PyUnicode_DecodeFSDefaultAndSize() uses surrogateescape error handler
This function is only used to decode Python module filenames, but Python
doesn't support surrogates in modules filenames yet. So nobody noticed this
minor bug.
diff --git a/Objects/unicodeobject.c b/Objects/unicodeobject.c
index 369306e..23b322f 100644
--- a/Objects/unicodeobject.c
+++ b/Objects/unicodeobject.c
@@ -1600,19 +1600,19 @@
if (Py_FileSystemDefaultEncoding) {
#if defined(MS_WINDOWS) && defined(HAVE_USABLE_WCHAR_T)
if (strcmp(Py_FileSystemDefaultEncoding, "mbcs") == 0) {
- return PyUnicode_DecodeMBCS(s, size, "replace");
+ return PyUnicode_DecodeMBCS(s, size, "surrogateescape");
}
#elif defined(__APPLE__)
if (strcmp(Py_FileSystemDefaultEncoding, "utf-8") == 0) {
- return PyUnicode_DecodeUTF8(s, size, "replace");
+ return PyUnicode_DecodeUTF8(s, size, "surrogateescape");
}
#endif
return PyUnicode_Decode(s, size,
Py_FileSystemDefaultEncoding,
- "replace");
+ "surrogateescape");
}
else {
- return PyUnicode_DecodeUTF8(s, size, "replace");
+ return PyUnicode_DecodeUTF8(s, size, "surrogateescape");
}
}