void AlphabeticIndex::addIndexExemplars(const Locale &locale, UErrorCode &status) {
LocalULocaleDataPointer uld(ulocdata_open(locale.getName(), &status));
if (U_FAILURE(status)) {
return;
}
UnicodeSet exemplars;
ulocdata_getExemplarSet(uld.getAlias(), exemplars.toUSet(), 0, ULOCDATA_ES_INDEX, &status);
if (U_SUCCESS(status)) {
initialLabels_->addAll(exemplars);
return;
}
status = U_ZERO_ERROR; // Clear out U_MISSING_RESOURCE_ERROR
// The locale data did not include explicit Index characters.
// Synthesize a set of them from the locale's standard exemplar characters.
ulocdata_getExemplarSet(uld.getAlias(), exemplars.toUSet(), 0, ULOCDATA_ES_STANDARD, &status);
if (U_FAILURE(status)) {
return;
}
// question: should we add auxiliary exemplars?
if (exemplars.containsSome(0x61, 0x7A) /* a-z */ || exemplars.size() == 0) {
exemplars.add(0x61, 0x7A);
}
if (exemplars.containsSome(0xAC00, 0xD7A3)) { // Hangul syllables
// cut down to small list
exemplars.remove(0xAC00, 0xD7A3).
add(0xAC00).add(0xB098).add(0xB2E4).add(0xB77C).
add(0xB9C8).add(0xBC14).add(0xC0AC).add(0xC544).
add(0xC790).add(0xCC28).add(0xCE74).add(0xD0C0).
add(0xD30C).add(0xD558);
}
if (exemplars.containsSome(0x1200, 0x137F)) { // Ethiopic block
// cut down to small list
// make use of the fact that Ethiopic is allocated in 8's, where
// the base is 0 mod 8.
UnicodeSet ethiopic(
UNICODE_STRING_SIMPLE("[[:Block=Ethiopic:]&[:Script=Ethiopic:]]"), status);
UnicodeSetIterator it(ethiopic);
while (it.next() && !it.isString()) {
if ((it.getCodepoint() & 0x7) != 0) {
exemplars.remove(it.getCodepoint());
}
}
}
// Upper-case any that aren't already so.
// (We only do this for synthesized index characters.)
UnicodeSetIterator it(exemplars);
UnicodeString upperC;
while (it.next()) {
const UnicodeString &exemplarC = it.getString();
upperC = exemplarC;
upperC.toUpper(locale);
initialLabels_->add(upperC);
}
}
std::string kiwix::ucFirst (const std::string &word) {
if (word.empty())
return "";
std::string ucFirstWord;
#ifdef __ANDROID__
ucFirstWord = word;
ucFirstWord[0] = toupper(ucFirstWord[0]);
#else
UnicodeString firstLetter = UnicodeString(word.substr(0, 1).c_str());
UnicodeString ucFirstLetter = firstLetter.toUpper();
ucFirstLetter.toUTF8String(ucFirstWord);
ucFirstWord += word.substr(1);
#endif
return ucFirstWord;
}
void
StringCaseTest::TestCaseConversion()
{
static const UChar uppercaseGreek[] =
{ 0x399, 0x395, 0x3a3, 0x3a5, 0x3a3, 0x20, 0x03a7, 0x3a1, 0x399, 0x3a3, 0x3a4,
0x39f, 0x3a3, 0 };
// "IESUS CHRISTOS"
static const UChar lowercaseGreek[] =
{ 0x3b9, 0x3b5, 0x3c3, 0x3c5, 0x3c2, 0x20, 0x03c7, 0x3c1, 0x3b9, 0x3c3, 0x3c4,
0x3bf, 0x3c2, 0 };
// "iesus christos"
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 };
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 };
UnicodeString expectedResult;
UnicodeString test3;
test3 += (UChar32)0x0130;
test3 += "STANBUL, NOT CONSTANTINOPLE!";
UnicodeString test4(test3);
test4.toLower(Locale(""));
expectedResult = UnicodeString("i\\u0307stanbul, not constantinople!", "").unescape();
if (test4 != expectedResult)
errln("1. toLower failed: expected \"" + expectedResult + "\", got \"" + test4 + "\".");
test4 = test3;
test4.toLower(Locale("tr", "TR"));
expectedResult = lowercaseTurkish;
if (test4 != expectedResult)
errln("2. toLower failed: expected \"" + expectedResult + "\", got \"" + test4 + "\".");
test3 = "topkap";
test3 += (UChar32)0x0131;
test3 += " palace, istanbul";
test4 = test3;
test4.toUpper(Locale(""));
expectedResult = "TOPKAPI PALACE, ISTANBUL";
if (test4 != expectedResult)
errln("toUpper failed: expected \"" + expectedResult + "\", got \"" + test4 + "\".");
test4 = test3;
test4.toUpper(Locale("tr", "TR"));
expectedResult = uppercaseTurkish;
if (test4 != expectedResult)
errln("toUpper failed: expected \"" + expectedResult + "\", got \"" + test4 + "\".");
test3 = CharsToUnicodeString("S\\u00FC\\u00DFmayrstra\\u00DFe");
test3.toUpper(Locale("de", "DE"));
expectedResult = CharsToUnicodeString("S\\u00DCSSMAYRSTRASSE");
