Don't fail on equal-but-differently-named cmaps in qt figureoptions. #29148
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
Currently, opening the Qt figureoptions UI for an image whose cmap is not registered in the colormap registry, has a name not matching any registry entry, but is actually equal (
==
, i.e. has the same LUT and colorbar-extension attributes) to a registry entry, leads to an error. A typical example would beand opening the qt figure options; this leads to the error "index 'bids:magma' is invalid ...".
Note that if the cmap is different from any registered cmap then we already add it to the UI combobox (try e.g.
cmap=cmap.Colormap("imagej:fire")
); the only problem was if it was equal to a registered cmap (this arises because when the code was originally written, cmap instance equality was by identity, not by comparing LUTs, so thecmap not in cm._colormaps.values()
check behaved differently).Fix that by checking whether the colormap name is registered. The behavior is still ill-defined in the opposite (theoretical) case of an unregistered cmap different from any registered cmap but with a matching name, but I'd argue that case is more pathological.
Test by running the above code and opening the qt figureoptions.
(aka how some code I wrote 9 years ago (#5469, a bit scary it's been that long) got broken by some code written 5.5 years later (#20227), and how to fix that with a one-line patch with 20 lines of explanation. I guess that's technically a regression :-))
PR summary
PR checklist