Wikipedia:Bot requests

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

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

A bot to help maintain and identify inspiration articlesEdit

At the idea lab a while back, I floated the idea of identifying inspiration articles for a given page, GAs or FAs on similar topics that could serve as inspiration for helping get that page to a higher level, and got some positive feedback. Following some additional discussion on Discord the other day, I've created {{Inspiration page}}. I'd like bot (or maybe tool) help for two things:

  1. Maintenance, such that if an article is demoted, any pages using it have their banner removed.
  2. Identifying new pages where it would be helpful to add the banner. This could possibly be done by searching for pages that share a substantial number of categories/wikiprojects, but where one is FA and the other is C-class or below.

Let me know if this sounds feasible! {{u|Sdkb}}talk 03:12, 18 June 2021 (UTC)[]

Task 1 is fairly trivial, and can be done with even something like AWB. Task 2 is slightly more difficult. 🐶 EpicPupper (he/him | talk, FAQ, contribs) 02:16, 13 August 2021 (UTC)[]
@Sdkb: I just came across phab:T290447 and, if implemented, the remaining logic to finish your task #2 would be trivial. While there are some additional ideas mentioned there, I don't think I personally have the time to work on it, and it doesn't look like anyone in particular has claimed that task. Vahurzpu (talk) 02:01, 7 September 2021 (UTC)[]

Restoring redirects to recreated pagesEdit

A fairly common phenomenon occurs where a page is deleted, then all the redirects to it are deleted per WP:G8, then the page is recreated but the redirects to it are lost. Could we get a bot that looks at recently recreated pages and restores the incoming redirects to them? {{u|Sdkb}}talk 07:12, 2 July 2021 (UTC)[]

@Sdkb: WP:BOTREQ is over that way. I don't see any policy problems on something like this, though it may have a large technical hurdle (specifically that a page has no easy way to show "what used to link here"). — xaosflux Talk 11:12, 2 July 2021 (UTC)[]
Oops, I thought I was posting at BOTREQ for some reason; thanks for doing the move, Xaosflux/Headbomb. {{u|Sdkb}}talk 16:21, 2 July 2021 (UTC)[]
Off the top of my head... It would need to be an adminbot, of course. The bot would probably need to watch for G8 deletions of redirects to maintain a database of deleted redirect → former target to be able to identify needed undeletions in the first place. The bot should probably also be prepared to avoid undeleting revisions from a previous deletion (e.g. if the history was "attack page → delete → redirect → G8", or "redirect → attack page → selective delete to restore the redirect → G8"), and should make sure the undeleted redirect isn't still tagged with {{db-redirnone}} or the like. Maybe only undelete revisions that begin with #REDIRECT? That would lose legitimate pre-deletion content though, is that ok? Should there be a time limit, e.g. if the redirects were deleted 10 years ago should they still be undeleted? Is there any discussion elsewhere of the problem, particularly from admins who normally handle this sort of undeletion manually? Anomie 12:02, 2 July 2021 (UTC)[]
To answer your last question, I discussed the instance that prompted this proposal briefly with @Liz, who I know does a lot of work with deletion. I also just added invite notices to WT:Deletion policy, WT:Deletion process, and WT:REFUND; hopefully that'll draw over some editors who have relevant thoughts. {{u|Sdkb}}talk 20:06, 2 July 2021 (UTC)[]

A related point: I have seen articles overwritten by a redirect to an unrelated target at AfD, as a way to make the article disappear per G8 when the target gets deleted. I occasionally check manually for new cases. (Current results are all good-faith false positives.) The bot proposed above might end up collecting the information required to automate that check. Certes (talk) 12:26, 2 July 2021 (UTC)[]

  • Is it not very context-dependent? Creating a page at the same name as a previously-deleted page is no guarantee that the topic is the same (for instance, biographies of homonyms) or that the redirects are still appropriate. That is probably a small proportion of cases, but that would be pretty bad, whereas the restoration of redirects is "only" a convenience, so even a low probability of failure can tip the risk/benefit ratio into the red. (It would be interesting to have numbers on this but I am not sure there is a practical way to obtain them.)
Adding in the concerns that Anomie raised about page history (I am not qualified to say if those are insurmountable or merely a pain to code around), I would suggest to make it semi-automated instead. The backend that generates the list of redirects to restore is the same, but actually restoring them requires human intervention. This could be either an additional popup in the undeletion process, which gives a "list of possible redirects" with a choice restore/do not restore/view history for each item and/or a "global restore" button (for a contested PROD, for instance), or simply a post to the talk page of the recreated article with "I am a bot, I found the following possible redirects". TigraanClick here for my talk page ("private" contact) 17:46, 13 July 2021 (UTC)[]
@Tigraan, good points. We should distinguish between undeletions (where an admin restores a page with its past revisions) and recreations (where any user begins a page at the same title as a page that was deleted in the past). For recreations, there is indeed the homonym problem, so I like your suggestion of a semi-automated process of some sort, but we should be safe for undeletions to make it fully automatic. {{u|Sdkb}}talk 21:42, 13 July 2021 (UTC)[]
For anything automated (and maybe semi-automated?) it would be useful if the bot kept a log for humans to check. It would only need to list the target page and a link to whatlinkshere with links and transclusions hidden. Maybe also a number of redirects restored, so perhaps looking something like
16 July 2021
I'll drop notes about this discussion at Wikipedia talk:WikiProject Redirect an Wikipedia talk:Redirects for discussion. Thryduulf (talk) 01:42, 16 July 2021 (UTC)[]
This will be so useful! If I'm not mistaken, redirects also get wiped out upon draftication, which happens quite often. I hope this is easy to code, but – as pointed out by Tigraan – the product will need to be at most a semi-automated process. – Uanfala (talk) 00:45, 18 July 2021 (UTC)[]
  • This would be very useful... except in the cases where it isn't so useful (ie where the new article is not a re-creation of the previous one, but an article on a topic which shares the title, for which some redirects may still be valid but others not - suppose the deleted article was for singer "Jane Mary Xyz", and the new one is for politician "Jane Mary Xyz", then the rd from her short name "Jane Xyz" would still be useful but not the ones from her married name "Jane Abc" or "Jane Mary Abc"). That's to say, yes, very useful to rescue and display the list of previous redirects, but they would need to be checked by an editor before being re-created. So, create them commented out at foot of the new page, or list them on the talk page, so that the person creating the new article can see them and they or someone else can easily check and decide which ones are still useful. As someone who often creates a lot of redirects from variations of someone's name I hate to see all that work disappear if the article is deleted or, worse in some ways, if it's draftified and will eventually be reinstated. Even where the new article is on the same topic, some of the redirects may no longer be valid because a new article may have been created in the interim, or a new link made which is not intended for this article, or a dab page may have been created, etc. But, in principle, Yes, Please. PamD 16:14, 24 July 2021 (UTC)[]
  • Checking back in on this, would anyone be interested in taking it on? {{u|Sdkb}}talk 17:26, 5 September 2021 (UTC)[]

Automatically create ending with“Archipelago”redirects for ending with“islands”class entriesEdit

I found that some of the islands artical did not create corresponding archipelago variants. Robots can automatically identify and create them.--q28 (talk) 01:35, 27 July 2021 (UTC)[]

According to incomplete statistics, at least 1000 redirects (exact amount depends on how it is determined) have not been created, so they are considered batch work and should be completed by robots.--q28 (talk) 01:36, 27 July 2021 (UTC)[]
It is worth noting that this is not only limited to the geography, it also applicable to foreign cartoons. According to the translator's preferences, some people use archipelago, and some use islands.--q28 (talk) 01:41, 27 July 2021 (UTC)[]
Is there really a one-to-one correspondence? For example, the Duke of York Archipelago is thousands of miles from the Duke of York Islands, and British Archipelago correctly redirects to British Isles rather than British Islands. Certes (talk) 02:15, 27 July 2021 (UTC)[]
  Impossible It seems likely that this will need to be done manually, Withdraw the request.--Here's 28 and did I make a mess? 11:03, 14 September 2021 (UTC)[]
@Q28: If you're filling in manually, other mismatches to beware of include Low Islands (Canada) vs Low Archipelago (Pacific), Marshall Islands (Pacific) vs Marshall Archipelago (Antarctic), and The Islands (British Columbia) vs The Archipelago (Ontario). There are also X Islands forming part of a larger X Archipelago, such as Riau Islands and Solomon Islands. Certes (talk) 13:45, 14 September 2021 (UTC)[]

Vital peopleEdit

I would like a bot that maintains the vital people here just like there is one for vital articles. I would like the bot's tasks to be the same as vital articles here and would like to make sure there are no duplicates in the vital people. Pinging Kanashimi, the bot operator for vital articles, if they could do the same thing for vital people. Interstellarity (talk) 20:29, 4 August 2021 (UTC)[]