if (test3 != expectedResult)
errln("toUpper failed: expected \"" + expectedResult + "\", got \"" + test3 + "\".");
test4.replace(0, test4.length(), uppercaseGreek);
test4.toLower(Locale("el", "GR"));
expectedResult = lowercaseGreek;
if (test4 != expectedResult)
errln("toLower failed: expected \"" + expectedResult + "\", got \"" + test4 + "\".");
test4.replace(0, test4.length(), lowercaseGreek);
test4.toUpper();
expectedResult = uppercaseGreek;
if (test4 != expectedResult)
errln("toUpper failed: expected \"" + expectedResult + "\", got \"" + test4 + "\".");
// more string case mapping tests with the new implementation
{
static const UChar
beforeLower[]= { 0x61, 0x42, 0x49, 0x3a3, 0xdf, 0x3a3, 0x2f, 0xd93f, 0xdfff },
lowerRoot[]= { 0x61, 0x62, 0x69, 0x3c3, 0xdf, 0x3c2, 0x2f, 0xd93f, 0xdfff },
lowerTurkish[]={ 0x61, 0x62, 0x131, 0x3c3, 0xdf, 0x3c2, 0x2f, 0xd93f, 0xdfff },
beforeUpper[]= { 0x61, 0x42, 0x69, 0x3c2, 0xdf, 0x3c3, 0x2f, 0xfb03, 0xfb03, 0xfb03, 0xd93f, 0xdfff },
upperRoot[]= { 0x41, 0x42, 0x49, 0x3a3, 0x53, 0x53, 0x3a3, 0x2f, 0x46, 0x46, 0x49, 0x46, 0x46, 0x49, 0x46, 0x46, 0x49, 0xd93f, 0xdfff },
upperTurkish[]={ 0x41, 0x42, 0x130, 0x3a3, 0x53, 0x53, 0x3a3, 0x2f, 0x46, 0x46, 0x49, 0x46, 0x46, 0x49, 0x46, 0x46, 0x49, 0xd93f, 0xdfff },
beforeMiniUpper[]= { 0xdf, 0x61 },
miniUpper[]= { 0x53, 0x53, 0x41 };
UnicodeString s;
/* lowercase with root locale */
s=UnicodeString(FALSE, beforeLower, (int32_t)(sizeof(beforeLower)/U_SIZEOF_UCHAR));
s.toLower("");
if( s.length()!=(sizeof(lowerRoot)/U_SIZEOF_UCHAR) ||
s!=UnicodeString(FALSE, lowerRoot, s.length())
) {
errln("error in toLower(root locale)=\"" + s + "\" expected \"" + UnicodeString(FALSE, lowerRoot, (int32_t)(sizeof(lowerRoot)/U_SIZEOF_UCHAR)) + "\"");
}
/* lowercase with turkish locale */
s=UnicodeString(FALSE, beforeLower, (int32_t)(sizeof(beforeLower)/U_SIZEOF_UCHAR));
s.setCharAt(0, beforeLower[0]).toLower(Locale("tr"));
if( s.length()!=(sizeof(lowerTurkish)/U_SIZEOF_UCHAR) ||
s!=UnicodeString(FALSE, lowerTurkish, s.length())
) {
errln("error in toLower(turkish locale)=\"" + s + "\" expected \"" + UnicodeString(FALSE, lowerTurkish, (int32_t)(sizeof(lowerTurkish)/U_SIZEOF_UCHAR)) + "\"");
}
/* uppercase with root locale */
s=UnicodeString(FALSE, beforeUpper, (int32_t)(sizeof(beforeUpper)/U_SIZEOF_UCHAR));
s.setCharAt(0, beforeUpper[0]).toUpper(Locale(""));
if( s.length()!=(sizeof(upperRoot)/U_SIZEOF_UCHAR) ||
s!=UnicodeString(FALSE, upperRoot, s.length())
) {
errln("error in toUpper(root locale)=\"" + s + "\" expected \"" + UnicodeString(FALSE, upperRoot, (int32_t)(sizeof(upperRoot)/U_SIZEOF_UCHAR)) + "\"");
}
/* uppercase with turkish locale */
s=UnicodeString(FALSE, beforeUpper, (int32_t)(sizeof(beforeUpper)/U_SIZEOF_UCHAR));
s.toUpper(Locale("tr"));
if( s.length()!=(sizeof(upperTurkish)/U_SIZEOF_UCHAR) ||
s!=UnicodeString(FALSE, upperTurkish, s.length())
) {
errln("error in toUpper(turkish locale)=\"" + s + "\" expected \"" + UnicodeString(FALSE, upperTurkish, (int32_t)(sizeof(upperTurkish)/U_SIZEOF_UCHAR)) + "\"");
}
/* uppercase a short string with root locale */
s=UnicodeString(FALSE, beforeMiniUpper, (int32_t)(sizeof(beforeMiniUpper)/U_SIZEOF_UCHAR));
s.setCharAt(0, beforeMiniUpper[0]).toUpper("");
if( s.length()!=(sizeof(miniUpper)/U_SIZEOF_UCHAR) ||
s!=UnicodeString(FALSE, miniUpper, s.length())
) {
errln("error in toUpper(root locale)=\"" + s + "\" expected \"" + UnicodeString(FALSE, miniUpper, (int32_t)(sizeof(miniUpper)/U_SIZEOF_UCHAR)) + "\"");
}
}
// test some supplementary characters (>= Unicode 3.1)
{
UnicodeString t;
UnicodeString
deseretInput=UnicodeString("\\U0001043C\\U00010414", "").unescape(),
deseretLower=UnicodeString("\\U0001043C\\U0001043C", "").unescape(),
deseretUpper=UnicodeString("\\U00010414\\U00010414", "").unescape();
(t=deseretInput).toLower();
if(t!=deseretLower) {
errln("error lowercasing Deseret (plane 1) characters");
}
(t=deseretInput).toUpper();
if(t!=deseretUpper) {
errln("error uppercasing Deseret (plane 1) characters");
}
}
