ICU-8972 replace gennorm with code in preparseucd.py

X-SVN-Rev: 31174
This commit is contained in:
Markus Scherer 2011-12-28 19:23:13 +00:00
parent 66ead436a8
commit 72559ba1cd
9 changed files with 109 additions and 1269 deletions

14
.gitignore vendored
View File

@ -996,20 +996,6 @@ tools/unicode/c/gennames/gennames.vcproj.*.*.user
tools/unicode/c/gennames/release
tools/unicode/c/gennames/x64
tools/unicode/c/gennames/x86
tools/unicode/c/gennorm/*.d
tools/unicode/c/gennorm/*.o
tools/unicode/c/gennorm/*.pdb
tools/unicode/c/gennorm/*.plg
tools/unicode/c/gennorm/Debug
tools/unicode/c/gennorm/Makefile
tools/unicode/c/gennorm/Release
tools/unicode/c/gennorm/debug
tools/unicode/c/gennorm/gennorm
tools/unicode/c/gennorm/gennorm.[0-9]
tools/unicode/c/gennorm/gennorm.vcproj.*.*.user
tools/unicode/c/gennorm/release
tools/unicode/c/gennorm/x64
tools/unicode/c/gennorm/x86
tools/unicode/c/genprops/*.d
tools/unicode/c/genprops/*.ncb
tools/unicode/c/genprops/*.o

View File

@ -19,7 +19,6 @@ include_directories(
link_directories(${ICU_INST_DIR}/lib)
add_subdirectory(gencase)
add_subdirectory(gennames)
add_subdirectory(gennorm)
add_subdirectory(genprops)
add_subdirectory(genuca)
add_subdirectory(genuts46)

View File

@ -1,9 +0,0 @@
# Copyright (C) 2010, International Business Machines
# Corporation and others. All Rights Reserved.
#
# created on: 2010jun03
# created by: Markus W. Scherer
# edited on: 2010jul20
# edited by: Stuart G. Gill
add_executable(gennorm gennorm.c store.c)
target_link_libraries(gennorm icuuc icutu)

View File

@ -1,97 +0,0 @@
## Makefile.in for ICU - tools/gennorm
## Copyright (c) 2001-2005, International Business Machines Corporation and
## others. All Rights Reserved.
## Steven R. Loomis/Markus W. Scherer
## Source directory information
srcdir = @srcdir@
top_srcdir = @top_srcdir@
top_builddir = ../..
include $(top_builddir)/icudefs.mk
## Build directory information
subdir = tools/gennorm
TARGET_STUB_NAME = gennorm
SECTION = 8
#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 = gennorm.o store.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) $(DESTDIR)$(sbindir)
# $(INSTALL) $(TARGET) $(DESTDIR)$(sbindir)
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

