✨ Track artifacts as inputs

[lazappi]

Related to: Fixes #98

Description

• Add Python lamindb to all instances (if available)
  • Still warn for the default instance if not
• Use Python in Artifact$cache() (if available)
  • Fall back to the API if not with a warning

Checklist

Before review

• Update and regenerate man pages
• Add/update tests
• Add/update examples in vignettes
• Pass CI checks

Before merge

• Update architecture vignette
• Update development vignette
• Update features in README
• Update CHANGELOG

[lazappi]

Example of doing this in an R script https://lamin.ai/laminlabs/lamindata/transform/Q5qWerxOc1R30000/lOBfMDTlIfZMeT4zvsza

@falexwolf I noticed the website says the code is Python and some of the formatting in the code is messed up (mostly <- gets converted to <-)

[falexwolf]

@falexwolf I noticed the website says the code is Python and some of the formatting in the code is messed up (mostly <- gets converted to <-)

This is solved in staging. Will go to prod later today.

[falexwolf]

Amazing that this seems to work already! 🤩

Looking through the code here now.

[falexwolf]

Is there every a case in which the Python package is not available? We already needed it for lamin connect etc. -- so I think it's always going to be available?

[rcannood]

Not necessarily -- if the user provides a correct current_user.env they could still connect to a lamindb instance without the python client. they just won't be able to track, create or delete records at this stage.

[lazappi]

I guess you need to CLI installed somewhere to do that but it doesn't necessarily mean it's an environment that {reticulate} can find.

[falexwolf]

Ok!

[falexwolf]

Great, @lazappi!

Does artifact$load() call artifact$cache() under-the-hood? If so, this is good to merge from my side.

[rcannood]

LGTM!

[lazappi]

Does artifact$load() call artifact$cache() under-the-hood? If so, this is good to merge from my side.

Yep. $cache() is used to get the file path and then we read it on the R side. That's easier than trying to read it in Python and convert to R.