int32 ChangeStringLinguisticCase(CaseFlags caseFlags, const char16* sourceString, uint32 sourceLength, char16* destString, uint32 destLength, ApiError* pErrorOut)
{
int32_t resultStringLength = 0;
UErrorCode errorCode = U_ZERO_ERROR;
static_assert(sizeof(UChar) == sizeof(char16), "Unexpected char type from ICU, function might have to be updated");
if (caseFlags == CaseFlagsUpper)
{
resultStringLength = u_strToUpper((UChar*) destString, destLength,
(UChar*) sourceString, sourceLength, NULL, &errorCode);
}
else if (caseFlags == CaseFlagsLower)
{
resultStringLength = u_strToLower((UChar*) destString, destLength,
(UChar*) sourceString, sourceLength, NULL, &errorCode);
}
else
{
Assert(false);
}
if (U_FAILURE(errorCode) &&
!(destLength == 0 && errorCode == U_BUFFER_OVERFLOW_ERROR))
{
*pErrorOut = TranslateUErrorCode(errorCode);
return -1;
}
// Todo: check for resultStringLength > destLength
// Return insufficient buffer in that case
return resultStringLength;
}
/*
** Implementations of scalar functions for case mapping - upper() and
** lower(). Function upper() converts its input to upper-case (ABC).
** Function lower() converts to lower-case (abc).
**
** ICU provides two types of case mapping, "general" case mapping and
** "language specific". Refer to ICU documentation for the differences
** between the two.
**
** To utilise "general" case mapping, the upper() or lower() scalar
** functions are invoked with one argument:
**
** upper('ABC') -> 'abc'
** lower('abc') -> 'ABC'
**
** To access ICU "language specific" case mapping, upper() or lower()
** should be invoked with two arguments. The second argument is the name
** of the locale to use. Passing an empty string ("") or SQL NULL value
** as the second argument is the same as invoking the 1 argument version
** of upper() or lower().
**
** lower('I', 'en_us') -> 'i'
** lower('I', 'tr_tr') -> '\u131' (small dotless i)
**
** http://www.icu-project.org/userguide/posix.html#case_mappings
*/
static void icuCaseFunc16(sqlite3_context *p, int nArg, sqlite3_value **apArg){
const UChar *zInput; /* Pointer to input string */
UChar *zOutput = 0; /* Pointer to output buffer */
int nInput; /* Size of utf-16 input string in bytes */
int nOut; /* Size of output buffer in bytes */
int cnt;
int bToUpper; /* True for toupper(), false for tolower() */
UErrorCode status;
const char *zLocale = 0;
assert(nArg==1 || nArg==2);
bToUpper = (sqlite3_user_data(p)!=0);
if( nArg==2 ){
zLocale = (const char *)sqlite3_value_text(apArg[1]);
}
zInput = sqlite3_value_text16(apArg[0]);
if( !zInput ){
return;
}
nOut = nInput = sqlite3_value_bytes16(apArg[0]);
if( nOut==0 ){
sqlite3_result_text16(p, "", 0, SQLITE_STATIC);
return;
}
for(cnt=0; cnt<2; cnt++){
UChar *zNew = sqlite3_realloc(zOutput, nOut);
if( zNew==0 ){
sqlite3_free(zOutput);
sqlite3_result_error_nomem(p);
return;
}
zOutput = zNew;
status = U_ZERO_ERROR;
if( bToUpper ){
nOut = 2*u_strToUpper(zOutput,nOut/2,zInput,nInput/2,zLocale,&status);
}else{
nOut = 2*u_strToLower(zOutput,nOut/2,zInput,nInput/2,zLocale,&status);
}
if( U_SUCCESS(status) ){
sqlite3_result_text16(p, zOutput, nOut, xFree);
}else if( status==U_BUFFER_OVERFLOW_ERROR ){
assert( cnt==0 );
continue;
}else{
icuFunctionError(p, bToUpper ? "u_strToUpper" : "u_strToLower", status);
}
return;
}
assert( 0 ); /* Unreachable */
}
/// Convert the unicode string to lowercase. This function will return the
/// required buffer length as a result. If this length does not match the
/// 'DestinationCapacity' this function must be called again with a buffer of
/// the required length to get a lowercase version of the string.
