Wikipedia talk:WikiProject Check Wikipedia/Archive 8

 Done

Hello! 34th error doesn't detects those template elements:

• {{NAMESPACENUMBER}}
• {{#tag:...}}

There're also a lot of other magic words; most of them shouldn't be used in main namespace, but I don't think that such words as {{CURRENTVERSION]} CAN be used in any namespace. Some words, such as {{SUBJECTSPACE}}, can, but very rarely. I can make a list of those words, if you need. Facenapalm (talk) 03:30, 23 August 2016 (UTC)[reply]

Facenapalm That's a good idea. Also include some crap VE is putting in articles, such as _noeditsection_ and _noindex_. {{#tag:...}} is valid per WP:REFNEST.

Could you make up the list and I'll see how common they are used. I'll need to talk to someone much smarter than me to see if they can be removed or not. Bgwhite (talk) 04:37, 23 August 2016 (UTC)[reply]

Your note about #tag sounds legit. Ok, I grab some statistics:

• NAMESPACENUMBER: 114 in ruwiki (I'll work with it), 0 in enwiki

Technical metadata:

• SITENAME: not used in ruwiki, idk for enwiki
• SERVER: used for both ruwiki and enwiki, I don't know how often. Example for enwiki: Ramee Mall
• SERVERNAME: don't used for both ruwiki and enwiki
• DIRMARK/DIRECTIONMARK: same
• SCRIPTPATH: used only in one article in enwiki, no matches in ruwiki
• STYLEPATH: don't used for both ruwiki and enwiki
• CURRENTVERSION: same
• CONTENTLANGUAGE/CONTENTLANG: one article in enwiki, none for ruwiki
• PAGEID, PROTECTIONLEVEL, PROTECTIONEXPIRY, CASCADINGSOURCES, REVISIONID etc - I'm not sure there's a sense to use those magicwords at all

Page names:

• SUBJECTPAGENAME/ARTICLEPAGENAME: one for ruwiki, one for enwiki
• TALKPAGENAME: 23 for ruwiki, 33 for enwiki
• ROOTPAGENAME: none for both ruwiki and enwiki

URL encoded page names doesn't used in ruwiki at all. Enwiki statistics:

• FULLPAGENAMEE: 11
• PAGENAMEE: 1
• BASEPAGENAMEE: 0
• SUBPAGENAMEE: 0
• SUBJECTPAGENAMEE: 0
• ARTICLEPAGENAMEE: 1
• TALKPAGENAMEE: 1
• ROOTPAGENAMEE: 0

Namespaces (SUBJECTSPACE/ARTICLESPACE, TALKSPACE) not used in both ruwiki and enwiki. Same on URL encoded namespaces: NAMESPACEE, SUBJECTSPACEE/ARTICLESPACEE, TALKSPACEE. Idk for "ns".

URL data:

• localurl: 4 for ruwiki, 1 for enwiki
• fullurl: 4 for ruwiki, none for enwiki
• canonicalurl: 0 in both wikis
• filepath: 0 for ruwiki, idk for enwiki - I think 0 too
• urlencode: there are uses in both wikis, but it seems they're legal
• anchorencode: 0 for both wikis

Formating and Localization is widely used in both wikis, no sence to ban :) I don't know statistics for Miscellaneous section - #language, #special.

IMHO, It's useful to detect all cases from technical metadata until PAGEID except CURRENTVERSION, page names, URL encoded page names and URL data except urlencode. Facenapalm (talk) 05:37, 23 August 2016 (UTC)[reply]

Hi. It will change nothing in your request or your statistics, but FYI only: revision magic word is very useful and very powerful, because it allows to create "anti"-editnotice - a page that has less text in edit mode than in reading mode, and we use it like that indeed. IKhitron (talk) 08:54, 23 August 2016 (UTC)[reply]

Hmm. May I recieve an example when that's needed? Facenapalm (talk) 11:42, 23 August 2016 (UTC)[reply]

It's up to you and your wiki needs. An example from our wiki: a template "read or edit" that has two parameters: what to show in read view and what to show in edit view. Each of them can also be empty, if you wish. It's much better than editnotice because you can use parameters conditions and so on. A concrete example for usage: We have a template to show a red articles that should be created immediately, in recentchanges. Tha read view shows a lot of explanations, using a lot of pictures, when and how you allowed to add some red article. It has no sections edit because there is only one place where user allowed to edit, using inputbox. So, we use the revisions mechanism to remove all the pictures in edit mode, leaving only the explanations, so the user should not scroll a lot until he finds where to edit. There are many other options how to write this, but we use revisions as the better one. IKhitron (talk) 12:07, 23 August 2016 (UTC)[reply]

There are 134 cases of _noeditsection_, mostly from VE. NAMESPACENUMBER, TALKPAGENAME and FULLPAGENAME look to be candidates to add to #34. If it is common, I'll add it. If it's not common, I'd rather not. I'll take a look and see if those three have a valid use for them. Bgwhite (talk) 22:18, 23 August 2016 (UTC)[reply]

FULLPAGENAME already detected, isn't it? Here's a difference between FULLPAGENAME and FULLPAGENAMEE. :) Facenapalm (talk) 00:36, 24 August 2016 (UTC)[reply]

Facenapalm Where are you getting the amount that NAMESPACENUMBER, TALKPAGENAME and FULLPAGENAMEE are used? I must be doing something wrong because I can't find any in articles. Bgwhite (talk) 21:28, 24 August 2016 (UTC)[reply]

The amount of NAMESPACENUMBER is zero for enwiki. TALKPAGENAME: insource:"TALKPAGENAME" search query. Same for FULLPAGENAMEE. Facenapalm (talk) 21:35, 24 August 2016 (UTC)[reply]

@Bgwhite: btw, as I understood, Check Wikipedia now doesnt' detects any unsolved parameters (such as {{{param}}}) in ruwiki, because /\{\{\{/ regexp doesn't works due to legal {{{!}}. Probably it's better to change this regexp to /\{\{\{[^!]/? I have found some pages with uncorrecttly substituted templates, which is not detected by checkwiki 'cause there's no template elements except params. Thanks! Facenapalm (talk) 12:24, 25 August 2016 (UTC)[reply]

Facenapalm you are correct about {{{param}}}. Do you have any example pages? I didn't know about "insource" as a search options, thank you. Will definitely be adding those three magic words to #34. Bgwhite (talk) 20:54, 25 August 2016 (UTC)[reply]

You can even search via regexp: insource:/\{\{PAGENAMEE?\}\}/. But be careful and don't trust in results count: regexp can find not all cases due to time limit. Regexps are very expensive, you know.
Yes, I'm talking about parameters like {{{param}}} in main namespace. Checkwiki detects them by searching "{{{" in wikicode for all wikis except ruwiki, because in ruwiki we usually insert tables into templates via {{{!}} code (instead of {|, which is not works correctly in templates). I think you remember: I wrote about it twice, because it led to thousands of false positives in 43rd and 34th errors. Your solution for 34th error was just to delete detection by "{{{". I ask you to re-add the detection via /\{\{\{[^!]/ regexp. (I missuderstand you, sorry) The example is here: there're a lot of template elemets {{{1|22px}}} but them don't detected by 34th error. But, as I noted just now, it detected by 43th one. 0_o Facenapalm (talk) 21:39, 25 August 2016 (UTC)[reply]

Btw, 34th error also detects incorrect substitutions ({{subst:). How about to find translated keywords as well? For example, in ruwiki templates can be substituted by {{подст: keyword. Facenapalm (talk) 16:13, 27 August 2016 (UTC)[reply]

@Bgwhite: any news? Simple and powerful idea: allow to specify your own regular expression in local translation file. Probably, somewhere {{PAGENAME}} is allowed, somewhere "{{{" isn't something bad (ruwiki), somewhere you can substitute templates via localized keywords. That's much better to specify all this stuff in local translation file instead of code. In other words, customization is always good. :) Facenapalm (talk) 22:38, 29 August 2016 (UTC)[reply]

Facenapalm I don't get to code too often, so I can be sloooow.
#34 now checks for {{namespacenumber}} {{talkpagename}} {{fullpagenamee}} __noindex__ __index__ __forcetoc__ __nonewsectionlink__

Your idea is a good one and will be adding it. Bgwhite (talk) 21:00, 30 August 2016 (UTC)[reply]

@Facenapalm, NicoV, and Magioladitis: It has been coded up and will add it tomorrow. To many things changed today. Should add whatever you are looking for to the translation file (ie error_034_templates_ruwiki=). The program adds {{ to the front, so adding "foo" to the translation file and checkwiki will look for {{foo. Bgwhite (talk) 22:29, 30 August 2016 (UTC)[reply]

error_034_templates is just list, not regexp? I wished to add "{{{[^!]" to regexp to see what happeneds... And I also not sure that __forcetoc__ in ruwiki is something bad. Well, ok, that solution isn't bad too. Can I write "подст:" to detect all cases like {{подст:PAGENAME}}? Facenapalm (talk) 23:48, 30 August 2016 (UTC)[reply]

Facenapalm I checked 20 cases of _forcetoc_ and none of them were needed. I do expect valid cases. If there are too many valid cases, I'll yank it out. Hmmm, if error_034_templates contained parenthesis (ie {{подст ), then the list could also contain magic words. I'll change it so you have to mention {{ in the list. Yes, writing подст: will catch all cases like {{подст:PAGENAME}}. Bgwhite (talk) 06:28, 31 August 2016 (UTC)[reply]

@Facenapalm and Magioladitis: It's coded up so you can add подст and _FORCETOC_ to the translation file. I've removed _FORCETOC_ from the code and added it into English's translation file. Did it so I and Magioladitis suffers if there are too many false positives rather than everybody else. Bgwhite (talk) 19:08, 31 August 2016 (UTC)[reply]

@Bgwhite: cool, really cool! Can you please explain the algorithm of error_032_templates_xxwiki lines parsing or update the code in github repository? Or, probably, there's another source where I can read the last version of checkwiki code? The algorithm is: any lines, which started with __, will be searched as is, all others - with "{{" at the beginning; or what? Facenapalm (talk) 21:47, 31 August 2016 (UTC)[reply]

Facenapalm There were 21 articles caught by Checkwiki that had _FORCETOC_. Only one was valid. {TALKPAGENAME} has a valid use in some templates, so I'll be removing that from the code. You can add it via your translation file.

