User talk:AutoFormat/2009

{{wikipedia}} placement

WT:GP#Broken WT:PREFS reminded me that if {{wikipedia}} is placed incorrectly (e.g. above all language sections) it messes with right-side TOCs and other elements over there. A quick scan reveals there are thousands of entries are this way. Is the logic required in moving {{wikipedia}} to the correct Template talk:wikipedia#Placement simple enough for AF to do it? I'd imagine that it can just be moved to immediately after the first language section (example). This would help if we make TOCs on the right for more users. Cheers. --Bequw → ¢ • τ 22:15, 24 December 2008 (UTC)

The same thing could also be done for images above the first L2. --Bequw → ¢ • τ 18:32, 27 October 2009 (UTC)

Check language heading against template use

In the last few days I've come across and fixed at least half a dozen articles French, Finnish, Icelandic, etc which wrongly had the English language heading. I spotted them because they didn't look English but a bot could spot them by the language codes in various templates. It would be helpful if Autoformat could spot such mismatches and add an ((rfc)) or cleanup category of some kind.

Another less reliable way to spot them would be language-specific categories but since categories are page level rather than entry level they will obviously not work as well. — hippietrail 00:58, 1 January 2009 (UTC)

That is a reasonable idea ... if {{xx-something}} or {{infl|xx|something}} and xx ≠ language code? The categories are not that hard, it is already sorting them into the correct language section if it exists, it is a matter of tagging them when it doesn't. (As in one of your fixes, if we have Category:Ewe nouns, and "Ewe" not in Lsects:). It won't catch if the wrong section is labelled "Ewe" of course. I'll look at these. Robert Ullmann 10:45, 1 January 2009 (UTC)

Misconstructions still.

In case you're not aware: http://en.wiktionary.org/wiki/mischievious?diff=5780736.

Happy New Year!

—Ruakh_TALK 14:39, 1 January 2009 (UTC)

P.S. I meant that comment for Robert, but Happy New Year to you as well, AF. ;-)   —Ruakh_TALK

Multiple Production sections

AutoFormat made this edit, stripping "Production 1" and "Production 2" both down to "Production" and then complaining about the heading level I used for the POS sections. I think my original structure was correct because it shows that the adjective sense "late" only applies to the first production section, while the adverb "not yet" sense only applies to the second production. I'd be happy to conform, though, if there is a different recommendation. Your advice? —Rod (A. Smith) 00:07, 7 January 2009 (UTC)

This parallels the problem with numbered Pronunciation sections, which have never been described in a way that could be added to ELE, being used ad hoc. I've moved Production up to standard, added Prod n headers (a few) in the same way, and set the tagging the same way. Robert Ullmann 08:43, 21 January 2009 (UTC)

Alternative spellings or Alternative forms?

Please take note of this; which does AutoFormat præfer: Alternative spellings or Alternative forms? –Or is it neutral on the issue? † ﴾^(u):Raifʻhār ^(t):Doremítzwr﴿ 05:30, 10 January 2009 (UTC)

It is neutral. Alternative spellings is standard and preferred, but (as HT points out) makes little sense in cases that are not "spellings"; so Alternative forms has come to be used. IMHO, we should allow both, prefer " ...spellings" when it makes sense, and things reading the data can treat them equivalently or differently as desired. (compare Pronunciation and Production, both standard)

AF does replace "forms and variants" and "other forms" with alt forms, and "other spellings" with alt spellings, see User:AutoFormat/Headers. Robert Ullmann 15:53, 10 January 2009 (UTC)

Non-Usage Notes

Twice now ([1] [2]) the bot has insisted on changing a Notes section to Usage Notes. However, the section has absolutely nothing to do with Usage. I'd rather not continually revert the bot. Any solutions? Bendono 10:18, 14 January 2009 (UTC)

There is no Notes header. There is only Usage notes It seems like you want this to be part ot the etymology, so put it there. There is no "see notes". I put the stuff in the ety. If you want it structured as before, you might consider the fact that the "notes" are in fact about historical usage, and that "Usage notes" is exactly correct; and you might as well use the standard header, it being right? Robert Ullmann 10:50, 14 January 2009 (UTC)

