loanpydatahub / ronataswestoldturkic Goto Github PK

"We can make this file beautiful and searchable if this error is corrected: Illegal quoting in line 11."

See glottolog/glottolog#899

In the current version I have extracted only data that is relevant for my own research, especially cols WOT, H and years. There is, however, much more interesting information in the pdf that could be included, like

• including Cumanian words from raw in CLDF as well
• adding the 39 Alanic borrowings on p. 1331-1339.
• Mongolic and sometimes Chinese etymologies for West Old Turkic words,
• the exact year and form of Hungarian words that appear in the written sources (phonetic transciptions of the various old-Hungarian orthographies are already provided in the pdf!),
• related words in the Turkic and other language families
• List of sources where it has been dealt with the given etymology

This could be a future project idea, maybe a good idea to get in touch with the authors of the original work and check if a database still exists somewhere and is available? Currently too much work for one person to parse + manually clean the pdf.

https://concepticon.clld.org/contributions/RonaTas-2011-431

@LinguList I added you as a collaborator for a general follow-up check to see what's left to do to make this repository ready to add it to the lexibank community :)

stick to this:

(https://en.wikipedia.org/wiki/Hungarian_phonology)

EAH vowel inventory: {'y', 'yː', 'ø', 'øː', 'i', 'iː', 'e', 'eː', 'ɛ', 'a', 'aː', 'ɒ', 'ɯ', 'u', 'uː', 'o' }

They are messing with the alignment and should not be there, e.g. "kk" should be "k:" They occur in very few words so can be ignored for now

have to identify into which one of the two the columns it should be integrated and make sure everything else stays in place

should only be "true" for recipient words (i.e. all lgs except WOT, since that's the donor lg)

I'm having troubles creating the borrowings-table. I have to connect EAH (the recipient words) and WOT (the donor words), but EAH is sometimes empty within a cognateset, so those I'd need to skip. I was experimenting around for a while now, trying to use a stack, but I assume there must be a much easier solution which I'm not aware of

Currently segmentation happens through orthography.py. All it would take to apply the clusterwise segmentation is from ipatok import clusterise and to replace tokenise with clusterise in line 12. But I don't know how to make that column appear in forms.csv eventually

There seems to be a small BIPA error hidden, but I can't find out which one. If I make a set of all the IPA-characters used, this comes out: ['', 'a', 'aː', 'b', 'c', 'd', 'd͡z', 'd͡ʒ', 'e', 'eː', 'f', 'h', 'i', 'iː', 'j', 'k', 'l', 'm', 'n', 'o', 'oː', 'p', 'r', 's', 't', 't͡s', 't͡ʃ', 'u', 'uː', 'v', 'w', 'y', 'yː', 'z', 'ø', 'øː', 'ŋ', 'ɐ', 'ɒ', 'ɛ', 'ɟ', 'ɡ', 'ɣ', 'ɥ', 'ɯ', 'ɲ', 'ʃ', 'ʎ', 'δ', 'χ']. I checked for every element if it's in the master list and it says yes. Which other strategies would there exist to pin down bad characters?

Since CLDF-validation is failing: Change 3.6 to 3.7 in https://github.com/martino-vic/ronataswestoldturkic/blob/165acbba26265bcfc51d8f498bce6fe7f71a5cc5/.github/workflows/python-package.yml
long term support version must have incremented by now from 3.6 to 3.7

Will need a comments.csv in the folder etc. referring to specific entrires, as well as general comments that go to the readme.

Dump from my personal notes:

• - kåråmbēl: the form with "n" must be a typo (kårånbēl),
• - changed csipa acc. to corrigenda, dzsartagan und dzarta zu yartagan und yarta,
• - ige: doesn't explicitely say WOT, but better putting it under WOT than skipping it all together.
• - add to readme: pages are not provided, entries must be looked up through the index that is provided in the end of the monograph WOT.
• - csigolya checken, gyón, köpönyeg, oktat, örmény
• - comment on the "V" like in čükV, esV -- loanpy should be able to handle "V"? Btw: Maybe this is causing the BIPA errors?
• - quote: "Berta suggested a T starting point for the word. However, its first documented appearance is fr the mid-17ʰ c. – certainly a weak point in a T etymology." -- Could use this to confirm the idea that col year can serve as a filter
• - böszörmény -- insert comment later that "WOT" must be just missing as a typo from the pdf
• - bǖbāy: explain why omitted from data
• explain in readme why dʹårmåt, tábor, teng were no taken into the data set

new mappings found here: https://github.com/martino-vic/rtbwestoldturkic/blob/main/etc/concepts.tsv
then re-run the lexibank script

half of the lines are wrong/superfluous: like EAH-0_carpenter-1 EAH-0_carpenter-1

I'm trying to add a column to forms.csv that counts how often each prosodic structure occurs in total in column "ProsodicStructure" in forms.csv and I can't figure out how to add this to the lexibank script. It's similar to this issue: I can't add the info from within the loop because I can count the number of occurrences only after the loop has ended. But somehow I don't manage to start a second loop at the bottom where I insert this info. Is there some kind of workaround for this @LinguList ?