzip header and trailer around the + compressed data instead of a zlib wrapper. The gzip header will have no + file name, no extra data, no comment, no modification time (set to zero), no + header crc, and the operating system will be set to 255 (unknown). If a + gzip stream is being written, strm->adler is a crc32 instead of an adler32. + + The memLevel parameter specifies how much memory should be allocated + for the internal compression state. memLevel=1 uses minimum memory but is + slow and reduces compression ratio; memLevel=9 uses maximum memory for + optimal speed. The default value is 8. See zconf.h for total memory usage + as a function of windowBits and memLevel. + + The strategy parameter is used to tune the compression algorithm. Use the + value Z_DEFAULT_STRATEGY for normal data, Z_FILTERED for data produced by a + filter (or predictor), Z_HUFFMAN_ONLY to force Huffman encoding only (no + string match), or Z_RLE to limit match distances to one (run-length + encoding). Filtered data consists mostly of small values with a somewhat + random distribution. In this case, the compression algorithm is tuned to + compress them better. The effect of Z_FILTERED is to force more Huffman + coding and less string matching; it is somewhat intermediate between + Z_DEFAULT_STRATEGY and Z_HUFFMAN_ONLY. Z_RLE is designed to be almost as + fast as Z_HUFFMAN_ONLY, but give better compression for PNG image data. The + strategy parameter only affects the compression ratio but not the + correctness of the compressed output even if it is not set appropriately. + Z_FIXED prevents the use of dynamic Huffman codes, allowing for a simpler + decoder for special applications. + + deflateInit2 returns Z_OK if success, Z_MEM_ERROR if there was not enough + memory, Z_STREAM_ERROR if any parameter is invalid (such as an invalid + method), or Z_VERSION_ERROR if the zlib library version (zlib_version) is + incompatible with the version assumed by the caller (ZLIB_VERSION). msg is + set to null if there is no error message. deflateInit2 does not perform any + compression: this will be done by deflate(). +*/ + +ZEXTERN int ZEXPORT deflateSetDictionary OF((z_streamp strm, + const Bytef *dictionary, + uInt dictLength)); +/* + Initializes the compression dictionary from the given byte sequence + without producing any compressed output. This function must be called + immediately after deflateInit, deflateInit2 or deflateReset, before any call + of deflate. The compressor and decompressor must use exactly the same + dictionary (see inflateSetDictionary). + + The dictionary should consist of strings (byte sequences) that are likely + to be encountered later in the data to be compressed, with the most commonly + used strings preferably put towards the end of the dictionary. Using a + dictionary is most useful when the data to be compressed is short and can be + predicted with good accuracy; the data can then be compressed better than + with the default empty dictionary. + + Depending on the size of the compression data structures selected by + deflateInit or deflateInit2, a part of the dictionary may in effect be + discarded, for example if the dictionary is larger than the window size + provided in deflateInit or deflateInit2. Thus the strings most likely to be + useful should be put at the end of the dictionary, not at the front. In + addition, the current implementation of deflate will use at most the window + size minus 262 bytes of the provided dictionary. + + Upon return of this function, strm->adler is set to the adler32 value + of the dictionary; the decompressor may later use this value to determine + which dictionary has been used by the compressor. (The adler32 value + applies to the whole dictionary even if only a subset of the dictionary is + actually used by the compressor.) If a raw deflate was requested, then the + adler32 value is not computed and strm->adler is not set. + + deflateSetDictionary returns Z_OK if success, or Z_STREAM_ERROR if a + parameter is invalid (e.g. dictionary being Z_NULL) or the stream state is + inconsistent (for example if deflate has already been called for this stream + or if the compression method is bsort). deflateSetDictionary does not + perform any compression: this will be done by deflate(). +*/ + +ZEXTERN int ZEXPORT deflateCopy OF((z_streamp dest, + z_streamp source)); +/* + Sets the destination stream as a complete copy of the source stream. + + This function can be useful when several compression strategies will be + tried, for example when there are several ways of pre-processing the input + data with a filter. The streams that will be discarded should then be freed + by calling deflateEnd. Note that deflateCopy duplicates the internal + compression state which can be quite large, so this strategy is slow and can + consume lots of memory. + + deflateCopy returns Z_OK if success, Z_MEM_ERROR if there was not + enough memory, Z_STREAM_ERROR if the source stream state was inconsistent + (such as zalloc being Z_NULL). msg is left unchanged in both source and + destination. +*/ + +ZEXTERN int ZEXPORT deflateReset OF((z_streamp strm)); +/* + This function is equivalent to deflateEnd followed by deflateInit, + but does not free and reallocate all the internal compression state. The + stream will keep the same compression level and any other attributes that + may have been set by deflateInit2. + + deflateReset returns Z_OK if success, or Z_STREAM_ERROR if the source + stream state was inconsistent (such as zalloc or state being Z_NULL). +*/ + +ZEXTERN int ZEXPORT deflateParams OF((z_streamp strm, + int level, + int strategy)); +/* + Dynamically update the compression level and compression strategy. The + interpretation of level and strategy is as in deflateInit2. This can be + used to switch between compression and straight copy of the input data, or + to switch to a different kind of input data requiring a different strategy. + If the compression level is changed, the input available so far is + compressed with the old level (and may be flushed); the new level will take + effect only at the next call of deflate(). + + Before the call of deflateParams, the stream state must be set as for + a call of deflate(), since the currently available input may have to