Module talk:User:Erutuon/grc-decl/testcases
Add topicLatest comment: 7 years ago by Erutuon
- @JohnC5: The dative plural of third declension nouns in -as, -antos should be -asi, not -ani (ἐλέφασι (eléphasi), not **ἐλέφανι). --Barytonesis (talk) 16:19, 1 August 2017 (UTC)
- And with a long vowel, like the nominative singular. — Eru·tuon 16:42, 1 August 2017 (UTC)
Done (awhile ago). — Eru·tuon 20:23, 4 August 2017 (UTC)
- @Erutuon, JohnC5: the code "ἁλιέως/ἁλιῶς" is not parsed properly, and is linking to ἁλιέως/ἁλιῶς --Barytonesis (talk) 11:53, 3 August 2017 (UTC)
- The vocative of πολίτης (polítēs) should be πολῖτᾰ (polîta), not **πόλιτα. --Barytonesis (talk) 11:58, 3 August 2017 (UTC)
- The problem actually runs a bit deeper than that. Save for a few exceptions, when the retraction of the accent is constrained in the nominative singular, its place is retained (analogically, I suppose) in other cases where it's theoretically unconstrained. So the nominative plural of πολίτης (polítēs) is πολῖται (polîtai) (with application of the sôtêra rule), not **πόλιται; νεανίας (neanías) > νεανίαι (neaníai), not **νεάνιαι; φιλία (philía) > φιλίαι (philíai), not **φίλιαι. The problem is that in the case of νεανίαι (neaníai) and φιλίαι (philíai), it could give the impression that it happens because -αι is long; that's not the case. Would there be a way to disambiguate this? --Barytonesis (talk) 12:37, 3 August 2017 (UTC)
- It sounds like we need to add code that accents first-declension adjectives differently from first-declension nouns. Not sure what you mean by disambiguate; add a mark to show that the diphthong is short? — Eru·tuon 16:12, 3 August 2017 (UTC)
- Yes. --Barytonesis (talk) 11:39, 4 August 2017 (UTC)
- Okay, the nouns are now correct, but the adjectives are still wrong (*ἀξίαι). — Eru·tuon 23:51, 3 August 2017 (UTC)
- It sounds like we need to add code that accents first-declension adjectives differently from first-declension nouns. Not sure what you mean by disambiguate; add a mark to show that the diphthong is short? — Eru·tuon 16:12, 3 August 2017 (UTC)
- The problem actually runs a bit deeper than that. Save for a few exceptions, when the retraction of the accent is constrained in the nominative singular, its place is retained (analogically, I suppose) in other cases where it's theoretically unconstrained. So the nominative plural of πολίτης (polítēs) is πολῖται (polîtai) (with application of the sôtêra rule), not **πόλιται; νεανίας (neanías) > νεανίαι (neaníai), not **νεάνιαι; φιλία (philía) > φιλίαι (philíai), not **φίλιαι. The problem is that in the case of νεανίαι (neaníai) and φιλίαι (philíai), it could give the impression that it happens because -αι is long; that's not the case. Would there be a way to disambiguate this? --Barytonesis (talk) 12:37, 3 August 2017 (UTC)
- @Erutuon Re this edit, I'm thinking of this problem Angr has mentioned about participles. Do we want to fix it at some point? --Barytonesis (talk) 12:06, 3 August 2017 (UTC)
- @Barytonesis: Hm, I could give it a try. It might already be fixed, but there don't seem to be any testcases in Appendix:Ancient Greek adjective declension tables. — Eru·tuon 16:13, 3 August 2017 (UTC)
- I added some participles, and they're failing right now... — Eru·tuon 16:19, 3 August 2017 (UTC)
- I definitely want to make the neuter of participles automatic. It's predictable, so there's no point having to manually specify it every time. — Eru·tuon 18:34, 3 August 2017 (UTC)
- Done. — Eru·tuon 23:51, 3 August 2017 (UTC)
- @Erutuon, JohnC5: The neuter of κακοδαίμων (kakodaímōn) should be κακοδαῖμον (kakodaîmon).
- Done — Eru·tuon 20:22, 4 August 2017 (UTC)
- @Erutuon, Barytonesis: Heeeeeyyy... Someone wanna add some test cases for adjectives? —JohnC5 06:31, 5 August 2017 (UTC)
- I'm going to see about making a function to print out the code, because it's quite laborious otherwise. — Eru·tuon 07:12, 5 August 2017 (UTC)
- Done:
{{#invoke:grc-decl/sandbox|show_adj_forms|βᾰρῠ́ς|βᾰρεῖᾰ}}, for instance. — Eru·tuon 07:30, 5 August 2017 (UTC)
- @Erutuon, JohnC5: at στᾱ́ς (stā́s), the module is applying the rule of accent shifting between stem and ending for monosyllabic (accusative στάντα–genitive **σταντός instead of the correct στάντος (stántos)). More generally, would it be possible to add a parameter to tell it that it's dealing with a participle rather than a normal adjective? --Barytonesis (talk) 09:52, 20 October 2017 (UTC)
- @Barytonesis: That problem is fixed in the sandbox module. I think the module can correctly determine that a monosyllabic third-declension word is not supposed to have accent shifting by looking for ντ at the end of the stem, unless there are nouns with that description that I haven't thought of. — Eru·tuon 22:25, 20 October 2017 (UTC)