int32_t
swift::_swift_stdlib_unicode_strToLower(uint16_t *Destination,
int32_t DestinationCapacity,
const uint16_t *Source,
int32_t SourceLength) {
UErrorCode ErrorCode = U_ZERO_ERROR;
#if defined(__CYGWIN__) || defined( _MSC_VER) || defined(__linux__)
uint32_t OutputLength = u_strToLower(reinterpret_cast<UChar *>(Destination),
DestinationCapacity,
reinterpret_cast<const UChar *>(Source),
SourceLength,
"", &ErrorCode);
#else
uint32_t OutputLength = u_strToLower(Destination, DestinationCapacity,
Source, SourceLength,
"", &ErrorCode);
#endif
if (U_FAILURE(ErrorCode) && ErrorCode != U_BUFFER_OVERFLOW_ERROR) {
swift::crash("u_strToLower: Unexpected error lowercasing unicode string.");
}
return OutputLength;
}
int helper_normalize_str(const char *src, char *dest, int dest_size)
{
int type = CTS_LANG_OTHERS;
int32_t size;
UErrorCode status = 0;
UChar tmp_result[CTS_SQL_MAX_LEN*2];
UChar result[CTS_SQL_MAX_LEN*2];
int i = 0;
int j = 0;
int str_len = strlen(src);
int char_len = 0;
for (i=0;i<str_len;i+=char_len) {
char char_src[10];
char_len = check_utf8(src[i]);
memcpy(char_src, &src[i], char_len);
char_src[char_len] = '\0';
u_strFromUTF8(tmp_result, array_sizeof(tmp_result), NULL, char_src, -1, &status);
h_retvm_if(U_FAILURE(status), CTS_ERR_ICU_FAILED,
"u_strFromUTF8() Failed(%s)", u_errorName(status));
u_strToLower(tmp_result, array_sizeof(tmp_result), tmp_result, -1, NULL, &status);
h_retvm_if(U_FAILURE(status), CTS_ERR_ICU_FAILED,
"u_strToLower() Failed(%s)", u_errorName(status));
size = unorm_normalize(tmp_result, -1, UNORM_NFD, 0,
(UChar *)result, array_sizeof(result), &status);
h_retvm_if(U_FAILURE(status), CTS_ERR_ICU_FAILED,
"unorm_normalize(%s) Failed(%s)", char_src, u_errorName(status));
if (0 == i)
type = helper_check_language(result);
helper_extra_normalize(result, size);
u_strToUTF8(&dest[j], dest_size-j, &size, result, -1, &status);
h_retvm_if(U_FAILURE(status), CTS_ERR_ICU_FAILED,
"u_strToUTF8() Failed(%s)", u_errorName(status));
j += size;
dest[j++] = 0x01;
}
dest[j]='\0';
HELPER_DBG("src(%s) is transformed(%s)", src, dest);
return type;
}
// lower {{{
static PyObject *
icu_lower(PyObject *self, PyObject *args) {
char *input, *ans, *buf3 = NULL;
const char *loc;
int32_t sz;
UChar *buf, *buf2;
PyObject *ret;
UErrorCode status = U_ZERO_ERROR;
if (!PyArg_ParseTuple(args, "ses", &loc, "UTF-8", &input)) return NULL;
sz = (int32_t)strlen(input);
buf = (UChar*)calloc(sz*4 + 1, sizeof(UChar));
buf2 = (UChar*)calloc(sz*8 + 1, sizeof(UChar));
if (buf == NULL || buf2 == NULL) return PyErr_NoMemory();
u_strFromUTF8(buf, sz*4, NULL, input, sz, &status);
u_strToLower(buf2, sz*8, buf, -1, loc, &status);
ans = input;
sz = u_strlen(buf2);
free(buf);
if (U_SUCCESS(status) && sz > 0) {
buf3 = (char*)calloc(sz*5+1, sizeof(char));
if (buf3 == NULL) return PyErr_NoMemory();
u_strToUTF8(buf3, sz*5, NULL, buf2, -1, &status);
if (U_SUCCESS(status)) ans = buf3;
}
ret = PyUnicode_DecodeUTF8(ans, strlen(ans), "replace");
if (ret == NULL) return PyErr_NoMemory();
free(buf2);
if (buf3 != NULL) free(buf3);
PyMem_Free(input);
return ret;
