Patch #636005: Filter unicode into unicode.
This commit is contained in:
parent
fc03a94aac
commit
8afd7571a1
|
@ -365,6 +365,12 @@ class BuiltinTest(unittest.TestCase):
|
|||
def __getitem__(self, index):
|
||||
raise ValueError
|
||||
self.assertRaises(ValueError, filter, lambda x: x >="3", badstr("1234"))
|
||||
if have_unicode:
|
||||
# test biltinmodule.c::filterstring()
|
||||
self.assertEqual(filter(None, unicode("12")), unicode("12"))
|
||||
self.assertEqual(filter(lambda x: x>="3", unicode("1234")), unicode("34"))
|
||||
self.assertRaises(TypeError, filter, 42, unicode("12"))
|
||||
self.assertRaises(ValueError, filter, lambda x: x >="3", badstr(unicode("1234")))
|
||||
|
||||
def test_float(self):
|
||||
self.assertEqual(float(3.14), 3.14)
|
||||
|
|
|
@ -12,6 +12,8 @@ What's New in Python 2.3 alpha 2?
|
|||
Core and builtins
|
||||
-----------------
|
||||
|
||||
- filter returns now Unicode results for Unicode arguments.
|
||||
|
||||
- raw_input can now return Unicode objects.
|
||||
|
||||
- List objects' sort() method now accepts None as the comparison function.
|
||||
|
|
|
@ -24,6 +24,9 @@ const char *Py_FileSystemDefaultEncoding = NULL; /* use default */
|
|||
|
||||
/* Forward */
|
||||
static PyObject *filterstring(PyObject *, PyObject *);
|
||||
#ifdef Py_USING_UNICODE
|
||||
static PyObject *filterunicode(PyObject *, PyObject *);
|
||||
#endif
|
||||
static PyObject *filtertuple (PyObject *, PyObject *);
|
||||
|
||||
static PyObject *
|
||||
|
@ -132,6 +135,10 @@ builtin_filter(PyObject *self, PyObject *args)
|
|||
/* Strings and tuples return a result of the same type. */
|
||||
if (PyString_Check(seq))
|
||||
return filterstring(func, seq);
|
||||
#ifdef Py_USING_UNICODE
|
||||
if (PyUnicode_Check(seq))
|
||||
return filterunicode(func, seq);
|
||||
#endif
|
||||
if (PyTuple_Check(seq))
|
||||
return filtertuple(func, seq);
|
||||
|
||||
|
@ -1926,3 +1933,58 @@ Fail_1:
|
|||
Py_DECREF(result);
|
||||
return NULL;
|
||||
}
|
||||
|
||||
#ifdef Py_USING_UNICODE
|
||||
/* Helper for filter(): filter a Unicode object through a function */
|
||||
|
||||
static PyObject *
|
||||
filterunicode(PyObject *func, PyObject *strobj)
|
||||
{
|
||||
PyObject *result;
|
||||
register int i, j;
|
||||
int len = PyUnicode_GetSize(strobj);
|
||||
|
||||
if (func == Py_None) {
|
||||
/* No character is ever false -- share input string */
|
||||
Py_INCREF(strobj);
|
||||
return strobj;
|
||||
}
|
||||
if ((result = PyUnicode_FromUnicode(NULL, len)) == NULL)
|
||||
return NULL;
|
||||
|
||||
for (i = j = 0; i < len; ++i) {
|
||||
PyObject *item, *arg, *good;
|
||||
int ok;
|
||||
|
||||
item = (*strobj->ob_type->tp_as_sequence->sq_item)(strobj, i);
|
||||
if (item == NULL)
|
||||
goto Fail_1;
|
||||
arg = Py_BuildValue("(O)", item);
|
||||
if (arg == NULL) {
|
||||
Py_DECREF(item);
|
||||
goto Fail_1;
|
||||
}
|
||||
good = PyEval_CallObject(func, arg);
|
||||
Py_DECREF(arg);
|
||||
if (good == NULL) {
|
||||
Py_DECREF(item);
|
||||
goto Fail_1;
|
||||
}
|
||||
ok = PyObject_IsTrue(good);
|
||||
Py_DECREF(good);
|
||||
if (ok)
|
||||
PyUnicode_AS_UNICODE((PyStringObject *)result)[j++] =
|
||||
PyUnicode_AS_UNICODE((PyStringObject *)item)[0];
|
||||
Py_DECREF(item);
|
||||
}
|
||||
|
||||
if (j < len)
|
||||
PyUnicode_Resize(&result, j);
|
||||
|
||||
return result;
|
||||
|
||||
Fail_1:
|
||||
Py_DECREF(result);
|
||||
return NULL;
|
||||
}
|
||||
#endif
|
||||
|
|
Loading…
Reference in New Issue