It's not etymology either. Removed. *sigh* Bendono 11:17, 14 January 2009 (UTC)

It's definitely not a usage note. It is part of the etymology. Years ago I suggested a "Word history" section for just this kind of thing as a sebsection of the Etymology section. I'll back up such a proposasl if you add such a section to this entry. There are many books on the topic of word history which is related to but distinct from etymology so there's no reason we shouldn't cover it too as part of a dictionary of very large scope. — hippietrail 13:02, 21 January 2009 (UTC)

It's not really etymology, either. It's more like "interesting trivia", that Japanese /h/ used to be pronounced [ɸ], that this can be seen in old riddles about the word 母 (haha, “mother”), and that more recent readers didn't understand these riddles. We do allow "Trivia" sections. —Ruakh_TALK 13:41, 21 January 2009 (UTC)

Proposal to separate AutoFormat into a text filtering backend and MediaWiki bot front end

Hi Robert. I have been thinking about some alternative uses for the solid and impressive AutoFormat code. It would be really handy if the code that does the actual conversion of the wikitext could be separated from the bot code so that I could write other front ends to use AF in various scripts, tools, and experiments. I don't currently speak Python sufficiently to grok if this is already the case or not. What do you think? — hippietrail 12:42, 17 January 2009 (UTC)

Any additions or removals from the Category:100 English basic words, etc.

We have a number of categories, like Category:100 English basic words which are based on static sources, and whose membership should not change, ever. However, people often don't notice this, or try to make trouble, by adding or removing these categories. It would be nice to add a feature to AutoFormat that fixed this. It could have a list of the proper contents of these categories, and revert any inclusions (or lack of inclusions) that wern't on the list. JesseW 09:15, 7 February 2009 (UTC)

Unnecessary edits

Several recent edits, such as this one, are unecessary since the change does not effect any change on the page. __meco 19:20, 10 February 2009 (UTC)

See the reason for these, they shouldn't have any effect on rendering, but will then permit the removal of the */bullet from the templates. Robert Ullmann 23:32, 10 February 2009 (UTC)

Very well. That's an acceptable reason. __meco 09:59, 11 February 2009 (UTC)

Converting language family Etymology templates to {etyl}

Now that {{etyl}} handles language family codes, could I add some entries to the User:AutoFormat/Ety_temps control file? Specifically, those from WT:RFDO#Unused etymology .28dot.29 templates (e.g. {{AA.}} → {{etyl|aus}}). --Bequw → ¢ • τ 14:18, 19 February 2009 (UTC)

bullet before {pedialite}

Hi, Otto. I know you put a bullet before {pedialite}, so thought you might want to know there's a new redirect to that template — template:Wikipedia-inline — that you can also put a bullet before.—msh210℠ 20:11, 4 March 2009 (UTC)

suffix categories

Would AF be able to make simple corrections such as this one? --Jackofclubs 20:14, 5 March 2009 (UTC)

Way to leave categories on sense lines?

I'd really prefer to have categories tied to a single sense be attached to that line, but AutoFormat likes to move them off the line and afterwards. See this edit for an example of what I mean. Will this require a minor code change, or would I be better off to bring up in the Beer Parlor the idea of a template to keep categories with sense lines. Carolina wren 00:15, 8 March 2009 (UTC)

I think you'd have to bring up in the BP the idea of putting categories on sense lines. My understanding is we don't do that currently. --Bequw → ¢ • τ 18:34, 27 October 2009 (UTC)

Clicks

In translation sections, language names starting with a click (ǀ, ǁ, ǂ, ǃ, ʘ) are sorted at the very end of the translations table. It would make more sense if the click letter was ignored when occuring initially. -- Prince Kassad 19:48, 29 March 2009 (UTC)