Are you sure you are talking about #32 for __ and {{? Do you mean #34. If #34, it's just a simple regex... if ( $text =~ /($temp)/i ) where $text is the article and $temp is _FORCETOC_ or {TALKPAGENAME}. It's set in a while loop. Bgwhite (talk) 05:34, 1 September 2016 (UTC)[reply]

• rev 10500 Fix for references inside small tags -- Magioladitis (talk) 20:23, 22 November 2014 (UTC)[reply]
• rev 10505, rev 10515 improves DoublePipeLink (error #32). -- Magioladitis (talk) 09:24, 1 December 2014 (UTC)[reply]
• rev 10777 extends error #1 fixes to redirects. -- Magioladitis (talk) 16:59, 16 January 2015 (UTC)[reply]
• rev 10800 for error #101. -- Magioladitis (talk) 16:21, 23 January 2015 (UTC)[reply]
• rev 10803 alerts for error #98 and #99. -- Magioladitis (talk) 21:25, 23 January 2015 (UTC)[reply]
• rev 10810 for ISBN format fixes. -- Magioladitis (talk) 23:15, 30 January 2015 (UTC)[reply]
• rev 10811 for the new error #2 additions. -- Magioladitis (talk) 12:29, 1 February 2015 (UTC)[reply]
• rev 11075 ISBN with tab (error #69).
• rev 11073 FixSyntax: handling of unbalanced bracket wikilink with trailing pipe. Unbalanced bracket fix was leading to error #64. now fixed. -- Magioladitis (talk) 18:35, 6 June 2015 (UTC)[reply]
• rev 11229 FixReferenceTags: handle nested empty ref tags (error #85). -- Magioladitis (talk) 15:18, 21 June 2015 (UTC)[reply]
• rev 11231 FixLinkWhitespace: fix whitespace on longer wikilinks, allow non-word character before link (error #86). -- Magioladitis (talk) 12:25, 22 June 2015 (UTC)[reply]
• rev 11238 Fix excess leading pipe in piped link, avoid malformatted image links (error #74). -- Magioladitis (talk) 15:08, 22 June 2015 (UTC)[reply]
• rev 11266 Simple closing bracket fixing to {{R...}} templates (error #43 and #47). -- Magioladitis (talk) 16:12, 23 June 2015 (UTC)[reply]
• rev 11278, rev 11279 Additional bracket fixes (error #43 and #47). -- Magioladitis (talk) 09:31, 26 June 2015 (UTC)[reply]
• rev 11283 catches more pages with double pipe links or links without target (error #32 and #74). -- Magioladitis (talk) 14:59, 30 June 2015 (UTC)[reply]
• rev 11291 more on unbalanced brackets. -- Magioladitis (talk) 19:15, 4 July 2015 (UTC)[reply]
• rev 11295 more on unbalanced brackets. -- Magioladitis (talk) 07:16, 5 July 2015 (UTC)[reply]
• rev 11300 common double pipe error fixes (error #32). -- Magioladitis (talk) 21:49, 6 July 2015 (UTC)[reply]
• rev 11324 fix incorrect closing ref (error #94). -- Magioladitis (talk) 11:12, 13 July 2015 (UTC)[reply]
• rev 11348 fix more bad ISBN (error #69). -- Magioladitis (talk) 08:05, 1 August 2015 (UTC)[reply]
• rev 11356 fix incorrect opening ref (error #94). -- Magioladitis (talk) 11:07, 6 August 2015 (UTC)[reply]
• rev 11367, rev 11373, rev 11377 fix for more simplify link cases (error #54). -- Magioladitis (talk) 12:13, 11 August 2015 (UTC)[reply]
• rev 11380 fix incorrect opening ref (error #94). -- Magioladitis (talk) 14:07, 16 August 2015 (UTC)[reply]
• rev 11381 more ref fixes (error #94). -- Magioladitis (talk) 14:27, 16 August 2015 (UTC)[reply]
• rev 11385, rev 11386, rev 11401, rev 11430 more on unbalanced brackets. -- Magioladitis (talk) 21:33, 27 August 2015 (UTC)[reply]
• rev 11429 more alerts for #32 and #74. -- Magioladitis (talk) 22:56, 5 September 2015 (UTC)[reply]
• rev 11620 more on #93. -- Magioladitis (talk) 11:49, 23 September 2015 (UTC)[reply]
• rev 117553 catches more pages with double pipe links or links without target (error #32 and #74). -- Magioladitis (talk) 09:12, 8 December 2015 (UTC)[reply]
• rev 11759 yet another #69 fix. -- Magioladitis (talk) 09:50, 21 December 2015 (UTC)[reply]
• rev 11840 more on unbalanced brackets. -- Magioladitis (talk) 21:16, 26 January 2016 (UTC)[reply]
• rev 11916 yet another #2 fix. -- Magioladitis (talk) 08:40, 21 February 2016 (UTC)[reply]
• rev 11924 fixes more from #54. -- Magioladitis (talk) 17:14, 21 February 2016 (UTC)[reply]
• rev 11931, rev 11939 duplicate category removal: can remove one cat if its sortkey is a substring of the duplicate cats one (#17). -- Magioladitis (talk) 18:07, 22 February 2016 (UTC)[reply]
• rev 11945 Removes ISBN from isbn field in Infoboxes (part of error #69). -- Magioladitis (talk) 10:36, 24 February 2016 (UTC)[reply]
• rev 11966 fixes issn with dash. (Error #106). -- Magioladitis (talk) 09:24, 6 March 2016 (UTC)[reply]
• rev 11967 fixes issn syntax inside cite magazine. (Error #106). -- Magioladitis (talk) 08:12, 8 March 2016 (UTC)[reply]
• rev 11968 fixes issn syntax inside cite report. (Error #106). -- Magioladitis (talk) 09:17, 8 March 2016 (UTC)[reply]
• rev 11970 more fixes on isbn/issn syntax (Errors #69 and #106). -- Magioladitis (talk) 09:35, 9 March 2016 (UTC)[reply]
• rev 11971 more error #69 fixes. -- Magioladitis (talk) 09:45, 9 March 2016 (UTC)[reply]
• rev 12006 more error #101 fixes. -- Magioladitis (talk) 07:05, 19 April 2016 (UTC)[reply]
• rev 12021 more error #94 fixes. -- Magioladitis (talk) 10:01, 6 June 2016 (UTC)[reply]
• rev 12025, rev 12028, rev 12029, rev 12030, rev 12036, rev 12040, rev 12041, rev 12051, rev 12052, rev 12057, rev 12058 more error #104 fixes. -- Magioladitis (talk) 11:25, 20 June 2016 (UTC)[reply]
• rev 12059 #101 for dates now works correctly. -- Magioladitis (talk) 15:26, 27 July 2016 (UTC)[reply]
• rev 12060, rev 12061, rev 12064, rev 12065, rev 12073, rev 12076 more on #2. -- Magioladitis (talk) 06:48, 28 July 2016 (UTC)[reply]
• rev 12062 fixes a bracket fix problem. -- Magioladitis (talk) 08:15, 31 July 2016 (UTC)[reply]
• rev 12078 critical change to #61 fixes to avoid false positives with colon. -- Magioladitis (talk) 14:14, 15 August 2016 (UTC)[reply]
• rev 12086, rev 12088 #69. -- Magioladitis (talk) 14:48, 23 August 2016 (UTC)[reply]
• rev 12090, rev 12091, rev 12093, rev 12094 #104. -- Magioladitis (talk) 14:02, 24 September 2016 (UTC)[reply]
• rev 12095 #69 -- Magioladitis (talk) 09:34, 25 September 2016 (UTC)[reply]
• rev 12098 #54 -- Magioladitis (talk) 11:04, 20 October 2016 (UTC)[reply]
• rev 12101, rev 12111, rev 12112, rev 12116, rev 12117, rev 12118, rev 12119 #104 -- Magioladitis (talk) 08:25, 13 November 2016 (UTC)[reply]
• rev 12109 #16. -- Magioladitis (talk) 16:27, 22 November 2016 (UTC)[reply]
• rev 12110 #85. -- Magioladitis (talk) 12:41, 24 November 2016 (UTC)[reply]
• rev 12114,rev 12115, rev 12121 #64. -- Magioladitis (talk) 23:30, 26 November 2016 (UTC)[reply]
• rev 12123, rev 12124 #69. -- Magioladitis (talk) 13:29, 30 November 2016 (UTC)[reply]

 Done

Hello. id 14 recognized unclosed source tags. This tag is deprecated. Does the #14 recognizes also syntaxhighlight unclosed tags? And can you create a new id that recognizes any source tags? Thank you very much. IKhitron (talk) 16:08, 29 July 2016 (UTC)[reply]

IKhitron While <source> tags have been deprecated, people are still using them. As adding syntaxhighlight would take 30 seconds to copy/paste and make minor tweeks, I've already added it. I've left it under #14 as the two tags do essentially the same thing. Bgwhite (talk) 21:39, 29 July 2016 (UTC)[reply]

Thanks a lot for adding it, Bgwhite. About people still using it - of course they do, this is why I need every time to find them in wikisearch and replace to the normal form. And this why I asked you if you can add id for recognition. But if it takes time - forget about it, I'll continue as before. IKhitron (talk) 23:58, 29 July 2016 (UTC)[reply]

Hello. Why the following errors are disabled?

• HTML named entities (11)
• Category's first letter is small (18)
• Unicode syntax (27)
• HTML text style element <big> (41)
• en dash or em dash (50)
• Arrow as ASCII art (56)
• Indented list (75)
• Reference duplication (81)
• HTML named entities without semicolon (87)
• Heading double (92)

Are they works correctly? Maybe it's good idea to write the reason for disabling this errors in comments in translation page to prevent such questions? Facenapalm (talk) 00:21, 1 May 2016 (UTC)[reply]

Facenapalm Some of these are before my time and my memory is bad. Maybe Magioladitis can help. Here it goes....

• #18 Deemed a "cosmetic edit" and can't be fixed by itself.
• #41 <big> needs to be replaced because its not in HTML5. However, there isn't a good replacement yet. {{big}} can't be used in all situations.
• #50 there is no consensus on enwiki for this.
• #56 there is no consensus on enwiki for this.
• #75 there is no consensus on enwiki for this. In some cases, it can only be done this way.
• #81 We turned it off. There were so many errors and there was no reasonable way to fix them all.
• I remember playing with 11, 27 and 87, but I can't remember why they are not activated.

Bgwhite (talk) 21:07, 6 May 2016 (UTC)[reply]

@Bgwhite: thank you for the answer! The vast majority of low priority errors is "cosmetic edits" which can't be fixed by itself - at least in ruwiki. On the other hand, there are some categories, which should be written with small first letter, for example, Category:iPhone... Ok, seems like all of these errors are working correctly, but disabled because of lack of consensus. Am I right? I'll try to discuss it. Thank you! Facenapalm (talk) 14:20, 7 May 2016 (UTC)[reply]

If 81 is turned on again, I could try and tackle them. But without a list or updatable database of these, it is nearly impossible...could it please be activated again? (t) Josve05a (c) 18:04, 7 May 2016 (UTC)[reply]

Facenapalm The reasons I've given are for enwiki only. Other wikis can turn off/on errors as they think are ok. Others wikis do have some of these turned on. Just use your best judgement.

Josve05a I can't remember the number, but it was a high amount. I think it was around 30,000 of these errors. Remind me so I don't forget, but I can turn it on for next month's dump. Bgwhite (talk) 20:39, 7 May 2016 (UTC)[reply]

Great! When should remind you? I could bug and ping you once a day, you know how annoying we punks are. (t) Josve05a (c) 21:01, 7 May 2016 (UTC)[reply]

@Bgwhite: Monthly reminder ;) (t) Josve05a (c) 22:10, 14 June 2016 (UTC)[reply]

81 contained a huge amount of false positives. -- Magioladitis (talk) 06:19, 13 May 2016 (UTC)[reply]

Resolved

Hi, I've seen false positives for #19 and #25 due to math tags and equal signs at the beginning of a line :

• Flux thermique
• Projection conique conforme de Lambert
• Signature d'une permutation
• Torseur

--NicoV ^{(Talk on frwiki)} 04:21, 16 September 2016 (UTC)[reply]

NicoV It was due to the update today and is already fixed. #47 has them too. Bgwhite (talk) 06:16, 16 September 2016 (UTC)[reply]

Errors #13, #14, #15, #98 and #99 have been updated to catch unbalanced closed tags too. -- Magioladitis (talk) 06:12, 16 September 2016 (UTC)[reply]

Like Like Like What about #5, #23, #24, #29? IKhitron (talk) 08:08, 16 September 2016 (UTC)[reply]

No sence to do it for #5. "-->" can used in articles, why not? Facenapalm (talk) 08:37, 16 September 2016 (UTC)[reply]

Very well, remove it. IKhitron (talk) 14:51, 16 September 2016 (UTC)[reply]

Follow up to the one above. -- Magioladitis (talk) 23:12, 16 September 2016 (UTC)[reply]

Belissimo. Even 28? I did not asked about it because I thought it's impossible. And what's wrong with #29? Div instead span? IKhitron (talk) 23:57, 16 September 2016 (UTC)[reply]