Comment: I am a bit busy these days. Please ping me if I miss something. Is the vital peoples using the same class mark listed in Category:Wikipedia vital articles by class? Kanashimi (talk) 22:44, 4 August 2021 (UTC)[]
@Kanashimi: Yes, that's correct. Interstellarity (talk) 23:28, 4 August 2021 (UTC)[]
@Interstellarity   BRFA filed Kanashimi (talk) 05:54, 8 August 2021 (UTC)[]
@Interstellarity:   Denied. GoingBatty (talk) 02:40, 21 September 2021 (UTC)[]

Bot for creating name redirectsEdit

I very often come across situations like the one I just did at Vanessa C. Tyson, where the middle name (Catherine) is given right in bold at the start of the article but the redirect from Vanessa Catherine Tyson has not been created. Sometimes there are other variations of this situation, such as if the redirect from Vanessa Tyson hadn't been created, or if the page was located at "Vanessa Tyson" but the redirect from "Vanessa C. Tyson" wasn't created. I don't expect a bot to be able to fix all of these, as in some cases there could be disambiguation concerns, but for many many situations, it should be possible for the bot to determine that only one notable person has a name and create redirects accordingly. Could we do that, and have it tag with {{r from short name}} and {{r from long name}} as needed? {{u|Sdkb}}talk 22:31, 7 August 2021 (UTC)[]

I'm just wondering: Why isn't there a semi-auto, human-in-the-loop system for tasks that might be sensitive to context and false positives like this? Having the bot find all the needed changes and apply them with approval would still save a ton of work from having humans do all the changes manually. Intralexical (talk) 12:28, 9 August 2021 (UTC)[]
We have WP:AWB and WP:JWB. They can accept or create a page list and suggest methodical changes to each page on the list. For example, I recently used JWB to change links to the ambiguous term Qu'Appelle. For the 90% about Regina—Qu'Appelle I just clicked Save; for the few exceptions I typed a better link in manually. Certes (talk) 12:45, 9 August 2021 (UTC)[]
I think data quality is a big concern here. It's well documented that middle names, birthdates, etc will often get added to biographies with no sourcing (or unacceptable sourcing) and persist for a very long time. A lot of them will turn out to be completely bogus, too! On the other hand, I think the idea of automatic name redirects for biographies with middle names has some potential. For example, if Vanessa C. Tyson didn't exist, there was no disambig at Vanessa Tyson, and no other articles were titled "Vanessa * Tyson", it would be helpful to make the redirect (or at least flag it for creation). jp×g 22:54, 20 August 2021 (UTC)[]
That is how the Order of the Companions of OR Tambo in Gold became awarded to Joseph Sepp Bellend Blatter.[1] Thincat (talk) 20:20, 9 October 2021 (UTC)[]

Post-Move BotEdit

A helpful message is shown after moving a page:


It seems like a lot of this could be be automated fairly straightforwardly.

It was pointed out on the Teahouse that Wikipedia:Double_redirects do get fixed by a bot, but the fair use rationales, navboxes, etc. also seem unnecessary to fix manually.

Intralexical (talk) 13:03, 9 August 2021 (UTC)[]

Although some of these tasks definitely look doable, there might be some WP:CONTEXTBOT issues. 🐶 EpicPupper (he/him | talk, FAQ, contribs) 02:09, 10 August 2021 (UTC)[]
@Intralexical, @EpicPupper: I think I could put together a bot to update fair use rationales. Tol (talk | contribs) @ 15:21, 4 October 2021 (UTC)[]

Convert templateEdit

Are there any existing bots that convert manually typed data to use {{Convert}}, like distance, temperature, among others? -- DaxServer (talk) 11:46, 20 August 2021 (UTC)[]

Don't think so, but what kinds of edits are you thinking (can you make a sample edit by hand for example)? It sounds like it would be a WP:CONTEXTBOT issue and could only be semi-automated (eg using an AWB Module) at best. ProcrastinatingReader (talk) 15:31, 20 August 2021 (UTC)[]
I have a couple of regexes in AWB which I ran last week on maybe 10-20 pages, couldn't remember exactly. (Couple of examples: [2], [3], manual search). These are the regexes I used, albeit incomplete:
  • Distance: (\d+(?:\.\d+)?)(?: |\s)?km(?!<sup>|2)(?:\s\(\d+(?:\.\d+)?\smi\))? would be replaced with {{convert|$1|km|abbr=on}}
  • Temperature: (\d+(?:.\d+)?)(?: |\s)?°C(?:\s\(\d+(?:\.\d+)? °F\)) would be replaced with {{convert|$1|C}}
  • Rainfall: (\d+(?:.\d+)?)(?: |\s)?millimetres(?:\s\(\d+(?:\.\d+)? in\))? would be replaced with {{convert|$1|mm|abbr=on}}
I chose the abbr=on because when I edit the articles [manually], I choose to go abbreviated. I ran only on Indian places and thus the regex is developed for km->mi and C->F and not the opposite. I don't know how to build AWB modules, but can develop a Pywikibot script and file for a supervised BRFA (already have two pending). -- DaxServer (talk) 16:43, 20 August 2021 (UTC)[]
DaxServer: I do have a solution that works with AWB and I am very happy to share it. My solution can perform minor miracles of convert template editing that includes dealing with ranges, worded numbers, fractions etc. but it is long and complicated and shows all the signs of a system that has slowly evolved. The solution is not just a one off search and replace but a long sequence. Although the current version is pretty reliable, things can go wrong and it does require each edit to be manually checked. What you will need is my copy of the XML file that defines all the AWB setup - I'm not sure how best to get it to you. Gaius Cornelius (talk) 05:09, 28 August 2021 (UTC)[]
@Gaius Cornelius Thanks. I am thinking of two ways to share, one being adding the contents of xml to a new/existing [sandboxy] page under your user page, and ping me. Or you can drop it here: User:DaxServer/test. Or if you don't want it in public record, you can email me. — DaxServer (talk to me) 09:22, 28 August 2021 (UTC)[]

Redlinked biographies that are potentially notable?Edit

Hi folks, not sure if this is a practical request, thought I'd ask anyway. The Science Fiction Encyclopedia has approximately 12,000 entries on people, most of whom are likely notable. As I discovered when writing Sarah LeFanu, at least some of them do not have Wikipedia articles. Is it practical for a bot to trawl through this list, check whether Wikipedia has an article on each entry, and save the entry to a list if it doesn't? Vanamonde (Talk) 07:33, 23 August 2021 (UTC)[]

I usually deal with this sort of thing in one of two ways:
  1. Format a list (Aaargh, Calvin → *[[Calvin Aaargh]]) and view it in a sandbox; grab the source HTML and filter on class="new" or "(page does not exist)" (either works)
  2. Paste the list into PetScan (tab "Other sources", section "Manual list"), set Wiki to enwiki and list pages which do exist, then find the difference.
I expect you're aware of the usual traps, such as people listed under a different name (Dr Acula is Forrest J Ackerman) and non-writers with similar names (John Peel isn't John Peel (writer)). Certes (talk) 12:31, 23 August 2021 (UTC)[]
Another possibility would be to use Wikidata. The Encyclopedia of Science Fiction ID (P5357) links Wikidata items to their associated entries; you can use this query to find items linked to an ID with no page on enwiki. Wikidata doesn't yet have complete coverage of the encyclopedia; that is tracked by mixnmatch:1330. This should help deal with the issue of mismatched names. Vahurzpu (talk) 13:00, 23 August 2021 (UTC)[]
That, especially the unmatched list, sounds much more reliable than my suggestions. Certes (talk) 13:11, 23 August 2021 (UTC)[]
Thanks, both; all those methods seem quite straightforward, and I did not know of any of them. Vanamonde (Talk) 13:24, 23 August 2021 (UTC)[]

Many projects having bad link to toolserverEdit

Hey all, hoping a bot (or an AWB master) could be deployed to help in the project space. Many, many WikiProject pages have code at the bottom that begins with:

[[tools:~dispenser/cgi-bin/ . . .

and of course that link is now bad (gives 404 error on toolserver); but since the pages use tools:, this is not really a URL change request. An example use is at WP:WikiProject Geelong#External watchlist.

All instances in the Wikipedia namespace that begin with the above string can be deleted, together with all other characters up until the closing ]]; there is no retcon that will fix it. Thanks in advance, UnitedStatesian (talk) 18:35, 23 August 2021 (UTC)[]

499 instances. Might be easier to get someone to do an AWB run. Primefac (talk) 19:59, 23 August 2021 (UTC)[]
Cool, Ill take it over to WP:AWBREQ. Thanks, you can close this here. UnitedStatesian (talk) 00:32, 24 August 2021 (UTC)[] ownership change, now ainews.xxxEdit

Referencing this discussion: Wikipedia:Help - Wrongly Indexed

It seems that was formerly "Adult Industry News", a news site for the porn industry, which has a lot of citations. The domain now belongs to "Artifical Intelligence News". Needless to say, the new owner doesn't want its domain linked in porn-related articles.

