Wikipedia:Bot requests

For the policy on bot requirements, see WP:BOTREQUIRE.

This is a page for requesting tasks to be done by bots per the bot policy. This is an appropriate place to put ideas for uncontroversial bot tasks, to get early feedback on ideas for bot tasks (controversial or not), and to seek bot operators for bot tasks. Consensus-building discussions requiring large community input (such as request for comments) should normally be held at WP:VPPROP or other relevant pages (such as a WikiProject's talk page).

You can check the "Commonly Requested Bots" box above to see if a suitable bot already exists for the task you have in mind. If you have a question about a particular bot, contact the bot operator directly via their talk page or the bot's talk page. If a bot is acting improperly, follow the guidance outlined in WP:BOTISSUE. For broader issues and general discussion about bots, see the bot noticeboard.

Before making a request, please see the list of frequently denied bots, either because they are too complicated to program, or do not have consensus from the Wikipedia community. If you are requesting that a template (such as a WikiProject banner) is added to all pages in a particular category, please be careful to check the category tree for any unwanted subcategories. It is best to give a complete list of categories that should be worked through individually, rather than one category to be analyzed recursively (see example difference).

Alternatives to bot requests

• WP:AWBREQ, for simple tasks that involve a handful of articles and/or only needs to be done once (e.g. adding a category to a few articles).
• WP:URLREQ, for tasks involving changing or updating URLs to prevent link rot (specialized bots deal with this).
• WP:SQLREQ, for tasks which might be solved with an SQL query (e.g. compiling a list of articles according to certain criteria).
• WP:TEMPREQ, to request a new template written in wiki code or Lua.
• WP:SCRIPTREQ, to request a new user script. Many useful scripts already exist, see Wikipedia:User scripts/List.
• WP:CITEBOTREQ, to request a new feature for WP:Citation bot, a user-initiated bot that fixes citations.

Note to bot operators: The {{BOTREQ}} template can be used to give common responses, and make it easier to keep track of the task's current status. If you complete a request, note that you did with {{BOTREQ|done}}, and archive the request after a few days (WP:1CA is useful here).

---

Please add your bot requests to the bottom of this page.
Make a new request

#	Bot request	Status	💬	👥	🙋 Last editor	🕒 (UTC)	🤖 Last botop editor	🕒 (UTC)
1	Bot to automatically revert date change vandalism		16	6	Pppery	2024-04-03 19:25	Primefac	2024-02-06 13:49
2	Implementing the outcome of Wikipedia talk:Naming conventions (television)#Follow-up RfC on TV season article titles	BRFA filed	18	9	Wikiwerner	2024-04-20 17:49	Primefac	2024-03-27 12:55
3	Auto-WP:NAVNOREDIRECT	Not a good task for a bot.	10	5	Wikiwerner	2024-04-28 12:22	Primefac	2024-03-13 18:37
4	Bot to add uncategorized tag to untagged uncategorized pages	Y Done	3	3	GoingBatty	2024-03-23 03:43	GoingBatty	2024-03-23 03:43
5	NFL Draft move downcasing cleanup	BRFA filed	22	4	Bsoyka	2024-03-23 02:55	Primefac	2024-03-18 10:12
6	Bot to clean up wikiproject templates		8	2	Cocobb8	2024-03-24 15:05	GoingBatty	2024-03-23 19:12
7	Update WP: maintaince pages		1	1	OrdinaryGiraffe	2024-03-21 23:43
8	IMDB Bot		1	1	BabbaQ	2024-03-29 13:27
9	Automatic NOGALLERY keyword for categories containing non-free files (again)		9	5	Wikiwerner	2024-05-15 18:43	Usernamekiran	2024-04-13 02:17
10	Automatically replace superscripts with sup and sub tags		5	3	Qwerfjkl	2024-04-06 19:54	Qwerfjkl	2024-04-06 19:54
11	Green Bay Packers draft picks (1936–1969) & Green Bay Packers draft picks (1970–present)	Y Done	3	2	Gonzo fan2007	2024-04-22 15:57
12	UTF-8 debugging		4	2	Qwerfjkl	2024-04-07 20:55	Qwerfjkl	2024-04-07 20:55
13	Long-dash URL		1	1	GreenC	2024-04-08 22:16
14	Can we have an AIV feed a bot posts on IRC?		6	2	Lofty abyss	2024-04-29 10:29	Usernamekiran	2024-04-15 11:27
15	Bot to sync talk page redirects with their corresponding page		10	5	Anomie	2024-04-17 11:40	Anomie	2024-04-17 11:40
16	Bot to update match reports to cite template		1	1	Yoblyblob	2024-04-16 13:01
17	Converting Category:Harold B. Lee Library-related articles to talk page categories	Y Done	10	3	HouseBlaster	2024-04-20 17:08	Primefac	2024-04-19 20:37
18	Bot to mass tag California State University sports seasons		2	2	Primefac	2024-04-19 18:13	Primefac	2024-04-19 18:13
19	Football league infoboxes		7	4	Bagumba	2024-04-25 13:43	Primefac	2024-04-25 12:01
20	Clear Category:Unlinked Wikidata redirects		6	3	A smart kitten	2024-04-23 10:56	DreamRimmer	2024-04-21 03:28
21	Find linkrot with a specific pattern		7	3	GreenC	2024-05-01 16:20
22	Converting positional parameters to named parameters	Done	10	2	MSGJ	2024-05-04 05:49
23	Fixing stub tag placement on new articles	Not a good task for a bot.	3	2	Paul 012	2024-05-10 14:28	Primefac	2024-05-10 11:14
24	Bot to change page links from Baetylus to Baetyl		2	2	Pppery	2024-05-12 23:45

Legend
In the last hour
In the last day
In the last week
In the last month
More than one month
Manual settings
When exceptions occur, please check the setting first.

SVG

Good day, can someone make a bot to run through this and append {{SVG-logo}} below the Non-free xxx template and add ==Summary== above the FUR template to files that don't have it? --Minorax^«¦talk¦» 11:22, 7 November 2023 (UTC)

Something like Special:Diff/1183936918 --Minorax^«¦talk¦» 11:23, 7 November 2023 (UTC)

Hey, @Minorax, sure I'll give it a shot. I'll let you know if I'm able to get something working and sent off to BRFA. Dr vulpes ^{(💬 • 📝)} 06:19, 20 December 2023 (UTC)

Disabling categories on drafts

Ever since the idea of immediately moving inadequate articles to draftspace emerged as a common alternative to deletion, the amount of time that has had to be invested in cleaning up polluted categories that have draftspace pages in them has gone way up, because the people who do the sandboxing frequently forget to remove or disable the categories in the process — so I wanted to ask if there's any way that a bot can be made to clean up any overlooked stuff.

Since there's already a bot, JJMC89bot, that detects main-to-draft page moves and tags them as {{Drafts moved from mainspace}}, the easiest thing would probably be to just have that bot automatically disable any categories on the page at the same time as it's tagging it — but when I directly approached that bot's maintainer earlier this year to ask if this could be implemented, they declined on the basis that the bot hadn't already been approved to perform that task, while failing to give me any explanation of why taking the steps necessary to get the bot approved to perform that task was somehow not an option. As an alternative, I then approached the maintainer of DannyS712bot, which catches and disables categories on drafts that are in the active AFC submission queue (which newly sandboxed former articles generally aren't, and thus don't get caught by it), but was basically told to buzz off and talk to JJMC89bot.

So, since I've already been rebuffed by the maintainers of both of the obvious candidate bots, I wanted to ask if there's any other way to either get one of those two bots on the task or make a new bot to go through Category:All content moved from mainspace to draftspace disabling any active categories, so that editors can cut down on the amount of time we have to spend on DRAFTNOCAT cleanup. If possible, such a bot would ideally also do an ifexist check, and outright remove any redlinked categories that don't even exist at all, though just disabling redlinks too would still be preferable to editors having to manually clean up hundreds of categorized drafts at a time — it's just that merely disabling the redlinks creates another load of cleanup work later on when the draft gets approved or moved by its own creator without AFC review or whatever, so killing redlinks right away is preferable to simply deferring them for a second round of future cleanup. Bearcat (talk) 16:18, 8 November 2023 (UTC)

Technically, this is doable — without interfering DannyS712bot's task. But I also would like to know why this was rejected by these two bot operators, and at the BRfA as well. —usernamekiran (talk) 18:22, 8 November 2023 (UTC)

This seems like something we can also tackle at the source. Hey MPGuy2824. Does the WP:MOVETODRAFT script disable categories when draftifying? If not we should consider adding this feature. If so we may need to look at diffs to see where these undisabled categories are coming from (manual moves? old script?) –Novem Linguae (talk) 19:02, 8 November 2023 (UTC)

The MTD script wraps categories within {{Draft categories}} which disables them. (e.g. [1]) The older script disables categories by adding a ":" before the word "Category:". That leaves two possible culprits. 1. manual moves and 2. the regex in my script isn't catching all categories. Let me see if I can narrow it down by running a quarry or two. -MPGuy2824 (talk) 04:07, 9 November 2023 (UTC)

