diff options
Diffstat (limited to 'Build/source/libs/icu/icu-xetex/tools/genbrk')
4 files changed, 779 insertions, 0 deletions
diff --git a/Build/source/libs/icu/icu-xetex/tools/genbrk/Makefile.in b/Build/source/libs/icu/icu-xetex/tools/genbrk/Makefile.in new file mode 100644 index 00000000000..5884cef323d --- /dev/null +++ b/Build/source/libs/icu/icu-xetex/tools/genbrk/Makefile.in @@ -0,0 +1,96 @@ +## Makefile.in for ICU - tools/genbrk +## Copyright (c) 2002-2006 International Business Machines Corporation and +## others. All Rights Reserved. + +## Source directory information +srcdir = @srcdir@ +top_srcdir = @top_srcdir@ + +top_builddir = ../.. + +include $(top_builddir)/icudefs.mk + +## Build directory information +subdir = tools/genbrk + +TARGET_STUB_NAME = genbrk + +SECTION = 1 + +MAN_FILES = $(TARGET_STUB_NAME).$(SECTION) + + +## Extra files to remove for 'make clean' +CLEANFILES = *~ $(DEPS) $(MAN_FILES) + +## Target information +TARGET = $(BINDIR)/$(TARGET_STUB_NAME)$(EXEEXT) + +ifneq ($(top_builddir),$(top_srcdir)) +CPPFLAGS += -I$(top_builddir)/common +endif +CPPFLAGS += -I$(top_srcdir)/common -I$(srcdir)/../toolutil +LIBS = $(LIBICUTOOLUTIL) $(LIBICUI18N) $(LIBICUUC) $(DEFAULT_LIBS) $(LIB_M) + +OBJECTS = genbrk.o + +DEPS = $(OBJECTS:.o=.d) + +## List of phony targets +.PHONY : all all-local install install-local clean clean-local \ +distclean distclean-local dist dist-local check check-local install-man + +## Clear suffix list +.SUFFIXES : + +## List of standard targets +all: all-local +install: install-local +clean: clean-local +distclean : distclean-local +dist: dist-local +check: all check-local + +all-local: $(TARGET) $(MAN_FILES) + +install-local: all-local install-man + $(MKINSTALLDIRS) $(DESTDIR)$(bindir) + $(INSTALL) $(TARGET) $(DESTDIR)$(bindir) + +install-man: $(MAN_FILES) + $(MKINSTALLDIRS) $(DESTDIR)$(mandir)/man$(SECTION) + $(INSTALL_DATA) $? $(DESTDIR)$(mandir)/man$(SECTION) + +dist-local: + +clean-local: + test -z "$(CLEANFILES)" || $(RMV) $(CLEANFILES) + $(RMV) $(TARGET) $(OBJECTS) + +distclean-local: clean-local + $(RMV) Makefile + +check-local: all-local + +Makefile: $(srcdir)/Makefile.in $(top_builddir)/config.status + cd $(top_builddir) \ + && CONFIG_FILES=$(subdir)/$@ CONFIG_HEADERS= $(SHELL) ./config.status + +$(TARGET) : $(OBJECTS) + $(LINK.cc) $(OUTOPT)$@ $^ $(LIBS) + $(POST_BUILD_STEP) + + +%.$(SECTION): $(srcdir)/%.$(SECTION).in + cd $(top_builddir) \ + && CONFIG_FILES=$(subdir)/$@ CONFIG_HEADERS= $(SHELL) ./config.status + + +ifeq (,$(MAKECMDGOALS)) +-include $(DEPS) +else +ifneq ($(patsubst %clean,,$(MAKECMDGOALS)),) +-include $(DEPS) +endif +endif + diff --git a/Build/source/libs/icu/icu-xetex/tools/genbrk/genbrk.1.in b/Build/source/libs/icu/icu-xetex/tools/genbrk/genbrk.1.in new file mode 100644 index 00000000000..a6e214a9791 --- /dev/null +++ b/Build/source/libs/icu/icu-xetex/tools/genbrk/genbrk.1.in @@ -0,0 +1,112 @@ +.\" Hey, Emacs! This is -*-nroff-*- you know... +.\" +.\" genbrk.1: manual page for the genbrk utility +.\" +.\" Copyright (C) 2005-2006 International Business Machines Corporation and others +.\" +.TH GENBRK 1 "2 December 2005" "ICU MANPAGE" "ICU @VERSION@ Manual" +.SH NAME +.B genbrk +\- Compiles ICU break iteration rules source files into binary data files +.SH SYNOPSIS +.B genbrk +[ +.BR "\-h\fP, \fB\-?