// test some more cases that looked like problems
{
UnicodeString t;
UnicodeString
ljInput=UnicodeString("ab'cD \\uFB00i\\u0131I\\u0130 \\u01C7\\u01C8\\u01C9 \\U0001043C\\U00010414", "").unescape(),
ljLower=UnicodeString("ab'cd \\uFB00i\\u0131ii\\u0307 \\u01C9\\u01C9\\u01C9 \\U0001043C\\U0001043C", "").unescape(),
ljUpper=UnicodeString("AB'CD FFIII\\u0130 \\u01C7\\u01C7\\u01C7 \\U00010414\\U00010414", "").unescape();
(t=ljInput).toLower("en");
if(t!=ljLower) {
errln("error lowercasing LJ characters");
}
(t=ljInput).toUpper("en");
if(t!=ljUpper) {
errln("error uppercasing LJ characters");
}
}
#if !UCONFIG_NO_NORMALIZATION
// some context-sensitive casing depends on normalization data being present
// Unicode 3.1.1 SpecialCasing tests
{
UnicodeString t;
// sigmas preceded and/or followed by cased letters
UnicodeString
sigmas=UnicodeString("i\\u0307\\u03a3\\u0308j \\u0307\\u03a3\\u0308j i\\u00ad\\u03a3\\u0308 \\u0307\\u03a3\\u0308 ", "").unescape(),
sigmasLower=UnicodeString("i\\u0307\\u03c3\\u0308j \\u0307\\u03c3\\u0308j i\\u00ad\\u03c2\\u0308 \\u0307\\u03c3\\u0308 ", "").unescape(),
sigmasUpper=UnicodeString("I\\u0307\\u03a3\\u0308J \\u0307\\u03a3\\u0308J I\\u00ad\\u03a3\\u0308 \\u0307\\u03a3\\u0308 ", "").unescape();
(t=sigmas).toLower();
if(t!=sigmasLower) {
errln("error in sigmas.toLower()=\"" + t + "\" expected \"" + sigmasLower + "\"");
}
(t=sigmas).toUpper(Locale(""));
if(t!=sigmasUpper) {
errln("error in sigmas.toUpper()=\"" + t + "\" expected \"" + sigmasUpper + "\"");
}
// turkish & azerbaijani dotless i & dotted I
// remove dot above if there was a capital I before and there are no more accents above
UnicodeString
dots=UnicodeString("I \\u0130 I\\u0307 I\\u0327\\u0307 I\\u0301\\u0307 I\\u0327\\u0307\\u0301", "").unescape(),
dotsTurkish=UnicodeString("\\u0131 i i i\\u0327 \\u0131\\u0301\\u0307 i\\u0327\\u0301", "").unescape(),
dotsDefault=UnicodeString("i i\\u0307 i\\u0307 i\\u0327\\u0307 i\\u0301\\u0307 i\\u0327\\u0307\\u0301", "").unescape();
(t=dots).toLower("tr");
if(t!=dotsTurkish) {
errln("error in dots.toLower(tr)=\"" + t + "\" expected \"" + dotsTurkish + "\"");
}
(t=dots).toLower("de");
if(t!=dotsDefault) {
errln("error in dots.toLower(de)=\"" + t + "\" expected \"" + dotsDefault + "\"");
}
}
// more Unicode 3.1.1 tests
{
UnicodeString t;
// lithuanian dot above in uppercasing
UnicodeString
dots=UnicodeString("a\\u0307 \\u0307 i\\u0307 j\\u0327\\u0307 j\\u0301\\u0307", "").unescape(),
dotsLithuanian=UnicodeString("A\\u0307 \\u0307 I J\\u0327 J\\u0301\\u0307", "").unescape(),
dotsDefault=UnicodeString("A\\u0307 \\u0307 I\\u0307 J\\u0327\\u0307 J\\u0301\\u0307", "").unescape();
(t=dots).toUpper("lt");
if(t!=dotsLithuanian) {
errln("error in dots.toUpper(lt)=\"" + t + "\" expected \"" + dotsLithuanian + "\"");
}
(t=dots).toUpper("de");
if(t!=dotsDefault) {
errln("error in dots.toUpper(de)=\"" + t + "\" expected \"" + dotsDefault + "\"");
}
// lithuanian adds dot above to i in lowercasing if there are more above accents
UnicodeString
i=UnicodeString("I I\\u0301 J J\\u0301 \\u012e \\u012e\\u0301 \\u00cc\\u00cd\\u0128", "").unescape(),
iLithuanian=UnicodeString("i i\\u0307\\u0301 j j\\u0307\\u0301 \\u012f \\u012f\\u0307\\u0301 i\\u0307\\u0300i\\u0307\\u0301i\\u0307\\u0303", "").unescape(),
iDefault=UnicodeString("i i\\u0301 j j\\u0301 \\u012f \\u012f\\u0301 \\u00ec\\u00ed\\u0129", "").unescape();
(t=i).toLower("lt");
if(t!=iLithuanian) {
errln("error in i.toLower(lt)=\"" + t + "\" expected \"" + iLithuanian + "\"");
}
(t=i).toLower("de");
if(t!=iDefault) {
errln("error in i.toLower(de)=\"" + t + "\" expected \"" + iDefault + "\"");
}
}
#endif
// test case folding
{
UnicodeString
s=UnicodeString("A\\u00df\\u00b5\\ufb03\\U0001040c\\u0130\\u0131", "").unescape(),
f=UnicodeString("ass\\u03bcffi\\U00010434i\\u0307\\u0131", "").unescape(),
g=UnicodeString("ass\\u03bcffi\\U00010434i\\u0131", "").unescape(),
t;
(t=s).foldCase();
if(f!=t) {
errln("error in foldCase(\"" + s + "\", default)=\"" + t + "\" but expected \"" + f + "\"");
}
