static int32_t parseSetNum(const char *setNumStr, UErrorCode &errorCode) {
if (U_FAILURE(errorCode)) { return -1; }
if (uprv_strncmp(setNumStr, "set", 3) != 0) {
errorCode = U_INVALID_FORMAT_ERROR;
return -1;
}
int32_t i = 3;
int32_t setNum = 0;
while (setNumStr[i] != 0) {
int32_t digit = setNumStr[i] - '0';
if (digit < 0 || 9 < digit) {
errorCode = U_INVALID_FORMAT_ERROR;
return -1;
}
setNum = 10 * setNum + digit;
++i;
}
// Rule set number must not be zero. (0 is used to indicate "not found" by hashmap.)
// Currently ICU data conveniently starts numbering rule sets from 1.
if (setNum == 0) {
errorCode = U_INVALID_FORMAT_ERROR;
return -1;
} else {
return setNum;
}
}
UBool
DigitList::operator==(const DigitList& that) const
{
return ((this == &that) ||
(fDecimalAt == that.fDecimalAt &&
fCount == that.fCount &&
fIsPositive == that.fIsPositive &&
fRoundingMode == that.fRoundingMode &&
uprv_strncmp(fDigits, that.fDigits, fCount) == 0));
}
static int32_t
_getStringOrCopyKey(const char *path, const char *locale,
const char *tableKey,
const char* subTableKey,
const char *itemKey,
const char *substitute,
UChar *dest, int32_t destCapacity,
UErrorCode *pErrorCode) {
const UChar *s = NULL;
int32_t length = 0;
if(itemKey==NULL) {
/* top-level item: normal resource bundle access */
UResourceBundle *rb;
rb=ures_open(path, locale, pErrorCode);
if(U_SUCCESS(*pErrorCode)) {
s=ures_getStringByKey(rb, tableKey, &length, pErrorCode);
/* see comment about closing rb near "return item;" in _res_getTableStringWithFallback() */
ures_close(rb);
}
} else {
/* Language code should not be a number. If it is, set the error code. */
if (!uprv_strncmp(tableKey, "Languages", 9) && uprv_strtol(itemKey, NULL, 10)) {
*pErrorCode = U_MISSING_RESOURCE_ERROR;
} else {
/* second-level item, use special fallback */
s=uloc_getTableStringWithFallback(path, locale,
tableKey,
subTableKey,
itemKey,
&length,
pErrorCode);
}
}
if(U_SUCCESS(*pErrorCode)) {
int32_t copyLength=uprv_min(length, destCapacity);
if(copyLength>0 && s != NULL) {
u_memcpy(dest, s, copyLength);
}
} else {
/* no string from a resource bundle: convert the substitute */
length=(int32_t)uprv_strlen(substitute);
u_charsToUChars(substitute, dest, uprv_min(length, destCapacity));
*pErrorCode=U_USING_DEFAULT_WARNING;
}
return u_terminateUChars(dest, destCapacity, length, pErrorCode);
}
static void U_CALLCONV
ageLineFn(void *context,
char *fields[][2], int32_t fieldCount,
UErrorCode *pErrorCode) {
char *s, *numberLimit;
uint32_t value, start, end, version;
u_parseCodePointRange(fields[0][0], &start, &end, pErrorCode);
if(U_FAILURE(*pErrorCode)) {
fprintf(stderr, "genprops: syntax error in DerivedAge.txt field 0 at %s\n", fields[0][0]);
exit(*pErrorCode);
}
/* ignore "unassigned" (the default is already set to 0.0) */
s=(char *)u_skipWhitespace(fields[1][0]);
if(0==uprv_strncmp(s, "unassigned", 10)) {
return;
}
/* parse version number */
value=(uint32_t)uprv_strtoul(s, &numberLimit, 10);
if(s==numberLimit || value==0 || value>15 || (*numberLimit!='.' && *numberLimit!=' ' && *numberLimit!='\t' && *numberLimit!=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(*numberLimit=='.') {
s=(char *)u_skipWhitespace(numberLimit+1);
value=(uint32_t)uprv_strtoul(s, &numberLimit, 10);
if(s==numberLimit || value>15 || (*numberLimit!=' ' && *numberLimit!='\t' && *numberLimit!=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(start==0 && end==0x10ffff) {
/* Also set bits for initialValue and errorValue. */
end=UPVEC_MAX_CP;
}
upvec_setValue(pv, start, end, 0, version<<UPROPS_AGE_SHIFT, UPROPS_AGE_MASK, pErrorCode);
if(U_FAILURE(*pErrorCode)) {
fprintf(stderr, "genprops error: unable to set character age: %s\n", u_errorName(*pErrorCode));
exit(*pErrorCode);
}
}
/*
* Extract the setting after the '=' and store it in flag excluding the newline character.
*/
static int32_t extractFlag(char* buffer, int32_t bufferSize, char* flag, int32_t flagSize, const char **flagNames, int32_t numOfFlags, UErrorCode *status) {
int32_t i, idx = -1;
char *pBuffer;
int32_t offset=0;
UBool bufferWritten = FALSE;
if (buffer[0] != 0) {
/* Get the offset (i.e. position after the '=') */
offset = getFlagOffset(buffer, bufferSize);
pBuffer = buffer+offset;
for(i = 0;;i++) {
if (i >= flagSize) {
*status = U_BUFFER_OVERFLOW_ERROR;
return -1;
}
if (pBuffer[i+1] == 0) {
/* Indicates a new line character. End here. */
flag[i] = 0;
break;
}
flag[i] = pBuffer[i];
if (i == 0) {
bufferWritten = TRUE;
}
}
}
if (!bufferWritten) {
flag[0] = 0;
}
if (flagNames != NULL && offset>0) {
offset--; /* Move offset back 1 because of '='*/
for (i = 0; i < numOfFlags; i++) {
if (uprv_strncmp(buffer, flagNames[i], offset) == 0) {
idx = i;
break;
}
}
}
return idx;
}
static const char *udata_pathiter_next(UDataPathIterator *iter)
{
const char *path = NULL;
uint32_t pathLen = 0;
const char *pathBasename;
do
{
if( iter->nextPath == NULL ) {
break;
}
path = iter->nextPath;
if(iter->nextPath == iter->itemPath) { /* we were processing item's path. */
iter->nextPath = iter->path; /* start with regular path next tm. */
pathLen = (int32_t)uprv_strlen(path);
} else {
/* fix up next for next time */
iter->nextPath = uprv_strchr(path, U_PATH_SEP_CHAR);
if(iter->nextPath == NULL) {
/* segment: entire path */
pathLen = (int32_t)uprv_strlen(path);
} else {
/* segment: until next segment */
pathLen = (int32_t)(iter->nextPath - path);
if(*iter->nextPath) { /* skip divider */
iter->nextPath ++;
}
}
}
if(pathLen == 0) {
continue;
}
#ifdef UDATA_DEBUG
fprintf(stderr, "rest of path (IDD) = %s\n", path);
fprintf(stderr, " ");
{
uint32_t qqq;
for(qqq=0;qqq<pathLen;qqq++)
{
fprintf(stderr, " ");
}
fprintf(stderr, "^\n");
}
#endif
uprv_strncpy(iter->pathBuffer, path, pathLen);
iter->pathBuffer[pathLen] = 0;
/* check for .dat files */
pathBasename = findBasename(iter->pathBuffer);
if(iter->checkLastFour == TRUE &&
(pathLen>=4) &&
uprv_strncmp(iter->pathBuffer +(pathLen-4),iter->suffix,4)==0 && /* suffix matches */