IKhitron It is me and my incapability to type correctly. I meant 29. -- Magioladitis (talk) 00:03, 17 September 2016 (UTC)[reply]

Great. Now you have only rephrase the desc vars for all these ids - add and vice versa or someghing like this. IKhitron (talk) 00:15, 17 September 2016 (UTC)[reply]

Hi Bgwhite, I'm starting a dump analysis on frwiki with WPC for errors related to categories. It should be completed in a few hours, but I'm already seeing a lot of pages detected by WPC, but not by CW. For some, I'm not sure if I should restrict the detection on my side, or if we should expand the detection on CW side. Examples:

• For #22 : fr:Arts visuels has [[Catégorie : Art visuel|* ]], so whitespace characters in almost every place : before the colon, after the colon, at the end of the sort key
• For #18 : fr:Cadmiage has [[catégorie:Construction aéronautique]], so a "catégorie" with a lowercase letter
• For #52 : fr:Senan (Yonne) has [[Catégorie:Commune de l'Yonne|Senan]] before the last title == Liens externes ==, so a "catégorie" before the last section

WPC will fill up the lists when it has finished its dump analysis (normally a few hours), I'm seeing many other detection in the logs, so there maybe other kinds of discrepancies between our tools. --NicoV ^{(Talk on frwiki)} 17:41, 29 June 2016 (UTC)[reply]

NicoV I've looked at the #22 one and I see why it is happening. Fixing it is another story. Bgwhite (talk) 00:24, 30 June 2016 (UTC)[reply]

Ok. The dump analysis apparently didn't complete, probably a problem with my computer (probably W10 deciding to reboot on its own as usual :-( ). I'll check this evening. --NicoV ^{(Talk on frwiki)} 08:17, 30 June 2016 (UTC)[reply]

Dump analysis is done for the following errors, and I've seen the following discrepancies:

• #17: some categories in double don't seem to be detected. I think this can be due to several reasons:
  • White space characters around the colon
  • Differences with underscore/white space
  • Accented first letter ?
• #18 (huge): also several reasons:
  • Lower case letter for "catégorie"
  • Accented first letter
• #22: various cases
• #52: only a few cases

--NicoV ^{(Talk on frwiki)} 08:48, 5 July 2016 (UTC)[reply]

Bgwhite, I've also added #9 to the dump analysis, a few articles found. --NicoV ^{(Talk on frwiki)} 05:32, 6 July 2016 (UTC)[reply]

Fixed

• #17 with underscores
• #17 with a space before the colon. Space after the colon was already working fine.
• #18 fixed for all accented letters.
• #18 added to test for lower case letter of "catégorie". The definition of #18 only checks for lower case letter on the category name and not "catégorie".
• #22 with a space before the colon. Space after the colon was already working fine.
• #22 with a space before the pipe.

Not fixed

• #52 fr:Senan (Yonne). Article had two identical section headers, == Liens externes ==. Algorithm is, 1) get a headline 2) find its position in the article 3) Go thru each category and find its position in the text 4) if headline position is greater than category position, call error.
  The problem is... when the call to find the position of the second identical headline, it starts from the beginning of the article, thus it finds the position of the first identical headline.
  FYI, I removed the second headline from fr:Senan (Yonne).
• #22 False positive: Leibniz-Zentrum für Informatik: [[Catégorie:Fondation en 1989<!--Pourquoi ceci ? : |Kategorie:Gegründet 1989-->]]

Not sure

• #17 fr:Obselidia: [[Catégorie:Prix Alfred P. Sloan]], [[Catégorie:Prix   Alfred P. Sloan]] Not sure what to do here. What do you do? Do you remove all spaces?

  Bgwhite : WPC normalizes titles before comparing them : replacing non breaking white space by regular white space, removing trailing 0x200E, replacing underscore by white space, replacing consecutive white space by one white space, trimming. --NicoV ^{(Talk on frwiki)} 04:38, 8 July 2016 (UTC)[reply]

Today, I first corrected all of the #53 errors. Then, I went to correct the #51 errors. To my initial surprise, I saw some of the same errors that I had corrected for #53 listed under #51 errors.

On further thought, I question why the Checkwiki software scans for two different "misplaced interwiki" errors. If an interwiki link does not appear after the last category, it's an error. No need to distinguish between "before last heading" and "before last category".

Knife-in-the-drawer (talk) 04:09, 26 March 2015 (UTC)[reply]

Simple, see {{Uncategorized}} Jerodlycett (talk) 13:32, 30 March 2015 (UTC)[reply]

I saw. What is the relation of {{Uncategorized}} to my issue of two different Checkwiki errors flagging basically the same condition?

Let me add an additional dimension to the problem, and maybe you'll see the issue. Not only do #51 and #53 flag the same basic error, but they also fail to flag an error when an interwiki occurs after all categories but before any stubs. See this older and more detailed version of Wikipedia:Manual of Style/Layout. Why was "Interlanguage links" removed from the MOS? Because "Interlanguage links" should be placed in Wikidata now. However, they still occur in many articles--thus, Checkwiki errors #51 and #53. Knife-in-the-drawer (talk) 02:11, 3 April 2015 (UTC)[reply]

Sorry, should have explained better. #51 gets triggered only if it's before the last heading. #53 only gets triggered if it's before the last category. There are several ways for #53 to occur without #51, if there are no headings (probably a stub) or if it's after the last heading, but before the last category. There are several ways to trigger #51 before #53, if the categories are misplaced (and put before the interwiki which is also misplaced, maybe a References section was added at the tail for example) or if the article is uncategorized (hence the reference to the template for that). If you have a situation like {{example]] you also get two errors (and it's disturbingly common), one for the template without a correct end, one for the square brackets without correct beginning. Correct one and the other is not updated. If you fix a bunch of errors in one article, you'll see it's still listed under all the other errors, which seems to me to be a missing feature in WPCleaner. Jerodlycett (talk) 03:12, 3 April 2015 (UTC)[reply]

Good--now, we're on the same page. Let me make a last pitch for "only one Checkwiki error is necessary"; then, we'll call the horse "dead" and stop beating it.

If interwiki at end of article

Then no error

Else

If only stubs after interwiki

Then no error

Else Checkwiki error #onlyoneerror

Knife-in-the-drawer (talk) 04:23, 3 April 2015 (UTC)[reply]

I counter with if(interwiki) Error: Should be on meta Also with magic missle. That's also a good way to counter. Jerodlycett (talk) 04:31, 3 April 2015 (UTC)[reply]

Some editors and ips needlessly add interwiki links where Wikidata has already established a link e.g. [1]. It would seem that the #51 and #53 errors would be better directed towards the removal rather than relocation of interwikis in those cases. Dl2000 (talk) 03:05, 8 April 2015 (UTC)[reply]

I agree with both of you. It's what I actually wanted from the beginning, but I didn't want to ask for too much initially, so I opted for the two-step approach. :) Knife-in-the-drawer (talk) 23:38, 9 April 2015 (UTC)[reply]

@Dl2000, Jerodlycett, and Knife-in-the-drawer: three things:

1. Some languages have one error turned off while the other is turned on. Different languages have different rules. Checkwiki was designed so each language can customize what errors to turn off and on.
2. I personally remove any interwiki links that are on wikidata. A lot of the time, it is to the wrong item on Wikidata and I fix that. Also, bots do go around and delete valid interwiki links. There are two helpful scripts that one can use:
   1. To add a link to Wikidata under the article's name, add to your following .js file: importScriptURI("//www.wikidata.org/w/index.php?title=User:Yair rand/WikidataInfo.js&action=raw&ctype=text/javascript");
   2. To add a script that will automatically delete any interwiki links already in wikidata, add to your following .js file: mw.loader.load('//www.wikidata.org/w/index.php?title=User:Yair_rand/checksitelinks.js&action=raw&ctype=text/javascript');
3. Around 1/2 the time, the "interwiki link" isn't one. It is a wikilink, just formatted wrong. The most common case is a missing ":" at the beginning. For example [[de:Germany]] should be [[:de:Germany]]

One can look at my User:Bgwhite/vector.js file to see how I added the above scripts. The scripts are at the bottom of the file. Bgwhite (talk) 23:03, 9 April 2015 (UTC)[reply]

I think our suggestion would still be valid. It should still be removed, not moved. Whether that's through adding a colon, or just removing it. Jerod Lycett (talk) 17:54, 10 April 2015 (UTC)[reply]

Hello. Could you please expand the id 34 by tags (open, full or closed) noinclude, includeonly and onlyinclude? Thank you. IKhitron (talk) 14:35, 15 November 2015 (UTC)[reply]

IKhitron On several wikis, these are used in articles when it comes to using a part of one article into another. So, they are allowed in articles. Any particular reason you need them scanned on hewiki? Bgwhite (talk) 06:08, 16 November 2015 (UTC)[reply]

Wierd, Bgwhite. The reason is they should not be there and destroy the article structure. Is there a possibility to add the as new ID and set default prio 0 so wiki that wants it will change it? Thank you, IKhitron (talk) 11:52, 16 November 2015 (UTC)[reply]

Bgwhite? IKhitron (talk) 13:24, 20 November 2015 (UTC)[reply]

IKhitron Thanks for pinging me again. I see a ping, but don't have time at that moment and then end up forgetting. It is possible to do a separate, new error. NicoV, you see other wiki's, what do you think? Bgwhite (talk) 18:38, 20 November 2015 (UTC)[reply]

Yes, why not. I'm not sure we will use on frwiki because I'm pretty sure some articles are using it, but I will maybe activate it only as a detection. --NicoV ^{(Talk on frwiki)} 13:39, 21 November 2015 (UTC)[reply]

It will be very nice, thank you. IKhitron (talk) 14:39, 22 November 2015 (UTC)[reply]

@Bgwhite: A user asked me to add a way of ignoring some texts for #103 in link labels : for example, the second division of Spain soccer championship is now named LaLiga 1|2|3, so it may often appear in link labels. What about adding a parameter for listing such normal texts containing a pipe, like the parameter error_067_abbreviations that exists for #67 ? I can easily add it to WPC, but it would become very useful only if CW only takes it into account. --NicoV ^{(Talk on frwiki)} 16:39, 21 August 2016 (UTC)[reply]

NicoV Segunda División does not register a #103 error. There are no {{!}} in the article and it would also have to show up between square brackets. If they were to use LaLiga 1|2|3 in a wikilink, then the article needs to be whitelisted. This would be rare as 99% of the articles just use "Segunda División" or "La Liga 2" The name changes every few years in La Liga 1 and La Liga 2 that its very rare to see the sponsorship name in a wikilink. It would actually be very dumb to use the sponsorship name because of the pain in changing roughly 5,000 articles every few years. Bgwhite (talk) 05:04, 22 August 2016 (UTC)[reply]

Bgwhite, maybe you would consider changing regex a little bit for #3? At lvwiki, equivalent for {{reflist}} is {{atsauces}}, for {{unreferenced}} - {{atsauces+}}. So if article has maintaince tag {{atsauces+}} and <ref>s, the article will be skipped in Checkwiki (as script checks only for {{atsauces). --Edgars2007 (talk/contribs) 15:29, 8 July 2016 (UTC)[reply]

Edgars2007 What's happening is in the translation file, under #03, you have atsauces listed. In the code, I take that and look for cases of "{{atsauces". There's nothing at the end of {{atsauces because there maybe parameters in the template. The code looking for "{{atsauces" is a regex. So, in theory, adding atsauces[^\+] to the template file should solve the problem. Could you add it to the template file and give me some articles/sandbox to check for a valid #3 error and some with atsauces+. I'll test and make sure it works. Bgwhite (talk) 21:51, 22 July 2016 (UTC)[reply]