be + compressed and flushed. In particular, strm->avail_out must be non-zero. + + deflateParams returns Z_OK if success, Z_STREAM_ERROR if the source + stream state was inconsistent or if a parameter was invalid, Z_BUF_ERROR if + strm->avail_out was zero. +*/ + +ZEXTERN int ZEXPORT deflateTune OF((z_streamp strm, + int good_length, + int max_lazy, + int nice_length, + int max_chain)); +/* + Fine tune deflate's internal compression parameters. This should only be + used by someone who understands the algorithm used by zlib's deflate for + searching for the best matching string, and even then only by the most + fanatic optimizer trying to squeeze out the last compressed bit for their + specific input data. Read the deflate.c source code for the meaning of the + max_lazy, good_length, nice_length, and max_chain parameters. + + deflateTune() can be called after deflateInit() or deflateInit2(), and + returns Z_OK on success, or Z_STREAM_ERROR for an invalid deflate stream. + */ + +ZEXTERN uLong ZEXPORT deflateBound OF((z_streamp strm, + uLong sourceLen)); +/* + deflateBound() returns an upper bound on the compressed size after + deflation of sourceLen bytes. It must be called after deflateInit() or + deflateInit2(), and after deflateSetHeader(), if used. This would be used + to allocate an output buffer for deflation in a single pass, and so would be + called before deflate(). +*/ + +ZEXTERN int ZEXPORT deflatePrime OF((z_streamp strm, + int bits, + int value)); +/* + deflatePrime() inserts bits in the deflate output stream. The intent + is that this function is used to start off the deflate output with the bits + leftover from a previous deflate stream when appending to it. As such, this + function can only be used for raw deflate, and must be used before the first + deflate() call after a deflateInit2() or deflateReset(). bits must be less + than or equal to 16, and that many of the least significant bits of value + will be inserted in the output. + + deflatePrime returns Z_OK if success, or Z_STREAM_ERROR if the source + stream state was inconsistent. +*/ + +ZEXTERN int ZEXPORT deflateSetHeader OF((z_streamp strm, + gz_headerp head)); +/* + deflateSetHeader() provides gzip header information for when a gzip + stream is requested by deflateInit2(). deflateSetHeader() may be called + after deflateInit2() or deflateReset() and before the first call of + deflate(). The text, time, os, extra field, name, and comment information + in the provided gz_header structure are written to the gzip header (xflag is + ignored -- the extra flags are set according to the compression level). The + caller must assure that, if not Z_NULL, name and comment are terminated with + a zero byte, and that if extra is not Z_NULL, that extra_len bytes are + available there. If hcrc is true, a gzip header crc is included. Note that + the current versions of the command-line version of gzip (up through version + 1.3.x) do not support header crc's, and will report that it is a "multi-part + gzip file" and give up. + + If deflateSetHeader is not used, the default gzip header has text false, + the time set to zero, and os set to 255, with no extra, name, or comment + fields. The gzip header is returned to the default state by deflateReset(). + + deflateSetHeader returns Z_OK if success, or Z_STREAM_ERROR if the source + stream state was inconsistent. +*/ + +/* +ZEXTERN int ZEXPORT inflateInit2 OF((z_streamp strm, + int windowBits)); + + This is another version of inflateInit with an extra parameter. The + fields next_in, avail_in, zalloc, zfree and opaque must be initialized + before by the caller. + + The windowBits parameter is the base two logarithm of the maximum window + size (the size of the history buffer). It should be in the range 8..15 for + this version of the library. The default value is 15 if inflateInit is used + instead. windowBits must be greater than or equal to the windowBits value + provided to deflateInit2() while compressing, or it must be equal to 15 if + deflateInit2() was not used. If a compressed stream with a larger window + size is given as input, inflate() will return with the error code + Z_DATA_ERROR instead of trying to allocate a larger window. + + windowBits can also be zero to request that inflate use the window size in + the zlib header of the compressed stream. + + windowBits can also be -8..-15 for raw inflate. In this case, -windowBits + determines the window size. inflate() will then process raw deflate data, + not looking for a zlib or gzip header, not generating a check value, and not + looking for any check values for comparison at the end of the stream. This + is for use with other formats that use the deflate compressed data format + such as zip. Those formats provide their own check values. If a custom + format is developed using the raw deflate format for compressed data, it is + recommended that a check value such as an adler32 or a crc32 be applied to + the uncompressed data as is done in the zlib, gzip, and zip formats. For + most applications, the zlib format should be used as is. Note that comments + above on the use in deflateInit2() applies to the magnitude of windowBits. + + windowBits can also be greater than 15 for optional gzip decoding. Add + 32 to windowBits to enable zlib and gzip decoding with automatic header + detection, or add 16 to decode only the gzip format (the zlib format will + return a Z_DATA_ERROR). If a gzip stream is being decoded, strm->adler is a + crc32 instead of an adler32. + + inflateInit2 returns Z_OK if success, Z_MEM_ERROR if there was not enough + memory, Z_VERSION_ERROR if the zlib library version is incompatible with the + version assumed by the caller, or Z_STREAM_ERROR if the parameters are + invalid, such as a null pointer to the structure. msg is set to null if + there is no error message. inflateInit2 does not perform any decompression + apart from possibly reading the zlib header if present: actual decompression + will be done by inflate(). (So next_in and avail_in may be modified, but + next_out and avail_out are unused and unchanged.) The current implementation + of inflateInit2() does not process any header information -- that is + deferred until inflate() is called. +*/ + +ZEXTERN int ZEXPORT inflateSetDictionary