diff --git a/icu4c/source/common/Makefile.in b/icu4c/source/common/Makefile.in index 36e3293eb4b..bf553105123 100644 --- a/icu4c/source/common/Makefile.in +++ b/icu4c/source/common/Makefile.in @@ -1,6 +1,6 @@ #****************************************************************************** # -# Copyright (C) 1999-2004, International Business Machines +# Copyright (C) 1999-2005, International Business Machines # Corporation and others. All Rights Reserved. # #****************************************************************************** @@ -67,7 +67,7 @@ ucnv.o ucnv_set.o ucnv_bld.o ucnv_cb.o ucnv_cnv.o ucnv_err.o ucnv_ext.o ucnv_io. ucnv_u7.o ucnv_u8.o ucnv_u16.o ucnv_u32.o ucnvscsu.o ucnvbocu.o \ ucnvmbcs.o ucnv2022.o ucnvhz.o ucnv_lmb.o ucnvisci.o ucnvdisp.o \ unistr.o unistr_case.o unistr_cnv.o unistr_props.o \ -utf_impl.o ustring.o ustr_cnv.o ustrcase.o cstring.o ustrfmt.o ustrtrns.o ustr_wcs.o uinvchar.o \ +utf_impl.o ustring.o ustr_cnv.o ustrcase.o ucasemap.o cstring.o ustrfmt.o ustrtrns.o ustr_wcs.o uinvchar.o \ normlzr.o unorm.o unormcmp.o unorm_it.o chariter.o schriter.o uchriter.o uiter.o \ uchar.o uprops.o ucase.o propname.o ubidi_props.o ubidi.o ubidiwrt.o ubidiln.o ushape.o unames.o \ uscript.o usc_impl.o uvector.o ustack.o uvectr32.o ucmp8.o \ diff --git a/icu4c/source/common/common.vcproj b/icu4c/source/common/common.vcproj index ef53007bac7..bc373e5da41 100644 --- a/icu4c/source/common/common.vcproj +++ b/icu4c/source/common/common.vcproj @@ -1751,6 +1751,26 @@ Outputs="..\..\include\unicode\$(InputFileName)"/> + + + + + + + + + + diff --git a/icu4c/source/common/ucase.c b/icu4c/source/common/ucase.c index 8620a38533d..8ea1488786b 100644 --- a/icu4c/source/common/ucase.c +++ b/icu4c/source/common/ucase.c @@ -918,13 +918,13 @@ enum { /* separator? */ #define is_sep(c) ((c)=='_' || (c)=='-' || (c)==0) -/* +/** * Requires non-NULL locale ID but otherwise does the equivalent of * checking for language codes as if uloc_getLanguage() were called: * Accepts both 2- and 3-letter codes and accepts case variants. */ -static int32_t -getCaseLocale(const char *locale, int32_t *locCache) { +U_CFUNC int32_t +ucase_getCaseLocale(const char *locale, int32_t *locCache) { int32_t result; char c; @@ -1171,7 +1171,7 @@ ucase_toFullLower(const UCaseProps *csp, UChar32 c, if(excWord&UCASE_EXC_CONDITIONAL_SPECIAL) { /* use hardcoded conditions and mappings */ - int32_t loc=getCaseLocale(locale, locCache); + int32_t loc=ucase_getCaseLocale(locale, locCache); /* * Test for conditional mappings first @@ -1318,7 +1318,7 @@ toUpperOrTitle(const UCaseProps *csp, UChar32 c, if(excWord&UCASE_EXC_CONDITIONAL_SPECIAL) { /* use hardcoded conditions and mappings */ - int32_t loc=getCaseLocale(locale, locCache); + int32_t loc=ucase_getCaseLocale(locale, locCache); if(loc==LOC_TURKISH && c==0x69) { /* diff --git a/icu4c/source/common/ucase.h b/icu4c/source/common/ucase.h index 999287ebab8..5781756692f 100644 --- a/icu4c/source/common/ucase.h +++ b/icu4c/source/common/ucase.h @@ -61,6 +61,14 @@ ucase_swap(const UDataSwapper *ds, U_CAPI void U_EXPORT2 ucase_addPropertyStarts(const UCaseProps *csp, const USetAdder *sa, UErrorCode *pErrorCode); +/** + * Requires non-NULL locale ID but otherwise does the equivalent of + * checking for language codes as if uloc_getLanguage() were called: + * Accepts both 2- and 3-letter codes and accepts case variants. + */ +U_CFUNC int32_t +ucase_getCaseLocale(const char *locale, int32_t *locCache); + /** * Bit mask for getting just the options from a string compare options word * that are relevant for case-insensitive string comparison. diff --git a/icu4c/source/common/ucasemap.c b/icu4c/source/common/ucasemap.c new file mode 100644 index 00000000000..02f94762815 --- /dev/null +++ b/icu4c/source/common/ucasemap.c @@ -0,0 +1,333 @@ +/* +******************************************************************************* +* +* Copyright (C) 2005, International Business Machines +* Corporation and others. All Rights Reserved. +* +******************************************************************************* +* file name: ucasemap.c +* encoding: US-ASCII +* tab size: 8 (not used) +* indentation:4 +* +* created on: 2005may06 +* created by: Markus W. Scherer +* +* Case mapping service object and functions using it. +*/ + +#include "unicode/utypes.h" +#include "unicode/uloc.h" +#include "unicode/ustring.h" +#include "unicode/ucasemap.h" +#include "cmemory.h" +#include "cstring.h" +#include "ucase.h" +#include "ustr_imp.h" + +/* UCaseMap service object -------------------------------------------------- */ + +struct UCaseMap { + const UCaseProps *csp; + char locale[32]; + int32_t locCache; + uint32_t options; +}; + +U_DRAFT UCaseMap * U_EXPORT2 +ucasemap_open(const char *locale, uint32_t options, UErrorCode *pErrorCode) { + UCaseMap *csm; + + if(U_FAILURE(*pErrorCode)) { + return NULL; + } + + csm=(UCaseMap *)uprv_malloc(sizeof(UCaseMap)); + if(csm==NULL) { + return NULL; + } + uprv_memset(csm, 0, sizeof(UCaseMap)); + + csm->csp=ucase_getSingleton(pErrorCode); + ucasemap_setLocale(csm, locale, pErrorCode); + if(U_FAILURE(*pErrorCode)) { + uprv_free(csm); + return NULL; + } + + csm->options=options; + return csm; +} + +U_DRAFT void U_EXPORT2 +ucasemap_close(UCaseMap *csm) { + if(csm!