/* ******************************************************************************* * * Copyright (C) 2002-2004, International Business Machines * Corporation and others. All Rights Reserved. * ******************************************************************************* * file name: props2.c * encoding: US-ASCII * tab size: 8 (not used) * indentation:4 * * created on: 2002feb24 * created by: Markus W. Scherer * * Parse more Unicode Character Database files and store * additional Unicode character properties in bit set vectors. */ #include #include "unicode/utypes.h" #include "unicode/uchar.h" #include "unicode/uscript.h" #include "cstring.h" #include "cmemory.h" #include "utrie.h" #include "uprops.h" #include "propsvec.h" #include "uparse.h" #include "genprops.h" #define LENGTHOF(array) (int32_t)(sizeof(array)/sizeof((array)[0])) /* data --------------------------------------------------------------------- */ static UNewTrie *trie; uint32_t *pv; static int32_t pvCount; /* miscellaneous ------------------------------------------------------------ */ static char * trimTerminateField(char *s, char *limit) { /* trim leading whitespace */ s=(char *)u_skipWhitespace(s); /* trim trailing whitespace */ while(sucdFile, fields[0][0]); exit(*pErrorCode); } ++limit; /* parse property alias */ s=trimTerminateField(fields[1][0], fields[1][1]); value=u_getPropertyValueEnum(sen->prop, s); if(value<0) { if(sen->prop==UCHAR_BLOCK) { if(isToken("Greek", s)) { value=UBLOCK_GREEK; /* Unicode 3.2 renames this to "Greek and Coptic" */ } else if(isToken("Combining Marks for Symbols", s)) { value=UBLOCK_COMBINING_MARKS_FOR_SYMBOLS; /* Unicode 3.2 renames this to "Combining Diacritical Marks for Symbols" */ } else if(isToken("Private Use", s)) { value=UBLOCK_PRIVATE_USE; /* Unicode 3.2 renames this to "Private Use Area" */ } } } if(value<0) { fprintf(stderr, "genprops error: unknown %s name in %s.txt field 1 at %s\n", sen->propName, sen->ucdFile, s); exit(U_PARSE_ERROR); } uv=(uint32_t)(value<vecShift); if((uv&sen->vecMask)!=uv) { fprintf(stderr, "genprops error: %s value overflow (0x%x) at %s\n", sen->propName, (int)uv, s); exit(U_INTERNAL_PROGRAM_ERROR); } if(!upvec_setValue(pv, start, limit, sen->vecWord, uv, sen->vecMask, pErrorCode)) { fprintf(stderr, "genprops error: unable to set %s code: %s\n", sen->propName, u_errorName(*pErrorCode)); exit(*pErrorCode); } } static void parseSingleEnumFile(char *filename, char *basename, const char *suffix, const SingleEnum *sen, UErrorCode *pErrorCode) { char *fields[2][2]; if(pErrorCode==NULL || U_FAILURE(*pErrorCode)) { return; } writeUCDFilename(basename, sen->ucdFile, suffix); u_parseDelimitedFile(filename, ';', fields, 2, singleEnumLineFn, (void *)sen, pErrorCode); if(U_FAILURE(*pErrorCode)) { fprintf(stderr, "error parsing %s.txt: %s\n", sen->ucdFile, u_errorName(*pErrorCode)); } } /* parse files with multiple binary properties ------------------------------ */ struct Binary { const char *propName; int32_t vecWord, vecShift; }; typedef struct Binary Binary; struct Binaries { const char *ucdFile; const Binary *binaries; int32_t binariesCount; }; typedef struct Binaries Binaries; static const Binary propListNames[]={ { "White_Space", 1, UPROPS_WHITE_SPACE }, { "Bidi_Control", 1, UPROPS_BIDI_CONTROL }, { "Join_Control", 1, UPROPS_JOIN_CONTROL }, { "Dash", 1, UPROPS_DASH }, { "Hyphen", 1, UPROPS_HYPHEN }, { "Quotation_Mark", 