AuroraMiddleware/scuffed-code
1
0
Fork 0
Code Issues Pull Requests Releases Wiki Activity

ICU-1245 Fixing problem with [before] handling... Still not finished

Browse Source 
X-SVN-Rev: 6194
This commit is contained in:
Vladimir Weinstein 2001-10-11 21:19:10 +00:00
parent e179814c37
commit 3b0d80266e
2 changed files with 95 additions and 49 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

21
icu4c/source/i18n/ucol_tok.cpp
View File

@ -652,9 +652,9 @@ const UChar *ucol_tok_parseNextToken(UColTokenParser *src,
return src->current;
}
inline void getVirginBefore(UColTokenParser *src, UColToken *sourceToken, uint32_t strength, uint32_t *charsOffset, uint32_t *newCharsLen, UErrorCode *status) {
inline UColToken *getVirginBefore(UColTokenParser *src, UColToken *sourceToken, uint32_t strength, uint32_t *charsOffset, uint32_t *newCharsLen, UErrorCode *status) {
if(U_FAILURE(*status)) {
return;
return NULL;
}
/* this is a virgin before - we need to fish the anchor from the UCA */
collIterate s;
@ -695,12 +695,13 @@ inline void getVirginBefore(UColTokenParser *src, UColToken *sourceToken, uint32
uint32_t key = (*newCharsLen << 24) | *charsOffset;
sourceToken = (UColToken *)uhash_geti(src->tailored, (int32_t)key);
return sourceToken;
// if we found a tailored thing, we have to get one further down the line
if(sourceToken != NULL && sourceToken->strength != UCOL_TOK_RESET) {
src->extraCurrent--;
getVirginBefore(src, sourceToken, strength, charsOffset, newCharsLen, status);
}
//if(sourceToken != NULL && sourceToken->strength != UCOL_TOK_RESET) {
//src->extraCurrent--;
//getVirginBefore(src, sourceToken, strength, charsOffset, newCharsLen, status);
//}
}
@ -949,12 +950,12 @@ uint32_t ucol_tok_assembleTokenList(UColTokenParser *src, UParseError *parseErro
} else { /* we hit NULL */
/* we should be doing the else part */
sourceToken = sourceToken->listHeader->reset;
getVirginBefore(src, sourceToken, strength, &charsOffset, &newCharsLen, status);
sourceToken = NULL;
sourceToken = getVirginBefore(src, sourceToken, strength, &charsOffset, &newCharsLen, status);
//sourceToken = NULL;
}
} else {
getVirginBefore(src, sourceToken, strength, &charsOffset, &newCharsLen, status);
sourceToken = NULL;
sourceToken = getVirginBefore(src, sourceToken, strength, &charsOffset, &newCharsLen, status);
//sourceToken = NULL;
}
}