I assume, they aren't case sensitive, right? Do you want me to give examples, which fits to your requirements now or 2016-07-01 (dump day)? --Edgars2007 (talk/contribs) 21:59, 22 July 2016 (UTC)[reply]

Edgars2007 Yes, they are not case sensitive. If you have any examples it would be good. I'd like to make sure everything is ok before the dump. I looked at today's (25th) results. How in the world did lv:Count Your Blessings get tagged as a #3? Ran the code manually and it didn't see anything wrong. Bgwhite (talk) 22:20, 26 July 2016 (UTC)[reply]

Bgwhite for example lv:Dalībnieks:Edgars2007/Reflist/1 should get tagged as #3, and this shouldn't. If you need some other examples to work with, you can simply create another sandbox page at lvwiki at my userspace. 2016-07-20 version of Count Your Blessings didn't have references template. --Edgars2007 (talk/contribs) 06:09, 27 July 2016 (UTC)[reply]

Edgars2007 Things look good on the test articles. Should be good to go. Bgwhite (talk) 19:48, 27 July 2016 (UTC)[reply]

 Resolved

After the update from 2016-09-01 there are 0 errors counted on alswiki. The update before had found hundreds of errors. Does anybody know why many projects are listed with zero errors now? --Holder (talk) 19:12, 7 September 2016 (UTC)[reply]

Holder Some of the dump files were messed up. Another dump run should be available around the 16th. Bgwhite (talk) 05:13, 8 September 2016 (UTC)[reply]

Thanks! --Holder (talk) 06:04, 8 September 2016 (UTC)[reply]

 Resolved

The problem is in dump or in our translation file? I decided to write the vast majority of magic words in error_034_templates_ruwiki and then disable some of them step-by-step if some of them will lead to false positives - I hope that's not the reason of failed scan. Can I be sure that the next dump will be scanned successfully? Facenapalm (talk) 12:52, 16 September 2016 (UTC)[reply]

Did you see the #Update September 2016? It can be the same problem. IKhitron (talk) 14:53, 16 September 2016 (UTC)[reply]

Yes I did. I just want to be sure that the problem doesn't depend on me. If so, I can wait, that's not a problem. Facenapalm (talk) 15:09, 16 September 2016 (UTC)[reply]

Facenapalm Looking at the dumps, it appears they are scheduled to start up again on the 20th. In theory, messing up the translation file for one error won't stop the program from running. At worst, it should only cause false positives or less errors for that particular error #.

You do have #34 wrong for the translation file. These are already in the program and no need to list again:

(#if:|#ifeq:|#switch:|#ifexist:|{{fullpagename}}|{{sitename}}|{{namespace}}|{{basepagename}}|{{pagename}}|{{subpagename}}|{{namespacenumber}}|{{talkpagename}}|{{fullpagenamee}}|__noindex__|__index__|__forcetoc__|__nonewsectionlink__|{{subst:)

Then add the rest one item per line. Program already adds | between each item. Bgwhite (talk) 23:35, 16 September 2016 (UTC)[reply]

I do not understand how exactly the subst can be a part of wiki code. It should be, you know, subst, shouldn't it? IKhitron (talk) 00:01, 17 September 2016 (UTC)[reply]

Like phab:T4700? Matěj Suchánek (talk) 08:47, 17 September 2016 (UTC)[reply]

First of all, subst: doesn't work with unexisting templates. See: {{subst:verylongtemplatenamethereisnosuchtemplate}}. Once we decided to delete template called "ab", so we should replace "{{ab|text}}" by "text". Botmaster X created template "User:X/0" with such code: {{{1|}}} and just replaced "{{ab" by "{{subst:User:X/0". He doesn't noticed that he broke template "{{abbr" as well, and we got a lot of "{{subst:User:X/0br" in articles. Checkwiki helps me to detect it. The other example is MediaWiki bugs: "subst:" not always works at least in refs. Facenapalm (talk) 12:07, 17 September 2016 (UTC)[reply]

I see. thank you. IKhitron (talk) 12:51, 17 September 2016 (UTC)[reply]

Understood, thanks. Well, one item "a|b" should works just as two items "a" and "b", so it's not an error, is it? Anyway, list of items are easier to understand. Facenapalm (talk) 12:10, 17 September 2016 (UTC)[reply]

 Resolved

There is a false positve on Telephone numbers in Indonesia due to their use of <prefix>. Not saying it's a good thing, but it's not the pre tag. Jerod Lycett (talk) 22:01, 18 September 2016 (UTC)[reply]

Jerod Lycett It's because CheckWiki only scans for <pre. There can be parameters in <pre> and people do forget the space. Two options:

1. Fix it somehow... remove the < > or some other way.
2. Add it to the whitelist. The list for #24 is located here.

Bgwhite (talk) 04:30, 19 September 2016 (UTC)[reply]

 Done

Consistent Overhead Byte Stuffing has a false positive for #15. Jerod Lycett (talk) 01:29, 19 September 2016 (UTC)[reply]

Jerodlycett I messed up the ping from your other message, so also look above.

This is a no win situation. In Checkwiki's code, I have it looking for <code> and then looking for <source>. Program runs, finds a broken code tag, then goes into the <source> section and blanks everything between <source>...</source>. If I had it the other way around, program wouldn't find a borked <code> because it was deleted.

Somethings have to run in a certain order... blanking what's between comment tags goes first. Blanking what goes between <nowiki> is second. I'll put the code section after source and see what happens... No more false positives, keep it where it is at. If more false positives show up, move the tag section back and white list Overhead Byte Stuffing. Bgwhite (talk) 04:48, 19 September 2016 (UTC)[reply]

But @Bgwhite:, you promissed to consider to replace such tags with some neutral string as "qwerty" instead of blanking. IKhitron (talk) 06:51, 19 September 2016 (UTC)[reply]

IKhitron You are correct. I have horrible memory, so always feel free to remind me about something. I added "CheckWiki" between the tags to make it obvious that CheckWiki added it. I've also added <pre> and <code> tags to #85. If all goes well will add more to #85. Bgwhite (talk) 22:48, 19 September 2016 (UTC)[reply]

Thanks! IKhitron (talk) 23:06, 19 September 2016 (UTC)[reply]

IKhitron I backed out adding "CheckWiki" between comment tags. It was causing false positives elsewhere. All others look good. Bgwhite (talk) 04:28, 20 September 2016 (UTC)[reply]

 Done

Hi. What do you think about [[abc|"abc"]], [[abc|'abc']] and so on? IKhitron (talk) 12:23, 19 October 2016 (UTC)[reply]

IKhitron I like easy request. Bgwhite (talk) 15:13, 19 October 2016 (UTC)[reply]

Thank you, Bgwhite! Hope you did not forgot «», ``, `', '`, „", "“, “" and „“. IKhitron (talk) 15:17, 19 October 2016 (UTC)[reply]

IKhitron Ok, now you are evil. :) I've added «», `` and „“. I haven't seen the others before. Are the same as others, but with one "curly" and one "straight" mark? Bgwhite (talk) 15:51, 19 October 2016 (UTC)[reply]

Yes, all kinds of options. I saw that in many science articles. What do you think about [«'`‘"„“”].*[»'`‘"„“”] (I added some more), Bgwhite IKhitron (talk) 15:57, 19 October 2016 (UTC)[reply]

Done for WPC. --NicoV ^{(Talk on frwiki)} 17:19, 19 October 2016 (UTC)[reply]

Good, but I can't use it. IKhitron (talk) 17:24, 19 October 2016 (UTC)[reply]

IKhitron Nico is running the latest WPC against the enwiki dump and I think frwiki. Will get a good idea of how many errors the newly added marks give and see what is needed to be added or not added. Nico can run one for hewiki. If he has updated the download, I can also run it. Bgwhite (talk) 20:26, 19 October 2016 (UTC)[reply]

Thank you both. Btw, if all you need is to know how many events, there is some linux command line command that returns the regular expressions search results statistics for all wiki sites, you do not need to run cw for this. IKhitron (talk) 20:33, 19 October 2016 (UTC)[reply]

Dump analysis for frwiki will be available here very soon, and for enwiki here probably in a few hours. --NicoV ^{(Talk on frwiki)} 21:39, 19 October 2016 (UTC)[reply]

Both lists have been generated by WPC. IKhitron, do they seem ok ? Bgwhite, in addition to the new quotes, it seems that WPC also detected cases that should have been picked up by CW previously both on enwiki (11th Air Army: [[64th Fighter Aviation Corps|64th Fighter Aviation Corps]], 2004–05 Michigan State Spartans men's basketball team: [[Big Ten - ACC Challenge|Big Ten - ACC Challenge]]...) and frwiki (19-2: [[Éric Hoziel|Éric Hoziel]], Alains: [[Iasses|Iasses]]...). I see a few false positives in WPC list, like Abbaye de Saint-Nicolas: [[Sainte-Anastasie (Gard)#Histoire|Sainte-Anastasie (Gard).]] --NicoV ^{(Talk on frwiki)} 06:19, 20 October 2016 (UTC)[reply]

Yes, it looks very nice, NicoV. It's even worth that I expected. Well, Bgwhite you can see much more apostroph kinds, and even unpaired. IKhitron (talk) 11:52, 20 October 2016 (UTC)[reply]

IKhitron, Bgwhite: I'm currently trying to fix all of them on frwiki, and I noticed some other cases with parenthesis, so I'm adding the detection of "(" before and ")" after to #64. --NicoV ^{(Talk on frwiki)} 07:44, 2 November 2016 (UTC)[reply]

IKhitron, Bgwhite: I've also added the detection of colon or semi-colon after and generated the list again at Wikipedia:CHECKWIKI/WPC 064 dump‎. --NicoV ^{(Talk on frwiki)} 09:10, 8 November 2016 (UTC)[reply]

Good idea, NicoV, thanks. You can try also [–־—]\s*(.+) and (.+)\s*[–־—]. IKhitron (talk) 11:34, 8 November 2016 (UTC)[reply]

IKhitron Major part came online last week. Everything is now up and running including the latest characters NicoV found. Bgwhite (talk) 19:55, 9 November 2016 (UTC)[reply]

Thank you! IKhitron (talk) 12:20, 10 November 2016 (UTC)[reply]

 Done

CHECKWIKI seems not to recognise a #9 error when the keyword "Category" is in lowercase. Examples: [2], [3]. -- Magioladitis (talk) 08:34, 20 October 2016 (UTC)[reply]

Magioladitis Done Bgwhite (talk) 21:40, 2 November 2016 (UTC)[reply]

 Done

WPCleaner detects categories that only differ in underscore while CHECKWIKI does not. Example. AWB removes the duplicated category. -- Magioladitis (talk) 08:49, 20 October 2016 (UTC)[reply]

Magioladitis Fixed Bgwhite (talk) 21:33, 2 November 2016 (UTC)[reply]

 Done

Currently, CHECKWIKI does not detect these: [4], [5]. -- Magioladitis (talk) 10:25, 20 October 2016 (UTC)[reply]

Magioladitis Done Bgwhite (talk) 21:28, 2 November 2016 (UTC)[reply]

 Resolved

@NicoV, Meno25, Edgars2007, Facenapalm, Josve05a, Matěj Suchánek, Magioladitis, and IKhitron:
WMF is taking labs down on November 2nd. Downtime is scheduled for 48 hours. They are moving to a new NFS (file system) redundant setup. The long 48-hour downtime is because of the time that may be needed to transfer files and any problems that crop up. Bgwhite (talk) 23:37, 21 October 2016 (UTC)[reply]

• Thanks, I wrote a warning. Facenapalm (talk) 23:50, 21 October 2016 (UTC)[reply]

Thanks for notification. --Edgars2007 (talk/contribs) 06:22, 22 October 2016 (UTC)[reply]