OF((z_streamp strm, + const Bytef *dictionary, + uInt dictLength)); +/* + Initializes the decompression dictionary from the given uncompressed byte + sequence. This function must be called immediately after a call of inflate, + if that call returned Z_NEED_DICT. The dictionary chosen by the compressor + can be determined from the adler32 value returned by that call of inflate. + The compressor and decompressor must use exactly the same dictionary (see + deflateSetDictionary). For raw inflate, this function can be called + immediately after inflateInit2() or inflateReset() and before any call of + inflate() to set the dictionary. The application must insure that the + dictionary that was used for compression is provided. + + inflateSetDictionary returns Z_OK if success, Z_STREAM_ERROR if a + parameter is invalid (e.g. dictionary being Z_NULL) or the stream state is + inconsistent, Z_DATA_ERROR if the given dictionary doesn't match the + expected one (incorrect adler32 value). inflateSetDictionary does not + perform any decompression: this will be done by subsequent calls of + inflate(). +*/ + +ZEXTERN int ZEXPORT inflateSync OF((z_streamp strm)); +/* + Skips invalid compressed data until a full flush point (see above the + description of deflate with Z_FULL_FLUSH) can be found, or until all + available input is skipped. No output is provided. + + inflateSync returns Z_OK if a full flush point has been found, Z_BUF_ERROR + if no more input was provided, Z_DATA_ERROR if no flush point has been + found, or Z_STREAM_ERROR if the stream structure was inconsistent. In the + success case, the application may save the current current value of total_in + which indicates where valid compressed data was found. In the error case, + the application may repeatedly call inflateSync, providing more input each + time, until success or end of the input data. +*/ + +ZEXTERN int ZEXPORT inflateCopy OF((z_streamp dest, + z_streamp source)); +/* + Sets the destination stream as a complete copy of the source stream. + + This function can be useful when randomly accessing a large stream. The + first pass through the stream can periodically record the inflate state, + allowing restarting inflate at those points when randomly accessing the + stream. + + inflateCopy returns Z_OK if success, Z_MEM_ERROR if there was not + enough memory, Z_STREAM_ERROR if the source stream state was inconsistent + (such as zalloc being Z_NULL). msg is left unchanged in both source and + destination. +*/ + +ZEXTERN int ZEXPORT inflateReset OF((z_streamp strm)); +/* + This function is equivalent to inflateEnd followed by inflateInit, + but does not free and reallocate all the internal decompression state. The + stream will keep attributes that may have been set by inflateInit2. + + inflateReset returns Z_OK if success, or Z_STREAM_ERROR if the source + stream state was inconsistent (such as zalloc or state being Z_NULL). +*/ + +ZEXTERN int ZEXPORT inflateReset2 OF((z_streamp strm, + int windowBits)); +/* + This function is the same as inflateReset, but it also permits changing + the wrap and window size requests. The windowBits parameter is interpreted + the same as it is for inflateInit2. + + inflateReset2 returns Z_OK if success, or Z_STREAM_ERROR if the source + stream state was inconsistent (such as zalloc or state being Z_NULL), or if + the windowBits parameter is invalid. +*/ + +ZEXTERN int ZEXPORT inflatePrime OF((z_streamp strm, + int bits, + int value)); +/* + This function inserts bits in the inflate input stream. The intent is + that this function is used to start inflating at a bit position in the + middle of a byte. The provided bits will be used before any bytes are used + from next_in. This function should only be used with raw inflate, and + should be used before the first inflate() call after inflateInit2() or + inflateReset(). bits must be less than or equal to 16, and that many of the + least significant bits of value will be inserted in the input. + + If bits is negative, then the input stream bit buffer is emptied. Then + inflatePrime() can be called again to put bits in the buffer. This is used + to clear out bits leftover after feeding inflate a block description prior + to feeding inflate codes. + + inflatePrime returns Z_OK if success, or Z_STREAM_ERROR if the source + stream state was inconsistent. +*/ + +ZEXTERN long ZEXPORT inflateMark OF((z_streamp strm)); +/* + This function returns two values, one in the lower 16 bits of the return + value, and the other in the remaining upper bits, obtained by shifting the + return value down 16 bits. If the upper value is -1 and the lower value is + zero, then inflate() is currently decoding information outside of a block. + If the upper value is -1 and the lower value is non-zero, then inflate is in + the middle of a stored block, with the lower value equaling the number of + bytes from the input remaining to copy. If the upper value is not -1, then + it is the number of bits back from the current bit position in the input of + the code (literal or length/distance pair) currently being processed. In + that case the lower value is the number of bytes already emitted for that + code. + + A code is being processed if inflate is waiting for more input to complete + decoding of the code, or if it has completed decoding but is waiting for + more output space to write the literal or match data. + + inflateMark() is used to mark locations in the input data for random + access, which may be at bit positions, and to note those cases where the + output of a code may span boundaries of random access blocks. The current + location in the input stream can be determined from avail_in and data_type + as noted in the description for the Z_BLOCK flush parameter for inflate. + + inflateMark returns the value noted above or -1 << 16 if the provided + source stream state was inconsistent. +*/ + +ZEXTERN int ZEXPORT inflateGetHeader OF((z_streamp strm, + gz_headerp head)); +/* + inflateGetHeader() requests that gzip header information be stored in the + provided gz_header structure. inflateGetHeader() may be called after + inflateInit2() or inflateReset(), and before the first call of inflate(). + As inflate() processes the gzip stream, head->done is zero until the header + is completed, at which time head->done is set to one. If a zlib stream is + being decoded, then head->done is set to -1 to indicate that there will be + no gzip header information forthcoming. Note that Z_BLOCK or Z_TREES can be + used to force inflate() to return immediately after header processing is + complete and before any actual data is decompressed. + + The text, time, xflags, and os fields are filled in with the gzip header + contents. hcrc is set to true if there is a header CRC. (The header CRC + was valid if done is set to one.) If extra is not Z_NULL, then extra_max + contains the maximum number of bytes to write to extra. Once done is true, + extra_len contains the actual extra field length, and extra contains the + extra field, or that field truncated if extra_max is less than extra_len. + If name is not Z_NULL, then up to name_max characters are written there, + terminated with a zero unless the length is greater than name_max. If + comment is not Z_NULL, then up to comm_max characters are written there, + terminated with a zero unless the length is greater than comm_max. When any + of extra, name, or comment are not Z_NULL and the respective field is not + present in the header, then that field is set to Z_NULL to signal its + absence. This allows the use of deflateSetHeader() with the returned + structure to duplicate the header. However if those fields are set to + allocated memory, then the application will need to save those pointers + elsewhere so that they can be eventually freed. + + If inflateGetHeader is not used, then the header information is simply + discarded. The header is always checked for validity, including the header + CRC if present. inflateReset() will reset the process to discard the header + information. The application would need to call inflateGetHeader() again to + retrieve the header from the next gzip stream. + + inflateGetHeader returns Z_OK if success, or Z_STREAM_ERROR if the source + stream state was inconsistent. +*/ + +/* +ZEXTERN int ZEXPORT inflateBackInit OF((z_streamp strm, int windowBits, + unsigned char FAR *window)); + + Initialize the internal stream state for decompression using inflateBack() + calls. The fields zalloc, zfree and opaque in strm must be initialized + before the call. If zalloc and zfree are Z_NULL, then the default library- + derived memory allocation routines are used. windowBits is the base two + logarithm of the window size, in the range 8..15. window is a caller + supplied buffer of that size. Except for special applications where it is + assured that deflate was used with small window sizes, windowBits must be 15 + and a 32K byte window must be supplied to be able to decompress general + deflate streams. + + See inflateBack() for the usage of these routines. + + inflateBackInit will return Z_OK on success, Z_STREAM_ERROR if any of + the paramaters are invalid, Z_MEM_ERROR if the internal state could not be + allocated, or Z_VERSION_ERROR if the version of the library does not match + the version of the header file. +*/ + +typedef unsigned (*in_func) OF((void FAR *, unsigned char FAR * FAR *)); +typedef int (*out_func) OF((void FAR *, unsigned char FAR *, unsigned)); + +ZEXTERN int ZEXPORT inflateBack OF((z_streamp strm, + in_func in, void FAR *in_desc, + out_func out, void FAR *out_desc)); +/* + inflateBack() does a raw inflate with a single call using a call-back + interface for input and output. This is more efficient than inflate() for + file i/o applications in that it avoids copying between the output and the + sliding window by simply making the window itself the output buffer. This + function trusts the application to not change the output buffer passed by + the output function, at least until inflateBack() returns. + + inflateBackInit() must be called first to allocate the internal state + and to initialize the state with the user-provided window buffer. + inflateBack() may then be used multiple times to inflate a complete, raw + deflate stream with each call. inflateBackEnd() is then called to free the + allocated state. + + A raw deflate stream is one with no zlib or gzip header or trailer. + This routine would normally be used in a utility that reads zip or gzip + files and writes out uncompressed files. The utility would decode the + header and process the trailer on its own, hence this routine expects only + the raw deflate stream to decompress. This is different from the normal + behavior of inflate(), which expects either a zlib or gzip header and + trailer around the deflate stream. + + inflateBack() uses two subroutines supplied by the caller that are then + called by inflateBack() for input and output. inflateBack() calls those + routines until it reads a complete deflate stream and writes out all of the + uncompressed data, or until it encounters an error. The function's + parameters and return types are defined above in the in_func and out_func + typedefs. inflateBack() will call in(in_desc, &buf) which should return the + number of bytes of provided input, and a pointer to that input in buf. If + there is no input available, in() must return zero--buf is ignored in that + case--and inflateBack() will return a buffer error. inflateBack() will call + out(out_desc, buf, len) to write the uncompressed data buf[0..len-1]. out() + should return zero on success, or non-zero on failure. If out() returns + non-zero, inflateBack() will return with an error. Neither in() nor out() + are permitted to change the contents of the window provided to + inflateBackInit(), which is also the buffer that out() uses to write from. + The length written by out() will be at most the window size. Any non-zero + amount of input may be provided by in(). + + For convenience, inflateBack() can be provided input on the first call by + setting strm->next_in and strm->avail_in. If that input is exhausted, then + in() will be called. Therefore strm->next_in must be initialized before + calling inflateBack(). If strm->next_in is Z_NULL, then in() will be called + immediately for input. If strm->next_in is not Z_NULL, then strm->avail_in + must also be initialized, and then if strm->avail_in is not zero, input will + initially be taken from strm->next_in[0 .. strm->avail_in - 1]. + + The in_desc and out_desc