[MRG] Improve check of scans.tsv for BrainVision files

[teonbrooks]

PR Description

This PR closes #1033.

Merge checklist

Maintainer, please confirm the following before merging:

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

[agramfort]

would it be possible to add a test or is it too much to ask?

[teonbrooks]

would it be possible to add a test or is it too much to ask?

@agramfort, yep, i can do that. i renamed it to wip, and i'll write a test in the morning. noticed this at neurohackademy when working through an ieeg example

[sappelhoff]

looks like you have some Windows issue:

ValueError: 'eeg\\sub-01_ses-01_task-testing_acq-01_run-01_eeg.vhdr' is not in list

probably as_posix needs to be called on the Path somewhere, to normalize to / instead of \ on win.

[codecov]

Codecov Report

Merging #1034 (085c382) into main (16a7b0c) will decrease coverage by 0.01%.
The diff coverage is 100.00%.

❗ Current head 085c382 differs from pull request most recent head cdaec9a. Consider uploading reports for the commit cdaec9a to get more accurate results

@@            Coverage Diff             @@
##             main    #1034      +/-   ##
==========================================
- Coverage   95.19%   95.17%   -0.02%     
==========================================
  Files          25       25              
  Lines        3805     3811       +6     
==========================================
+ Hits         3622     3627       +5     
- Misses        183      184       +1     

Impacted Files	Coverage Δ
mne_bids/read.py	96.05% <100.00%> (+0.05%)	⬆️
mne_bids/write.py	97.04% <0.00%> (-0.11%)	⬇️

Help us with your feedback. Take ten seconds to tell us how you rate us. Have a feature suggestion? Share it here.

[teonbrooks]

looks like you have some Windows issue:

ValueError: 'eeg\\sub-01_ses-01_task-testing_acq-01_run-01_eeg.vhdr' is not in list

probably as_posix needs to be called on the Path somewhere, to normalize to / instead of \ on win.

so i'm looking into this and it seems like we read in the tsv file but we don't change filename from str to a filepath obj that would be insensitive to the os. should we be working with filepath all over the place?

also, if you see an easy fix, please let me know, i keep bumping into different windows errors on the testing front. not sure if i have any more time to debug it at the moment

[agramfort]

ok all good !

CI failure with doc is also in main branch

[agramfort]

thanks @teonbrooks !