Patch #1071: Improve unicode.translate() so that you can pass unicode
characters as mapping keys and invalid mapping keys are recognized
and raise an error.
diff --git a/Doc/library/stdtypes.rst b/Doc/library/stdtypes.rst
index 61b589f..06eb1cf 100644
--- a/Doc/library/stdtypes.rst
+++ b/Doc/library/stdtypes.rst
@@ -931,7 +931,7 @@
Return a copy of the *s* where all characters have been mapped through the
*map* which must be a dictionary of characters (strings of length 1) or
Unicode ordinals (integers) to Unicode ordinals, strings or ``None``.
- Unmapped characters are left untouched. Characters mapped to ``None`` are
+ Unmapped characters are left untouched. Characters mapped to ``None`` are
deleted.
.. note::
diff --git a/Lib/test/test_unicode.py b/Lib/test/test_unicode.py
index 04dddaa..9aad59a 100644
--- a/Lib/test/test_unicode.py
+++ b/Lib/test/test_unicode.py
@@ -160,12 +160,14 @@
self.checkequalnofix('bbbc', 'abababc', 'translate', {ord('a'):None})
self.checkequalnofix('iiic', 'abababc', 'translate', {ord('a'):None, ord('b'):ord('i')})
self.checkequalnofix('iiix', 'abababc', 'translate', {ord('a'):None, ord('b'):ord('i'), ord('c'):'x'})
- self.checkequalnofix('<i><i><i>c', 'abababc', 'translate', {ord('a'):None, ord('b'):'<i>'})
+ self.checkequalnofix('<i><i><i>c', 'abababc', 'translate', {'a':None, 'b':'<i>'})
self.checkequalnofix('c', 'abababc', 'translate', {ord('a'):None, ord('b'):''})
self.checkequalnofix('xyyx', 'xzx', 'translate', {ord('z'):'yy'})
self.assertRaises(TypeError, 'hello'.translate)
self.assertRaises(TypeError, 'abababc'.translate, 'abc', 'xyz')
+ self.assertRaises(ValueError, 'abababc'.translate, {'xy':2})
+ self.assertRaises(TypeError, 'abababc'.translate, {(1,):2})
def test_split(self):
string_tests.CommonTest.test_split(self)
diff --git a/Objects/unicodeobject.c b/Objects/unicodeobject.c
index 13644b0..61a2320 100644
--- a/Objects/unicodeobject.c
+++ b/Objects/unicodeobject.c
@@ -7810,10 +7810,54 @@
static PyObject*
unicode_translate(PyUnicodeObject *self, PyObject *table)
{
- return PyUnicode_TranslateCharmap(self->str,
- self->length,
- table,
- "ignore");
+ PyObject *newtable = NULL;
+ Py_ssize_t i = 0;
+ PyObject *key, *value, *result;
+
+ if (!PyDict_Check(table)) {
+ PyErr_SetString(PyExc_TypeError, "translate argument must be a dict");
+ return NULL;
+ }
+ /* fixup the table -- allow size-1 string keys instead of only int keys */
+ newtable = PyDict_Copy(table);
+ if (!newtable) return NULL;
+ while (PyDict_Next(table, &i, &key, &value)) {
+ if (PyUnicode_Check(key)) {
+ /* convert string keys to integer keys */
+ PyObject *newkey;
+ int res;
+ if (PyUnicode_GET_SIZE(key) != 1) {
+ PyErr_SetString(PyExc_ValueError, "string items in translate "
+ "table must be 1 element long");
+ goto err;
+ }
+ newkey = PyInt_FromLong(PyUnicode_AS_UNICODE(key)[0]);
+ if (!newkey)
+ goto err;
+ res = PyDict_SetItem(newtable, newkey, value);
+ Py_DECREF(newkey);
+ if (res < 0)
+ goto err;
+ } else if (PyInt_Check(key)) {
+ /* just keep integer keys */
+ if (PyDict_SetItem(newtable, key, value) < 0)
+ goto err;
+ } else {
+ PyErr_SetString(PyExc_TypeError, "items in translate table must be "
+ "strings or integers");
+ goto err;
+ }
+ }
+
+ result = PyUnicode_TranslateCharmap(self->str,
+ self->length,
+ newtable,
+ "ignore");
+ Py_DECREF(newtable);
+ return result;
+ err:
+ Py_DECREF(newtable);
+ return NULL;
}
PyDoc_STRVAR(upper__doc__,