uprv_strncmp(findBasename(iter->pathBuffer),iter->basename,iter->basenameLen)==0 && /* base matches */
uprv_strlen(pathBasename)==(iter->basenameLen+4)) { /* base+suffix = full len */
#ifdef UDATA_DEBUG
fprintf(stderr, "Have %s file on the path: %s\n", iter->suffix, iter->pathBuffer);
#endif
/* do nothing */
}
else
{ /* regular dir path */
if(iter->pathBuffer[pathLen-1] != U_FILE_SEP_CHAR) {
if((pathLen>=4) &&
uprv_strncmp(iter->pathBuffer+(pathLen-4), ".dat", 4) == 0)
{
#ifdef UDATA_DEBUG
fprintf(stderr, "skipping non-directory .dat file %s\n", iter->pathBuffer);
#endif
continue;
}
/* Check if it is a directory with the same name as our package */
if(iter->packageStubLen &&
(pathLen > iter->packageStubLen) &&
!uprv_strcmp(iter->pathBuffer + pathLen - iter->packageStubLen, iter->packageStub)) {
#ifdef UDATA_DEBUG
fprintf(stderr, "Found stub %s ( will add package %s of len %d)\n", iter->packageStub, iter->basename, iter->basenameLen);
#endif
pathLen -= iter->packageStubLen;
}
iter->pathBuffer[pathLen++] = U_FILE_SEP_CHAR;
}
uprv_strncpy(iter->pathBuffer + pathLen, /* + basename */
iter->packageStub+1,
iter->packageStubLen-1);
pathLen += iter->packageStubLen-1;
if(*iter->suffix) /* tack on suffix */
{
uprv_strcpy(iter->pathBuffer + pathLen,
iter->suffix);
pathLen += (int32_t)uprv_strlen(iter->suffix);
}
}
#ifdef UDATA_DEBUG
fprintf(stderr, " --> %s\n", iter->pathBuffer);
#endif
return iter->pathBuffer;
} while(iter->path);
/* fell way off the end */
return NULL;
}
static void TestFileStream(void){
int32_t c = 0;
int32_t c1=0;
UErrorCode status = U_ZERO_ERROR;
const char* testdatapath = loadTestData(&status);
char* fileName = (char*) malloc(uprv_strlen(testdatapath) +10);
FileStream* stream = NULL;
/* these should not be closed */
FileStream* pStdin = T_FileStream_stdin();
FileStream* pStdout = T_FileStream_stdout();
FileStream* pStderr = T_FileStream_stderr();
const char* testline = "This is a test line";
int32_t bufLen = (int32_t)strlen(testline)+10;
char* buf = (char*) malloc(bufLen);
int32_t retLen = 0;
if(pStdin==NULL){
log_err("failed to get T_FileStream_stdin()");
}
if(pStdout==NULL){
log_err("failed to get T_FileStream_stdout()");
}
if(pStderr==NULL){
log_err("failed to get T_FileStream_stderr()");
}
uprv_strcpy(fileName,testdatapath);
uprv_strcat(fileName,".dat");
stream = T_FileStream_open(fileName, "r");
if(stream==NULL){
log_data_err("T_FileStream_open failed to open %s\n",fileName);
} else {
if(!T_FileStream_file_exists(fileName)){
log_data_err("T_FileStream_file_exists failed to verify existence of %s \n",fileName);
}
retLen=T_FileStream_read(stream,&c,1);
if(retLen==0){
log_data_err("T_FileStream_read failed to read from %s \n",fileName);
}
retLen=0;
T_FileStream_rewind(stream);
T_FileStream_read(stream,&c1,1);
if(c!=c1){
log_data_err("T_FileStream_rewind failed to rewind %s \n",fileName);
}
T_FileStream_rewind(stream);
c1 = T_FileStream_peek(stream);
if(c!=c1){
log_data_err("T_FileStream_peek failed to peekd %s \n",fileName);
}
c = T_FileStream_getc(stream);
T_FileStream_ungetc(c,stream);
if(c!= T_FileStream_getc(stream)){
log_data_err("T_FileStream_ungetc failed to d %s \n",fileName);
}
if(T_FileStream_size(stream)<=0){
log_data_err("T_FileStream_size failed to d %s \n",fileName);
}
if(T_FileStream_error(stream)){
log_data_err("T_FileStream_error shouldn't have an error %s\n",fileName);
}
if(!T_FileStream_error(NULL)){
log_err("T_FileStream_error didn't get an error %s\n",fileName);
}
T_FileStream_putc(stream, 0x20);
if(!T_FileStream_error(stream)){
/*
Warning
writing to a read-only file may not consistently fail on all platforms
(e.g. HP-UX, FreeBSD, MacOSX)
*/
log_verbose("T_FileStream_error didn't get an error when writing to a readonly file %s\n",fileName);
}
T_FileStream_close(stream);
}
/* test writing function */
stream=NULL;
uprv_strcpy(fileName,testdatapath);
uprv_strcat(fileName,".tmp");
stream = T_FileStream_open(fileName,"w+");
if(stream == NULL){
log_data_err("Could not open %s for writing\n",fileName);
} else {
c= '$';
T_FileStream_putc(stream,c);
T_FileStream_rewind(stream);
if(c != T_FileStream_getc(stream)){
log_data_err("T_FileStream_putc failed %s\n",fileName);
}
T_FileStream_rewind(stream);
T_FileStream_writeLine(stream,testline);
T_FileStream_rewind(stream);
T_FileStream_readLine(stream,buf,bufLen);
if(uprv_strncmp(testline, buf,uprv_strlen(buf))!=0){
log_data_err("T_FileStream_writeLine failed %s\n",fileName);
}
T_FileStream_rewind(stream);
T_FileStream_write(stream,testline,(int32_t)strlen(testline));
T_FileStream_rewind(stream);
retLen = T_FileStream_read(stream, buf, bufLen);
if(uprv_strncmp(testline, buf,retLen)!=0){
log_data_err("T_FileStream_write failed %s\n",fileName);
}
T_FileStream_close(stream);
}
if(!T_FileStream_remove(fileName)){
log_data_err("T_FileStream_remove failed to delete %s\n",fileName);
}
free(fileName);
free(buf);
}
//-------------------------------------------------------------------------------
//
// Read a text data file, convert it from UTF-8 to UChars, and return the data
// in one big UChar * buffer, which the caller must delete.
//
// (Lightly modified version of a similar function in regextst.cpp)
//
//--------------------------------------------------------------------------------
UChar *DecimalFormatTest::ReadAndConvertFile(const char *fileName, int32_t &ulen,
UErrorCode &status) {
UChar *retPtr = NULL;
char *fileBuf = NULL;
const char *fileBufNoBOM = NULL;
FILE *f = NULL;
ulen = 0;
if (U_FAILURE(status)) {
return retPtr;
}
//
// Open the file.
//
f = fopen(fileName, "rb");
if (f == 0) {
dataerrln("Error opening test data file %s\n", fileName);
status = U_FILE_ACCESS_ERROR;
return NULL;
}
//
// Read it in
//
int32_t fileSize;
int32_t amtRead;
int32_t amtReadNoBOM;
fseek( f, 0, SEEK_END);
fileSize = ftell(f);
fileBuf = new char[fileSize];
fseek(f, 0, SEEK_SET);
amtRead = fread(fileBuf, 1, fileSize, f);
if (amtRead != fileSize || fileSize <= 0) {
errln("Error reading test data file.");
goto cleanUpAndReturn;
}
//
// Look for a UTF-8 BOM on the data just read.