Maybe Bearcat can provide some diffs for us to examine. Would be interesting to see if any of these are being created by MoveToDraft, or something else. –Novem Linguae (talk) 05:24, 9 November 2023 (UTC)

I've only ever seen manual moves; in fact, I wasn't aware that MTD even existed. Bearcat (talk) 12:41, 9 November 2023 (UTC)

@Bearcat: I already have a bot task approved for adding {{Draft categories}}. The challenge for me is identifying which drafts have article categories. There's Wikipedia:Database reports/Polluted categories to find user pages with article categories, but I'm not aware of anything comparable for drafts. Going through Category:Pending AfC submissions looked like an easy start, especially since it's so small at the moment, but I didn't find any drafts with article categories. I'm open to further discussion. GoingBatty (talk) 21:36, 8 November 2023 (UTC)

Wikipedia:Database reports/Polluted categories (2) catches categories with drafts in them; the minor flaw is that it's currently not correctly recognizing the {{Polluted category}} template that used to flag maintenance categories as "don't bother with this because we don't care about it", so it's picking up things it doesn't need to pick up like Category:Miscellany to be merged and Category:Wikipedia Student Program. But even if that report is failing to react to that template properly, a bot could potentially be programmed to react to that template around whatever's stopping that report from reacting to it.

In the case of the particular issue I was asking about, just working directly with Category:All content moved from mainspace to draftspace itself is also an option: have a bot go through that, and disable categories that are on the pages in that category. That won't catch all categorized drafts by itself, but it will certainly catch the ones that are categorized because they're former articles that got moved into draftspace without the mover disabling the categories in the process — and at this point, that accounts for the majority of categorized drafts, so it would become easier for human editors to catch whatever's still left if we only have to deal with 25 or 30 per cent as many pages as we do now.

I genuinely doubt that there's any way to make a bot perfect at catching all improperly categorized drafts without ever missing any — but if we can get bots to deal with as many as possible, that still reduces the amount of time that human editors have to invest in worrying about it. So I don't think we need to shoot for "the magic bullet that will make a bot infallible at instantly catching every draft that ever gets categorized at all" — let's just aim for "where can a bot make as many dents in the problem as feasibly possible by working on defined targets". Bearcat (talk) 22:48, 8 November 2023 (UTC)

@Bearcat: Aha - thanks for telling me about that report! I ran the bot over the report, and manually cleaned up some drafts that had incorrect categories. I've added it to my favorites, so I can run the bot when it gets republished. Thanks! GoingBatty (talk) 04:16, 9 November 2023 (UTC)

Ah, ok. I think the reason you didn't know about it might be that the report you already knew about used to catch both user-polluted and draft-polluted categories in the same place — but then they were split up into two separate reports later on for whatever reason, so you might simply never have found out about the newer draft report. Bearcat (talk) 12:41, 9 November 2023 (UTC)

If you don't want to have your bot churn through 8000 pages most of which don't need action, then Category:Content moved from mainspace to draftspace from November 2023 (using the current month instead) would also work fairly well. * Pppery * _{it has begun...} 23:32, 8 November 2023 (UTC)

Actually, yeah, that's a good alternative too. In reality, categorized drafts will virtually always be new pages that became categorized drafts within the past couple of days (and humans can catch the less common exceptions where a much older draft gets recategorized more than a month later), so the dated categories are likely more manageable chunks for a bot to grind through. Bearcat (talk) 23:48, 8 November 2023 (UTC)

• The other thing is, just having a bot go through the polluted category reports doesn't solve the problem all by itself — sometimes drafts are categorized not by direct declaration of categories on the page itself, but by artificially smuggling in transcluded categories from a template or infobox. (There is a rule that templates aren't supposed to be transcluding categories at all, but people don't actually obey it, so in reality it happens quite frequently anyway.)
  A bot can't fix cases like that, which means that even if a bot is pounding through that report a human still has to go through it every few days anyway to suppress any transcluded categories that the bot couldn't fix — but if most of the categories on the report have already been cleaned up by a bot, but I still have to inspect all of them anyway to look for one or two pages that might have been missed, then that's a huge imposition on my time. So what we also need is a bot that's catching and killing categories on drafts before they even get picked up by that report at all, so that the size of the report itself is reduced — which means that there does need to be a bot that whips through Category:All content moved from mainspace to draftspace, or a dated subcategory of that, on a daily basis so that a larger percentage of categorized drafts get resolved before they even show up on the polluted categories report in the first place. Bearcat (talk) 13:43, 9 November 2023 (UTC)

  Does something like searching the Draft namespace with insource:/\[\[\s*[Cc]ategory\s*:/ -hastemplate:"Draft categories" -insource:/\[\[\s*[Cc]ategory\s*:[^]]*[Dd]raft/ or a variation do the job? Wikiwerner (talk) 20:48, 2 December 2023 (UTC)

Major task: Update Template:Ct to Template:UCI team code

Please see the Template_talk:UCI_team_code#Requested_move_30_October_2023 where there was rough consensus to usurp the ct shortcut. Essentially, replace the 12,000 transclusions to bypass the redirect so that Template:Ct can redirect to Template:Contentious topics. Awesome Aasim 23:08, 9 November 2023 (UTC)

Running through the 98 remaining templates that transclude {{ct}} might make a nice dent in the overall page count. I did eight or ten of the core ones that needed human attention. – Jonesey95 (talk) 00:07, 10 November 2023 (UTC)

Have worked on the rest of the templates with AWB. – robertsky (talk) 00:47, 12 November 2023 (UTC)

BRFA filed. – robertsky (talk) 00:54, 12 November 2023 (UTC)

Fixing duplicate banners

I asked at the help desk and I was told to ask here. The Organized crime task force and the Serial killer task force banners recently got added to the banner of Template:WikiProject Crime and Criminal Biography using parameters. The previous banners (as wrappers of the new one with the parameters) were mass substituted. This has left ~6700 (see Category:Unknown-importance Crime-related articles, not counting ones that didn't have an initial basic crime importance) duplicates, that have both the original crime importance and task force importance but split between two duplicate banners.

Is there any bot that can merge the importance values on the pages that have both templates so there aren't so many duplicates (for example if there's two duplicate banners, one of which has the importance for wp crime and one which has the task force importance, add them together)? Of course the ones that were not initially tagged with the original crime ones will have to be manually tagged as they don't have the basic importance parameter, but that's less than 500 which isn't as bad (compared to 6700 that already HAVE all the required importance parameters) PARAKANYAA (talk) 17:56, 11 November 2023 (UTC)

@PARAKANYAA: I just ran my bot against Category:Pages using WikiProject banner shell with duplicate banner templates, which edited about 550 talk pages (many WP:CRIME related) to remove a WikiProject template only if every parameter is included in a duplicate WikiProject template on the same page. You'll need someone else to clean up those where the two templates have different parameters that need merging. GoingBatty (talk) 06:16, 12 November 2023 (UTC)

@GoingBatty Someone else meaning it must be done manually or someone else's bot? PARAKANYAA (talk) 04:03, 16 November 2023 (UTC)

@PARAKANYAA: Other bot operators may be able to fulfill your requests. Doing things manually is also an option. GoingBatty (talk) 04:17, 16 November 2023 (UTC)

Well, I sure hope that someone's bot can... that would take a while. PARAKANYAA (talk) 04:21, 16 November 2023 (UTC)

to elaborate further on the problem and what I imagine the solution to be

a hypothetical bot should merge

{{WikiProject Crime and Criminal Biography|serialkiller=yes|serialkiller-imp=low|organizedcrime=yes|organizedcrime-imp=low}} (or having 1 or the other task forc parameters, just showing both for sake of example)

on the same page as either

{{WikiProject Crime|importance=low}} OR {{WikiProject Criminal Biography|importance=low}} (also called WikiProject Criminal which iirc has quite a few transclusions)

would combine to be

{{WikiProject Crime and Criminal Biography|importance=low|serialkiller=yes|serialkiller-imp=/nowiki>low|organizedcrime=yes|organizedcrime-imp=low}}

95% of the articles I've seen that have a crime duplicate have either the basic crimebio banner or normal crime (which doesn't matter, Crime/Crimebio/Crime and Crimebio are all the same now) with a crime importance parameter, and a separate banner with either the serial killer importance or the organized crime importance that lacks the crime project importance, so wiki categorizes it as unknown importance (though strangely WP 1.0's bot does not) PARAKANYAA (talk) 04:43, 21 November 2023 (UTC)

Y Done. Primefac (talk) 12:58, 24 December 2023 (UTC)

Date ranges for noticeboard archives

This seems uncontroversial.

