mirrors/cpython
1
0
Fork 0
mirror of https://github.com/python/cpython.git synced 2025-08-30 21:48:47 +00:00
Code Issues Projects Releases Packages Wiki Activity Actions 7

merge 3.3 (#18183)

Browse source
This commit is contained in:
Benjamin Peterson 2013-06-10 09:24:01 -07:00
commit 3164f5d565
3 changed files with 25 additions and 24 deletions
Download patch file Download diff file Expand all files Collapse all files

3
Lib/test/test_unicode.py
View file

@ -577,6 +577,9 @@ class UnicodeTest(string_tests.CommonTest,
self.assertEqual('\U0008fffe'.lower(), '\U0008fffe') self.assertEqual('\U0008fffe'.lower(), '\U0008fffe')
self.assertEqual('\u2177'.lower(), '\u2177') self.assertEqual('\u2177'.lower(), '\u2177')
# See issue #18183 for this one.
'\U00010000\U00100000'.lower()
def test_casefold(self): def test_casefold(self):
self.assertEqual('hello'.casefold(), 'hello') self.assertEqual('hello'.casefold(), 'hello')
self.assertEqual('hELlo'.casefold(), 'hello') self.assertEqual('hELlo'.casefold(), 'hello')

3
Misc/NEWS
View file

@ -10,6 +10,9 @@ What's New in Python 3.4.0 Alpha 1?
Core and Builtins Core and Builtins
----------------- -----------------
- Issue #18183: Fix various unicode operations on strings with large unicode
codepoints.
- Issue #18180: Fix ref leak in _PyImport_GetDynLoadWindows(). - Issue #18180: Fix ref leak in _PyImport_GetDynLoadWindows().
- Issue #18038: SyntaxError raised during compilation sources with illegal - Issue #18038: SyntaxError raised during compilation sources with illegal