} // }}}
/*
** Implementations of scalar functions for case mapping - upper() and
** lower(). Function upper() converts its input to upper-case (ABC).
** Function lower() converts to lower-case (abc).
**
** ICU provides two types of case mapping, "general" case mapping and
** "language specific". Refer to ICU documentation for the differences
** between the two.
**
** To utilise "general" case mapping, the upper() or lower() scalar
** functions are invoked with one argument:
**
** upper('ABC') -> 'abc'
** lower('abc') -> 'ABC'
**
** To access ICU "language specific" case mapping, upper() or lower()
** should be invoked with two arguments. The second argument is the name
** of the locale to use. Passing an empty string ("") or SQL NULL value
** as the second argument is the same as invoking the 1 argument version
** of upper() or lower().
**
** lower('I', 'en_us') -> 'i'
** lower('I', 'tr_tr') -> 'ı' (small dotless i)
**
** http://www.icu-project.org/userguide/posix.html#case_mappings
*/
static void icuCaseFunc16(sqlite3_context *p, int nArg, sqlite3_value **apArg){
const UChar *zInput;
UChar *zOutput;
int nInput;
int nOutput;
UErrorCode status = U_ZERO_ERROR;
const char *zLocale = 0;
assert(nArg==1 || nArg==2);
if( nArg==2 ){
zLocale = (const char *)sqlite3_value_text(apArg[1]);
}
zInput = sqlite3_value_text16(apArg[0]);
if( !zInput ){
return;
}
nInput = sqlite3_value_bytes16(apArg[0]);
nOutput = nInput * 2 + 2;
zOutput = sqlite3_malloc(nOutput);
if( !zOutput ){
return;
}
if( sqlite3_user_data(p) ){
u_strToUpper(zOutput, nOutput/2, zInput, nInput/2, zLocale, &status);
}else{
u_strToLower(zOutput, nOutput/2, zInput, nInput/2, zLocale, &status);
}
if( !U_SUCCESS(status) ){
icuFunctionError(p, "u_strToLower()/u_strToUpper", status);
return;
}
sqlite3_result_text16(p, zOutput, -1, xFree);
}
static PyObject* icu_change_case(PyObject *self, PyObject *args) {
char *locale = NULL;
PyObject *input = NULL, *result = NULL;
int which = UPPER_CASE;
UErrorCode status = U_ZERO_ERROR;
UChar *input_buf = NULL, *output_buf = NULL;
int32_t sz = 0;
if (!PyArg_ParseTuple(args, "Oiz", &input, &which, &locale)) return NULL;
if (locale == NULL) {
PyErr_SetString(PyExc_NotImplementedError, "You must specify a locale"); // We deliberately use NotImplementedError so that this error can be unambiguously identified
return NULL;
}
input_buf = python_to_icu(input, &sz, 1);
if (input_buf == NULL) goto end;
output_buf = (UChar*) calloc(3 * sz, sizeof(UChar));
if (output_buf == NULL) { PyErr_NoMemory(); goto end; }
switch (which) {
case TITLE_CASE:
sz = u_strToTitle(output_buf, 3 * sz, input_buf, sz, NULL, locale, &status);
break;
case UPPER_CASE:
sz = u_strToUpper(output_buf, 3 * sz, input_buf, sz, locale, &status);
break;
default:
sz = u_strToLower(output_buf, 3 * sz, input_buf, sz, locale, &status);
}
if (U_FAILURE(status)) { PyErr_SetString(PyExc_ValueError, u_errorName(status)); goto end; }
result = icu_to_python(output_buf, sz);
end:
if (input_buf != NULL) free(input_buf);
if (output_buf != NULL) free(output_buf);
return result;
} // }}}
static uint32_t icu_utf16_sub(struct icu_buf_utf16 *dest16,
struct icu_buf_utf16 *src16,
const char *locale, char action,
UErrorCode *status)
{
switch (action)
{
case 'l':
case 'L':
return u_strToLower(dest16->utf16, dest16->utf16_cap,
src16->utf16, src16->utf16_len,
locale, status);
case 'u':
case 'U':
return u_strToUpper(dest16->utf16, dest16->utf16_cap,
src16->utf16, src16->utf16_len,
locale, status);
break;
case 't':
case 'T':
return u_strToTitle(dest16->utf16, dest16->utf16_cap,
src16->utf16, src16->utf16_len,
0, locale, status);
break;
case 'f':
case 'F':
return u_strFoldCase(dest16->utf16, dest16->utf16_cap,
src16->utf16, src16->utf16_len,
U_FOLD_CASE_DEFAULT, status);
break;
default:
*status = U_UNSUPPORTED_ERROR;
break;
}
return 0;
}
static void
TestCaseLower(void) {
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 };
UChar buffer[32];
int32_t length;
UErrorCode errorCode;
/* lowercase with root locale and separate buffers */
buffer[0]=0xabcd;
errorCode=U_ZERO_ERROR;
length=u_strToLower(buffer, sizeof(buffer)/U_SIZEOF_UCHAR,
beforeLower, sizeof(beforeLower)/U_SIZEOF_UCHAR,
"",
&errorCode);
if( U_FAILURE(errorCode) ||
length!=(sizeof(lowerRoot)/U_SIZEOF_UCHAR) ||
uprv_memcmp(lowerRoot, buffer, length*U_SIZEOF_UCHAR)!=0 ||
buffer[length]!=0
) {
log_err("error in u_strToLower(root locale)=%ld error=%s string matches: %s\t\nlowerRoot=%s\t\nbuffer=%s\n",