Thank you very much for this notification. How is it possible that I haven't yet heard of this takedown? Matěj Suchánek (talk) 07:42, 22 October 2016 (UTC)[reply]

@NicoV, Meno25, Edgars2007, Facenapalm, Josve05a, Matěj Suchánek, Magioladitis, and IKhitron:
They are postponing the down time till November 14.
I'm on the labs-announce mailing list, which is how I was notified. I haven't seen it posted anywhere else. Bgwhite (talk) 20:43, 1 November 2016 (UTC)[reply]

Thanks (again). Yeah, it's weird, that this hasn't been announced anywhere, even on VPT. --Edgars2007 (talk/contribs) 01:20, 2 November 2016 (UTC)[reply]

Edgars2007 Concerning VPT, It's now done. Wikipedia:Village pump (technical)#Planned labs outage. Bgwhite (talk) 05:46, 2 November 2016 (UTC)[reply]

 Resolved

Would be good to catch articles, which have ref tag after last {{reflist}} template/references tag. --Edgars2007 (talk/contribs) 07:24, 29 October 2016 (UTC)[reply]

Edgars2007 WPC already does this, see for example fr:Projet:Correction syntaxique/Analyse 003. --NicoV ^{(Talk on frwiki)} 09:50, 3 November 2016 (UTC)[reply]

Thanks. So lvwiki is clean in this area... Hmm, weird, but nice :) --Edgars2007 (talk/contribs) 11:15, 3 November 2016 (UTC)[reply]

Edgars2007 This will be listed as error #110. I should have time this week to actually code. Bgwhite (talk) 21:40, 8 November 2016 (UTC)[reply]

Edgars2007 Done, for both CW and WPC, as error #111. --NicoV ^{(Talk on frwiki)} 11:57, 16 November 2016 (UTC)[reply]

 Done Bgwhite and NicoV please add U+FFFC to the list. -- Magioladitis (talk) 07:57, 29 October 2016 (UTC)[reply]

 Done Done for WPC. --NicoV ^{(Talk on frwiki)} 11:32, 1 November 2016 (UTC)[reply]

 Done

Hello Bgwhite, I found that the script checks for HTML entity &plusm; which doesn't exist. The correct one is &plusmn; (±). And when you are around, could you also add &acute; (´), &times; (×), &sbquo; (‚), &prime; (′), &Prime; (″) and &minus; (−)? Thank you for everything. Matěj Suchánek (talk) 17:47, 30 October 2016 (UTC)[reply]

Matěj Suchánek Done Bgwhite (talk) 21:20, 2 November 2016 (UTC)[reply]

 Done

Starting from yesterday checkwiki adds many articles on .de containing <div style="clear:xxx;"></div>. Why? It cannot be written in unary form and it must be empty. --Hadibe (talk) 04:45, 10 November 2016 (UTC)[reply]

Hadibe Change was made to find more <div ...></div> (and <span>) type errors. I've excluded <div id="foo"></div> as this is valid to use. On enwiki, there's {{clear}} that replaces <div style="clear...> and it should always be used over any div tags. It appears this is also true for dewiki... de:Vorlage:Absatz Diese Vorlagen bitte nicht substituieren.

@NicoV and Magioladitis: Any thoughts? I see the same code showing up on frwiki. I also see alot of<span class="cx-segment" data-segmentid="77"></span>, but I think that's content transcrapulator's doing. Bgwhite (talk) 05:26, 10 November 2016 (UTC)[reply]

@Bgwhite and Magioladitis: For CX, sure, we should definitely detect it and report it (this tool is a shame, put into production more than a year ago and still unable to remove its own garbage from the result...). For the div tag, I think we can also exclude special cases with style="clear:xxx" because it's valid HTML (even if many wikis prefer to use a {{clear}} template to replace it. Currently, in WPC, I'm reporting empty tags even if they have an attribute except some cases (name in ref for the moment), but I should add a few other cases. --NicoV ^{(Talk on frwiki)} 09:27, 10 November 2016 (UTC)[reply]

@NicoV and Hadibe: It should not pick up any <div style="clear:xxx;"></div> in the next run, which starts in about ~90 minutes. Bgwhite (talk) 22:26, 10 November 2016 (UTC)[reply]

It didn't. Thanks for your help. --Hadibe (talk) 05:01, 11 November 2016 (UTC)[reply]

 Resolved

(pinging Magioladitis) #85 seems to want to delete all of the line colour <div></div> tags in the Tube map article; is this a bug or deliberate? Jc86035 (talk) Use {{re|Jc86035}}
to reply to me 11:43, 10 November 2016 (UTC)[reply]

@NicoV: to check the script. @Frietjes: to check the page. -- Magioladitis (talk) 18:09, 10 November 2016 (UTC)[reply]

@Jc86035, Frietjes, NicoV, and Magioladitis: Yes, #85 picked up Tube Map. This is also the first day since #85 was "updated" with more checks. I'm not sure what one should be used, <div> or <hr>. The examples and their results:

<div style="border-style: solid; border-width: 6px 0px 0px 0px; border-color:#{{LUL colour|Bakerloo}};">

 

<hr style="height:6px; background-color:#{{LUL colour|Bakerloo}};">

---

If <hr> is not the better option, then we can whitelist the article. Personally, from just looking at it, the <hr> option seems easier to understand, especially for novices. Bgwhite (talk) 19:26, 10 November 2016 (UTC)[reply]

not all the div tags are closed on that page. Frietjes (talk) 19:37, 10 November 2016 (UTC)[reply]

now templated to simplify the text, and avoid problems with unclosed div tags. Frietjes (talk) 20:48, 10 November 2016 (UTC)[reply]

Thank you Frietjes. Bgwhite (talk) 21:48, 10 November 2016 (UTC)[reply]

 Resolved

#48 is now first letter case-insensitive and also will catch cases where target and link differ only in underscores. -- Magioladitis (talk) 07:33, 29 November 2016 (UTC)[reply]

Same for WPC (was for a long time). --NicoV ^{(Talk on frwiki)} 09:18, 29 November 2016 (UTC)[reply]

 Resolved

CW has been updated to catch [[ISBN ...]] and ISBN inside nowiki tags. -- Magioladitis (talk) 06:45, 29 November 2016 (UTC)[reply]

Same for WPC. --NicoV ^{(Talk on frwiki)} 09:17, 29 November 2016 (UTC)[reply]

 Resolved

Hi. There is something that can take you a couple of days to implement, but it will be very useful for many people and wikis. Could you please:

1. Create a new function, nulledit.
2. It will run on the list, open each page and force saving without changing anything.
3. It will not wait for each edit, as it was a bot run.
4. It will work for any registered user, not just for those with AWB rights.
5. It will skip automatically pages that this user has no rights to edit.

Please take a year, if you need, to think about this in positive direction. Thank you very much in advance. IKhitron (talk) 14:21, 22 August 2016 (UTC)[reply]

It shouldn't be included into Check Wikipedia project: Check Wikipedia finds the errors, that's all. If you want to do nulledits, you can download python, install pywikibot framework and write something like that:

import pywikibot

def main():
    site = pywikibot.Site()
    for line in open("pages.txt", encoding="utf-8").readlines():
        page = pywikibot.Page(site, line)
        page.save("Null edit.")

if __name__ == "__main__":
    main()

This code will make nulledits in every page from pages.txt file. Facenapalm (talk) 14:59, 22 August 2016 (UTC)[reply]

There is page.touch(), btw. And shouldn't open("pages.txt", encoding="utf-8").readlines() be outside of the loop? --Edgars2007 (talk/contribs) 15:19, 22 August 2016 (UTC)[reply]

Never heard about touch() before, thanks. No sense to open file outside the loop, it will be processed only once, I've checked that. Anyway, it's just an example (despite the fact that I used this code before). It's good idea to catch exceptions too, for example. Facenapalm (talk) 15:29, 22 August 2016 (UTC)[reply]

I have this bot, of course. I ask for AWB to use it's make list mechanism. And you can't say it does not fix errors, it fixes a lot of errors during this run, especially of "forgotten to insert to category". IKhitron (talk) 15:33, 22 August 2016 (UTC)[reply]

It's a good idea to at least once go trough page.py file and maybe others, too (generators and site.py) :) --Edgars2007 (talk/contribs) 18:58, 22 August 2016 (UTC)[reply]

Don't use .readlines(); file objects are iterable in python and using the readlines method forces all of the data to be loaded into memory at once. Pppery 20:52, 14 October 2016 (UTC)[reply]

 Resolved

Hi. It ran today, a long time before the new dumps were completed. It's a bug? IKhitron (talk) 00:29, 6 August 2016 (UTC)[reply]

IKhitron Before, WMFLabs wouldn't transfer all the dump files over until all the files were completed. Now, when a file is completed, it is transferred over. The file CheckWiki uses is towards the beginning of the dump cycle. Here's the latest dump progress of hewiki. The file CheckWiki uses is "pages-articles". Bgwhite (talk) 07:40, 6 August 2016 (UTC)[reply]

I see: you have protection.:-) Thank you, IKhitron (talk) 14:39, 6 August 2016 (UTC)[reply]

 Resolved

Among the crap regularly produced by CX, the content-translation tool, I often see <ref>...</ref> tags containing a <cite>...</cite> tag (fix at the end of the diff): it seems totally useless and redundant, it's not in the original article, it often contains useless attributes (contenteditable..., class with a CSS class from the original wiki)... @Bgwhite and Magioladitis: and others: should we detect this as an other error number? --NicoV ^{(Talk on frwiki)} 17:45, 17 November 2016 (UTC)[reply]

NicoV <cite> is only to be used for the title of a work. In bare refs like your example had, <cite> could be valid if it were just around the title of the book or journal article. However, the example has the entire ref is inside the cite tags. There are no cases of this on enwiki, but there are 330 on frwiki. Your example was from a year ago, but fr:Bear Bergman was created two months ago and it has the problem. dewiki has 8 cases and eswiki has 640. Why the descrepincy between enwiki and dewiki vs frwiki and eswiki? Whatamidoing (WMF), do you have any answers? Bgwhite (talk) 20:50, 17 November 2016 (UTC)[reply]

The discrepancy is probably because enwiki and dewiki have been more strict against CX usage. This tool is a nightmare with developers simply ignoring the problems it creates. --NicoV ^{(Talk on frwiki)} 12:01, 18 November 2016 (UTC)[reply]

It looks like the translation has lost most of the content from the citation template. It should say something similar to <ref name=Nelson2001>{{cite book|last=Nelson|first=Charles A.|title=Handbook of Developmental Cognitive Neuroscience|year=2001|publisher=The MIT Press|isbn=978-0262140737|edition=2|author2=Monica Luciana}}</ref>, but only <ref name="Nelson2001">Nelson, Charles A.; Monica Luciana (2001).</ref> survives. I wonder if this will be fixed when the template system is deployed (unknown date). Whatamidoing (WMF) (talk) 09:24, 22 November 2016 (UTC)[reply]

Yes, this will be fixed with the template system. But it has some other bugs, so unless this is so widespread that you'd rather have these bugs instead of those bugs (and the incomplete citations, which make me twitch), then it may be a while. Whatamidoing (WMF) (talk) 18:15, 22 November 2016 (UTC)[reply]

CW and WPC now detect <ref><cite> as part of #2. --NicoV ^{(Talk on frwiki)} 09:17, 29 November 2016 (UTC)[reply]

With the addition of this detection in #2, I ran a dump analysis on frwiki: more than 500 articles still damaged for this single problem by CX... Whatamidoing (WMF) When will WMF do really something to fix this broken tool ? --NicoV ^{(Talk on frwiki)} 13:50, 29 November 2016 (UTC)[reply]