// The test data file is UTF-8.
// The BOM needs to be there in the source file to keep the Windows &
// EBCDIC machines happy, so force an error if it goes missing.
// Many Linux editors will silently strip it.
//
fileBufNoBOM = fileBuf + 3;
amtReadNoBOM = amtRead - 3;
if (fileSize<3 || uprv_strncmp(fileBuf, "\xEF\xBB\xBF", 3) != 0) {
// TODO: restore this check.
errln("Test data file %s is missing its BOM", fileName);
fileBufNoBOM = fileBuf;
amtReadNoBOM = amtRead;
}
//
// Find the length of the input in UTF-16 UChars
// (by preflighting the conversion)
//
u_strFromUTF8(NULL, 0, &ulen, fileBufNoBOM, amtReadNoBOM, &status);
//
// Convert file contents from UTF-8 to UTF-16
//
if (status == U_BUFFER_OVERFLOW_ERROR) {
// Buffer Overflow is expected from the preflight operation.
status = U_ZERO_ERROR;
retPtr = new UChar[ulen+1];
u_strFromUTF8(retPtr, ulen+1, NULL, fileBufNoBOM, amtReadNoBOM, &status);
}
cleanUpAndReturn:
fclose(f);
delete[] fileBuf;
if (U_FAILURE(status)) {
errln("ICU Error \"%s\"\n", u_errorName(status));
delete retPtr;
retPtr = NULL;
};
return retPtr;
}
/*
* state table row grammar (ebnf-style):
* (whitespace is allowed between all tokens)
*
* row=[[firstentry ','] entry (',' entry)*]
* firstentry="initial" | "surrogates"
* (initial state (default for state 0), output is all surrogate pairs)
* entry=range [':' nextstate] ['.' action]
* range=number ['-' number]
* nextstate=number
* (0..7f)
* action='u' | 's' | 'p' | 'i'
* (unassigned, state change only, surrogate pair, illegal)
* number=(1- or 2-digit hexadecimal number)
*/
static const char *
parseState(const char *s, int32_t state[256], uint32_t *pFlags) {
const char *t;
uint32_t start, end, i;
int32_t entry;
/* initialize the state: all illegal with U+ffff */
for(i=0; i<256; ++i) {
state[i]=MBCS_ENTRY_FINAL(0, MBCS_STATE_ILLEGAL, 0xffff);
}
/* skip leading white space */
s=u_skipWhitespace(s);
/* is there an "initial" or "surrogates" directive? */
if(uprv_strncmp("initial", s, 7)==0) {
*pFlags=MBCS_STATE_FLAG_DIRECT;
s=u_skipWhitespace(s+7);
if(*s++!=',') {
return s-1;
}
} else if(*pFlags==0 && uprv_strncmp("surrogates", s, 10)==0) {
*pFlags=MBCS_STATE_FLAG_SURROGATES;
s=u_skipWhitespace(s+10);
if(*s++!=',') {
return s-1;
}
} else if(*s==0) {
/* empty state row: all-illegal */
return NULL;
}
for(;;) {
/* read an entry, the start of the range first */
s=u_skipWhitespace(s);
start=uprv_strtoul(s, (char **)&t, 16);
if(s==t || 0xff<start) {
return s;
}
s=u_skipWhitespace(t);
/* read the end of the range if there is one */
if(*s=='-') {
s=u_skipWhitespace(s+1);
end=uprv_strtoul(s, (char **)&t, 16);
if(s==t || end<start || 0xff<end) {
return s;
}
s=u_skipWhitespace(t);
} else {
end=start;
}
/* determine the state entrys for this range */
if(*s!=':' && *s!='.') {
/* the default is: final state with valid entries */
entry=MBCS_ENTRY_FINAL(0, MBCS_STATE_VALID_16, 0);
} else {
entry=MBCS_ENTRY_TRANSITION(0, 0);
if(*s==':') {
/* get the next state, default to 0 */
s=u_skipWhitespace(s+1);
i=uprv_strtoul(s, (char **)&t, 16);
if(s!=t) {
if(0x7f<i) {
return s;
}
s=u_skipWhitespace(t);
entry=MBCS_ENTRY_SET_STATE(entry, i);
}
}
/* get the state action, default to valid */
if(*s=='.') {
/* this is a final state */
entry=MBCS_ENTRY_SET_FINAL(entry);
s=u_skipWhitespace(s+1);
if(*s=='u') {
/* unassigned set U+fffe */
entry=MBCS_ENTRY_FINAL_SET_ACTION_VALUE(entry, MBCS_STATE_UNASSIGNED, 0xfffe);
s=u_skipWhitespace(s+1);
} else if(*s=='p') {
if(*pFlags!=MBCS_STATE_FLAG_DIRECT) {
entry=MBCS_ENTRY_FINAL_SET_ACTION(entry, MBCS_STATE_VALID_16_PAIR);
} else {
entry=MBCS_ENTRY_FINAL_SET_ACTION(entry, MBCS_STATE_VALID_16);
}
s=u_skipWhitespace(s+1);
} else if(*s=='s') {
entry=MBCS_ENTRY_FINAL_SET_ACTION(entry, MBCS_STATE_CHANGE_ONLY);
s=u_skipWhitespace(s+1);
} else if(*s=='i') {
/* illegal set U+ffff */
entry=MBCS_ENTRY_FINAL_SET_ACTION_VALUE(entry, MBCS_STATE_ILLEGAL, 0xffff);
s=u_skipWhitespace(s+1);
} else {
/* default to valid */
entry=MBCS_ENTRY_FINAL_SET_ACTION(entry, MBCS_STATE_VALID_16);
}
} else {
/* this is an intermediate state, nothing to do */
}
}
/* adjust "final valid" states according to the state flags */
if(MBCS_ENTRY_FINAL_ACTION(entry)==MBCS_STATE_VALID_16) {
switch(*pFlags) {
case 0:
/* no adjustment */
break;
case MBCS_STATE_FLAG_DIRECT:
/* set the valid-direct code point to "unassigned"==0xfffe */
entry=MBCS_ENTRY_FINAL_SET_ACTION_VALUE(entry, MBCS_STATE_VALID_DIRECT_16, 0xfffe);
break;
case MBCS_STATE_FLAG_SURROGATES:
entry=MBCS_ENTRY_FINAL_SET_ACTION_VALUE(entry, MBCS_STATE_VALID_16_PAIR, 0);
break;
default:
break;
}
}
/* set this entry for the range */
for(i=start; i<=end; ++i) {
state[i]=entry;
}
if(*s==',') {
++s;
} else {
return *s==0 ? NULL : s;
}
}
}
U_CAPI const char* U_EXPORT2
uprv_tzname(int n)
{
#ifdef U_WINDOWS
const char *id = uprv_detectWindowsTimeZone();
if (id != NULL) {
return id;
}
#else
const char *tzenv = NULL;
/*#if defined(U_DARWIN)
int ret;
tzenv = getenv("TZFILE");
if (tzenv != NULL) {
return tzenv;
}
#endif*/
tzenv = getenv("TZ");
if (tzenv != NULL && isValidOlsonID(tzenv))
{
/* This might be a good Olson ID. */
if (uprv_strncmp(tzenv, "posix/", 6) == 0
|| uprv_strncmp(tzenv, "right/", 6) == 0)
{
/* Remove the posix/ or right/ prefix. */
tzenv += 6;
}
return tzenv;
}
/* else U_TZNAME will give a better result. */
#if defined(CHECK_LOCALTIME_LINK)
/* Caller must handle threading issues */
if (gTimeZoneBufferPtr == NULL) {
/*
This is a trick to look at the name of the link to get the Olson ID
because the tzfile contents is underspecified.