1, UPROPS_QUOTATION_MARK }, { "Terminal_Punctuation", 1, UPROPS_TERMINAL_PUNCTUATION }, { "Hex_Digit", 1, UPROPS_HEX_DIGIT }, { "ASCII_Hex_Digit", 1, UPROPS_ASCII_HEX_DIGIT }, { "Ideographic", 1, UPROPS_IDEOGRAPHIC }, { "Diacritic", 1, UPROPS_DIACRITIC }, { "Extender", 1, UPROPS_EXTENDER }, { "Noncharacter_Code_Point", 1, UPROPS_NONCHARACTER_CODE_POINT }, { "Grapheme_Link", 1, UPROPS_GRAPHEME_LINK }, { "IDS_Binary_Operator", 1, UPROPS_IDS_BINARY_OPERATOR }, { "IDS_Trinary_Operator", 1, UPROPS_IDS_TRINARY_OPERATOR }, { "Radical", 1, UPROPS_RADICAL }, { "Unified_Ideograph", 1, UPROPS_UNIFIED_IDEOGRAPH }, { "Deprecated", 1, UPROPS_DEPRECATED }, { "Soft_Dotted", 1, UPROPS_SOFT_DOTTED }, { "Logical_Order_Exception", 1, UPROPS_LOGICAL_ORDER_EXCEPTION }, /* new properties in Unicode 4.0.1 */ { "STerm", 2, UPROPS_V2_S_TERM }, { "Variation_Selector", 2, UPROPS_V2_VARIATION_SELECTOR } }; static const Binaries propListBinaries={ "PropList", propListNames, LENGTHOF(propListNames) }; static const Binary derCorePropsNames[]={ { "XID_Start", 1, UPROPS_XID_START }, { "XID_Continue", 1, UPROPS_XID_CONTINUE }, /* before Unicode 4/ICU 2.6/format version 3.2, these used to be Other_XYZ from PropList.txt */ { "Math", 1, UPROPS_MATH }, { "Alphabetic", 1, UPROPS_ALPHABETIC }, { "Lowercase", 1, UPROPS_LOWERCASE }, { "Uppercase", 1, UPROPS_UPPERCASE }, { "Grapheme_Extend", 1, UPROPS_GRAPHEME_EXTEND }, { "Default_Ignorable_Code_Point", 1, UPROPS_DEFAULT_IGNORABLE_CODE_POINT }, /* new properties bits in ICU 2.6/format version 3.2 */ { "ID_Start", 1, UPROPS_ID_START }, { "ID_Continue", 1, UPROPS_ID_CONTINUE }, { "Grapheme_Base", 1, UPROPS_GRAPHEME_BASE } }; static const Binaries derCorePropsBinaries={ "DerivedCoreProperties", derCorePropsNames, LENGTHOF(derCorePropsNames) }; static char ignoredProps[100][64]; static int32_t ignoredPropsCount; static void addIgnoredProp(char *s, char *limit) { int32_t i; s=trimTerminateField(s, limit); for(i=0; iucdFile, fields[0][0]); exit(*pErrorCode); } ++limit; /* parse binary property name */ s=(char *)u_skipWhitespace(fields[1][0]); for(i=0;; ++i) { if(i==bin->binariesCount) { /* ignore unrecognized properties */ addIgnoredProp(s, fields[1][1]); return; } if(isToken(bin->binaries[i].propName, s)) { break; } } if(bin->binaries[i].vecShift>=32) { fprintf(stderr, "genprops error: shift value %d>=32 for %s %s\n", (int)bin->binaries[i].vecShift, bin->ucdFile, bin->binaries[i].propName); exit(U_INTERNAL_PROGRAM_ERROR); } uv=U_MASK(bin->binaries[i].vecShift); if(!upvec_setValue(pv, start, limit, bin->binaries[i].vecWord, uv, uv, pErrorCode)) { fprintf(stderr, "genprops error: unable to set %s code: %s\n", bin->binaries[i].propName, u_errorName(*pErrorCode)); exit(*pErrorCode); } } static void parseBinariesFile(char *filename, char *basename, const char *suffix, const Binaries *bin, UErrorCode *pErrorCode) { char *fields[2][2]; int32_t i; if(pErrorCode==NULL || U_FAILURE(*pErrorCode)) { return; } writeUCDFilename(basename, bin->ucdFile, suffix); ignoredPropsCount=0; u_parseDelimitedFile(filename, ';', fields, 2, binariesLineFn, (void *)bin, pErrorCode); if(U_FAILURE(*pErrorCode)) { fprintf(stderr, "error parsing %s.txt: %s\n", bin->ucdFile, u_errorName(*pErrorCode)); } for(i=0; iucdFile); } } /* -------------------------------------------------------------------------- */ U_CFUNC void initAdditionalProperties() { pv=upvec_open(UPROPS_VECTOR_WORDS, 