FIX: skip minification tests on M1/M2 macs and run them otherwise

[kaczmarj]

This adds back in the minification tests to CI. These tests passed on my Debian Linux machine (intel 64-bit). See docker/for-mac#5191 for an explanation of why ptrace does not work on arm macs.

[codecov]

Codecov Report

Base: 83.67% // Head: 88.60% // Increases project coverage by +4.93% 🎉

Coverage data is based on head (927b960) compared to base (95f8f91).
Patch coverage: 100.00% of modified lines in pull request are covered.

Additional details and impacted files

@@            Coverage Diff             @@
##           master     #497      +/-   ##
==========================================
+ Coverage   83.67%   88.60%   +4.93%     
==========================================
  Files          11       11              
  Lines        1017     1027      +10     
==========================================
+ Hits          851      910      +59     
+ Misses        166      117      -49     

Impacted Files	Coverage Δ
neurodocker/cli/cli.py	90.00% <100.00%> (-10.00%)	⬇️
neurodocker/cli/minify/trace.py	83.90% <0.00%> (+58.62%)	⬆️

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

☔ View full report at Codecov.
📢 Do you have feedback about the report comment? Let us know in this issue.

[kaczmarj]

@satra - is this change ok with you? i use the following code to test if we are running on an m1/m2 mac:

import platform

def _arm_on_mac():
    """Return True if on an ARM processor (M1/M2) in macos operating system."""
    is_mac = platform.system().lower() == "darwin"
    is_arm = platform.processor().lower() == "arm"
    return is_mac and is_arm

[satra]

this PR should technically update, the CLI as well. see /~https://github.com/ReproNim/neurodocker/blob/master/neurodocker/cli/cli.py#L13 and also later in that file. and perhaps it should use the same function to determine if it's available.

also we should consider checking if trace can be carried out inside a singularity instance with a mounted mamba+reprozip. (not for this PR).

[kaczmarj]

this PR should technically update, the CLI as well. see /~https://github.com/ReproNim/neurodocker/blob/master/neurodocker/cli/cli.py#L13 and also later in that file. and perhaps it should use the same function to determine if it's available.

i added that, so neurodocker minify is not available on arm-based macs.

[kaczmarj]

also we should consider checking if trace can be carried out inside a singularity instance with a mounted mamba+reprozip. (not for this PR).

interesting idea. how would we save the minified singularity container? perhaps we could copy the minified directories into some tarball, and then inject them into a new container?

[satra]

how would we save the minified singularity container? perhaps we could copy the minified directories into some tarball, and then inject them into a new container?

yup, something like that.

[kaczmarj]

@satra - ready to merge on my end. let me know if you want any other changes.

[satra]

lgtm. you can decide which PR merge should trigger a release: add release label and minor or patch depending on the type of release.

if you are not planning to merge anymore PRs yet i would suggest labeling this one with release + patch and merging.