This isn't guaranteed to work because it may not be a symlink.
*/
int32_t ret = (int32_t)readlink(TZZONELINK, gTimeZoneBuffer, sizeof(gTimeZoneBuffer));
if (0 < ret) {
int32_t tzZoneInfoLen = uprv_strlen(TZZONEINFO);
gTimeZoneBuffer[ret] = 0;
if (uprv_strncmp(gTimeZoneBuffer, TZZONEINFO, tzZoneInfoLen) == 0
&& isValidOlsonID(gTimeZoneBuffer + tzZoneInfoLen))
{
return (gTimeZoneBufferPtr = gTimeZoneBuffer + tzZoneInfoLen);
}
}
}
else {
return gTimeZoneBufferPtr;
}
#endif
#endif
#ifdef U_TZNAME
/*
U_TZNAME is usually a non-unique abbreviation,
which isn't normally usable.
*/
return U_TZNAME[n];
#else
return "";
#endif
}
/* NOTE: The caller should handle thread safety */
U_CAPI const char* U_EXPORT2
uprv_getDefaultLocaleID()
{
#if U_POSIX_LOCALE
/*
Note that: (a '!' means the ID is improper somehow)
LC_ALL ----> default_loc codepage
--------------------------------------------------------
ab.CD ab CD
ab@CD ab__CD -
[email protected] ab__CD EF
ab_CD.EF@GH ab_CD_GH EF
Some 'improper' ways to do the same as above:
! [email protected] ab_CD_GH EF
! [email protected] ab_CD_GH EF
! [email protected]@GH.IJ ab_CD_GH EF
_CD@GH _CD_GH -
_CD.EF@GH _CD_GH EF
The variant cannot have dots in it.
The 'rightmost' variant (@xxx) wins.
The leftmost codepage (.xxx) wins.
*/
char *correctedPOSIXLocale = 0;
const char* posixID = uprv_getPOSIXID();
const char *p;
const char *q;
int32_t len;
/* Format: (no spaces)
ll [ _CC ] [ . MM ] [ @ VV]
l = lang, C = ctry, M = charmap, V = variant
*/
if (gCorrectedPOSIXLocale != NULL) {
return gCorrectedPOSIXLocale;
}
if ((p = uprv_strchr(posixID, '.')) != NULL) {
/* assume new locale can't be larger than old one? */
correctedPOSIXLocale = (char*) uprv_malloc(uprv_strlen(posixID)+1);
uprv_strncpy(correctedPOSIXLocale, posixID, p-posixID);
correctedPOSIXLocale[p-posixID] = 0;
/* do not copy after the @ */
if ((p = uprv_strchr(correctedPOSIXLocale, '@')) != NULL) {
correctedPOSIXLocale[p-correctedPOSIXLocale] = 0;
}
}
/* Note that we scan the *uncorrected* ID. */
if ((p = uprv_strrchr(posixID, '@')) != NULL) {
if (correctedPOSIXLocale == NULL) {
correctedPOSIXLocale = (char*) uprv_malloc(uprv_strlen(posixID)+1);
uprv_strncpy(correctedPOSIXLocale, posixID, p-posixID);
correctedPOSIXLocale[p-posixID] = 0;
}
p++;
/* Take care of any special cases here.. */
if (!uprv_strcmp(p, "nynorsk")) {
p = "NY";
/* Don't worry about no__NY. In practice, it won't appear. */
}
if (uprv_strchr(correctedPOSIXLocale,'_') == NULL) {
uprv_strcat(correctedPOSIXLocale, "__"); /* aa@b -> aa__b */
}
else {
uprv_strcat(correctedPOSIXLocale, "_"); /* aa_CC@b -> aa_CC_b */
}
if ((q = uprv_strchr(p, '.')) != NULL) {
/* How big will the resulting string be? */
len = (int32_t)(uprv_strlen(correctedPOSIXLocale) + (q-p));
uprv_strncat(correctedPOSIXLocale, p, q-p);
correctedPOSIXLocale[len] = 0;
}
else {
/* Anything following the @ sign */
uprv_strcat(correctedPOSIXLocale, p);
}
/* Should there be a map from 'no@nynorsk' -> no_NO_NY here?
* How about 'russian' -> 'ru'?
* Many of the other locales using ISO codes will be handled by the
* canonicalization functions in uloc_getDefault.
*/
}
/* Was a correction made? */
if (correctedPOSIXLocale != NULL) {
posixID = correctedPOSIXLocale;
}
else {
/* copy it, just in case the original pointer goes away. See j2395 */
correctedPOSIXLocale = (char *)uprv_malloc(uprv_strlen(posixID) + 1);
posixID = uprv_strcpy(correctedPOSIXLocale, posixID);
}
if (gCorrectedPOSIXLocale == NULL) {
gCorrectedPOSIXLocale = correctedPOSIXLocale;
ucln_common_registerCleanup(UCLN_COMMON_PUTIL, putil_cleanup);
correctedPOSIXLocale = NULL;
}
if (correctedPOSIXLocale != NULL) { /* Was already set - clean up. */
uprv_free(correctedPOSIXLocale);
}
return posixID;
#elif defined(U_WINDOWS)
UErrorCode status = U_ZERO_ERROR;
LCID id = GetThreadLocale();
const char* locID = uprv_convertToPosix(id, &status);
if (U_FAILURE(status)) {
locID = "en_US";
}
return locID;
#elif defined(XP_MAC)
int32_t script = MAC_LC_INIT_NUMBER;
/* = IntlScript(); or GetScriptManagerVariable(smSysScript);*/
int32_t region = MAC_LC_INIT_NUMBER;
/* = GetScriptManagerVariable(smRegionCode);*/
int32_t lang = MAC_LC_INIT_NUMBER;
/* = GetScriptManagerVariable(smScriptLang);*/
int32_t date_region = MAC_LC_INIT_NUMBER;
const char* posixID = 0;
int32_t count = sizeof(mac_lc_recs) / sizeof(mac_lc_rec);
int32_t i;
Intl1Hndl ih;
ih = (Intl1Hndl) GetIntlResource(1);
if (ih)
date_region = ((uint16_t)(*ih)->intl1Vers) >> 8;
for (i = 0; i < count; i++) {
if ( ((mac_lc_recs[i].script == MAC_LC_MAGIC_NUMBER)
|| (mac_lc_recs[i].script == script))
&& ((mac_lc_recs[i].region == MAC_LC_MAGIC_NUMBER)
|| (mac_lc_recs[i].region == region))
&& ((mac_lc_recs[i].lang == MAC_LC_MAGIC_NUMBER)
|| (mac_lc_recs[i].lang == lang))
&& ((mac_lc_recs[i].date_region == MAC_LC_MAGIC_NUMBER)
|| (mac_lc_recs[i].date_region == date_region))
)
{
posixID = mac_lc_recs[i].posixID;
break;
}
}
return posixID;
#elif defined(OS400)
/* locales are process scoped and are by definition thread safe */
static char correctedLocale[64];
const char *localeID = getenv("LC_ALL");
char *p;
if (localeID == NULL)
localeID = getenv("LANG");
if (localeID == NULL)
localeID = setlocale(LC_ALL, NULL);
/* Make sure we have something... */
if (localeID == NULL)
return "en_US_POSIX";
/* Extract the locale name from the path. */
if((p = uprv_strrchr(localeID, '/')) != NULL)
{
/* Increment p to start of locale name. */
p++;
localeID = p;
}
/* Copy to work location. */
uprv_strcpy(correctedLocale, localeID);
/* Strip off the '.locale' extension. */
if((p = uprv_strchr(correctedLocale, '.')) != NULL) {
*p = 0;
}
/* Upper case the locale name. */
T_CString_toUpperCase(correctedLocale);
/* See if we are using the POSIX locale. Any of the
* following are equivalent and use the same QLGPGCMA
* (POSIX) locale.