\fP, \fB\-\-help" +] +[ +.BR "\-V\fP, \fB\-\-version" +] +[ +.BR "\-c\fP, \fB\-\-copyright" +] +[ +.BR "\-v\fP, \fB\-\-verbose" +] +[ +.BI "\-d\fP, \fB\-\-destdir" " destination" +] +[ +.BI "\-i\fP, \fB\-\-icudatadir" " directory" +] +.BI "\-r\fP, \fB\-\-rules" " rule\-file" +.BI "\-o\fP, \fB\-\-out" " output\-file" +.SH DESCRIPTION +.B genbrk +reads the break (boundary) rule source code from +.I rule-file +and creates a break iteration data file. Normally this data file has the +.B .brk +extension. +.PP +The details of the rule syntax can be found in ICU's User Guide. +.SH OPTIONS +.TP +.BR "\-h\fP, \fB\-?\fP, \fB\-\-help" +Print help about usage and exit. +.TP +.BR "\-V\fP, \fB\-\-version" +Print the version of +.B genbrk +and exit. +.TP +.BR "\-c\fP, \fB\-\-copyright" +Embeds the standard ICU copyright into the +.IR output-file . +.TP +.BR "\-v\fP, \fB\-\-verbose" +Display extra informative messages during execution. +.TP +.BI "\-d\fP, \fB\-\-destdir" " destination" +Set the destination directory of the +.IR output-file +to +.IR destination . +.TP +.BI "\-i\fP, \fB\-\-icudatadir" " directory" +Look for any necessary ICU data files in +.IR directory . +For example, the file +.B pnames.icu +must be located when ICU's data is not built as a shared library. +The default ICU data directory is specified by the environment variable +.BR ICU_DATA . +Most configurations of ICU do not require this argument. +.TP +.BI "\-r\fP, \fB\-\-rules" " rule\-file" +The source file to read. +.TP +.BI "\-o\fP, \fB\-\-out" " output\-file" +The output data file to write. +.SH CAVEATS +When the +.IR rule-file +contains a byte order mark (BOM) at the beginning of the file, which is the Unicode character +.B U+FEFF, +then the +.IR rule-file +is interpreted as Unicode. Without the BOM, +the file is interpreted in the current operating system default codepage. +In order to eliminate any ambiguity of the encoding for how the +.IR rule-file +was written, it is recommended that you write this file in UTF-8 +with the BOM. +.SH ENVIRONMENT +.TP 10 +.B ICU_DATA +Specifies the directory containing ICU data. Defaults to +.BR @thepkgicudatadir@/@PACKAGE@/@VERSION@/ . +Some tools in ICU depend on the presence of the trailing slash. It is thus +important to make sure that it is present if +.B ICU_DATA +is set. +.SH AUTHORS +George Rhoten +.br +Andy Heninger +.SH VERSION +1.0 +.SH COPYRIGHT +Copyright (C) 2005 International Business Machines Corporation and others +.SH SEE ALSO +.BR http://www.icu-project.org/userguide/boundaryAnalysis.html + diff --git a/Build/source/libs/icu/icu-xetex/tools/genbrk/genbrk.cpp b/Build/source/libs/icu/icu-xetex/tools/genbrk/genbrk.cpp new file mode 100644 index 00000000000..ddbcfc2d575 --- /dev/null +++ b/Build/source/libs/icu/icu-xetex/tools/genbrk/genbrk.cpp @@ -0,0 +1,345 @@ +/* +********************************************************************** +* Copyright (C) 2002-2006, International Business Machines +* Corporation and others. All Rights Reserved. +********************************************************************** +* +* File genbrk.c +*/ + +//-------------------------------------------------------------------- +// +// Tool for generating RuleBasedBreakIterator data files (.brk files). +// .brk files contain the precompiled rules for standard types +// of iterators - word, line, sentence, etc. +// +// Usage: genbrk [options] -r rule-file.txt -o output-file.brk +// +// options: -v verbose +// -? or -h help +// +// The input rule file is a plain text file containing break rules +// in the input format accepted by RuleBasedBreakIterators. The +// file can be encoded as utf-8, or utf-16 (either endian), or +// in the default code page (platform dependent.). utf encoded +// files must include a BOM. +// +//-------------------------------------------------------------------- + +#include "unicode/utypes.h" +#include "unicode/ucnv.h" +#include "unicode/unistr.h" +#include "unicode/rbbi.h" +#include "unicode/uclean.h" +#include "unicode/udata.h" +#include "unicode/putil.h" + +#include "uoptions.h" +#include "unewdata.h" +#include "ucmndata.h" +#include "rbbidata.h" +#include "cmemory.h" + +#include <stdio.h> +#include <stdlib.h> +#include <string.h> + +U_NAMESPACE_USE + +static char *progName; +static UOption options[]={ + UOPTION_HELP_H, /* 0 */ + UOPTION_HELP_QUESTION_MARK, /* 1 */ + UOPTION_VERBOSE, /* 2 */ + { "rules", NULL, NULL, NULL, 'r', UOPT_REQUIRES_ARG, 0 }, /* 3 */ + { "out", NULL, NULL, NULL, 'o', UOPT_REQUIRES_ARG, 0 }, /* 4 */ + UOPTION_ICUDATADIR, /* 5 */ + UOPTION_DESTDIR, /* 6 */ + UOPTION_COPYRIGHT, /* 7 */ +}; + +void usageAndDie(int retCode) { + printf("Usage: %s [-v] [-options] -r rule-file -o output-file\n", progName); + printf("\tRead in break iteration rules text and write out the binary data\n" + "options:\n" + "\t-h or -? or --help this usage text\n" + "\t-V or --version show a version message\n" + "\t-c or --copyright include a copyright notice\n" + "\t-v or --verbose turn on verbose output\n" + "\t-i or --icudatadir directory for locating any needed intermediate data files,\n" + "\t followed by path, defaults to %s\n" + "\t-d or --destdir destination directory, followed by the path\n", + u_getDataDirectory()); + exit (retCode); +} + + +#if UCONFIG_NO_BREAK_ITERATION + +/* dummy UDataInfo cf. udata.h */ +static UDataInfo dummyDataInfo = { + sizeof(UDataInfo), + 0, + + U_IS_BIG_ENDIAN, + U_CHARSET_FAMILY, + U_SIZEOF_UCHAR, + 0, + + { 0, 0, 0, 0 }, /* dummy dataFormat */ + { 0, 0, 0, 0 }, /* dummy formatVersion */ + { 0, 0, 0, 0 } /* dummy dataVersion */ +}; + +#else + +// +// Set up the ICU data header, defined in ucmndata.h +// +DataHeader dh ={ + {sizeof(DataHeader), // Struct MappedData + 0xda, + 0x27}, + + { // struct UDataInfo + sizeof(UDataInfo), // size + 0, // reserved + U_IS_BIG_ENDIAN, + U_CHARSET_FAMILY, + U_SIZEOF_UCHAR, + 0, // reserved + + { 0x42, 0x72, 0x6b, 0x20 }, // dataFormat="Brk " + { 0xff, 0, 0, 0 }, // formatVersion. Filled in later with values + // from the RBBI rule builder. The values declared + // here should never appear in any real RBBI data. + { 4, 1, 0, 0 } // dataVersion (Unicode version) + }}; + +#endif + +//---------------------------------------------------------------------------- +// +// main for genbrk +// +//---------------------------------------------------------------------------- +int main(int argc, char **argv) { + UErrorCode status = U_ZERO_ERROR; + const char *ruleFileName; + const char *outFileName; + const char *outDir = NULL; + const char *copyright = NULL; + + // + // Pick up and check the command line arguments, + // using the standard ICU tool utils option handling. + // + U_MAIN_INIT_ARGS(argc, argv); + progName = argv[0]; + argc=u_parseArgs(argc, argv, sizeof(options)/sizeof(options[0]), options); + if(argc<0) { + // Unrecognized option + fprintf(stderr, "error in command line argument \"%s\"\n", argv[-argc]); + usageAndDie(U_ILLEGAL_ARGUMENT_ERROR); + } + + if(options[0].doesOccur || options[1].doesOccur) { + // -? or -h for help. + usageAndDie(0); + } + + if (!