[MAINT] Offload interfaces with help formatting

[oesteban]

This is the first on a series of PRs to clean up Nipype interfaces.

EDIT: also removed load_template from nipype.interfaces.base.support because it was deprecated since 1.0.0 and scheduled for removal in 1.1.0

[satra]

thanks for this - we can re-use this in the pydra refactor.

[codecov-io]

Codecov Report

Merging #2797 into master will decrease coverage by 3.46%.
The diff coverage is 86.73%.

@@            Coverage Diff             @@
##           master    #2797      +/-   ##
==========================================
- Coverage   67.51%   64.05%   -3.47%     
==========================================
  Files         341      339       -2     
  Lines       43297    43296       -1     
  Branches     5369     5374       +5     
==========================================
- Hits        29234    27732    -1502     
- Misses      13363    14495    +1132     
- Partials      700     1069     +369

Flag	Coverage Δ
#smoketests	?
#unittests	64.05% <86.73%> (-0.89%)	⬇️

Impacted Files	Coverage Δ
nipype/interfaces/base/__init__.py	100% <100%> (ø)	⬆️
nipype/scripts/utils.py	27.16% <50%> (+0.91%)	⬆️
nipype/utils/nipype_cmd.py	42.62% <50%> (+0.95%)	⬆️
nipype/interfaces/base/support.py	79.65% <88.15%> (+8.65%)	⬆️
nipype/interfaces/base/core.py	84.25% <88.23%> (-2.85%)	⬇️
nipype/interfaces/nilearn.py	40% <0%> (-56.67%)	⬇️
nipype/utils/spm_docs.py	25.92% <0%> (-44.45%)	⬇️
nipype/interfaces/freesurfer/base.py	50% <0%> (-30.51%)	⬇️
nipype/utils/logger.py	59.7% <0%> (-29.86%)	⬇️
nipype/algorithms/rapidart.py	35% <0%> (-29.61%)	⬇️
... and 57 more

---

Continue to review full report at Codecov.

Legend - Click here to learn more
Δ = absolute <relative> (impact), ø = not affected, ? = missing data
Powered by Codecov. Last update fc85fef...7861a4a. Read the comment docs.

[oesteban]

Anyone is currently/planning to reviewing this or is it good to go?

[effigies]

I haven't had a chance to look at any of these refactors. If they're inert, that's fine, but I'll admit I'm a little wary of a huge battery of refactors immediately before a release.

[oesteban]

Oh sure, let's hold on to after the release.

[effigies]

@oesteban Is this good to go? Realizing that 1.1.7 is just about here, and trying to decide what to push off to next month. (It's going to be most things, but if this is just a quick review, I can do it tomorrow.)

If the docs are produced from the help text, that will make checking the results of this easy. @satra Is that the case?

[satra]

docs are produced from the help text. (a side bug: our release version docs don't list the release version properly)

[satra]

this builds docs are here: https://1400-17021985-gh.circle-artifacts.com/0/home/circleci/work/docs/index.html

[oesteban]

This one is ready to go 👍

@satra, what you mean by our release version docs don't list the release version properly?

[satra]

@effigies - take a look at the release/devel versions on the side panel in these two links:

https://nipype.readthedocs.io/en/1.1.6/
https://nipype.readthedocs.io/en/latest/

[satra]

i meant @oesteban - sorry

[effigies]

@satra Sorry, missed that in the last release. Pushed a fix master, so we'll get it for a couple days, but no going back to fix it on the release tag.

[effigies]

@oesteban Compare:

Old

New

I agree with the move away from alphabetizing inputs/outputs, as those are usually already sorted by the interface author's notion of importance. (This was possibly accidental... I've suggested removing sorted in a couple more places.)

I think we should preserve the extra newline between mandatory and optional sections, and I think the extra indentation helps the input names pop out, so I've made suggestions I think will resolve these.

A further note, it looks like we have not been generating references correctly (see "References:: None" under *CompCor) but now we're not generating them at all. Fixing generation is probably a problem for another PR and thus another release, but I've at least suggested a fix I think will restore the indicator that there should be a reference.

[oesteban]

A further note, it looks like we have not been generating references correctly (see "References:: None" under *CompCor) but now we're not generating them at all. Fixing generation is probably a problem for another PR and thus another release, but I've at least suggested a fix I think will restore the indicator that there should be a reference.

Yep, I think I assumed we would go back to this issue at some point, and I didn't like having that "References:: None". But I'm fine either way.

[effigies]

Hmm. References are back, but not rendered.

https://1491-17021985-gh.circle-artifacts.com/0/home/circleci/work/docs/interfaces/generated/nipype.algorithms.confounds.html

[effigies]

But otherwise this looks good.