* QLGPGCMA2 means UCS2
* QLGPGCMA_4 means UTF-32
* QLGPGCMA_8 means UTF-8
*/
if ((uprv_strcmp("C", correctedLocale) == 0) ||
(uprv_strcmp("POSIX", correctedLocale) == 0) ||
(uprv_strncmp("QLGPGCMA", correctedLocale, 8) == 0))
{
uprv_strcpy(correctedLocale, "en_US_POSIX");
}
else
{
int16_t LocaleLen;
/* Lower case the lang portion. */
for(p = correctedLocale; *p != 0 && *p != '_'; p++)
{
*p = uprv_tolower(*p);
}
/* Adjust for Euro. After '_E' add 'URO'. */
LocaleLen = uprv_strlen(correctedLocale);
if (correctedLocale[LocaleLen - 2] == '_' &&
correctedLocale[LocaleLen - 1] == 'E')
{
uprv_strcat(correctedLocale, "URO");
}
/* If using Lotus-based locale then convert to
* equivalent non Lotus.
*/
else if (correctedLocale[LocaleLen - 2] == '_' &&
correctedLocale[LocaleLen - 1] == 'L')
{
correctedLocale[LocaleLen - 2] = 0;
}
/* There are separate simplified and traditional
* locales called zh_HK_S and zh_HK_T.
*/
else if (uprv_strncmp(correctedLocale, "zh_HK", 5) == 0)
{
uprv_strcpy(correctedLocale, "zh_HK");
}
/* A special zh_CN_GBK locale...
*/
else if (uprv_strcmp(correctedLocale, "zh_CN_GBK") == 0)
{
uprv_strcpy(correctedLocale, "zh_CN");
}
}
return correctedLocale;
#endif
}
static void TestGetKeywordValuesForLocale(void) {
#define PREFERRED_SIZE 15
#define MAX_NUMBER_OF_KEYWORDS 4
const char *PREFERRED[PREFERRED_SIZE][MAX_NUMBER_OF_KEYWORDS] = {
{ "root", "USD", "USN", NULL },
{ "und", "USD", "USN", NULL },
/* { "und_ZZ", "USD", NULL, NULL }, -- temporarily remove as this locale now has 15 entries */
{ "en_US", "USD", "USN", NULL },
{ "en_029", "USD", "USN", NULL },
{ "en_TH", "THB", NULL, NULL },
{ "de", "EUR", NULL, NULL },
{ "de_DE", "EUR", NULL, NULL },
{ "ar", "EGP", NULL, NULL },
{ "ar_PS", "ILS", "JOD", NULL },
{ "en@currency=CAD", "USD", "USN", NULL },
{ "fr@currency=zzz", "EUR", NULL, NULL },
{ "de_DE@currency=DEM", "EUR", NULL, NULL },
{ "en_US@rg=THZZZZ", "THB", NULL, NULL },
{ "de@rg=USZZZZ", "USD", "USN", NULL },
{ "en_US@currency=CAD;rg=THZZZZ", "THB", NULL, NULL },
};
const int32_t EXPECTED_SIZE[PREFERRED_SIZE] = {
2, 2, 2, 2, 1, 1, 1, 1, 2, 2, 1, 1, 1, 2, 1
};
/* ucurr_forLocale results for same locales; "" if no result expected */
const char *FORLOCALE[PREFERRED_SIZE] = {
"", "", "USD", "",
"THB", "", "EUR", "",
"ILS", "CAD", "ZZZ", "DEM",
"THB", "USD", "CAD"
};
UErrorCode status = U_ZERO_ERROR;
int32_t i, j, size;
UEnumeration *pref, *all;
const char *loc = NULL;
UBool matchPref, matchAll;
const char *value = NULL;
int32_t valueLength = 0;
UList *ALLList = NULL;
UEnumeration *ALL = ucurr_getKeywordValuesForLocale("currency", uloc_getDefault(), FALSE, &status);
if (ALL == NULL) {
log_err_status(status, "ERROR getting keyword value for default locale. -> %s\n", u_errorName(status));
return;
}
for (i = 0; i < PREFERRED_SIZE; i++) {
UChar getCurrU[4];
int32_t getCurrLen;
status = U_ZERO_ERROR;
pref = NULL;
all = NULL;
loc = PREFERRED[i][0];
pref = ucurr_getKeywordValuesForLocale("currency", loc, TRUE, &status);
matchPref = FALSE;
matchAll = FALSE;
size = uenum_count(pref, &status);
if (size == EXPECTED_SIZE[i]) {
matchPref = TRUE;
for (j = 0; j < size; j++) {
if ((value = uenum_next(pref, &valueLength, &status)) != NULL && U_SUCCESS(status)) {
if (uprv_strcmp(value, PREFERRED[i][j+1]) != 0) {
log_err("ERROR: locale %s got keywords #%d %s expected %s\n", loc, j, value, PREFERRED[i][j+1]);
matchPref = FALSE;
break;
}
} else {
matchPref = FALSE;
log_err("ERROR getting keyword value for locale \"%s\"\n", loc);
break;
}
}
} else {
log_err("FAIL: size of locale \"%s\" %d does not match expected size %d\n", loc, size, EXPECTED_SIZE[i]);
}
if (!matchPref) {
log_err("FAIL: Preferred values for locale \"%s\" does not match expected.\n", loc);
break;
}
uenum_close(pref);
all = ucurr_getKeywordValuesForLocale("currency", loc, FALSE, &status);
size = uenum_count(all, &status);
if (U_SUCCESS(status) && size == uenum_count(ALL, &status)) {
matchAll = TRUE;
ALLList = ulist_getListFromEnum(ALL);
for (j = 0; j < size; j++) {
if ((value = uenum_next(all, &valueLength, &status)) != NULL && U_SUCCESS(status)) {
if (!ulist_containsString(ALLList, value, uprv_strlen(value))) {
log_err("Locale %s have %s not in ALL\n", loc, value);
matchAll = FALSE;
break;
}
} else {
matchAll = FALSE;
log_err("ERROR getting \"all\" keyword value for locale \"%s\"\n", loc);
break;
}
}
if (!matchAll) {
log_err("FAIL: All values for locale \"%s\" does not match expected.\n", loc);
}
} else {
if(U_FAILURE(status)) {
log_err("ERROR: %s\n", u_errorName(status));
} else if(size!=uenum_count(ALL, &status)) {
log_err("ERROR: got size of %d, wanted %d\n", size, uenum_count(ALL, &status));
}
}
uenum_close(all);
status = U_ZERO_ERROR;
getCurrLen = ucurr_forLocale(loc, getCurrU, 4, &status);
if(U_FAILURE(status)) {
if (FORLOCALE[i][0] != 0) {
log_err("ERROR: ucurr_forLocale %s, status %s\n", loc, u_errorName(status));
}
} else if (getCurrLen != 3) {
if (FORLOCALE[i][0] != 0 || getCurrLen != -1) {
log_err("ERROR: ucurr_forLocale %s, returned len %d\n", loc, getCurrLen);
}
} else {
char getCurrB[4];
u_UCharsToChars(getCurrU, getCurrB, 4);
if ( uprv_strncmp(getCurrB, FORLOCALE[i], 4) != 0 ) {
log_err("ERROR: ucurr_forLocale %s, expected %s, got %s\n", loc, FORLOCALE[i], getCurrB);
}
}
}
uenum_close(ALL);
}
static UDataMemory *
doOpenChoice(const char *path, const char *type, const char *name,
UDataMemoryIsAcceptable *isAcceptable, void *context,
UErrorCode *pErrorCode)
{
UDataMemory *retVal = NULL;
TinyString tocEntryName; /* entry name in tree format. ex: 'icudt28b/coll/ar.res' */
