added conditional code in web2c to support xetex

git-svn-id: svn://tug.org/texlive/trunk@1708 c570f23f-e606-0410-a88d-b1316a301751

5 files changed, 246 insertions, 8 deletions

diff --git a/Build/source/texk/web2c/ChangeLog b/Build/source/texk/web2c/ChangeLog
index 8571a64a2e9..c3f95eddda3 100644
--- a/Build/source/texk/web2c/ChangeLog
+++ b/Build/source/texk/web2c/ChangeLog
@@ -1,3 +1,8 @@
+2006-06-20  Jonathan Kew  <jonathan_kew@sil.org>
+
+	* texmfmem.h, texmfmp.h, texmfmp-help.h, lib/texmfmp.c: Added
+	conditional fragments needed for building XeTeX.
+
 2005-09-02  Olaf Weber  <olaf@infovore.xs4all.nl>
 
 	* tex.ch [53.1370]: Since we generate a string in all cases, we
diff --git a/Build/source/texk/web2c/lib/texmfmp.c b/Build/source/texk/web2c/lib/texmfmp.c
index 8511256ba11..0671bac0f0e 100644
--- a/Build/source/texk/web2c/lib/texmfmp.c
+++ b/Build/source/texk/web2c/lib/texmfmp.c
@@ -38,7 +38,9 @@
    Unfortunately there's no way to get the banner into this code, so
    just repeat the text.  */
 #ifdef TeX
-#if defined (eTeX)
+#if defined(XeTeX)
+#include <xetexdir/xetexextra.h>
+#elif defined (eTeX)
 #include <etexdir/etexextra.h>
 #elif defined (pdfTeX)
 #include <pdftexdir/pdftexextra.h>
@@ -103,6 +105,11 @@
 
 /* The main program, etc.  */
 
+#ifdef XeTeX
+#include "xetexdir/XeTeX_ext.h"
+#endif
+
+
 /* What we were invoked as and with.  */
 char **argv;
 int argc;
@@ -260,10 +267,12 @@ maininit P2C(int, ac, string *, av)
     if (mltexp) {
       fprintf(stderr, "-mltex only works with -ini\n");
     }
+#if !defined(XeTeX)
     if (enctexp) {
       fprintf(stderr, "-enc only works with -ini\n");
     }
 #endif
+#endif
 #if defined(eTeX) || defined(pdfeTeX) || defined(Aleph)
     if (etexp) {
       fprintf(stderr, "-etex only works with -ini\n");
@@ -354,17 +363,56 @@ topenin P1H(void)
 {
   int i;
 
+#ifdef XeTeX
+  static UFILE termin_file;
+  if (termin == 0) {
+    termin = &termin_file;
+    termin->f = stdin;
+    termin->savedChar = -1;
+    termin->skipNextLF = 0;
+    termin->encodingMode = UTF8;
+    termin->conversionData = 0;
+    inputfile[0] = termin;
+  }
+#endif
+
   buffer[first] = 0; /* In case there are no arguments.  */
 
   if (optind < argc) { /* We have command line arguments.  */
     int k = first;
     for (i = optind; i < argc; i++) {
+#ifdef XeTeX
+      unsigned char *ptr = (unsigned char *)&(argv[i][0]);
+      /* need to interpret UTF8 from the command line */
+      UInt32 rval;
+      while (rval = *(ptr++)) {
+        UInt16 extraBytes = bytesFromUTF8[rval];
+        switch (extraBytes) { /* note: code falls through cases! */
+          case 5: rval <<= 6; if (*ptr) rval += *(ptr++);
+          case 4: rval <<= 6; if (*ptr) rval += *(ptr++);
+          case 3: rval <<= 6; if (*ptr) rval += *(ptr++);
+          case 2: rval <<= 6; if (*ptr) rval += *(ptr++);
+          case 1: rval <<= 6; if (*ptr) rval += *(ptr++);
+          case 0: ;
+        };
+        rval -= offsetsFromUTF8[extraBytes];
+        /* now rval is a USV; if it's >=64K, we need to put surrogates in the buffer */
+        if (rval > 0xFFFF) {
+          rval -= 0x10000;
+          buffer[k++] = 0xd800 + rval / 0x0400;
+          buffer[k++] = 0xdc00 + rval % 0x0400;
+        }
+        else
+          buffer[k++] = rval;
+      }
+#else
       char *ptr = &(argv[i][0]);
       /* Don't use strcat, since in Omega the buffer elements aren't
          single bytes.  */
       while (*ptr) {
         buffer[k++] = *(ptr++);
       }
+#endif
       buffer[k++] = ' ';
     }
     argc = 0;	/* Don't do this again.  */
@@ -384,12 +432,13 @@ topenin P1H(void)
 
