diff options
| -rw-r--r-- | Doc/c-api/unicode.rst | 3 | ||||
| -rw-r--r-- | Lib/test/test_unicode.py | 39 | ||||
| -rw-r--r-- | Modules/_testcapimodule.c | 72 | ||||
| -rw-r--r-- | Objects/unicodeobject.c | 2 |
4 files changed, 105 insertions, 11 deletions
diff --git a/Doc/c-api/unicode.rst b/Doc/c-api/unicode.rst index 50a1254..45c46aa 100644 --- a/Doc/c-api/unicode.rst +++ b/Doc/c-api/unicode.rst @@ -466,7 +466,8 @@ wchar_t support for platforms which support it: Convert the Unicode object to a wide character string. The output string always ends with a nul character. If *size* is not *NULL*, write the number - of wide characters (including the nul character) into *\*size*. + of wide characters (excluding the trailing 0-termination character) into + *\*size*. Returns a buffer allocated by :cfunc:`PyMem_Alloc` (use :cfunc:`PyMem_Free` to free it) on success. On error, returns *NULL*, *\*size* is undefined and diff --git a/Lib/test/test_unicode.py b/Lib/test/test_unicode.py index f6c38dd..2ac79fb 100644 --- a/Lib/test/test_unicode.py +++ b/Lib/test/test_unicode.py @@ -1394,6 +1394,45 @@ class UnicodeTest(string_tests.CommonTest, 'string, got a non-ASCII byte: 0xe9$', format_unicode, b'unicode\xe9=%s', 'ascii') + # Test PyUnicode_AsWideChar() + def test_aswidechar(self): + from _testcapi import test_aswidechar + from ctypes import c_wchar, sizeof + + wchar, size = test_aswidechar('abcdef', 2) + self.assertEquals(size, 2) + self.assertEquals(wchar, 'ab') + + wchar, size = test_aswidechar('abc', 3) + self.assertEquals(size, 3) + self.assertEquals(wchar, 'abc') + + wchar, size = test_aswidechar('abc', 4) + self.assertEquals(size, 3) + self.assertEquals(wchar, 'abc\0') + + wchar, size = test_aswidechar('abc', 10) + self.assertEquals(size, 3) + self.assertEquals(wchar, 'abc\0') + + wchar, size = test_aswidechar('abc\0def', 20) + self.assertEquals(size, 7) + self.assertEquals(wchar, 'abc\0def\0') + + # Test PyUnicode_AsWideCharString() + def test_aswidecharstring(self): + from _testcapi import test_aswidecharstring + from ctypes import c_wchar, sizeof + + wchar, size = test_aswidecharstring('abc') + self.assertEquals(size, 3) + self.assertEquals(wchar, 'abc\0') + + wchar, size = test_aswidecharstring('abc\0def') + self.assertEquals(size, 7) + self.assertEquals(wchar, 'abc\0def\0') + + def test_main(): support.run_unittest(__name__) diff --git a/Modules/_testcapimodule.c b/Modules/_testcapimodule.c index 20887b1..d2d3b19 100644 --- a/Modules/_testcapimodule.c +++ b/Modules/_testcapimodule.c @@ -1386,6 +1386,58 @@ test_widechar(PyObject *self) } static PyObject * +test_aswidechar(PyObject *self, PyObject *args) +{ + PyObject *unicode, *result; + Py_ssize_t buflen, size; + wchar_t *buffer; + + if (!PyArg_ParseTuple(args, "Un", &unicode, &buflen)) + return NULL; + buffer = PyMem_Malloc(buflen * sizeof(wchar_t)); + if (buffer == NULL) + return PyErr_NoMemory(); + + size = PyUnicode_AsWideChar((PyUnicodeObject*)unicode, buffer, buflen); + if (size == -1) { + PyMem_Free(buffer); + return NULL; + } + + if (size < buflen) + buflen = size + 1; + else + buflen = size; + result = PyUnicode_FromWideChar(buffer, buflen); + PyMem_Free(buffer); + if (result == NULL) + return NULL; + + return Py_BuildValue("(Nn)", result, size); +} + +static PyObject * +test_aswidecharstring(PyObject *self, PyObject *args) +{ + PyObject *unicode, *result; + Py_ssize_t size; + wchar_t *buffer; + + if (!PyArg_ParseTuple(args, "U", &unicode)) + return NULL; + + buffer = PyUnicode_AsWideCharString((PyUnicodeObject*)unicode, &size); + if (buffer == NULL) + return NULL; + + result = PyUnicode_FromWideChar(buffer, size + 1); + PyMem_Free(buffer); + if (result == NULL) + return NULL; + return Py_BuildValue("(Nn)", result, size); +} + +static PyObject * getargs_w_star(PyObject *self, PyObject *args) { Py_buffer buffer; @@ -2262,28 +2314,30 @@ static PyMethodDef TestMethods[] = { {"getargs_Z_hash", getargs_Z_hash, METH_VARARGS}, {"getargs_w_star", getargs_w_star, METH_VARARGS}, {"codec_incrementalencoder", - (PyCFunction)codec_incrementalencoder, METH_VARARGS}, + (PyCFunction)codec_incrementalencoder, METH_VARARGS}, {"codec_incrementaldecoder", - (PyCFunction)codec_incrementaldecoder, METH_VARARGS}, + (PyCFunction)codec_incrementaldecoder, METH_VARARGS}, {"test_s_code", (PyCFunction)test_s_code, METH_NOARGS}, {"test_u_code", (PyCFunction)test_u_code, METH_NOARGS}, {"test_Z_code", (PyCFunction)test_Z_code, METH_NOARGS}, {"test_widechar", (PyCFunction)test_widechar, METH_NOARGS}, + {"test_aswidechar", test_aswidechar, METH_VARARGS}, + {"test_aswidecharstring", test_aswidecharstring, METH_VARARGS}, #ifdef WITH_THREAD - {"_test_thread_state", test_thread_state, METH_VARARGS}, + {"_test_thread_state", test_thread_state, METH_VARARGS}, {"_pending_threadfunc", pending_threadfunc, METH_VARARGS}, #endif #ifdef HAVE_GETTIMEOFDAY - {"profile_int", profile_int, METH_NOARGS}, + {"profile_int", profile_int, METH_NOARGS}, #endif - {"traceback_print", traceback_print, METH_VARARGS}, - {"exception_print", exception_print, METH_VARARGS}, - {"argparsing", argparsing, METH_VARARGS}, - {"code_newempty", code_newempty, METH_VARARGS}, + {"traceback_print", traceback_print, METH_VARARGS}, + {"exception_print", exception_print, METH_VARARGS}, + {"argparsing", argparsing, METH_VARARGS}, + {"code_newempty", code_newempty, METH_VARARGS}, {"make_exception_with_doc", (PyCFunction)make_exception_with_doc, METH_VARARGS | METH_KEYWORDS}, {"crash_no_current_thread", (PyCFunction)crash_no_current_thread, METH_NOARGS}, - {"format_unicode", format_unicode, METH_VARARGS}, + {"format_unicode", format_unicode, METH_VARARGS}, {NULL, NULL} /* sentinel */ }; diff --git a/Objects/unicodeobject.c b/Objects/unicodeobject.c index 29404c3..1c083b2 100644 --- a/Objects/unicodeobject.c +++ b/Objects/unicodeobject.c @@ -1216,7 +1216,7 @@ PyUnicode_AsWideCharString(PyUnicodeObject *unicode, } unicode_aswidechar(unicode, buffer, buflen); if (size) - *size = buflen; + *size = buflen - 1; return buffer; } |