20000); } U_CFUNC void generateAdditionalProperties(char *filename, const char *suffix, UErrorCode *pErrorCode) { char *basename; basename=filename+uprv_strlen(filename); /* process various UCD .txt files */ /* add Han numeric types & values */ parseMultiFieldFile(filename, basename, "DerivedNumericValues", suffix, 2, numericLineFn, pErrorCode); /* set proper bidi class for unassigned code points (Cn) */ parseTwoFieldFile(filename, basename, "DerivedBidiClass", suffix, bidiClassLineFn, pErrorCode); parseTwoFieldFile(filename, basename, "DerivedAge", suffix, ageLineFn, pErrorCode); /* * UTR 24 says: * Section 2: * "Common - For characters that may be used * within multiple scripts, * or any unassigned code points." * * Section 4: * "The value COMMON is the default value, * given to all code points that are not * explicitly mentioned in the data file." * * COMMON==USCRIPT_COMMON==0 - nothing to do */ parseSingleEnumFile(filename, basename, suffix, &scriptSingleEnum, pErrorCode); parseSingleEnumFile(filename, basename, suffix, &blockSingleEnum, pErrorCode); parseBinariesFile(filename, basename, suffix, &propListBinaries, pErrorCode); parseBinariesFile(filename, basename, suffix, &derCorePropsBinaries, pErrorCode); /* * LineBreak-4.0.0.txt: * - All code points, assigned and unassigned, that are not listed * explicitly are given the value "XX". * * XX==U_LB_UNKNOWN==0 - nothing to do */ parseSingleEnumFile(filename, basename, suffix, &lineBreakSingleEnum, pErrorCode); parseSingleEnumFile(filename, basename, suffix, &jtSingleEnum, pErrorCode); parseSingleEnumFile(filename, basename, suffix, &jgSingleEnum, pErrorCode); /* * Preset East Asian Width defaults: * * http://www.unicode.org/reports/tr11/#Unassigned * 7.1 Unassigned and Private Use characters * * All unassigned characters are by default classified as non-East Asian neutral, * except for the range U+20000 to U+2FFFD, * since all code positions from U+20000 to U+2FFFD are intended for CJK ideographs (W). * All Private use characters are by default classified as ambiguous, * since their definition depends on context. * * N for all ==0 - nothing to do * A for Private Use * W for plane 2 */ *pErrorCode=U_ZERO_ERROR; if( !upvec_setValue(pv, 0xe000, 0xf900, 0, (uint32_t)(U_EA_AMBIGUOUS<15 || (*end!='.' && *end!=' ' && *end!='\t' && *end!=0)) { fprintf(stderr, "genprops: syntax error in DerivedAge.txt field 1 at %s\n", fields[1][0]); *pErrorCode=U_PARSE_ERROR; exit(U_PARSE_ERROR); } version=value<<4; /* parse minor version number */ if(*end=='.') { s=(char *)u_skipWhitespace(end+1); value=(uint32_t)uprv_strtoul(s, &end, 10); if(s==end || value>15 || (*end!=' ' && *end!='\t' && *end!=0)) { fprintf(stderr, "genprops: syntax error in DerivedAge.txt field 1 at %s\n", fields[1][0]); *pErrorCode=U_PARSE_ERROR; exit(U_PARSE_ERROR); } version|=value; } if(!upvec_setValue(pv, start, limit, 0, version<=0x80000000) { fprintf(stderr, "genprops: syntax error in DerivedNumericValues.txt field 1 at %s\n", fields[0][0]); exit(U_PARSE_ERROR); } } /* * Unicode 4.0.1 removes the third column that used to list the numeric type. * Assume that either the data is the same as in UnicodeData.txt, * or else that the numeric type is "numeric". * This should work because we only expect to add numeric values for * Han characters; for those, UnicodeData.txt lists only ranges without * specific properties for single characters. */ for(; start