   /* One more time, this time converting to TeX's internal character
      representation.  */
-#if !defined(Omega) && !defined(eOmega) && !defined(Aleph)
+#if !defined(Omega) && !defined(eOmega) && !defined(Aleph) && !defined(XeTeX)
   for (i = first; i < last; i++)
     buffer[i] = xord[buffer[i]];
 #endif
 }
 
+
 /* IPC for TeX.  By Tom Rokicki for the NeXT; it makes TeX ship out the
    DVI file in a pipe to TeXView so that the output can be displayed
    incrementally.  Shamim Mohamed adapted it for Web2c.  */
@@ -583,7 +632,7 @@ ipcpage P1C(int, is_eof)
 
 #if defined (TeX) || defined (MF) || defined (MP)
   /* TCX and Omega get along like sparks and gunpowder. */
-#if !defined(Omega) && !defined(eOmega) && !defined(Aleph)
+#if !defined(Omega) && !defined(eOmega) && !defined(Aleph) && !defined(XeTeX)
 
 /* Return the next number following START, setting POST to the following
    character, as in strtol.  Issue a warning and return -1 if no number
@@ -684,7 +733,7 @@ readtcxfile P1H(void)
     WARNING1 ("Could not open char translation file `%s'", orig_filename);
   }
 }
-#endif /* !Omega && !eOmega && !Aleph */
+#endif /* !Omega && !eOmega && !Aleph && !XeTeX */
 #endif /* TeX || MF || MP [character translation] */
 
 /* Normalize quoting of filename -- that is, only quote if there is a space,
@@ -772,7 +821,9 @@ static struct option long_options[]
 #endif /* IPC */
 #if !defined(Omega) && !defined(eOmega) && !defined(Aleph)
       { "mltex",                     0, &mltexp, 1 },
+#if !defined(XeTeX)
       { "enc",                       0, &enctexp, 1 },
+#endif /* !XeTeX */
 #endif /* !Omega && !eOmega && !Aleph */
 #if defined (eTeX) || defined(pdfeTeX) || defined(Aleph)
       { "etex",                      0, &etexp, 1 },
@@ -799,6 +850,11 @@ static struct option long_options[]
       { "translate-file",            1, 0, 0 },
       { "default-translate-file",    1, 0, 0 },
       { "8bit",                      0, &eightbitp, 1 },
+#if defined(XeTeX)
+      { "no-pdf",                 0, &nopdfoutput, 1 },
+      { "output-driver",          1, 0, 0 },
+      { "papersize",              1, 0, 0 },
+#endif /* XeTeX */
 #endif /* TeX || MF || MP */
 #if defined (TeX) || defined (MF)
       { "mktex",                     1, 0, 0 },
@@ -834,6 +890,13 @@ parse_options P2C(int, argc,  string *, argv)
     if (ARGUMENT_IS ("kpathsea-debug")) {
       kpathsea_debug |= atoi (optarg);
 
+#ifdef XeTeX
+    } else if (ARGUMENT_IS ("papersize")) {
+      papersize = optarg;
+    } else if (ARGUMENT_IS ("output-driver")) {
+      outputdriver = optarg;
+#endif
+
     } else if (ARGUMENT_IS ("progname")) {
       user_progname = optarg;
 
@@ -1436,6 +1499,7 @@ getrandomseed()
    to eof.  Otherwise, we return `true' and set last = first +
    length(line except trailing whitespace).  */
 
+#ifndef XeTeX /* for XeTeX, we have a replacement function in XeTeX_ext.c */
 boolean
 input_line P1C(FILE *, f)
 {
@@ -1481,6 +1545,7 @@ input_line P1C(FILE *, f)
 
     return true;
 }
+#endif /* !XeTeX */
 
 /* This string specifies what the `e' option does in response to an
    error message.  */ 
@@ -1506,7 +1571,11 @@ calledit P4C(packedASCIIcode *, filename,
 