=NULL) { + uprv_free(csm); + } +} + +U_DRAFT const char * U_EXPORT2 +ucasemap_getLocale(const UCaseMap *csm) { + return csm->locale; +} + +U_DRAFT uint32_t U_EXPORT2 +ucasemap_getOptions(const UCaseMap *csm) { + return csm->options; +} + +U_DRAFT void U_EXPORT2 +ucasemap_setLocale(UCaseMap *csm, const char *locale, UErrorCode *pErrorCode) { + int32_t length; + + if(U_FAILURE(*pErrorCode)) { + return; + } + + length=uloc_getName(locale, csm->locale, (int32_t)sizeof(csm->locale), pErrorCode); + if(*pErrorCode==U_BUFFER_OVERFLOW_ERROR || length==sizeof(csm->locale)) { + *pErrorCode=U_ZERO_ERROR; + /* we only really need the language code for case mappings */ + length=uloc_getLanguage(locale, csm->locale, (int32_t)sizeof(csm->locale), pErrorCode); + } + if(length==sizeof(csm->locale)) { + *pErrorCode=U_BUFFER_OVERFLOW_ERROR; + } + csm->locCache=0; + if(U_SUCCESS(*pErrorCode)) { + ucase_getCaseLocale(csm->locale, &csm->locCache); + } else { + csm->locale[0]=0; + } +} + +U_DRAFT void U_EXPORT2 +ucasemap_setOptions(UCaseMap *csm, uint32_t options, UErrorCode *pErrorCode) { + csm->options=options; +} + +/* UTF-8 string case mappings ----------------------------------------------- */ + +/* append a full case mapping result, see UCASE_MAX_STRING_LENGTH */ +static U_INLINE int32_t +appendResult(uint8_t *dest, int32_t destIndex, int32_t destCapacity, + int32_t result, const UChar *s) { + UChar32 c; + int32_t length, destLength; + UErrorCode errorCode; + + /* decode the result */ + if(result<0) { + /* (not) original code point */ + c=~result; + length=-1; + } else if(result<=UCASE_MAX_STRING_LENGTH) { + c=U_SENTINEL; + length=result; + } else { + c=result; + length=-1; + } + + if(destIndexindex=csc->cpStart; + csc->dir=dir; + } else if(dir>0) { + /* reset for forward iteration */ + csc->index=csc->cpLimit; + csc->dir=dir; + } else { + /* continue current iteration direction */ + dir=csc->dir; + } + + if(dir<0) { + if(csc->startindex) { + U8_PREV((const uint8_t *)csc->p, csc->start, csc->index, c); + return c; + } + } else { + if(csc->indexlimit) { + U8_NEXT((const uint8_t *)csc->p, csc->index, csc->limit, c); + return c; + } + } + return U_SENTINEL; +} + +typedef int32_t U_CALLCONV +UCaseMapFull(const UCaseProps *csp, UChar32 c, + UCaseContextIterator *iter, void *context, + const UChar **pString, + const char *locale, int32_t *locCache); + +/* + * Case-maps [srcStart..srcLimit[ but takes + * context [0..srcLength[ into account. + */ +static int32_t +_caseMap(const UCaseMap *csm, UCaseMapFull *map, + uint8_t *dest, int32_t destCapacity, + const uint8_t *src, UCaseContext *csc, + int32_t srcStart, int32_t srcLimit, + UErrorCode *pErrorCode) { + const UChar *s; + UChar32 c; + int32_t srcIndex, destIndex; + int32_t locCache; + + locCache=csm->locCache; + + /* case mapping loop */ + srcIndex=srcStart; + destIndex=0; + while(srcIndexcpStart=srcIndex; + U8_NEXT(src, srcIndex, srcLimit, c); + csc->cpLimit=srcIndex; + c=map(csm->csp, c, utf8_caseContextIterator, csc, &s, csm->locale, &locCache); + destIndex=appendResult(dest, destIndex, destCapacity, c, s); + } + + if(destIndex>destCapacity) { + *pErrorCode=U_BUFFER_OVERFLOW_ERROR; + } + return destIndex; +} + +/* + * Implement argument checking and buffer handling + * for string case mapping as a common function. + */ +enum { + TO_LOWER, + TO_UPPER, + TO_TITLE, + FOLD_CASE +}; + +/* common internal function for public API functions */ + +static int32_t +caseMap(const UCaseMap *csm, + uint8_t *dest, int32_t destCapacity, + const uint8_t *src, int32_t srcLength, + int32_t toWhichCase, + UErrorCode *pErrorCode) { + UCaseContext csc={ NULL }; + int32_t destLength; + + /* check argument values */ + if(U_FAILURE(*pErrorCode)) { + return 0; + } + if( destCapacity<0 || + (dest==NULL && destCapacity>0) || + src==NULL || + srcLength<-1 + ) { + *pErrorCode=U_ILLEGAL_ARGUMENT_ERROR; + return 0; + } + + /* get the string length */ + if(srcLength==-1) { + srcLength=uprv_strlen((const char *)src); + } + + /* check for overlapping source and destination */ + if( dest!