43
Objects/unicodeobject.c
View file

@ -104,11 +104,6 @@ extern "C" {
#define _PyUnicode_DATA_ANY(op) \ #define _PyUnicode_DATA_ANY(op) \
(((PyUnicodeObject*)(op))->data.any) (((PyUnicodeObject*)(op))->data.any)
/* Optimized version of Py_MAX() to compute the maximum character:
use it when your are computing the second argument of PyUnicode_New() */
#define MAX_MAXCHAR(maxchar1, maxchar2) \
((maxchar1) | (maxchar2))
#undef PyUnicode_READY #undef PyUnicode_READY
#define PyUnicode_READY(op) \ #define PyUnicode_READY(op) \
(assert(_PyUnicode_CHECK(op)), \ (assert(_PyUnicode_CHECK(op)), \
@ -8609,11 +8604,11 @@ fix_decimal_and_space_to_ascii(PyObject *self)
} }
if (fixed != 0) { if (fixed != 0) {
modified = 1; modified = 1;
maxchar = MAX_MAXCHAR(maxchar, fixed); maxchar = Py_MAX(maxchar, fixed);
PyUnicode_WRITE(kind, data, i, fixed); PyUnicode_WRITE(kind, data, i, fixed);
} }
else else
maxchar = MAX_MAXCHAR(maxchar, ch); maxchar = Py_MAX(maxchar, ch);
} }
} }
@ -8654,7 +8649,7 @@ PyUnicode_TransformDecimalToASCII(Py_UNICODE *s,
int decimal = Py_UNICODE_TODECIMAL(ch); int decimal = Py_UNICODE_TODECIMAL(ch);
if (decimal >= 0) if (decimal >= 0)
ch = '0' + decimal; ch = '0' + decimal;
maxchar = MAX_MAXCHAR(maxchar, ch); maxchar = Py_MAX(maxchar, ch);
} }
} }
@ -8895,7 +8890,7 @@ _PyUnicode_InsertThousandsGrouping(
if (unicode == NULL) { if (unicode == NULL) {
*maxchar = 127; *maxchar = 127;
if (len != n_digits) { if (len != n_digits) {
*maxchar = MAX_MAXCHAR(*maxchar, *maxchar = Py_MAX(*maxchar,
PyUnicode_MAX_CHAR_VALUE(thousands_sep)); PyUnicode_MAX_CHAR_VALUE(thousands_sep));
} }
} }
@ -9289,14 +9284,14 @@ do_capitalize(int kind, void *data, Py_ssize_t length, Py_UCS4 *res, Py_UCS4 *ma
c = PyUnicode_READ(kind, data, 0); c = PyUnicode_READ(kind, data, 0);
n_res = _PyUnicode_ToUpperFull(c, mapped); n_res = _PyUnicode_ToUpperFull(c, mapped);
for (j = 0; j < n_res; j++) { for (j = 0; j < n_res; j++) {
*maxchar = MAX_MAXCHAR(*maxchar, mapped[j]); *maxchar = Py_MAX(*maxchar, mapped[j]);
res[k++] = mapped[j]; res[k++] = mapped[j];
} }
for (i = 1; i < length; i++) { for (i = 1; i < length; i++) {
c = PyUnicode_READ(kind, data, i); c = PyUnicode_READ(kind, data, i);
n_res = lower_ucs4(kind, data, length, i, c, mapped); n_res = lower_ucs4(kind, data, length, i, c, mapped);
for (j = 0; j < n_res; j++) { for (j = 0; j < n_res; j++) {
*maxchar = MAX_MAXCHAR(*maxchar, mapped[j]); *maxchar = Py_MAX(*maxchar, mapped[j]);
res[k++] = mapped[j]; res[k++] = mapped[j];
} }
} }
@ -9321,7 +9316,7 @@ do_swapcase(int kind, void *data, Py_ssize_t length, Py_UCS4 *res, Py_UCS4 *maxc
mapped[0] = c; mapped[0] = c;
} }
for (j = 0; j < n_res; j++) { for (j = 0; j < n_res; j++) {
*maxchar = MAX_MAXCHAR(*maxchar, mapped[j]); *maxchar = Py_MAX(*maxchar, mapped[j]);
res[k++] = mapped[j]; res[k++] = mapped[j];
} }
} }
@ -9342,7 +9337,7 @@ do_upper_or_lower(int kind, void *data, Py_ssize_t length, Py_UCS4 *res,
else else
n_res = _PyUnicode_ToUpperFull(c, mapped); n_res = _PyUnicode_ToUpperFull(c, mapped);
for (j = 0; j < n_res; j++) { for (j = 0; j < n_res; j++) {
*maxchar = MAX_MAXCHAR(*maxchar, mapped[j]); *maxchar = Py_MAX(*maxchar, mapped[j]);
res[k++] = mapped[j]; res[k++] = mapped[j];
} }
} }
@ -9371,7 +9366,7 @@ do_casefold(int kind, void *data, Py_ssize_t length, Py_UCS4 *res, Py_UCS4 *maxc
Py_UCS4 mapped[3]; Py_UCS4 mapped[3];
int j, n_res = _PyUnicode_ToFoldedFull(c, mapped); int j, n_res = _PyUnicode_ToFoldedFull(c, mapped);
for (j = 0; j < n_res; j++) { for (j = 0; j < n_res; j++) {
*maxchar = MAX_MAXCHAR(*maxchar, mapped[j]); *maxchar = Py_MAX(*maxchar, mapped[j]);
res[k++] = mapped[j]; res[k++] = mapped[j];
} }
} }
@ -9396,7 +9391,7 @@ do_title(int kind, void *data, Py_ssize_t length, Py_UCS4 *res, Py_UCS4 *maxchar