length,
u_errorName(errorCode),
uprv_memcmp(lowerRoot, buffer, length*U_SIZEOF_UCHAR)==0 &&
buffer[length]==0 ? "yes" : "no",
aescstrdup(lowerRoot,-1),
aescstrdup(buffer,-1));
}
/* lowercase with turkish locale and in the same buffer */
uprv_memcpy(buffer, beforeLower, sizeof(beforeLower));
buffer[sizeof(beforeLower)/U_SIZEOF_UCHAR]=0;
errorCode=U_ZERO_ERROR;
length=u_strToLower(buffer, sizeof(buffer)/U_SIZEOF_UCHAR,
buffer, -1, /* implicit srcLength */
"tr",
&errorCode);
if( U_FAILURE(errorCode) ||
length!=(sizeof(lowerTurkish)/U_SIZEOF_UCHAR) ||
uprv_memcmp(lowerTurkish, buffer, length*U_SIZEOF_UCHAR)!=0 ||
buffer[length]!=0
) {
log_err("error in u_strToLower(turkish locale)=%ld error=%s string matches: %s\n",
length,
u_errorName(errorCode),
uprv_memcmp(lowerTurkish, buffer, length*U_SIZEOF_UCHAR)==0 && buffer[length]==0 ? "yes" : "no");
}
/* test preflighting */
buffer[0]=buffer[2]=0xabcd;
errorCode=U_ZERO_ERROR;
length=u_strToLower(buffer, 2, /* set destCapacity=2 */
beforeLower, sizeof(beforeLower)/U_SIZEOF_UCHAR,
"",
&errorCode);
if( errorCode!=U_BUFFER_OVERFLOW_ERROR ||
length!=(sizeof(lowerRoot)/U_SIZEOF_UCHAR) ||
uprv_memcmp(lowerRoot, buffer, 2*U_SIZEOF_UCHAR)!=0 ||
buffer[2]!=0xabcd
) {
log_err("error in u_strToLower(root locale preflighting)=%ld error=%s string matches: %s\n",
length,
u_errorName(errorCode),
uprv_memcmp(lowerRoot, buffer, 2*U_SIZEOF_UCHAR)==0 && buffer[2]==0xabcd ? "yes" : "no");
}
/* test error handling */
errorCode=U_ZERO_ERROR;
length=u_strToLower(NULL, sizeof(buffer)/U_SIZEOF_UCHAR,
beforeLower, sizeof(beforeLower)/U_SIZEOF_UCHAR,
"",
&errorCode);
if(errorCode!=U_ILLEGAL_ARGUMENT_ERROR) {
log_err("error in u_strToLower(root locale dest=NULL)=%ld error=%s\n",
length,
u_errorName(errorCode));
}
buffer[0]=0xabcd;
errorCode=U_ZERO_ERROR;
length=u_strToLower(buffer, -1,
beforeLower, sizeof(beforeLower)/U_SIZEOF_UCHAR,
"",
&errorCode);
if( errorCode!=U_ILLEGAL_ARGUMENT_ERROR ||
buffer[0]!=0xabcd
) {
log_err("error in u_strToLower(root locale destCapacity=-1)=%ld error=%s buffer[0]==0x%lx\n",
length,
u_errorName(errorCode),
buffer[0]);
}
}
int32_t
u_sprintf_scientific_handler(u_localized_string *output,
const u_sprintf_spec_info *info,
const ufmt_args *args)
{
double num = (double) (args[0].doubleValue);
UNumberFormat *format;
UChar result [USPRINTF_BUFFER_SIZE];
int32_t minDecimalDigits;
int32_t maxDecimalDigits;
UErrorCode status = U_ZERO_ERROR;
UChar srcExpBuf[USPRINTF_SYMBOL_BUFFER_SIZE];
int32_t srcLen, expLen;
UChar expBuf[USPRINTF_SYMBOL_BUFFER_SIZE];
/* mask off any necessary bits */
/* if(! info->fIsLongDouble)
num &= DBL_MAX;*/
/* get the formatter */
format = u_locbund_getScientificFormat(output->fBundle);
/* handle error */
if(format == 0)
return 0;
/* set the appropriate flags on the formatter */
/* clone the stream's bundle if it isn't owned */
if(! output->fOwnBundle) {
output->fBundle = u_locbund_clone(output->fBundle);
output->fOwnBundle = TRUE;
format = u_locbund_getScientificFormat(output->fBundle);
}
srcLen = unum_getSymbol(format,
UNUM_EXPONENTIAL_SYMBOL,
srcExpBuf,
sizeof(srcExpBuf),
&status);
/* Upper/lower case the e */
if (info->fSpec == (UChar)0x65 /* e */) {
expLen = u_strToLower(expBuf, (int32_t)sizeof(expBuf),
srcExpBuf, srcLen,
output->fBundle->fLocale,
&status);
}
else {
expLen = u_strToUpper(expBuf, (int32_t)sizeof(expBuf),
srcExpBuf, srcLen,
output->fBundle->fLocale,
&status);
}
unum_setSymbol(format,
UNUM_EXPONENTIAL_SYMBOL,
expBuf,
expLen,
&status);
/* set the number of decimal digits */
/* save the formatter's state */
minDecimalDigits = unum_getAttribute(format, UNUM_MIN_FRACTION_DIGITS);
maxDecimalDigits = unum_getAttribute(format, UNUM_MAX_FRACTION_DIGITS);
if(info->fPrecision != -1) {
/* set the # of decimal digits */
unum_setAttribute(format, UNUM_FRACTION_DIGITS, info->fPrecision);
}
else if(info->fPrecision == 0 && ! info->fAlt) {
/* no decimal point in this case */
unum_setAttribute(format, UNUM_FRACTION_DIGITS, 0);
}
else if(info->fAlt) {
/* '#' means always show decimal point */
/* copy of printf behavior on Solaris - '#' shows 6 digits */
unum_setAttribute(format, UNUM_FRACTION_DIGITS, 6);
}
else {
/* # of decimal digits is 6 if precision not specified */
unum_setAttribute(format, UNUM_FRACTION_DIGITS, 6);
}
/* set whether to show the sign */
u_sprintf_set_sign(format, info, &status);
/* format the number */