   /* Close any open input files, since we're going to kill the job.  */
   for (i = 1; i <= inopen; i++)
+#ifdef XeTeX
+    xfclose (inputfile[i]->f, "inputfile");
+#else
     xfclose (inputfile[i], "inputfile");
+#endif
 
   /* Replace the default with the value of the appropriate environment
      variable or config file value, if it's set.  */
@@ -1737,15 +1806,45 @@ checkpoolpointer (poolpointer poolptr, size_t len)
 }
 
 #if !defined(pdfTeX) && !defined(pdfeTeX)
-static int
+#ifndef XeTeX	/* XeTeX uses this from XeTeX_mac.c */
+static
+#endif
+int
 maketexstring(const_string s)
 {
   size_t len;
+#ifdef XeTeX
+  UInt32 rval;
+  unsigned char* cp = (unsigned char*)s;
+#endif
   assert (s != 0);
   len = strlen(s);
-  checkpoolpointer (poolptr, len);
+  checkpoolpointer (poolptr, len); /* in the XeTeX case, this may be more than enough */
+#ifdef XeTeX
+  while (rval = *(cp++)) {
+  UInt16 extraBytes = bytesFromUTF8[rval];
+  switch (extraBytes) { /* note: code falls through cases! */
+    case 5: rval <<= 6; if (*cp) rval += *(cp++);
+    case 4: rval <<= 6; if (*cp) rval += *(cp++);
+    case 3: rval <<= 6; if (*cp) rval += *(cp++);
+    case 2: rval <<= 6; if (*cp) rval += *(cp++);
+    case 1: rval <<= 6; if (*cp) rval += *(cp++);
+    case 0: ;
+  };
+  rval -= offsetsFromUTF8[extraBytes];
+  if (rval > 0xffff) {
+    rval -= 0x10000;
+    strpool[poolptr++] = 0xd800 + rval / 0x0400;
+    strpool[poolptr++] = 0xdc00 + rval % 0x0400;
+  }
+  else
+    strpool[poolptr++] = rval;
+  }
+#else
   while (len-- > 0)
     strpool[poolptr++] = *s++;
+#endif
+
   return (makestring());
 }
 #endif
@@ -1786,18 +1885,22 @@ compare_paths P2C(const_string, p1, const_string, p2)
   return ret;
 }
 
+#ifdef XeTeX
+#define strstartar strstart
+#endif
+
 string
 gettexstring P1C(strnumber, s)
 {
   poolpointer i, len;
   string name;
-#if !defined(Omega) && !defined(eOmega) && !defined(Aleph)
+#if !defined(Omega) && !defined(eOmega) && !defined(Aleph) && !defined(XeTeX)
   len = strstart[s + 1] - strstart[s];
 #else
   len = strstartar[s + 1 - 65536L] - strstartar[s - 65536L];
 #endif
   name = (string)xmalloc (len + 1);
-#if !defined(Omega) && !defined(eOmega) && !defined(Aleph)
+#if !defined(Omega) && !defined(eOmega) && !defined(Aleph) && !defined(XeTeX)
   strncpy (name, (string)&strpool[strstart[s]], len);
 #else
   /* Don't use strncpy.  The strpool is not made up of chars. */
@@ -1807,6 +1910,10 @@ gettexstring P1C(strnumber, s)
   return name;
 }
 
+#ifdef XeTeX
+#undef strstartar
+#endif
+
 boolean
 isnewsource P2C(strnumber, srcfilename, int, lineno)
 {
diff --git a/Build/source/texk/web2c/texmfmem.h b/Build/source/texk/web2c/texmfmem.h
index 32c41e08215..78bf73fb787 100644
--- a/Build/source/texk/web2c/texmfmem.h
+++ b/Build/source/texk/web2c/texmfmem.h
@@ -113,9 +113,24 @@ typedef union
   twohalves hhfield;
 #endif
 #ifdef WORDS_BIGENDIAN
+#ifdef XeTeX
+  struct
+  {
+    integer CINT;
+    integer CINT1;
+  } x;
+#else
   integer cint;
+#endif
   fourquarters qqqq;
 #else /* not WORDS_BIGENDIAN */
+#ifdef XeTeX
+  struct
+  {
+    integer CINT1;
+    integer CINT;
+  } x;
+#else
   struct
   {
 #if defined (TeX) && !defined (SMALLTeX) || defined (MF) && !defined (SMALLMF) || defined (MP) && !defined (SMALLMP)
@@ -123,12 +138,15 @@ typedef union
 #endif /* big {TeX,MF,MP} */
     integer CINT;
   } u;
+#endif
 