(options[3].doesOccur && options[4].doesOccur)) { + fprintf(stderr, "rule file and output file must both be specified.\n"); + usageAndDie(U_ILLEGAL_ARGUMENT_ERROR); + } + ruleFileName = options[3].value; + outFileName = options[4].value; + + if (options[5].doesOccur) { + u_setDataDirectory(options[5].value); + } + + /* Initialize ICU */ + u_init(&status); + if (U_FAILURE(status)) { + fprintf(stderr, "%s: can not initialize ICU. status = %s\n", + argv[0], u_errorName(status)); + exit(1); + } + status = U_ZERO_ERROR; + + /* Combine the directory with the file name */ + if(options[6].doesOccur) { + outDir = options[6].value; + } + if (options[7].doesOccur) { + copyright = U_COPYRIGHT_STRING; + } + +#if UCONFIG_NO_BREAK_ITERATION + + UNewDataMemory *pData; + char msg[1024]; + + /* write message with just the name */ + sprintf(msg, "genbrk writes dummy %s because of UCONFIG_NO_BREAK_ITERATION, see uconfig.h", outFileName); + fprintf(stderr, "%s\n", msg); + + /* write the dummy data file */ + pData = udata_create(outDir, NULL, outFileName, &dummyDataInfo, NULL, &status); + udata_writeBlock(pData, msg, strlen(msg)); + udata_finish(pData, &status); + return (int)status; + +#else + + // + // Read in the rule source file + // + long result; + long ruleFileSize; + FILE *file; + char *ruleBufferC; + + file = fopen(ruleFileName, "rb"); + if( file == 0 ) { + fprintf(stderr, "Could not open file \"%s\"\n", ruleFileName); + exit(-1); + } + fseek(file, 0, SEEK_END); + ruleFileSize = ftell(file); + fseek(file, 0, SEEK_SET); + ruleBufferC = new char[ruleFileSize+10]; + + result = (long)fread(ruleBufferC, 1, ruleFileSize, file); + if (result != ruleFileSize) { + fprintf(stderr, "Error reading file \"%s\"\n", ruleFileName); + exit (-1); + } + ruleBufferC[ruleFileSize]=0; + fclose(file); + + // + // Look for a Unicode Signature (BOM) on the rule file + // + int32_t signatureLength; + const char * ruleSourceC = ruleBufferC; + const char* encoding = ucnv_detectUnicodeSignature( + ruleSourceC, ruleFileSize, &signatureLength, &status); + if (U_FAILURE(status)) { + exit(status); + } + if(encoding!=NULL ){ + ruleSourceC += signatureLength; + ruleFileSize -= signatureLength; + } + + // + // Open a converter to take the rule file to UTF-16 + // + UConverter* conv; + conv = ucnv_open(encoding, &status); + if (U_FAILURE(status)) { + fprintf(stderr, "ucnv_open: ICU Error \"%s\"\n", u_errorName(status)); + exit(status); + } + + // + // Convert the rules to UChar. + // Preflight first to determine required buffer size. + // + uint32_t destCap = ucnv_toUChars(conv, + NULL, // dest, + 0, // destCapacity, + ruleSourceC, + ruleFileSize, + &status); + if (status != U_BUFFER_OVERFLOW_ERROR) { + fprintf(stderr, "ucnv_toUChars: ICU Error \"%s\"\n", u_errorName(status)); + exit(status); + }; + + status = U_ZERO_ERROR; + UChar *ruleSourceU = new UChar[destCap+1]; + ucnv_toUChars(conv, + ruleSourceU, // dest, + destCap+1, + ruleSourceC, + ruleFileSize, + &status); + if (U_FAILURE(status)) { + fprintf(stderr, "ucnv_toUChars: ICU Error \"%s\"\n", u_errorName(status)); + exit(status); + }; + ucnv_close(conv); + + + // + // Put the