n_res = _PyUnicode_ToTitleFull(c, mapped); n_res = _PyUnicode_ToTitleFull(c, mapped);
for (j = 0; j < n_res; j++) { for (j = 0; j < n_res; j++) {
*maxchar = MAX_MAXCHAR(*maxchar, mapped[j]); *maxchar = Py_MAX(*maxchar, mapped[j]);
res[k++] = mapped[j]; res[k++] = mapped[j];
} }
@ -9551,7 +9546,7 @@ PyUnicode_Join(PyObject *separator, PyObject *seq)
goto onError; goto onError;
sz += PyUnicode_GET_LENGTH(item); sz += PyUnicode_GET_LENGTH(item);
item_maxchar = PyUnicode_MAX_CHAR_VALUE(item); item_maxchar = PyUnicode_MAX_CHAR_VALUE(item);
maxchar = MAX_MAXCHAR(maxchar, item_maxchar); maxchar = Py_MAX(maxchar, item_maxchar);
if (i != 0) if (i != 0)
sz += seplen; sz += seplen;
if (sz < old_sz || sz > PY_SSIZE_T_MAX) { if (sz < old_sz || sz > PY_SSIZE_T_MAX) {
@ -9735,7 +9730,7 @@ pad(PyObject *self,
return NULL; return NULL;
} }
maxchar = PyUnicode_MAX_CHAR_VALUE(self); maxchar = PyUnicode_MAX_CHAR_VALUE(self);
maxchar = MAX_MAXCHAR(maxchar, fill); maxchar = Py_MAX(maxchar, fill);
u = PyUnicode_New(left + _PyUnicode_LENGTH(self) + right, maxchar); u = PyUnicode_New(left + _PyUnicode_LENGTH(self) + right, maxchar);
if (!u) if (!u)
return NULL; return NULL;
@ -10075,7 +10070,7 @@ replace(PyObject *self, PyObject *str1,
/* Replacing str1 with str2 may cause a maxchar reduction in the /* Replacing str1 with str2 may cause a maxchar reduction in the
result string. */ result string. */
mayshrink = (maxchar_str2 < maxchar_str1) && (maxchar == maxchar_str1); mayshrink = (maxchar_str2 < maxchar_str1) && (maxchar == maxchar_str1);
maxchar = MAX_MAXCHAR(maxchar, maxchar_str2); maxchar = Py_MAX(maxchar, maxchar_str2);
if (len1 == len2) { if (len1 == len2) {
/* same length */ /* same length */
@ -10749,7 +10744,7 @@ PyUnicode_Concat(PyObject *left, PyObject *right)
maxchar = PyUnicode_MAX_CHAR_VALUE(u); maxchar = PyUnicode_MAX_CHAR_VALUE(u);
maxchar2 = PyUnicode_MAX_CHAR_VALUE(v); maxchar2 = PyUnicode_MAX_CHAR_VALUE(v);
maxchar = MAX_MAXCHAR(maxchar, maxchar2); maxchar = Py_MAX(maxchar, maxchar2);
/* Concat the two Unicode strings */ /* Concat the two Unicode strings */
w = PyUnicode_New(new_len, maxchar); w = PyUnicode_New(new_len, maxchar);
@ -10831,7 +10826,7 @@ PyUnicode_Append(PyObject **p_left, PyObject *right)
else { else {
maxchar = PyUnicode_MAX_CHAR_VALUE(left); maxchar = PyUnicode_MAX_CHAR_VALUE(left);
maxchar2 = PyUnicode_MAX_CHAR_VALUE(right); maxchar2 = PyUnicode_MAX_CHAR_VALUE(right);
maxchar = MAX_MAXCHAR(maxchar, maxchar2); maxchar = Py_MAX(maxchar, maxchar2);
/* Concat the two Unicode strings */ /* Concat the two Unicode strings */
res = PyUnicode_New(new_len, maxchar); res = PyUnicode_New(new_len, maxchar);
@ -12993,7 +12988,7 @@ _PyUnicodeWriter_PrepareInternal(_PyUnicodeWriter *writer,
} }
newlen = writer->pos + length; newlen = writer->pos + length;
maxchar = MAX_MAXCHAR(maxchar, writer->min_char); maxchar = Py_MAX(maxchar, writer->min_char);
if (writer->buffer == NULL) { if (writer->buffer == NULL) {
assert(!writer->readonly); assert(!writer->readonly);
@ -14133,16 +14128,16 @@ unicode_format_arg_output(struct unicode_formatter_t *ctx,
if (!(arg->flags & F_LJUST)) { if (!(arg->flags & F_LJUST)) {
if (arg->sign) { if (arg->sign) {
if ((arg->width-1) > len) if ((arg->width-1) > len)
maxchar = MAX_MAXCHAR(maxchar, fill); maxchar = Py_MAX(maxchar, fill);
} }
else { else {
if (arg->width > len) if (arg->width > len)
maxchar = MAX_MAXCHAR(maxchar, fill); maxchar = Py_MAX(maxchar, fill);
} }
} }
if (PyUnicode_MAX_CHAR_VALUE(str) > maxchar) { if (PyUnicode_MAX_CHAR_VALUE(str) > maxchar) {
Py_UCS4 strmaxchar = _PyUnicode_FindMaxChar(str, 0, pindex+len); Py_UCS4 strmaxchar = _PyUnicode_FindMaxChar(str, 0, pindex+len);
maxchar = MAX_MAXCHAR(maxchar, strmaxchar); maxchar = Py_MAX(maxchar, strmaxchar);
} }
buflen = arg->width; buflen = arg->width;