// alternate handling for dotted I/dotless i (U+0130, U+0131)
(t=s).foldCase(U_FOLD_CASE_EXCLUDE_SPECIAL_I);
if(g!=t) {
errln("error in foldCase(\"" + s + "\", U_FOLD_CASE_EXCLUDE_SPECIAL_I)=\"" + t + "\" but expected \"" + g + "\"");
}
}
}
void
StringCaseTest::TestCasingImpl(const UnicodeString &input,
const UnicodeString &output,
int32_t whichCase,
void *iter, const char *localeID, uint32_t options) {
// UnicodeString
UnicodeString result;
const char *name;
Locale locale(localeID);
result=input;
switch(whichCase) {
case TEST_LOWER:
name="toLower";
result.toLower(locale);
break;
case TEST_UPPER:
name="toUpper";
result.toUpper(locale);
break;
#if !UCONFIG_NO_BREAK_ITERATION
case TEST_TITLE:
name="toTitle";
result.toTitle((BreakIterator *)iter, locale, options);
break;
#endif
case TEST_FOLD:
name="foldCase";
result.foldCase(options);
break;
default:
name="";
break; // won't happen
}
if(result!=output) {
dataerrln("error: UnicodeString.%s() got a wrong result for a test case from casing.res", name);
}
#if !UCONFIG_NO_BREAK_ITERATION
if(whichCase==TEST_TITLE && options==0) {
result=input;
result.toTitle((BreakIterator *)iter, locale);
if(result!=output) {
dataerrln("error: UnicodeString.toTitle(options=0) got a wrong result for a test case from casing.res");
}
}
#endif
// UTF-8
char utf8In[100], utf8Out[100];
int32_t utf8InLength, utf8OutLength, resultLength;
UChar *buffer;
IcuTestErrorCode errorCode(*this, "TestCasingImpl");
LocalUCaseMapPointer csm(ucasemap_open(localeID, options, errorCode));
#if !UCONFIG_NO_BREAK_ITERATION
if(iter!=NULL) {
// Clone the break iterator so that the UCaseMap can safely adopt it.
UBreakIterator *clone=ubrk_safeClone((UBreakIterator *)iter, NULL, NULL, errorCode);
ucasemap_setBreakIterator(csm.getAlias(), clone, errorCode);
}
#endif
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.getAlias(),
utf8Out, (int32_t)sizeof(utf8Out),
utf8In, utf8InLength, errorCode);
break;
case TEST_UPPER:
name="ucasemap_utf8ToUpper";
utf8OutLength=ucasemap_utf8ToUpper(csm.getAlias(),
utf8Out, (int32_t)sizeof(utf8Out),
utf8In, utf8InLength, errorCode);
break;
#if !UCONFIG_NO_BREAK_ITERATION
case TEST_TITLE:
name="ucasemap_utf8ToTitle";
utf8OutLength=ucasemap_utf8ToTitle(csm.getAlias(),
utf8Out, (int32_t)sizeof(utf8Out),
utf8In, utf8InLength, errorCode);
break;
#endif
case TEST_FOLD:
name="ucasemap_utf8FoldCase";
utf8OutLength=ucasemap_utf8FoldCase(csm.getAlias(),
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(errorCode.isSuccess() ? resultLength : 0);
if(errorCode.isFailure()) {
errcheckln(errorCode, "error: %s() got an error for a test case from casing.res - %s", name, u_errorName(errorCode));
errorCode.reset();
} else if(result!=output) {
errln("error: %s() got a wrong result for a test case from casing.res", name);
errln("expected \"" + output + "\" got \"" + result + "\"" );
}
}
void
StringCaseTest::TestCasingImpl(const UnicodeString &input,
const UnicodeString &output,
int32_t whichCase,
void *iter, const char *localeID, uint32_t options) {
// UnicodeString
UnicodeString result;
const char *name;
Locale locale(localeID);
result=input;
switch(whichCase) {
case TEST_LOWER:
name="toLower";
result.toLower(locale);
break;
case TEST_UPPER:
name="toUpper";
result.toUpper(locale);
break;
#if !UCONFIG_NO_BREAK_ITERATION
case TEST_TITLE:
name="toTitle";
result.toTitle((BreakIterator *)iter, locale, options);
break;
#endif
case TEST_FOLD:
name="foldCase";
result.foldCase(options);
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);
}
#if !UCONFIG_NO_BREAK_ITERATION
if(whichCase==TEST_TITLE && options==0) {
result=input;
result.toTitle((BreakIterator *)iter, locale);
if(result!=output) {
errln("error: UnicodeString.toTitle(options=0) got a wrong result for a test case from casing.res");
}
}
#endif
// 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);
#if !UCONFIG_NO_BREAK_ITERATION
if(iter!=NULL) {
// Clone the break iterator so that the UCaseMap can safely adopt it.
int32_t size=1; // Not 0 because that only gives preflighting.
