Module talk:grc-headword

It's funny because sc is not used at all in the first place, nor will it be necessary in this process to my knowledge. —JohnC5 04:16, 31 March 2017 (UTC)Reply

@Erutuon, once we get this template up and running, are you fine with moving the accentation categorization to the headword instead of the inflection tables? That will ensure that only pagenames with the correct accentation will show up in the categories. —JohnC5 09:55, 27 October 2017 (UTC)Reply

@JohnC5: Definitely. — Eru·tuon 18:38, 27 October 2017 (UTC)Reply

@Erutuon: I've converted all the templates in Category:Ancient Greek headword-line templates except for {{grc-diacritical mark}}. If there are any upgrades or categorization you'd like to add, please do. Otherwise, I think we're good to remove the accent categorization from the declension templates. —JohnC5 23:34, 29 October 2017 (UTC)Reply

@JohnC5: I removed the accent categorization from Module:grc-decl/table, and I had removed it already from Module:grc-decl/sandbox/table. I've considered the idea of making {{grc-noun}} detect declensions, and the adjective templates generate the neuter or feminine and neuter forms, but I don't feel like doing that right now. — Eru·tuon 02:18, 5 November 2017 (UTC)Reply

@Erutuon: Thank you for changing that. I think detection would be cool, but is not necessary at the moment. I was also thinking of making headword templates for the more obscure lemma types like particles, but I'm too busy at the moment. —JohnC5 07:46, 6 November 2017 (UTC)Reply

@JohnC5: I finally remembered an idea I had: allowing the noun and adjective templates to use the first parameter for the headword if there are ambiguous vowels. So, {{grc-noun|δήμου|m|second}}, but {{grc-noun|ῐ̔́ππος|ἵππου|m|second}}. This would be more elegant than using the |head= parameter {{grc-noun|head=ῐ̔́ππος|ἵππου|m|second}}. — Eru·tuon 17:37, 10 November 2017 (UTC)Reply

@Erutuon: I am very excited about this idea. —JohnC5 22:45, 10 November 2017 (UTC)Reply

@JohnC5: Done. Hopefully no bugs. — Eru·tuon 02:59, 11 November 2017 (UTC)Reply

Re declension detection, how would it work? We manually input the nom. and gen., and the template assigns the declension class (first/second/third) from there, or would it be the other way around: manually input first/second/third, and it generates the gen. from there? Or something else? --Barytonesis (talk) 08:46, 11 November 2017 (UTC)Reply

It would have to be similar to {{grc-decl}}, where we must input the nominative and genitive. Look at Module:grc-decl/decl/staticdata; it's not always possible to determine the genitive from the nominative and declension. — Eru·tuon 09:00, 11 November 2017 (UTC)Reply

Headword can now be placed in the first parameter: {{grc-noun|ψῡχή|ψῡχῆς|f|first}}, {{grc-noun|λόγου|m|second}}.

However, the method might not be the best, as it's difficult to accommodate indeclinable nouns. Indeclinable nouns that need a headword are fine ({{grc-proper noun|Ἰσρᾱήλ|m}}), but I am not sure if it will be possible to make {{grc-proper noun|gender}} work for those that do not. — Eru·tuon 20:39, 11 November 2017 (UTC)Reply

I rewrote the logic, and now the numbered parameters are much more flexible. For instance, in {{grc-noun}} and {{grc-proper noun}}:

• {{grc-noun|ψῡχή|ψῡχῆς|f|first}} in ψῡχή
• {{grc-noun|λόγου|m|second}} in λόγος
• {{grc-proper noun|Ἰσρᾱήλ|m}} in Ἰσρᾱήλ (Isrāḗl)
• {{grc-proper noun|m}} in Ἡσαῦ (Hēsaû)

The new logic might be a little confusing because one numbered parameter does not correspond to one meaning, but I think it's intuitive: you simply put the headword, genitive, gender, or declension, in that order, in numbered parameters, and the template should handle them correctly.

The logic is similar with the adjective, participle, verb, and preposition headword templates, only the list of items that can be given in numbered parameters is different. — Eru·tuon 20:14, 14 November 2017 (UTC)Reply

I assume in the first line of your message above you mean {{grc-noun}}, not {{grc-decl}}, right? —Aɴɢʀ (talk) 23:37, 15 November 2017 (UTC)Reply

@Angr: Yep. I keep making that mistake. Fixed. — Eru·tuon 23:39, 15 November 2017 (UTC)Reply

@Erutuon, Benwing2: With {{grc-verb form}} at least, and probably others as well, if |1= is not specified, the entry gets added to CAT:Ancient Greek terms with redundant head parameter, which seems suboptimal. Can this be fixed so the category gets added only if |1= is actually redundant and not merely absent? —Mahāgaja · talk 12:53, 20 July 2023 (UTC)Reply

@Mahagaja: Yes, done, I think. Benwing2 fixed the redundant headword categorization for nouns, and I applied the same fix to all parts of speech. — Eru·tuon 15:54, 20 July 2023 (UTC)Reply

I'm considering how to treat σκότος (skótos) and other nouns with two headword templates that can be found by searching incategory:"Ancient Greek nouns" insource:/\{\{grc\-noun[^{]+\{\{grc-noun/ (with a few false positives). Most of them are nouns in -ος (-os) that are in both the masculine second declension and neuter third declension.

@JohnC5 noted that {{grc-noun|m|second|decl2=third|g2=n}} doesn't work, because that puts the noun in the categories for "masculine second-declension nouns", "neuter second-declension nouns", "masculine third-declension nouns", and "neuter third-declension nouns, when only masculine second declension and neuter third declension are correct.

I'm thinking maybe something like {{grc-noun|...|second|decl1g=m|decl2=third|decl2g=n}}. — Eru·tuon 02:37, 25 January 2018 (UTC)Reply

@Erutuon: Another issue to consider is the fact that (with the current layout) it may not be immediately apparent which sets of genders in and declension belong together. Thus "m, n" at the beginning of the headword could be associated with the wrong genitives and the wrong declension labels. Any layout which would fix this would need to disentangle all three outputs. For this reason, I think keeping them separate is easier. —*i̯óh₁nC^[5] 04:44, 25 January 2018 (UTC)Reply

@Erutuon, Benwing2 This module seems to be putting basically all Ancient Greek entries into CAT:Ancient Greek terms with redundant script codes. Is there some way to fix that? Does the line local sc = require("Module:scripts").getByCode("Polyt") have something to do with it? —Mahāgaja · talk 15:32, 10 July 2024 (UTC)Reply

@Mahagaja: Yep, that's the cause. Manual script codes weren't checked before. I've removed them, though I don't think there was any harm in supplying them when they are correct. — Eru·tuon 18:19, 11 July 2024 (UTC)Reply

Thanks! —Mahāgaja · talk 18:57, 11 July 2024 (UTC)Reply