   struct
   {
+#ifndef XeTeX
 #if defined (TeX) && !defined (SMALLTeX) || defined (MF) && !defined (SMALLMF) || defined (MP) && !defined (SMALLMP)
     halfword junk;
 #endif /* big {TeX,MF,MP} */
+#endif
     fourquarters QQQQ;
   } v;
 #endif /* not WORDS_BIGENDIAN */
@@ -141,13 +159,28 @@ typedef union
 typedef union
 {
 #ifdef WORDS_BIGENDIAN
+#ifdef XeTeX
+  struct
+  {
+    integer CINT;
+  } x;
+#else
   integer cint;
+#endif
   fourquarters qqqq;
 #else /* not WORDS_BIGENDIAN */
+#ifdef XeTeX
+  struct
+  {
+    halfword junk;
+    integer CINT;
+  } x;
+#else
   struct
   {
     integer CINT;
   } u;
+#endif
 
   struct
   {
@@ -166,8 +199,16 @@ typedef union
 #define rh v.RH
 #define lhfield	v.LH
 
+#ifdef XeTeX
+#define cint  x.CINT
+#define cint1 x.CINT1
+#else
 #ifndef WORDS_BIGENDIAN
 #define cint u.CINT
+#endif
+#endif /* XeTeX */
+
+#ifndef WORDS_BIGENDIAN
 #define qqqq v.QQQQ
 #endif
 
diff --git a/Build/source/texk/web2c/texmfmp-help.h b/Build/source/texk/web2c/texmfmp-help.h
index c0c4525a1e4..942a882e0f4 100644
--- a/Build/source/texk/web2c/texmfmp-help.h
+++ b/Build/source/texk/web2c/texmfmp-help.h
@@ -450,6 +450,66 @@ const_string PDFETEXHELP[] = {
 };
 #endif /* pdfeTeX */
 
