Template talk:R:cau:NCED

Links not working

@Erutuon, Vahagn Petrosyan: The link to entries don't seem to work very well, see Reconstruction:Proto-North Caucasian/mɨ̆rꝗwă. I wonder if it has something to do with {{urlencode:{{{1}}}}}. --{{victar|talk}} 18:25, 21 May 2020 (UTC)

The URL encoding seems to be fine; the letters come through. The problem seems to be that there are no matches for either mɨ̆rꝗwă or mɨ̆rq_wă in this database. — Eru·tuon 19:04, 21 May 2020 (UTC)

@Erutuon: Nah bro, check it. I had to add &method_proto=substring, but even if I add that URL param to this template, it still doesn't work. --{{victar|talk}} 04:24, 22 May 2020 (UTC)

@Victar: Aha, I got it. The database uses LATIN SMALL LETTER A, COMBINING BREVE (ă) as the last letter, while Wiktionary has LATIN SMALL LETTER A WITH BREVE (ă) because of NFC normalization. No matches for the latter. So the template has to decompose (NFD normalization) – at least in this case, hopefully all the time. The website should treat NFC and NFD normalizations alike. Anyway, I'll see about making a decompose–URL encode Lua function to deal with this. (I'm confused because searching for ă in the database doesn't get any matches.)

Okay, actually the database uses both NFC (composed) and NFD (decomposed) forms (like ō) so I'll just have to play whack-a-mole with this one letter, or any others that turn up. I've written Module:StarLing to deal with this, but the "equals" matching method isn't working even if I copy and paste the headword into the query string (or try removing the *). — Eru·tuon 04:57, 22 May 2020 (UTC)

@Erutuon: Oof, sounds like a pain. Thanks for any efforts you might be able to put into it. --{{victar|talk}} 05:10, 22 May 2020 (UTC)

@Victar: The substring method works, though it might give unwanted results sometimes. — Eru·tuon 05:22, 22 May 2020 (UTC)

Awesomesauce! Seeing as there's absolutely no way to search for *mɨ̆rq_wă ( ~ -ā-, -ǝ̆), we don't seem to have a choice. Thanks, @Erutuon!

@Erutuon this link doesn't seem to work for some reason. --{{victar|talk}} 02:00, 25 May 2020 (UTC)

@Victar: Fixed! Hadn't considered the possibility of the uppercase. — Eru·tuon 02:04, 25 May 2020 (UTC)

@Erutuon: Thanks! --{{victar|talk}} 02:29, 25 May 2020 (UTC)

The automatic search almost never works. I manually link the page using url=, in this case https://starling.rinet.ru/cgi-bin/response.cgi?single=1&basename=%2fdata%2fcauc%2fcaucet&text_number=+127&root=config. By the way, Victar, I vaguely remember "Proto-North-Caucasian" being banned on Wiktionary, because its existence is uncertain like that of "Proto-Altaic". In any case, NCED should be used very carefully. It is not reliable. --Vahag (talk) 08:16, 22 May 2020 (UTC)

Thanks, @Vahagn Petrosyan:, I found the thread on that. I'd be in support of banning PNC, but someone need to create a vote to delete that language code. Nothing was finalized in the discussion. @Crom_daba --{{victar|talk}} 15:10, 22 May 2020 (UTC)

Orthography

In A North Caucasian Etymological Dictionary, are breve marks simply used to indicate short vowels? @Vahagn Petrosyan, Crom_daba? --{{victar|talk}} 15:54, 22 May 2020 (UTC)

I don't know. The preface should contain an answer. --Vahag (talk) 15:59, 22 May 2020 (UTC)

Thanks, it does. --{{victar|talk}} 16:07, 22 May 2020 (UTC)