Update .md docs in docs/

[anna-stacey]

Some parts of these are totally out-of-date e.g., discussing OPDDatabase

[anna-stacey]

This has been largely overhauled, but there's a couple lingering things:

• Add some more examples/details about flag diacritics to docs/fst.md. I removed the spot where a list of all flag diacritics would go (because there are like 2000), but I think some more varied examples could help. I don't even understand what all the crazy long equals signs ones are about, e.g., @U.Order.Other@ NAD_Class=NAD_Cw_Prefix=G_Order=Other_Endings ;. Basically I'll need to ask Miikka about this.
• Update the flowchart in docs/README.md
• docs/morphological_source.md talks extensively about the config (.json) files in OjiweMorph. There's a couple features there that I believe are unused and could be removed. There's also inconsistencies in terms of what can be left out vs. what is 'required' but can be sent to NONE or "" or whatever. These minor lingering weirdnesses are mentioned in these docs when going through the features in the config files. Could either talk to M or just remove the unused features. Would also be good to get his advice on whether it's better to require all features be listed or not.
• In docs/morphological_source.md, the preverb section is just very descriptive about what the OM proverb spreadsheets look like, cf. discussing the potential to have these kind of spreadsheets in a language-general way. Plus, something needs to be added about enclitics.
• In docs/morphological_source.md, the .xfst section discusses the rules and their syntax, but I'm still pretty unsure about what else happens in these files. Considering lack of resources, probably need to check w Miikka.
• In docs/FSTmorph.md, the step-by-step verb lexicon example is incomplete. I could probably attempt to fill this in on my own, and then just run it past Miikka. The image itself will also need to be updated to include enclitics, so ask how to do that as well.
• I just noticed that despite seemingly going through the FST creation from start to finish, I never interacted with extract_tag_combinations.py. It generates verb_tags.json, etc. in FST/generated/, which as far as I can tell are not used. For now I have commented out the creation of these files (and thus any use of extract_tag_combinations.py from the Makefile in OM. Before removing this code entirely, I'd like to confirm this deletion with Miikka.

[anna-stacey]

Updated flowchart with b149f1d