View File

@ -1,399 +0,0 @@
/*
*******************************************************************************
*
* Copyright (C) 2001-2010, International Business Machines
* Corporation and others. All Rights Reserved.
*
*******************************************************************************
* file name: gennorm.c
* encoding: US-ASCII
* tab size: 8 (not used)
* indentation:4
*
* created on: 2001may25
* created by: Markus W. Scherer
*
* This program reads the Unicode character database text file,
* parses it, and extracts the data for normalization.
* It then preprocesses it and writes a binary file for efficient use
* in various Unicode text normalization processes.
*/
#include <stdio.h>
#include <stdlib.h>
#include "unicode/utypes.h"
#include "unicode/uchar.h"
#include "unicode/ustring.h"
#include "unicode/putil.h"
#include "unicode/uclean.h"
#include "unicode/udata.h"
#include "unicode/uset.h"
#include "cmemory.h"
#include "cstring.h"
#include "unewdata.h"
#include "uoptions.h"
#include "uparse.h"
#include "unormimp.h"
U_CDECL_BEGIN
#include "gennorm.h"
U_CDECL_END
UBool beVerbose=FALSE;
/* prototypes --------------------------------------------------------------- */
static void
parseDerivedNormalizationProperties(const char *filename, UErrorCode *pErrorCode, UBool reportError);
static void
parseDB(const char *filename, UErrorCode *pErrorCode);
/* -------------------------------------------------------------------------- */
enum {
HELP_H,
HELP_QUESTION_MARK,
VERBOSE,
DESTDIR,
SOURCEDIR,
ICUDATADIR
};
static UOption options[]={
UOPTION_HELP_H,
UOPTION_HELP_QUESTION_MARK,
UOPTION_VERBOSE,
UOPTION_DESTDIR,
UOPTION_SOURCEDIR,
UOPTION_ICUDATADIR
};
extern int
main(int argc, char* argv[]) {
#if !UCONFIG_NO_NORMALIZATION
char filename[300];
#endif
const char *srcDir=NULL, *destDir=NULL, *suffix=NULL;
char *basename=NULL;
UErrorCode errorCode=U_ZERO_ERROR;
U_MAIN_INIT_ARGS(argc, argv);
/* preset then read command line options */
options[DESTDIR].value=u_getDataDirectory();
options[SOURCEDIR].value="";
options[ICUDATADIR].value=u_getDataDirectory();
argc=u_parseArgs(argc, argv, sizeof(options)/sizeof(options[0]), options);
/* error handling, printing usage message */
if(argc<0) {
fprintf(stderr,
"error in command line argument \"%s\"\n",
argv[-argc]);
}
if(argc<0 || options[HELP_H].doesOccur || options[HELP_QUESTION_MARK].doesOccur) {
/*
* Broken into chucks because the C89 standard says the minimum
* required supported string length is 509 bytes.
*/
fprintf(stderr,
"Usage: %s [-options] [suffix]\n"
"\n"
"Read the UnicodeData.txt file and other Unicode properties files and\n"
"write nfc.txt and nfkc.txt files for gennorm2\n"
"\n",
argv[0]);
fprintf(stderr,
"Options:\n"
"\t-h or -? or --help this usage text\n"
"\t-v or --verbose verbose output\n");
fprintf(stderr,
"\t-d or --destdir destination directory, followed by the path\n"
"\t-s or --sourcedir source directory, followed by the path\n"
"\t-i or --icudatadir directory for locating any needed intermediate data files,\n"
"\t followed by path, defaults to <%s>\n"
"\tsuffix suffix that is to be appended with a '-'\n"
"\t to the source file basenames before opening;\n"
"\t 'gennorm new' will read UnicodeData-new.txt etc.\n",
u_getDataDirectory());
return argc<0 ? U_ILLEGAL_ARGUMENT_ERROR : U_ZERO_ERROR;
}
/* get the options values */
beVerbose=options[VERBOSE].doesOccur;
srcDir=options[SOURCEDIR].value;
destDir=options[DESTDIR].value;
