diff options
| author | Victor Stinner <victor.stinner@gmail.com> | 2015-10-01 21:54:51 +0200 |
|---|---|---|
| committer | Victor Stinner <victor.stinner@gmail.com> | 2015-10-01 21:54:51 +0200 |
| commit | 01ada3996bded57d1baf9c54b050fc55907d9b13 (patch) | |
| tree | 5f795a91278beb5ae831a75eef3f4b20c5c3a6f8 /Objects/unicodeobject.c | |
| parent | 29a1445136c7353543b516a085c38b8be9ce5109 (diff) | |
| download | cpython-git-01ada3996bded57d1baf9c54b050fc55907d9b13.tar.gz | |
Issue #25267: The UTF-8 encoder is now up to 75 times as fast for error
handlers: ``ignore``, ``replace``, ``surrogateescape``, ``surrogatepass``.
Patch co-written with Serhiy Storchaka.
Diffstat (limited to 'Objects/unicodeobject.c')
| -rw-r--r-- | Objects/unicodeobject.c | 7 |
1 files changed, 5 insertions, 2 deletions
diff --git a/Objects/unicodeobject.c b/Objects/unicodeobject.c index 6657cd4310..df3a1b5069 100644 --- a/Objects/unicodeobject.c +++ b/Objects/unicodeobject.c @@ -297,6 +297,7 @@ typedef enum { _Py_ERROR_UNKNOWN=0, _Py_ERROR_STRICT, _Py_ERROR_SURROGATEESCAPE, + _Py_ERROR_SURROGATEPASS, _Py_ERROR_REPLACE, _Py_ERROR_IGNORE, _Py_ERROR_XMLCHARREFREPLACE, @@ -312,6 +313,8 @@ get_error_handler(const char *errors) return _Py_ERROR_STRICT; if (strcmp(errors, "surrogateescape") == 0) return _Py_ERROR_SURROGATEESCAPE; + if (strcmp(errors, "surrogatepass") == 0) + return _Py_ERROR_SURROGATEPASS; if (strcmp(errors, "ignore") == 0) return _Py_ERROR_IGNORE; if (strcmp(errors, "replace") == 0) @@ -6479,8 +6482,8 @@ unicode_encode_ucs1(PyObject *unicode, goto onError; case _Py_ERROR_REPLACE: - while (collstart++ < collend) - *str++ = '?'; + memset(str, '?', collend - collstart); + str += (collend - collstart); /* fall through ignore error handler */ case _Py_ERROR_IGNORE: pos = collend; |