UBreakIterator *clone=ubrk_safeClone((UBreakIterator *)iter, NULL, &size, &errorCode);
ucasemap_setBreakIterator(csm, clone, &errorCode);
}
#endif
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;
#if !UCONFIG_NO_BREAK_ITERATION
case TEST_TITLE:
name="ucasemap_utf8ToTitle";
utf8OutLength=ucasemap_utf8ToTitle(csm,
utf8Out, (int32_t)sizeof(utf8Out),
utf8In, utf8InLength, &errorCode);
break;
#endif
case TEST_FOLD:
name="ucasemap_utf8FoldCase";
utf8OutLength=ucasemap_utf8FoldCase(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);
}
UBool
TimeZone::parseCustomID(const UnicodeString& id, int32_t& sign,
int32_t& hour, int32_t& min, int32_t& sec) {
static const int32_t kParseFailed = -99999;
NumberFormat* numberFormat = 0;
UnicodeString idUppercase = id;
idUppercase.toUpper();
if (id.length() > GMT_ID_LENGTH &&
idUppercase.startsWith(GMT_ID))
{
ParsePosition pos(GMT_ID_LENGTH);
sign = 1;
hour = 0;
min = 0;
sec = 0;
if (id[pos.getIndex()] == MINUS /*'-'*/) {
sign = -1;
} else if (id[pos.getIndex()] != PLUS /*'+'*/) {
return FALSE;
}
pos.setIndex(pos.getIndex() + 1);
UErrorCode success = U_ZERO_ERROR;
numberFormat = NumberFormat::createInstance(success);
if(U_FAILURE(success)){
return FALSE;
}
numberFormat->setParseIntegerOnly(TRUE);
// Look for either hh:mm, hhmm, or hh
int32_t start = pos.getIndex();
Formattable n(kParseFailed);
numberFormat->parse(id, n, pos);
if (pos.getIndex() == start) {
delete numberFormat;
return FALSE;
}
hour = n.getLong();
if (pos.getIndex() < id.length()) {
if (pos.getIndex() - start > 2
|| id[pos.getIndex()] != COLON) {
delete numberFormat;
return FALSE;
}
// hh:mm
pos.setIndex(pos.getIndex() + 1);
int32_t oldPos = pos.getIndex();
n.setLong(kParseFailed);
numberFormat->parse(id, n, pos);
if ((pos.getIndex() - oldPos) != 2) {
// must be 2 digits
delete numberFormat;
return FALSE;
}
min = n.getLong();
if (pos.getIndex() < id.length()) {
if (id[pos.getIndex()] != COLON) {
delete numberFormat;
return FALSE;
}
// [:ss]
pos.setIndex(pos.getIndex() + 1);
oldPos = pos.getIndex();
n.setLong(kParseFailed);
numberFormat->parse(id, n, pos);
if (pos.getIndex() != id.length()
|| (pos.getIndex() - oldPos) != 2) {
delete numberFormat;
return FALSE;
}
sec = n.getLong();
}
} else {
// Supported formats are below -
//
// HHmmss
// Hmmss
// HHmm
// Hmm
// HH
// H
int32_t length = pos.getIndex() - start;
if (length <= 0 || 6 < length) {
// invalid length
delete numberFormat;
return FALSE;
}
switch (length) {
case 1:
case 2:
// already set to hour
break;
case 3:
case 4:
min = hour % 100;
hour /= 100;
break;
case 5:
case 6:
sec = hour % 100;
min = (hour/100) % 100;
hour /= 10000;
break;
}
}
delete numberFormat;
if (hour > kMAX_CUSTOM_HOUR || min > kMAX_CUSTOM_MIN || sec > kMAX_CUSTOM_SEC) {
return FALSE;
}
return TRUE;
}
return FALSE;
}
void AlphabeticIndex::addIndexExemplars(const Locale &locale, UErrorCode &status) {
if (U_FAILURE(status)) { return; }
// Chinese index characters, which are specific to each of the several Chinese tailorings,
// take precedence over the single locale data exemplar set per language.
const char *language = locale.getLanguage();
if (uprv_strcmp(language, "zh") == 0 || uprv_strcmp(language, "ja") == 0 ||
uprv_strcmp(language, "ko") == 0) {
// TODO: This should be done regardless of the language, but it's expensive.
// We should add a Collator function (can be @internal)
// to enumerate just the contractions that start with a given code point or string.
if (addChineseIndexCharacters(status) || U_FAILURE(status)) {
return;
}
}
LocalULocaleDataPointer uld(ulocdata_open(locale.getName(), &status));
if (U_FAILURE(status)) {
return;
}
UnicodeSet exemplars;
ulocdata_getExemplarSet(uld.getAlias(), exemplars.toUSet(), 0, ULOCDATA_ES_INDEX, &status);
if (U_SUCCESS(status)) {
initialLabels_->addAll(exemplars);
return;
}
status = U_ZERO_ERROR; // Clear out U_MISSING_RESOURCE_ERROR
// The locale data did not include explicit Index characters.
// Synthesize a set of them from the locale's standard exemplar characters.
ulocdata_getExemplarSet(uld.getAlias(), exemplars.toUSet(), 0, ULOCDATA_ES_STANDARD, &status);
if (U_FAILURE(status)) {
return;
}
// question: should we add auxiliary exemplars?
if (exemplars.containsSome(0x61, 0x7A) /* a-z */ || exemplars.size() == 0) {
exemplars.add(0x61, 0x7A);
}
if (exemplars.containsSome(0xAC00, 0xD7A3)) { // Hangul syllables
// cut down to small list
exemplars.remove(0xAC00, 0xD7A3).
add(0xAC00).add(0xB098).add(0xB2E4).add(0xB77C).
add(0xB9C8).add(0xBC14).add(0xC0AC).add(0xC544).
add(0xC790).add(0xCC28).add(0xCE74).add(0xD0C0).
add(0xD30C).add(0xD558);
}
if (exemplars.containsSome(0x1200, 0x137F)) { // Ethiopic block
// cut down to small list
// make use of the fact that Ethiopic is allocated in 8's, where
// the base is 0 mod 8.