Adult Industry News is now

Experimenting with some of the links from* it seems that one cannot simply substitute .com with .xxx. The pages must be found on

@GreenC: I am not sure if InternetArchiveBot would handle this unless someone went through all ~130 links and tagged them with {{dead link}}. I don't know of another bot that comes close. ~Anachronist (talk) 15:29, 27 August 2021 (UTC)[]

Sounds like a WP:URLREQ task. Primefac (talk) 15:32, 27 August 2021 (UTC)[]
Thanks, I'll put this request there. ~Anachronist (talk) 15:35, 27 August 2021 (UTC)[]

Fix past weekly subpages by re-substituting the "preload" templateEdit

I have fixed Wikipedia:Articles for improvement/Schedule/Preload so that it does not incorrectly show the year 2019 again. Also, the image from Wikipedia:Articles for improvement/2014/8 and earlier subpages (which forgot to be included at Wikipedia:Articles for improvement/2014/9) is now back. Finally, I have used "titleparts" to allow for re-substitution onto past weekly subpages. So now, someone (perhaps, a bot) needs to fix all past weekly subpages of Wikipedia:Articles for improvement by re-substituting the "preload" template, i.e. replacing the entire text of the page with {{subst:Wikipedia:Articles for improvement/Schedule/Preload}}.

The bot MusikBot is the one who is currently creating the weekly subpages. So, if this needs to be done by a bot, then the best choice for the bot would be MusikBot.

This will fix the following problems:

GeoffreyT2000 (talk) 02:12, 1 September 2021 (UTC)[]

AFI used to involve more than one article, such as Wikipedia:Articles for improvement/2013/22. The switch to using a single article appears to have started with Wikipedia:Articles for improvement/2013/48, but it may have gone back to multiple articles for some time later, I'm not sure. Anyway, I suppose the red links in the multi-article selections will have to be fixed manually, but frankly I think it's fine to leave them broken... they are only visible in very old archives.
I don't know that I will have time to work on this in the near future, but if you can compile a list of pages that can safely be replaced with {{subst:Wikipedia:Articles for improvement/Schedule/Preload}}, I suspect they could easily be fixed using AWB. MusikAnimal talk 15:43, 1 September 2021 (UTC)[]

reflist talkEdit

Would it be possible to have a bot add {{reflist talk}} to talk page threads which have refs?(I am not watching this page, so please ping me if you want my attention.) SSSB (talk) 08:39, 3 September 2021 (UTC)[]

@SSSB: You may be interested in User:GreenC bot/Job 8. Certes (talk) 10:37, 3 September 2021 (UTC)[]
@Certes: - didn't know it was already done. Thanks, SSSB (talk) 10:44, 3 September 2021 (UTC)[]
@SSSB: It ran one time 2.5 years ago. It checks every single talk page (6+ million). Maybe time to run again. -- GreenC 05:22, 7 September 2021 (UTC)[]
@GreenC: it would probably be worth it to run it somewhat regularly (monthly?), especially if a way can be found to streamline the pages it has to check (only talk pages that had been edited since the last run?) Wouldn't do anything for the first run, but thereafter... SSSB (talk) 07:26, 7 September 2021 (UTC)[]
The first (next) run could check only talk pages edited in the last 2.5 years. There are a lot of inactive ones to skip. Certes (talk) 10:12, 7 September 2021 (UTC)[]
@SSSB and Certes: That is a good idea as it would avoid downloading the page and issuing a API:Revisions (rvsections) call. Unfortunately I'm busy right now and rather then investing more programming time, or delaying the run, I went ahead and started it up. It's a single thread running about 2.4/second is 207,360 pages/day should take 30 days for 6.4 million. Watchable at Special:Contributions/GreenC_bot. I'll get the time check done before next run, and depending how this went consider automating it (more steps to fully automate). -- GreenC 03:52, 13 September 2021 (UTC)[]
Forgot I already had a function to check last edit time in another program, so dropped it in. Seems to be running at 6.66 pages a second which should finish in about 11 days. Significant improvement. Most of them are skipped even after 2.5 years. -- GreenC 16:23, 13 September 2021 (UTC)[]
@SSSB and Certes: The bot completed its run, it took about 14 days and edited 5,086 pages. I am hand editing about 150 pages due to formatting oddness. It's been 30 months since last run / 5086 = about 170 per month on average. Given the resources required to run the bot (millions of API calls) waiting 6 months would be a backlog of 1,020 between runs seems like a decent trade off. -- GreenC 15:12, 27 September 2021 (UTC)[]

Bot for creating redirects from alternative names or sort namesEdit

This would be a difficult bot to code, but one that I think would have a big impact if it could be created.

It's regrettably common for editors to create a biography page, not realizing that it already exists at a different title (such as with a middle initial or middle name). Likewise, it's common as a reader to search for a person and end up at the results page because you used/didn't use a middle initial, whereas the page did the opposite. These issues could be remedied if we did a better job creating redirects. For a Robert Quincy Smith, all of the following are possibilities:

  • Robert Quincy Smith
  • Robert Q. Smith
  • Robert Smith
  • Smith, Robert Quincy
  • Smith, Robert Q.
  • Smith, Robert

If he goes by "Bob", that adds another whole set. It would be awesome to have a bot that could help create these redirects. However, there are two pitfalls I see: (1) Human names are quite complex and can sometimes be difficult to automatically parse. (2) When multiple people share a name, some terms may need to get pointed at a disambiguation page instead. To address these things, the bot could (at least initially) limit itself to clear instances, where pages have Wikidata name information, sort name data, and bolding that all lines up, and there are no other pages with similar names that might conflict.

Would anyone be interested in taking on this challenge, or is it just too difficult to be even partially feasible? {{u|Sdkb}}talk 17:43, 5 September 2021 (UTC)[]

I'm not sure this makes much sense as a bot. It might make sense as a script / preview template a bit like {{JAH}} however. Headbomb {t · c · p · b} 17:54, 5 September 2021 (UTC)[]
Definitely problematic as a bot; hell, we had to create an entire category of CSD for a user who did this sort of thing. Primefac (talk) 19:40, 5 September 2021 (UTC)[]
I have been batch-creating sort name redirects manually for several years, only for human names consisting of a first name and last name with no middle initial or postnominal material or the like, and I run into quirks and oddities requiring human attention fairly frequently (such as two-name constructions that are actually a single name, common in Asia and the Middle East). I think this is doable if relatively small batches (a few thousand at a time) are eyeball-reviewed for common issues to be manually removed before setting a bot to do the task work. I also note that there is an occasional problem of redirects being made for articles on the brink of deletion, particularly where something else is slated to be moved to the original article title post-deletion. BD2412 T 19:50, 5 September 2021 (UTC)[]

Bot to groom the Mentor listEdit

Hi, I think it would be a good idea for a bot to add grooming of Wikipedia:Growth Team features/Mentor list to a periodic (perhaps hourly) job. That page is integrated with growth features and is fragile to specific wikitext layout. — xaosflux Talk 09:53, 21 September 2021 (UTC) Initial requirements:[]

  • Ensure that the initial --------- identifier is in place
  • Ensure there is only one --------- identifier on the page
  • Ensure each line follows an exact format:
    • * [[User:Username]]|ddd
      • Where:
        • Username is a valid username with proper capitalization
        • A space is present between the * and the Username
        • No space is present between the Username link and the required pipe character
        • ddd is plain text from 0-239 characters in length, not containing any wikimarkup or external links
  • Should an invalid line be discovered, the line should be removed (this will include removing usernames that are (no longer) valid

Optional features:

  • Sort the list of usernames

Open for comments. — xaosflux Talk 09:53, 21 September 2021 (UTC)[]