It's on the train for this week (so approximately now, if you're at mediawiki.org, but Thursday for the Wikipedias). Expect new bugs. Whatamidoing (WMF) (talk) 19:46, 29 November 2016 (UTC)[reply]

Whatamidoing (WMF) I was not speaking just about this problem, but the many problems that CX produces on articles. It has been released for about 1 year 1/2, there are random popups advertising it when you edit wp, but it's basically still unable to produce clean wikitext... So, my question was more general. --NicoV ^{(Talk on frwiki)} 17:43, 1 December 2016 (UTC)[reply]

 Resolved

At list 91 it shows some articles which doesn't have any Interwiki link written as an external link. in my pinion it should check if http://en.wikipedia.org/wiki/.bm%7Cعنوان is inside {{}} so it isn't internal linkYamaha5 (talk) 20:22, 28 June 2016 (UTC)[reply]

Also list 90 has the same False positive bugYamaha5 (talk) 20:24, 28 June 2016 (UTC)[reply]

Yamaha5 I'm confused. You said "Interwiki link" and #90 and #91 doesn't check for interwiki links. #90 and #91 should be checking for external links and references. I haven't seen a problem with enwiki on the ~10,000 articles I've fixed except for when a reference has a Wikipedia link and the article is about something on Wikipedia. Could you give a better example. Bgwhite (talk) 20:36, 28 June 2016 (UTC)[reply]

please check list 91 it's title is Interwiki link written as an external link.

for example: at the first row page fa:.bm has {{یادکرد-ویکی|پیوند = http://en.wikipedia.org/wiki/.bm|عنوان = .bm|زبان = انگلیسی|بازیابی =30 آپریل ۲۰12}} which is in english equal to {{citation|link= http://en.wikipedia.org/wiki/.bm|title= .bm|lang= en|date =30 April 2012}}.

Now checkwiki assumes http://en.wikipedia.org/wiki/.bm|عنوان as Interwiki link written as an external link which is not correct it is inside a template and it doesn't have any [[]]. please check articles that are listed on list 91 and list 90 you see the False positive bugYamaha5 (talk) 00:04, 29 June 2016 (UTC)[reply]

Yamaha5 If I'm understanding correctly.... This is still an error. One can't use Wikipedia as a reference. 90/91 checks references that used Wikipedia as a source. The 90/91 definition didn't reflect that and that is my fault. I've updated the English translation file. Bgwhite (talk) 00:30, 29 June 2016 (UTC)[reply]

Thanks. would you please update fawiki's data? I cleaned more than 30k pages and I wait for new updateYamaha5 (talk) 00:37, 29 June 2016 (UTC)[reply]

Yamaha5 The new dumps start up on the 1st. If all goes well, checkwiki will update fawiki around the 4th. Bgwhite (talk) 18:26, 29 June 2016 (UTC)[reply]

 Resolved

Hello, Bgwhite. Could you, please, explain me what is the problem in the first article in [6]? It's there a couple of months. I can't understand the second one two, but it's less important because it's new. Thank you in advance, IKhitron (talk) 18:46, 9 June 2016 (UTC)[reply]

IKhitron Most of the time, #60 will give an error if there are unbalanced brackets. It will also give an error if there is a bad parameter in a template. |date*= or |date:= would be errors as * or : can't be part of a parameter's name. I can't tell if that is an issue in the Eviatar Nevo article because of your wacky right-left, funny character language. :) Bgwhite (talk) 22:02, 9 June 2016 (UTC)[reply]

OK, Bgwhite, here is a kind of translation.

The text in the article:

{{templatename|templatetext}}

The text in Notice column:

{{templatename,templatetext}}

And this is the same every month. I have no idea where checkwiki finds this comma, and why does it loose the pipe. Thank you, IKhitron (talk) 22:30, 9 June 2016 (UTC)[reply]

IKhitron I only looked at the first article. I'll look at the second one too. Bgwhite (talk) 04:55, 10 June 2016 (UTC)[reply]

But this is in the first one, Bgwhite. IKhitron (talk) 08:51, 10 June 2016 (UTC)[reply]

@Bgwhite: IKhitron (talk) 12:09, 14 June 2016 (UTC)[reply]

 Resolved

The fix for this error seems to break AfD tag syntax - it moves the {{Article for deletion}} template outside of the comments it comes with when it replaces/alphabetises the tags at the top of the page. This in turn breaks the commonly used AfD closing script User talk:Mr.Z-man/closeAFD.

For an example see: [7].

Pinging Bgwhite and Magioladitis, who run BG19bot and Yobot, which run these fixes. ansh666 05:08, 26 May 2016 (UTC)[reply]

Ansh666 This is actually an AWB question. That is the tool causing all the fun. CheckWiki only finds problems, not fixes them. I swear this was brought up recently but I can't find it at any of Magioladitis' talk pages. He'll have an answer. Bgwhite (talk) 05:29, 26 May 2016 (UTC)[reply]

Ansh666 Magioladitis is on his phone and not at his laptop. He asked me to type up the issue in phabricator. Bgwhite (talk) 06:43, 26 May 2016 (UTC)[reply]

I 'll be back at my laptop in 4 days. Thanks for reporting it to Phabricator. -- Magioladitis (talk) 16:24, 26 May 2016 (UTC)[reply]

Alright, thanks guys. ansh666 17:19, 26 May 2016 (UTC)[reply]

 Resolved

On sv.wp CHECKWIKI detects the following false positives:

• On sv:Pox Special: ISBN-märkning
• On sv:Kaliumvätetartrat: ISBN-nummer

(t) Josve05a (c) 13:13, 28 June 2015 (UTC)[reply]

Josve05a can't you just use {{Not a typo}}? -- Magioladitis (talk) 13:16, 28 June 2015 (UTC)[reply]

Will that help against CHECKWIKI errors? Also this is perhaps a common issue on svwp since we do tend to use dashes to merge words together. (t) Josve05a (c) 13:30, 28 June 2015 (UTC)[reply]

Josve05a was this problem fixed/handled somehow? -- Magioladitis (talk) 11:38, 10 March 2016 (UTC)[reply]

ISBN were replaced with [[ISBN]]. Matěj Suchánek (talk) 19:35, 1 December 2016 (UTC)[reply]

 Resolved

Hi. Could you, please, insert the "Notice" column in checkwiki project pages to <span style="unicode-bidi:bidi-override;">...</span>? It will not affect most of wikipedias, but for rtl ones the text with rtl and ltr parts will become readable. Thank you. IKhitron (talk) 17:10, 25 November 2016 (UTC)[reply]

@IKhitron: Done, I think. Tell me if it is working ok. Bgwhite (talk) 22:22, 1 December 2016 (UTC)[reply]

Thank you very very very much, Bgwhite! Works great. IKhitron (talk) 11:20, 2 December 2016 (UTC)[reply]

Hello !
On french wiki, last update has been done on March 26. Usually there is a daily update, but none on March 27 and 28, means that errors cannot be corrected because not reported.
Could somebody check that ?
Thanks. Croquemort Nestor (talk) 05:22, 28 March 2021 (UTC)[reply]
PS : (I copy @NicoV:)

@Croquemort Nestor: The bot was hung-up since March 25th. It is now running again and will update errors tonight. All errors from March 25th until I restarted it were lost. The lost errors will get found during the next dump scan after April 1st, except for enwiki and dewiki. --Bamyers99 (talk) 15:00, 28 March 2021 (UTC)[reply]

thanks ! --Croquemort Nestor (talk) 15:19, 28 March 2021 (UTC)[reply]

Check Wikipedia found a headline with capital letters (error 058) in the article pl:Pralki Polar even though the article is on the whitelist. The same happened with the article pl:Funkcja anonimowa on error 056 (whitelist). Bamyers99, could you check it? --The Polish (talk) 06:57, 28 April 2021 (UTC)[reply]

Hi The Polish. It's a problem in the configuration page on plwiki : error_058_whitelist_plwiki=Wikiprojekt:Check_Wikipedia/Whitelist_058 is incorrect, you should use error_058_whitelistpage_plwiki=Wikiprojekt:Check_Wikipedia/Whitelist_058 (added "page"), whitelist is for giving directly the list, not a page containing the list. --NicoV ^{(Talk on frwiki)} 08:02, 28 April 2021 (UTC)[reply]

Check Wikipedia found on May 2 a bug in articles pl:Zdzisław Czeszejko-Sochacki (ID 80) and pl:Piotr Ożga (zm. 1671) (ID 94). I corrected the mistakes: here are two editions (here and here). I did not mark it as done because all edits from pl.wiki are checked every day at midnight. There are more such problems, almost this whole category. It seems that the edits of the day od dump scanning (or maybe the day after) are not checked and these articles will not be removed from the list until the next dump. Is it possible to fix this error? The Polish (talk) 14:21, 4 May 2021 (UTC)[reply]

@The Polish: The problem is that any fixes made between the 1st of the month and the dump scan run (May 3 for plwiki) are not seen because the dump scan completely replaces the error reports with the data asof the 1st of the month. This is why the enwiki and dewiki dumps are not scanned because the dumps are not ready until several days after the 1st of the month. The reason the dump scans are needed even for the daily scanned wikis is that sometimes the daily scans miss article changes because of errors reading the recent changes log or because the daily scanner gets stuck and needs to be restarted to get it scanning again. I can turn the dump scan off for plwiki if there is consensus to do that. --Bamyers99 (talk) 17:47, 4 May 2021 (UTC)[reply]

@Bamyers99: I think you can leave it as is. It makes no sense to change it then. Thanks for the clarification. The Polish (talk) 19:54, 4 May 2021 (UTC)[reply]

Is it possible to exclude the content of Template:Chem2 from 'Arrow as ASCII art' checking? The template is designed to use ASCII arrows and they are listed as correct ones in the template documentation. Michał Sobkowski (talk) 10:48, 27 April 2021 (UTC)[reply]

@Michał Sobkowski: ASCII arrows are now excluded if found in Chem2 template. The error 56 listing will get cleaned up when the dump scan runs at the beginning of May. --Bamyers99 (talk) 20:10, 28 April 2021 (UTC)[reply]

Great thanks! Michał Sobkowski (talk) 20:46, 28 April 2021 (UTC)[reply]

Bamyers99, should you exclude pl:Template:Reakcja too from Arrow as ASCII art checking? --The Polish (talk) 18:03, 5 May 2021 (UTC)[reply]

@The Polish: It has been excluded via the config page with this edit --Bamyers99 (talk) 20:40, 5 May 2021 (UTC)[reply]

It looks like the range of valid ISBNs is again outdated (last update was done, afaik, here): The german list of ISBN-13 with wrong Checksum (ID 73) [8] reports many false positives with ISBNs starting 979-8. Could anybody update the list of valid ISBNs as done in December 2020 (I do not know where and how this has to be done)? Many thanks --Bicycle Tourer (talk) 06:01, 7 May 2021 (UTC)[reply]

@Bicycle Tourer: I have just updated the ISBN ranges. I have marked the false positives on the report as done. --Bamyers99 (talk) 15:18, 7 May 2021 (UTC)[reply]

@Bamyers99: Many Thanks! It looks good now. --Bicycle Tourer (talk) 04:29, 9 May 2021 (UTC)[reply]

Is it possible to exclude the content of pl:Template:Wydawnictwo infobox from ISBN with wrong length? The template is designed to use the publisher ISBN prefix so the lenght obviously is wrong. The Polish (talk) 09:42, 5 May 2021 (UTC)[reply]

@The Polish: It has been excluded for 070 via the config page with this edit --Bamyers99 (talk) 20:39, 5 May 2021 (UTC)[reply]

Thanks. The Polish (talk) 21:38, 7 May 2021 (UTC)[reply]

