feat: add a modules mulled command

[Midnighter]

Add a command nf-core modules mulled that can generate image names from given tool specifications. See the discussion at https://nfcore.slack.com/archives/C01LNCGJBMH/p1649431242727599

PR checklist

• This comment contains a description of changes (with reason)
• CHANGELOG.md is updated
• If you've fixed a bug or added code that should be tested, add tests!
• Documentation in docs is updated

[codecov]

Codecov Report

Merging #1519 (893fc9a) into dev (faf4716) will increase coverage by 0.02%.
The diff coverage is 78.57%.

@@            Coverage Diff             @@
##              dev    #1519      +/-   ##
==========================================
+ Coverage   64.91%   64.94%   +0.02%     
==========================================
  Files          52       53       +1     
  Lines        6063     6105      +42     
==========================================
+ Hits         3936     3965      +29     
- Misses       2127     2140      +13     

Impacted Files	Coverage Δ
nf_core/__main__.py	51.40% <40.00%> (-0.28%)	⬇️
nf_core/modules/mulled.py	90.32% <90.32%> (ø)
nf_core/modules/__init__.py	100.00% <100.00%> (ø)
nf_core/modules/create.py	79.27% <0.00%> (-1.04%)	⬇️
nf_core/utils.py	83.00% <0.00%> (-0.50%)	⬇️

---

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 faf4716...893fc9a. Read the comment docs.

[Midnighter]

I don't think that the failing pipeline lint test is my fault.

[JoseEspinosa]

Very nice @Midnighter!
My only concern is that the hash of the mulled container is generated even if the container is not yet available. I am not aware that this could be checked using the Biocontainers API so maybe we could make a get request to the singularity registry and if the status code response is not 200 then print a warning? Or maybe we can just print always a warning that the hash is generated even if the image is not yet available?

[Midnighter]

My only concern is that the hash of the mulled container is generated even if the container is not yet available. I am not aware that this could be checked using the Biocontainers API so maybe we could make a get request to the singularity registry and if the status code response is not 200 then print a warning? Or maybe we can just print always a warning that the hash is generated even if the image is not yet available?

That's a fair point. I've been struggling with this as well. My solution would have been to write in the docs that you should use the mulled-search to find an image but if you have an entry in the hash.tsv, you can use this tool.

[JoseEspinosa]

Fair enough 😃 let's see what others think

[Midnighter]

Please take another look. I'm now polling quay.io for the existence of the image.

[JoseEspinosa]

Just made a small suggestion, it looks great to me now! Thanks @Midnighter 🚀

[jfy133]

Late to the game (sorry have a big backlog) but I think this would be really useful as I had this issue exactly yesterday trying to find something.

Only comments would be:

1. I would expect someone would argue that this is not a nf-core specific tool, more towards biocontainers, so someone may ask to make it standalone
2. I would say mulled as a subcommand alone is rather uninformative. Make biocontainers-search-mulled or something would be better (also I would say making the name longer would make it sort of clearer it's less nf-core specific)

[Midnighter]

1. Integrating it into the tools was an explicit wish by @ewels and @JoseEspinosa, I don't care myself.
2. I don't have strong feelings regarding the command name, however, I would not name it search because it doesn't search anything. It generates the name by hashing the inputs and then verifies that this name exists.

[JoseEspinosa]

Here is the link to the discussion we had for completeness: https://nfcore.slack.com/archives/C01LNCGJBMH/p1649431242727599

1. Maybe we should check what others think?
2. I don't have any strong opinion, I know my limitations and naming commands is not one of my strengths 😛 But I agree that search is misleading since in fact, the script creates the hash even if the mulled container has not yet been created.

[ewels]

I'm happy enough with this, it's a super obscure name but also a pretty obscure function...

This PR is missing any contributions to the docs though! Needs adding to README.md at least, with some long-form discussion about what it is and how to use it.

Ideally that would also be complemented with some docs on the website / in the modules readme if possible.