Add write support for parquet

[charles-turner-1]

Change Summary

• In Add parquet driver #728, I added support for reading .parquet files, but completely forgot to add support for serialising catalogs back to them. This PR fixes that.
• Cleans up an issue with available compression associated with that PR.
• Tests we can deserialise and reserialise (from {.csv,.parquet} => {.csv, .parquet}).
• Adds write_kwargs, a la read_kwargs in Add parquet driver #728, plus soft deprecation of the to_csv_kwargs keywods argument when serialising.

N.B: Deserialisation & reserialisation tests round trip twice, it looks like the defaults for how empty catalog field options (in catalog.json) have changed since that catalog was created.

Checklist

• Unit tests for the changes exist
• Tests pass on CI
• Documentation reflects the changes where applicable
  • Will need to update ecgtools first to allow for quickstart guide update.
  • Relevant API reference bits updated.

---

I will no longer make pull requests when jetlagged.

[mgrover1]

@charles-turner-1 - let me know when you would like a review here

[charles-turner-1]

Sorry Max, completely forgot to reply to you here!

We've been using a fork the past month or so to allow us to go further/faster with development on this - I think there are a couple of minor tweaks there that need merging back into here. I'll ping you for a review once I'm satisfied that everything is working as intended!

(N.B I'm mostly treating our fork as something of a staging area - intention is to sync developments back in here.)

[mgrover1]

No worries @charles-turner-1 👍 sounds like a plan!!

[charles-turner-1]

(We implemented this & the concomitant change to ecgtools on our forks ~ 6 weeks ago with no hitches. I've been slammed with other stuff but I'll check push those changes back to here & ecgtools either this week or early next week).