diff options
| author | Victor Stinner <victor.stinner@gmail.com> | 2016-03-01 20:59:58 (GMT) |
|---|---|---|
| committer | Victor Stinner <victor.stinner@gmail.com> | 2016-03-01 20:59:58 (GMT) |
| commit | 337986740f26ade866dbe79e35188f72caa8fe8d (patch) | |
| tree | 0dfe877dfcfe031ede9b96798c8f3645cb4a25e5 | |
| parent | 6c9aa8f2bf46e28de74ce11d85bcd448c0d0d529 (diff) | |
| download | cpython-337986740f26ade866dbe79e35188f72caa8fe8d.zip cpython-337986740f26ade866dbe79e35188f72caa8fe8d.tar.gz cpython-337986740f26ade866dbe79e35188f72caa8fe8d.tar.bz2 | |
Issue #26464: Fix unicode_fast_translate() again
Initialize i variable if the string is non-ASCII.
| -rw-r--r-- | Lib/test/test_unicode.py | 14 | ||||
| -rw-r--r-- | Objects/unicodeobject.c | 21 |
2 files changed, 22 insertions, 13 deletions
diff --git a/Lib/test/test_unicode.py b/Lib/test/test_unicode.py index b258db1..c30310e 100644 --- a/Lib/test/test_unicode.py +++ b/Lib/test/test_unicode.py @@ -341,16 +341,22 @@ class UnicodeTest(string_tests.CommonTest, "[XXX]") self.assertEqual("[a]".translate(str.maketrans({'a': '\xe9'})), "[\xe9]") + self.assertEqual('axb'.translate(str.maketrans({'a': None, 'b': '123'})), + "x123") + self.assertEqual('axb'.translate(str.maketrans({'a': None, 'b': '\xe9'})), + "x\xe9") + + # test non-ASCII (don't take the fast-path) self.assertEqual("[a]".translate(str.maketrans({'a': '<\xe9>'})), "[<\xe9>]") self.assertEqual("[\xe9]".translate(str.maketrans({'\xe9': 'a'})), "[a]") self.assertEqual("[\xe9]".translate(str.maketrans({'\xe9': None})), "[]") - self.assertEqual('axb'.translate(str.maketrans({'a': None, 'b': '123'})), - "x123") - self.assertEqual('axb'.translate(str.maketrans({'a': None, 'b': '\xe9'})), - "x\xe9") + self.assertEqual("[\xe9]".translate(str.maketrans({'\xe9': '123'})), + "[123]") + self.assertEqual("[a\xe9]".translate(str.maketrans({'a': '<\u20ac>'})), + "[<\u20ac>\xe9]") # invalid Unicode characters invalid_char = 0x10ffff+1 diff --git a/Objects/unicodeobject.c b/Objects/unicodeobject.c index a1ee776..adc4615 100644 --- a/Objects/unicodeobject.c +++ b/Objects/unicodeobject.c @@ -8582,10 +8582,6 @@ unicode_fast_translate(PyObject *input, PyObject *mapping, Py_UCS1 *in, *end, *out; int res = 0; - if (PyUnicode_READY(input) == -1) - return -1; - if (!PyUnicode_IS_ASCII(input)) - return 0; len = PyUnicode_GET_LENGTH(input); memset(ascii_table, 0xff, 128); @@ -8668,13 +8664,20 @@ _PyUnicode_TranslateCharmap(PyObject *input, ignore = (errors != NULL && strcmp(errors, "ignore") == 0); - res = unicode_fast_translate(input, mapping, &writer, ignore, &i); - if (res < 0) { - _PyUnicodeWriter_Dealloc(&writer); + if (PyUnicode_READY(input) == -1) return NULL; + if (PyUnicode_IS_ASCII(input)) { + res = unicode_fast_translate(input, mapping, &writer, ignore, &i); + if (res < 0) { + _PyUnicodeWriter_Dealloc(&writer); + return NULL; + } + if (res == 1) + return _PyUnicodeWriter_Finish(&writer); + } + else { + i = 0; } - if (res == 1) - return _PyUnicodeWriter_Finish(&writer); while (i<size) { /* try to encode it */ |