UnicodeSet ethiopic(
UNICODE_STRING_SIMPLE("[[:Block=Ethiopic:]&[:Script=Ethiopic:]]"), status);
UnicodeSetIterator it(ethiopic);
while (it.next() && !it.isString()) {
if ((it.getCodepoint() & 0x7) != 0) {
exemplars.remove(it.getCodepoint());
}
}
}
// Upper-case any that aren't already so.
// (We only do this for synthesized index characters.)
UnicodeSetIterator it(exemplars);
UnicodeString upperC;
while (it.next()) {
const UnicodeString &exemplarC = it.getString();
upperC = exemplarC;
upperC.toUpper(locale);
initialLabels_->add(upperC);
}
}
void AlphabeticIndex::getIndexExemplars(UnicodeSet &dest, const Locale &locale, UErrorCode &status) {
if (U_FAILURE(status)) {
return;
}
LocalULocaleDataPointer uld(ulocdata_open(locale.getName(), &status));
UnicodeSet exemplars;
ulocdata_getExemplarSet(uld.getAlias(), exemplars.toUSet(), 0, ULOCDATA_ES_INDEX, &status);
if (U_SUCCESS(status)) {
dest.addAll(exemplars);
return;
}
status = U_ZERO_ERROR; // Clear out U_MISSING_RESOURCE_ERROR
// Locale data did not include explicit Index characters.
// Synthesize a set of them from the locale's standard exemplar characters.
ulocdata_getExemplarSet(uld.getAlias(), exemplars.toUSet(), 0, ULOCDATA_ES_STANDARD, &status);
if (U_FAILURE(status)) {
return;
}
// Upper-case any that aren't already so.
// (We only do this for synthesized index characters.)
UnicodeSetIterator it(exemplars);
UnicodeString upperC;
UnicodeSet lowersToRemove;
UnicodeSet uppersToAdd;
while (it.next()) {
const UnicodeString &exemplarC = it.getString();
upperC = exemplarC;
upperC.toUpper(locale);
if (exemplarC != upperC) {
lowersToRemove.add(exemplarC);
uppersToAdd.add(upperC);
}
}
exemplars.removeAll(lowersToRemove);
exemplars.addAll(uppersToAdd);
// get the exemplars, and handle special cases
// question: should we add auxiliary exemplars?
if (exemplars.containsSome(*CORE_LATIN)) {
exemplars.addAll(*CORE_LATIN);
}
if (exemplars.containsSome(*HANGUL)) {
// cut down to small list
UnicodeSet BLOCK_HANGUL_SYLLABLES(UNICODE_STRING_SIMPLE("[:block=hangul_syllables:]"), status);
exemplars.removeAll(BLOCK_HANGUL_SYLLABLES);
exemplars.addAll(*HANGUL);
}
if (exemplars.containsSome(*ETHIOPIC)) {
// cut down to small list
// make use of the fact that Ethiopic is allocated in 8's, where
// the base is 0 mod 8.
UnicodeSetIterator it(*ETHIOPIC);
while (it.next() && !it.isString()) {
if ((it.getCodepoint() & 0x7) != 0) {
exemplars.remove(it.getCodepoint());
}
}
}
dest.addAll(exemplars);
}
void grid_renderer<T>::process(text_symbolizer const& sym,
Feature const& feature,
proj_transform const& prj_trans)
{
typedef coord_transform2<CoordTransform,geometry_type> path_type;
bool placement_found = false;
text_placement_info_ptr placement_options = sym.get_placement_options()->get_placement_info();
while (!placement_found && placement_options->next())
{
expression_ptr name_expr = sym.get_name();
if (!name_expr) return;
value_type result = boost::apply_visitor(evaluate<Feature,value_type>(feature),*name_expr);
UnicodeString text = result.to_unicode();
if ( sym.get_text_transform() == UPPERCASE)
{
text = text.toUpper();
}
else if ( sym.get_text_transform() == LOWERCASE)
{
text = text.toLower();
}
else if ( sym.get_text_transform() == CAPITALIZE)
{
text = text.toTitle(NULL);
}
if ( text.length() <= 0 ) continue;
color const& fill = sym.get_fill();
face_set_ptr faces;
if (sym.get_fontset().size() > 0)
{
faces = font_manager_.get_face_set(sym.get_fontset());
}
else
{
faces = font_manager_.get_face_set(sym.get_face_name());
}
stroker_ptr strk = font_manager_.get_stroker();
if (!(faces->size() > 0 && strk))
{
throw config_error("Unable to find specified font face '" + sym.get_face_name() + "'");
}
text_renderer<T> ren(pixmap_, faces, *strk);
ren.set_pixel_size(placement_options->text_size * (scale_factor_ * (1.0/pixmap_.get_resolution())));
ren.set_fill(fill);
ren.set_halo_fill(sym.get_halo_fill());
ren.set_halo_radius(sym.get_halo_radius() * scale_factor_);
ren.set_opacity(sym.get_text_opacity());
// /pixmap_.get_resolution() ?