parameters of inflateBack() is passed as the + first parameter of in() and out() respectively when they are called. These + descriptors can be optionally used to pass any information that the caller- + supplied in() and out() functions need to do their job. + + On return, inflateBack() will set strm->next_in and strm->avail_in to + pass back any unused input that was provided by the last in() call. The + return values of inflateBack() can be Z_STREAM_END on success, Z_BUF_ERROR + if in() or out() returned an error, Z_DATA_ERROR if there was a format error + in the deflate stream (in which case strm->msg is set to indicate the nature + of the error), or Z_STREAM_ERROR if the stream was not properly initialized. + In the case of Z_BUF_ERROR, an input or output error can be distinguished + using strm->next_in which will be Z_NULL only if in() returned an error. If + strm->next_in is not Z_NULL, then the Z_BUF_ERROR was due to out() returning + non-zero. (in() will always be called before out(), so strm->next_in is + assured to be defined if out() returns non-zero.) Note that inflateBack() + cannot return Z_OK. +*/ + +ZEXTERN int ZEXPORT inflateBackEnd OF((z_streamp strm)); +/* + All memory allocated by inflateBackInit() is freed. + + inflateBackEnd() returns Z_OK on success, or Z_STREAM_ERROR if the stream + state was inconsistent. +*/ + +ZEXTERN uLong ZEXPORT zlibCompileFlags OF((void)); +/* Return flags indicating compile-time options. + + Type sizes, two bits each, 00 = 16 bits, 01 = 32, 10 = 64, 11 = other: + 1.0: size of uInt + 3.2: size of uLong + 5.4: size of voidpf (pointer) + 7.6: size of z_off_t + + Compiler, assembler, and debug options: + 8: DEBUG + 9: ASMV or ASMINF -- use ASM code + 10: ZLIB_WINAPI -- exported functions use the WINAPI calling convention + 11: 0 (reserved) + + One-time table building (smaller code, but not thread-safe if true): + 12: BUILDFIXED -- build static block decoding tables when needed + 13: DYNAMIC_CRC_TABLE -- build CRC calculation tables when needed + 14,15: 0 (reserved) + + Library content (indicates missing functionality): + 16: NO_GZCOMPRESS -- gz* functions cannot compress (to avoid linking + deflate code when not needed) + 17: NO_GZIP -- deflate can't write gzip streams, and inflate can't detect + and decode gzip streams (to avoid linking crc code) + 18-19: 0 (reserved) + + Operation variations (changes in library functionality): + 20: PKZIP_BUG_WORKAROUND -- slightly more permissive inflate + 21: FASTEST -- deflate algorithm with only one, lowest compression level + 22,23: 0 (reserved) + + The sprintf variant used by gzprintf (zero is best): + 24: 0 = vs*, 1 = s* -- 1 means limited to 20 arguments after the format + 25: 0 = *nprintf, 1 = *printf -- 1 means gzprintf() not secure! + 26: 0 = returns value, 1 = void -- 1 means inferred string length returned + + Remainder: + 27-31: 0 (reserved) + */ + + + /* utility functions */ + +/* + The following utility functions are implemented on top of the basic + stream-oriented functions. To simplify the interface, some default options + are assumed (compression level and memory usage, standard memory allocation + functions). The source code of these utility functions can be modified if + you need special options. +*/ + +ZEXTERN int ZEXPORT compress OF((Bytef *dest, uLongf *destLen, + const Bytef *source, uLong sourceLen)); +/* + Compresses the source buffer into the destination buffer. sourceLen is + the byte length of the source buffer. Upon entry, destLen is the total size + of the destination buffer, which must be at least the value returned by + compressBound(sourceLen). Upon exit, destLen is the actual size of the + compressed buffer. + + compress returns Z_OK if success, Z_MEM_ERROR if there was not + enough memory, Z_BUF_ERROR if there was not enough room in the output + buffer. +*/ + +ZEXTERN int ZEXPORT compress2 OF((Bytef *dest, uLongf *destLen, + const Bytef *source, uLong sourceLen, + int level)); +/* + Compresses the source buffer into the destination buffer. The level + parameter has the same meaning as in deflateInit. sourceLen is the byte + length of the source buffer. Upon entry, destLen is the total size of the + destination buffer, which must be at least the value returned by + compressBound(sourceLen). Upon exit, destLen is the actual size of the + compressed buffer. + + compress2 returns Z_OK if success, Z_MEM_ERROR if there was not enough + memory, Z_BUF_ERROR if there was not enough room in the output buffer, + Z_STREAM_ERROR if the level parameter is invalid. +*/ + +ZEXTERN uLong ZEXPORT compressBound OF((uLong sourceLen)); +/* + compressBound() returns an upper bound on the compressed size after + compress() or compress2() on sourceLen bytes. It would be used before a + compress() or compress2() call to allocate the destination buffer. +*/ + +ZEXTERN int ZEXPORT uncompress OF((Bytef *dest, uLongf *destLen, + const Bytef *source, uLong sourceLen)); +/* + Decompresses the source buffer into the destination buffer. sourceLen is + the byte length of the source buffer. Upon entry, destLen is the total size + of the destination buffer, which must be large enough to hold the entire + uncompressed data. (The size of the uncompressed data must have been saved + previously by the compressor and transmitted to the decompressor by some + mechanism outside the scope of this compression library.) Upon exit, destLen + is the actual size of the uncompressed buffer. + + uncompress returns Z_OK if success, Z_MEM_ERROR if there was not + enough memory, Z_BUF_ERROR if there was not enough room in the output + buffer, or Z_DATA_ERROR if the input data was corrupted or incomplete. +*/ + + + /* gzip file access functions */ + +/* + This library supports reading and writing files in gzip (.gz) format with + an interface similar to that of stdio, using the functions that start with + "gz". The gzip format is different from the zlib format. gzip is a gzip + wrapper, documented in RFC 1952, wrapped around a deflate stream. +*/ + +typedef voidp gzFile; /* opaque gzip file descriptor */ + +/* +ZEXTERN gzFile ZEXPORT gzopen OF((const char *path, const char *mode)); + + Opens a gzip (.gz) file for reading or writing. The mode parameter is as + in fopen ("rb" or "wb") but can also include a