if(argc>=2) {
suffix=argv[1];
} else {
suffix=NULL;
}
#if !UCONFIG_NO_NORMALIZATION
if (options[ICUDATADIR].doesOccur) {
u_setDataDirectory(options[ICUDATADIR].value);
}
/*
* Verify that we can work with properties
* but don't call u_init() because that needs unorm.icu which we are just
* going to build here.
*/
{
U_STRING_DECL(ideo, "[:Ideographic:]", 15);
USet *set;
U_STRING_INIT(ideo, "[:Ideographic:]", 15);
set=uset_openPattern(ideo, -1, &errorCode);
if(U_FAILURE(errorCode) || !uset_contains(set, 0xf900)) {
fprintf(stderr, "gennorm is unable to work with properties (uprops.icu): %s\n", u_errorName(errorCode));
exit(errorCode);
}
uset_close(set);
}
/* prepare the filename beginning with the source dir */
uprv_strcpy(filename, srcDir);
basename=filename+uprv_strlen(filename);
if(basename>filename && *(basename-1)!=U_FILE_SEP_CHAR && *(basename-1)!=U_FILE_ALT_SEP_CHAR) {
*basename++=U_FILE_SEP_CHAR;
}
/* initialize */
init();
/* process DerivedNormalizationProps.txt (name changed for Unicode 3.2, to <=31 characters) */
if(suffix==NULL) {
uprv_strcpy(basename, "DerivedNormalizationProps.txt");
} else {
uprv_strcpy(basename, "DerivedNormalizationProps");
basename[30]='-';
uprv_strcpy(basename+31, suffix);
uprv_strcat(basename+31, ".txt");
}
parseDerivedNormalizationProperties(filename, &errorCode, FALSE);
if(U_FAILURE(errorCode)) {
/* can be only U_FILE_ACCESS_ERROR - try filename from before Unicode 3.2 */
if(suffix==NULL) {
uprv_strcpy(basename, "DerivedNormalizationProperties.txt");
} else {
uprv_strcpy(basename, "DerivedNormalizationProperties");
basename[30]='-';
uprv_strcpy(basename+31, suffix);
uprv_strcat(basename+31, ".txt");
}
parseDerivedNormalizationProperties(filename, &errorCode, TRUE);
}
/* process UnicodeData.txt */
if(suffix==NULL) {
uprv_strcpy(basename, "UnicodeData.txt");
} else {
uprv_strcpy(basename, "UnicodeData");
basename[11]='-';
uprv_strcpy(basename+12, suffix);
uprv_strcat(basename+12, ".txt");
}
parseDB(filename, &errorCode);
/* process parsed data */
if(U_SUCCESS(errorCode)) {
writeNorm2(destDir);
cleanUpData();
}
#endif
return errorCode;
}
#if !UCONFIG_NO_NORMALIZATION
/* parser for DerivedNormalizationProperties.txt ---------------------------- */
static void U_CALLCONV
derivedNormalizationPropertiesLineFn(void *context,
char *fields[][2], int32_t fieldCount,
UErrorCode *pErrorCode) {
UChar string[32];
char *s;
uint32_t start, end;
int32_t count;
/* get code point range */
count=u_parseCodePointRange(fields[0][0], &start, &end, pErrorCode);
if(U_FAILURE(*pErrorCode)) {
fprintf(stderr, "gennorm: error parsing DerivedNormalizationProperties.txt mapping at %s\n", fields[0][0]);
exit(*pErrorCode);
}
/* ignore hangul - handle explicitly */
if(start==0xac00) {
return;
}
/* get property - ignore unrecognized ones */
s=(char *)u_skipWhitespace(fields[1][0]);
if(0==uprv_memcmp(s, "Comp_Ex", 7) || 0==uprv_memcmp(s, "Full_Composition_Exclusion", 26)) {
/* full composition exclusion */
while(start<=end) {
setCompositionExclusion(start++);
}
}
}
static void
parseDerivedNormalizationProperties(const char *filename, UErrorCode *pErrorCode, UBool reportError) {
char *fields[2][2];
if(pErrorCode==NULL || U_FAILURE(*pErrorCode)) {
return;
}
u_parseDelimitedFile(filename, ';', fields, 2, derivedNormalizationPropertiesLineFn, NULL, pErrorCode);
if(U_FAILURE(*pErrorCode) && (reportError || *pErrorCode!=U_FILE_ACCESS_ERROR)) {
