ENH: Allow BIDS-style slice timings to be passed directly to TShift

[effigies]

This is kind of an upstream from fMRIPrep. It replaces a helper Function interface we use before TShift.

This refines the -tpattern inputs to permit both verifying file existence and directly passing a list of floats, which will create a slice timing file to pass as an argument.

Changes proposed in this pull request

• Create slice_timing input, mutually exclusive with tpattern, which accepts a file
  • If a list of floats is passed, write a file and pass its name to 3dTshift
• Change tpattern to an Enum, and list valid strings
  • A genericStr is kept as an option, to avoid breaking code that uses @<fname>, but warns user that they should use slice_timing

[codecov-io]

Codecov Report

Merging #2608 into master will increase coverage by 0.05%.
The diff coverage is 100%.

@@            Coverage Diff             @@
##           master    #2608      +/-   ##
==========================================
+ Coverage   67.59%   67.65%   +0.05%     
==========================================
  Files         340      340              
  Lines       43007    43030      +23     
  Branches     5321     5325       +4     
==========================================
+ Hits        29072    29111      +39     
+ Misses      13236    13218      -18     
- Partials      699      701       +2

Flag	Coverage Δ
#smoketests	50.52% <ø> (ø)	⬆️
#unittests	65.11% <100%> (+0.06%)	⬆️

Impacted Files	Coverage Δ
nipype/interfaces/afni/preprocess.py	82.01% <100%> (+0.5%)	⬆️
nipype/pipeline/plugins/legacymultiproc.py	61.3% <0%> (+5.35%)	⬆️
nipype/interfaces/afni/base.py	68.54% <0%> (+5.64%)	⬆️

---

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 7fa55ae...afecc94. Read the comment docs.

[effigies]

Don't have time to do the proper tests on fmriprep today, so punting for this release.

[effigies]

Just a note that I'm testing this on ds000030/sub-10159. Should be good to go today. Anybody up for a code review?

[effigies]

Patching this branch and nipreps/fmriprep#1160 into the poldracklab/fmriprep:1.1.2 Docker image produced slice-timing corrected files that differed only in timestamp.

[mgxd]

if we generate this file, perhaps we should include it in the interface's outputs

[effigies]

Seems reasonable.

[mgxd]

looks good to me - one minor suggestion