unum_formatDouble(format, num, result, USPRINTF_BUFFER_SIZE, 0, &status);
/* restore the number format */
unum_setAttribute(format, UNUM_MIN_FRACTION_DIGITS, minDecimalDigits);
unum_setAttribute(format, UNUM_MAX_FRACTION_DIGITS, maxDecimalDigits);
/* Since we clone the fBundle and we're only using the scientific
format, we don't need to save the old exponent value. */
/*unum_setSymbol(format,
UNUM_EXPONENTIAL_SYMBOL,
srcExpBuf,
srcLen,
&status);*/
return u_sprintf_pad_and_justify(output, info, result, u_strlen(result));
}
static int32_t
u_scanf_scientific_handler(UFILE *input,
u_scanf_spec_info *info,
ufmt_args *args,
const UChar *fmt,
int32_t *fmtConsumed,
int32_t *argConverted)
{
int32_t len;
double num;
UNumberFormat *format;
int32_t parsePos = 0;
int32_t skipped;
UErrorCode status = U_ZERO_ERROR;
UChar srcExpBuf[UPRINTF_SYMBOL_BUFFER_SIZE];
int32_t srcLen, expLen;
UChar expBuf[UPRINTF_SYMBOL_BUFFER_SIZE];
/* skip all ws in the input */
skipped = u_scanf_skip_leading_ws(input, info->fPadChar);
/* fill the input's internal buffer */
ufile_fill_uchar_buffer(input);
/* determine the size of the input's buffer */
len = (int32_t)(input->str.fLimit - input->str.fPos);
/* truncate to the width, if specified */
if(info->fWidth != -1)
len = ufmt_min(len, info->fWidth);
/* get the formatter */
format = u_locbund_getNumberFormat(&input->str.fBundle, UNUM_SCIENTIFIC);
/* handle error */
if(format == 0)
return 0;
/* set the appropriate flags on the formatter */
srcLen = unum_getSymbol(format,
UNUM_EXPONENTIAL_SYMBOL,
srcExpBuf,
sizeof(srcExpBuf),
&status);
/* Upper/lower case the e */
if (info->fSpec == (UChar)0x65 /* e */) {
expLen = u_strToLower(expBuf, (int32_t)sizeof(expBuf),
srcExpBuf, srcLen,
input->str.fBundle.fLocale,
&status);
}
else {
expLen = u_strToUpper(expBuf, (int32_t)sizeof(expBuf),
srcExpBuf, srcLen,
input->str.fBundle.fLocale,
&status);
}
unum_setSymbol(format,
UNUM_EXPONENTIAL_SYMBOL,
expBuf,
expLen,
&status);
/* Skip the positive prefix. ICU normally can't handle this due to strict parsing. */
skipped += u_scanf_skip_leading_positive_sign(input, format, &status);
/* parse the number */
num = unum_parseDouble(format, input->str.fPos, len, &parsePos, &status);
if (!info->fSkipArg) {
if (info->fIsLong)
*(double*)(args[0].ptrValue) = num;
else if (info->fIsLongDouble)
*(long double*)(args[0].ptrValue) = num;
else
*(float*)(args[0].ptrValue) = (float)num;
}
/* mask off any necessary bits */
/* if(! info->fIsLong_double)
num &= DBL_MAX;*/
/* update the input's position to reflect consumed data */
input->str.fPos += parsePos;
/* we converted 1 arg */
*argConverted = !info->fSkipArg;
return parsePos + skipped;
}
static int32_t
u_printf_scientific_handler(const u_printf_stream_handler *handler,
void *context,
ULocaleBundle *formatBundle,
const u_printf_spec_info *info,
const ufmt_args *args)
{
double num = (double) (args[0].doubleValue);
UNumberFormat *format;
UChar result[UPRINTF_BUFFER_SIZE];
UChar prefixBuffer[UPRINTF_BUFFER_SIZE];
int32_t prefixBufferLen = sizeof(prefixBuffer);
int32_t minDecimalDigits;
int32_t maxDecimalDigits;
UErrorCode status = U_ZERO_ERROR;
UChar srcExpBuf[UPRINTF_SYMBOL_BUFFER_SIZE];
int32_t srcLen, expLen;
int32_t resultLen;
UChar expBuf[UPRINTF_SYMBOL_BUFFER_SIZE];
prefixBuffer[0] = 0;
/* mask off any necessary bits */
/* if(! info->fIsLongDouble)
num &= DBL_MAX;*/
/* get the formatter */
format = u_locbund_getNumberFormat(formatBundle, UNUM_SCIENTIFIC);
/* handle error */
if(format == 0)
return 0;
/* set the appropriate flags on the formatter */
srcLen = unum_getSymbol(format,
UNUM_EXPONENTIAL_SYMBOL,
srcExpBuf,
sizeof(srcExpBuf),
&status);
/* Upper/lower case the e */
if (info->fSpec == (UChar)0x65 /* e */) {
expLen = u_strToLower(expBuf, (int32_t)sizeof(expBuf),
srcExpBuf, srcLen,
formatBundle->fLocale,
&status);
}
else {
expLen = u_strToUpper(expBuf, (int32_t)sizeof(expBuf),
srcExpBuf, srcLen,
formatBundle->fLocale,
&status);
}
unum_setSymbol(format,
UNUM_EXPONENTIAL_SYMBOL,
expBuf,
expLen,
&status);
/* save the formatter's state */
minDecimalDigits = unum_getAttribute(format, UNUM_MIN_FRACTION_DIGITS);
maxDecimalDigits = unum_getAttribute(format, UNUM_MAX_FRACTION_DIGITS);
/* set the appropriate flags and number of decimal digits on the formatter */
if(info->fPrecision != -1) {
/* set the # of decimal digits */
if (info->fOrigSpec == (UChar)0x65 /* e */ || info->fOrigSpec == (UChar)0x45 /* E */) {
unum_setAttribute(format, UNUM_FRACTION_DIGITS, info->fPrecision);
