Ofer Schwarz added the comment:
I'm working on this now (patch hopefully coming soon), and I've hit an
interesting issue:
PySequence methods take care of negative indexes by adding +len before calling
the subclass method (s.t. -1 becomes len-1 etc.). This means that if the
subclass raises
Changes by Ofer Schwarz os.uran...@gmail.com:
--
keywords: +patch
Added file: http://bugs.python.org/file39055/indexerror.patch
___
Python tracker rep...@bugs.python.org
http://bugs.python.org/issue18162
Changes by Ofer Schwarz os.uran...@gmail.com:
Added file: http://bugs.python.org/file39056/indexerror.tests
___
Python tracker rep...@bugs.python.org
http://bugs.python.org/issue18162
Changes by Ofer Schwarz os.uran...@gmail.com:
Added file: http://bugs.python.org/file39060/indexerror_all.patch
___
Python tracker rep...@bugs.python.org
http://bugs.python.org/issue18162
Ofer Schwarz added the comment:
There's actually no advantage for a weakref here, since the exception object
already holds an indirect reference to all the raising frame's locals through
__traceback__.
--
nosy: +blackfawn
___
Python tracker rep
Ofer Schwarz added the comment:
So, this is actually impossible to do.
Since exceptions are c-level classes, adding members to ValueError means
anything that multiple-inherits from ValueError and any other complex exception
gets a multiple bases layout conflict. There's already one such class
New submission from Ofer Schwarz:
When creating kwargs to construct the ImportError, if PyDict_SetItemString
fails the function returns without decref'ing the already-created locals.
--
components: Interpreter Core
files: importerrorcleanup.patch
keywords: patch
messages: 241052
nosy
Ofer Schwarz added the comment:
I couldn't reproduce this on my machine, but I have no idea if I'm using the
wrong sequence of operations, or type of pool, or if my machine is just faster
or something else entirely.
Johannes, can you upload a repro code for this?
--
nosy: +blackfawn
