Mercurial > cpython-lru_cache-weakref
diff Objects/unicodeobject.c @ 64346:5a3101f558ed
Issue #7415: PyUnicode_FromEncodedObject() now uses the new buffer API
properly. Patch by Stefan Behnel.
| author | Antoine Pitrou <solipsis@pitrou.net> |
|---|---|
| date | Wed, 01 Sep 2010 15:10:12 +0000 |
| parents | 867699c9b63f |
| children | 944232521624 |
line wrap: on
line diff
--- a/Objects/unicodeobject.c Wed Sep 01 12:58:21 2010 +0000 +++ b/Objects/unicodeobject.c Wed Sep 01 15:10:12 2010 +0000 @@ -1234,8 +1234,7 @@ const char *encoding, const char *errors) { - const char *s = NULL; - Py_ssize_t len; + Py_buffer buffer; PyObject *v; if (obj == NULL) { @@ -1243,44 +1242,44 @@ return NULL; } + /* Decoding bytes objects is the most common case and should be fast */ + if (PyBytes_Check(obj)) { + if (PyBytes_GET_SIZE(obj) == 0) { + Py_INCREF(unicode_empty); + v = (PyObject *) unicode_empty; + } + else { + v = PyUnicode_Decode( + PyBytes_AS_STRING(obj), PyBytes_GET_SIZE(obj), + encoding, errors); + } + return v; + } + if (PyUnicode_Check(obj)) { PyErr_SetString(PyExc_TypeError, "decoding str is not supported"); return NULL; } - /* Coerce object */ - if (PyBytes_Check(obj)) { - s = PyBytes_AS_STRING(obj); - len = PyBytes_GET_SIZE(obj); - } - else if (PyByteArray_Check(obj)) { - s = PyByteArray_AS_STRING(obj); - len = PyByteArray_GET_SIZE(obj); - } - else if (PyObject_AsCharBuffer(obj, &s, &len)) { - /* Overwrite the error message with something more useful in - case of a TypeError. */ - if (PyErr_ExceptionMatches(PyExc_TypeError)) - PyErr_Format(PyExc_TypeError, - "coercing to str: need bytes, bytearray or char buffer, " - "%.80s found", - Py_TYPE(obj)->tp_name); - goto onError; - } - - /* Convert to Unicode */ - if (len == 0) { + /* Retrieve a bytes buffer view through the PEP 3118 buffer interface */ + if (PyObject_GetBuffer(obj, &buffer, PyBUF_SIMPLE) < 0) { + PyErr_Format(PyExc_TypeError, + "coercing to str: need bytes, bytearray " + "or buffer-like object, %.80s found", + Py_TYPE(obj)->tp_name); + return NULL; + } + + if (buffer.len == 0) { Py_INCREF(unicode_empty); - v = (PyObject *)unicode_empty; + v = (PyObject *) unicode_empty; } else - v = PyUnicode_Decode(s, len, encoding, errors); - + v = PyUnicode_Decode((char*) buffer.buf, buffer.len, encoding, errors); + + PyBuffer_Release(&buffer); return v; - - onError: - return NULL; } /* Convert encoding to lower case and replace '_' with '-' in order to
