// BreakIterator.split {{{
static PyObject *
icu_BreakIterator_split(icu_BreakIterator *self, PyObject *args, PyObject *kwargs) {
int32_t prev = 0, p = 0, sz = 0;
PyObject *ans = NULL, *token = NULL;
ans = PyList_New(0);
if (ans == NULL) return PyErr_NoMemory();
p = ubrk_first(self->break_iterator);
while (p != UBRK_DONE) {
prev = p; p = ubrk_next(self->break_iterator);
if (self->type == UBRK_WORD && ubrk_getRuleStatus(self->break_iterator) == UBRK_WORD_NONE)
continue; // We are not at the start of a word
sz = (p == UBRK_DONE) ? self->text_len - prev : p - prev;
if (sz > 0) {
token = icu_to_python(self->text + prev, sz);
if (token == NULL) {
Py_DECREF(ans); ans = NULL; break;
}
if (PyList_Append(ans, token) != 0) {
Py_DECREF(token); Py_DECREF(ans); ans = NULL; break;
}
Py_DECREF(token);
}
}
return ans;
} // }}}
static PyObject* icu_swap_case(PyObject *self, PyObject *input) {
PyObject *result = NULL;
UErrorCode status = U_ZERO_ERROR;
UChar *input_buf = NULL, *output_buf = NULL;
UChar32 *buf = NULL;
int32_t sz = 0, sz32 = 0, i = 0;
input_buf = python_to_icu(input, &sz);
if (input_buf == NULL) goto end;
output_buf = (UChar*) calloc(3 * sz, sizeof(UChar));
buf = (UChar32*) calloc(2 * sz, sizeof(UChar32));
if (output_buf == NULL || buf == NULL) { PyErr_NoMemory(); goto end; }
u_strToUTF32(buf, 2 * sz, &sz32, input_buf, sz, &status);
for (i = 0; i < sz32; i++) {
if (u_islower(buf[i])) buf[i] = u_toupper(buf[i]);
else if (u_isupper(buf[i])) buf[i] = u_tolower(buf[i]);
}
u_strFromUTF32(output_buf, 3*sz, &sz, buf, sz32, &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);
if (buf != NULL) free(buf);
return result;
} // }}}
static PyObject *
icu_normalize(PyObject *self, PyObject *args) {
UErrorCode status = U_ZERO_ERROR;
int32_t sz = 0, cap = 0, rsz = 0;
NORM_MODES mode;
UChar *dest = NULL, *source = NULL;
PyObject *ret = NULL, *src = NULL;
if (!PyArg_ParseTuple(args, "iO", &mode, &src)) return NULL;
const UNormalizer2 *n = NULL;
switch (mode) {
case NFC:
n = unorm2_getNFCInstance(&status);
break;
case NFKC:
n = unorm2_getNFKCInstance(&status);
break;
case NFD:
n = unorm2_getNFDInstance(&status);
break;
case NFKD:
n = unorm2_getNFKDInstance(&status);
break;
}
if (U_FAILURE(status)) {
PyErr_SetString(PyExc_ValueError, u_errorName(status));
goto end;
}
source = python_to_icu(src, &sz);
if (source == NULL) goto end;
cap = 2 * sz;
dest = (UChar*) calloc(cap, sizeof(UChar));
if (dest == NULL) { PyErr_NoMemory(); goto end; }
while (1) {
rsz = unorm2_normalize(n, source, sz, dest, cap, &status);
if (status == U_BUFFER_OVERFLOW_ERROR) {
cap *= 2;
dest = (UChar*) realloc(dest, cap*sizeof(UChar));
if (dest == NULL) { PyErr_NoMemory(); goto end; }
continue;
}
break;
}
if (U_FAILURE(status)) {
PyErr_SetString(PyExc_ValueError, u_errorName(status));
goto end;
}
ret = icu_to_python(dest, rsz);
end:
if (source != NULL) free(source);
if (dest != NULL) free(dest);
return ret;
} // }}}
// chr {{{
static PyObject *
icu_chr(PyObject *self, PyObject *args) {
UErrorCode status = U_ZERO_ERROR;
UChar32 code = 0;
UChar buf[5] = {0};
int32_t sz = 0;
if (!PyArg_ParseTuple(args, "I", &code)) return NULL;
u_strFromUTF32(buf, 4, &sz, &code, 1, &status);
if (U_FAILURE(status)) { PyErr_SetString(PyExc_ValueError, "arg not in range(0x110000)"); return NULL; }
return icu_to_python(buf, sz);
} // }}}