It would be pretty cool if some kind bot could go through the pre-current (should never change) archives of the boards listed in the dramaboard archivebox series, extract the earliest and latest timestamps, truncate them to dates, and use those dates to annotate the links somehow. Inactive archives at time of writing are:

• 355 at Wikipedia:Administrators' noticeboard/Archives
• 1140 at Wikipedia:Administrators' noticeboard/IncidentArchives
• 474 at Wikipedia:Administrators' noticeboard/3RRArchives
• 323 at Wikipedia:Arbitration/Requests/Enforcement/Archive

User story: I was recently trying to find an archived conversation from a few months ago, and the best tools I had available were a scattershot "tap an archive number, wait for the entire page to load, check top and bottom timestamps" and "search archives for exact string matched date". Improved navigability gained from annotating the archive links with date ranges should save people time.

Implementation ideas: The quickest implementation would just be a plaintext date range edited onto the archive list pages linked above. A further step could be to add a |date-span= (or similar) to {{Administrators' noticeboard navbox all}} which, if present, would display the date range of comments posted at the top of the page itself, so the information is available both on the archive page and the index of archives. The most elegant, stupid, and expensive implementation would be to add {{shortdesc}} to all the archives, set the |1= to the date range, and convert the indices to use {{annotated link}}.

Anyway though: Anyway though the first step is getting the date ranges. Maybe this is already in a report somewhere? Folly Mox (talk) 18:30, 11 November 2023 (UTC)

@Folly Mox, I suspect you'd need consensus to go through with this. Perhaps trying asking at those noticeboards first? — Qwerfjkltalk 19:06, 11 November 2023 (UTC)

Oh right I tacked on all those expanded scope ideas in the process of making the edit. Folly Mox (talk) 20:22, 11 November 2023 (UTC)

Just noting, an Index of some variety would probably be easier than going through the thousands of archives and amending them. Primefac (talk) 08:10, 12 November 2023 (UTC)

Would this need a bot?

Does the task at Wikipedia:Village pump (technical)#Implementation of Template:Refideas editnotice require a bot, or is there another way to accomplish that? You can respond there if you like. BOZ (talk) 05:42, 14 November 2023 (UTC)

Automatic NOGALLERY keyword for categories containing non-free files

I have noticed that many categories, especially content categories, include non-free files without the __NOGALLERY__ magic word, which is against WP:NFCC#9. I'd suggest using a bot to auto-tag such categories, skipping a whitelist for those categories covered by WP:NFEXMP (generally those categories concerning reviews of questionable files, such as CAT:FFD, and some maintenance categories that should contain no non-free files). –LaundryPizza03 (dc̄) 12:01, 14 November 2023 (UTC)

I'm pretty sure that most categories on en.wikipedia should have __NOGALLERY__ and it would be actually smarter and less work to disable image showing on all categories by default (without requiring any code per page or bot work) and have a __YESGALLERY__ magic word for the much less instances of categories that actually could show images. Gonnym (talk) 12:38, 14 November 2023 (UTC)

I doubt that would be possible, since as a MediaWiki tweak that would have to apply to all Wikimedia wikis, many of which don't allow non-free files. –LaundryPizza03 (dc̄) 02:48, 15 November 2023 (UTC)

From a coding perspective it could just be a setting that can be activated per wiki. I'd be opposed to any bot before other solutions are researched. --Gonnym (talk) 06:23, 15 November 2023 (UTC)

@Gonnym: Might want to report this request at https://phabricator.wikimedia.org/? I'd still like to hear from other people on whether a bot would be a good solution in the short term. –LaundryPizza03 (dc̄) 01:22, 15 December 2023 (UTC)

Archiving washingtonindependent.com

See Wikipedia:Reliable sources/Noticeboard#Washington Independent.
But archiving those links is not quite straightforward. We should probably get rid of any link to the live domain (which is garbage) and we should only use archive.org snapshots that are older than, say, 2016. When there is no older snapshot, the link/reference should be removed entirely. — Alexis Jazz (talk or ping me) 01:34, 15 November 2023 (UTC)

I think WP:URLREQ is the correct board for this. I think GreenC has a bot that can easily fix usurped domains en masse, and that is the board that they monitor for this. –Novem Linguae (talk) 07:17, 15 November 2023 (UTC)

Novem Linguae, thanks: Wikipedia:Link rot/URL change requests#washingtonindependent.com — Alexis Jazz (talk or ping me) 14:40, 15 November 2023 (UTC)

Deferred. Queen of Hearts ❤️ (no relation) 03:19, 19 November 2023 (UTC)

Implement project-independent quality assessments

Looking for a willing bot operator to implement WP:PIQA by migrating quality assessments from WikiProject banners into {{WikiProject banner shell}}. To be more precise,

1. If there is a banner shell already on the page, then add |class= parameter and remove from project banners, e.g. [2]
2. If there is no banner shell, then add it and move class rating from project banners, e.g. [3]
3. If there are no assessments on page, then add empty |class= parameter to encourage editors to add a rating, e.g. [4]
4. If assessments of projects differ, then add the majority rating to the banner shell and leave any different assessments on those banners, e.g. [5]. These will be manually reviewed by human editors.
5. If assessments of projects differ, but there is no majority rating, then add banner shell with empty |class= parameter. These will be tracked and reviewed manually.
6. If the page has {{WikiProject biography}} with |living=yes or |blp=yes then add |blp=yes to {{WikiProject banner shell}}.
7. If any project banner has |listas= then move this to {{WikiProject banner shell}} and remove from project banners, e.g. [6]
8. For any of the projects which have opted out, the class parameter should not be changed or removed.

Thanks in advance — Martin (MSGJ · talk) 22:18, 15 November 2023 (UTC)

@MSGJ: How should the bot (or a manual user) handle a banner shell with two {{WikiProject Articles for creation}}, like Talk:1975–76 Notre Dame Fighting Irish men's basketball team? GoingBatty (talk) 04:47, 16 November 2023 (UTC)

For WikiProject tags that have different ratings but located on the same talk page, perhaps the bot should note these on a log page somewhere, then skip them. –Novem Linguae (talk) 05:31, 16 November 2023 (UTC)

I've previously tested code at [7] that could add additional parameters to handle this situation. Gonnym (talk) 06:25, 16 November 2023 (UTC)

We are tracking these (and many others) at Category:Pages using WikiProject banner shell with duplicate banner templates but have not yet made any serious effort to fix them. Last time I asked the AfC project, they were happy for any duplicates to be removed and I would suggest leaving the one with the most recent time stamp — Martin (MSGJ · talk) 07:57, 16 November 2023 (UTC)

Uh...false. We don't want duplicates removed, we indicated that they should be merged so that no information is lost. Primefac (talk) 13:01, 16 November 2023 (UTC)

Novem Linguae said "You can just use your best judgment on which one to save" and you agreed with that ... but okay if you want to merge/consolidate somehow then go ahead — Martin (MSGJ · talk) 13:44, 16 November 2023 (UTC)

NL also said it'd be fine to boldly consolidate duplicate AFC banners on article talk pages. There are two possibilities - actual duplicated banners (in which case one should be removed) and two different banners, which should be merged, to best save the review history. Primefac (talk) 13:48, 16 November 2023 (UTC)

@MSGJ: How should the bot (or a manual user) handle a banner shell with two different {{WIR}} templates, like Talk:Ruth M. Davis? GoingBatty (talk) 04:52, 16 November 2023 (UTC)

These two WIR templates don't appear to have any class or importance specified, so I don't think it would confuse the bot. –Novem Linguae (talk) 05:39, 16 November 2023 (UTC)

Agree with Novem Linguae. Would like to get these merged one day, but hopefully will not affect this task and this task is already complicated enough! — Martin (MSGJ · talk) 07:55, 16 November 2023 (UTC)

Hopefully when running, the bot can also preform User:Magioladitis/WikiProjects. Gonnym (talk) 15:04, 16 November 2023 (UTC)

@MSGJ, BRFA filed. — Qwerfjkltalk 18:07, 19 November 2023 (UTC)

Convert links to Wikimedia sites in body text to normal interwiki links

A lot of pages have links to Wiktionary pages in the body text. This is fine, though I think the links are supposed to be like this (interwiki) and not this (external).