fprintf(stderr, "gennorm error: u_parseDelimitedFile(\"%s\") failed - %s\n", filename, u_errorName(*pErrorCode));
exit(*pErrorCode);
}
}
/* parser for UnicodeData.txt ----------------------------------------------- */
static void U_CALLCONV
unicodeDataLineFn(void *context,
char *fields[][2], int32_t fieldCount,
UErrorCode *pErrorCode) {
uint32_t decomp[40];
Norm norm;
const char *s;
char *end;
uint32_t code, value;
int32_t length;
UBool isCompat, something=FALSE;
/* ignore First and Last entries for ranges */
if( *fields[1][0]=='<' &&
(length=(int32_t)(fields[1][1]-fields[1][0]))>=9 &&
(0==uprv_memcmp(", First>", fields[1][1]-8, 8) || 0==uprv_memcmp(", Last>", fields[1][1]-7, 7))
) {
return;
}
/* reset the properties */
uprv_memset(&norm, 0, sizeof(Norm));
/* get the character code, field 0 */
code=(uint32_t)uprv_strtoul(fields[0][0], &end, 16);
if(end<=fields[0][0] || end!=fields[0][1]) {
fprintf(stderr, "gennorm: syntax error in field 0 at %s\n", fields[0][0]);
*pErrorCode=U_PARSE_ERROR;
exit(U_PARSE_ERROR);
}
/* get canonical combining class, field 3 */
value=(uint32_t)uprv_strtoul(fields[3][0], &end, 10);
if(end<=fields[3][0] || end!=fields[3][1] || value>0xff) {
fprintf(stderr, "gennorm: syntax error in field 3 at %s\n", fields[0][0]);
*pErrorCode=U_PARSE_ERROR;
exit(U_PARSE_ERROR);
}
if(value>0) {
norm.udataCC=(uint8_t)value;
something=TRUE;
}
/* get the decomposition, field 5 */
if(fields[5][0]<fields[5][1]) {
if(*(s=fields[5][0])=='<') {
++s;
isCompat=TRUE;
/* skip and ignore the compatibility type name */
do {
if(s==fields[5][1]) {
/* missing '>' */
fprintf(stderr, "gennorm: syntax error in field 5 at %s\n", fields[0][0]);
*pErrorCode=U_PARSE_ERROR;
exit(U_PARSE_ERROR);
}
} while(*s++!='>');
} else {
isCompat=FALSE;
}
/* parse the decomposition string */
length=u_parseCodePoints(s, decomp, sizeof(decomp)/4, pErrorCode);
if(U_FAILURE(*pErrorCode)) {
fprintf(stderr, "gennorm error parsing UnicodeData.txt decomposition of U+%04lx - %s\n",
(long)code, u_errorName(*pErrorCode));
exit(*pErrorCode);
}
/* store the string */
if(length>0) {
something=TRUE;
if(isCompat) {
norm.lenNFKD=(uint8_t)length;
norm.nfkd=decomp;
} else {
if(length>2) {
fprintf(stderr, "gennorm: error - length of NFD(U+%04lx) = %ld >2 in UnicodeData - illegal\n",
(long)code, (long)length);
*pErrorCode=U_PARSE_ERROR;
exit(U_PARSE_ERROR);
}
norm.lenNFD=(uint8_t)length;
norm.nfd=decomp;
}
}
}
/* check for non-character code points */
if((code&0xfffe)==0xfffe || (uint32_t)(code-0xfdd0)<0x20 || code>0x10ffff) {
fprintf(stderr, "gennorm: error - properties for non-character code point U+%04lx\n",
(long)code);
*pErrorCode=U_PARSE_ERROR;
exit(U_PARSE_ERROR);
}
if(something) {
/* there are normalization values, so store them */
#if 0
if(beVerbose) {
printf("store values for U+%04lx: cc=%d, lenNFD=%ld, lenNFKD=%ld\n",
(long)code, norm.udataCC, (long)norm.lenNFD, (long)norm.lenNFKD);
}
#endif
storeNorm(code, &norm);
}
}
static void
parseDB(const char *filename, UErrorCode *pErrorCode) {
char *fields[15][2];
if(pErrorCode==NULL || U_FAILURE(*pErrorCode)) {
return;
}
u_parseDelimitedFile(filename, ';', fields, 15, unicodeDataLineFn, NULL, pErrorCode);
if(U_FAILURE(*pErrorCode)) {
fprintf(stderr, "gennorm error: u_parseDelimitedFile(\"%s\") failed - %s\n", filename, u_errorName(*pErrorCode));
exit(*pErrorCode);
}
}
#endif /* #if !UCONFIG_NO_NORMALIZATION */
/*
* Hey, Emacs, please set the following:
*
* Local Variables:
* indent-tabs-mode: nil
* End:
*
*/