=NULL && + ((src>=dest && src<(dest+destCapacity)) || + (dest>=src && dest<(src+srcLength))) + ) { + *pErrorCode=U_ILLEGAL_ARGUMENT_ERROR; + return 0; + } + + destLength=0; + + csc.p=(void *)src; + csc.limit=srcLength; + + if(toWhichCase==TO_LOWER) { + destLength=_caseMap(csm, ucase_toFullLower, + dest, destCapacity, + src, &csc, + 0, srcLength, + pErrorCode); + } else /* if(toWhichCase==TO_UPPER) */ { + destLength=_caseMap(csm, ucase_toFullUpper, + dest, destCapacity, + src, &csc, + 0, srcLength, + pErrorCode); + } + + return u_terminateChars((char *)dest, destCapacity, destLength, pErrorCode); +} + +/* public API functions */ + +U_DRAFT int32_t U_EXPORT2 +ucasemap_utf8ToLower(const UCaseMap *csm, + char *dest, int32_t destCapacity, + const char *src, int32_t srcLength, + UErrorCode *pErrorCode) { + return caseMap(csm, + (uint8_t *)dest, destCapacity, + (const uint8_t *)src, srcLength, + TO_LOWER, pErrorCode); +} + +U_DRAFT int32_t U_EXPORT2 +ucasemap_utf8ToUpper(const UCaseMap *csm, + char *dest, int32_t destCapacity, + const char *src, int32_t srcLength, + UErrorCode *pErrorCode) { + return caseMap(csm, + (uint8_t *)dest, destCapacity, + (const uint8_t *)src, srcLength, + TO_UPPER, pErrorCode); +} diff --git a/icu4c/source/common/unicode/ucasemap.h b/icu4c/source/common/unicode/ucasemap.h new file mode 100644 index 00000000000..17c8f601f75 --- /dev/null +++ b/icu4c/source/common/unicode/ucasemap.h @@ -0,0 +1,180 @@ +/* +******************************************************************************* +* +* Copyright (C) 2005, International Business Machines +* Corporation and others. All Rights Reserved. +* +******************************************************************************* +* file name: ucasemap.h +* encoding: US-ASCII +* tab size: 8 (not used) +* indentation:4 +* +* created on: 2005may06 +* created by: Markus W. Scherer +* +* Case mapping service object and functions using it. +*/ + +#ifndef __UCASEMAP_H__ +#define __UCASEMAP_H__ + +#include "unicode/utypes.h" +#include "unicode/ustring.h" + +/** + * \file + * \brief C API: Unicode case mapping functions using a UCaseMap service object. + * + * The service object takes care of memory allocations, data loading, and setup + * for the attributes, as usual. + * + * Currently, the functionality provided here does not overlap with uchar.h + * and ustring.h. + * + * ucasemap_utf8ToLower() and ucasemap_utf8ToUpper() operate directly on + * UTF-8 strings. + */ + +/** + * UCaseMap is an opaque service object for newer ICU case mapping functions. + * Older functions did not use a service object. + * @draft ICU 3.4 + */ +struct UCaseMap; +typedef struct UCaseMap UCaseMap; /**< C typedef for struct UCaseMap. @draft ICU 3.4 */ + +/** + * Open a UCaseMap service object for a locale and a set of options. + * The locale ID and options are preprocessed so that functions using the + * service object need not process them in each call. + * + * @param locale ICU locale ID, used for language-dependent + * upper-/lower-/title-casing according to the Unicode standard. + * Usual semantics: ""=root, NULL=default locale, etc. + * @param options Options bit set, used for case folding and string comparisons. + * Same flags as for u_foldCase(), u_strFoldCase(), + * u_strCaseCompare(), etc. + * Use 0 or U_FOLD_CASE_DEFAULT for default behavior. + * @param pErrorCode Must be a valid pointer to an error code value, + * which must not indicate a failure before the function call. + * @return Pointer to a UCaseMap service object, if successful. + * + * @draft ICU 3.4 + */ +U_DRAFT UCaseMap * U_EXPORT2 +ucasemap_open(const char *locale, uint32_t options, UErrorCode *pErrorCode); + +/** + * Close a UCaseMap service object. + * @param csm Object to be closed. + * @draft ICU 3.4 + */ +U_DRAFT void U_EXPORT2 +ucasemap_close(UCaseMap *csm); + +/** + * Get the locale ID that is used for language-dependent case mappings. + * @param csm UCaseMap service object. + * @return locale ID + * @draft ICU 3.4 + */ +U_DRAFT const char * U_EXPORT2 +ucasemap_getLocale(const UCaseMap *csm); + +/** + * Get the options bit set that is used for case folding and string comparisons. + * @param csm UCaseMap service object. + * @return options bit set + * @draft ICU 3.4 + */ +U_DRAFT uint32_t U_EXPORT2 +ucasemap_getOptions(const UCaseMap *csm); + +/** + * Set the locale ID that is used for language-dependent case mappings. + * + * @param csm UCaseMap service object. + * @param locale Locale ID, see ucasemap_open(). + * @param pErrorCode Must be a valid pointer to an error code value, + * which must not indicate a failure before the function call. + * + * @see ucasemap_open + * @draft ICU 3.4 + */ +U_DRAFT void U_EXPORT2 +ucasemap_setLocale(UCaseMap *csm, const char *locale, UErrorCode *pErrorCode); + +/** + * Set the options bit set that is used for case folding and string comparisons. + * + * @param csm UCaseMap service object. + * @param options Options bit set, see ucasemap_open(). + * @param pErrorCode Must be a valid pointer to an error code value, + * which must not indicate a failure before the function call. + * + * @see ucasemap_open + * @draft ICU 3.4 + */ +U_DRAFT void U_EXPORT2 +ucasemap_setOptions(UCaseMap *csm, uint32_t options, UErrorCode *pErrorCode); + +/** + * Lowercase the characters in a UTF-8 string. + * Casing is locale-dependent and context-sensitive. + * The result may be longer or shorter than the original. + * The source string and the destination buffer must not overlap. + * + * @param csm UCaseMap service object. + * @param dest A buffer for the result string. The result will be NUL-terminated if + * the buffer is large enough. + * The contents is undefined in case of failure. + * @param destCapacity The size of the buffer (number of bytes). If it is 0, then + * dest may be NULL and the function will only return the length of the result + * without writing any of the result string. + * @param src The original string + * @param srcLength The length of the original string. If -1, then src must be NUL-terminated. + * @param pErrorCode Must be a valid pointer to an error code value, + * which must not indicate a failure before the function call. + * @return The length of the result string, if successful - or in case of a buffer overflow, + * in which case it will be greater than destCapacity. + * + * @see u_strToLower + * @draft ICU 3.4 + */ +U_DRAFT int32_t U_EXPORT2 +ucasemap_utf8ToLower(const UCaseMap *csm, + char *dest, int32_t destCapacity, + const char *src, int32_t srcLength, + UErrorCode *pErrorCode); + +/** + * Uppercase the characters in a UTF-8 string. + * Casing is locale-dependent and context-sensitive. + * The result may be longer or shorter than the original. + * The source string and the destination buffer must not overlap. + * + * @param csm UCaseMap service object. + * @param dest A buffer for the result string. The result will be NUL-terminated if + * the buffer is large enough. + * The contents is undefined in case of failure. + * @param destCapacity The size of the buffer (number of bytes). If it is 0, then + * dest may be NULL and the function will only return the length of the result + * without writing any of the result string. + * @param src The original string + * @param srcLength The length of the original string. If -1, then src must be NUL-terminated. + * @param pErrorCode Must be a valid pointer to an error code value, + * which must not indicate a failure before the function call. + * @return The length of the result string, if successful - or in case of a buffer overflow, + * in which case it will be greater than destCapacity. + * + * @see u_strToUpper + * @draft ICU 3.4 + */ +U_DRAFT int32_t U_EXPORT2 +ucasemap_utf8ToUpper(const UCaseMap *csm, + char *dest, int32_t destCapacity, + const char *src, int32_t srcLength, + UErrorCode *pErrorCode); + +#endif diff --git a/icu4c/source/common/ustrcase.c b/icu4c/source/common/ustrcase.c index 6946d5e4a5a..aedae0a3538 100644 --- a/icu4c/source/common/ustrcase.c +++ b/icu4c/source/common/ustrcase.c @@ -1,7 +1,7 @@ /* ******************************************************************************* * -* Copyright (C) 2001-2004, International Business Machines +* Copyright (C) 2001-2005, International Business Machines * Corporation and others. All Rights Reserved. * ******************************************************************************* @@ -121,7 +121,7 @@ UCaseMapFull(const UCaseProps *csp, UChar32 c, const char *locale, int32_t *locCache); /* - * Lowercases [srcStart..srcLimit[ but takes + * Case-maps [srcStart..srcLimit[ but takes * context [0..srcLength[ into account. */ static int32_t @@ -132,7 +132,7 @@ _caseMap(const UCaseProps *csp, UCaseMapFull *map, const char *locale, int32_t *locCache, UErrorCode *pErrorCode) { const UChar *s; - UChar32 c; + UChar32 c, c2; int32_t srcIndex, destIndex; /* case mapping loop */ @@ -143,7 +143,12 @@ _caseMap(const UCaseProps *csp, UCaseMapFull *map, U16_NEXT(src, srcIndex, srcLimit, c); csc->cpLimit=srcIndex; c=map(csp, c, utf16_caseContextIterator, csc, &s, locale, locCache); - destIndex=appendResult(dest, destIndex, destCapacity, c, s); + if((destIndexdestCapacity) { @@ -293,14 +298,19 @@ ustr_foldCase(const UCaseProps *csp, int32_t srcIndex, destIndex; const UChar *s; - UChar32 c; + UChar32 c, c2; /* case mapping loop */ srcIndex=destIndex=0; while(srcIndexdestCapacity) { diff --git a/icu4c/source/test/cintltst/cstrcase.c b/icu4c/source/test/cintltst/cstrcase.c index dcc6047fcf6..57256e6c52d 100644 --- a/icu4c/source/test/cintltst/cstrcase.c +++ b/icu4c/source/test/cintltst/cstrcase.c @@ -16,11 +16,13 @@ * Test file for string casing C API functions. */ +#include #include "unicode/utypes.h" #include "unicode/uchar.h" #include "unicode/ustring.h" #include "unicode/uloc.h" #include "unicode/ubrk.h" +#include "unicode/ucasemap.h" #include "cmemory.h" #include "cintltst.h" #include "cucdtst.h" @@ -643,3 +645,138 @@ TestCaseCompare() { log_err("error: u_memcasecmp(mixed, different, 5, default)=%ld instead of positive\n", result); } } + +/* test UCaseMap ------------------------------------------------------------ */ + +/* + * API test for UCaseMap; + * test cases for actual case mappings using UCaseMap see + * intltest utility/UnicodeStringTest/StringCaseTest/TestCasing + */ +U_CFUNC void +TestUCaseMap() { + static const char + aBc[] ={ 0x61, 0x42, 0x63, 0 }, + abc[] ={ 0x61, 0x62, 0x63, 0 }, + ABC[] ={ 0x41, 0x42, 0x43, 0 }, + ABCg[]={ 0x41, 0x42, 0x43, 0x67, 0 }, + defg[]={ 0x64, 0x65, 0x66, 0x67, 0 }; + char utf8Out[8]; + + UCaseMap *csm; + const char *locale; + uint32_t options; + int32_t length; + UErrorCode errorCode; + + errorCode=U_ZERO_ERROR; + csm=ucasemap_open("tur", 0xa5, &errorCode); + if(U_FAILURE(errorCode)) { + log_err("ucasemap_open(\"tur\") failed - %s\n", u_errorName(errorCode)); + return; + } + locale=ucasemap_getLocale(csm); + if(0!=strcmp(locale, "tr")) { + log_err("ucasemap_getLocale(ucasemap_open(\"tur\"))==%s!=\"tr\"\n", locale); + } + /* overly long locale IDs get truncated to their language code to avoid unnecessary allocation */ + ucasemap_setLocale(csm, "I-kLInGOn-the-quick-brown-fox-jumps-over-the-lazy-dog", &errorCode); + locale=ucasemap_getLocale(csm); + if(0!=strcmp(locale, "i-klingon")) { + log_err("ucasemap_getLocale(ucasemap_setLocale(\"I-kLInGOn-the-quick-br...\"))==%s!=\"i-klingon\"\n", locale); + } + + errorCode=U_ZERO_ERROR; + options=ucasemap_getOptions(csm); + if(options!=0xa5) { + log_err("ucasemap_getOptions(ucasemap_open(0xa5))==0x%lx!=0xa5\n", (long)options); + } + ucasemap_setOptions(csm, 0x333333, &errorCode); + options=ucasemap_getOptions(csm); + if(options!=0x333333) { + log_err("ucasemap_getOptions(ucasemap_setOptions(0x333333))==0x%lx!=0x333333\n", (long)options); + } + + /* test case mapping API; not all permutations necessary due to shared implementation code */ + + /* NUL terminated source */ + errorCode=U_ZERO_ERROR; + length=ucasemap_utf8ToLower(csm, utf8Out, (int32_t)sizeof(utf8Out), aBc, -1, &errorCode); + if(U_FAILURE(errorCode) || length!=3 || 0!=strcmp(abc, utf8Out)) { + log_err("ucasemap_utf8ToLower(aBc\\0) failed\n"); + } + + /* incoming failure code */ + errorCode=U_PARSE_ERROR; + strcpy(utf8Out, defg); + length=ucasemap_utf8ToLower(csm, utf8Out, (int32_t)sizeof(utf8Out), aBc, -1, &errorCode); + if(errorCode!=U_PARSE_ERROR || 0!=strcmp(defg, utf8Out)) { + log_err("ucasemap_utf8ToLower(failure) failed\n"); + } + + /* overlapping input & output */ + errorCode=U_ZERO_ERROR; + strcpy(utf8Out, aBc); + length=ucasemap_utf8ToUpper(csm, utf8Out, 2, utf8Out+1, 2, &errorCode); + if(errorCode!=U_ILLEGAL_ARGUMENT_ERROR || 0!=strcmp(aBc, utf8Out)) { + log_err("ucasemap_utf8ToUpper(overlap 1) failed\n"); + } + + /* overlap in the other direction */ + errorCode=U_ZERO_ERROR; + strcpy(utf8Out, aBc); + length=ucasemap_utf8ToUpper(csm, utf8Out+1, 2, utf8Out, 2, &errorCode); + if(errorCode!=U_ILLEGAL_ARGUMENT_ERROR || 0!=strcmp(aBc, utf8Out)) { + log_err("ucasemap_utf8ToUpper(overlap 2) failed\n"); + } + + /* NULL destination */ + errorCode=U_ZERO_ERROR; + strcpy(utf8Out, defg); + length=ucasemap_utf8ToLower(csm, NULL, (int32_t)sizeof(utf8Out), aBc, -1, &errorCode); + if(errorCode!=U_ILLEGAL_ARGUMENT_ERROR || 0!=strcmp(defg, utf8Out)) { + log_err("ucasemap_utf8ToLower(dest=NULL) failed\n"); + } + + /* destCapacity<0 */ + errorCode=U_ZERO_ERROR; + strcpy(utf8Out, defg); + length=ucasemap_utf8ToLower(csm, utf8Out, -2, aBc, -1, &errorCode); + if(errorCode!=U_ILLEGAL_ARGUMENT_ERROR || 0!=strcmp(defg, utf8Out)) { + log_err("ucasemap_utf8ToLower(destCapacity<0) failed\n"); + } + + /* NULL source */ + errorCode=U_ZERO_ERROR; + strcpy(utf8Out, defg); + length=ucasemap_utf8ToLower(csm, utf8Out, (int32_t)sizeof(utf8Out), NULL, -1, &errorCode); + if(errorCode!=U_ILLEGAL_ARGUMENT_ERROR || 0!=strcmp(defg, utf8Out)) { + log_err("ucasemap_utf8ToLower(src=NULL) failed\n"); + } + + /* srcLength<-1 */ + errorCode=U_ZERO_ERROR; + strcpy(utf8Out, defg); + length=ucasemap_utf8ToLower(csm, utf8Out, (int32_t)sizeof(utf8Out), aBc, -2, &errorCode); + if(errorCode!=U_ILLEGAL_ARGUMENT_ERROR || 0!=strcmp(defg, utf8Out)) { + log_err("ucasemap_utf8ToLower(srcLength<-1) failed\n"); + } + + /* buffer overflow */ + errorCode=U_ZERO_ERROR; + strcpy(utf8Out, defg); + length=ucasemap_utf8ToUpper(csm, utf8Out, 2, aBc, 3, &errorCode); + if(errorCode!=U_BUFFER_OVERFLOW_ERROR || length!=3 || 0!=strcmp(defg+2, utf8Out+2)) { + log_err("ucasemap_utf8ToUpper(overflow) failed\n"); + } + + /* dest not terminated (leaves g from defg alone) */ + errorCode=U_ZERO_ERROR; + strcpy(utf8Out, defg); + length=ucasemap_utf8ToUpper(csm, utf8Out, 3, aBc, 3, &errorCode); + if(errorCode!=U_STRING_NOT_TERMINATED_WARNING || length!=3 || 0!