Bug #1704793: Return UTF-16 pair if unicodedata.lookup cannot
represent the result in a single character.
diff --git a/Lib/test/test_unicodedata.py b/Lib/test/test_unicodedata.py
index 0023bf4..574178d 100644
--- a/Lib/test/test_unicodedata.py
+++ b/Lib/test/test_unicodedata.py
@@ -214,6 +214,9 @@
count += 1
self.assert_(count >= 10) # should have tested at least the ASCII digits
+ def test_bug_1704793(self):
+ self.assertEquals(self.db.lookup("GOTHIC LETTER FAIHU"), u'\U00010346')
+
def test_main():
test.test_support.run_unittest(
UnicodeMiscTest,
diff --git a/Misc/NEWS b/Misc/NEWS
index 4764daf..55ebfa5 100644
--- a/Misc/NEWS
+++ b/Misc/NEWS
@@ -238,6 +238,9 @@
Library
-------
+- Bug #1704793: Return UTF-16 pair if unicodedata.lookup cannot
+ represent the result in a single character.
+
- Bug #978833: Close https sockets by releasing the _ssl object.
- Change location of the package index to pypi.python.org/pypi
diff --git a/Modules/unicodedata.c b/Modules/unicodedata.c
index fac9adc..a075693 100644
--- a/Modules/unicodedata.c
+++ b/Modules/unicodedata.c
@@ -1077,8 +1077,7 @@
unicodedata_lookup(PyObject* self, PyObject* args)
{
Py_UCS4 code;
- Py_UNICODE str[1];
- char errbuf[256];
+ Py_UNICODE str[2];
char* name;
int namelen;
@@ -1086,24 +1085,20 @@
return NULL;
if (!_getcode(self, name, namelen, &code)) {
- /* XXX(nnorwitz): why are we allocating for the error msg?
- Why not always use snprintf? */
- char fmt[] = "undefined character name '%s'";
- char *buf = PyMem_MALLOC(sizeof(fmt) + namelen);
- if (buf)
- sprintf(buf, fmt, name);
- else {
- buf = errbuf;
- PyOS_snprintf(buf, sizeof(errbuf), fmt, name);
- }
- PyErr_SetString(PyExc_KeyError, buf);
- if (buf != errbuf)
- PyMem_FREE(buf);
+ PyErr_Format(PyExc_KeyError, "undefined character name '%s'",
+ name);
return NULL;
}
+#ifndef Py_UNICODE_WIDE
+ if (code >= 0x10000) {
+ str[0] = 0xd800 + ((code - 0x10000) >> 10);
+ str[1] = 0xdc00 + ((code - 0x10000) & 0x3ff);
+ return PyUnicode_FromUnicode(str, 2);
+ }
+#endif
str[0] = (Py_UNICODE) code;
- return PyUnicode_FromUnicode(str, 1);
+ return PyUnicode_FromUnicode(str, 1);
}
/* XXX Add doc strings. */