// roundtrip {{{
static PyObject *
icu_roundtrip(PyObject *self, PyObject *src) {
int32_t sz = 0;
UChar *icu = NULL;
PyObject *ret = NULL;
icu = python_to_icu(src, &sz);
if (icu != NULL) {
ret = icu_to_python(icu, sz);
free(icu);
}
return ret;
} // }}}
// roundtrip {{{
static PyObject *
icu_roundtrip(PyObject *self, PyObject *args) {
int32_t sz = 0;
UChar *icu = NULL;
PyObject *ret = NULL, *src = NULL;
if (!PyArg_ParseTuple(args, "O", &src)) return NULL;
icu = python_to_icu(src, &sz, 1);
if (icu != NULL) {
ret = icu_to_python(icu, sz);
free(icu);
}
return ret;
} // }}}
// Collator.display_name {{{
static PyObject *
icu_Collator_display_name(icu_Collator *self, void *closure) {
const char *loc = NULL;
UErrorCode status = U_ZERO_ERROR;
UChar dname[400];
int32_t sz = 0;
loc = ucol_getLocaleByType(self->collator, ULOC_ACTUAL_LOCALE, &status);
if (loc == NULL) {
PyErr_SetString(PyExc_Exception, "Failed to get actual locale"); return NULL;
}
sz = ucol_getDisplayName(loc, "en", dname, sizeof(dname), &status);
if (U_FAILURE(status)) {PyErr_SetString(PyExc_ValueError, u_errorName(status)); return NULL; }
return icu_to_python(dname, sz);
}
// Collator.contractions {{{
static PyObject *
icu_Collator_contractions(icu_Collator *self, PyObject *args, PyObject *kwargs) {
UErrorCode status = U_ZERO_ERROR;
UChar *str = NULL;
UChar32 start=0, end=0;
int32_t count = 0, len = 0, i;
PyObject *ans = Py_None, *pbuf;
if (self->contractions == NULL) {
self->contractions = uset_open(1, 0);
if (self->contractions == NULL) return PyErr_NoMemory();
self->contractions = ucol_getTailoredSet(self->collator, &status);
}
status = U_ZERO_ERROR;
count = uset_getItemCount(self->contractions);
str = (UChar*)calloc(100, sizeof(UChar));
if (str == NULL) { PyErr_NoMemory(); goto end; }
ans = PyTuple_New(count);
if (ans == NULL) { goto end; }
for (i = 0; i < count; i++) {
len = uset_getItem(self->contractions, i, &start, &end, str, 1000, &status);
if (len >= 2) {
// We have a string
status = U_ZERO_ERROR;
pbuf = icu_to_python(str, len);
if (pbuf == NULL) { Py_DECREF(ans); ans = NULL; goto end; }
PyTuple_SetItem(ans, i, pbuf);
} else {
// Ranges dont make sense for contractions, ignore them
PyTuple_SetItem(ans, i, Py_None); Py_INCREF(Py_None);
}
}
end:
if (str != NULL) free(str);
return ans;
} // }}}
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;
} // }}}
// normalize {{{
static PyObject *
icu_normalize(PyObject *self, PyObject *args) {
UErrorCode status = U_ZERO_ERROR;
int32_t sz = 0, mode = UNORM_DEFAULT, cap = 0, rsz = 0;
UChar *dest = NULL, *source = NULL;
PyObject *ret = NULL, *src = NULL;
if (!PyArg_ParseTuple(args, "iO", &mode, &src)) return NULL;
source = python_to_icu(src, &sz, 1);
if (source == NULL) goto end;
cap = 2 * sz;
dest = (UChar*) calloc(cap, sizeof(UChar));
if (dest == NULL) { PyErr_NoMemory(); goto end; }
while (1) {
rsz = unorm_normalize(source, sz, (UNormalizationMode)mode, 0, dest, cap, &status);
if (status == U_BUFFER_OVERFLOW_ERROR) {
cap *= 2;
dest = (UChar*) realloc(dest, cap*sizeof(UChar));
if (dest == NULL) { PyErr_NoMemory(); goto end; }
continue;
}
break;
}
if (U_FAILURE(status)) {
PyErr_SetString(PyExc_ValueError, u_errorName(status));
goto end;
}
ret = icu_to_python(dest, rsz);
end:
if (source != NULL) free(source);
if (dest != NULL) free(dest);
return ret;
} // }}}