Feature: optional loading of an EPOCH input deck

[Dartrisen]

Now you can get access to the input.deck file by simply doing this:

import sdf_xarray as sdf
import xarray as xr
import epydeck

TEST_FILES_DIR = sdf.download.fetch_dataset("test_files_1D")
ds = xr.open_dataset(TEST_FILES_DIR / "0000.sdf", load_deck=True)
ds.attrs["deck"]

The package epydeck should be installed in your virtual environment to use this. The default value of load_deck is False.

Resolves #91

[JoelLucaAdams]

Hi @Dartrisen, thank you very much for implementing the new feature!

This is a great first start although it would be nice to also allow for this parameter would accept bool | str | Path with the default being False. If the user passes in a str | Path then it would first check if the path specified is a full file path using Path.is_absolute(), otherwise if the path is relative then look for it in the same directory as the SDF file being opened.

On top of adding this feature could you also add some additional documentation to highlight how this works? Probably a small section after this https://sdf-xarray.readthedocs.io/en/latest/key_functionality.html#loading-specific-variables or potentially a new page if you prefer.

[JoelLucaAdams]

FYI, you might want to additionally pull in branch #89 so that it works with datatrees as well

[JoelLucaAdams]

Thank you for the feature implementation! There are a few things that need ironing out that I have listed below.

On top of those changes you also need to add some tests please

[JoelLucaAdams]

Looks great now, nice job!

[LiamPattinson]

Hi, this pretty much all looks good to me. I just have a few comments, but two of them would require a fairly major overhaul of the branch, so it would be understandable if you would rather merge now and consider coming back to it later. Setting epydeck as an optional dependency would be pretty easy though.

[JoelLucaAdams]

There are lots of duplicate tests in here, for now we can ignore this problem as it will be resolved when someone tackles #100

[LiamPattinson]

Nice, this is looking a lot better! Just a few minor issues from here, one of which was an existing issue in the code that may as well be fixed while we're at it.

[JoelLucaAdams]

@Dartrisen can you also add your info to the authors section in the pyproject.toml and to the end of the CITATION.cff file please?

[Dartrisen]

are we good to merge? @LiamPattinson @JoelLucaAdams

[LiamPattinson]

Looks good to me, sorry that we got a bit lost in the minutiae there!