Nick Coghlan <[email protected]> wrote:
> 2. the current check is too strict (it should just check for obj !=
> NULL, not obj == &exporter)
Yes. For anyone who is interested, see issue #14181.
> 3. the current check is in the wrong place (it should be in
> PyObject_GetBuffer)
Agreed, since it's not memoryview specific. But I don't think we even
need to check for obj != NULL. view.obj was undocumented, and since 3.0
Include/object.h contains this:
typedef struct bufferinfo {
void *buf;
PyObject *obj; /* owned reference */
So it would be somewhat audacious to set this field to NULL. But even if
existing code uses the view.obj==NULL scheme from PyBuffer_FillInfo()
correctly, it will still work in the new implementation.
I'd just prefer to forbid this in the documentation, because it's much
easier to remember: getbuffer "returns" a new reference or NULL.
Stefan Krah
_______________________________________________
Python-Dev mailing list
[email protected]
http://mail.python.org/mailman/listinfo/python-dev
Unsubscribe:
http://mail.python.org/mailman/options/python-dev/archive%40mail-archive.com