View File

@ -1,50 +0,0 @@
/*
*******************************************************************************
*
* Copyright (C) 1999-2010, International Business Machines
* Corporation and others. All Rights Reserved.
*
*******************************************************************************
* file name: gennorm.h
* encoding: US-ASCII
* tab size: 8 (not used)
* indentation:4
*
* created on: 2001may25
* created by: Markus W. Scherer
*/
#ifndef __GENPROPS_H__
#define __GENPROPS_H__
#include "unicode/utypes.h"
/*
* data structure that holds the normalization properties for one or more
* code point(s) at build time
*/
typedef struct Norm {
uint8_t udataCC, lenNFD, lenNFKD;
uint32_t *nfd, *nfkd;
} Norm;
/* global flags */
extern UBool beVerbose;
/* prototypes */
extern void
init(void);
extern void
storeNorm(uint32_t code, Norm *norm);
extern void
setCompositionExclusion(uint32_t code);
extern void
writeNorm2(const char *dataDir);
extern void
cleanUpData(void);
#endif

View File

@ -1,407 +0,0 @@
<?xml version="1.0" encoding="Windows-1252"?>
<VisualStudioProject
ProjectType="Visual C++"
Version="9.00"
Name="gennorm"
ProjectGUID="{F5213103-6CBE-46E6-B4CC-2570B6837D86}"
TargetFrameworkVersion="131072"
>
<Platforms>
<Platform
Name="Win32"
/>
<Platform
Name="x64"
/>
</Platforms>
<ToolFiles>
</ToolFiles>
<Configurations>
<Configuration
Name="Release|Win32"
OutputDirectory=".\x86\Release"
IntermediateDirectory=".\x86\Release"
ConfigurationType="1"
InheritedPropertySheets="$(VCInstallDir)VCProjectDefaults\UpgradeFromVC71.vsprops"
UseOfMFC="0"
ATLMinimizesCRunTimeLibraryUsage="false"
CharacterSet="2"
>
<Tool
Name="VCPreBuildEventTool"
/>
<Tool
Name="VCCustomBuildTool"
CommandLine="copy &quot;$(TargetPath)&quot; ..\..\..\bin&#x0D;&#x0A;"
Outputs="..\..\..\bin\$(TargetFileName)"
/>
<Tool
Name="VCXMLDataGeneratorTool"
/>
<Tool
Name="VCWebServiceProxyGeneratorTool"
/>
<Tool
Name="VCMIDLTool"
TypeLibraryName=".\x86\Release/gennorm.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=".\x86\Release/gennorm.pch"
AssemblerListingLocation=".\x86\Release/"
ObjectFile=".\x86\Release/"
ProgramDataBaseFileName=".\x86\Release/"
WarningLevel="3"
SuppressStartupBanner="true"
CompileAs="0"
/>
<Tool
Name="VCManagedResourceCompilerTool"
/>
<Tool
Name="VCResourceCompilerTool"
PreprocessorDefinitions="NDEBUG"
Culture="1033"
/>
<Tool
Name="VCPreLinkEventTool"
/>
<Tool
Name="VCLinkerTool"
OutputFile=".\x86\Release/gennorm.exe"
LinkIncremental="1"
SuppressStartupBanner="true"
ProgramDatabaseFile=".\x86\Release/gennorm.pdb"
SubSystem="1"
RandomizedBaseAddress="1"
DataExecutionPrevention="0"
/>
<Tool
Name="VCALinkTool"
/>
<Tool
Name="VCManifestTool"
/>
<Tool
Name="VCXDCMakeTool"
/>
<Tool
Name="VCBscMakeTool"
/>
<Tool
Name="VCFxCopTool"
/>
<Tool
Name="VCAppVerifierTool"
/>
<Tool
Name="VCPostBuildEventTool"
/>
</Configuration>
<Configuration
Name="Debug|Win32"
OutputDirectory=".\x86\Debug"
IntermediateDirectory=".\x86\Debug"
ConfigurationType="1"
InheritedPropertySheets="$(VCInstallDir)VCProjectDefaults\UpgradeFromVC71.vsprops"
UseOfMFC="0"
ATLMinimizesCRunTimeLibraryUsage="false"
CharacterSet="2"
>
<Tool
Name="VCPreBuildEventTool"
/>
<Tool
Name="VCCustomBuildTool"
CommandLine="copy &quot;$(TargetPath)&quot; ..\..\..\bin&#x0D;&#x0A;"
Outputs="..\..\..\bin\$(TargetFileName)"
/>
<Tool
Name="VCXMLDataGeneratorTool"
/>
<Tool
Name="VCWebServiceProxyGeneratorTool"
/>
<Tool
Name="VCMIDLTool"
TypeLibraryName=".\x86\Debug/gennorm.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=".\x86\Debug/gennorm.pch"
AssemblerListingLocation=".\x86\Debug/"
ObjectFile=".\x86\Debug/"
ProgramDataBaseFileName=".\x86\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=".\x86\Debug/gennorm.exe"
LinkIncremental="2"
SuppressStartupBanner="true"
GenerateDebugInformation="true"
