add ga4gh refget digest functionality

[nsheff]

Adds a new digests module that provides md5 and ga4gh digest computation for strings and fasta files.

I did not implement a CLI for it, just rust library functions and python bindings.

I confirmed the python bindings work for me locally.

[nsheff]

Do you know why I can't do this?

from gtars.digests import digest_fasta
Traceback (most recent call last):
  File "<stdin>", line 1, in <module>
ModuleNotFoundError: No module named 'gtars.digests'

When I can do this?

from gtars import digests
>>> digests
<module 'digests'>
>>> digests.digest_fasta()

[nleroy917]

Two things:

1. When I test, I make sure that I am in a different directory, otherwise python can be weird about imports (because there is a gtars folder already there so it tries that first...)
2. You might need to create a bindings/python/gtars/digests/__init__.py file with this inside:

from .gtars.digests import *

You can place a __init__.pyi file next to it to add type-annotations/docs for intellisense inside IDE's too. Check out the other folders for examples. Tokenizers is probably a good example.

The reason this __init__.py file is needed is not totally clear to me... but its what I got to work after looking at other examples of Rust + Python bindings for pyo3 with a module system and then also looking at the pyo3 discussion forum

[nsheff]

yeah I had figured that out, but even after adding that I still get the same issue...

[nsheff]

But now the error is different:

from gtars.digests import md5_digest
Traceback (most recent call last):
  File "<stdin>", line 1, in <module>
  File "/home/nsheff/.local/lib/python3.12/site-packages/gtars/digests/__init__.py", line 1, in <module>
    from .gtars.digests import *  # noqa: F403
    ^^^^^^^^^^^^^^^^^^^^^^^^^^^^
ModuleNotFoundError: No module named 'gtars.digests.gtars'

[nleroy917]

I think we are missing sys_modules.set_item("")... line 35 in bindings/python/src/lib.rs

    // set names of submodules
    sys_modules.set_item("gtars.tokenizers", m.getattr("tokenizers")?)?;
    sys_modules.set_item("gtars.ailist", m.getattr("ailist")?)?;
    sys_modules.set_item("gtars.utils", m.getattr("utils")?)?;
    sys_modules.set_item("gtars.models", m.getattr("models")?)?;
    // <--------    add your module here? 

[nsheff]

Yep that was what I was missing. it works now! thanks.

[nleroy917]

Do we want to merge this soon?

[donaldcampbelljr]

Great. Thanks for adding tests and docs, too.