Bamyers99, could you check please why wydawnictwo infobox is not excluded from ID 72? It should be excluded, see: pl:Wikiprojekt:Check Wikipedia/Tłumaczenie. --The Polish (talk) 10:32, 22 May 2021 (UTC)[reply]

@The Polish: Adding template exclusions for a particular error requires additional programming. I have added it for #72. --Bamyers99 (talk) 16:57, 22 May 2021 (UTC)[reply]

On the main project page, the table is sortable according to the number of articles done and to-do, while on the individual project pages the table is not sortable. Can it be fixed? The Polish (talk) 13:15, 7 June 2021 (UTC)[reply]

I have removed the div and table markup at the top of this page since it makes it very difficult to read this page in mobile. Only 3/4th of a line is visible on screen and I have to horizontally scroll the page back and forth to read a full line. It is ironic that talk page of checkwiki project is so difficult to read. The project page and list of errors page has this problem too. Can someone please modify it so that it is readable in mobile? At least for this page, it is better to simply keep it as a normal talk page. ಮಲ್ನಾಡಾಚ್ ಕೊಂಕ್ಣೊ (talk) 13:18, 5 July 2021 (UTC)[reply]

Hello,
We have strange results this morning on frwiki:
List of errors is a partial list of yesterday's errors, which have been already corrected.
I mean "partial" because this is not the full list of yesterdays'errors : for exemple we had yesterday around 30 errors #3, and this morning we have only 13 from yesterday.
Thanks for your help.
--Croquemort Nestor (talk) 04:41, 22 July 2021 (UTC)[reply]

@Croquemort Nestor: The July 20 database dump was scanned at Wed Jul 21 12:01:41, so it would have reported again any errors that were fixed between July 20 and the dump scan. --Bamyers99 (talk) 13:41, 22 July 2021 (UTC)[reply]

Hi NicoV, Can we include mentioned wikis on Check Wikipedia tools?

• bnwikibooks b:bn:উইকিবই:WikiProject Check Wikipedia/Translation
• bnwikisource s:bn:উইকিসংকলন:WikiProject Check Wikipedia/Translation
• bnwikivoyage voy:bn:উইকিভ্রমণ:WikiProject Check Wikipedia/Translation
• bnwiktionary wikt:bn:উইকিঅভিধান:WikiProject Check Wikipedia/Translation
• commonswiki c:Commons:WikiProject Check Wikipedia/Translation (if possible)

The first step has done, created a configuration page for each wiki. Warm Regards, ZI Jony ^(Talk) 16:31, 27 July 2021 (UTC)[reply]

Hi ZI Jony. It's rather Bamyers99 that can add the wikis to Check Wikipedia. I can only add them to WPCleaner if you want. --NicoV ^{(Talk on frwiki)} 20:10, 27 July 2021 (UTC)[reply]

@ZI Jony: bnwikisource, bnwikivoyage, bnwiktionary have been added. Checkwiki doesn't support any wikibooks languages nor commonswiki. --Bamyers99 (talk) 21:28, 27 July 2021 (UTC)[reply]

Bamyers99, thanks for your work. Warm Regards, ZI Jony ^(Talk) 14:21, 28 July 2021 (UTC)[reply]

Hi ZI Jony. If you're interested, I've also configured WPCleaner to work on the 5 wikis (list), and link its configuration to Check Wikipedia configuration. Note that Check Wikipedia configuration on wikisource has been deleted. --NicoV ^{(Talk on frwiki)} 13:33, 30 July 2021 (UTC)[reply]

Hi NicoV, Are you talking about run WPCleaner bot on those wikis? If yes, I can't say yes at the moment! because, community approval is required. Thanks also for your note. Warm Regards, ZI Jony ^(Talk) 15:39, 30 July 2021 (UTC)[reply]

Hi ZI Jony. I was talking about running WPCleaner manually with its user interface (like Wikipedia:WPCleaner/Check wiki), not as a bot. If you want to run WPCleaner as a bot on those wikis, you will have to do it yourself (but now WPCleaner can run there) because I don't operate my bot on wikis where I don't speak the language . --NicoV ^{(Talk on frwiki)} 16:30, 30 July 2021 (UTC)[reply]

Hi NicoV, Thanks for your helping thoughts, you are welcome to run WPCleaner manually. I'm the only one person who is doing Check Wikipedia on those wikis! Warm Regards, ZI Jony ^(Talk) 16:36, 30 July 2021 (UTC)[reply]

Hi ZI Jony. Sorry, I wasn't clear: I won't run it manually (same as with running a bot, I don't when I don't speak the language), but you can use it yourself if it helps you fix the errors. --NicoV ^{(Talk on frwiki)} 16:40, 30 July 2021 (UTC)[reply]

Hi NicoV, it will definitely helps me to fix the errors, but I'm not well aware that how to start and run WPCleaner! Warm Regards, ZI Jony ^(Talk) 17:02, 30 July 2021 (UTC)[reply]

Hi ZI Jony: you have Wikipedia:WPCleaner/Installation if you're interested. --NicoV ^{(Talk on frwiki)} 20:34, 30 July 2021 (UTC)[reply]

Hi NicoV, thanks, I've run WPCleaner itself on enwiki, its work for me here. However, Its not working for bnwikis! I can't recognize all Bangla words, maybe we need to add Bangla fonts (like, AWB "Set edit box font") for WPCleaner. Warm Regards, ZI Jony ^(Talk) 16:08, 1 August 2021 (UTC)[reply]

Hi ZI Jony. Ah, yes, it's difficult for some languages to display them properly with Java. I tried one thing, but it's far from perfect: I download unifont.ttf from Unifont project, and installed this font on my Windows 10 computer (right click on the file and select Install). Then I started WPCleaner, and in the Options (formatting tab), I selected Unifont as the font to use. There are still many places in WPCleaner were the Bangla characters will appear as squares, but the main edit area seem to be able to display both Bangla and Latin characters. Having WPCleaner properly display texts everywhere will require some work on my end. --NicoV ^{(Talk on frwiki)} 11:24, 2 August 2021 (UTC)[reply]

Hi NicoV, I've tried same thing, I use Kalpurush (it is our Bangla wiki font) font in formatting tab but same issue. Please do the needful and let me know when you done. Warm Regards, ZI Jony ^(Talk) 16:13, 2 August 2021 (UTC)[reply]

Hi Bamyers99, could you please clarify that which dump(s) you are using for updating Check Wikipedia? because, all Bangla Wikis Check Wikipedia updated based on July 20, (19 days ago). Could you please make the configuration that Check Wikipedia will update based on most recent dump(s), nor every week? Warm Regards, ZI Jony ^(Talk) 01:32, 8 August 2021 (UTC)[reply]

@ZI Jony: Most of the dumps are stalled per Wikimedia Downloads. --Bamyers99 (talk) 13:40, 8 August 2021 (UTC)[reply]

Hi Bamyers99, could you please check again, per most recent dumps last updated dumps for all Bangla Wikis is 2021-08-01, but per Check Wikipedia Last scanned dump 2021-07-20 (19 days old) and Last Check Wikipedia updated on 2021-07-21. Please try to update Check Wikipedia if possible. Warm Regards, ZI Jony ^(Talk) 14:34, 8 August 2021 (UTC)[reply]

@ZI Jony: Notice how it says "Partial dump" next the the bn... entries. That means that the dump has not finished yet. If you click on the link for bnwiktionary and search for the file that Checkwiki uses: bnwiktionary-20210801-pages-articles.xml.bz2, just above that line it says "waiting". --Bamyers99 (talk) 15:19, 8 August 2021 (UTC)[reply]

Thanks for the clarification. Warm Regards, ZI Jony ^(Talk) 15:52, 8 August 2021 (UTC)[reply]

Hi. Please add fandom to id91. Thank you. IKhitron (talk) 11:06, 23 August 2021 (UTC)[reply]

@IKhitron: id91 only looks for wikipedia.org language sub-domains. Example: https://fr.wikipedia.org/wiki/ --Bamyers99 (talk) 19:32, 24 August 2021 (UTC)[reply]

This is why I've suggested to extend it. IKhitron (talk) 12:46, 25 August 2021 (UTC)[reply]

Hi IKhitron. If you're interested, WPCleaner can also look for any external link that matches an interwiki link, provided the configuration is added to activate that check (error_091_only_local_enwiki and error_091_only_language_enwiki in Wikipedia:WikiProject_Check_Wikipedia/Translation#Errors 091 to 100, see fr:Projet:Correction syntaxique/Traduction#Erreurs 091 à 100). But the list of errors (see Wikipedia:CHECKWIKI/WPC 091 dump), is already big with only the language sub-domains. --NicoV ^{(Talk on frwiki)} 12:26, 30 August 2021 (UTC)[reply]

I see. Thanks, I'll think about this. IKhitron (talk) 15:49, 30 August 2021 (UTC)[reply]

Why does WPC 002 dump include non-empty and not-obviously broken instances of <cite> in it? All those entries also don't match toolforge.

I don't mind cite in article space being tracked somewhere -- it's almost always an error or should be in a template, like currently-off 566 -- I'm just pretty sure it shouldn't be there. Izno (talk) 00:23, 31 August 2021 (UTC)[reply]

Hi Izno. I've added it to WPC because indeed cite tags directly in articles are almost always an error (most of the time probably due to CX2 adding junk to translated articles). I can extract it from error #2 and create a new error if you think it should be better. --NicoV ^{(Talk on frwiki)} 18:46, 1 September 2021 (UTC)[reply]

Could you please? It is also a different fix usually than the 'broken tags' WPC 2 (fix the tag != revert the whole thing and turn it back into a template).

Separately, I don't know if it's possible, but it would be nice if I could get a one-off list of all articles with an HTML tag in the wikitext with the citation class on it. Izno (talk) 18:51, 1 September 2021 (UTC)[reply]

Hi Izno. Cite tags will appear in WPC 571 dump instead of WPC 002 dump. I will see if I can generate the list you're asking for. --NicoV ^{(Talk on frwiki)} 17:01, 3 September 2021 (UTC)[reply]

class="citation" and equivalent (class=citation, class="citation web", etc...) are indicative of the same kind of issues as <cite> appearing in wikitext, so you might consider putting those in the same report permanently. Up to you. Izno (talk) 20:14, 3 September 2021 (UTC)[reply]

Hi Izno. I will do the first generation in WPC 999 dump, and after I'll see if I group it with 571. --NicoV ^{(Talk on frwiki)} 08:26, 4 September 2021 (UTC)[reply]

It is remarkable that the first run only caught the one page that Special:Search catches now (I took care of several hundred others with Special:Search). Thanks! Izno (talk) 14:33, 5 September 2021 (UTC)[reply]

The Science Page statistics appears to display a spurious error --Ancheta Wis   (talk | contribs) 23:46, 8 September 2021 (UTC)[reply]

@Ancheta Wis: Checkwiki does not understand nested {{...{{'s, so it reported the wrong }}. Funny enough, it was an edit that you made on May 23, 2019 that introduced the error. I have fixed the Science article. --Bamyers99 (talk) 01:20, 9 September 2021 (UTC)[reply]

Thank you for explaining this. --Ancheta Wis   (talk | contribs) 01:45, 9 September 2021 (UTC)[reply]

I recently fixed TemplateData on Template:Infobox organization but did not find an easy way to search for other potentially affected templates with errors, particularly high traffic ones. Is there a tracking category for this? This mainly impacts the editing experience when people add templates with the modern (2017?) visual editor ~ 🦝 Shushugah (he/him • talk) 16:40, 30 September 2021 (UTC)[reply]

@Shushugah: You could perform an insource search in the Template space to search for similar code. For example, searching for insource:"proporcionar" in the Template space will find Template:Short description/doc. Hope this helps, and thanks for keeping the templates up to date! GoingBatty (talk) 17:54, 30 September 2021 (UTC)[reply]