}
else {
unum_setAttribute(format, UNUM_MIN_FRACTION_DIGITS, 1);
unum_setAttribute(format, UNUM_MAX_FRACTION_DIGITS, info->fPrecision);
}
}
else if(info->fAlt) {
/* '#' means always show decimal point */
/* copy of printf behavior on Solaris - '#' shows 6 digits */
unum_setAttribute(format, UNUM_FRACTION_DIGITS, 6);
}
else {
/* # of decimal digits is 6 if precision not specified */
unum_setAttribute(format, UNUM_FRACTION_DIGITS, 6);
}
/* set whether to show the sign */
if (info->fShowSign) {
u_printf_set_sign(format, info, prefixBuffer, &prefixBufferLen, &status);
}
/* format the number */
resultLen = unum_formatDouble(format, num, result, UPRINTF_BUFFER_SIZE, 0, &status);
if (U_FAILURE(status)) {
resultLen = 0;
}
/* restore the number format */
/* TODO: Is this needed? */
unum_setAttribute(format, UNUM_MIN_FRACTION_DIGITS, minDecimalDigits);
unum_setAttribute(format, UNUM_MAX_FRACTION_DIGITS, maxDecimalDigits);
/* Since we're the only one using the scientific
format, we don't need to save the old exponent value. */
/*unum_setSymbol(format,
UNUM_EXPONENTIAL_SYMBOL,
srcExpBuf,
srcLen,
&status);*/
if (info->fShowSign) {
/* Reset back to original value regardless of what the error was */
UErrorCode localStatus = U_ZERO_ERROR;
u_printf_reset_sign(format, info, prefixBuffer, &prefixBufferLen, &localStatus);
}
return handler->pad_and_justify(context, info, result, resultLen);
}
static void demoCaseMapInC() {
/*
* input=
* "aB<capital sigma>"
* "iI<small dotless i><capital dotted I> "
* "<sharp s> <small lig. ffi>"
* "<small final sigma><small sigma><capital sigma>"
*/
static const UChar input[]={
0x61, 0x42, 0x3a3,
0x69, 0x49, 0x131, 0x130, 0x20,
0xdf, 0x20, 0xfb03,
0x3c2, 0x3c3, 0x3a3, 0
};
UChar buffer[32];
UErrorCode errorCode;
UChar32 c;
int32_t i, j, length;
UBool isError;
printf("\n* demoCaseMapInC() ----------------- ***\n\n");
/*
* First, use simple case mapping functions which provide
* 1:1 code point mappings without context/locale ID.
*
* Note that some mappings will not be "right" because some "real"
* case mappings require context, depend on the locale ID,
* and/or result in a change in the number of code points.
*/
printUString("input string: ", input, -1);
/* uppercase */
isError=FALSE;
for(i=j=0; j<UPRV_LENGTHOF(buffer) && !isError; /* U16_NEXT post-increments */) {
U16_NEXT(input, i, INT32_MAX, c); /* without length because NUL-terminated */
if(c==0) {
break; /* stop at terminating NUL, no need to terminate buffer */
}
c=u_toupper(c);
U16_APPEND(buffer, j, UPRV_LENGTHOF(buffer), c, isError);
}
printUString("simple-uppercased: ", buffer, j);
/* lowercase */
isError=FALSE;
for(i=j=0; j<UPRV_LENGTHOF(buffer) && !isError; /* U16_NEXT post-increments */) {
U16_NEXT(input, i, INT32_MAX, c); /* without length because NUL-terminated */
if(c==0) {
break; /* stop at terminating NUL, no need to terminate buffer */
}
c=u_tolower(c);
U16_APPEND(buffer, j, UPRV_LENGTHOF(buffer), c, isError);
}
printUString("simple-lowercased: ", buffer, j);
/* titlecase */
isError=FALSE;
for(i=j=0; j<UPRV_LENGTHOF(buffer) && !isError; /* U16_NEXT post-increments */) {
U16_NEXT(input, i, INT32_MAX, c); /* without length because NUL-terminated */
if(c==0) {
break; /* stop at terminating NUL, no need to terminate buffer */
}
c=u_totitle(c);
U16_APPEND(buffer, j, UPRV_LENGTHOF(buffer), c, isError);
}
printUString("simple-titlecased: ", buffer, j);
/* case-fold/default */
isError=FALSE;
for(i=j=0; j<UPRV_LENGTHOF(buffer) && !isError; /* U16_NEXT post-increments */) {
U16_NEXT(input, i, INT32_MAX, c); /* without length because NUL-terminated */
if(c==0) {
break; /* stop at terminating NUL, no need to terminate buffer */
}
c=u_foldCase(c, U_FOLD_CASE_DEFAULT);
U16_APPEND(buffer, j, UPRV_LENGTHOF(buffer), c, isError);
}
printUString("simple-case-folded/default: ", buffer, j);
/* case-fold/Turkic */
isError=FALSE;
for(i=j=0; j<UPRV_LENGTHOF(buffer) && !isError; /* U16_NEXT post-increments */) {
U16_NEXT(input, i, INT32_MAX, c); /* without length because NUL-terminated */
if(c==0) {
break; /* stop at terminating NUL, no need to terminate buffer */
}
c=u_foldCase(c, U_FOLD_CASE_EXCLUDE_SPECIAL_I);
U16_APPEND(buffer, j, UPRV_LENGTHOF(buffer), c, isError);
}
printUString("simple-case-folded/Turkic: ", buffer, j);
/*
* Second, use full case mapping functions which provide
* 1:n code point mappings (n can be 0!) and are sensitive to context and locale ID.