Would it be possible to create a bot that turns these into the first example? LOOKSQUARE (👤️·🗨️) talk 21:05, 16 November 2023 (UTC)

I'm curious why you're using the heavy formatting. Your two links are this and this without them, which to me appear exactly the same. Primefac (talk) 21:07, 16 November 2023 (UTC)

oops it didn't work 💀

this diff is what I meant LOOKSQUARE (👤️·🗨️) talk 21:22, 16 November 2023 (UTC)

You mean this: Special:Diff/1162303860/1185448120 -- GreenC 21:49, 16 November 2023 (UTC)

*falls over on the floor*

*life support unplugs*

yes LOOKSQUARE (👤️·🗨️) talk 21:57, 16 November 2023 (UTC)

@LOOKSQUARE: Like this edit and this edit for the English Wiktionary, and this edit for the German Wiktionary? GoingBatty (talk) 23:00, 16 November 2023 (UTC)

As the section heading implies, there are also plenty of "external" links to en.wikipedia.org/.../Article_title, and presumably the mobile and non-English variants too. (A tiny number of them may be valid primary-source citations of Wikipedia in articles about Wikipedia.) Certes (talk) 23:12, 16 November 2023 (UTC)

all by itself, it is cosmetic editing. —usernamekiran (talk) 23:08, 16 November 2023 (UTC)

oh ok nvm LOOKSQUARE (👤️·🗨️) talk 23:13, 16 November 2023 (UTC)

I'm not sure it is merely cosmetic. It's replacing external links by wikilinks, which has a positive visual effect and clarifies that the target is internal to the WMF ecosystem and possibly this wiki. Certes (talk) 23:38, 16 November 2023 (UTC)

I agree, it's not cosmetic. There are only around 2,000, not all should be converted. -- GreenC 02:19, 17 November 2023 (UTC)

@GreenC: Where should consensus be built to decide which should be converted and which are acceptable as is? GoingBatty (talk) 03:40, 17 November 2023 (UTC)

@GreenC: Your search is only for en.XXXX.org. There are links to many other languages that could be changed (e.g. de.wiktionary.org). Expanding to this search gets us to over 4,200. GoingBatty (talk) 03:44, 17 November 2023 (UTC)

Add "wikipedia" and that is another ~500 results. Gonnym (talk) 06:54, 17 November 2023 (UTC)

yes, just saw this conversation/diffs from computer again, this is not cosmetic. —usernamekiran (talk) 02:42, 17 November 2023 (UTC)

LOOKSQUARE, consider using a gadget or userscript to take care of this. [spam] User:Alexis Jazz/Factotum converts many external links automatically to wikilinks. Special:Diff/1185452015, Wikipedia:Bot requests (history), m:User:LOOKSQUARE, etc. is just me wikt:en:copy-pasting URLS. If you enter an additional custom regex in the settings it could also update existing named links. (normally it avoids named links)[/spam]
As for having a bot do this: hmmm I don't know. You need to be very careful as sometimes it's on purpose, especially when there are extra parameters like ?useskin=monobook or ?uselang=de. But it's also used to show the difference between Wikipedia's mobile site and the desktop site. If the URL is a template parameter value you also shouldn't change it. It may seem easy but there are quite some edge cases. — Alexis Jazz (talk or ping me) 07:00, 17 November 2023 (UTC)

Yeah, maybe there could be some kind of template, similar to {{Not a typo}} so that bots didn't accidentally convert an external link to normal. LOOKSQUARE (👤️·🗨️) talk 10:35, 17 November 2023 (UTC)

@Alexis Jazz: Are those edge cases in articlespace? I'd imagine there could be many in Help: and/or Wikipedia: and/or talk pages. GoingBatty (talk) 16:49, 17 November 2023 (UTC)

We have a few, such as the last item in List of Wikipedia controversies#Further reading. Of course, there are many Wikipedia URLs within citations (some legitimate, others unreliably circular sourcing) which I assume are not covered by this exercise. Certes (talk) 17:07, 17 November 2023 (UTC)

@Certes: Thank you for providing the example. Why shouldn't that link be changed to Wikipedia:Wikipedia as a court source to make it obvious that it is a page that exists within Wikipedia? GoingBatty (talk) 17:11, 17 November 2023 (UTC)

Good question. It probably should. My usual yardstick is "if we created Forkpedia, would we want this link to lead to Forkpedia or to Wikipedia"? If Forkpedia then it should be an internal link; if Wikipedia then an external link. I think my example is a misplaced "See also" entry rather than further reading, so it should be internal. There's a better example in Circular reporting#Examples on Wikipedia. The 2008 link is about Wikipedia (and so would be external on Forkpedia) but it has an oldid= URL parameter and we can perhaps exempt those from any bot. Certes (talk) 17:28, 17 November 2023 (UTC)

Other examples about Wikipedia include List of Wikipedia pages banned in Russia. We also have technical articles where Wikipedia is used as an example website, e.g. URL redirection#Redirect chains, HTML#Elements. Certes (talk) 17:38, 17 November 2023 (UTC)

Yes, I meant that it would only replace Wikimedia links in the article text, not the "Further reading" or "References" sections. LOOKSQUARE (👤️·🗨️) talk 18:48, 17 November 2023 (UTC)

I wonder if the task can be split into two parts, one suitable for a bot and the other needing manual help with AWB or similar. Certes (talk) 18:59, 17 November 2023 (UTC)

Replacement for MalnadachBot?

Not sure if this is the right place to ask this, but I noticed that when User:MalnadachBot was procedurally blocked, its tasks 12 and 13 were still marked as active. I don't know if it has finished running through all lint errors on wiki, but task 13 is certainly an ongoing effort. Would we need a replacement bot to pick up these tasks, or do we have existing bots/procedures handling these things? Liu1126 (talk) 20:17, 19 November 2023 (UTC)

@Liu1126, this was previously discussed at Wikipedia:Bots/Noticeboard/Archive 18#MalnadachBot -- owner indeffed, what now? — Qwerfjkltalk 21:49, 19 November 2023 (UTC)

Thanks, I somehow failed to find that discussion when searching the archives earlier. I was sure I wouldn't be the first person to notice this! Liu1126 (talk) 09:27, 20 November 2023 (UTC)

Translate Article Pages from English to Gagana Sāmoa

I am in the midst of translating articles from English to Gagana Sāmoa which is a very necessary task given the massive inequality in the available information in each language. My hope is that more Sāmoa users will find Wikipedia to be a more hospitable site and access it to find information in their language. There is a massive disparity not only between English and Gagana Sāmoa but even between other languages and the languages of the Pasifika by and large. I would like to request a bot to help translate these articles as this task is overwhelming and this disparity will only grow given the population, internet access, and specialization of Sāmoa users. Something has to be done otherwise the language will likely go the way of 'Ōlelo Hawaiʻi. IonaPatamea (talk) 20:57, 20 November 2023 (UTC)

Machine translations usually do a poor job. From other languages to English, they may create fluent-sounding text, but when you read it a little more closely, it doesn't make sense and contradicts itself. I imagine the same is true from English to other languages. Are you sure you want to do a bunch of mass translations on your wiki? –Novem Linguae (talk) 21:02, 20 November 2023 (UTC)

Yes, I completely understand the concern. My only qualm is that there is a dearth of users translating pages to Gagana Sāmoa and my concern is that this is a collective action dilemma, or something like it, insofar as Sāmoa users will not necessarily use or contribute to Wikipedia until there is a sufficient amount of material available in the language. I feel that once that threshold is crossed there will be more incentive to both contribute, edit, and admin articles. Although I appreciate the concern for accuracy, which I also share, I think that the situation of the disparity between English and Gagana Sāmoa as sources of knowledge is not truly appreciated by those unfamiliar with the context. Sāmoa users are shifting their sources of information, and thus knowledge, from Gagana Sāmoa to English on an unprecedented level and this shift is having an unprecedented effect on the social structure, cultural continuity, and even viability of the Gagana Sāmoa itself which once lost will likely disappear for future generations. So, for the short-term, and in short, I feel that using machine translations are the best solution and I would like to use them in lieu of other options, which I must admit, I am not familiar with. Perhaps there are other suggestions? If not I would argue machine translation would be a good option. IonaPatamea (talk) 21:23, 20 November 2023 (UTC)

@IonaPatamea: Hi. We had similar issues on Marathi Wikipedia, and we organically reached 70k articles a couple of years ago. Currently we have around 75k articles, but most of these articles are promotional. Marathi has huge number of native users in the world — it's among the top languages with rich history, and still, very few are genuinely interested in Wikipedia. —usernamekiran (talk) 21:51, 20 November 2023 (UTC)

Thanks for the message! How was this content in Marathi generated organically if I may ask? Was there a concerted effort to get users to contribute or was this something that happened over time? Yes, as you stated the community for Sāmoa is much smaller. I hope that more content for Marathi will continue to be generated. Although I cannot speak the language it undoubtedly has a rich history and much to offer users. IonaPatamea (talk) 07:08, 21 November 2023 (UTC)

Everyone's tolerances of machine translations differs. I would argue that machine translations can be used as a basis of writing new articles, but whatever that's generated from these translations should be vetted through, and edited, before publishing.

As for requesting for a bot that will ultimately affect Gagana Sāmoa wiki, this is not the right venue since the bot requests here are generally for edits on enwiki.