123
icu4c/source/test/cintltst/cmsccoll.c
View File

@ -3034,7 +3034,7 @@ static void TestNewJapanese() {
"\\u3077\\u309d",
"\\u30d7\\u30fd",
"\\u3077\\u3075",
};
};
static const char *test2[] = {
"\\u306f\\u309d", /* H\\u309d */
@ -3072,48 +3072,40 @@ static void TestNewJapanese() {
};
static const char *test3[] = {
"\\u3071\\u309d", /* H\\u309c\\u309d */
"\\u30d1\\u30fd", /* K\\u309c\\u30fd */
"\\u3074\\u309d",
"\\u30d4\\u30fd",
"\\u3077\\u309d",
"\\u30d7\\u30fd",
"\\u307a\\u309d",
"\\u30da\\u30fd",
"\\u221er\\u221e",
"\\u221eR#",
"\\u221et\\u221e",
"#r\\u221e",
"#R#",
"#t%",
"#T%",
"8t\\u221e",
"8T\\u221e",
"8t#",
"8T#",
"8t%",
"8T%",
"8t8",
"8T8",
"\\u03c9r\\u221e",
"\\u03a9R%",
"rr\\u221e",
"rR\\u221e",
"Rr\\u221e",
"RR\\u221e",
"RT%",
"rt8",
"tr\\u221e",
"tr8",
"TR8",
"tt8",
"\\u30b7\\u30e3\\u30fc\\u30ec",
};
genericLocaleStarter("ja_JP_JIS", test1, sizeof(test1)/sizeof(test1[0]));
genericLocaleStarter("ja_JP_JIS", test2, sizeof(test2)/sizeof(test2[0]));
/*
genericLocaleStarter("ja_JP_JIS", test3, sizeof(test3)/sizeof(test3[0]));
*/
/*genericLocaleStarter("ja_JP_JIS", test3, sizeof(test3)/sizeof(test3[0]));*/
}
/* this peace of code should be in some sort of verbose mode */
/* it gets the collation elements for elements and prints them */
/* This is useful when trying to see whether the problem is */
/*
uint32_t i = 0;
UCollationElements *it = NULL;
uint32_t CE;
coll = ucol_open("ja_JP_JIS", &status);
it = ucol_openElements(coll, string, 0, &status);
for(i = 0; i < sizeof(test)/sizeof(test[0]); i++) {
log_verbose("%s\n", test[i]);
uStringLen = u_unescape(test[i], string, 256);
ucol_setText(it, string, uStringLen, &status);
while((CE=ucol_next(it, &status)) != UCOL_NULLORDER) {
log_verbose("%08X\n", CE);
}
log_verbose("\n");
}
ucol_closeElements(it);
ucol_close(coll);
*/
static void TestStrCollIdenticalPrefix() {
const char* rule = "&\\ud9b0\\udc70=\\ud9b0\\udc71";
@ -3124,9 +3116,62 @@ static void TestStrCollIdenticalPrefix() {
genericRulesTestWithResult(rule, test, sizeof(test)/sizeof(test[0]), UCOL_EQUAL);
}
static void TestBeforePrefixFailure() {
const char* rule1 =
"&\\u30A7=\\u30A7=\\u3047=\\uff6a"
"&\\u30A8=\\u30A8=\\u3048=\\uff74"
"&[before 3]\\u30a7<<<\\u30c6|\\u30fc";
const char* rule2 =
"&[before 3]\\u30a7<<<\\u30c6|\\u30fc"
"&\\u30A7=\\u30A7=\\u3047=\\uff6a"
"&\\u30A8=\\u30A8=\\u3048=\\uff74";
const char* test[] = {
"\\u30c6\\u30fc\\u30bf",
"\\u30c6\\u30a7\\u30bf",
};
genericRulesStarter(rule1, test, sizeof(test)/sizeof(test[0]));
genericRulesStarter(rule2, test, sizeof(test)/sizeof(test[0]));
/* this peace of code should be in some sort of verbose mode */
/* it gets the collation elements for elements and prints them */
/* This is useful when trying to see whether the problem is */
{
UErrorCode status = U_ZERO_ERROR;
uint32_t i = 0;
UCollationElements *it = NULL;
uint32_t CE;
UChar string[256];
uint32_t uStringLen;
UCollator *coll = NULL;
uStringLen = u_unescape(rule1, string, 256);
coll = ucol_openRules(string, uStringLen, UCOL_DEFAULT, UCOL_DEFAULT, NULL, &status);
/*coll = ucol_open("ja_JP_JIS", &status);*/
it = ucol_openElements(coll, string, 0, &status);
for(i = 0; i < sizeof(test)/sizeof(test[0]); i++) {
log_verbose("%s\n", test[i]);
uStringLen = u_unescape(test[i], string, 256);
ucol_setText(it, string, uStringLen, &status);
while((CE=ucol_next(it, &status)) != UCOL_NULLORDER) {
log_verbose("%08X\n", CE);
}
log_verbose("\n");
}
ucol_closeElements(it);
ucol_close(coll);
}
}
void addMiscCollTest(TestNode** root)
{
/*addTest(root, &TestBeforePrefixFailure, "tscoll/cmsccoll/TestBeforePrefixFailure");*/
addTest(root, &TestStrCollIdenticalPrefix, "tscoll/cmsccoll/TestStrCollIdenticalPrefix");
addTest(root, &TestPrefix, "tscoll/cmsccoll/TestPrefix");
addTest(root, &TestNewJapanese, "tscoll/cmsccoll/TestNewJapanese");