source rules into a UnicodeString + // + UnicodeString ruleSourceS(FALSE, ruleSourceU, destCap); + + // + // Create the break iterator from the rules + // This will compile the rules. + // + UParseError parseError; + parseError.line = 0; + parseError.offset = 0; + RuleBasedBreakIterator *bi = new RuleBasedBreakIterator(ruleSourceS, parseError, status); + if (U_FAILURE(status)) { + fprintf(stderr, "createRuleBasedBreakIterator: ICU Error \"%s\" at line %d, column %d\n", + u_errorName(status), (int)parseError.line, (int)parseError.offset); + exit(status); + }; + + + // + // Get the compiled rule data from the break iterator. + // + uint32_t outDataSize; + const uint8_t *outData; + outData = bi->getBinaryRules(outDataSize); + + // Copy the data format version numbers from the RBBI data header into the UDataMemory header. + uprv_memcpy(dh.info.formatVersion, ((RBBIDataHeader *)outData)->fFormatVersion, sizeof(dh.info.formatVersion)); + + // + // Create the output file + // + size_t bytesWritten; + UNewDataMemory *pData; + pData = udata_create(outDir, NULL, outFileName, &(dh.info), copyright, &status); + if(U_FAILURE(status)) { + fprintf(stderr, "genbrk: Could not open output file \"%s\", \"%s\"\n", + outFileName, u_errorName(status)); + exit(status); + } + + + // Write the data itself. + udata_writeBlock(pData, outData, outDataSize); + // finish up + bytesWritten = udata_finish(pData, &status); + if(U_FAILURE(status)) { + fprintf(stderr, "genbrk: error %d writing the output file\n", status); + exit(status); + } + + if (bytesWritten != outDataSize) { + fprintf(stderr, "Error writing to output file \"%s\"\n", outFileName); + exit(-1); + } + + delete bi; + delete[] ruleSourceU; + delete[] ruleBufferC; + u_cleanup(); + + + printf("genbrk: tool completed successfully.\n"); + return 0; + +#endif /* #if !UCONFIG_NO_BREAK_ITERATION */ +} + diff --git a/Build/source/libs/icu/icu-xetex/tools/genbrk/genbrk.vcproj b/Build/source/libs/icu/icu-xetex/tools/genbrk/genbrk.vcproj new file mode 100644 index 00000000000..2289a3c54c5 --- /dev/null +++ b/Build/source/libs/icu/icu-xetex/tools/genbrk/genbrk.vcproj @@ -0,0 +1,226 @@ +<?xml version="1.0" encoding="Windows-1252"?> +<VisualStudioProject + ProjectType="Visual C++" + Version="8.00" + Name="genbrk" + ProjectGUID="{C2BE5000-7501-4E87-9724-B8D82494FAE6}" + > + <Platforms> + <Platform + Name="Win32" + /> + </Platforms> + <ToolFiles> + </ToolFiles> + <Configurations> + <Configuration + Name="Release|Win32" + OutputDirectory=".\Release" + IntermediateDirectory=".\Release" + ConfigurationType="1" + InheritedPropertySheets="$(VCInstallDir)VCProjectDefaults\UpgradeFromVC71.vsprops" + UseOfMFC="0" + ATLMinimizesCRunTimeLibraryUsage="false" + CharacterSet="2" + > + <Tool + Name="VCPreBuildEventTool" + /> + <Tool + Name="VCCustomBuildTool" + CommandLine="copy "$(TargetPath)" ..\..\..\bin
" + Outputs="..\..\..\bin\$(TargetFileName)" + /> + <Tool + Name="VCXMLDataGeneratorTool" + /> + <Tool + Name="VCWebServiceProxyGeneratorTool" + /> + <Tool + Name="VCMIDLTool" + TypeLibraryName=".\Release/genbrk.tlb" + /> + <Tool + Name="VCCLCompilerTool" + AdditionalIncludeDirectories="..\..\common;..\toolutil" + PreprocessorDefinitions="WIN32;NDEBUG;_CRT_SECURE_NO_DEPRECATE" + StringPooling="true" + RuntimeLibrary="2" + EnableFunctionLevelLinking="true" + DisableLanguageExtensions="true" + TreatWChar_tAsBuiltInType="true" + PrecompiledHeaderFile=".