Alternatively what you can do is to create stub articles like what the Minnan wiki is doing and hopefully one day someone will come by to expand the content. In this manner, you can generate interests when getting press attention (X wiki has now N number of articles in effort to <do something with the language>) and as well as having more existing content as start points for anyone else whose interested in that particular topic while having a mild interest in the language. – robertsky (talk) 05:05, 21 November 2023 (UTC)

Talofa! Thank you for your response and very helpful suggestion. I believe this sounds like a good way to generate, and sustain, enough interest for users. I appreciate the help and I will be sure to look into Minnan wiki. Fortunately, I am currently at one place where there are academics that teach Gagana Sāmoa at the tertiary or university level so they may be able to help with generating interest and press attention. The stub articles would be an efficient way to start getting more information out to contributors, academics, and the press alike. IonaPatamea (talk) 07:12, 21 November 2023 (UTC)

author-link

We have a lot of citations that could be improved using |author-link= eg. Special:Diff/1186331321/1186342802. The problem is it's difficult to match the correct author, it requires a human. Thus wondering if/how this might be automated in certain cases. It doesn't require every case only those it can match with greater certainty. For example we know, per the above diff, there is only one Steven Poole there is no dab page. And we know Steven Poole writes for a publication called Quercus. Thus any other cites that match those criteria, is a good bet that is the same person, and where an |author-link= could be added.

Is this method 100% foolproof? Probably not, but is it at least 99% accurate in matching names? Probably. I think a test run would show how reliable it is. I don't have the time right now but wanted to mention in case anyone wants to run an experiment. Or had other ideas. A dump of CS1|2 citations on enwiki - not including cite web - can be found here. I currently have updates disabled, but can restart if anyone wants. -- GreenC 20:27, 22 November 2023 (UTC)

One could involve Wikidata: if the corresponding item is a writer, then the author link can be added. Wikiwerner (talk) 20:58, 22 November 2023 (UTC)

I don't think that will help that much. Many authors cited on Wikipedia aren't "writers": probably as many or even more are scientists of some vein.

I'm not sure I have a better idea how to automate this though. Apparently Growth Team have a pretty good algorithm for suggesting appropriate wikilinks given topic and text, and I doubt their application will interface with template parameters, but solving the "what should link here" problem – of which this is a subtype – is not easy. Folly Mox (talk) 14:57, 1 December 2023 (UTC)

For a first pass, you might limit the suggestions to publications mentioned in the author's article. Perhaps exclude citations, so we don't link to a footballer of the same name whose transfer was reported in another edition of the same newspaper. Certes (talk) 15:21, 1 December 2023 (UTC)

I think look at existing uses of |author-link= and build a 2-column database: "Steven Pool = Quercus". Then find all other citations that cite Steven Pool and Quercus, and add the |author-link= if missing. It works backwards from what is know to be true. -- GreenC 19:42, 5 December 2023 (UTC)

Good plan. For extra safety, any names with two author-links that don't redirect to the same article can be filtered out into a log file without attempting to fix them. There probably won't be many. Certes (talk) 23:00, 5 December 2023 (UTC)

I really like this idea. Folly Mox (talk) 14:29, 6 December 2023 (UTC)

Incidentally, I did ask whether the future Suggested Links task will interface with citation template parameters, and although I got a response I'm still not sure. The conversation is at :mw:Talk:Growth/Feature summary § Suggested Links and interaction with non-prose text, in case anyone is interested. Folly Mox (talk) 13:20, 14 December 2023 (UTC)

Extending copyright expiration date on Pablo Picasso's works from 70 years (January 1, 2044) to 80 years (either April 8, 2053 or January 1, 2054)

I have seen Pablo Picasso's files on English Wikipedia that his works will be transferred to Wikimedia Commons on January 1, 2044, 70 years after his death. However under Spanish copyright law, the copyright term for Spanish authors who died before December 7, 1987, including Picasso who died on April 8, 1973, have life term plus 80 years, and for those who died otherwise have life term plus 70 years, though it is unclear if the copyright expires on his 80th death anniversary or January 1 following it. So for sure, there is no room to have his copyright expired on January 1, 2044. It's either April 8, 2053 or January 1, 2054. I am hoping for a consensus if his copyright expired on these bolded dates. Here are his works to have his copyright expiration date edited or added (if not) one-by-one. Ishagaturo (talk) 09:07, 23 November 2023 (UTC)

A better forum for this question is Wikipedia:Media copyright questions. Suggest moving this discussion there. (FWIW I think copyright on Wikimedia servers follows US law because that's where the content resides.) -- GreenC 16:24, 23 November 2023 (UTC)

Not on Commons, though, where all the files will be deleted immediately after transfer. Adam Cuerden ^(talk)_{Has about 8.6% of all FPs.} 20:01, 23 November 2023 (UTC)

Early idea

"The three most dangerous things in the world are a programmer with a soldering iron, a hardware type with a program patch, and a user with an idea."

— Rick Cook, The Wizardry Consulted

So I have an idea, and...

Is it possible for a bot to find articles that:

• contain a number of words of readable prose (e.g., as calculated by Wikipedia:Prosesize) between x and y (e.g., 150 and 2,000) and also
• contain fewer internal links per word than some simple mathematical formula (e.g., "less than three total" or "less than one link per 100 words")?

If a bot could automatically detect such articles, then I'd like to have it add the {{underlinked}} template, on a schedule of perhaps a few articles being tagged per hour, to feed the seemingly popular Category:Underlinked articles for the Wikipedia:Growth Team features, without giving a large number of articles to the first editor and then leaving none for anyone else.

I realize that this would require a demonstration of consensus, but I don't want to make the suggestion, get people's hopes up, and then find out that bots can't count the number of words or links in an article. WhatamIdoing (talk) 22:21, 30 November 2023 (UTC)

The tagging part seems like a CONTEXTBOT problem, but I can imagine a bot-generated report that listed the 1,000 articles with the fewest links per 100 prose words. Humans could then look through the report and refine the bot's criteria. If somehow the bot can be made to correctly identify underlinked articles without false positives, tagging might be possible. – Jonesey95 (talk) 01:22, 1 December 2023 (UTC)

Related question, for WhatamIdoing: What can we do to provide feedback about this newcomer linking activity? I see that at least some edits are adding undesirable disambiguation links. Is the tool suggesting these links? – Jonesey95 (talk) 01:25, 1 December 2023 (UTC)

The tool doesn't suggest any specific link. It pops up a box that says things like "All you need to do is add one or two links to make a difference." They're using the visual editor, so its link tool will de-prioritize (in the search results) and label dab links (so you can see that it's not the kind of page you were expecting). However, it doesn't tell you what a redirect points to, so if you have a redirect to a dab page, then you'll see 'redirect' and not know that it's a redirect to a dab page. WhatamIdoing (talk) 02:25, 1 December 2023 (UTC)

It might be worth mentioning that the Suggested Links task, scheduled to be enabled on en.wp next year, does not use maintenance templates as an inclusion criterion. So if timetables are not further extended, any success in this effort will apply only to a few months of newcomer activity. Folly Mox (talk) 15:01, 1 December 2023 (UTC)

Is word-counting and link-counting a realistic task, then? WhatamIdoing (talk) 02:28, 1 December 2023 (UTC)

WhatamIdoing, I don't know of any way to do it using queries, but running a bot on a database dump probably wouldn't be that hard. — Qwerfjkltalk 14:23, 1 December 2023 (UTC)

From a technical perspective, word counting and link counting are pretty straightforward to do. I explained how to implement prosesize word counts on my blog a while back, and that technique is used to power, among other things, Wikipedia:Database reports/Featured articles by size. Link counts are a simple database query or extraction from page HTML/wikitext. Unfortunately much of this work is blocked on the fact that the HTML dumps are currently created using proprietary source code. Legoktm (talk) 07:20, 5 December 2023 (UTC)

Thanks for that, @Legoktm. It sounds like word-counting could be done "today" (i.e., by adapting existing code). I'm not sure how to summarize what you said about link-counting. On the one hand, you say it's a "simple" query, but on the other hand, that it's blocked.

Is the database report for FAs the size at time of promotion, or the size today? Tpbradbury had been looking into that recently. (He's been hoping to find out whether there was a trend in FA size over time.) WhatamIdoing (talk) 17:58, 5 December 2023 (UTC)

The report I threw together shows the size at time of promotion, as requested. There may be other reports based on current size. Certes (talk) 18:56, 5 December 2023 (UTC)

Sorry, I wasn't very clear @WhatamIdoing. Individually, getting an article's prose size and link count is simple. Finding articles out of the entire wiki that meet those criteria isn't really feasible right now because of the lack of HTML database dumps. So if there's some other way to limit the number of articles to check, e.g. just looking at a few categories, that's probably doable. Legoktm (talk) 07:43, 7 December 2023 (UTC)

If I wanted to get this for, say, all the articles in Category:WikiProject Medicine articles, excluding articles in Category:Society and medicine task force articles, then it sounds like we (i.e., you/someone/not me) could make a one-time report that lists each article and the number of words and links in it, but an ongoing "monitoring" process would be less feasible. Am I closer to understanding this now? WhatamIdoing (talk) 18:54, 8 December 2023 (UTC)

@WhatamIdoing: I just tried writing some code and seeing how it went and here you go: User:WhatamIdoing/Possibly underlinked medicine articles (feel free to move elsewhere, etc.). It took about 15 minutes to generate the listing; definitely surprised me how fast it was. So I think it's fine to run as a regular thing on categories of roughly that magnitude, probably weekly?