box2d<double> dims(0,0,width_,height_);
placement_finder<label_collision_detector4> finder(detector_,dims);
string_info info(text);
faces->get_string_info(info);
unsigned num_geom = feature.num_geometries();
for (unsigned i=0; i<num_geom; ++i)
{
geometry_type const& geom = feature.get_geometry(i);
if (geom.num_points() == 0) continue; // don't bother with empty geometries
while (!placement_found && placement_options->next_position_only())
{
placement text_placement(info, sym, scale_factor_);
text_placement.avoid_edges = sym.get_avoid_edges();
if (sym.get_label_placement() == POINT_PLACEMENT ||
sym.get_label_placement() == INTERIOR_PLACEMENT)
{
double label_x, label_y, z=0.0;
if (sym.get_label_placement() == POINT_PLACEMENT)
geom.label_position(&label_x, &label_y);
else
geom.label_interior_position(&label_x, &label_y);
prj_trans.backward(label_x,label_y, z);
t_.forward(&label_x,&label_y);
double angle = 0.0;
expression_ptr angle_expr = sym.get_orientation();
if (angle_expr)
{
// apply rotation
value_type result = boost::apply_visitor(evaluate<Feature,value_type>(feature),*angle_expr);
angle = result.to_double();
}
finder.find_point_placement(text_placement, placement_options,
label_x, label_y,
angle, sym.get_line_spacing(),
sym.get_character_spacing());
finder.update_detector(text_placement);
}
else if ( geom.num_points() > 1 && sym.get_label_placement() == LINE_PLACEMENT)
{
path_type path(t_,geom,prj_trans);
finder.find_line_placements<path_type>(text_placement, placement_options, path);
}
if (!text_placement.placements.size()) continue;
placement_found = true;
for (unsigned int ii = 0; ii < text_placement.placements.size(); ++ii)
{
double x = text_placement.placements[ii].starting_x;
double y = text_placement.placements[ii].starting_y;
ren.prepare_glyphs(&text_placement.placements[ii]);
ren.render_id(feature.id(),x,y,2);
}
}
}
}
if (placement_found)
pixmap_.add_feature(feature);
}
/**
* Parse a custom time zone identifier and return a corresponding zone.
* @param id a string of the form GMT[+-]hh:mm, GMT[+-]hhmm, or
* GMT[+-]hh.
* @return a newly created SimpleTimeZone with the given offset and
* no Daylight Savings Time, or null if the id cannot be parsed.
*/
TimeZone*
TimeZone::createCustomTimeZone(const UnicodeString& id)
{
static const int32_t kParseFailed = -99999;
NumberFormat* numberFormat = 0;
UnicodeString idUppercase = id;
idUppercase.toUpper();
if (id.length() > GMT_ID_LENGTH &&
idUppercase.startsWith(GMT_ID))
{
ParsePosition pos(GMT_ID_LENGTH);
UBool negative = FALSE;
int32_t offset;
if (id[pos.getIndex()] == 0x002D /*'-'*/)
negative = TRUE;
else if (id[pos.getIndex()] != 0x002B /*'+'*/)
return 0;
pos.setIndex(pos.getIndex() + 1);
UErrorCode success = U_ZERO_ERROR;
numberFormat = NumberFormat::createInstance(success);
numberFormat->setParseIntegerOnly(TRUE);
// Look for either hh:mm, hhmm, or hh
int32_t start = pos.getIndex();
Formattable n(kParseFailed);
numberFormat->parse(id, n, pos);
if (pos.getIndex() == start) {
delete numberFormat;
return 0;
}
offset = n.getLong();
if (pos.getIndex() < id.length() &&
id[pos.getIndex()] == 0x003A /*':'*/)
{
// hh:mm
offset *= 60;
pos.setIndex(pos.getIndex() + 1);
int32_t oldPos = pos.getIndex();
n.setLong(kParseFailed);
numberFormat->parse(id, n, pos);
if (pos.getIndex() == oldPos) {
delete numberFormat;
return 0;
}
offset += n.getLong();
}
else
{
// hhmm or hh
// Be strict about interpreting something as hh; it must be
// an offset < 30, and it must be one or two digits. Thus
// 0010 is interpreted as 00:10, but 10 is interpreted as
// 10:00.
if (offset < 30 && (pos.getIndex() - start) <= 2)
offset *= 60; // hh, from 00 to 29; 30 is 00:30
else
offset = offset % 100 + offset / 100 * 60; // hhmm
}
if(negative)
offset = -offset;
delete numberFormat;
return new SimpleTimeZone(offset * 60000, CUSTOM_ID);
}
return 0;
}
void agg_renderer<T>::process(shield_symbolizer const& sym,
Feature const& feature,
proj_transform const& prj_trans)
{
typedef coord_transform2<CoordTransform,geometry_type> path_type;
text_placement_info_ptr placement_options = sym.get_placement_options()->get_placement_info();
placement_options->next();
placement_options->next_position_only();
UnicodeString text;
if( sym.get_no_text() )
text = UnicodeString( " " ); // TODO: fix->use 'space' as the text to render
else
{
expression_ptr name_expr = sym.get_name();
if (!name_expr) return;
value_type result = boost::apply_visitor(evaluate<Feature,value_type>(feature),*name_expr);
text = result.to_unicode();
}
if ( sym.get_text_transform() == UPPERCASE)
{
text = text.toUpper();
}
else if ( sym.get_text_transform() == LOWERCASE)
{
text = text.toLower();
}
else if ( sym.get_text_transform() == CAPITALIZE)
{
text = text.toTitle(NULL);
}
agg::trans_affine tr;
boost::array<double,6> const& m = sym.get_transform();
tr.load_from(&m[0]);
std::string filename = path_processor_type::evaluate( *sym.get_filename(), feature);
boost::optional<mapnik::marker_ptr> marker;
if ( !filename.empty() )
{
marker = marker_cache::instance()->find(filename, true);
}
else
{
marker.reset(boost::make_shared<mapnik::marker>());