TinyString tocEntryPath; /* entry name in path format. ex: 'icudt28b\\coll\\ar.res' */
TinyString pkgName;
TinyString treeName;
#if (U_FILE_SEP_CHAR != U_FILE_ALT_SEP_CHAR) /* '/' vs '\' */
TinyString altSepPath;
#endif
const char *dataPath;
int32_t tocEntrySuffixIndex;
const char *tocEntryPathSuffix;
UErrorCode subErrorCode=U_ZERO_ERROR;
const char *treeChar;
UBool isICUData = FALSE;
/* Is this path ICU data? */
if(path == NULL ||
!strcmp(path, U_ICUDATA_ALIAS) || /* "ICUDATA" */
!uprv_strncmp(path, U_ICUDATA_NAME U_TREE_SEPARATOR_STRING, /* "icudt26e-" */
uprv_strlen(U_ICUDATA_NAME U_TREE_SEPARATOR_STRING)) ||
!uprv_strncmp(path, U_ICUDATA_ALIAS U_TREE_SEPARATOR_STRING, /* "ICUDATA-" */
uprv_strlen(U_ICUDATA_ALIAS U_TREE_SEPARATOR_STRING))) {
isICUData = TRUE;
}
#if (U_FILE_SEP_CHAR != U_FILE_ALT_SEP_CHAR) /* Windows: try "foo\bar" and "foo/bar" */
/* remap from alternate path char to the main one */
TinyString_init(&altSepPath);
if(path) {
char *p;
if((p=uprv_strchr(path,U_FILE_ALT_SEP_CHAR))) {
TinyString_append(&altSepPath, path);
while((p=uprv_strchr(altSepPath.s,U_FILE_ALT_SEP_CHAR))) {
*p = U_FILE_SEP_CHAR;
}
#if defined (UDATA_DEBUG)
fprintf(stderr, "Changed path from [%s] to [%s]\n", path, altSepPath.s);
#endif
path = altSepPath.s;
}
}
#endif
TinyString_init(&tocEntryName);
TinyString_init(&tocEntryPath);
TinyString_init(&pkgName);
TinyString_init(&treeName);
/* ======= Set up strings */
if(path==NULL) {
TinyString_append(&pkgName, U_ICUDATA_NAME);
} else {
const char *pkg;
const char *first;
pkg = uprv_strrchr(path, U_FILE_SEP_CHAR);
first = uprv_strchr(path, U_FILE_SEP_CHAR);
if(uprv_pathIsAbsolute(path) || (pkg != first)) { /* more than one slash in the path- not a tree name */
/* see if this is an /absolute/path/to/package path */
if(pkg) {
TinyString_append(&pkgName, pkg+1);
} else {
TinyString_append(&pkgName, path);
}
} else {
treeChar = uprv_strchr(path, U_TREE_SEPARATOR);
if(treeChar) {
TinyString_append(&treeName, treeChar+1); /* following '-' */
if(isICUData) {
TinyString_append(&pkgName, U_ICUDATA_NAME);
} else {
TinyString_appendn(&pkgName, path, (int32_t)(treeChar-path));
if (first == NULL) {
/*
This user data has no path, but there is a tree name.
Look up the correct path from the data cache later.
*/
path = pkgName.s;
}
}
} else {
if(isICUData) {
TinyString_append(&pkgName, U_ICUDATA_NAME);
} else {
TinyString_append(&pkgName, path);
}
}
}
}
#ifdef UDATA_DEBUG
fprintf(stderr, " P=%s T=%s\n", pkgName.s, treeName.s);
#endif
/* setting up the entry name and file name
* Make up a full name by appending the type to the supplied
* name, assuming that a type was supplied.
*/
/* prepend the package */
TinyString_append(&tocEntryName, pkgName.s);
TinyString_append(&tocEntryPath, pkgName.s);
tocEntrySuffixIndex = tocEntryName.length;
if(treeName.s[0]) {
TinyString_append(&tocEntryName, U_TREE_ENTRY_SEP_STRING);
TinyString_append(&tocEntryName, treeName.s);
TinyString_append(&tocEntryPath, U_FILE_SEP_STRING);
TinyString_append(&tocEntryPath, treeName.s);
}
TinyString_append(&tocEntryName, U_TREE_ENTRY_SEP_STRING);
TinyString_append(&tocEntryPath, U_FILE_SEP_STRING);
TinyString_append(&tocEntryName, name);
TinyString_append(&tocEntryPath, name);
if(type!=NULL && *type!=0) {
TinyString_append(&tocEntryName, ".");
TinyString_append(&tocEntryName, type);
TinyString_append(&tocEntryPath, ".");
TinyString_append(&tocEntryPath, type);
}
tocEntryPathSuffix = tocEntryPath.s+tocEntrySuffixIndex; /* suffix starts here */
#ifdef UDATA_DEBUG
fprintf(stderr, " tocEntryName = %s\n", tocEntryName.s);
fprintf(stderr, " tocEntryPath = %s\n", tocEntryName.s);
#endif
if(path == NULL) {
path = COMMON_DATA_NAME; /* "icudt26e" */
}
/************************ Begin loop looking for ind. files ***************/
#ifdef UDATA_DEBUG
fprintf(stderr, "IND: inBasename = %s, pkg=%s\n", "(n/a)", packageNameFromPath(path));
#endif
/* End of dealing with a null basename */
dataPath = u_getDataDirectory();
/**** COMMON PACKAGE - only if packages are first. */
if(gDataFileAccess == UDATA_PACKAGES_FIRST) {
#ifdef UDATA_DEBUG
fprintf(stderr, "Trying packages (UDATA_PACKAGES_FIRST)\n");
#endif
/* #2 */
retVal = doLoadFromCommonData(isICUData,
pkgName.s, dataPath, tocEntryPathSuffix, tocEntryName.s,
path, type, name, isAcceptable, context, &subErrorCode, pErrorCode);
if((retVal != NULL) || U_FAILURE(*pErrorCode)) {
goto commonReturn;
}
}
/**** INDIVIDUAL FILES */
if((gDataFileAccess==UDATA_PACKAGES_FIRST) ||
(gDataFileAccess==UDATA_FILES_FIRST)) {
#ifdef UDATA_DEBUG
fprintf(stderr, "Trying individual files\n");
#endif
/* Check to make sure that there is a dataPath to iterate over */
if ((dataPath && *dataPath) || !isICUData) {
retVal = doLoadFromIndividualFiles(pkgName.s, dataPath, tocEntryPathSuffix,
path, type, name, isAcceptable, context, &subErrorCode, pErrorCode);
if((retVal != NULL) || U_FAILURE(*pErrorCode)) {
goto commonReturn;
}
}
}
/**** COMMON PACKAGE */
if((gDataFileAccess==UDATA_ONLY_PACKAGES) ||
(gDataFileAccess==UDATA_FILES_FIRST)) {
#ifdef UDATA_DEBUG
fprintf(stderr, "Trying packages (UDATA_ONLY_PACKAGES || UDATA_FILES_FIRST)\n");
#endif
retVal = doLoadFromCommonData(isICUData,
pkgName.s, dataPath, tocEntryPathSuffix, tocEntryName.s,
path, type, name, isAcceptable, context, &subErrorCode, pErrorCode);
if((retVal != NULL) || U_FAILURE(*pErrorCode)) {
goto commonReturn;
}
}
/* Load from DLL. If we haven't attempted package load, we also haven't had any chance to
try a DLL (static or setCommonData/etc) load.