@Xaosflux: - it's a substantive request, one I'll be asking for on the Growth talk page shortly, but is there a standard bot capacity to check when an editor last edited and remove them from the list under certain conditions, or does it have to be done manually? As inactive mentors starts doing immediate harm (as opposed to, say, an inactive admin) it would be a good thing to be able to do frequent checks on, which we'd probably struggle to do manually. Nosebagbear (talk) 10:39, 21 September 2021 (UTC)[]
I believe User:Yapperbot/Pruner is what you're looking for. Primefac (talk) 10:47, 21 September 2021 (UTC)[]
(EC with User:Primefac) @Nosebagbear: I thought about that for at least a possibly "part 2" optional feature. Wikipedia:AutoWikiBrowser/CheckPageJSON clerking by Musikbot2 has this feature already (example) - so this could be a good task for User:MusikAnimal to pick up if they are so inclined. As far as actually removing "inactive" mentors - that part should probably get some more discussion - but it seems reasonable to at least removing someone from the list if they haven't edited in a long time. — xaosflux Talk 10:48, 21 September 2021 (UTC)[]
I'm interested. Some of these things shouldn't matter though, and if they do, they should be fixed in mw:Extension:GrowthExperiments. For instance the spacing between *, the username, and the pipe, etc. It looks like the regex already ignores the spacing [4]. The horizontal rule --------- also shouldn't matter. The code simply looks for links to user pages in the parser output [5] to collect the list of mentors. Bearing that in mind, I think at minimum we'd only need the bot for validating the mentor intro text (length and absence of markup), the optional alphabetical sorting, and perhaps removal of inactive editors. I believe invalid usernames simply get ignored, so it might be better to leave them in place so maintainers of the mentor list page can fix them.
Alternatively, we could make this page sysop-protected and require non-admins request to be added on the talk page, similar to Wikipedia:WikiProject Articles for creation/Participants and the AWB CheckPage. That system seems to work well as far as ensuring proper data entry, and also allows us to vet who is listed as a mentor. It would create yet another backlog, but I doubt it would be that burdensome. What do you think? MusikAnimal talk 17:39, 21 September 2021 (UTC)[]
There are real negatives to the AfC method of doing things (eg the backlogs of 1600 pending submissions). I think premature optimisation is generally a bad idea, in this context that being premature sysop protection unless there's evidence that's actually necessary due to a systematic problem if it isn't a protected page. ProcrastinatingReader (talk) 17:50, 21 September 2021 (UTC)[]
That's a little harsh, no? The backlog has absolutely nothing to do with the fact that people have to ask to be reviewers, so saying "AFC is shite, we shouldn't be following their model" is a pretty unfair. MusikAnimal is wondering if protection would be necessary to stop trolls and other troublemakers from adding themselves to the mentors list. Primefac (talk) 19:01, 21 September 2021 (UTC)[]
Yeah basically, and also that it being sysop-protected by itself could be enough to ensure the correct syntax is used. Not that non-admins don't know how to read and follow instructions, but the pink background seen when editing a sysop or template-protected page to most means "be careful". To my knowledge we haven't had many if any issues with breaking changes being made to the AfC participants page, AWB CheckPage, spam blackist, etc. -- all of these are also fragile wikitext pages that don't rely on a bot to enforce adherence to a specific format (there is User:MusikBot II/AWBListMan but it doesn't do anything as far as validating syntax).
I also wonder if trolls etc. would add themselves to the list, as well as just overzealous users who think they are qualified to be a mentor when they really aren't. Perhaps adding a barrier to becoming a mentor is too bureaucratic, but if it only introduced a minimal backlog for admins (as I would expect), then it might be worth it. Just an idea :) MusikAnimal talk 20:04, 21 September 2021 (UTC)[]
There are probably lots of people that would do irregular reviews of AfC drafts, but won't be regular reviewers so don't request the permission (either 'hat collecting' reasons, or feeling a sense of commitment when asking for a perm, etc). There's a related discussion somewhere in my user talk archives. The same idea (in the opposite direction) is visible in the admin corps -- the position that the tools are part of a 'kit' and admins should be able to explore new areas, and indeed some admins do irregular work in other admin areas without having to request the permission. I do think the fact that most editors have to request permissions for AfC does probably contribute towards the backlogs. How significant that is, I don't know. Either way, I didn't say AfC is shite, just that its permission system is probably not ideal, and shouldn't be adopted as a rule.
A sysop protected page wouldn't just stop trolls and troublemakers. It means a sysop needs to take an affirmative action to add a user to a page. That would naturally result in de facto requirements beyond just "this edit is not vandalism" (see Pending Changes and the higher standard to approve edits, or indeed the fact that approving AfC has higher standards than AfD; requiring a 'vouch' of sorts has always resulted in higher requirements on here). If we just want to stop trolls and vandals, ECP protection should be sufficient. ProcrastinatingReader (talk) 21:59, 21 September 2021 (UTC)[]
Basically agree with ProcrastinatingReader. I think it's unnecessary to require an admin to act as a gatekeeper – they could just as well watchlist the page and fix any malformed entries. The page already has ext-conf protection which should stop the "trolls and other troublemakers". Better regulating the syntax can also be done by creating Template:Editnotices/Page/Wikipedia:Growth Team features/Mentor list with a "STOP AND READ THE FOLLOWING!" kind of message. Hopefully, ext-conf users are capable of following the instructions when prominently visible. – SD0001 (talk) 12:07, 23 September 2021 (UTC)[]

Bot to fix indentsEdit

This is probably going to be quite challenging, but putting out the idea nevertheless – it would be great to have a bot that corrects indent characters used in talk page discussions. For example,

Comment. – Editor A
::Comment. – Editor B

^this would make the bot comes along and remove the extra indent (:: --> :)

*Comment – Editor A
::Comment. – Editor B

^this would make the bot come along and replace :: with :**: to comply with MOS:INDENTMIX.

This should be restricted to comments being posted from now on and not be done for historical discussions. – SD0001 (talk) 16:13, 23 September 2021 (UTC)[]