\Release/genbrk.pch" + AssemblerListingLocation=".\Release/" + ObjectFile=".\Release/" + ProgramDataBaseFileName=".\Release/" + WarningLevel="3" + SuppressStartupBanner="true" + CompileAs="0" + /> + <Tool + Name="VCManagedResourceCompilerTool" + /> + <Tool + Name="VCResourceCompilerTool" + PreprocessorDefinitions="NDEBUG" + Culture="1033" + /> + <Tool + Name="VCPreLinkEventTool" + /> + <Tool + Name="VCLinkerTool" + OutputFile=".\Release/genbrk.exe" + LinkIncremental="1" + SuppressStartupBanner="true" + ProgramDatabaseFile=".\Release/genbrk.pdb" + SubSystem="1" + OptimizeForWindows98="1" + /> + <Tool + Name="VCALinkTool" + /> + <Tool + Name="VCManifestTool" + /> + <Tool + Name="VCXDCMakeTool" + /> + <Tool + Name="VCBscMakeTool" + /> + <Tool + Name="VCFxCopTool" + /> + <Tool + Name="VCAppVerifierTool" + /> + <Tool + Name="VCWebDeploymentTool" + /> + <Tool + Name="VCPostBuildEventTool" + /> + </Configuration> + <Configuration + Name="Debug|Win32" + OutputDirectory=".\Debug" + IntermediateDirectory=".\Debug" + ConfigurationType="1" + InheritedPropertySheets="$(VCInstallDir)VCProjectDefaults\UpgradeFromVC71.vsprops" + UseOfMFC="0" + ATLMinimizesCRunTimeLibraryUsage="false" + CharacterSet="2" + > + <Tool + Name="VCPreBuildEventTool" + /> + <Tool + Name="VCCustomBuildTool" + CommandLine="copy "$(TargetPath)" ..\..\..\bin
" + Outputs="..\..\..\bin\$(TargetFileName)" + /> + <Tool + Name="VCXMLDataGeneratorTool" + /> + <Tool + Name="VCWebServiceProxyGeneratorTool" + /> + <Tool + Name="VCMIDLTool" + TypeLibraryName=".\Debug/genbrk.tlb" + /> + <Tool + Name="VCCLCompilerTool" + Optimization="0" + AdditionalIncludeDirectories="..\..\common;..\toolutil" + PreprocessorDefinitions="WIN32;_DEBUG;_CRT_SECURE_NO_DEPRECATE" + BasicRuntimeChecks="3" + RuntimeLibrary="3" + BufferSecurityCheck="true" + DisableLanguageExtensions="true" + TreatWChar_tAsBuiltInType="true" + PrecompiledHeaderFile=".\Debug/genbrk.pch" + AssemblerListingLocation=".\Debug/" + ObjectFile=".\Debug/" + ProgramDataBaseFileName=".\Debug/" + BrowseInformation="1" + WarningLevel="3" + SuppressStartupBanner="true" + DebugInformationFormat="4" + CompileAs="0" + /> + <Tool + Name="VCManagedResourceCompilerTool" + /> + <Tool + Name="VCResourceCompilerTool" + PreprocessorDefinitions="_DEBUG" + Culture="1033" + /> + <Tool + Name="VCPreLinkEventTool" + /> + <Tool + Name="VCLinkerTool" + OutputFile=".\Debug/genbrk.exe" + LinkIncremental="2" + SuppressStartupBanner="true" + GenerateDebugInformation="true" + ProgramDatabaseFile=".\Debug/genbrk.pdb" + SubSystem="1" + /> + <Tool + Name="VCALinkTool" + /> + <Tool + Name="VCManifestTool" + UseFAT32Workaround="true" + /> + <Tool + Name="VCXDCMakeTool" + /> + <Tool + Name="VCBscMakeTool" + /> + <Tool + Name="VCFxCopTool" + /> + <Tool + Name="VCAppVerifierTool" + /> + <Tool + Name="VCWebDeploymentTool" + /> + <Tool + Name="VCPostBuildEventTool" + /> + </Configuration> + </Configurations> + <References> + </References> + <Files> + <Filter + Name="Source Files" + Filter="cpp;c;cxx;rc;def;r;odl;idl;hpj;bat" + > + <File + RelativePath=".\genbrk.cpp" + > + </File> + </Filter> + <Filter + Name="Header Files" + Filter="h;hpp;hxx;hm;inl" + > + </Filter> + <Filter + Name="Resource Files" + Filter="ico;cur;bmp;dlg;rc2;rct;bin;rgs;gif;jpg;jpeg;jpe" + > + </Filter> + </Files> + <Globals> + </Globals> +</VisualStudioProject> |