Module talk:R:Perseus

Sources

The table at W:Greek_Extended proved useful. --Dan Polansky (talk) 14:14, 16 March 2014 (UTC)

The guillemets

@I'm so meta even this acronym, Isomorphyc: Per this diff, could we discuss the use of guillemets? —JohnC5 07:35, 16 July 2016 (UTC)

I obviously oppose using these guillemets. You might want to kick off the discussion by providing a rationale for using what is not the usual English typography, as far as I know. --Dan Polansky (talk) 08:14, 16 July 2016 (UTC)

Hi @Dan Polansky, @JohnC5: I don't have a strong opinion about this. Headword linking is not a bibliographical practice. The question is whether to follow the convention of the source or target langauge. A contra-guillemet datum would be that while French is normally a guillemet language, the template R:TLFi uses double-quotes. The original rationale was to follow the Modern Greek practice, as is the precedent here: Ἀριστοφάνης (Aristophánēs) and elsewhere, as cited in this discussion. I.S.M.E.T.A. points out elsewhere on this page that the guillemets avoid the potential orthographical confusion between the dasia-oxia (marking some uppercase letters) and the English-style quotation marks. Since this discussion was linked from the original vote, it should be accorded at least a modest degree of respect. Isomorphyc (talk) 10:17, 16 July 2016 (UTC)

Assuming I understand correctly that you are saying that modern Greek uses guillemets for quoting, why would an English dictionary use typographical quoting conventions of another language? I don't understand what it is for a discussion on a user's talk page to get "degree of respect"; does the discussion contain any convincing arguments or evidence? If so, where in that discussion (User talk:I'm so meta even this acronym#R:LSJ and the Perseus Resolve Form) are these arguments and evidence; can they be quoted here? --Dan Polansky (talk) 12:34, 16 July 2016 (UTC)

Hi @Dan Polansky: I summarised what was the most convincing argument on the talk page already; here it is in full: I substituted the “” quotation marks with guillemets because (Modern) Greek uses guillemets as its primary quotation marks and because there exists the potential for confusion between ⟨ “ ⟩ and ⟨ ῞ ⟩, especially where that diacritic (dasia-oxia) marks a capital letter. --User:I'm so meta even this acronym. The talk page discussion should be accorded modest respect because it is linked from the second BP discussion regarding the vote. Of the five Greek editors who participated extensively in the references discussion, two strongly favoured guillemets, I had no strong opinion, and neither of the other two till now expressed an opinion. I'm willing to implement whatever most people agree on, but I'd like to be consistent across the Greek reference templates. Isomorphyc (talk) 13:17, 16 July 2016 (UTC)

The quoted argument is invalid on two grounds: 1) we have the option of using no quoting, so no confusion can arise--indeed, that is the status quo ante for R:LSJ; 2) the convention used in Greek is irrelevant for quoting in the middle of English text, and segment of the form '"X" in Y' is English, using English preposition "in", even if X is Greek. Similarly, for Duden we do not want to use German quoting conventions, for Czech reference templates we do not want to use Czech quoting conventions, likewise for French, etc. --Dan Polansky (talk) 13:24, 16 July 2016 (UTC)

I agree with you, but it is still 2-2. I don't know who else is an interested party, but if others with opinions can be found we should be able to solve this. Isomorphyc (talk) 13:39, 16 July 2016 (UTC)

To be fair, I could go with guillemets or with nothing at all. Quotation marks are obviously bad, and it does look nice to show that text is quoted, but I'm fine leaving them off. I just thought we should talk a bit first and make sure we came to an agreement. Before we change the rest, however, I think we should give ISMETA some time to respond. —JohnC5 16:31, 16 July 2016 (UTC)

@JohnC5, Dan Polansky, Isomorphyc: To begin, I'd like to clarify that the argument of mine quoted by Isomorphyc is from User talk:I'm so meta even this acronym#dēlēlicet, in my post timestamped 21:57, 20 March 2015, specifically point № 1 thereof; JohnC5 wrote "I like the guillemets that you added (they look super classy), but is there a reason why?", to which I responded "I substituted the “” quotation marks with guillemets because (Modern) Greek uses guillemets as its primary quotation marks and because there exists the potential for confusion between ⟨ “ ⟩ and ⟨ ῞ ⟩, especially where that diacritic (dasia-oxia) marks a capital letter.". User talk:I'm so meta even this acronym#R:LSJ and the Perseus Resolve Form includes, in my post timestamped 15:15, 1 July 2016, my request to Isomorphyc: "would it be possible to enclose the cited headword in guillemets: « » per the practice of Modern Greek punctuation, per {{R:DGE}}, and per the citations of those two authorities at Ἀριστοφάνης, Ἰφιγένεια, Πόλυβος, Ἀντίκλεια, ὄσσε, Εὐρύκλεια, ἠριγένεια, Ἠριγένεια, and κεχηνώς, please?". I also wish to thank JohnC5 for advocating giving me ample time to respond; as I have mentioned elsewhere, I no longer have Internet access at home, so I am far less able to involve myself in the cut-and-thrust of discussions here.