Two things. First, I think you mean *:. Second, I feel like this would have huge context issues; we would have to find pretty much all of the expected exceptions (such as multiple replies to a single comment with varying indents). That being said, I suppose it could work. Primefac (talk) 16:19, 23 September 2021 (UTC)[]
This seems too context-dependent. Sometimes, a block of text, like a reflist, or a block of code, or a template with div tags in it, will have to be zero-indented in order to avoid display or syntax problems. Those things should not be indented. – Jonesey95 (talk) 16:41, 23 September 2021 (UTC)[]
Sometimes things like this you can try it, see what breaks and add exceptions (skip if contains div tags) until it runs clean-enough. Maybe there are too many exceptions or maybe just a dozen or two. Hard to say without trying it. -- GreenC 15:40, 26 September 2021 (UTC)[]
Indeed, when I wrote out my initial reply I tried to find exceptions for potentially accurate-but-would-trigger-the-bot syntax but actually ended up finding none (which is why it's not in my reply above), so "try it and see" might be a good way to proceed. Primefac (talk) 15:56, 26 September 2021 (UTC)[]
You mentioned the scenario of multiple replies to a single comment with varying indents. How would a bot handle that? The bot wouldn't be able to determine which comment is being replied to right? Winston (talk) 21:42, 2 October 2021 (UTC)[]
:Comment 1
::Reply to line 1 (2)
:::Reply to line 2 (3)
::::Reply to line 3 (4)
::Reply to line 1 (5)
:::Reply to line 5 (6)
I think as long as the bot looks for increasing (and not decreasing) indents, it should be fine. That being said, if there was a reply to line 4 after line 6 (with the correct indents) along the lines of
:::::Reply to line 4 (7)
That would be a GIGO issue that mis-indents the replies and potentially causes problems. Can't fix that though. Primefac (talk) 22:12, 2 October 2021 (UTC)[]
  • HERE is an example diff of a bot that removes gaps between indented lines, enforces consistent indentation style, and removes extra indentation, assuming indents consist only of colons and asterisks. A gap is a sequence of whitespace lines sandwiched between two indented lines. Each line's indent style is modified to match the previous line's indent style, going from beginning to end. If INDENT_LEVEL(line 2) > INDENT_LEVEL(line 1) + 1, the contiguous block of indented lines beginning with line 2 which have indent level >= INDENT_LEVEL(line 2) is shifted to the left by INDENT_LEVEL(line 2)-INDENT_LEVEL(line 1)+1. I think this is the best a naive bot can do. Note that some of the stuff in this diff is just for demonstration and is essentially garbage in, garbage out. Is this about right? Winston (talk) 02:02, 3 October 2021 (UTC)[]
    @Notsniwiast Yes that looks quite promising. At the top I'm not sure if we should change *Simple example 2 to :Simple example 2 (Doesn't seem required per MOS:INDENTMIX – though I could be mistaken here). In general, I think it would be better to have a conservative bot that fixes only what's necessary rather than re-format everything. – SD0001 (talk) 15:25, 9 October 2021 (UTC)[]
    The reason is that *Simple example 2 is considered to be part of the same list as ::Hello, since only blank lines separate the two lines. The bot first removes the blank lines, then enforces consistent indentation style which results in *Simple example 2 becoming :Simple example 2. Then ::Hello is changed to :Hello.
    Relevant links regarding gaps MOS/Lists#List styles, MOS:INDENTMIX, MOS:INDENTGAP, MOS:LISTGAP.
    But there are choices to be made regarding which intervening blank lines to remove. I see two behaviors here that can be configured:
    1. Whether to remove intervening blank lines if there is more than one intervening blank line.
    2. Whether to remove intervening blank lines preceding a line with INDENT_LEVEL 1. This is because a line with INDENT_LEVEL 1 with a preceding blank line is treated by MediaWiki as a new list, which may or may not be the intended behavior of the editor. If there is just a single blank line then it looks like part of the previous list. If there is more than one blank line then there is extra spacing and it looks separated from the previous list. Thus 1. and 2. are related.
    The diff I gave in my preceding comment uses a TRUE, TRUE configuration, i.e. both 1. and 2. are removed since it just naively removes intervening blank lines between any two indented lines. However, now that I think about it a FALSE, TRUE configuration might be better, i.e. only remove single blank line gaps, ignore multiline blank line gaps. Let me give example diffs for each configuration:
    For each diff, it helps to open the original in one tab and the new version in a second tab and switch back and forth to see what changed visually.
    Winston (talk) 22:57, 9 October 2021 (UTC)[]
    Agreed that a FALSE, TRUE configuration would be better. Multiple blank lines suggest the user intentionally wanted whitespace to be visible (for whatever reason), so maybe we shouldn't remove that. Not sure about the 2nd though. – SD0001 (talk) 14:30, 10 October 2021 (UTC)[]
  • As a developer of a tool that has to go through all kinds of complex markup to detect where comments start and end, I would eagerly support this. But knowing how widespread this incorrect markup (talking about the second example now) is, I'm not sure it would be acceptable. Open any huge talk page that has RfCs, like Wikipedia:Village pump (policy) or Talk:Donald Trump, — you will see a lot of markup like this:
    * Comment. [signature]
    :: Reply. [signature]
    I'm a afraid the bot would be overwhelmed with work, and this could be irritating for users. It's easier for most people who edit wikitext, not use a tool like CD or DT, to type a known number of : instead of repeating the previous indentation.
    Also note that this markup would break:
    *: {|
    | Text.
    While this would not:
    :: {|
    | Text.
    Jack who built the house (talk) 11:26, 10 October 2021 (UTC)[]
    That's an interesting edge case.
    For the busier talk page, maybe the bot could be set to make at most 1 edit per 12 hours (say). But I've seen enough instances of people manually making indentation fixes in others' posts – sometimes even complaining about it in the discussion itself – that I think it would help to have these done automatically. That being said, @Notsniwiast could you make a pass of your code on some of these big talk pages so that we can see the extent of changes it brings about? – SD0001 (talk) 14:31, 10 October 2021 (UTC)[]
    @SD0001 Here's a few. WP:BOTREQ, Village Pump Technical, Village Pump Policy, Help Desk, Teahouse, Village Pump Proposals Winston (talk) 23:35, 10 October 2021 (UTC)[]
    @Notsniwiast Thanks, there's a bug in the fixes around <pre> tags in this discussion (first diff). Pre tags are notorious (syntaxhighlight tag doesn't cause the same issue) for breaking the apparent layout (as seen from wikitext). – SD0001 (talk) 07:59, 11 October 2021 (UTC)[]
    I'm not sure it's really a bug, since the list was broken by putting a new line character (the one before the pre tag) without starting a new indented line. If the newlines before the pre tag are deleted (which I will now do), the formatting looks more correct. However, if it's common to start pre tags on a new line in discussions, then I can build in this edge case. Winston (talk) 08:17, 11 October 2021 (UTC)[]
    Just kidding... apparently deleting the newlines only looks right in the preview. Don't know much about html. Winston (talk) 08:22, 11 October 2021 (UTC)[]
    @SD0001 I've fixed <pre> tag edge case by not treating a newline character which immediately precedes a <pre> tag as a line delimiter. Are there any other tags that should be treated similarly? If so, I can just do the same for all tags. Winston (talk) 08:49, 11 October 2021 (UTC)[]
    No I think, but @Jack who built the house can confirm. – SD0001 (talk) 08:56, 11 October 2021 (UTC)[]
    The problem here is not limited to <pre>...</pre> tags and even tags as such. There is a lot of markup in templates that doesn't stand being moved to a new line, like tables (navboxes, for example, or quotation templates). People can also use <syntaxhighlight>...</syntaxhighlight> like this, despite it is OK with newlines inside of it. <gallery>...</gallery> is sometimes used this way too. I believe doing the same for all tags and templates would be the safest solution here. Jack who built the house (talk) 09:38, 11 October 2021 (UTC)[]
    Yeah I was thinking we might as well do the same for all tags and templates just to be safe, even though I'm not knowledgeable enough to know what exactly might break. I've modified the bot so that newline characters immediately preceding a tag or template are not used as line delimiters. Winston (talk) 10:18, 11 October 2021 (UTC)[]
  • I have added '#' as an indentation character as well. Originally was only using checking asterisks and colons. Winston (talk) 11:00, 11 October 2021 (UTC)[]
    From the 2nd diff (VPT), I think this shows a bug. Indentation of the final comment is changed from **: (which doesn't show a bullet) to :** (which shows a bullet). So shouldn't it have been :*: ? – SD0001 (talk) 11:47, 11 October 2021 (UTC)[]
    Hmm, I think I will modify the algorithm a little. Note that some conversions from bullet points to non-bullet points and vice versa are inevitable (since we shouldn't be alternating back and forth), but I think I can make the algo a bit smarter. Winston (talk) 22:55, 11 October 2021 (UTC)[]
    @SD0001 New diffs: WP:BOTREQ, Village Pump Technical, Village Pump Policy, Village Pump Proposals, Teahouse, Talk:Donald Trump.
    Note that the specific comment you pointed out is still showing a bullet point. This is because Graham87's comment, which is the first comment in the reply chain to Jonesey95's comment, uses a bullet point. To have a consistent indentation style, Izno's comment about a scriptlet must also begin with a bullet point.
    More specifically, the indent style fixing algorithm is now as follows. We iterate over the lines from beginning to end. Let Line A and Line B be consecutive lines. If lvl(Line B) > lvl(Line A), then the first lvl(Line A) characters of Line B are replaced with the first lvl(Line A) characters of Line A. If lvl(Line B) <= lvl(Line A), then the indentation characters of Line B are replaced with the indentation characters of the closest previous line with the same indentation level.
    Ultimately, if we are to enforce consistent indentation style, some comments will inevitably be changed to use or not use a bullet point, e.g. in a comment chain with alternating : and *. Winston (talk) 00:50, 12 October 2021 (UTC)[]
    The Teahouse diff incorrectly 'fixed' a comment:
I just thought of a fourth case - you might want a list. Lists can be numbered, bulleted, or (in your case) unbulleted. This is done using colons, depending on the indentation you want, as (see code or there are details at WP:PLAINLIST):
::This is line 1.
:::::This is line 2.
:::::This is line 3.
If I didn't answer your question, please ask .... --Gronk Oz (talk) 06:17, 10 October 2021 (UTC)[]
 ― Qwerfjkltalk 19:51, 12 October 2021 (UTC)[]
(I've messed up the formatting but you get the idea.) ― Qwerfjkltalk 19:53, 12 October 2021 (UTC)[]
In this case there's not much we can do, since the bot cannot tell when an editor has intentionally over-indented. In this instance at least, the "fix" still looks ok. Winston (talk) 22:33, 12 October 2021 (UTC)[]
  • @SD0001: What do ya think, should I file a BRFA? Winston (talk) 07:35, 14 October 2021 (UTC)[]
    I'm not SD, but I think the only way to find more of the border cases will be to get some trials in. Primefac (talk) 07:38, 14 October 2021 (UTC)[]
    Just filed a BRFA. Winston (talk) 03:36, 15 October 2021 (UTC)[]

Bot to handle unsigned commentsEdit

For a long time this task was handled by SineBot, which has been down for 2 months. Negative effects of this have been visible throughout discussion namespaces. The operator Slakr has not made any edits since last October. There is no indication that SineBot will be operational soon. Is someone else willing to take over this task? ಮಲ್ನಾಡಾಚ್ ಕೊಂಕ್ಣೊ (talk) 15:27, 26 September 2021 (UTC)[]

