MRG: Correctly handle reading when BIDS channel order differs from Raw

[hoechenberger]

Fixes #1170

@moritz-gerster does this fix your problem?

Merge checklist

Maintainer, please confirm the following before merging.
If applicable:

• All comments are resolved
• This is not your own PR
• All CIs are happy
• PR title starts with [MRG]
• whats_new.rst is updated
• New contributors have been added to CITATION.cff
• PR description includes phrase "closes <#issue-number>"

[codecov]

Codecov Report

Merging #1171 (3c9fbaf) into main (2de8536) will decrease coverage by 0.06%.
The diff coverage is 100.00%.

@@            Coverage Diff             @@
##             main    #1171      +/-   ##
==========================================
- Coverage   97.63%   97.57%   -0.06%     
==========================================
  Files          40       40              
  Lines        8655     8662       +7     
==========================================
+ Hits         8450     8452       +2     
- Misses        205      210       +5     

Files	Coverage Δ
mne_bids/read.py	95.87% <100.00%> (-0.02%)	⬇️
mne_bids/tests/test_read.py	99.71% <100.00%> (+<0.01%)	⬆️

... and 3 files with indirect coverage changes

[moritz-gerster]

@moritz-gerster does this fix your problem?

Yes, that works! :)

[sappelhoff]

I'm in favor, but we need to be aware that unfortunately, the channel order in channels.tsv only SHOULD (not MUST) correspond to the ordering in the raw file. Don't ask me why, I don't like it either.

So the change we are doing here may lead to breakage, but I think that's fine, as IMHO the channels.tsv file channel order really should should should correspond to that in the raw file.

[hoechenberger]

I believe a common use case is that you realize two EEG electrodes were accidentally swapped, hence you change their order in channels.tsv. With this PR branch, you don't have to worry about fixing the raw file too