+#ifdef XeTeX
+const_string XETEXHELP[] = {
+    "Usage: xetex [OPTION]... [TEXNAME[.tex]] [COMMANDS]",
+    "   or: xetex [OPTION]... \\FIRST-LINE",
+    "   or: xetex [OPTION]... &FMT ARGS",
+    "  Run XeTeX on TEXNAME, usually creating TEXNAME.pdf.",
+    "  Any remaining COMMANDS are processed as XeTeX input, after TEXNAME is read.",
+    "  If the first line of TEXNAME is %&FMT, and FMT is an existing .fmt file,",
+    "  use it.  Else use `NAME.fmt', where NAME is the program invocation name,",
+    "  most commonly `xetex'.",
+    "",
+    "  Alternatively, if the first non-option argument begins with a backslash,",
+    "  interpret all non-option arguments as a line of XeTeX input.",
+    "",
+    "  Alternatively, if the first non-option argument begins with a &, the",
+    "  next word is taken as the FMT to read, overriding all else.  Any",
+    "  remaining arguments are processed as above.",
+    "",
+    "  If no arguments or options are specified, prompt for input.",
+    "",
+    "[-no]-file-line-error   disable/enable file:line:error style messages",
+    "-fmt=FMTNAME            use FMTNAME instead of program name or a %& line",
+    "-halt-on-error          stop processing at the first error",
+    "-ini                    be xeinitex, for dumping formats; this is implicitly",
+    "                          true if the program name is `xeinitex'",
+    "-interaction=STRING     set interaction mode (STRING=batchmode/nonstopmode/",
+    "                          scrollmode/errorstopmode)",
+    "-jobname=STRING         set the job name to STRING",
+    "-kpathsea-debug=NUMBER  set path searching debugging flags according to",
+    "                          the bits of NUMBER",
+    "[-no]-mktex=FMT         disable/enable mktexFMT generation (FMT=tex/tfm)",
+    "-mltex                  enable MLTeX extensions such as \\charsubdef",
+    "-output-comment=STRING  use STRING for XDV file comment instead of date",
+    "-output-directory=DIR   use DIR as the directory to write files to",
+    "-output-driver=CMD      use CMD as the XDV-to-PDF driver instead of "
+#ifdef XETEX_MAC
+                             "xdv2pdf"
+#else
+                             "xdvipdfmx"
+#endif
+                             ,
+    "-no-pdf                 generate XDV (extended DVI) output rather than PDF",
+    "[-no]-parse-first-line  disable/enable parsing of the first line of the",
+    "                          input file",
+    "-papersize=STRING       set PDF media size to STRING",
+    "-progname=STRING        set program (and fmt) name to STRING",
+    "-recorder               enable filename recorder",
+    "[-no]-shell-escape      disable/enable \\write18{SHELL COMMAND}",
+    "-src-specials           insert source specials into the XDV file",
+    "-src-specials=WHERE     insert source specials in certain places of",
+    "                          the XDV file. WHERE is a comma-separated value",
+    "                          list: cr display hbox math par parend vbox",
+    "-translate-file=TCXNAME (ignored)",
+    "-8bit                   make all characters printable, don't use ^^X sequences",
+    "-help                   display this help and exit",
+    "-version                output version information and exit",
+    NULL
+};
+#endif /* XeTeX */
+
 #ifdef TeX
 const_string TEXHELP[] = {
     "Usage: tex [OPTION]... [TEXNAME[.tex]] [COMMANDS]",
diff --git a/Build/source/texk/web2c/texmfmp.h b/Build/source/texk/web2c/texmfmp.h
index d000afc7cd7..64816111163 100644
--- a/Build/source/texk/web2c/texmfmp.h
+++ b/Build/source/texk/web2c/texmfmp.h
@@ -6,6 +6,21 @@
 #include <kpathsea/c-pathch.h> /* for IS_DIR_SEP, used in the change files */
 #include <kpathsea/tex-make.h> /* for kpse_make_tex_discard_errors */
 
+#ifdef XeTeX
+/* added typedefs for unicodefile and voidpointer */
+#define XETEX_UNICODE_FILE_DEFINED	1
+typedef struct {
+  FILE *f;
+  long  savedChar;
+  short skipNextLF;
+  short encodingMode;
+  void *conversionData;
+} UFILE;
+typedef UFILE* unicodefile;
+
+typedef void* voidpointer;
+#endif
+
 /* If we have these macros, use them, as they provide a better guide to
    the endianess when cross-compiling. */
 #if defined (BYTE_ORDER) && defined (BIG_ENDIAN) && defined (LITTLE_ENDIAN)
@@ -37,6 +52,9 @@
 #elif defined (eTeX)
 #define TEXMFPOOLNAME "etex.pool"
 #define TEXMFENGINENAME "etex"
+#elif defined (XeTeX)
+#define TEXMFPOOLNAME "xetex.pool"
+#define TEXMFENGINENAME "xetex"
 #elif defined (Omega)
 #define TEXMFPOOLNAME "omega.pool"
 #define TEXMFENGINENAME "omega"
@@ -153,7 +171,11 @@ extern void ipcpage P1H(int);
 
 /* Read a line of input as quickly as possible.  */
 #define	inputln(stream, flag) input_line (stream)
+#ifdef XeTeX
+extern boolean input_line P1H(UFILE *);
+#else
 extern boolean input_line P1H(FILE *);
+#endif
 
 /* This routine has to return four values.  */
 #define	dateandtime(i,j,k,l) get_date_and_time (&(i), &(j), &(k), &(l))
@@ -201,6 +223,9 @@ extern void setupboundvariable P3H(integer *, const_string, integer);
 #define wopenin(f)	open_input (&(f), DUMP_FORMAT, FOPEN_RBIN_MODE)
 #define wopenout	bopenout
 #define wclose		aclose
+#ifdef XeTeX
+#define uopenin(f,p,m,d) u_open_in(&(f), p, FOPEN_RBIN_MODE, m, d)
+#endif
 
 /* Used in tex.ch (section 1338) to get a core dump in debugging mode.  */
 #ifdef unix