That bot/function is awesome. -- GreenC 15:35, 26 September 2021 (UTC)[]
I thought I was having to do more unsigned edits than usual lately... seconding the desire to either get in touch with Slakr or take over the bot's functionality. Primefac (talk) 15:37, 26 September 2021 (UTC)[]
Too bad that code for the task is not open source. At some point I think User:Ritchie333 was working on an open-source python implementation. – SD0001 (talk) 15:55, 26 September 2021 (UTC)[]
Yes, the bot should be open source, otherwise this will happen. I have some work in progress code at User:Ritchie333/, however it's currently log only and has bugs. Ritchie333 (talk) (cont) 16:36, 26 September 2021 (UTC)[]
I'd be willing to write a task for this. I'll put together some code. Tol (talk | contribs) @ 00:10, 27 September 2021 (UTC)[]
@Tol: That's great. Feel free to use whatever code in my prototype you think is helpful. I think we need something running ASAP and is well-tested and is open source. I'm not an rms fanboy, but I do think closed-source bots are entirely incompatible with a free encyclopaedia. Ritchie333 (talk) (cont) 08:29, 27 September 2021 (UTC)[]
@Ritchie333: Thanks! I've got a basic implementation; I'm currently testing it and will file a BRFA shortly. Tol (talk | contribs) @ 21:15, 27 September 2021 (UTC)[]
I'm now working on a better diff, and also bundling API calls when checking user pages for preferences. I should have a BRFA by the end of the week. Tol (talk | contribs) @ 20:40, 28 September 2021 (UTC)[]
SinBot's Changelog gives some idea of issues that came up during development. Looks like a lot of edge cases. -- GreenC 14:59, 27 September 2021 (UTC)[]
Has anyone tried emailing Slakr to see if they'd be willing to open the source at this point if they've lost interest in maintaining the bot? --Ahecht (TALK
) 14:57, 27 September 2021 (UTC)[]
I'll be happy to nudge them via email another time (I see you've already emailed Slackr). 🐶 EpicPupper (he/him | talk) 23:29, 27 September 2021 (UTC)[]
Slakr was editing today, I saw his name on the Deletion log and I noticed because he's not one of the regular admins who patrol CSDs and XfDs. Liz Read! Talk! 22:59, 28 September 2021 (UTC)[]
I have already run a sign bot on 6 wiki projects for years. I would be honored if I could transplant these codes to English Wikipedia as a backup of SineBot. Kanashimi (talk) 03:41, 29 September 2021 (UTC)[]
That might be another useful option. As mentioned in SineBot's BRFA(s), having more tools ready is always good in case something like this happens again. 🐶 EpicPupper (he/him | talk) 18:56, 29 September 2021 (UTC)[]
That would be easier (and more well tested) than making my own bot from scratch! I would certainly appreciate if you could do that. Tol (talk | contribs) @ 20:44, 29 September 2021 (UTC)[]
Weird... it thought it was still running but wasn't editing, so it never notified me to restart it. I've restarted it. Sorry for the absence; life's been a bit more chaotic than usual. --slakrtalk / 21:19, 30 September 2021 (UTC)[]
No problem; thank you! Tol (talk | contribs) @ 22:07, 30 September 2021 (UTC)[]
Thanks slakr! It's good to hear from you. — The Earwig (talk) 00:51, 1 October 2021 (UTC)[]
@Slakr: For continuity and redundancy, would you consider publishing the source code for this task? ProcrastinatingReader (talk) 18:54, 15 October 2021 (UTC)[]
For when it's not working, I'd advise people to use User:Anomie/unsignedhelper, which is perhaps one of my most useful scripts. Sdrqaz (talk) 00:08, 1 October 2021 (UTC)[]
Sdrqaz Tried it and have to say that is super cool, and fun. Look forward to finding unsigned comments (not yet caught by SineBot which is pretty fleet). -- — Preceding unsigned comment added by GreenC (talkcontribs) 18:22, 11 October 2021 (UTC)[]

Contribution ProjectEdit

hello i am currently running a Project, which counts User's Contributions on daily basis, The Project works on several wikis such as (ckbwiki, SimpleWiki, ksWiki, ArWiki, jawiki) it works by checks User's Contributions and comparing it to previous Contribution, Ranking Top User's Accordingly, if a user is less active than before The Comparison will change to Red otherwise Green, it also Shows User Rights along with their contributions,

i need someone's Help, to make a bot Specially for the Project, cuz i am doing it Manually by myself and it takes so much time and energy, Can anyone help me to make a script for,   i really appreciate it. —— 🌸 Sakura emad 💖 (talk) 19:48, 1 October 2021 (UTC)[]

We have Wikipedia:List of Wikipedians by number of edits which is updated by User:BernsteinBot; is this not what you're looking for? Primefac (talk) 17:57, 2 October 2021 (UTC)[]
@Primefac that's right but still i want to install my own script because i need to run it on other projects too, it really helps me alot if someone lends me a hand with coding Besides my project a little bit different than above. —— 🌸 Sakura emad 💖 (talk) 11:53, 3 October 2021 (UTC)[]
See if MZMcBride will share the bot's code? Primefac (talk) 12:01, 3 October 2021 (UTC)[]
@Primefac Thank you alot. —— 🌸 Sakura emad 💖 (talk) 19:20, 3 October 2021 (UTC)[]

Hi Primefac and Sakura emad. Wikipedia:List of Wikipedians by number of edits/Configuration is the bot's source code for this report. --MZMcBride (talk) 22:36, 13 October 2021 (UTC)[]

Categorising redirectsEdit

Hi, I presume this has been requested previously, but is there any reason why a bot couldn't categorise redirects? I think at least {{R to diacritic}}/{{R from diacritic}} could be done, maybe {{R to section}} and {{R to anchor}}. Some more can also probably be used, but I can't think of them. ― Qwerfjkltalk 22:01, 2 October 2021 (UTC)[]

Quite a few redirect templates could be added, preferably within {{Rcat shell}}. Others are tempting but probably best left to humans, e.g. deciding between {{R from other capitalisation}} and {{R from miscapitalisation}}. This does sound like a perennial request but I can't find previous discussions. WP:WikiProject Redirect may be more helpful. Certes (talk) 00:29, 3 October 2021 (UTC)[]
@Qwerfjkl: AWB's general fixes include a Redirect tagger which could be run as a bot. If you could point to a conversation where there in consensus to run it as a bot, I'd submit the BRFA. GoingBatty (talk) 15:27, 3 October 2021 (UTC)[]

Bot to change the afghan flag used on the sports pages between 2013-21Edit

The flag template for the afghan flag was changed a few days back to reflect one that is being currently used. Due to this change, all the previous pages (particularly the sports pages between 2013-21) are now also showing the current flag instead of the one which was being used at that point of time.

A bot which tracks content by date/year so as to replace   with  , since the latter is the one which was being used at that point in time. Please ping for a reply.--Anbans 585 (talk) 18:28, 3 October 2021 (UTC)[]

@Anbans 585:   Doing... via AWB. GoingBatty (talk) 19:36, 3 October 2021 (UTC)[]
@Anbans 585:   Done! I updated the instances of {{flagicon}}, {{flagcountry}}, {{flagdeco}}, and {{flagu}} containing the value "AFG" or "Afghanistan" where it was easy to tell that the article was about a specific period of time, such as having a year in the article title. If you want to compile a list of other articles that need updating (along with the appropriate year per Template:Country data Afghanistan), feel free to let me know. Thanks! GoingBatty (talk) 21:29, 3 October 2021 (UTC)[]

@GoingBatty: Great. Can you please do the same on {{fb}}, {{fbw}}, {{futsal}}, {{fsw}}, {{beachsoccer}}, {{fsu}}, {{fbu}}, {{fbwu}}, where a number would already be there after fbu, fsu and fbwu (example 17), and 2013 would have to be appended after that, for the same time period.--Anbans 585 (talk) 13:26, 4 October 2021 (UTC)[]

@Anbans 585:   Doing... GoingBatty (talk) 14:04, 4 October 2021 (UTC)[]
@Anbans 585:   Done! GoingBatty (talk) 15:20, 4 October 2021 (UTC)[]

@GoingBatty: Thanks again. Some more please, found this as I was reviewing the pages. {{fb-rt}}, in short '-rt' appended to all the ones I mentioned in the previous one. Another example {{fsu-rt}}.--Anbans 585 (talk) 16:03, 4 October 2021 (UTC)[]

@Anbans 585:   Doing... GoingBatty (talk) 16:09, 4 October 2021 (UTC)[]
@Anbans 585:   Done! GoingBatty (talk) 16:24, 4 October 2021 (UTC)[]

@Anbans 585: I'm also adding years to {{flag|Afghanistan}}, and converting {{AFG}} to {{flag|Afghanistan|2013}} when needed. GoingBatty (talk) 14:03, 5 October 2021 (UTC)[]

Unused Templates reportsEdit

This is now my fourth time trying to get the necessary feedback. This is a request I made on the talk page of the Database reports page, and on Request a query, but it hasn't been answered anywhere. It is part of the overall proposal discussion I have started on the WikiProject Templates talk page to create an Unused Templates Task Force to deal with the backlog of unused templates.

I've requested four reports per the original discussion and I'm going to relist them here:

I would like the report to run for at least two months as the task force is currently in an idea stage. When the reports are going to expire, if possible I would like to be notified of when it will happen. I need four reports from the Unused Templates database:

1) All unused templates that are not stubs or redirects.