=strcmp(ABCg, utf8Out)) { + log_err("ucasemap_utf8ToUpper(overflow) failed\n"); + } + + ucasemap_close(csm); +} diff --git a/icu4c/source/test/cintltst/cucdtst.h b/icu4c/source/test/cintltst/cucdtst.h index 336476a0892..ef3a68c53bc 100644 --- a/icu4c/source/test/cintltst/cucdtst.h +++ b/icu4c/source/test/cintltst/cucdtst.h @@ -1,6 +1,6 @@ /******************************************************************** * COPYRIGHT: - * Copyright (c) 1997-2001, International Business Machines Corporation and + * Copyright (c) 1997-2005, International Business Machines Corporation and * others. All Rights Reserved. ********************************************************************/ /******************************************************************************** @@ -23,5 +23,6 @@ U_CFUNC void TestCaseUpper(void); U_CFUNC void TestCaseTitle(void); U_CFUNC void TestCaseFolding(void); U_CFUNC void TestCaseCompare(void); +U_CFUNC void TestUCaseMap(void); #endif diff --git a/icu4c/source/test/cintltst/custrtst.c b/icu4c/source/test/cintltst/custrtst.c index 358dd006472..c5f1a22b182 100644 --- a/icu4c/source/test/cintltst/custrtst.c +++ b/icu4c/source/test/cintltst/custrtst.c @@ -1,7 +1,7 @@ /* ****************************************************************************** * -* Copyright (C) 2002-2004, International Business Machines +* Copyright (C) 2002-2005, International Business Machines * Corporation and others. All Rights Reserved. * ****************************************************************************** @@ -64,6 +64,7 @@ void addUStringTest(TestNode** root) #endif addTest(root, &TestCaseFolding, "tsutil/custrtst/TestCaseFolding"); addTest(root, &TestCaseCompare, "tsutil/custrtst/TestCaseCompare"); + addTest(root, &TestUCaseMap, "tsutil/custrtst/TestUCaseMap"); } /* test data for TestStringFunctions ---------------------------------------- */ diff --git a/icu4c/source/test/intltest/strcase.cpp b/icu4c/source/test/intltest/strcase.cpp index a2dc70d7ee6..f6d5b6102d1 100644 --- a/icu4c/source/test/intltest/strcase.cpp +++ b/icu4c/source/test/intltest/strcase.cpp @@ -1,7 +1,7 @@ /* ******************************************************************************* * -* Copyright (C) 2002-2004, International Business Machines +* Copyright (C) 2002-2005, International Business Machines * Corporation and others. All Rights Reserved. * ******************************************************************************* @@ -21,6 +21,8 @@ #include "unicode/uloc.h" #include "unicode/locid.h" #include "unicode/ubrk.h" +#include "unicode/unistr.h" +#include "unicode/ucasemap.h" #include "ustrtest.h" #include "unicode/tstdtmod.h" @@ -32,9 +34,9 @@ StringCaseTest::runIndexedTest(int32_t index, UBool exec, const char *&name, cha switch (index) { case 0: name = "TestCaseConversion"; if (exec) TestCaseConversion(); break; case 1: - name = "TestTitleCasing"; + name = "TestCasing"; #if !UCONFIG_NO_BREAK_ITERATION - if(exec) TestTitleCasing(); + if(exec) TestCasing(); #endif break; @@ -45,21 +47,21 @@ StringCaseTest::runIndexedTest(int32_t index, UBool exec, const char *&name, cha void StringCaseTest::TestCaseConversion() { - UChar uppercaseGreek[] = + static const UChar uppercaseGreek[] = { 0x399, 0x395, 0x3a3, 0x3a5, 0x3a3, 0x20, 0x03a7, 0x3a1, 0x399, 0x3a3, 0x3a4, 0x39f, 0x3a3, 0 }; // "IESUS CHRISTOS" - UChar lowercaseGreek[] = + static const UChar lowercaseGreek[] = { 0x3b9, 0x3b5, 0x3c3, 0x3c5, 0x3c2, 0x20, 0x03c7, 0x3c1, 0x3b9, 0x3c3, 0x3c4, 0x3bf, 0x3c2, 0 }; // "iesus christos" - UChar lowercaseTurkish[] = + static const UChar lowercaseTurkish[] = { 0x69, 0x73, 0x74, 0x61, 0x6e, 0x62, 0x75, 0x6c, 0x2c, 0x20, 0x6e, 0x6f, 0x74, 0x20, 0x63, 0x6f, 0x6e, 0x73, 0x74, 0x61, 0x6e, 0x74, 0x0131, 0x6e, 0x6f, 0x70, 0x6c, 0x65, 0x21, 0 }; - UChar uppercaseTurkish[] = + static const UChar uppercaseTurkish[] = { 0x54, 0x4f, 0x50, 0x4b, 0x41, 0x50, 0x49, 0x20, 0x50, 0x41, 0x4c, 0x41, 0x43, 0x45, 0x2c, 0x20, 0x0130, 0x53, 0x54, 0x41, 0x4e, 0x42, 0x55, 0x4c, 0 }; @@ -318,45 +320,176 @@ StringCaseTest::TestCaseConversion() } } +// data-driven case mapping tests ------------------------------------------ *** + +enum { + TEST_LOWER, + TEST_UPPER, +#if !UCONFIG_NO_BREAK_ITERATION + TEST_TITLE, +#endif + TEST_COUNT +}; + +// names of TestData children in casing.txt +static const char *const dataNames[TEST_COUNT+1]={ + "lowercasing", + "uppercasing", +#if !UCONFIG_NO_BREAK_ITERATION + "titlecasing", +#endif + "" +}; + +void +StringCaseTest::TestCasingImpl(const UnicodeString &input, + const UnicodeString &output, + int32_t whichCase, + const char *localeID, uint32_t options) { + // UnicodeString + UnicodeString result; + const char *name; + + result=input; + switch(whichCase) { + case TEST_LOWER: + name="toLower"; + result.toLower(Locale(localeID)); + break; + case TEST_UPPER: + name="toUpper"; + result.toUpper(Locale(localeID)); + break; + default: + name=""; + break; // won't happen + } + if(result!=output) { + errln("error: UnicodeString.