BUG : fix eps corruption when using clipping#3526
Conversation
tacaswell
added
the
Release critical
_get_clip_path was returning '<built-in function id>' instead of the ps-function name like it should have been. This was introduced in 9b9c0c6 in PR matplotlib#2927. Closes matplotlib#3523
tacaswell
force-pushed
the
fix_eps_clip_paths
branch
from
660dc47 to
3424506
Compare
|
If you don't mind, I would like to add @bryan-woods to this discussion. I have been meaning to show him how a PR works, and this is a good way to see how it works. |
There was a problem hiding this comment.
didn't the PEP8 tool notice this? I know pylint would have warned about namespace collisions. Then again, I guess it isn't really a collision as much as a typo.
|
Also, why didn't any of our tests see this? What was it about basemap that triggered this? |
|
Looking back at that referenced PR, I see that the original code did use "id" and it was changed to "pid" everywhere except in the return. Lovely... |
BUG : fix eps corruption when using clipping
|
Tests didn't catch it because we don't test the ps backend and apparently no one ever tried to open the file that was generating the issue in the other PR. pep8 does not do any linting, just white space. |
3 participants
_get_clip_path was returning '' instead
of the ps-function name like it should have been. This was introduced
in 9b9c0c6 in PR #2927.
Closes #3523