It might make more sense to sort ǀ, ǂ, and ǃ as t; ǁ as l; and ʘ as p. These are letters, after all, and represent sounds; they are not something like stress markers or what-have-you. Personally, I like the current situation best of all.—msh210℠ 02:35, 31 March 2009 (UTC)

μῆδος

I have reverted AF here (no doubt he'll try again, the bot is damned persistent :P). The inflection is set at L3 because it applies to both etymologies. It might be nice if there was a template which could be inserted to tell AF to back off when a human editor has an entry the way they like it. It could even insert a category like "Category:Completely wrong and terrible entries which AF is not allowed to fix" or something, if that would make you happy. -Atelaes λάλει ἐμοί 23:19, 9 April 2009 (UTC)

context tags out of, er, context

Hey, Otto, do you, or can you, catch context tags used in lists of -nyms, derived terms, descendants, see-alsos, and the like, and convert them to {{qualifier}}s, as in diff, so they don't categorize?—msh210℠ 22:23, 14 May 2009 (UTC)

Auto adding of * in translations sections?

Would it be unreasonable to have AF add a missing * at the beginning of a translation line where such an addition by itself would make the entry something it accepts? See for example this edit. It's certainly bot-doable, but you'd have a better idea of whether the extra overhead is worth it. — Carolina wren discussió 23:39, 14 May 2009 (UTC)

A reasonable idea. One does want to be careful not to try to coerce everything into format; it might mean something else entirely. AF does add : after a recognized language name; would be reasonable to add * before. Robert Ullmann 13:32, 20 May 2009 (UTC)

Template:zh-zh-p in trans tables

AF seems to object to {{zh-zh-p}} in the translation table at Qinghai, but it looks like something that should be added to the allowed list to me. Thryduulf 11:41, 16 May 2009 (UTC)

Where did that piece of crap come from? We don't use "Chinese:" by itself in translations tables, and the references should be using {{t}} properly. More stuff to clean up ...

The "unknown line format" is because the character at the start isn't "*", but some variant (Arabic start), so isn't the wikitext formatting character. Robert Ullmann 12:58, 20 May 2009 (UTC)

adding lang= to {homophones} (à la context tags)

Hi, Robert. Would it be feasible to have AF add a lang= parameter to {{homophones}} the way it does to context tags? And would you mind having it do so? That would be to categorize. (Discussion on categorization — or mention of it, anyway — is currently in the Grease pit (permalink to revision). Even if there is no consensus to categorize all languages, English homophones already have a category, so lang=fr would help in template-based categorizing of English homophones by preventing inclusion of French words into the English category.) Rsvp.​—msh210℠ 16:58, 24 June 2009 (UTC)

Translations titles

If you have the time, could you persuade AutoFormat to add a ====Translations==== header at the appropriate depth before the first translation table, I find 10-20 entries with this error every time I run the indexer. Thanks Conrad.Irwin 00:10, 1 August 2009 (UTC)

confusion at tyst

What does the bot want to express with this edit? In its summary it has written: Adjective at L4+ not in L3 Ety section, but the adjective at L4 section is subordinate to the L3 Etymology section - firstly language header, then pronunciation(L3), then Etymology(L3) including the adjective(L4). The uſer hight Bogorm converſation 07:43, 27 August 2009 (UTC)

When there is only one Ety section, the POS goes at level 3; the usual case. POS only go at L4 when nested under Ety N sections. See DCDuring's subsequent edit. Robert Ullmann 15:34, 28 August 2009 (UTC)

Latin letter names, etc.

In this edit [3] and possibly others, AutoFormat made a well meaning but mistaken move of a category.

Category:Latin letter names is not a Latin category, but an English one as it contains the English names of letters of the Latin alphabet. I imagine that it would have the same issue with Category:Greek letter names, Category:Hebrew letter names, and any further subcats of Category:Letter names. The corresponding Latin category would be Category:la:Latin letter names etc.