%s() got a wrong result for a test case from casing.res", name); + } + + // UTF-8 + char utf8In[100], utf8Out[100]; + int32_t utf8InLength, utf8OutLength, resultLength; + UChar *buffer; + + UCaseMap *csm; + UErrorCode errorCode; + + errorCode=U_ZERO_ERROR; + csm=ucasemap_open(localeID, options, &errorCode); + + u_strToUTF8(utf8In, (int32_t)sizeof(utf8In), &utf8InLength, input.getBuffer(), input.length(), &errorCode); + switch(whichCase) { + case TEST_LOWER: + name="ucasemap_utf8ToLower"; + utf8OutLength=ucasemap_utf8ToLower(csm, + utf8Out, (int32_t)sizeof(utf8Out), + utf8In, utf8InLength, &errorCode); + break; + case TEST_UPPER: + name="ucasemap_utf8ToUpper"; + utf8OutLength=ucasemap_utf8ToUpper(csm, + utf8Out, (int32_t)sizeof(utf8Out), + utf8In, utf8InLength, &errorCode); + break; + default: + name=""; + utf8OutLength=0; + break; // won't happen + } + buffer=result.getBuffer(utf8OutLength); + u_strFromUTF8(buffer, result.getCapacity(), &resultLength, utf8Out, utf8OutLength, &errorCode); + result.releaseBuffer(U_SUCCESS(errorCode) ? resultLength : 0); + + if(U_FAILURE(errorCode)) { + errln("error: %s() got an error for a test case from casing.res - %s", name, u_errorName(errorCode)); + } else if(result!=output) { + errln("error: %s() got a wrong result for a test case from casing.res", name); + } + ucasemap_close(csm); +} + #if !UCONFIG_NO_BREAK_ITERATION void -StringCaseTest::TestTitleCasing() { +StringCaseTest::TestTitleCasing(const UnicodeString &input, + const UnicodeString &output, + const char *localeID, + UBreakIterator *iter) { + UnicodeString result; + + result=input; + result.toTitle((BreakIterator *)iter, Locale(localeID)); + if(result!=output) { + errln("error: UnicodeString.toTitle() got a wrong result for a test case from casing.res"); + } +} + +#endif + +void +StringCaseTest::TestCasing() { UErrorCode status = U_ZERO_ERROR; +#if !UCONFIG_NO_BREAK_ITERATION UBreakIterator *iter; +#endif char cLocaleID[100]; - UnicodeString locale, input, result; - int32_t type; + UnicodeString locale, input, output, result; + int32_t whichCase, type; TestDataModule *driver = TestDataModule::getTestDataModule("casing", *this, status); if(U_SUCCESS(status)) { - TestData *casingTest = driver->createTestData("titlecasing", status); - const DataMap *myCase = NULL; - while(casingTest->nextCase(myCase, status)) { - locale = myCase->getString("Locale", status); - locale.extract(0, 0x7fffffff, cLocaleID, sizeof(cLocaleID), ""); - type = myCase->getInt("Type", status); - - - input = myCase->getString("Input", status); - if(type<0) { - iter=0; - } else { - iter=ubrk_open((UBreakIteratorType)type, cLocaleID, NULL, 0, &status); - } - + for(whichCase=0; whichCasecreateTestData(dataNames[whichCase], status); if(U_FAILURE(status)) { - errln("error: TestTitleCasing() ubrk_open(%d) failed for test case from casing.res: %s", type, u_errorName(status)); - status = U_ZERO_ERROR; - } else { - result=input; - result.toTitle((BreakIterator *)iter, Locale(cLocaleID)); - if(result!=myCase->getString("Output", status)) { - errln("error: TestTitleCasing() got a wrong result for test case from casing.res"); - } - ubrk_close(iter); + errln("TestCasing failed to createTestData(%s) - %s", dataNames[whichCase], u_errorName(status)); + break; } + const DataMap *myCase = NULL; + while(casingTest->nextCase(myCase, status)) { + locale = myCase->getString("Locale", status); + locale.extract(0, 0x7fffffff, cLocaleID, sizeof(cLocaleID), ""); + + input = myCase->getString("Input", status); + output = myCase->getString("Output", status); + +#if !UCONFIG_NO_BREAK_ITERATION + iter=NULL; + if(whichCase==TEST_TITLE) { + type = myCase->getInt("Type", status); + if(type>=0) { + iter=ubrk_open((UBreakIteratorType)type, cLocaleID, NULL, 0, &status); + } + } +#endif + + if(U_FAILURE(status)) { + errln("error: TestCasing() setup failed for %s test case from casing.res: %s", dataNames[whichCase], u_errorName(status)); + status = U_ZERO_ERROR; + } else { + switch(whichCase) { + case TEST_LOWER: + case TEST_UPPER: + TestCasingImpl(input, output, whichCase, cLocaleID, 0); + break; +#if !UCONFIG_NO_BREAK_ITERATION + case TEST_TITLE: + TestTitleCasing(input, output, cLocaleID, iter); + break; +#endif + default: + break; // won't happen + } + } + +#if !UCONFIG_NO_BREAK_ITERATION + if(iter!=NULL) { + ubrk_close(iter); + } +#endif + } + delete casingTest; } - delete casingTest; } delete driver; @@ -367,77 +500,4 @@ StringCaseTest::TestTitleCasing() { if(result!