[ENH] Add compressed TSV files to the common principles

[oesteban]

Their description is hedged within the physiological recordings so upcasting them to the common principles seems important.

[codecov]

Codecov Report

All modified and coverable lines are covered by tests ✅

Project coverage is 87.93%. Comparing base (bd08602) to head (cab6e9d).

Additional details and impacted files

@@           Coverage Diff           @@
##           master    #1749   +/-   ##
=======================================
  Coverage   87.93%   87.93%           
=======================================
  Files          16       16           
  Lines        1351     1351           
=======================================
  Hits         1188     1188           
  Misses        163      163           

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.

[Remi-Gau]

Slightly related to #644

[Remi-Gau]

I think that this reverts some of #1699, no?
Specifically the exception to the rule for tsv for motion data as motion.tsv are header less.

[oesteban]

It feels lines 46-48 of the diff should stay. The first parenthesis is addressed, but can be made more explicit either here or (better, IMHO) within motion. I can take care of harmonizing this.

…

On Mon, Mar 25, 2024, 13:05 Remi Gau ***@***.***> wrote: I think that this reverts some of #1699 <#1699>, no? Specifically the exception to the rule for tsv for motion data as motion.tsv are header less. — Reply to this email directly, view it on GitHub <#1749 (comment)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/AAESDRRFSSQD2DHQ4JZKKC3Y2AHIBAVCNFSM6AAAAABFGOEVG6VHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMZDAMJXHA2TINRRHE> . You are receiving this because you authored the thread.Message ID: ***@***.***>

[oesteban]

I think that this reverts some of #1699, no? Specifically the exception to the rule for tsv for motion data as motion.tsv are header less.

Okay, I'm realizing a problem -- motion's tsv files ARE NOT compressed, is that right?

I'll push a commit improving some of my changes not to undo the motion exception.

[Remi-Gau]

I think that this reverts some of #1699, no? Specifically the exception to the rule for tsv for motion data as motion.tsv are header less.

Okay, I'm realizing a problem -- motion's tsv files ARE NOT compressed, is that right?

I'll push a commit improving some of my changes not to undo the motion exception.

Yup. Sorry. Pre coffee me was not clear.
This is an "inconsistency" we realized only recently .

[oesteban]

Okay, compatibilization with #1699 is done (c4ee8d6).

Slightly related to #644

I doubled down with 4eadad5. I am happy to see that changed if #644 leads the table somewhere else later on.

[effigies]

Okay, reading through, I think this is good overall. Some specific proposals.

[sappelhoff]

I'll try to think how to make it more clear.

Thanks, I'll mark the parts that I find a bit ambiguous.

[oesteban]

@sappelhoff please have a look and maybe iterate over my suggestions to make the reading of the specs unequivocal regarding TSV vs TSVGZ

[oesteban]

@effigies @Remi-Gau @sappelhoff please check the changes to make unambiguous that .tsv and .tsv.gz are not interchangeable 48ffe9d

It feels too verbose to me, but I agree with @sappelhoff we need to avert any diverting readings of the specs. TBH, understanding that tsv.gz can be used instead of tsv at every instance felt a bit forced to me, but it's better to make the interpretation absolutely unambiguous.

[sappelhoff]

Thanks for the adjustments and the general proposal! I am happy to see this merged either with or without my proposed subheading change.

[effigies]

I think this is a reasonably compact update that eliminates the possible misreading.

[oesteban]

Considering that the minimum period for discussion has been safely satisfied, the very incremental nature of the proposal, the absence of major side effects on the specs, and the possibility of ironing around yet-to-be-identified wrinkles through the PR mechanism anytime in the future, I think I can go ahead and merge.

There are a few aspects to follow up from here:

• [FIX] Move _stim.<ext> specification within the Task Events module #1750 following upon this PR improving _stim.tsv.gz specs
• Make extensions identifiers contain . to separate extensions where multiple used #1755 normalizing extensions as terms in the glossary
• [ENH] BEP 020 Eye Tracking #1128 BEP020, where we are looking at physio specs with fresh eyes and trying to solidify BIDS foundations and consistency.