compression level ("wb9") or + a strategy: 'f' for filtered data as in "wb6f", 'h' for Huffman-only + compression as in "wb1h", 'R' for run-length encoding as in "wb1R", or 'F' + for fixed code compression as in "wb9F". (See the description of + deflateInit2 for more information about the strategy parameter.) Also "a" + can be used instead of "w" to request that the gzip stream that will be + written be appended to the file. "+" will result in an error, since reading + and writing to the same gzip file is not supported. + + gzopen can be used to read a file which is not in gzip format; in this + case gzread will directly read from the file without decompression. + + gzopen returns NULL if the file could not be opened, if there was + insufficient memory to allocate the gzFile state, or if an invalid mode was + specified (an 'r', 'w', or 'a' was not provided, or '+' was provided). + errno can be checked to determine if the reason gzopen failed was that the + file could not be opened. +*/ + +ZEXTERN gzFile ZEXPORT gzdopen OF((int fd, const char *mode)); +/* + gzdopen associates a gzFile with the file descriptor fd. File descriptors + are obtained from calls like open, dup, creat, pipe or fileno (if the file + has been previously opened with fopen). The mode parameter is as in gzopen. + + The next call of gzclose on the returned gzFile will also close the file + descriptor fd, just like fclose(fdopen(fd, mode)) closes the file descriptor + fd. If you want to keep fd open, use fd = dup(fd_keep); gz = gzdopen(fd, + mode);. The duplicated descriptor should be saved to avoid a leak, since + gzdopen does not close fd if it fails. + + gzdopen returns NULL if there was insufficient memory to allocate the + gzFile state, if an invalid mode was specified (an 'r', 'w', or 'a' was not + provided, or '+' was provided), or if fd is -1. The file descriptor is not + used until the next gz* read, write, seek, or close operation, so gzdopen + will not detect if fd is invalid (unless fd is -1). +*/ + +ZEXTERN int ZEXPORT gzbuffer OF((gzFile file, unsigned size)); +/* + Set the internal buffer size used by this library's functions. The + default buffer size is 8192 bytes. This function must be called after + gzopen() or gzdopen(), and before any other calls that read or write the + file. The buffer memory allocation is always deferred to the first read or + write. Two buffers are allocated, either both of the specified size when + writing, or one of the specified size and the other twice that size when + reading. A larger buffer size of, for example, 64K or 128K bytes will + noticeably increase the speed of decompression (reading). + + The new buffer size also affects the maximum length for gzprintf(). + + gzbuffer() returns 0 on success, or -1 on failure, such as being called + too late. +*/ + +ZEXTERN int ZEXPORT gzsetparams OF((gzFile file, int level, int strategy)); +/* + Dynamically update the compression level or strategy. See the description + of deflateInit2 for the meaning of these parameters. + + gzsetparams returns Z_OK if success, or Z_STREAM_ERROR if the file was not + opened for writing. +*/ + +ZEXTERN int ZEXPORT gzread OF((gzFile file, voidp buf, unsigned len)); +/* + Reads the given number of uncompressed bytes from the compressed file. If + the input file was not in gzip format, gzread copies the given number of + bytes into the buffer. + + After reaching the end of a gzip stream in the input, gzread will continue + to read, looking for another gzip stream, or failing that, reading the rest + of the input file directly without decompression. The entire input file + will be read if gzread is called until it returns less than the requested + len. + + gzread returns the number of uncompressed bytes actually read, less than + len for end of file, or -1 for error. +*/ + +ZEXTERN int ZEXPORT gzwrite OF((gzFile file, + voidpc buf, unsigned len)); +/* + Writes the given number of uncompressed bytes into the compressed file. + gzwrite returns the number of uncompressed bytes written or 0 in case of + error. +*/ + +ZEXTERN int ZEXPORTVA gzprintf OF((gzFile file, const char *format, ...)); +/* + Converts, formats, and writes the arguments to the compressed file under + control of the format string, as in fprintf. gzprintf returns the number of + uncompressed bytes actually written, or 0 in case of error. The number of + uncompressed bytes written is limited to 8191, or one less than the buffer + size given to gzbuffer(). The caller should assure that this limit is not + exceeded. If it is exceeded, then gzprintf() will return an error (0) with + nothing written. In this case, there may also be a buffer overflow with + unpredictable consequences, which is possible only if zlib was compiled with + the insecure functions sprintf() or vsprintf() because the secure snprintf() + or vsnprintf() functions were not available. This can be determined using + zlibCompileFlags(). +*/ + +ZEXTERN int ZEXPORT gzputs OF((gzFile file, const char *s)); +/* + Writes the given null-terminated string to the compressed file, excluding + the terminating null character. + + gzputs returns the number of characters written, or -1 in case of error. +*/ + +ZEXTERN char * ZEXPORT gzgets OF((gzFile file, char *buf, int len)); +/* + Reads bytes from the compressed file until len-1 characters are read, or a + newline character is read and transferred to buf, or an end-of-file + condition is encountered. If any characters are read or if len == 1, the + string is terminated with a null character. If no characters are read due + to an end-of-file or len < 1, then the buffer is left untouched. + + gzgets returns buf which is a null-terminated string, or it returns NULL + for end-of-file or in case of error. If there was an error, the contents at + buf are indeterminate. +*/ + +ZEXTERN int ZEXPORT gzputc OF((gzFile file, int c)); +/* + Writes c, converted to an unsigned char, into the compressed file. gzputc + returns the value that was written, or -1 in case of error. +*/ + +ZEXTERN int ZEXPORT gzgetc OF((gzFile file)); +/* + Reads one byte from the compressed file. gzgetc returns this byte or -1 + in case of end of file or error. +*/ + +ZEXTERN int ZEXPORT gzungetc OF((int c, gzFile file)); +/* + Push one character back onto