}
if (text.length() > 0 && marker)
{
int w = (*marker)->width();
int h = (*marker)->height();
double px0 = - 0.5 * w;
double py0 = - 0.5 * h;
double px1 = 0.5 * w;
double py1 = 0.5 * h;
double px2 = px1;
double py2 = py0;
double px3 = px0;
double py3 = py1;
tr.transform(&px0,&py0);
tr.transform(&px1,&py1);
tr.transform(&px2,&py2);
tr.transform(&px3,&py3);
box2d<double> label_ext (px0, py0, px1, py1);
label_ext.expand_to_include(px2, py2);
label_ext.expand_to_include(px3, py3);
face_set_ptr faces;
if (sym.get_fontset().size() > 0)
{
faces = font_manager_.get_face_set(sym.get_fontset());
}
else
{
faces = font_manager_.get_face_set(sym.get_face_name());
}
stroker_ptr strk = font_manager_.get_stroker();
if (strk && faces->size() > 0)
{
text_renderer<T> ren(pixmap_, faces, *strk);
ren.set_pixel_size(sym.get_text_size() * scale_factor_);
ren.set_fill(sym.get_fill());
ren.set_halo_fill(sym.get_halo_fill());
ren.set_halo_radius(sym.get_halo_radius() * scale_factor_);
ren.set_opacity(sym.get_text_opacity());
placement_finder<label_collision_detector4> finder(detector_);
string_info info(text);
faces->get_string_info(info);
metawriter_with_properties writer = sym.get_metawriter();
for (unsigned i = 0; i < feature.num_geometries(); ++i)
{
geometry_type const& geom = feature.get_geometry(i);
if (geom.num_points() > 0 )
{
path_type path(t_,geom,prj_trans);
label_placement_enum how_placed = sym.get_label_placement();
if (how_placed == POINT_PLACEMENT || how_placed == VERTEX_PLACEMENT || how_placed == INTERIOR_PLACEMENT)
{
// for every vertex, try and place a shield/text
geom.rewind(0);
placement text_placement(info, sym, scale_factor_, w, h, false);
text_placement.avoid_edges = sym.get_avoid_edges();
text_placement.allow_overlap = sym.get_allow_overlap();
if (writer.first)
text_placement.collect_extents =true; // needed for inmem metawriter
position const& pos = sym.get_displacement();
position const& shield_pos = sym.get_shield_displacement();
for( unsigned jj = 0; jj < geom.num_points(); jj++ )
{
double label_x;
double label_y;
double z=0.0;
if( how_placed == VERTEX_PLACEMENT )
geom.vertex(&label_x,&label_y); // by vertex
else if( how_placed == INTERIOR_PLACEMENT )
geom.label_interior_position(&label_x,&label_y);
else
geom.label_position(&label_x, &label_y); // by middle of line or by point
prj_trans.backward(label_x,label_y, z);
t_.forward(&label_x,&label_y);
label_x += boost::get<0>(shield_pos);
label_y += boost::get<1>(shield_pos);
finder.find_point_placement( text_placement, placement_options,
label_x, label_y, 0.0,
sym.get_line_spacing(),
sym.get_character_spacing());
// check to see if image overlaps anything too, there is only ever 1 placement found for points and verticies
if( text_placement.placements.size() > 0)
{
double x = floor(text_placement.placements[0].starting_x);
double y = floor(text_placement.placements[0].starting_y);
int px;
int py;
if( !sym.get_unlock_image() )
{
// center image at text center position
// remove displacement from image label
double lx = x - boost::get<0>(pos);
double ly = y - boost::get<1>(pos);
px=int(floor(lx - (0.5 * w))) + 1;
py=int(floor(ly - (0.5 * h))) + 1;
label_ext.re_center(lx,ly);
}
else
{ // center image at reference location
px=int(floor(label_x - 0.5 * w));
py=int(floor(label_y - 0.5 * h));
label_ext.re_center(label_x,label_y);
}
if ( sym.get_allow_overlap() || detector_.has_placement(label_ext) )
{
render_marker(px,py,**marker,tr,sym.get_opacity());
box2d<double> dim = ren.prepare_glyphs(&text_placement.placements[0]);
ren.render(x,y);
detector_.insert(label_ext);
finder.update_detector(text_placement);
if (writer.first) {
writer.first->add_box(label_ext, feature, t_, writer.second);
writer.first->add_text(text_placement, faces, feature, t_, writer.second);
}
}
}
}
}
else if (geom.num_points() > 1 && how_placed == LINE_PLACEMENT)
{
placement text_placement(info, sym, scale_factor_, w, h, false);
position const& pos = sym.get_displacement();
text_placement.avoid_edges = sym.get_avoid_edges();
text_placement.additional_boxes.push_back(
box2d<double>(-0.5 * label_ext.width() - boost::get<0>(pos),
-0.5 * label_ext.height() - boost::get<1>(pos),
0.5 * label_ext.width() - boost::get<0>(pos),
0.5 * label_ext.height() - boost::get<1>(pos)));
finder.find_point_placements<path_type>(text_placement, placement_options, path);
for (unsigned int ii = 0; ii < text_placement.placements.size(); ++ ii)
{
double x = floor(text_placement.placements[ii].starting_x);
double y = floor(text_placement.placements[ii].starting_y);
double lx = x - boost::get<0>(pos);
double ly = y - boost::get<1>(pos);
int px=int(floor(lx - (0.5*w))) + 1;
int py=int(floor(ly - (0.5*h))) + 1;
label_ext.re_center(lx, ly);
render_marker(px,py,**marker,tr,sym.get_opacity());
box2d<double> dim = ren.prepare_glyphs(&text_placement.placements[ii]);
ren.render(x,y);
if (writer.first) writer.first->add_box(label_ext, feature, t_, writer.second);
}
finder.update_detector(text_placement);
if (writer.first) writer.first->add_text(text_placement, faces, feature, t_, writer.second);
}
}
}
}
}
}