I used the criteria you suggested at the beginning (between 150-2000 words and less than 3 links total), but we can change that based on what you find useful without really affecting the runtime.

So let me know if that list of articles is useful and if/how you'd like to move forward :) Legoktm (talk) 04:47, 16 December 2023 (UTC)

@Legoktm, thank you for this. Could you take a look at Motoric Cognitive Risk and Igor Smirnov (engineer), which are reported as having two links, but appear to have none? WhatamIdoing (talk) 02:56, 17 December 2023 (UTC)

I tagged all nine of the identified articles as being underlinked. Less than 48 hours later, all of them were improved, and all of the tags had been removed by the (apparently many) watchers of the category. WhatamIdoing (talk) 22:12, 18 December 2023 (UTC)

Thanks for pointing out those two articles - I found a bug in my code, so I re-ran the report and surfaced ~280 more articles that meet the criterion and have updated the listing. Should I turn this into a regular database report that updates every week? Legoktm (talk) 06:18, 29 December 2023 (UTC)

Probably not yet. Why is Academy of Medical Royal Colleges and Faculties in Scotland in the list? WhatamIdoing (talk) 06:35, 29 December 2023 (UTC)

List items don't count for prose size (see this), so those links aren't being counted. Legoktm (talk) 06:54, 29 December 2023 (UTC)

I'm not sure that using Prosesize's limitations makes sense for counting links. It's a bit tricky, because you don't want a "List of 100 most popular songs", with links to 100 articles about songs, in this list (and Prosesize will do that), but you also don't want a medium-sized ==See also== section to exclude an article with completely unlinked text (and counting all links in list-formatted text will do that).

Unless you have ideas about how to get around this, it's possible that this is not a suitable task for full bot automation. WhatamIdoing (talk) 18:03, 29 December 2023 (UTC)

Right, I think there are pros and cons both ways. The initial list missed a number of articles that were underlinked because the link was in a reference or something.

I'm also not interested in pursuing full automation, I think, if you find the list useful (but not necessarily perfect), to just regularly generate the list and let humans review it and decide whether to tag it or not. Legoktm (talk) 23:53, 29 December 2023 (UTC)

Protection fallback adminbot

Simple idea: monitor the protection log, and any time the protection level is increased, but the expiration time is decreased, wait until a few minutes before the expiration, and restore the status quo. If it really is the intention of the protecting admin to leave the page unprotected at expiry, they can leave a keyword like NOFALLBACK or something in the protection summary. An obvious complication would arise if the bot is lagging, and some edits slip in before protection can be restored, but that's a minor detail. Yes, I know about the PC trick, but people sometimes forget, and sometimes PC is isn't enough. Suffusion of Yellow (talk) 03:58, 2 December 2023 (UTC)

From WP:ADMINBOT this needs a wider discussion on WP:AN or WP:VP, though I think this is a good idea. 0xDeadbeef→∞ (talk to me) 06:08, 2 December 2023 (UTC)

Wouldn't it be best to find someone willing to operate the bot before proposing it to the wider community? Suffusion of Yellow (talk) 19:39, 2 December 2023 (UTC)

It's a chicken-and-egg problem, right? Operators don't want to invest time until there's consensus, and it's a waste of time to determine consensus if there's no operator...

It would be nice if we had a collective group of admin bot operators, so we don't need to rely on one single person volunteering, i.e. I'd be happy to contribute and work on such a protection bot but it would not be a good idea for me to do it alone.

P.S. Might be be easier to find an operator if this was a blue link. Legoktm (talk) 05:26, 16 December 2023 (UTC)

+1— Qwerfjkltalk 22:00, 16 December 2023 (UTC)

Subtle hint there, Legoktm. I like it. @Suffusion of Yellow: feel free to take them up on that advice . More admin bot ops couldn't hurt. As for this idea, that does definitely seem like a doable one (technically speaking), though I am not entirely sure at just past 1am local time (my timezone) how I would implement that...but TheSandBot is an adminbot and I would be happy to entertain the idea.

I have played around with monitoring event logs on Commons, so I am sure I could find the log. The main question would honestly be keeping track of the changes and how to jump/kick-start the system. Though, as I think this through writing this, I guess it wouldn't necessarily need to have data pre-populated as it could just create entries etc (i.e. in a database) based on protection level changes that come in. Before this went to a community discussion, I would definitely like to hammer out some of the details/concepts before attaching my name to it as a bot-op. If you have any thoughts on either implementation or conceptually how this would work further, I am definitely all ears (seriously/no sarcasm). TheSandDoctor ^Talk 09:11, 24 December 2023 (UTC)

Oh and, just re-reading your message @Legoktm:, I would be happy to tag-team something as well, potentially. I am thinking that the bot itself might not actually have to be hyper complicated. Could potentially just, at least for part of it, watch/listen for events, shove them in a database table with some sort of action date field, and another component makes some sort of a change at that date/time. Hmm... TheSandDoctor ^Talk 09:15, 24 December 2023 (UTC)

Yeah, that's roughly what I was thinking as well. One suggestion I'd make is to store the "database" on a wiki page. This would provide 2 features: 1) it keeps all information on-wiki, which makes transferring the bot much easier since there's no separate database and 2) it provides an easily understandable opt-out feature for admins, since they could just remove it from the wiki page. So my idea of the workflow is:

• Bot watches event stream (or polls the protection log), to identify instances of protection level being increased and expiration time being decreased.
  • If found, the bot will add an entry to a fully protected user page (aka the "database page"), and ping the protecting admin in the edit summary
  • Admins can undo the bot right away if they don't want it to apply. And if they don't want notifications, they can mute the bot.
• The second part of the bot just watches the database page, identifies when the next protection change is needed, and sleeps until then.
  • When it's time, it restores the original protection level+expiry, and then removes it from the database page. And then sleeps until the next instance.

Seem reasonable and hopefully not too complicated? Legoktm (talk) 06:27, 29 December 2023 (UTC)

Thanks; that's much better than anything I could have come up with! And the first part (maintaining the page) can be done without needing community approval. Suffusion of Yellow (talk) 22:00, 31 December 2023 (UTC)

Automatic tagging of some easily-identified non-English language text

This probably should not be implemented trivially for languages written in the Latin script, but with a few caveats, it seems pretty doable to write a bot that scours articles, and while staying out of appropriate templates, tags text using existing templates like {{lang}} as either being in a specific language, or at least being in some language written in a particular script, e.g. und-Hani or und-Cyrl as per the obligatory HTML |lang= parameter and ISO 639. If there is und text already tagged, it makes it much easier to see whether 漢字 is lang=ja-Hani or lang=zh-Hant, and also to quickly retag everything en masse.
If we are getting dangerous, I can think of multiple ways to further discriminate between, say, Japanese and Chinese-language text beyond simple checking for strings of CJK ideographs. Remsense留 21:16, 6 December 2023 (UTC)

I'm wondering if this may fall afoul of WP:COSMETICBOT, since it doesn't alter the rendering of the page, but I feel like COSMETBOT exceptions have been carved out before for changes that alter the presentation of a page via speech synthesis, which this would do (and, AFAIA is the primary reason we tag non-English terms like this). As a minor note, I don't think it's necessary to specify in these templates the distinction between zh-hant and zh-hans or whatever they're called in the appropriate standards. Folly Mox (talk) 13:29, 7 December 2023 (UTC)

Oh no, I think a cosmetic argument is a non-starter, since it's a fundamentally semantic change core to the HTML standard itself. Not to be overly dramatic, but every HTML page that doesn't tag foreign language content is meaningfully running afoul of the standard, because it has likely explicitly declared at the top that the whole thing is in English. And no—I don't think Hans versus Hant is useful for most end users, but it's a further specification one could make that I decided to spell out for some reason. Remsense留 14:24, 7 December 2023 (UTC)

template change

Hi, I would like to know whether a bot would be able to do this particular task or not. The task is to replace the existing format with the template like I did here on my sandbox to explain it better: [8]

The following articles: 2004 Andhra Pradesh Legislative Assembly election and 2009 Andhra Pradesh Legislative Assembly election require these template changes. Since I am finding this monotonous task quite difficult to do it myself, I am looking for help probably a bot might help I believe? Any info or help is appreciated. Thank you 456legend (talk) 05:52, 7 December 2023 (UTC)

Depending how many pages need editing, this could be a good job for AWB. Certes (talk) 09:39, 7 December 2023 (UTC)

@Certes All the following articles will need this template changes:

1. 2004 Andhra Pradesh Legislative Assembly election

2. 2009 Andhra Pradesh Legislative Assembly election

3. 2014 Andhra Pradesh Legislative Assembly election

4. 2004 Karnataka Legislative Assembly election

5. 2008 Karnataka Legislative Assembly election

6. 2013 Karnataka Legislative Assembly election

7. 2021 Tamil Nadu Legislative Assembly election

8. 2016 Tamil Nadu Legislative Assembly election

And there are few more in addition to these articles.. 456legend (talk) 13:41, 7 December 2023 (UTC)