ProgramDatabaseFile=".\x86\Debug/gennorm.pdb"
SubSystem="1"
RandomizedBaseAddress="1"
DataExecutionPrevention="0"
/>
<Tool
Name="VCALinkTool"
/>
<Tool
Name="VCManifestTool"
UseFAT32Workaround="true"
/>
<Tool
Name="VCXDCMakeTool"
/>
<Tool
Name="VCBscMakeTool"
/>
<Tool
Name="VCFxCopTool"
/>
<Tool
Name="VCAppVerifierTool"
/>
<Tool
Name="VCPostBuildEventTool"
/>
</Configuration>
<Configuration
Name="Release|x64"
OutputDirectory=".\x64\Release"
IntermediateDirectory=".\x64\Release"
ConfigurationType="1"
InheritedPropertySheets="$(VCInstallDir)VCProjectDefaults\UpgradeFromVC71.vsprops"
UseOfMFC="0"
ATLMinimizesCRunTimeLibraryUsage="false"
CharacterSet="2"
>
<Tool
Name="VCPreBuildEventTool"
/>
<Tool
Name="VCCustomBuildTool"
CommandLine="copy &quot;$(TargetPath)&quot; ..\..\..\bin64&#x0D;&#x0A;"
Outputs="..\..\..\bin64\$(TargetFileName)"
/>
<Tool
Name="VCXMLDataGeneratorTool"
/>
<Tool
Name="VCWebServiceProxyGeneratorTool"
/>
<Tool
Name="VCMIDLTool"
TargetEnvironment="3"
TypeLibraryName=".\x64\Release/gennorm.tlb"
/>
<Tool
Name="VCCLCompilerTool"
AdditionalIncludeDirectories="..\..\common;..\toolutil"
PreprocessorDefinitions="WIN64;WIN32;NDEBUG;_CRT_SECURE_NO_DEPRECATE"
StringPooling="true"
RuntimeLibrary="2"
EnableFunctionLevelLinking="true"
DisableLanguageExtensions="true"
TreatWChar_tAsBuiltInType="true"
PrecompiledHeaderFile=".\x64\Release/gennorm.pch"
AssemblerListingLocation=".\x64\Release/"
ObjectFile=".\x64\Release/"
ProgramDataBaseFileName=".\x64\Release/"
WarningLevel="3"
SuppressStartupBanner="true"
CompileAs="0"
/>
<Tool
Name="VCManagedResourceCompilerTool"
/>
<Tool
Name="VCResourceCompilerTool"
PreprocessorDefinitions="NDEBUG"
Culture="1033"
/>
<Tool
Name="VCPreLinkEventTool"
/>
<Tool
Name="VCLinkerTool"
OutputFile=".\x64\Release/gennorm.exe"
LinkIncremental="1"
SuppressStartupBanner="true"
ProgramDatabaseFile=".\x64\Release/gennorm.pdb"
SubSystem="1"
TargetMachine="17"
/>
<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|x64"
OutputDirectory=".\x64\Debug"
IntermediateDirectory=".\x64\Debug"
ConfigurationType="1"
InheritedPropertySheets="$(VCInstallDir)VCProjectDefaults\UpgradeFromVC71.vsprops"
UseOfMFC="0"
ATLMinimizesCRunTimeLibraryUsage="false"
CharacterSet="2"
>
<Tool
Name="VCPreBuildEventTool"
/>
<Tool
Name="VCCustomBuildTool"
CommandLine="copy &quot;$(TargetPath)&quot; ..\..\..\bin64&#x0D;&#x0A;"
Outputs="..\..\..\bin64\$(TargetFileName)"
/>
<Tool
Name="VCXMLDataGeneratorTool"
/>
<Tool
Name="VCWebServiceProxyGeneratorTool"
/>
<Tool
Name="VCMIDLTool"
TargetEnvironment="3"
TypeLibraryName=".\x64\Debug/gennorm.tlb"
/>
<Tool
Name="VCCLCompilerTool"
Optimization="0"
AdditionalIncludeDirectories="..\..\common;..\toolutil"
PreprocessorDefinitions="WIN64;WIN32;_DEBUG;_CRT_SECURE_NO_DEPRECATE"
BasicRuntimeChecks="3"
RuntimeLibrary="3"
BufferSecurityCheck="true"
DisableLanguageExtensions="true"
TreatWChar_tAsBuiltInType="true"
PrecompiledHeaderFile=".\x64\Debug/gennorm.pch"
AssemblerListingLocation=".\x64\Debug/"
ObjectFile=".\x64\Debug/"
ProgramDataBaseFileName=".\x64\Debug/"
BrowseInformation="1"
WarningLevel="3"
SuppressStartupBanner="true"
DebugInformationFormat="3"
CompileAs="0"
/>
<Tool
Name="VCManagedResourceCompilerTool"
/>
<Tool
Name="VCResourceCompilerTool"
PreprocessorDefinitions="_DEBUG"
Culture="1033"
/>
<Tool
Name="VCPreLinkEventTool"
/>
<Tool
Name="VCLinkerTool"
OutputFile=".\x64\Debug/gennorm.exe"
LinkIncremental="2"
SuppressStartupBanner="true"
GenerateDebugInformation="true"
ProgramDatabaseFile=".\x64\Debug/gennorm.pdb"
SubSystem="1"
TargetMachine="17"
/>
<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>
<File
RelativePath=".\gennorm.c"
>
</File>
<File
RelativePath=".\gennorm.h"
>
</File>
<File
RelativePath=".\store.c"
>
</File>
</Files>
<Globals>
</Globals>
</VisualStudioProject>

