organicmaps/icu
13
Fork 0
mirror of https://github.com/unicode-org/icu.git synced 2025-04-06 22:15:31 +00:00
Code Issues 1 Projects Releases Wiki Activity Actions 1

ICU-22520 Update all users of ulocimp_get*() to ulocimp_getSubtags().

Browse source 
This simplifies the code by removing the need for finding the positions
of the subtags, all that logic is now in just one single place.
This commit is contained in:
Fredrik Roubert 2024-01-03 10:20:21 +09:00 • committed by Fredrik Roubert
parent dc70b5a056
commit 1b768edbdf
3 changed files with 55 additions and 124 deletions
Show stats Download patch file Download diff file Expand all files Collapse all files

37
icu4c/source/common/loclikely.cpp
View file

@ -156,52 +156,27 @@ parseTagString(
icu::CharString& region,
UErrorCode* err)
{
icu::CharString variant;
const char* position = localeID;
if (U_FAILURE(*err) || localeID == nullptr) {
goto error;
}
lang = ulocimp_getLanguage(position, &position, *err);
ulocimp_getSubtags(localeID, &lang, &script, &region, &variant, &position, *err);
/*
* Note that we explicit consider U_STRING_NOT_TERMINATED_WARNING
* to be an error, because it indicates the user-supplied tag is
* not well-formed.
*/
if(U_FAILURE(*err)) {
goto error;
}
/*
* If no language was present, use the empty string instead.
* Otherwise, move past any separator.
*/
if (!variant.isEmpty()) {
position -= 1 + variant.length();
}
if (_isIDSeparator(*position)) {
++position;
}
script = ulocimp_getScript(position, &position, *err);
if(U_FAILURE(*err)) {
goto error;
}
if (!script.isEmpty()) {
/*
* Move past any separator.
*/
if (_isIDSeparator(*position)) {
++position;
}
}
region = ulocimp_getCountry(position, &position, *err);
if(U_FAILURE(*err)) {
goto error;
}
if (region.isEmpty() && *position != 0 && *position != '@') {
/* back up over consumed trailing separator */
--position;