=UNICODE_STRING_SIMPLE("Stra\\u00dfe").unescape()) { errln("UnicodeString::toTitle(NULL) failed"); } - -#if 0 - char cLocaleID[100]; - UnicodeString in, expect, result, localeID; - UResourceBundle *casing, *titlecasing, *test, *res; - UErrorCode errorCode; - int32_t testIndex, type; - - errorCode=U_ZERO_ERROR; - loadTestData(errorCode); - casing=ures_openDirect("testdata", "casing", &errorCode); - if(U_FAILURE(errorCode)) { - errln("error: TestTitleCasing() is unable to open casing.res: %s", u_errorName(errorCode)); - return; - } - - // titlecasing tests - titlecasing=ures_getByKey(casing, "titlecasing", 0, &errorCode); - if(U_FAILURE(errorCode)) { - logln("TestTitleCasing() is unable to open get casing.res/titlecasing: %s", u_errorName(errorCode)); - } else { - UBreakIterator *iter; - - for(testIndex=0;; ++testIndex) { - // get test case - test=ures_getByIndex(titlecasing, testIndex, 0, &errorCode); - if(U_FAILURE(errorCode)) { - break; // done - } - - // get test case data - in=ures_getUnicodeStringByIndex(test, 0, &errorCode); - expect=ures_getUnicodeStringByIndex(test, 1, &errorCode); - localeID=ures_getUnicodeStringByIndex(test, 2, &errorCode); - - res=ures_getByIndex(test, 3, 0, &errorCode); - type=ures_getInt(res, &errorCode); - ures_close(res); - - if(U_FAILURE(errorCode)) { - errln("error: TestTitleCasing() is unable to get data for test case %ld from casing.res: %s", testIndex, u_errorName(errorCode)); - continue; // skip this test case - } - - // run this test case - localeID.extract(0, 0x7fffffff, cLocaleID, sizeof(cLocaleID), ""); - if(type<0) { - iter=0; - } else { - iter=ubrk_open((UBreakIteratorType)type, cLocaleID, in.getBuffer(), in.length(), &errorCode); - } - - if(U_FAILURE(errorCode)) { - errln("error: TestTitleCasing() ubrk_open(%d) failed for test case %d from casing.res: %s", type, testIndex, u_errorName(errorCode)); - } else { - result=in; - result.toTitle((BreakIterator *)iter, Locale(cLocaleID)); - if(result!=expect) { - errln("error: TestTitleCasing() got a wrong result for test case %ld from casing.res", testIndex); - } - } - - // clean up - ubrk_close(iter); - ures_close(test); - } - ures_close(titlecasing); - logln("TestTitleCasing() processed %ld test cases", testIndex); - } - - ures_close(casing); -#endif } -#endif diff --git a/icu4c/source/test/intltest/ustrtest.h b/icu4c/source/test/intltest/ustrtest.h index 8fc751ec4a1..c3f9f39947e 100644 --- a/icu4c/source/test/intltest/ustrtest.h +++ b/icu4c/source/test/intltest/ustrtest.h @@ -1,12 +1,14 @@ /******************************************************************** * COPYRIGHT: - * Copyright (c) 1997-2003, International Business Machines Corporation and + * Copyright (c) 1997-2005, International Business Machines Corporation and * others. All Rights Reserved. ********************************************************************/ #ifndef UNICODESTRINGTEST_H #define UNICODESTRINGTEST_H +#include "unicode/unistr.h" +#include "unicode/ubrk.h" #include "intltest.h" /** @@ -83,9 +85,18 @@ public: void runIndexedTest(int32_t index, UBool exec, const char *&name, char *par=0); void TestCaseConversion(); + + void TestCasingImpl(const UnicodeString &input, + const UnicodeString &output, + int32_t whichCase, + const char *localeID, uint32_t options); #if !UCONFIG_NO_BREAK_ITERATION - void TestTitleCasing(); + void TestTitleCasing(const UnicodeString &input, + const UnicodeString &output, + const char *localeID, + UBreakIterator *iter); #endif + void TestCasing(); }; #endif diff --git a/icu4c/source/test/testdata/casing.txt b/icu4c/source/test/testdata/casing.txt index c4c3bb144be..1e1c95c23e6 100644 --- a/icu4c/source/test/testdata/casing.txt +++ b/icu4c/source/test/testdata/casing.txt @@ -1,21 +1,38 @@ //******************************************************************************* //* -//* Copyright (C) 2002, International Business Machines +//* Copyright (C) 2002-2005, International Business Machines //* Corporation and others. All Rights Reserved. //* //******************************************************************************* casing { Info { - Description { "This is test data file for string casing" } + Description { "This is test data file for string casing." } - LongDescription { "each item is an array with" - "input string, result string, locale ID, break iterator" - "the break iterator is specified as an int, same as in UBreakIteratorType:" - "0=UBRK_CHARACTER 1=UBRK_WORD 2=UBRK_LINE 3=UBRK_SENTENCE 4=UBRK_TITLE -1=default" - } + LongDescription { + "each item is an array with\n" + "input string, result string, locale ID[, break iterator]\n" + "the break iterator (only for titlecasing) is specified as an int, same as in UBreakIteratorType:\n" + "0=UBRK_CHARACTER 1=UBRK_WORD 2=UBRK_LINE 3=UBRK_SENTENCE 4=UBRK_TITLE -1=default\n" + } } TestData { + lowercasing { + Headers { "Input", "Output", "Locale" } + Cases { + { " tHe QUIcK bRoWn", " the quick brown", "" }, + { "aBIΣßΣ/񟿿𐐅", "abiσßς/񟿿𐐭", "" }, + { "aBIΣßΣ/񟿿𐐅", "abıσßς/񟿿𐐭", "tur" } // tur: 3-letter code for Turkish + } + } + uppercasing { + Headers { "Input", "Output", "Locale" } + Cases { + { " tHe QUIcK bRoWn", " THE QUICK BROWN", "" }, + { "aBiσßς/ffi񟿿𐐭", "ABIΣSSΣ/FFI񟿿𐐅", "" }, + { "aBiσßς/ffi񟿿𐐭", "ABİΣSSΣ/FFI񟿿𐐅", "az" } // az same casing as tr + } + } titlecasing { Headers { "Input", "Output", "Locale", "Type" } Cases {