Issue 473: change how output directory is generated for extract_xri script

[rminsil]

This PR addresses one point raised in the code review feedback from Damien from this comment:

#491 (review)

I would make the output directory an optional command-line argument. By default, you should extract files to SIL_NLP_ENV.mt_corpora_dir. This should point to the MT/corpora folder in the S3 bucket, which is where we would store this kind of corpus.

I wasn't clear on the directory structure to use within the corpora dir, so I kept the existing logic of generating a unique folder based on the cli inputs + timestamp.

Note that this PR is built off the back of #506. When that one is merged I'll rebase this one and have it target master.

Once this PR is merged, all code review comments from #491 are addressed and balance is returned to the force.

---

This change is 

[ddaspit]

Reviewed 1 of 1 files at r1, all commit messages.
Reviewable status: all files reviewed, 1 unresolved discussion (waiting on @rminsil)

---

silnlp/common/extract_xri.py line 116 at r1 (raw file):

    if cli_input.output is None:
        unique_dir = f"{cli_input.source_iso}-{cli_input.target_iso}-{cli_input.dataset_descriptor}-{time.strftime('%Y%m%d-%H%M%S')}"
        output_dir = Path(os.path.join(SIL_NLP_ENV.mt_corpora_dir, unique_dir))

SIL_NLP_ENV.mt_corpora_dir is already a Path, so you should be able to do:

output_dir = SIL_NLP_ENV.mt_corpora_dir / unique_dir

[rminsil]

SIL_NLP_ENV.mt_corpora_dir is already a Path, so you should be able to do:

output_dir = SIL_NLP_ENV.mt_corpora_dir / unique_dir

Thanks for the tip @ddaspit, there's some newfangled things added to python since I last used it.

I've updated my PR.

[ddaspit]

Reviewed 2 of 2 files at r3, all commit messages.
Reviewable status: complete! all files reviewed, all discussions resolved (waiting on @rminsil)