115
icu4c/source/common/uloc.cpp
View file

@ -1569,32 +1569,18 @@ uloc_openKeywords(const char* localeID,
tmpLocaleID=localeID;
}
/* Skip the language */
ulocimp_getLanguage(tmpLocaleID, &tmpLocaleID, *status);
ulocimp_getSubtags(
tmpLocaleID,
nullptr,
nullptr,
nullptr,
nullptr,
&tmpLocaleID,
*status);
if (U_FAILURE(*status)) {
return 0;
}
if(_isIDSeparator(*tmpLocaleID)) {
const char *scriptID;
/* Skip the script if available */
ulocimp_getScript(tmpLocaleID+1, &scriptID, *status);
if (U_FAILURE(*status)) {
return 0;
}
if(scriptID != tmpLocaleID+1) {
/* Found optional script */
tmpLocaleID = scriptID;
}
/* Skip the Country */
if (_isIDSeparator(*tmpLocaleID)) {
ulocimp_getCountry(tmpLocaleID+1, &tmpLocaleID, *status);
if (U_FAILURE(*status)) {
return 0;
}
}
}
/* keywords are located after '@' */
if((tmpLocaleID = locale_getKeywordsStart(tmpLocaleID)) != nullptr) {
CharString keywords;
@ -1634,7 +1620,7 @@ _canonicalize(const char* localeID,
return;
}
int32_t j, fieldCount=0, scriptSize=0, variantSize=0;
int32_t j, fieldCount=0;
CharString tempBuffer; // if localeID has a BCP47 extension, tmpLocaleID points to this
CharString localeIDWithHyphens; // if localeID has a BPC47 extension and have _, tmpLocaleID points to this
const char* origLocaleID;
@ -1671,57 +1657,41 @@ _canonicalize(const char* localeID,
origLocaleID=tmpLocaleID;
/* get all pieces, one after another, and separate with '_' */
CharString tag = ulocimp_getLanguage(tmpLocaleID, &tmpLocaleID, *err);
CharString tag;
CharString script;
CharString country;
CharString variant;
ulocimp_getSubtags(
tmpLocaleID,
&tag,
&script,
&country,
&variant,
&tmpLocaleID,
*err);
if (tag.length() == I_DEFAULT_LENGTH &&
uprv_strncmp(origLocaleID, i_default, I_DEFAULT_LENGTH) == 0) {
tag.clear();
tag.append(uloc_getDefault(), *err);
} else if(_isIDSeparator(*tmpLocaleID)) {
const char *scriptID;
++fieldCount;
tag.append('_', *err);
CharString script = ulocimp_getScript(tmpLocaleID+1, &scriptID, *err);
tag.append(script, *err);
scriptSize = script.length();
if(scriptSize > 0) {
/* Found optional script */
tmpLocaleID = scriptID;
} else {
if (!script.isEmpty()) {
++fieldCount;
if (_isIDSeparator(*tmpLocaleID)) {
/* If there is something else, then we add the _ */
tag.append('_', *err);
tag.append(script, *err);
}
if (!country.isEmpty()) {
++fieldCount;
tag.append('_', *err);
tag.append(country, *err);
}
if (!variant.isEmpty()) {
++fieldCount;
if (country.isEmpty()) {
tag.append('_', *err);
}
}
if (_isIDSeparator(*tmpLocaleID)) {
const char *cntryID;
CharString country = ulocimp_getCountry(tmpLocaleID+1, &cntryID, *err);
tag.append(country, *err);
if (!country.isEmpty()) {
/* Found optional country */
tmpLocaleID = cntryID;
}
if(_isIDSeparator(*tmpLocaleID)) {
/* If there is something else, then we add the _ if we found country before. */
if (!_isIDSeparator(*(tmpLocaleID+1))) {
++fieldCount;
tag.append('_', *err);
}
variantSize = -tag.length();
{
CharStringByteSink s(&tag);
_getVariant(tmpLocaleID+1, *tmpLocaleID, &s, nullptr, false);
}
variantSize += tag.length();
if (variantSize > 0) {
tmpLocaleID += variantSize + 1; /* skip '_' and variant */
}
}
tag.append('_', *err);
tag.append(variant, *err);
}
}
@ -1767,22 +1737,15 @@ _canonicalize(const char* localeID,
/* Handle @FOO variant if @ is present and not followed by = */
if (tmpLocaleID!=nullptr && keywordAssign==nullptr) {
/* Add missing '_' if needed */
if (fieldCount < 2 || (fieldCount < 3 && scriptSize > 0)) {
if (fieldCount < 2 || (fieldCount < 3 && !script.isEmpty())) {
do {
tag.append('_', *err);
++fieldCount;
} while(fieldCount<2);
}
int32_t posixVariantSize = -tag.length();
{
CharStringByteSink s(&tag);
_getVariant(tmpLocaleID+1, '@', &s, nullptr, (UBool)(variantSize > 0));
}
posixVariantSize += tag.length();
if (posixVariantSize > 0) {
variantSize += posixVariantSize;
}
CharStringByteSink s(&tag);
_getVariant(tmpLocaleID+1, '@', &s, nullptr, !variant.isEmpty());
}
/* Look up the ID in the canonicalization map */

27
icu4c/source/common/uresbund.cpp
View file

@ -209,17 +209,11 @@ static bool getParentLocaleID(char *name, const char *origName, UResOpenType ope
}
UErrorCode err = U_ZERO_ERROR;
const char* tempNamePtr = name;
CharString language = ulocimp_getLanguage(tempNamePtr, &tempNamePtr, err);
if (*tempNamePtr == '_') {
++tempNamePtr;
}
CharString script = ulocimp_getScript(tempNamePtr, &tempNamePtr, err);
if (*tempNamePtr == '_') {
++tempNamePtr;
}
CharString region = ulocimp_getCountry(tempNamePtr, &tempNamePtr, err);
CharString workingLocale;
CharString language;
CharString script;
CharString region;
ulocimp_getSubtags(name, &language, &script, &region, nullptr, nullptr, err);
if (U_FAILURE(err)) {
// hopefully this never happens...
return chopLocale(name);
@ -238,6 +232,8 @@ static bool getParentLocaleID(char *name, const char *origName, UResOpenType ope
}
}
CharString workingLocale;
// if it's not in the parent locale table, figure out the fallback script algorithmically
// (see CLDR-15265 for an explanation of the algorithm)
if (!script.isEmpty() && !region.isEmpty()) {
@ -254,12 +250,9 @@ static bool getParentLocaleID(char *name, const char *origName, UResOpenType ope
// - if yes, replace the region with the script from the original locale ID
// - if no, replace the region with the default script for that language and region
UErrorCode err = U_ZERO_ERROR;
tempNamePtr = origName;
CharString origNameLanguage = ulocimp_getLanguage(tempNamePtr, &tempNamePtr, err);
if (*tempNamePtr == '_') {
++tempNamePtr;
}
CharString origNameScript = ulocimp_getScript(origName, nullptr, err);
CharString origNameLanguage;
CharString origNameScript;
ulocimp_getSubtags(origName, &origNameLanguage, &origNameScript, nullptr, nullptr, nullptr, err);
if (!origNameScript.isEmpty()) {
workingLocale.append(language, err).append("_", err).append(origNameScript, err);
} else {