Thanks. With a list that size, using AWB will be much quicker than writing a bot. Once you have a final list of articles, WP:AWB/Tasks should be able to help. Alternatively, a good programmers' text editor with regexp features should be able to do the job. Certes (talk) 13:57, 7 December 2023 (UTC)

Not a good task for a bot. too small in scale/scope, as discussed above. Feel free to bring back here if scale/scope (in terms of number of pages impacted) massively increases. --TheSandDoctor ^Talk 09:03, 24 December 2023 (UTC)

A bot to clear Template Sandboxes X21 to X71

Hello, it has come to my attention that template sandboxes X21 to X71 are not automatically cleared by Cyberbot I, which clears template sandboxes X1 to X20, and the main template sandbox. So I think there should be a bot that clears the rest of the template sandboxes. This bot would be called "SandBot", and it would clear the template sandboxes at 00:00 UTC and 12:00 UTC every day. It would do additional help for Cyberbot I for clearing template sandboxes X21 to X71. This is only a proposed bot I had the idea to create. RandomWikiPerson_277^{talk page or something} 19:58, 12 December 2023 (UTC)

RandomWikiPerson277, have you asked the operator of Cyberbot if they'd be willing to do this? — Qwerfjkltalk 20:40, 12 December 2023 (UTC)

Well, I did contact the operator of Cyberbot, I haven't got a message back yet/ RandomWikiPerson_277^{talk page or something} 21:22, 12 December 2023 (UTC)

Just delete them. There's absolutely no good reason to have so many sandboxes, and in particular your creation of 53-71 seems to serve no purpose other than wasting others' time. * Pppery * _{it has begun...} 21:24, 12 December 2023 (UTC)

I did not mean to waste others time. I just decided to add some more sandboxes for some reason. RandomWikiPerson_277^{talk page or something} 15:01, 13 December 2023 (UTC)

Mass WP:TFD maybe? –Novem Linguae (talk) 20:35, 13 December 2023 (UTC)

21-52 survived Wikipedia:Miscellany for deletion/Template sandboxes X21-X52 in the past, although I think the outcome of that discussion would have been different if it had been at TfD since MfD is more inclusionist. * Pppery * _{it has begun...} 20:47, 13 December 2023 (UTC)

Wikipedia:Templates for discussion/Log/2023 December 14#Template:X21 –Novem Linguae (talk) 09:14, 14 December 2023 (UTC)

• Y Done task moot as templates deleted. --TheSandDoctor ^Talk 09:01, 24 December 2023 (UTC)

Question

Moved from Wikipedia:Bots/Noticeboard § Question

Can a bot make this kind of changes to multiple pages?

• Before: {{abcd|ᚠ|ᚡ|ᚢ |ᚣ|ᚤ|ᚥ|ᚦ| ᚧ|ᚨ|ᚩ}}, {{abcd|Ꭰ|Ꭱ|Ꭲ|Ꭳ|Ꭴ}}
• After: ᚠᚡᚢ ᚣᚤᚥᚦ ᚧᚨᚩ, {{abcd|Ꭰ|Ꭱ|Ꭲ|Ꭳ|Ꭴ}}

That is,

1. Check if Template:abcd only contains [ ]?[ᚠ-ᛸ][ ]? in each parameter.
2. If so, remove Template:abcd and |, but retain the text entered as parameters (including spaces).
3. If not (= if Template:abcd contains characters other than [ ]?[ᚠ-ᛸ][ ]?), leave it as-is.

172.58.208.108 (talk) 19:19, 16 December 2023 (UTC)

Of course an unsupervised script can make a regex substitution across multiple pages, but the problem statement has been generalised beyond the point where a meaningful discussion can take place. (There is no Template:Abcd, and so it's also unclear whether you're actually talking about Futhark or using it as a placeholder.)

Can you give us an actual example of an article you'd like to make this sort of mass edit at, and describe the effect you'd like to achieve by doing this? Folly Mox (talk) 20:26, 16 December 2023 (UTC)

The template name "abcd" and the character range [ᚠ-ᛸ] are just placeholders. The actual request will use a different template name and a different character range. Anyway, I would like to know for sure if it is possible for a bot to make changes like what I wrote above. (The discussion has not begun, so I cannot give you more details at this moment.) 172.58.208.125 (talk) 23:02, 16 December 2023 (UTC)

Yes, there's no technical barrier to that sort of text processing. Folly Mox (talk) 23:21, 16 December 2023 (UTC)

Update links from http: to https:

My website runeberg.org just recently moved from http: to https: so it would be nice if someone could update the 11,000 links accordingly. This is not urgent, as everything works fine with automatic redirects, but it would be nice. Thank you. -- LA2 (talk) 22:33, 17 December 2023 (UTC)

@LA2: Is this a job for WP:URLREQ? Certes (talk) 22:53, 17 December 2023 (UTC)

@Certes: Indeed, it might well be. I'll post it there. --LA2 (talk) 22:55, 17 December 2023 (UTC)

Small Question

This is my first time posting here, so no idea if this should be done by a bot. So, the "IPAlink" template has another variation "IPA link" (notice the space). The official representation is "IPA link" but I find the "IPAlink" variation also is quite predominant. This isn't urgent ("IPAlink" redirects to "IPA link"), but would a bot fix this sort of thing? PharyngealImplosive7 (talk) 18:08, 21 December 2023 (UTC)

The links are not broken, and changing them would be purely cosmetic. The small improvement might not justify the human or machine effort and the resulting pollution to the page histories. Certes (talk) 18:27, 21 December 2023 (UTC)

Yeah, that's what I thought. If my AWB request at PERM gets approved, I might do it myself, but it's fine. PharyngealImplosive7 (talk) 21:44, 21 December 2023 (UTC)

@PharyngealImplosive7: This is Not a good task for a bot. for the reason above, and doing so with AWB manually would go against the WP:AWBRULES because it would have the same pollution to the page histories. You could update Wikipedia:AutoWikiBrowser/Template redirects to add a rule to change {{IPAlink}} to {{IPA link}}, and then AWB editors and bots will make the update when doing something else that improves the rendered page. GoingBatty (talk) 22:21, 21 December 2023 (UTC)

Ok. I see, and I won't update it manually. I'll just add it to the Template redirect AWB links you mentioned before. PharyngealImplosive7 (talk) 22:32, 21 December 2023 (UTC)

External links bot

There are quite a few external links templates created in recent years (See Category:Social media external link templates) and when used they offer a consistent style and allow for error tracking among other things. However there are still quite a lot of external links that don't use these. Sometimes they are bare links, while others have some kind of text with them. Would it possible for a bot to convert external links in the external links section (links in the body should be ignored as I'm not sure if these templates work in the body correctly or not) to use one of the listed templates at the bottom? Here is an example of an edit with IMDb title.

Templates:

• Template:Bandcamp
• Template:Blogger
• Template:Cyworld
• Template:Facebook
• Template:Facebook page
• Template:Instagram
• Template:Linkedin
• Template:LinkedIn page
• Template:LinkedIn URL
• Template:LiveJournal
• Template:Mastodon user
• Template:Medium.com
• Template:Muck Rack
• Template:Panoramio
• Template:Patreon
• Template:Pinterest
• Template:ResearchGate
• Template:ReverbNation
• Template:SoundCloud
• Template:Telegram
• Template:Threads
• Template:TikTok
• Template:Tumblr
• Template:Twitter
• Template:Vimeo
• Template:Vimeo user
• Template:Vine
• Template:VK user
• Template:Weibo
• Template:WordPress
• Template:YouTube
• Template:YouTube channel

If this is controversial and needs discussion, please point me to where it should be held. Gonnym (talk) 15:45, 28 December 2023 (UTC)

@Gonnym: Template:IMDb name and Template:Official website are two I manually add frequently. Template:Rotten Tomatoes and Template:Metacritic are two more to consider. There are almost 600 entries in Category:External link templates. GoingBatty (talk) 18:58, 28 December 2023 (UTC)

@Gonnym ...plus more in its subcategories. GoingBatty (talk) 19:08, 28 December 2023 (UTC)

Maybe pick one template and start with that, to decrease the massiveness of this task. Get a consensus on a talk page somewhere, then someone can start working on a (now much smaller) bot task, then WP:BRFA it. Template:Google Scholar ID and some of the IMDB templates are the ones I find myself converting the most. –Novem Linguae (talk) 22:21, 28 December 2023 (UTC)

While the task is big I think having a bot do only one at a time runs the risk of it becomming a spam bot (and having editors complaining) as a lot of times there are more than one of these on a page (Coco Lee as an example of one). Gonnym (talk) 22:41, 28 December 2023 (UTC)

My thinking is to do two two BRFAs: one that just does one template and gets the process started, then a big one for everything else later. This avoids WP:TRAINWRECK issues with deciding which templates to cleanup, minimizes the amount of bugs that are likely to crop up during the first bot run, etc. Up to y'all though. –Novem Linguae (talk) 23:05, 28 December 2023 (UTC)

Two cents, take it or leave it: In my workplace for something like this we'd start with a basic test of one, then do a broader test of, say, 3-5, then engage in more widescale implementation. DonIago (talk) 14:56, 29 December 2023 (UTC)

Replacing TWL OCLC proxy links in articles

Hey all! For some background here, for TWL users to access Newspapers.com, the library sends them through a proxied domain at https://www-newspapers-com.wikipedialibrary.idm.oclc.org/. This often results in this domain name making its way into the mainspace, which is problematic because it can only be accessed by those with access to TWL.