*
* Note that lower/upper/titlecasing take a locale ID while case-folding
* has bit flag options instead, by design of the Unicode SpecialCasing.txt UCD file.
*
* Also, string titlecasing requires a BreakIterator to find starts of words.
* The sample code here passes in a NULL pointer; u_strToTitle() will open and close a default
* titlecasing BreakIterator automatically.
* For production code where many strings are titlecased it would be more efficient
* to open a BreakIterator externally and pass it in.
*/
printUString("\ninput string: ", input, -1);
/* lowercase/English */
errorCode=U_ZERO_ERROR;
length=u_strToLower(buffer, UPRV_LENGTHOF(buffer), input, -1, "en", &errorCode);
if(U_SUCCESS(errorCode)) {
printUString("full-lowercased/en: ", buffer, length);
} else {
printf("error in u_strToLower(en)=%ld error=%s\n", length, u_errorName(errorCode));
}
/* lowercase/Turkish */
errorCode=U_ZERO_ERROR;
length=u_strToLower(buffer, UPRV_LENGTHOF(buffer), input, -1, "tr", &errorCode);
if(U_SUCCESS(errorCode)) {
printUString("full-lowercased/tr: ", buffer, length);
} else {
printf("error in u_strToLower(tr)=%ld error=%s\n", length, u_errorName(errorCode));
}
/* uppercase/English */
errorCode=U_ZERO_ERROR;
length=u_strToUpper(buffer, UPRV_LENGTHOF(buffer), input, -1, "en", &errorCode);
if(U_SUCCESS(errorCode)) {
printUString("full-uppercased/en: ", buffer, length);
} else {
printf("error in u_strToUpper(en)=%ld error=%s\n", length, u_errorName(errorCode));
}
/* uppercase/Turkish */
errorCode=U_ZERO_ERROR;
length=u_strToUpper(buffer, UPRV_LENGTHOF(buffer), input, -1, "tr", &errorCode);
if(U_SUCCESS(errorCode)) {
printUString("full-uppercased/tr: ", buffer, length);
} else {
printf("error in u_strToUpper(tr)=%ld error=%s\n", length, u_errorName(errorCode));
}
/* titlecase/English */
errorCode=U_ZERO_ERROR;
length=u_strToTitle(buffer, UPRV_LENGTHOF(buffer), input, -1, NULL, "en", &errorCode);
if(U_SUCCESS(errorCode)) {
printUString("full-titlecased/en: ", buffer, length);
} else {
printf("error in u_strToTitle(en)=%ld error=%s\n", length, u_errorName(errorCode));
}
/* titlecase/Turkish */
errorCode=U_ZERO_ERROR;
length=u_strToTitle(buffer, UPRV_LENGTHOF(buffer), input, -1, NULL, "tr", &errorCode);
if(U_SUCCESS(errorCode)) {
printUString("full-titlecased/tr: ", buffer, length);
} else {
printf("error in u_strToTitle(tr)=%ld error=%s\n", length, u_errorName(errorCode));
}
/* case-fold/default */
errorCode=U_ZERO_ERROR;
length=u_strFoldCase(buffer, UPRV_LENGTHOF(buffer), input, -1, U_FOLD_CASE_DEFAULT, &errorCode);
if(U_SUCCESS(errorCode)) {
printUString("full-case-folded/default: ", buffer, length);
} else {
printf("error in u_strFoldCase(default)=%ld error=%s\n", length, u_errorName(errorCode));
}
/* case-fold/Turkic */
errorCode=U_ZERO_ERROR;
length=u_strFoldCase(buffer, UPRV_LENGTHOF(buffer), input, -1, U_FOLD_CASE_EXCLUDE_SPECIAL_I, &errorCode);
if(U_SUCCESS(errorCode)) {
printUString("full-case-folded/Turkic: ", buffer, length);
} else {
printf("error in u_strFoldCase(Turkic)=%ld error=%s\n", length, u_errorName(errorCode));
}
}
/**
* psl_str_to_utf8lower:
* @str: string to convert
* @encoding: charset encoding of @str, e.g. 'iso-8859-1' or %NULL
* @locale: locale of @str for to lowercase conversion, e.g. 'de' or %NULL
* @lower: return value containing the converted string
*
* This helper function converts a string to lowercase UTF-8 representation.
* Lowercase UTF-8 is needed as input to the domain checking functions.
*
* @lower is set to %NULL on error.
*
* The return value 'lower' must be freed after usage.
*
* Returns: psl_error_t value.
* PSL_SUCCESS: Success
* PSL_ERR_INVALID_ARG: @str is a %NULL value.