2) All stub templates listed as unused. According to one of the users on the talk page discussion, there are either exactly or about 1,000 stub templates.

3) All redirects listed as unused. According to one of the users on the talk page discussion, there are either exactly or about 69,000 redirects.

4) Templates that were created and/or edited in the last year and the present year. --WikiCleanerMan (talk) 22:53, 3 October 2021 (UTC)[]

@WikiCleanerMan: Is Wikipedia:Database reports/Unused templates what you're looking for? Tol (talk | contribs) @ 04:14, 4 October 2021 (UTC)[]
That report could be a starting point for working on the above requests, but it contains too much noise to be useful. See this discussion for details; it led to this request. – Jonesey95 (talk) 05:32, 4 October 2021 (UTC)[]
User:Zackmann08 before he retired started doing some work at User:Zackmann08/unused templates and related sub-pages, removing various pages from the "unused" criteria. Maybe that can help. Gonnym (talk) 12:39, 4 October 2021 (UTC)[]
Gonnym, the pages by Zackmann is a start, but it's not really speific to what I need, but I'll definitely use those pages. Tol, techinally yes, but my request is looking for separate individual reports that includes the templates I need, but outside the main unused templates database report(s). --WikiCleanerMan (talk) 14:25, 4 October 2021 (UTC)[]
Another thing to point out is that Tim.landscheidt also has similar subpages of unused templates. MZMcBride, your bot, User:BernsteinBot, created those reports. Do you think the reports I'm looking for can be created? --WikiCleanerMan (talk) 15:57, 4 October 2021 (UTC)[]

Hi WikiCleanerMan and Jonesey95. Thank you both for your efforts in cleaning out unused templates. I took a look at Wikipedia:Database reports/Unused templates/Configuration and I remembered that we basically already tried to accommodate all of these requests previously. We sort non-redirects first, then redirects, then stubs within the report. We also include metadata about each unused template, including last edit date and number of unique authors. I'm not really sure what else is needed here currently. --MZMcBride (talk) 22:45, 13 October 2021 (UTC)[]

MZMcBride, we want broken up reports from the main database report. Each individual one for unused templates, then redirects, then the stub templates. --WikiCleanerMan (talk) 23:27, 13 October 2021 (UTC)[]
I'm going to as the potentially dumb question, but if they're already sorted into those values, why do they need to explicitly be split into separate tables? Primefac (talk) 11:56, 14 October 2021 (UTC)[]
To navigate easier so we don't get confused with what is unused stub templates, redirects, and just regular templates. Since regular templates make up the first two and a part of third database reports. And regular unused templates are up to number 12205. Thus, what we are looking for could be on one database page or two. --WikiCleanerMan (talk) 14:29, 14 October 2021 (UTC)[]
If you already know the first redirect (which to me looks green anyway), then you already know which values you're looking for (since they're "everything before that template"). If you know the first two pages (and a bit) of the dbase report are exactly what you're looking for, why does it need to be split? Primefac (talk) 20:08, 14 October 2021 (UTC)[]
So we don't have a database report that includes all unused, redirects, and stubs as part of the same report. --WikiCleanerMan (talk) 22:17, 14 October 2021 (UTC)[]

Jonesey has created a subpage dealing with just the non-stub and non-redirect templates. Really it now stands at 6,800 templates. --WikiCleanerMan (talk) 03:15, 17 October 2021 (UTC)[]

This request can be closed thanks to Jonesey95 who was able to create the filtered report I was looking for. --WikiCleanerMan (talk) 18:56, 18 October 2021 (UTC)[]

Talk Pages Missing Project BannersEdit

I've noticed that sometimes some articles, categories, and templates have no related WikiProject banners for the subject of the related articles, categories, and templates talk pages. If possible, I would like a bot that can create essentially a database report of all articles, categories, and templates talk pages that don't have the project banners. As part of the report, it can list all uncreated talk pages to differentiate. This would save me and other editors a lot of time in helping organize relevant pages for their respective projects. --WikiCleanerMan (talk) 19:47, 9 October 2021 (UTC)[]

One category that the bot could go ahead and tag are the disambiguation pages. Some of those pages don't even have the talk page created. — DaxServer (talk to me) 20:51, 9 October 2021 (UTC)[]
WikiProject Disambiguation has decided not to create talk pages simply to carry its banner. We only add the banner if the page already exists, whether to carry other projects' banners or for other reasons. Certes (talk) 21:03, 9 October 2021 (UTC)[]
That's interesting. I've been wondering why many of those pages lack the talk page, when there are bots and AWB gen fixes running. — DaxServer (talk to me) 21:10, 9 October 2021 (UTC)[]
The dab talk template is often used when the page would otherwise become blank. For example, perhaps Foo was moved to Foo (novel) because someone created Foo (album) and there is no longer a primary topic. The move leaves behind a redirect FooFoo (novel), which we edit into a dab called Foo listing both articles. The move also leaves behind an unwanted redirect Talk:FooTalk:Foo (novel), which we edit into a project banner placeholder. Certes (talk) 00:35, 10 October 2021 (UTC)[]
@DaxServer:   Doing... - I haven't run BattyBot 34 to tag disambiguation existing pages since May, so I'll run it now for you. GoingBatty (talk) 01:31, 10 October 2021 (UTC)[]
@DaxServer:   Done - added {{WikiProject Disambiguation}} to 432 existing talk pages. GoingBatty (talk) 05:03, 10 October 2021 (UTC)[]
Thanks for the explanation, Certes, and the run, GoingBatty! — DaxServer (talk to me) 08:36, 10 October 2021 (UTC)[]

I forgot to add if the report also includes empty categories, then it should differentiate that by listing those categories as empty on a separate list or section. --WikiCleanerMan (talk) 14:41, 10 October 2021 (UTC)[]

A related idea I have is that a bot like this can create suggestions for talk pages of articles, categories, and templates for certain project banners that may be missing from the talk page. I've noticed throughout Wikipedia where the relevant project banner isn't there until I add it. Can save a lot of time with a list to improve and help improve the WikiProjects objectives. It should also include taskforces of the overall project as it is hard to sometimes know what taskforces are part of the project. --WikiCleanerMan (talk) 19:10, 13 October 2021 (UTC)[]

@WikiCleanerMan: What criteria would the bot use to make suggestions? Could you please give some examples? Thanks! GoingBatty (talk) 20:36, 13 October 2021 (UTC)[]
GoingBatty, not sure of the criteria, but there are these WikiProject taggig bots that would add the banners on the talk page for the related projects. Most are inactive at this point. If a bot can add or in my request suggest projects and task forces in a database report, then I don't think it would be all that different. Although, they were designed for a specific project like WikiProject Germany, Film, etc... --WikiCleanerMan (talk) 21:10, 13 October 2021 (UTC)[]
There is a regular Empty Category list that gets issued daily until Bernstein Bot got hiccups recently. It's not a long list as Empty Categories are regularly tagged and deleted unless they fit one of the few exceptions (disambiguation categories, redirect categories, categories that are part of a CFD discussion, maintenance categories, etc.). I'm not sure of the connection you are making between WikiProject banners and empty categories. Unless there is a major category restructuring going on or a sockpuppet has gone on a category-creation rampage, categories typically become empty when the articles, templates and other categories that they contained are deleted which happens on an irregular basis, depending on AFD, TFD and CFD decisions.
A recent exception to this predictability is that over the summer UnitedStatesian and I went through defunct WikiProjects and had hundreds of empty assessment categories deleted but this action focused on categories that had been created and abandoned over the past 15's unlikely there will be a need for an action like this to happen again for a long, long time as WikiProject activity has declined since about 2012. I don't see other big projects like this on the horizon. Liz Read! Talk! 01:28, 14 October 2021 (UTC)[]
The connection is that usually empty categories don't have project banners on their talk pages. Sometimes these categories are newer and yet to have any articles in them. The users who create such newer and empty categories are probably not aware of WikiProject banners and thus don't place them on the talk page. --WikiCleanerMan (talk) 01:40, 14 October 2021 (UTC)[]

Bot notifications when Template:GetMop is published on a user talk pageEdit

I would like a bot to notify users listed at Wikipedia:Administrators without tools/Endorsers whenever a user talk page is published with {{subst:GetMop}} in place. This is to give an opportunity, to those who have signed the page, to endorse the original posting with their own encouragement. Past usage of this template has shown that secondary endorsements increase the likelihood of positive results. If the bot could run this task once a day, the objective can be afhieved (twice a day would be better). Thank you for considering this request. Best regards.--John Cline (talk) 00:47, 14 October 2021 (UTC)[]

