Talk:arbeir

From Wiktionary, the free dictionary
Jump to navigation Jump to search

Requested fixes[edit]

@Mellohi!, Mahagaja The argument subjunctive_ps_deut is duplicated in the call to {{sga-conj-complex}}. Could one of you fix this? BTW this template really should (but does not) automatically link its arguments using {{l|sga|...}}, so that the caller doesn't have to do this. Any objections to me making this change? Benwing2 (talk) 03:44, 5 November 2020 (UTC)[reply]

@Benwing2: Fixed the subjunctive passive duplicate problem - one of the forms was supposed to be plural, not singular. — Ceso femmuin mbolgaig mbung, mellohi! (投稿) 04:04, 5 November 2020 (UTC)[reply]
@Benwing2, Mellohi!: the reasons I originally wrote {{sga-conj-complex}} to require the editor to type {{l|sga|...}} manually were (1) to make it possible to include qualifiers like {{small|{{q|relative}}}} or {{small|{{q|{{m|sga|ro-}}form}}}} or {{small|{{q|with infixed pronoun {{m|sga|id-}}}}}}, and (2) to make it easier to include multiple variants in a single line. The work of generating the template has since been moved to Module:sga-verbs, which means I no longer know how to edit it, but if someone wants to update the module to make it possible to add qualifiers to specific forms (maybe via parameters that look like |subjunctive_ps_deut_q1= or the like) and to specify a variety of spellings for given forms, that would be great. It would be especially nice if Module:sga-verbs worked like Module:grc-conj in resolving parentheses, so that the editor could type, say, |present_2p_deut=as·be(i)rid or |perfect_3p_deut=as·rubartat(ar) and the software would automagically render those as "{{l|sga|as·berid}}, {{l|sga|as·beirid}}" and "{{l|sga|as·rubartat}}, {{l|sga|as·rubartatar}}". —Mahāgaja · talk 09:32, 5 November 2020 (UTC)[reply]
@Mahagaja The actual code to generate the table is still at Template:sga-conj-table-complex, not in Lua. The issue of specifying several forms does not preclude automatically linking with {{l|sga}}, the caller would just use raw links around the multiple forms and everything will work fine. The issue of qualifiers is different; that could be done in the template code using separate arguments but probably easier in Lua. Let me see about moving the code to Lua and implementing qualifiers, multiple forms and auto-expansion of parens. Benwing2 (talk) 13:31, 5 November 2020 (UTC)[reply]
Actually the current situation is a bit more complex that what I just described, but no matter, I can fix it. Benwing2 (talk) 13:49, 5 November 2020 (UTC)[reply]
@Benwing2: That would be great, thanks! Also, while today it's true that {{l|sga|[[as·beirid]], [[as·berid]]}} renders correctly, I'm pretty sure that back in 2012 when I first wrote the conjugation table templates, that syntax did not work. —Mahāgaja · talk 14:27, 5 November 2020 (UTC)[reply]
@Mahagaja, Mellohi! This is implemented. I fixed arbeir to use the new format. The code will leave alone any cases with a "<" in it, so that existing entries already formatted with {{l|sga}} or with a manually entered qualifier won't be modified. Paren expansion is also implemented (including multiple sets of parens in a single word), as is splitting on commas; to enter multiple forms, either put them at e.g. |past_participle=, |past_participle2=, or enter both at |past_participle=, comma-separated. (However, only the former format allows you to attach separate qualifiers to each form, respectively at |past_participle_q= and |past_participle2_q=.) Benwing2 (talk) 14:52, 5 November 2020 (UTC)[reply]
@Benwing2: Wow, that was fast! It looks great! But you've renamed the inflection class categories, e.g. CAT:Old Irish t preterite verbs is now CAT:Old Irish class t preterite verbs. Was that on purpose? —Mahāgaja · talk 14:56, 5 November 2020 (UTC)[reply]
@Mahagaja Oops, no that wasn't intentional, will fix. Benwing2 (talk) 14:58, 5 November 2020 (UTC)[reply]
@Mahagaja Should be fixed. Benwing2 (talk) 15:00, 5 November 2020 (UTC)[reply]
@Benwing2: Could variant forms generated by separate parameters like |<whatever>=, |<whatever>2=, |<whatever>3= etc. be separated by semicolons instead of commas? I find myself using the comma within a parameter for minor things like spelling differences and separate parameters for bigger differences like ro-forms and forms with infixed pronouns. —Mahāgaja · talk 16:34, 5 November 2020 (UTC)[reply]