* PSL_ERR_CONVERTER: Failed to open the unicode converter with name @encoding
* PSL_ERR_TO_UTF16: Failed to convert @str to unicode
* PSL_ERR_TO_LOWER: Failed to convert unicode to lowercase
* PSL_ERR_TO_UTF8: Failed to convert unicode to UTF-8
*
* Since: 0.4
*/
psl_error_t psl_str_to_utf8lower(const char *str, const char *encoding, const char *locale, char **lower)
{
int ret = PSL_ERR_INVALID_ARG;
if (lower)
*lower = NULL;
if (!str)
return PSL_ERR_INVALID_ARG;
/* shortcut to avoid costly conversion */
if (_str_is_ascii(str)) {
if (lower) {
char *p;
*lower = strdup(str);
/* convert ASCII string to lowercase */
for (p = *lower; *p; p++)
if (isupper(*p))
*p = tolower(*p);
}
return PSL_SUCCESS;
}
#ifdef WITH_LIBICU
do {
size_t str_length = strlen(str);
UErrorCode status = 0;
UChar *utf16_dst, *utf16_lower;
int32_t utf16_dst_length;
char *utf8_lower;
UConverter *uconv;
/* C89 allocation */
utf16_dst = alloca(sizeof(UChar) * (str_length * 2 + 1));
utf16_lower = alloca(sizeof(UChar) * (str_length * 2 + 1));
utf8_lower = alloca(str_length * 2 + 1);
uconv = ucnv_open(encoding, &status);
if (U_SUCCESS(status)) {
utf16_dst_length = ucnv_toUChars(uconv, utf16_dst, str_length * 2 + 1, str, str_length, &status);
ucnv_close(uconv);
if (U_SUCCESS(status)) {
int32_t utf16_lower_length = u_strToLower(utf16_lower, str_length * 2 + 1, utf16_dst, utf16_dst_length, locale, &status);
if (U_SUCCESS(status)) {
u_strToUTF8(utf8_lower, str_length * 8 + 1, NULL, utf16_lower, utf16_lower_length, &status);
if (U_SUCCESS(status)) {
if (lower)
*lower = strdup(utf8_lower);
ret = PSL_SUCCESS;
} else {
ret = PSL_ERR_TO_UTF8;
/* fprintf(stderr, "Failed to convert UTF-16 to UTF-8 (status %d)\n", status); */
}
} else {
ret = PSL_ERR_TO_LOWER;
/* fprintf(stderr, "Failed to convert UTF-16 to lowercase (status %d)\n", status); */
}
} else {
ret = PSL_ERR_TO_UTF16;
/* fprintf(stderr, "Failed to convert string to UTF-16 (status %d)\n", status); */
}
} else {
ret = PSL_ERR_CONVERTER;
/* fprintf(stderr, "Failed to open converter for '%s' (status %d)\n", encoding, status); */
}
} while (0);
#elif defined(WITH_LIBIDN2) || defined(WITH_LIBIDN)
do {
/* find out local charset encoding */
if (!encoding) {
encoding = nl_langinfo(CODESET);
if (!encoding || !*encoding)
encoding = "ASCII";
}
/* convert to UTF-8 */
if (strcasecmp(encoding, "utf-8")) {
iconv_t cd = iconv_open("utf-8", encoding);
if (cd != (iconv_t)-1) {
char *tmp = (char *)str; /* iconv won't change where str points to, but changes tmp itself */
size_t tmp_len = strlen(str);
size_t dst_len = tmp_len * 6, dst_len_tmp = dst_len;
char *dst = malloc(dst_len + 1), *dst_tmp = dst;
if (iconv(cd, &tmp, &tmp_len, &dst_tmp, &dst_len_tmp) != (size_t)-1) {
uint8_t *resbuf = malloc(dst_len * 2 + 1);
size_t len = dst_len * 2; /* leave space for additional \0 byte */
if ((dst = (char *)u8_tolower((uint8_t *)dst, dst_len - dst_len_tmp, 0, UNINORM_NFKC, resbuf, &len))) {
/* u8_tolower() does not terminate the result string */
if (lower)
*lower = strndup((char *)dst, len);
} else {
ret = PSL_ERR_TO_LOWER;
/* fprintf(stderr, "Failed to convert UTF-8 to lowercase (errno %d)\n", errno); */
}
if (lower)
*lower = strndup(dst, dst_len - dst_len_tmp);
ret = PSL_SUCCESS;
} else {
ret = PSL_ERR_TO_UTF8;
/* fprintf(stderr, "Failed to convert '%s' string into '%s' (%d)\n", src_encoding, dst_encoding, errno); */
}
free(dst);
iconv_close(cd);
} else {
ret = PSL_ERR_TO_UTF8;
/* fprintf(stderr, "Failed to prepare encoding '%s' into '%s' (%d)\n", src_encoding, dst_encoding, errno); */
}
} else
ret = PSL_SUCCESS;
/* convert to lowercase */
if (ret == PSL_SUCCESS) {
uint8_t *dst, resbuf[256];
size_t len = sizeof(resbuf) - 1; /* leave space for additional \0 byte */
/* we need a conversion to lowercase */
if ((dst = u8_tolower((uint8_t *)str, u8_strlen((uint8_t *)str), 0, UNINORM_NFKC, resbuf, &len))) {
/* u8_tolower() does not terminate the result string */
if (lower)
*lower = strndup((char *)dst, len);
} else {
ret = PSL_ERR_TO_LOWER;
/* fprintf(stderr, "Failed to convert UTF-8 to lowercase (errno %d)\n", errno); */
}
}
} while (0);
#endif
return ret;
}
int32_t __hs_u_strToLower(UChar *dest, int32_t destCapacity,
const UChar *src, int32_t srcLength,
const char *locale, UErrorCode *pErrorCode)
{
return u_strToLower(dest, destCapacity, src, srcLength, locale, pErrorCode);
}