If we ever have a "UDATA_ONLY_FILES", add it to the or list here. */
if(gDataFileAccess==UDATA_NO_FILES) {
#ifdef UDATA_DEBUG
fprintf(stderr, "Trying common data (UDATA_NO_FILES)\n");
#endif
retVal = doLoadFromCommonData(isICUData,
pkgName.s, "", tocEntryPathSuffix, tocEntryName.s,
path, type, name, isAcceptable, context, &subErrorCode, pErrorCode);
if((retVal != NULL) || U_FAILURE(*pErrorCode)) {
goto commonReturn;
}
}
/* data not found */
if(U_SUCCESS(*pErrorCode)) {
if(U_SUCCESS(subErrorCode)) {
/* file not found */
*pErrorCode=U_FILE_ACCESS_ERROR;
} else {
/* entry point not found or rejected */
*pErrorCode=subErrorCode;
}
}
commonReturn:
TinyString_dt(&tocEntryName);
TinyString_dt(&tocEntryPath);
TinyString_dt(&pkgName);
TinyString_dt(&treeName);
#if (U_FILE_SEP_CHAR != U_FILE_ALT_SEP_CHAR)
TinyString_dt(&altSepPath);
#endif
return retVal;
}
UHashtable*
ZoneMeta::createMetaToOlsonMap(void) {
UErrorCode status = U_ZERO_ERROR;
UHashtable *metaToOlson = NULL;
UResourceBundle *metazones = NULL;
UResourceBundle *mz = NULL;
metaToOlson = uhash_open(uhash_hashUChars, uhash_compareUChars, NULL, &status);
if (U_FAILURE(status)) {
return NULL;
}
uhash_setKeyDeleter(metaToOlson, deleteUCharString);
uhash_setValueDeleter(metaToOlson, deleteUVector);
metazones = ures_openDirect(NULL, gSupplementalData, &status);
metazones = ures_getByKey(metazones, gMapTimezonesTag, metazones, &status);
metazones = ures_getByKey(metazones, gMetazonesTag, metazones, &status);
if (U_FAILURE(status)) {
goto error_cleanup;
}
while (ures_hasNext(metazones)) {
mz = ures_getNextResource(metazones, mz, &status);
if (U_FAILURE(status)) {
status = U_ZERO_ERROR;
continue;
}
const char *mzkey = ures_getKey(mz);
if (uprv_strncmp(mzkey, gMetazoneIdPrefix, MZID_PREFIX_LEN) == 0) {
const char *mzid = mzkey + MZID_PREFIX_LEN;
const char *territory = uprv_strrchr(mzid, '_');
int32_t mzidLen = 0;
int32_t territoryLen = 0;
if (territory) {
mzidLen = territory - mzid;
territory++;
territoryLen = uprv_strlen(territory);
}
if (mzidLen > 0 && territoryLen > 0) {
int32_t tzidLen;
const UChar *tzid = ures_getStringByIndex(mz, 0, &tzidLen, &status);
if (U_SUCCESS(status)) {
// Create MetaToOlsonMappingEntry
MetaToOlsonMappingEntry *entry = (MetaToOlsonMappingEntry*)uprv_malloc(sizeof(MetaToOlsonMappingEntry));
if (entry == NULL) {
status = U_MEMORY_ALLOCATION_ERROR;
goto error_cleanup;
}
entry->id = tzid;
entry->territory = (UChar*)uprv_malloc((territoryLen + 1) * sizeof(UChar));
if (entry->territory == NULL) {
status = U_MEMORY_ALLOCATION_ERROR;
uprv_free(entry);
goto error_cleanup;
}
u_charsToUChars(territory, entry->territory, territoryLen + 1);
// Check if mapping entries for metazone is already available
if (mzidLen < ZID_KEY_MAX) {
UChar mzidUChars[ZID_KEY_MAX];
u_charsToUChars(mzid, mzidUChars, mzidLen);
mzidUChars[mzidLen++] = 0; // Add NUL terminator
UVector *tzMappings = (UVector*)uhash_get(metaToOlson, mzidUChars);
if (tzMappings == NULL) {
// Create new UVector and put it into the hashtable
tzMappings = new UVector(deleteMetaToOlsonMappingEntry, NULL, status);
if (U_FAILURE(status)) {
deleteMetaToOlsonMappingEntry(entry);
goto error_cleanup;
}
UChar *key = (UChar*)uprv_malloc(mzidLen * sizeof(UChar));
if (key == NULL) {
status = U_MEMORY_ALLOCATION_ERROR;
delete tzMappings;
deleteMetaToOlsonMappingEntry(entry);
goto error_cleanup;
}
u_strncpy(key, mzidUChars, mzidLen);
uhash_put(metaToOlson, key, tzMappings, &status);
if (U_FAILURE(status)) {
goto error_cleanup;
}
}
tzMappings->addElement(entry, status);
if (U_FAILURE(status)) {
goto error_cleanup;
}
} else {
deleteMetaToOlsonMappingEntry(entry);
}
} else {
status = U_ZERO_ERROR;
}
}
}
}
normal_cleanup:
ures_close(mz);
ures_close(metazones);
return metaToOlson;
error_cleanup:
if (metaToOlson != NULL) {
uhash_close(metaToOlson);
metaToOlson = NULL;
}
goto normal_cleanup;
}
static void U_CALLCONV
strprepProfileLineFn(void *context,
char *fields[][2], int32_t fieldCount,
UErrorCode *pErrorCode) {
uint32_t mapping[40];
char *end, *map;
uint32_t code;
int32_t length;
/*UBool* mapWithNorm = (UBool*) context;*/
const char* typeName;
uint32_t rangeStart=0,rangeEnd =0;
const char* filename = (const char*) context;
const char *s;
s = u_skipWhitespace(fields[0][0]);
if (*s == '@') {
/* special directive */
s++;
length = fields[0][1] - s;
if (length >= NORMALIZE_DIRECTIVE_LEN
&& uprv_strncmp(s, NORMALIZE_DIRECTIVE, NORMALIZE_DIRECTIVE_LEN) == 0) {
options[NORMALIZE].doesOccur = TRUE;
return;
}
else if (length >= CHECK_BIDI_DIRECTIVE_LEN
&& uprv_strncmp(s, CHECK_BIDI_DIRECTIVE, CHECK_BIDI_DIRECTIVE_LEN) == 0) {
options[CHECK_BIDI].doesOccur = TRUE;
return;
}
else {
fprintf(stderr, "gensprep error parsing a directive %s.", fields[0][0]);
}
}
typeName = fields[2][0];
map = fields[1][0];
if(uprv_strstr(typeName, usprepTypeNames[USPREP_UNASSIGNED])!=NULL){
u_parseCodePointRange(s, &rangeStart,&rangeEnd, pErrorCode);
if(U_FAILURE(*pErrorCode)){
fprintf(stderr, "Could not parse code point range. Error: %s\n",u_errorName(*pErrorCode));
return;
}
/* store the range */
