Port over DLC conversion utils

[CodyCBakerPhD]

@h-mayorquin and I talked about this, and it's come up before in the past...

The DLC team has historically taken far too long to even respond to issues, let alone push fixes or review PRs on dlc2nwb

So we're just porting over the util functions here so we have full control over making minor bug fixes or improvements like the recent container name exposure (and no doubt others we'll find over time)

cc @vigji I'll absorb your fixes from DeepLabCut/DLC2NWB#23 and for rebase your NeuroConv PR to this one (assuming I can)

[vigji]

@h-mayorquin and I talked about this, and it's come up before in the past...

The DLC team has historically taken far too long to even respond to issues, let alone push fixes or review PRs on dlc2nwb

So we're just porting over the util functions here so we have full control over making minor bug fixes or improvements like the recent container name exposure (and no doubt others we'll find over time)

cc @vigji I'll absorb your fixes from DeepLabCut/DLC2NWB#23 and for rebase your NeuroConv PR to this one (assuming I can)

Sounds great, good for me ofc! Thanks!

[CodyCBakerPhD]

@h-mayorquin The live service failures should have nothing to do with this I'd think (some out of sync release on DANDI side)

[h-mayorquin]

OK, some questions. I will give it another read tomorrow but this seems like a simple copy-paste which should work. We can merge and then slowly improve as we need it.

[h-mayorquin]

Shouldn't this import be protected? I was expecting it to throw imports errors.

[CodyCBakerPhD]

Ah, sure

[CodyCBakerPhD]

Good catch

[CodyCBakerPhD]

But as to why it was not throwing error, it's because we don't test global importability of this private submodule and it's not triggered by anything global (the import of this submodule is itself lazy in the interface usage)

[h-mayorquin]

This is because of the protection with __all__ on the __init__ files?

[CodyCBakerPhD]

I didn't read much about the reasoning, only that it triggered on all the ones not contained in the __all__ of that file (which is most if not all of them)

[h-mayorquin]

OK, LGTM.

[codecov]

Codecov Report

Attention: Patch coverage is 59.28571% with 57 lines in your changes missing coverage. Please review.

Project coverage is 90.14%. Comparing base (0866c30) to head (d3ea2a1).
Report is 118 commits behind head on main.

Files with missing lines	Patch %	Lines
...v/datainterfaces/behavior/deeplabcut/_dlc_utils.py	58.08%	57 Missing ⚠️

Additional details and impacted files

@@            Coverage Diff             @@
##             main     #946      +/-   ##
==========================================
- Coverage   90.76%   90.14%   -0.63%     
==========================================
  Files         125      126       +1     
  Lines        7085     7214     +129     
==========================================
+ Hits         6431     6503      +72     
- Misses        654      711      +57     

Flag	Coverage Δ
unittests	90.14% <59.28%> (-0.63%)	⬇️

Flags with carried forward coverage won't be shown. Click here to find out more.

Files with missing lines	Coverage Δ
...ces/behavior/deeplabcut/deeplabcutdatainterface.py	92.50% <100.00%> (-1.12%)	⬇️
...v/datainterfaces/behavior/deeplabcut/_dlc_utils.py	58.08% <58.08%> (ø)