the stream to be read as the first character + on the next read. At least one character of push-back is allowed. + gzungetc() returns the character pushed, or -1 on failure. gzungetc() will + fail if c is -1, and may fail if a character has been pushed but not read + yet. If gzungetc is used immediately after gzopen or gzdopen, at least the + output buffer size of pushed characters is allowed. (See gzbuffer above.) + The pushed character will be discarded if the stream is repositioned with + gzseek() or gzrewind(). +*/ + +ZEXTERN int ZEXPORT gzflush OF((gzFile file, int flush)); +/* + Flushes all pending output into the compressed file. The parameter flush + is as in the deflate() function. The return value is the zlib error number + (see function gzerror below). gzflush is only permitted when writing. + + If the flush parameter is Z_FINISH, the remaining data is written and the + gzip stream is completed in the output. If gzwrite() is called again, a new + gzip stream will be started in the output. gzread() is able to read such + concatented gzip streams. + + gzflush should be called only when strictly necessary because it will + degrade compression if called too often. +*/ + +/* +ZEXTERN z_off_t ZEXPORT gzseek OF((gzFile file, + z_off_t offset, int whence)); + + Sets the starting position for the next gzread or gzwrite on the given + compressed file. The offset represents a number of bytes in the + uncompressed data stream. The whence parameter is defined as in lseek(2); + the value SEEK_END is not supported. + + If the file is opened for reading, this function is emulated but can be + extremely slow. If the file is opened for writing, only forward seeks are + supported; gzseek then compresses a sequence of zeroes up to the new + starting position. + + gzseek returns the resulting offset location as measured in bytes from + the beginning of the uncompressed stream, or -1 in case of error, in + particular if the file is opened for writing and the new starting position + would be before the current position. +*/ + +ZEXTERN int ZEXPORT gzrewind OF((gzFile file)); +/* + Rewinds the given file. This function is supported only for reading. + + gzrewind(file) is equivalent to (int)gzseek(file, 0L, SEEK_SET) +*/ + +/* +ZEXTERN z_off_t ZEXPORT gztell OF((gzFile file)); + + Returns the starting position for the next gzread or gzwrite on the given + compressed file. This position represents a number of bytes in the + uncompressed data stream, and is zero when starting, even if appending or + reading a gzip stream from the middle of a file using gzdopen(). + + gztell(file) is equivalent to gzseek(file, 0L, SEEK_CUR) +*/ + +/* +ZEXTERN z_off_t ZEXPORT gzoffset OF((gzFile file)); + + Returns the current offset in the file being read or written. This offset + includes the count of bytes that precede the gzip stream, for example when + appending or when using gzdopen() for reading. When reading, the offset + does not include as yet unused buffered input. This information can be used + for a progress indicator. On error, gzoffset() returns -1. +*/ + +ZEXTERN int ZEXPORT gzeof OF((gzFile file)); +/* + Returns true (1) if the end-of-file indicator has been set while reading, + false (0) otherwise. Note that the end-of-file indicator is set only if the + read tried to go past the end of the input, but came up short. Therefore, + just like feof(), gzeof() may return false even if there is no more data to + read, in the event that the last read request was for the exact number of + bytes remaining in the input file. This will happen if the input file size + is an exact multiple of the buffer size. + + If gzeof() returns true, then the read functions will return no more data, + unless the end-of-file indicator is reset by gzclearerr() and the input file + has grown since the previous end of file was detected. +*/ + +ZEXTERN int ZEXPORT gzdirect OF((gzFile file)); +/* + Returns true (1) if file is being copied directly while reading, or false + (0) if file is a gzip stream being decompressed. This state can change from + false to true while reading the input file if the end of a gzip stream is + reached, but is followed by data that is not another gzip stream. + + If the input file is empty, gzdirect() will return true, since the input + does not contain a gzip stream. + + If gzdirect() is used immediately after gzopen() or gzdopen() it will + cause buffers to be allocated to allow reading the file to determine if it + is a gzip file. Therefore if gzbuffer() is used, it should be called before + gzdirect(). +*/ + +ZEXTERN int ZEXPORT gzclose OF((gzFile file)); +/* + Flushes all pending output if necessary, closes the compressed file and + deallocates the (de)compression state. Note that once file is closed, you + cannot call gzerror with file, since its structures have been deallocated. + gzclose must not be called more than once on the same file, just as free + must not be called more than once on the same allocation. + + gzclose will return Z_STREAM_ERROR if file is not valid, Z_ERRNO on a + file operation error, or Z_OK on success. +*/ + +ZEXTERN int ZEXPORT gzclose_r OF((gzFile file)); +ZEXTERN int ZEXPORT gzclose_w OF((gzFile file)); +/* + Same as gzclose(), but gzclose_r() is only for use when reading, and + gzclose_w() is only for use when writing or appending. The advantage to + using these instead of gzclose() is that they avoid linking in zlib + compression or decompression code that is not used when only reading or only + writing respectively. If gzclose() is used, then both compression and + decompression code will be included the application when linking to a static + zlib library. +*/ + +ZEXTERN const char * ZEXPORT gzerror OF((gzFile file, int *errnum)); +/* + Returns the error message for the last error which occurred on the given + compressed file. errnum is set to zlib error number. If an error occurred + in the file system and not in the compression library, errnum is set to + Z_ERRNO and the application may consult errno to get the exact error code. + + The application must not modify the returned string. Future calls to + this function may invalidate the previously returned string. If file is + closed, then the string previously returned by gzerror will no longer be + available. + + gzerror() should be used to distinguish errors from end-of-file for those + functions above that do not distinguish those cases in their return values. +*/ + +ZEXTERN void ZEXPORT gzclearerr OF((gzFile file)); +/* + Clears the error and end-of-file flags for file. This is analogous to the + clearerr() function in stdio. This is useful for continuing to read a gzip + file that is being written concurrently. +*/ + + + /* checksum functions */ + +/* + These functions are not related to compression but are exported + anyway because they might be useful in applications using the compression + library. +*/ + +ZEXTERN uLong ZEXPORT adler32 OF((uLong adler, const Bytef *buf, uInt len)); +/* + Update a running Adler-32 checksum with the bytes buf[0..len-1] and + return the updated checksum. If buf is Z_NULL, this function returns the + required initial value for the checksum. + + An Adler-32 checksum is almost as reliable as a CRC32 but can be computed + much faster. + + Usage example: + + uLong adler = adler32(0L, Z_NULL, 0); + + while (read_buffer(buffer, length) != EOF) { + adler = adler32(adler, buffer, length); + } + if (adler != original_adler) error(); +*/ + +/* +ZEXTERN uLong ZEXPORT adler32_combine OF((uLong adler1, uLong adler2, + z_off_t len2)); + + Combine two Adler-32 checksums into one. For two sequences of bytes, seq1 + and seq2 with lengths len1 and len2, Adler-32 checksums were calculated for + each, adler1 and adler2. adler32_combine() returns the Adler-32 checksum of + seq1 and seq2 concatenated, requiring only adler1, adler2, and len2. +*/ + +ZEXTERN uLong ZEXPORT crc32 OF((uLong crc, const Bytef *buf, uInt len)); +/* + Update a running CRC-32 with the bytes buf[0..len-1] and return the + updated CRC-32. If buf is Z_NULL, this function returns the required + initial value for the for the crc. Pre- and post-conditioning (one's + complement) is performed within this function so it shouldn't be done by the + application. + + Usage example: + + uLong crc = crc32(0L, Z_NULL, 0); + + while (read_buffer(buffer, length) != EOF) { + crc = crc32(crc, buffer, length); + } + if (crc != original_crc) error(); +*/ + +/* +ZEXTERN uLong ZEXPORT crc32_combine OF((uLong crc1, uLong crc2, z_off_t len2)); + + Combine two CRC-32 check values into one. For two sequences of bytes, + seq1 and seq2 with lengths len1 and len2, CRC-32 check values were + calculated for each, crc1 and crc2. crc32_combine() returns the CRC-32 + check value of seq1 and seq2 concatenated, requiring only crc1, crc2, and + len2. +*/ + + + /* various hacks, don't look :) */ + +/* deflateInit and inflateInit are macros to allow checking the zlib version + * and the compiler's view of z_stream: + */ +ZEXTERN int ZEXPORT deflateInit_ OF((z_streamp strm, int level, + const char *version, int stream_size)); +ZEXTERN int ZEXPORT inflateInit_ OF((z_streamp strm, + const char *version, int stream_size)); +ZEXTERN int ZEXPORT deflateInit2_ OF((z_streamp strm, int level, int method, + int windowBits, int memLevel, + int strategy, const char *version, + int stream_size)); +ZEXTERN int ZEXPORT inflateInit2_ OF((z_streamp strm, int windowBits, + const char *version, int stream_size)); +ZEXTERN int ZEXPORT inflateBackInit_ OF((z_streamp strm, int windowBits, + unsigned char FAR *window, + const char *version, + int stream_size)); +#define deflateInit(strm, level) \ + deflateInit_((strm), (level), ZLIB_VERSION, sizeof(z_stream)) +#define inflateInit(strm) \ + inflateInit_((strm), ZLIB_VERSION, sizeof(z_stream)) +#define deflateInit2(strm, level, method, windowBits, memLevel, strategy) \ + deflateInit2_((strm),(level),(method),(windowBits),(memLevel),\ + (strategy), ZLIB_VERSION, sizeof(z_stream)) +#define inflateInit2(strm, windowBits) \ + inflateInit2_((strm), (windowBits), ZLIB_VERSION, sizeof(z_stream)) +#define inflateBackInit(strm, windowBits, window) \ + inflateBackInit_((strm), (windowBits), (window), \ + ZLIB_VERSION, sizeof(z_stream)) + +/* provide 64-bit offset functions if _LARGEFILE64_SOURCE defined, and/or + * change the regular functions to 64 bits if _FILE_OFFSET_BITS is 64 (if + * both are true, the application gets the *64 functions, and the regular + * functions are changed to 64 bits) -- in case these are set on systems + * without large file support, _LFS64_LARGEFILE must also be true + */ +#if defined(_LARGEFILE64_SOURCE) && _LFS64_LARGEFILE-0 + ZEXTERN gzFile ZEXPORT gzopen64 OF((const char *, const char *)); + ZEXTERN z_off64_t ZEXPORT gzseek64 OF((gzFile, z_off64_t, int)); + ZEXTERN z_off64_t ZEXPORT gztell64 OF((gzFile)); + ZEXTERN z_off64_t ZEXPORT gzoffset64 OF((gzFile)); + ZEXTERN uLong ZEXPORT adler32_combine64 OF((uLong, uLong, z_off64_t)); + ZEXTERN uLong ZEXPORT crc32_combine64 OF((uLong, uLong, z_off64_t)); +#endif + +#if !defined(ZLIB_INTERNAL) && _FILE_OFFSET_BITS-0 == 64 && _LFS64_LARGEFILE-0 +# define gzopen gzopen64 +# define gzseek gzseek64 +# define gztell gztell64 +# define gzoffset gzoffset64 +# define adler32_combine adler32_combine64 +# define crc32_combine crc32_combine64 +# ifdef _LARGEFILE64_SOURCE + ZEXTERN gzFile ZEXPORT gzopen64 OF((const char *, const char *)); + ZEXTERN z_off_t ZEXPORT gzseek64 OF((gzFile, z_off_t, int)); + ZEXTERN z_off_t ZEXPORT gztell64 OF((gzFile)); + ZEXTERN z_off_t ZEXPORT gzoffset64 OF((gzFile)); + ZEXTERN uLong ZEXPORT adler32_combine64 OF((uLong, uLong, z_off_t)); + ZEXTERN uLong ZEXPORT crc32_combine64 OF((uLong, uLong, z_off_t)); +# endif +#else + ZEXTERN gzFile ZEXPORT gzopen OF((const char *, const char *)); + ZEXTERN z_off_t ZEXPORT gzseek OF((gzFile, z_off_t, int)); + ZEXTERN z_off_t ZEXPORT gztell OF((gzFile)); + ZEXTERN z_off_t ZEXPORT gzoffset OF((gzFile)); + ZEXTERN uLong ZEXPORT adler32_combine OF((uLong, uLong, z_off_t)); + ZEXTERN uLong ZEXPORT crc32_combine OF((uLong, uLong, z_off_t)); +#endif + +/* hack for buggy compilers */ +#if !defined(ZUTIL_H) && !defined(NO_DUMMY_DECL) + struct internal_state {int dummy;}; +#endif + +/* undocumented functions */ +ZEXTERN const char * ZEXPORT zError OF((int)); +ZEXTERN int ZEXPORT inflateSyncPoint OF((z_streamp)); +ZEXTERN const uLongf * ZEXPORT get_crc_table OF((void)); +ZEXTERN int ZEXPORT inflateUndermine OF((z_streamp, int)); + +#ifdef __cplusplus +} +#endif + +#endif /* ZLIB_H */ |