storeRange(rangeStart,rangeEnd,USPREP_UNASSIGNED, pErrorCode);
}else if(uprv_strstr(typeName, usprepTypeNames[USPREP_PROHIBITED])!=NULL){
u_parseCodePointRange(s, &rangeStart,&rangeEnd, pErrorCode);
if(U_FAILURE(*pErrorCode)){
fprintf(stderr, "Could not parse code point range. Error: %s\n",u_errorName(*pErrorCode));
return;
}
/* store the range */
storeRange(rangeStart,rangeEnd,USPREP_PROHIBITED, pErrorCode);
}else if(uprv_strstr(typeName, usprepTypeNames[USPREP_MAP])!=NULL){
/* get the character code, field 0 */
code=(uint32_t)uprv_strtoul(s, &end, 16);
if(end<=s || end!=fields[0][1]) {
fprintf(stderr, "gensprep: syntax error in field 0 at %s\n", fields[0][0]);
*pErrorCode=U_PARSE_ERROR;
exit(U_PARSE_ERROR);
}
/* parse the mapping string */
length=u_parseCodePoints(map, mapping, sizeof(mapping)/4, pErrorCode);
/* store the mapping */
storeMapping(code,mapping, length,USPREP_MAP, pErrorCode);
}else{
*pErrorCode = U_INVALID_FORMAT_ERROR;
}
if(U_FAILURE(*pErrorCode)) {
fprintf(stderr, "gensprep error parsing %s line %s at %s. Error: %s\n",filename,
fields[0][0],fields[2][0],u_errorName(*pErrorCode));
exit(*pErrorCode);
}
}
U_CAPI int32_t U_EXPORT2
ualoc_getAppleParent(const char* localeID,
char * parent,
int32_t parentCapacity,
UErrorCode* err)
{
UResourceBundle *rb;
int32_t len;
UErrorCode tempStatus;
char locbuf[ULOC_FULLNAME_CAPACITY+1];
char * foundDoubleUnderscore;
if (U_FAILURE(*err)) {
return 0;
}
if ( (parent==NULL)? parentCapacity!=0: parentCapacity<0 ) {
*err = U_ILLEGAL_ARGUMENT_ERROR;
return 0;
}
len = uloc_getBaseName(localeID, locbuf, ULOC_FULLNAME_CAPACITY, err); /* canonicalize and strip keywords */
if (U_FAILURE(*err)) {
return 0;
}
if (*err == U_STRING_NOT_TERMINATED_WARNING) {
locbuf[ULOC_FULLNAME_CAPACITY] = 0;
*err = U_ZERO_ERROR;
}
foundDoubleUnderscore = uprv_strstr(locbuf, "__"); /* __ comes from bad/missing subtag or variant */
if (foundDoubleUnderscore != NULL) {
*foundDoubleUnderscore = 0; /* terminate at the __ */
len = uprv_strlen(locbuf);
}
if (len >= 2 && uprv_strncmp(locbuf, "zh", 2) == 0) {
const char ** forceParentPtr = forceParent;
const char * testCurLoc;
while ( (testCurLoc = *forceParentPtr++) != NULL ) {
int cmp = uprv_strcmp(locbuf, testCurLoc);
if (cmp <= 0) {
if (cmp == 0) {
len = uprv_strlen(*forceParentPtr);
if (len < parentCapacity) {
uprv_strcpy(parent, *forceParentPtr);
} else {
*err = U_BUFFER_OVERFLOW_ERROR;
}
return len;
}
break;
}
forceParentPtr++;
}
}
tempStatus = U_ZERO_ERROR;
rb = ures_openDirect(NULL, locbuf, &tempStatus);
if (U_SUCCESS(tempStatus)) {
const char * actualLocale = ures_getLocaleByType(rb, ULOC_ACTUAL_LOCALE, &tempStatus);
if (U_SUCCESS(tempStatus) && uprv_strcmp(locbuf, actualLocale) != 0) {
// we have followed an alias
len = uprv_strlen(actualLocale);
if (len < parentCapacity) {
uprv_strcpy(parent, actualLocale);
} else {
*err = U_BUFFER_OVERFLOW_ERROR;
}
ures_close(rb);
return len;
}
tempStatus = U_ZERO_ERROR;
const UChar * parentUName = ures_getStringByKey(rb, "%%Parent", &len, &tempStatus);
if (U_SUCCESS(tempStatus) && tempStatus != U_USING_FALLBACK_WARNING) {
if (len < parentCapacity) {
u_UCharsToChars(parentUName, parent, len + 1);
} else {
*err = U_BUFFER_OVERFLOW_ERROR;
}
ures_close(rb);
return len;
}
ures_close(rb);
}
len = uloc_getParent(locbuf, parent, parentCapacity, err);
if (U_SUCCESS(*err) && len == 0) {
len = 4;
if (len < parentCapacity) {
uprv_strcpy(parent, "root");
} else {
*err = U_BUFFER_OVERFLOW_ERROR;
}
}
return len;
}
static void
parseConverterOptions(const char *inName,
char *cnvName,
char *locale,
uint32_t *pFlags,
UErrorCode *err)
{
char c;
int32_t len = 0;
/* copy the converter name itself to cnvName */
while((c=*inName)!=0 && c!=UCNV_OPTION_SEP_CHAR) {
if (++len>=UCNV_MAX_CONVERTER_NAME_LENGTH) {
*err = U_ILLEGAL_ARGUMENT_ERROR; /* bad name */
*cnvName=0;
return;
}
*cnvName++=c;
inName++;
}
*cnvName=0;
/* parse options. No more name copying should occur. */
while((c=*inName)!=0) {
if(c==UCNV_OPTION_SEP_CHAR) {
++inName;
}
/* inName is behind an option separator */
if(uprv_strncmp(inName, "locale=", 7)==0) {
/* do not modify locale itself in case we have multiple locale options */
char *dest=locale;
/* copy the locale option value */
inName+=7;
len=0;
while((c=*inName)!=0 && c!=UCNV_OPTION_SEP_CHAR) {
++inName;
if(++len>=ULOC_FULLNAME_CAPACITY) {
*err=U_ILLEGAL_ARGUMENT_ERROR; /* bad name */
*locale=0;
return;
}
*dest++=c;
}
*dest=0;
} else if(uprv_strncmp(inName, "version=", 8)==0) {
/* copy the version option value into bits 3..0 of *pFlags */
inName+=8;
c=*inName;
if(c==0) {
*pFlags&=~UCNV_OPTION_VERSION;
return;
} else if((uint8_t)(c-'0')<10) {
*pFlags=(*pFlags&~UCNV_OPTION_VERSION)|(uint32_t)(c-'0');
++inName;
}
} else if(uprv_strncmp(inName, "swaplfnl", 8)==0) {
inName+=8;
*pFlags|=UCNV_OPTION_SWAP_LFNL;
/* add processing for new options here with another } else if(uprv_strncmp(inName, "option-name=", XX)==0) { */
} else {
/* ignore any other options until we define some */
while(((c = *inName++) != 0) && (c != UCNV_OPTION_SEP_CHAR)) {
}
if(c==0) {
return;
}
}
}
}