View File

@ -1,289 +0,0 @@
/*
*******************************************************************************
*
* Copyright (C) 1999-2010, International Business Machines
* Corporation and others. All Rights Reserved.
*
*******************************************************************************
* file name: store.c
* encoding: US-ASCII
* tab size: 8 (not used)
* indentation:4
*
* created on: 2001may25
* created by: Markus W. Scherer
*
* Store Unicode normalization data in a memory-mappable file.
*/
#include <stdio.h>
#include <stdlib.h>
#include "unicode/utypes.h"
#include "unicode/udata.h"
#include "unicode/uset.h"
#include "cmemory.h"
#include "filestrm.h"
#include "utrie.h"
#include "toolutil.h"
#include "writesrc.h"
#include "unormimp.h"
#include "gennorm.h"
#define DO_DEBUG_OUT 0
#define LENGTHOF(array) (int32_t)(sizeof(array)/sizeof((array)[0]))
#if !UCONFIG_NO_NORMALIZATION
/* builder data ------------------------------------------------------------- */
static UNewTrie *normTrie;
static UToolMemory *normMem, *utf32Mem;
static Norm *norms;
static USet *compositionExclusions;
/* allocate and initialize a Norm unit */
static Norm *
allocNorm() {
/* allocate Norm */
Norm *p=(Norm *)utm_alloc(normMem);
return p;
}
extern void
init() {
uint16_t *p16;
normTrie = (UNewTrie *)uprv_malloc(sizeof(UNewTrie));
uprv_memset(normTrie, 0, sizeof(UNewTrie));
/* initialize the two tries */
if(NULL==utrie_open(normTrie, NULL, 30000, 0, 0, FALSE)) {
fprintf(stderr, "error: failed to initialize tries\n");
exit(U_MEMORY_ALLOCATION_ERROR);
}
/* allocate Norm structures and reset the first one */
normMem=utm_open("gennorm normalization structs", 20000, 20000, sizeof(Norm));
norms=allocNorm();
/* allocate UTF-32 string memory */
utf32Mem=utm_open("gennorm UTF-32 strings", 30000, 30000, 4);
compositionExclusions=uset_openEmpty();
}
/*
* get or create a Norm unit;
* get or create the intermediate trie entries for it as well
*/
static Norm *
createNorm(uint32_t code) {
Norm *p;
uint32_t i;
i=utrie_get32(normTrie, (UChar32)code, NULL);
if(i!=0) {
p=norms+i;
} else {
/* allocate Norm */
p=allocNorm();
if(!utrie_set32(normTrie, (UChar32)code, (uint32_t)(p-norms))) {
fprintf(stderr, "error: too many normalization entries\n");
exit(U_BUFFER_OVERFLOW_ERROR);
}
}
return p;
}
/* processing incoming normalization data ----------------------------------- */
/*
* process the data for one code point listed in UnicodeData;
* UnicodeData itself never maps a code point to both NFD and NFKD
*/
extern void
storeNorm(uint32_t code, Norm *norm) {
Norm *p=createNorm(code);
/* store the data */
uprv_memcpy(p, norm, sizeof(Norm));
/* store the decomposition string if there is one here */
if(norm->lenNFD!=0) {
uint32_t *s32=utm_allocN(utf32Mem, norm->lenNFD);
uprv_memcpy(s32, norm->nfd, norm->lenNFD*4);
p->nfd=s32;
} else if(norm->lenNFKD!=0) {
uint32_t *s32=utm_allocN(utf32Mem, norm->lenNFKD);
uprv_memcpy(s32, norm->nfkd, norm->lenNFKD*4);
p->nfkd=s32;
}
}
extern void
setCompositionExclusion(uint32_t code) {
uset_add(compositionExclusions, (UChar32)code);
}
static void
writeAllCC(FILE *f) {
uint32_t i;
UChar32 prevCode, code;
uint8_t prevCC, cc;
UBool isInBlockZero;
fprintf(f, "# Canonical_Combining_Class (ccc) values\n");
prevCode=0;
prevCC=0;
for(code=0; code<=0x110000;) {
if(code==0x110000) {
cc=0;
} else {
i=utrie_get32(normTrie, code, &isInBlockZero);
if(i==0 || isInBlockZero) {
cc=0;
} else {
cc=norms[i].udataCC;
}
}
if(prevCC!=cc) {
if(prevCC!=0) {
uint32_t lastCode=code-1;
if(prevCode==lastCode) {
fprintf(f, "%04lX:%d\n", (long)lastCode, prevCC);
} else {
fprintf(f, "%04lX..%04lX:%d\n",
(long)prevCode, (long)lastCode, prevCC);
}
}
prevCode=code;
prevCC=cc;
}
if(isInBlockZero) {
code+=UTRIE_DATA_BLOCK_LENGTH;
} else {
++code;
}
}
}
static UBool
hasMapping(uint32_t code) {
Norm *norm=norms+utrie_get32(normTrie, code, NULL);
return norm->lenNFD!=0 || norm->lenNFKD!=0;
}
static UBool
hasOneWayMapping(uint32_t code, UBool withCompat) {
for(;;) {
Norm *norm=norms+utrie_get32(normTrie, code, NULL);
uint8_t length;
if((length=norm->lenNFD)!=0) {
/*
* The canonical decomposition is a one-way mapping if
* - it does not map to exactly two code points
* - the code has ccc!=0
* - the code has the Composition_Exclusion property
* - its starter has a one-way mapping (loop for this)
* - its non-starter decomposes
*/
if( length!=2 ||
norm->udataCC!=0 ||
uset_contains(compositionExclusions, (UChar32)code) ||
hasMapping(norm->nfd[1])
) {
return TRUE;
}
code=norm->nfd[0]; /* continue */
} else if(withCompat && norm->lenNFKD!=0) {
return TRUE;
} else {
return FALSE;
}
}
}
static void
writeAllMappings(FILE *f, UBool withCompat) {
uint32_t i, code;
UBool isInBlockZero;
if(withCompat) {
fprintf(f, "\n# Canonical and compatibility decomposition mappings\n");
} else {
fprintf(f, "\n# Canonical decomposition mappings\n");
}
for(code=0; code<=0x10ffff;) {
i=utrie_get32(normTrie, code, &isInBlockZero);
if(isInBlockZero) {
code+=UTRIE_DATA_BLOCK_LENGTH;
} else {
if(i!=0) {
uint32_t *s32;
uint8_t length;
char separator;
if((length=norms[i].lenNFD)!=0) {
s32=norms[i].nfd;
separator= hasOneWayMapping(code, withCompat) ? '>' : '=';
} else if(withCompat && (length=norms[i].lenNFKD)!=0) {
s32=norms[i].nfkd;
separator='>';
}
if(length!=0) {
uint8_t j;
fprintf(f, "%04lX%c", (long)code, separator);
for(j=0; j<length; ++j) {
if(j!=0) {
fputc(' ', f);
}
fprintf(f, "%04lX", (long)s32[j]);
}
fputc('\n', f);
}
}
++code;
}
}
}
static void
writeNorm2TextFile(const char *path, const char *filename, UBool withCompat) {
FILE *f=usrc_createTextData(path, filename);
if(f==NULL) {
exit(U_FILE_ACCESS_ERROR);
}
writeAllCC(f);
writeAllMappings(f, withCompat);
fclose(f);
}
extern void
writeNorm2(const char *dataDir) {
writeNorm2TextFile(dataDir, "nfc.txt", FALSE);
writeNorm2TextFile(dataDir, "nfkc.txt", TRUE);
}
extern void
cleanUpData(void) {
utm_close(normMem);
utm_close(utf32Mem);
utrie_close(normTrie);
uprv_free(normTrie);
uset_close(compositionExclusions);
}
#endif /* #if !UCONFIG_NO_NORMALIZATION */
/*
* Hey, Emacs, please set the following:
*
* Local Variables:
* indent-tabs-mode: nil
* End:
*
*/