I conclude from my reading of the discussions so far that the point of contention is whether to use guillemets or not to use any quotation marks whatsoever. Re Isomorphyc's hypothesis that "[a] contra-guillemet datum would be that while French is normally a guillemet language, the template R:TLFi uses double-quotes" presupposes that the choice of quotation mark is a per-language consideration. A quick look at w:Quotation mark#Summary table should be enough to disuade advocates of per-language variation of the chaos that would ensue from our following a per-language policy for quotation marks. I advocate a per-script policy. I advocate the use of “ ” for the Latin script, « » for the Greek script, 「 」 for Han-derived scripts, and so on. Beside the aforementioned issue of mark-confuseability (as in the case of the dasia-oxia: ⟨ ῞ ⟩ versus the double opening quotation mark: ⟨ “ ⟩), there is also the issue of the ambiguity of cites works' parts; for example, what if we wanted to cite a sub-entry in Valerius Harpocration's Lexicon in decem oratores Atticos? How would we distinguish the main entry from a sub-entry? But beyond that, the “” quotation marks simply look wrong when used with the Greek script or with Han-derived scripts. Because of their proximity to the enclosed terms, the quotation marks simply look a lot better if they are script-matched.

That's my opinion, anyhow; I appreciate the opportunity to voice it. — I.S.M.E.T.A. 22:46, 14 August 2016 (UTC)

Code entrance point

@Isomorphyc: Might I recommend adding separate create functions for AG and Latin? This would save a fair amount of silly logic and would only involve calling different functions in the respective templates. Whad'ya think? —JohnC5 17:35, 3 August 2016 (UTC)

@JohnC5: I'm not incredibly excited about this suggestion. A clunky function call sounds to me like an invitation for someone else to rewrite our silly logic, whereas I can use a diaeresis just by talking about moving the logic to a silly logic zoö. Isomorphyc (talk)

@Isomorphyc: Okey dokey. —JohnC5 18:01, 3 August 2016 (UTC)

@JohnC5: Apologies; that came out about as silly as I intended, but I intended it to be sillier than I should have. If I can't make the logic look decent I will do this. But I recently realized Perseus has Zoega's dictionary, so per-language entry points will likely lead to more than two. Moreover, this enables inconsistent template interfaces, which I would prefer to ameliorate, discourage, grandfather or extirpate as appropriate. My real mistake was doing this in the main space rather than a sandbox first. It hasn't been a catastrophe, but the mess turned out to be about half a standard deviation worse than I expected. Isomorphyc (talk) 15:55, 4 August 2016 (UTC)

Edit: if you are wondering about my inconsistent spelling, it is because different user interfaces have different automatic spell-check settings which I have more-or-less stopped learning how to turn off. Isomorphyc (talk) 16:00, 4 August 2016 (UTC)

@Isomorphyc: This all makes sense, and thanks for the longer explication. —JohnC5 18:12, 4 August 2016 (UTC)

Tagging language

Perhaps this is not the place to suggest this, but it would be really nice if this and other Ancient Greek reference templates tagged Greek text with <span lang="grc"></span>. It is somewhat nitpicky, but I do like to see Greek formatted with the fonts that I've selected in my common.css. I would try to make the change myself, but I am not sure I could do it without breaking something. — Eru·tuon 07:15, 1 October 2016 (UTC)

@Erutuon: I added this for all of the Perseus templates and for Strong's (which prints Greek text in cases where no Strong's number is available). Please let me know if it doesn't work, or if you have a list of other templates which need this HTML tag; it is not so easy for me to test and view because I do not use javascript myself in Wiktionary. Thanks. Isomorphyc (talk) 23:46, 7 October 2016 (UTC)

Edit: I have also added these tags to : R:Cunliffe, R:Bailly, and R:DGE. Isomorphyc (talk) 00:12, 8 October 2016 (UTC)

It's better to use the tag_text(title, lang) function in Module:script utilities. DTLHS (talk) 00:22, 8 October 2016 (UTC)

When the LSJ entry is not the same as the Wiktionary entry

In εὐκτική (euktikḗ), I am not sure how to link to the LSJ entry for εὐκτικός (euktikós), which describes the word. I've tried using many permutations of |1=, |2=, and |w=, but none give the desired combination of a link with the beta code eu)tiko/s and the link text εὐκτικός. I wish I could simply give the Greek and have the module use it as link text while converting it to Beta Code to generate the URL for the link. — Eru·tuon 22:28, 7 October 2016 (UTC)

Not sure who to ping: @ObsequiousNewt, @Isomorphyc? — Eru·tuon 22:29, 7 October 2016 (UTC)

@Erutuon: I have added this feature. I think it was never implemented because the original template expected a beta code argument in the first position. Now it performs the expected behaviour when a polytonic script is detected in that argument position. Please let me know if there are issues. Isomorphyc (talk) 23:53, 7 October 2016 (UTC)

@Isomorphyc: Thank you! It works now at εὐκτική (euktikḗ) at least. — Eru·tuon 00:05, 8 October 2016 (UTC)

Errors

@Erutuon: There seem to be some new errors in CAT:E. —JohnC5 05:52, 26 February 2017 (UTC)

@JohnC5: Ahh, I was simplifying the regex and didn't finish. Now it works. — Eru·tuon 05:58, 26 February 2017 (UTC)

Changes to parameters

I've changed the logic of parameters. Completely gone is the |w= parameter, which seemed to be only or mainly used in {{R:LSJ}}. See Wiktionary talk:About Ancient Greek#Template:R:LSJ and Module:R:Perseus for more information. — Eru·tuon 08:54, 4 December 2017 (UTC)