In a more serious concern, in that same edit it moved categories out of comments and into non-comment space. — Carolina wren discussió 15:20, 29 September 2009 (UTC)

okay, "... letter names" should be treated as with "... derivations". Fixed.

in the other case, there are a number of things that shouldn't be in comments, because the wikitext is parsed by all sorts of software, written by people with all sorts of varying capabilities. In particular, blocks of an entry with headers etc shouldn't be commented out. The HTML comment syntax is just a hack in wikitext anyway (;-) albeit commonly used. (AF might have function added to tag some of these nasty cases ;-) Robert Ullmann 15:31, 29 September 2009 (UTC)

novel

This entry had a clueless but trying to be helpful IP poster who was adding Latin translations to the entry notice an "incorrect" "Latin" translation and delete it. [4] It was actually a the Latin script version of the Serbian translation. Certainly nothing AutoFormat can do about that, but as you may have already noticed, when he did that he also removed the colon from the Cyrillic line, causing AF to think it was a translation for the "Cyrillic" language, and move it (oddly enough, in his case just under Croatian ^_v^ ).

Any such encounters by AF would be the result of either an error or vandalism. Low priority, as any such errors by IP editors should be caught anyway, but if you think it's worth the effort, when AF encounters "Cyrillic" as a language in a translation table, instead of moving it, could it instead mark it with {{rfc-auto}}? — Carolina wren discussió 23:33, 7 October 2009 (UTC)

This is very amusing. We encourage (specify, see WT:ELE) using "Roman" for the script rather than "Latin", to avoid this sort of confusion. But in this case, the word itself (Serbian for "novel") is "roman"! (:-)

Yes, catching "Cyrillic" as a language name (or Roman) and tagging it isn't hard. (done, with next restart sometime). Tx, Robert Ullmann 11:56, 9 October 2009 (UTC)

L4 Header "Mutation" in Irish

I would like to request to the powers that be that L4 header "Mutation" be changed to allow for L3 header. Why I ask is that in Irish, a lemma may be two parts of speech (i.e. "mol" that is both a noun and a verb) that have the same mutation; it is unnecessary to tag both "Noun: mol" and "Verb: mol" with a Mutation header, and it is wrong that one lexeme take the header when the same kind of mutation applies to both. I've had AutoFormat change L3 to L4, which makes them subject to only one lexeme.

collapsing multiple {also}s into just one

Could AF do this? --Bequw → ¢ • τ 20:06, 26 October 2009 (UTC)

Entries with nonstandard headers

I've always wondered why these need to be categorixzed by the individual headers. Most of the time, these categories never get created, or if they do they need speedy deletion afterwards. What's the advantage of this system over grouping everything together, or by language? Mglovesfun (talk) 18:06, 14 November 2009 (UTC)

Fixing broken tables

Could AF add {{trans-mid}} and {{trans-bot}} when it finds things like [5]? Conrad.Irwin 14:24, 5 December 2009 (UTC)

Cmene

Autoformat doesn't have "Cmene" in its list of accepted headers and has been adding {{rfc-header}} to them. Could this be fixed, please? --Yair rand 05:23, 27 December 2009 (UTC)

Help wanted

Hi, AutoFormat!

First, I would like to thank you for your helpful edits at for your edits at “troll”.

But it makes me wonder —

Trans links

Why are there trans links for Macedonian and N. Sami, while there are none for others?

Templates

I may be dumb — actually: I 'am, when it comes to such matters — but, despite efforts to understand, I do not know yet what are:

• {{'''t'''|language ISO|word|etc}}
• {{'''t-'''|language ISO|word|etc}}
• {{'''t+'''|language ISO|word|etc}}

Thank you forward for helping me.

Have a nice day. ৵ Kąġi Oȟąko ^Ƭ 11:11, 27 December 2009 (UTC)

"If the translated word does not exist on the foreign language wikt, use {{t-}}, if you know it does, use {{t+}}, and if the foreign wiki doesn't exist, use {{tø}}. If you don't know whether it exists, use {{t}}. In any case, these are changed by automation to correct them, so don't worry about it." (from Template talk:t).

Commonly known languages are not wikified, see Wiktionary:Translations/Wikification. Maro 16:53, 27 December 2009 (UTC)