JPxG has set up a way to replace these links with the unproxied domain using JWB (see more info and an example edit), but I feel like this is an area where a bot could step in.

Citation bot is able to clean these links up automatically (see an example edit), but it has to be triggered manually. These proxy URLs are not automatically placed in a category, which means a human editor would need to assemble a list of pages to be fixed for Citation bot to even look at them. Citation bot also wouldn't deal with these links outside of citations, such as with external links.

It's worth noting that I've previously filed a tangentially similar BRFA, which was denied as Citation bot would be easier to use and give better results. With these links, however, I don't think that's the case, mainly because Citation bot is tedious to trigger on these pages, but also because Citation bot doesn't even touch other proxied URLs, only Newspapers.com.

I'd love to make this happen using Pywikibot, but based on my previous BRFA I wanted to see some thoughts on this being fully automated. This task is already being done semi-automatically way through JWB, so I think it might as well be fully automated, potentially expanding additionally to other TWL-proxied sites. (Citation bot doesn't even touch other proxied URLs, only Newspapers.com.)

(CCing Headbomb for your thorough comments on the previous BRFA—would love to hear your opinion especially.) Bsoyka (talk) 18:32, 28 December 2023 (UTC)

Yeah, there should be a bot for this. The JWB regex is dead-on-the-marmot simple, and never pops false positives (there's simply never any legitimate reason to link to www-newspapers-com.wikipedialibrary.idm.oclc.org in mainspace so you can just replace it in ns0 indiscrim[...]ly). I run through every couple weeks and people always give me like 14 thankses for it so it seems like a particularly loved task. jp×g🗯️ 18:36, 28 December 2023 (UTC)

I support this bot and would further support a task like Qwerfjkl bot Task 17, to post on the usertalk of editors who leave links like this and tell them to stop doing it. Folly Mox (talk) 18:54, 28 December 2023 (UTC)

Folly Mox, my bot only tracks categories. Is there an error category for this? — Qwerfjkltalk 19:04, 28 December 2023 (UTC)

@Qwerfjkl: Not as far as I know. The replacement process itself would be built on a simple regex find/replace. Bsoyka (talk) 19:05, 28 December 2023 (UTC)

Bsoyka, I know, I'm talking about notifying users. But thinking about it now, there's not much point if they can just be fixed. It would be trivial to setup a daily run on this with a bot. — Qwerfjkltalk 19:29, 28 December 2023 (UTC)

hrrmery mayhapsicles it could mention https://github.com/jp-x-g/PressPass which autoformats newspapers.com citations on firefox and chrome and autofixes this exact issue B^) jp×g🗯️ 19:33, 28 December 2023 (UTC)

JPxG, didn't recognise you for a moment there! I'm used to a green username, not blue. Congratulations on your adminship! — Qwerfjkltalk 19:33, 28 December 2023 (UTC)

Seems like this has some general support to become fully automated. I'll give this a bit more time to sit for anyone else who wants to comment then work up some quick code and get a BRFA going. (Just trying to avoid what happened with my last one—thanks for all the feedback so far, everyone!) Bsoyka (talk) 19:33, 28 December 2023 (UTC)

As far as categories go, tracking categories could be built-in CS1/CS2 templates like Category:CS1 maint: unflagged free DOI and then Citation bot can run over them. The place to raise that is at Help talk:CS1. This is my preferred option, personally, as far as my own individual opinion is concerned.

As for bots doing this, two options. You build a list of articles and feed it to Citation bot (either via a page of links, or separated by pipes). Or you have a dedicated bot fixing that stuff.

Or you can do both, but the tracking category is what needs the least coding. Headbomb {t · c · p · b} 22:28, 28 December 2023 (UTC)

@Headbomb: While I like the idea of a CS1 maintenance category paired with Citation bot to fix these, that wouldn't work for any TWL links other than Newspapers.com as far as I know.

Take, for example, this version of an article and its references 11 and 14. They link to doi.org through the TWL proxy, but Citation bot doesn't even touch them and they have to be replaced manually (or semi-automatically similar to JPxG's JWB settings mentioned above). (Edit: Probably not the best example since the DOIs are invalid to begin with, but I think the idea is still there.)

This is why I think a separate bot task would be useful—Citation bot only deals with proxied Newspapers.com links, but there are tens of other sources going through TWL proxies that it won't handle. Bsoyka (t • c • g) 22:45, 28 December 2023 (UTC)

It deals with many TWL proxied links, not just newspaper.com ones. It might not deal with all of them, which is a great argument to improve Citation bot by providing it with the full list of proxies used by TWL.

But again, a separate bot specifically on this is also not a bad idea. We'd just lose many of CB's other fixes, but proxied links are bad enough to be fixed on their own. Headbomb {t · c · p · b} 22:48, 28 December 2023 (UTC)

Oh, it does? Apologies, I had tested it a few times with other proxied sites and didn't get any more hits.

Appreciate the feedback though—I'll get started on a script and BRFA for now and we'll see where it goes. Bsoyka (t • c • g) 22:53, 28 December 2023 (UTC)

BRFA filed: Thanks for the feedback everyone! I'll point you in the direction of this new BRFA for further discussion. Bsoyka (t • c • g) 01:32, 29 December 2023 (UTC)

I will go vouch for it if necessary. It's always fun to push the big button on a JWB run but it's imperative that this problem have a solution beyond "JPxG makes an embarrassing mistake every couple weeks and wants to hide it behind a couple dozen JWB edits" jp×g🗯️ 10:55, 29 December 2023 (UTC)

It appears quite a lot in other namespaces (see Special:LinkSearch/www-newspapers-com.wikipedialibrary.idm.oclc.org), I assume it should also be fixed there? — Qwerfjkltalk 11:22, 29 December 2023 (UTC)

No, TWL links in other namespaces are likely intentional and not bad practice. I'll often drop a TWL link on a talk page if I'm discussing a source, with the knowledge that the intended audience (other Wikipedia editors) has proxy access. Folly Mox (talk) 11:32, 29 December 2023 (UTC)

Is there a bot

Category:Pages using WikiProject Film with unknown parameters, a maintenance category which exists to flag problems where a use of {{WikiProject Film}} on a talk page is calling parameters that don't exist to be called, currently has 4,808 articles in it — and after looking at it and cleaning up the tiny single-digits handful of exceptions that existed anywhere after the letter B, I was able to determine that the remaining contents all relate entirely to an old, long-deprecated practice whereby B-Class articles in that queue were each also tagged as b1=[y/n], b2=[y/n], b3=[y/n], b4=[y/n] and b5=[y/n] for their individual success or failure in meeting each of the five B-Class criteria listed at Wikipedia:WikiProject Film/Assessment. That's long since been deprecated and isn't done anymore, which is why those are landing as unknown parameters now — but with 4,808 articles to deal with, actually cleaning them up is more work than any human editor would ever actually be inclined to undertake.

Accordingly, I wanted to ask if there's any bot that can be set loose on the task of stripping b#= parameters from the contents of that category. Bearcat (talk) 17:20, 1 January 2024 (UTC)

Yes. I'll try to get to this in the next few days. Primefac (talk) 17:42, 1 January 2024 (UTC)

@Bearcat: BattyBot 79 is going through all the subcategories of Category:WikiProject templates with unknown parameters, and will get to this if Primefac doesn't get to it first. GoingBatty (talk) 17:51, 1 January 2024 (UTC)

Okay, no worries. It isn't urgent or anything, I was just wondering if it was possible — so if it is, it's all good. Bearcat (talk) 17:53, 1 January 2024 (UTC)

Heh, thanks for the reminder GoingBatty, somehow managed to forget I approved that task... yesterday... Primefac (talk) 17:54, 1 January 2024 (UTC)

@Primefac: No worries - it's hard to remember what we did last year.  ;-) GoingBatty (talk) 00:15, 2 January 2024 (UTC)

Add parameter for WikiProject Africa/The 10,000 Challenge

For the 5832 articles listed at Wikipedia:WikiProject Africa/The 10,000 Challenge please add |AFR10k=yes to the project banner {{WikiProject Africa}} on the talk page. This adds a note to the banner and also populates Category:Articles created or improved during the WikiProject Africa 10,000 Challenge. Thanks — Martin (MSGJ · talk) 19:19, 3 January 2024 (UTC)