View File

@ -80,8 +80,10 @@ _ignored_properties = set((
"OMath",
"OUpper",
# Further properties that just contribute to others.
"CE", # Composition_Exclusion just contributes to Full_Composition_Exclusion.
"JSN",
# These properties just don't seem useful.
# They are deprecated since Unicode 6.0.
"XO_NFC",
"XO_NFD",
"XO_NFKC",
@ -235,7 +237,7 @@ _alg_names_ranges = []
# Starts with one range for all of Unicode without any properties.
# Setting values subdivides ranges.
_starts = array.array('l', [0, 0x110000]) # array of int32_t
_props = [{}] # props for 0 but not 110000
_props = [{}, {}] # props for 0 and 110000
def FindRange(x):
""" Binary search for x in the inversion map.
@ -243,6 +245,11 @@ def FindRange(x):
return bisect.bisect(_starts, x) - 1
def GetProps(c):
i = FindRange(c)
return _props[i]
def UpdateProps(start, end, update):
assert 0 <= start <= end <= 0x10ffff
(need_to_update, do_update, u) = (update[0], update[1], update[2])
@ -1108,6 +1115,101 @@ def WritePreparsedUCD(out_file):
if props:
WriteFieldsRangeProps(["cp"], start, end, props, out_file)
# Write Normalizer2 input files -------------------------------------------- ***
# Ported from genprops/store.c.
def WriteAllCC(out_file):
out_file.write("# Canonical_Combining_Class (ccc) values\n");
prev_start = 0
prev_cc = 0
for i in xrange(len(_starts)):
start = _starts[i]
props = _props[i]
cc = props.get("ccc")
if not cc: cc = 0
if prev_cc != cc:
if prev_cc != 0:
last_code_point = start - 1
if prev_start == last_code_point:
out_file.write("%04X:%d\n" % (last_code_point, prev_cc))
else:
out_file.write("%04X..%04X:%d\n" %
(prev_start, last_code_point, prev_cc))
prev_start = start
prev_cc = cc
def HasMapping(c):
props = GetProps(c)
dt = props.get("dt")
return dt and dt != "None"
def HasOneWayMapping(c, with_compat):
while True:
props = GetProps(c)
dt = props.get("dt")
if not dt or dt == "None":
return False # no mapping
elif dt == "Can":
# The canonical decomposition is a one-way mapping if
# - it does not map to exactly two code points
# - c has ccc!=0
# - c has the Composition_Exclusion property
# - its starter has a one-way mapping (loop for this)
# - its non-starter decomposes
nfd = props["dm"].split()
if (len(nfd) != 2 or
props.get("ccc") or
props.get("Comp_Ex") or
HasMapping(int(nfd[1], 16))):
return True
c = int(nfd[0], 16) # continue
else:
# c has a compatibility mapping.
return with_compat
def WriteAllMappings(out_file, with_compat):
if with_compat:
out_file.write("\n# Canonical and compatibility decomposition mappings\n")
else:
out_file.write("\n# Canonical decomposition mappings\n")
for i in xrange(len(_starts) - 1):
start = _starts[i]
end = _starts[i + 1] - 1
props = _props[i]
dm = props.get("dm")
if dm and dm[0] != '<' and (with_compat or props["dt"] == "Can"):
assert start == end
separator = '>' if HasOneWayMapping(start, with_compat) else '='
out_file.write("%04X%s%s\n" % (start, separator, dm))
def WriteNorm2TextFile(path, filename, with_compat):
year = datetime.date.today().strftime("%Y")
with open(os.path.join(path, filename), "w") as out_file:
out_file.write(
"""# Copyright (C) 1999-""" + year +
""", International Business Machines
# Corporation and others. All Rights Reserved.
#
# file name: """ + filename + """
#
# machine-generated by ICU preparseucd.py
#
""")
type = "NFKC" if with_compat else "NFC"
out_file.write("# Complete data for Unicode " + type + " normalization.\n")
out_file.write("# Unicode " + _ucd_version + "\n\n")
WriteAllCC(out_file)
WriteAllMappings(out_file, with_compat)
def WriteNorm2(path):
WriteNorm2TextFile(path, "nfc.txt", False)
WriteNorm2TextFile(path, "nfkc.txt", True)
# Preprocessing ------------------------------------------------------------ ***
_strip_re = re.compile("([0-9a-fA-F]+.+?) *#.*")
@ -1778,11 +1880,15 @@ def main():
if pnv:
raise Exception("no default values (@missing lines) for " +
"some Catalog or Enumerated properties: %s " % pnv)
# Write Normalizer2 input text files.
# Do this before compacting the data so that we need not handle fallbacks.
unidata_path = os.path.join(icu_src_root, "source", "data", "unidata")
norm2_path = os.path.join(unidata_path, "norm2")
if not os.path.exists(norm2_path): os.makedirs(norm2_path)
WriteNorm2(norm2_path)
# Optimize block vs. cp properties.
CompactBlocks()
# Write the ppucd.txt output file.
unidata_path = os.path.join(icu_src_root, "source", "data", "unidata")
if not os.path.exists(unidata_path): os.makedirs(unidata_path)
out_path = os.path.join(unidata_path, "ppucd.txt")
with open(out_path, "w") as out_file:
WritePreparsedUCD(out_file)