@John Cline: I could do this. That page looks like a redlink — is it going to be created later? Tol (talk | contribs) @ 02:01, 14 October 2021 (UTC)[]
Thank you Tol. Yes, in fact I was in the process of creating it now. And then I am going to post it to Wikipedia:Requests for adminship/2021 review/Brainstorming#Discussion (An underused ready-resource) to invite others to sign the page. Let me know if you need anything from me to help get this up and running. Thanks again. --John Cline (talk) 02:23, 14 October 2021 (UTC)[]
@John Cline: Alright; I should have a working implementation over the weekend. Just so you know, it will look for any links to userspace (not just in the list in the section). Tol (talk | contribs) @ 04:54, 14 October 2021 (UTC)[]
I'm not sure I understand what you mean here. Quite often users allow this to be archived and at times they will add it to their user page in an barnstar Hall of Fame or some such. I wouldn't want this to promt the bot to post messages. That's why I hope it will know when the template has been substituted because that only happens durring the initial posting. --John Cline (talk) 05:10, 14 October 2021 (UTC)[]
I believe Tol means that any user talk page links on Wikipedia:Administrators without tools/Endorsers will get the notification. So don't create a second list of "Users who definitely don't want a notification"! I think that's a perfectly good solution, this is a definite opt-in method. Tol, do you need a notification template for the talk page notifications? WormTT(talk) 08:53, 14 October 2021 (UTC)[]
Yes, I'm sure that you are correct. It makes sense to me now. Thank you.--John Cline (talk) 09:37, 14 October 2021 (UTC)[]
@Worm That Turned: Yes, you are correct. A notification template would be helpful. Thank you! Tol (talk | contribs) @ 17:56, 14 October 2021 (UTC)[]
@John Cline: I've finished writing the code. Is there a template (like Worm That Turned mentioned) that the bot could use? Tol (talk | contribs) @ 00:58, 16 October 2021 (UTC)[]
Thank you for your diligence in this regard, I really appreciate it. I haven't written the message yet, I will ensure it's ready for your review, and approval, no later than Monday. I'd be happy if Worm That Turned was willing to write it, he writes particularly well (in my opinion). I do have 1 question for you Tol: Would it be best for you for the notification template to be a separate, stand alone template, like {{GetMop notification}}, for example (perhaps for reasons like page protection)? Or would you be ok using switched output from {{GetMop}}, for example, {{GetMop|notice}} could be used. Or I could create a sub template at {{GetMop/Notice}}. Let me know what you think is best. Thank you again. Sincerely.--John Cline (talk) 01:36, 16 October 2021 (UTC)[]
@John Cline: You're welcome! Personally, I think that a subpage would be best: it's related to Template:GetMop, but I think passing a parameter that would cause the template to do something completely different is suboptimal. Having an entirely separate template would also work. Tol (talk | contribs) @ 02:22, 16 October 2021 (UTC)[]
Okay Tol, I've created {{GetMop/Notice}}, and worked up a message box and message that I am satisfied with. I tested the templat's substitution output and everything works fine It's ready for you to do what you do. I am looking forward to seeing it's working use. Best regards.--John Cline (talk) 13:57, 16 October 2021 (UTC)[]
@John Cline: Alright,   BRFA filed. I've also tidied up that template (and added a parameter so that the user(s) who have had Template:GetMop on their talk page can be listed inside the box). Is there a reason the message itself is includeonly? I think it would be easier to just display the template on the template page. Tol (talk | contribs) @ 18:09, 16 October 2021 (UTC)[]
Many thanks again. I'll remove the includeonly tags, there's no compelling reason for them and it's more habit than anything else. I've convinced myself that, in some small way, it discourages drive by vandalism under the concept: if they don't see it, they don't mess with it. It may be an imaginary precaution with no foundation, but it has affected my habit. I appreciate any copyedits you gave or ever give, it's how I learn better ways. Have a good one and be well.--John Cline (talk) 19:54, 16 October 2021 (UTC)[]
  • Hello Tol. I looked at the   BRFA filed, and I do have one concern. In Function details:, where It says: "It runs daily and checks if there have been any user talk pages added to Category:Administrators without tools since its last run. If there have been, it then sends notifications". My concern is that when the discussion archives, I wouldn't want the bot to recognize the archive as a new talk page being added. To be certain, it must only be "user talk pages where {{#ifeq: {{FULLPAGENAMEE}} | User_talk:{{ROOTPAGENAMEE}} |yes|no}} returns yes for the new user talk page being added. Another issue that may arise involves formating the user name and discussion links to append through parameter 1. Assuming the possibility of something standard, like: "John Cline (talk) at There is a mop reserved in your name", and because the section link will always be the same title, that format would be too long to keep the entry on one line for almost every username used. To preempt, I wanted to suggest possibly using the following format: "John Cline (talk) at this discussion" or something similar. Keeping the entry on one line is important because the centered text looks bad when only a few words fall off the end. See [[User:John Cline/sandbox|my sandbox for a visualization. Thank you.--John Cline (talk) 03:28, 17 October 2021 (UTC)[]
    @John Cline: If you have concerns, it's better to raise them at the BRFA itself. I had set up a filter to blacklist any pages with "/Archive" in them, and this caught all current instances of archives in the category. The link itself would currently be formatted as the username, linking to the discussion (like John Cline, which links to User talk:John Cline#There is a mop reserved in your name). Tol (talk | contribs) @ 03:57, 17 October 2021 (UTC)[]

Bot for maintenance at WP:CREdit

Wikipedia:Closure requests wants the requests in chronological order and the {{Initiated}} with full timestamp instead of just dmy ([6] [7]). Some miss the initiated template [8]. Could a bot takeover for these tasks? — DaxServer (talk to me) 13:41, 15 October 2021 (UTC)[]

How often are people misreading the guidelines and thus necessitating the page be edited to conform to standard practice? If it's less than once a day, I think manual cleanup is perfectly acceptable. Primefac (talk) 14:28, 15 October 2021 (UTC)[]
It looks like it's not as often as I thought — DaxServer (talk) 08:19, 17 October 2021 (UTC)[]

Grammar bot or spelling bot?Edit

Wondering if there is a grammar bot or a spelling bot for clear mistakes? I can't seem to find one. Also WP:List of bots doesn't seem to show what the bots do. Thanks much, Facts707 (talk) 09:16, 18 October 2021 (UTC)[]

@Facts707: Both these points, regarding grammar and spelling, have been discussed at Wikipedia:Bot requests/Frequently denied bots#Bots to automatically spell-check articles. Regarding the list of bots, I will soon get back to you :) —usernamekiran (talk) sign the (guestbook) 09:23, 18 October 2021 (UTC)[]
Agree on the list of bots, there should be more documentation added on what this list is supposed to include (e.g. not all bots?), and what each bot does. Maybe operator talk page links would also be nice :) . 🐶 EpicPupper (he/him | talk) 15:37, 18 October 2021 (UTC)[]
WP:AWB and WP:JWB also fix many spelling mistakes and a few grammatical errors as and when contributors edit a page for other reasons. They encourage a human to verify the change, to preserve genuine use such as foreign words, olde worlde spellinges and verbatim quotes of careless journalists. Certes (talk) 12:19, 18 October 2021 (UTC)[]
  • dummy edit. I am not being shown as a bot-op editor like I used to be in the table of contents above. —usernamekiran (talk) sign the (guestbook) 16:38, 18 October 2021 (UTC)[]
    dummy edit after updating the signature. —usernamekiran • sign the guestbook(talk) 16:59, 18 October 2021 (UTC)[]
  • @Facts707: Wikipedia:Bots/Status seems to be the one you are looking for, but it needs to be updated. —usernamekiran • sign the guestbook(talk) 05:48, 19 October 2021 (UTC)[]
    That page is manual so there's no way it can be updated without hours of effort (and it would be outdated soon after that). There's a bot-maintained WP:BAM but it's incomplete and needs listing of remaining bots. – SD0001 (talk) 07:16, 19 October 2021 (UTC)[]
  • Thanks everyone for the info and support! WP:SPELLBOT is a shortcut to Wikipedia:Bot requests/Frequently denied bots#Bots to automatically spell-check articles that @Usernamekiran: mentioned that explains why fully-automated such bots aren't a great idea. I added a hatnote there to WP:Typo Team – lots of help there. Will look at bot lists some more a bit later. Cheers, Facts707 (talk) 09:55, 19 October 2021 (UTC)[]
Update: my hatnote to WP:Typo Team was reverted with the comment "that's not what this section is about". I will try WP:SPELLING or somewhere else to add that. It can be hard sometimes to find things in WP space – I try to add a redirect or a hatnote sometimes to make it a little bit easier for the next editor... Facts707 (talk) 10:13, 19 October 2021 (UTC)[]

Bot to add Top 25 report and Annual report to talk pagesEdit

I think this would be helpful, as this is perennial. Thoughts? (Or is there a bot already?) — DaxServer (talk) 07:04, 20 October 2021 (UTC)[]