Conversation
…s dropped) Co-authored-by: Andrew Murray <radarhere@users.noreply.github.com>
f4729e4 to
de31565
Compare
| assert im_pa.get_flattened_data() == ((0, 240), (1, 220)) # Matches the alpha band | ||
|
|
||
| im_p = im_pa.convert("P") | ||
| assert im_p.im.getpalettemode() == "RGB" |
There was a problem hiding this comment.
This drops the check for im_p.palette.mode?
| im_p = im_pa.convert("P") | ||
| assert im_p.im.getpalettemode() == "RGB" | ||
| assert im_p.im.getpalette("RGB") == bytes([255, 0, 0, 0, 255, 0]) | ||
|
|
There was a problem hiding this comment.
I'd still like the whole rgba_data test too 🤔 It tests that converting the palette-no-alpha image gets converted to solid colors.
|
We appear to have a difference of opinion here. You would like to test that the palette doesn't affect pixel data, the Python palette mode and what happens after you convert the image to RGBA. I consider these to all be more than the unit test needs to address, and that the fix is verified without them. It's not a significant point of disagreement, so I don't mind if someone else merges the PR. I just hope you understand that my review is going to stall here, as I would rather add not in code that I don't think needs to be there. |
|
@radarhere I appreciate the comment. My idea is that this is also a regression test, since Pillow had a bug (the two palette representations being discrepant). The RGBA end result is what matters for display in modern use cases, which is why I'd like to also assert on it. |
2 participants
Suggestion for python-pillow#9834