Hi all. What is necessary to add a project to Check Wiki? I'd like to have astwiki here. YoaR (talk) 15:56, 7 October 2021 (UTC)[reply]

Hi YoaR. I think it's mostly:

• Create a configuration page (like Wikipedia:WikiProject Check Wikipedia/Translation) on astwiki, configure it, and report here what the page name is
• Bamyers99 will add astwiki to the list of wikis
• I can add it also to WPCleaner if interested

--NicoV ^{(Talk on frwiki)} 16:08, 7 October 2021 (UTC)[reply]

@NicoV: Ok, thanks. I'm going to create that page and I'll come back here. YoaR (talk) 21:32, 7 October 2021 (UTC)[reply]

This change [9] messed up the format of a list within a Template:Expand section by merging the final item in a list with the text following the list. I can't really fault the script too much, because ending the list with a <br> was not correct, but the fact remains that it looked ok before the change and not after. What if anything should have been done here? GA-RT-22 (talk) 14:18, 9 November 2021 (UTC)[reply]

Hello. Is it possible to add a search for usage of mw-content-ltr/rtl styles without usage of dir and lang altogether? Thank you. IKhitron (talk) 15:47, 30 July 2021 (UTC)[reply]

Hi IKhitron. Is this list Wikipedia:CHECKWIKI/WPC 570 dump seems ok to you? --NicoV ^{(Talk on frwiki)} 05:12, 2 August 2021 (UTC)[reply]

Sorry for being late. Looks good, but why they all are only divs? Thank you. IKhitron (talk) 21:12, 16 August 2021 (UTC)[reply]

Hi IKhitron. I'm not filtering on the type of tag, so they should be reported also. Do you think it happens for other types of tags? --NicoV ^{(Talk on frwiki)} 12:15, 30 August 2021 (UTC)[reply]

I'm surprized that span and code are not there. IKhitron (talk) 15:49, 30 August 2021 (UTC)[reply]

Hi IKhitron. A simple search returns results for div tags outside main namespace, but not for span tags or code tags.

https://en.wikipedia.org/w/index.php?title=Special:Search&search=insource%3A%2F\%3Cdiv+class\%3D[%22%27]mw-content-(ltr|rtl)[%22%27]\%3E%2F&profile=advanced&fulltext=1&ns0=1&ns1=1&ns2=1&ns3=1&ns4=1&ns5=1&ns6=1&ns7=1&ns8=1&ns9=1&ns10=1&ns11=1&ns12=1&ns13=1&ns14=1&ns15=1&ns100=1&ns101=1 (for div tags)

This search may miss some instances, but it seems normal that my bot doesn't find any. --NicoV ^{(Talk on frwiki)} 17:32, 30 August 2021 (UTC)[reply]

You can rerun the original bot on hewiki, there is one there. IKhitron (talk) 18:39, 30 August 2021 (UTC)[reply]

Hi IKhitron. Yes, there's one article on hewiki with a span tag, see the dump analysis. --NicoV ^{(Talk on frwiki)} 19:31, 30 August 2021 (UTC)[reply]

Great, thanks. Now, does it check classes combinations? For example, <div style="..." class="... ...rtl ..." lang=...>. IKhitron (talk) 17:07, 9 September 2021 (UTC)[reply]

Hi IKhitron. Normally, yes, I check if class contains mw-content-ltr or mw-content-rtl, so class combination should be ok. There's just a theoretical risk of false positive if there was a class named mw-content-ltr2... --NicoV ^{(Talk on frwiki)} 11:03, 10 September 2021 (UTC)[reply]

Thanks a lot. I have no more questions. IKhitron (talk) 17:23, 11 September 2021 (UTC)[reply]

Hi. Is there something new with this? IKhitron (talk) 17:17, 30 September 2021 (UTC)[reply]

Any news, @NicoV? Thanks. IKhitron (talk) 02:24, 7 December 2021 (UTC)[reply]

Hi IKhitron. No, I thought everything was done. What news did you expect? --NicoV ^{(Talk on frwiki)} 08:36, 7 December 2021 (UTC)[reply]

Well, @NicoV, turning it on would be nice, I think. IKhitron (talk) 03:21, 15 December 2021 (UTC)[reply]

Hi IKhitron. On enWP, it's on, last update. If you mean heWP, dump analysis takes time, so I try to restrict myself to enWP and frWP. I can explain how to run it on other wikis if someone is interested. --NicoV ^{(Talk on frwiki)} 07:58, 15 December 2021 (UTC)[reply]

What about a page ID in CheckWiki? IKhitron (talk) 20:58, 15 December 2021 (UTC)[reply]

Bamyers99 ? --NicoV ^{(Talk on frwiki)} 06:31, 16 December 2021 (UTC)[reply]

@NicoV and IKhitron: I only signed on as a maintainer to keep the program running. I am not interested in adding new reports. --Bamyers99 (talk) 15:22, 16 December 2021 (UTC)[reply]

So, looks like we can't use it. A pity. IKhitron (talk) 23:06, 21 December 2021 (UTC)[reply]

Hello. In category &id=8 a lot of nornal headings, such as "== Известные носители ==</noinclude>" and "== Известные носители ==<onlyinclude>". It's work and requires no fix. Sorry for my english, --Jaguar K (talk) 20:49, 27 December 2021 (UTC)[reply]

@Jaguar K: I have modified checkwiki to ignore noinclude, onlyinclude, includeonly tags for error #8. --Bamyers99 (talk) 21:07, 28 December 2021 (UTC)[reply]

Wikipedia:CHECKWIKI/WPC 551 dump is listing articles that have errors in commented text, like Bijoy Krishna Handique. Can that check be set up to ignore errors in commented text? – Jonesey95 (talk) 16:19, 5 January 2022 (UTC)[reply]

Hi Jonesey95. Yes, it will be in the next release. --NicoV ^{(Talk on frwiki)} 16:58, 5 January 2022 (UTC)[reply]

Hello! I'd like to request bswiki be added to CheckWiki. @Bamyers99:: configuration is here: bs:Korisnik:Ivi104/CheckWiki. @NicoV: I would also like to request WPCleaner support. Thank you! Ivi104 (talk) 13:31, 4 January 2022 (UTC)[reply]

Hi Ivi104. I've added bswiki to WPCleaner, and I've created a basic configuration page for it bs:Korisnik:NicoV/WikiCleanerConfiguration. --NicoV ^{(Talk on frwiki)} 07:00, 5 January 2022 (UTC)[reply]

@NicoV: Thank you, I will modify it. Small correction: in WPCleaner, when selecting the wiki, it says "bs - Slobodna enciklopedija", which translates to "bs - Free encyclopedia". The correct name is "Wikipedia na bosanskom jeziku". Similarly, please change the name for hrwiki to "Wikipedija na hrvatskom jeziku" (notice the j in Wikipedija here). All the best! Ivi104 (talk) 13:09, 5 January 2022 (UTC)[reply]

Hi Ivi104. Done. --NicoV ^{(Talk on frwiki)} 14:03, 5 January 2022 (UTC)[reply]

bswiki has been added to CheckWiki, see: bswiki. --Bamyers99 (talk) 16:31, 5 January 2022 (UTC)[reply]

@Bamyers99: Thank you! Please take a look at error #95 on bswiki. I do not see a link to userspace in any of the listed mainspace pages (197.871 in total). Is this a bug of some sort? Ivi104 (talk) 02:19, 6 January 2022 (UTC)[reply]

@Ivi104: Sorry about that, I forgot to do a configuration step. The dump scan has been rerun. --Bamyers99 (talk) 05:08, 6 January 2022 (UTC)[reply]

Bamyers99 could you please change the configuration for bnwiki like arwiki and cswiki? Dump scan will remain same, but list will update daily based on last dump. Because not possible to mark done every time manually, and it makes confused Warm Regards, ZI Jony ^(Talk) 08:27, 13 February 2022 (UTC)[reply]

@ZI Jony: bnwiki is now scanning daily. --Bamyers99 (talk) 14:43, 17 February 2022 (UTC)[reply]

Bamyers99, Thanks. Could you please disable permanently 6 and 37 for all Bangla Wikis, both are not working for Bangla. It's showing wrong info. Warm Regards, ZI Jony ^(Talk) 08:15, 28 February 2022 (UTC)[reply]

@ZI Jony: An error message can be disabled by setting its priority to -2 on the Translation page. Example: error_006_prio_bnwiki=-2 --Bamyers99 (talk) 19:43, 28 February 2022 (UTC)[reply]

First off, I am unsure if this is actually a bug. I am pretty sure that what I was doing with wikitext is nonstandard. Nevertheless, I had a table in a footnote, and the WikiCleaner bot did something weird that made the table gobble up the next footnote? Here's the diff that I take issue with. If you compare the before to after, I don't know any way to describe it other than "the table gobbled up the next footnote". I made a simplified example in my user sandbox of the issue with the script (specifically the issue is replacing }}} with }}). Here's the diff, here's the before, here's the after. As you can see, in the before, the table works just fine, but after WikiCleaner, the table has an extra column for some reason, and it has pulled the unrelated other footnote into it. I know this is a super weird edge case so I am not too concerned (I would not be surprised if this is only an issue on Space Launch System and no other page on the whole site), but perhaps the bot could be programmed to leave this specific pattern alone? Leijurv (talk) 00:13, 22 March 2022 (UTC)[reply]

Also I'm totally willing to accept the answer of "three curly brackets in a row is always bad practice / syntax error / not a good idea", if so I'd just like to know some alternate way to put the table in the footnote that doesn't need }}}. Leijurv (talk) 00:28, 22 March 2022 (UTC)[reply]

Also, I got the example of how to put a table in a footnote from Template:!, and the example there does indeed have the triple curly brackets at the end. Leijurv (talk) 00:33, 22 March 2022 (UTC)[reply]

Hi Leijurv. Yes, this seems to be a weird edge case... I tried an edit on your simplified example that seems to work and will normally prevent my tool from changing the syntax. I can also look into modifying my tool for ignoring the magic word {{!}}. --NicoV ^{(Talk on frwiki)} 17:17, 22 March 2022 (UTC)[reply]

Hi Leijurv. I checked: the edit was manual, WPCleaner is only reporting the strange syntax and suggesting to remove the extra curly bracket. I still changed the code so that it won't report {{!}} at the beginning of a line to avoid the risk of replacing a table end. --NicoV ^{(Talk on frwiki)} 19:00, 22 March 2022 (UTC)[reply]

The webserver and backend jobs have been migrated from the Grid Engine to Kubernetes. There should be no changes from an end user perspective. Leave a message here if something doesn't look right. --Bamyers99 (talk) 02:48, 28 May 2022 (UTC)[reply]

Could this be the cause of the slow update of local projects this month? There is still a handful of projects that have not been updated with the June 1 dump. For svwp that would usually have been achieved by June 3. Plumbum208 (talk) 11:56, 13 June 2022 (UTC)[reply]

@Plumbum208: Fewer dumps can run simultaneously so it is taking longer to process them. I am switching more wikis to daily scans and stopping dump scans for all daily scanned. The new list of daily scans is: enwiki dewiki eswiki frwiki arwiki cswiki plwiki bnwiki nlwiki nowiki cawiki hewiki ruwiki. --Bamyers99 (talk) 13:25, 13 June 2022 (UTC)[reply]

Hello. I don't know where to report the bug, so I report it here. CW Error 3 doesn't exclude redirects of {{reflist}}. --Yining Chen (talk) 07:02, 24 June 2022 (UTC)[reply]

@Yining Chen: All of the redirects have to be added to the translation page for a project in the error_003_templates_<wikiname> section. For example on the enwiki translation page, see the error_003_templates_enwiki section. A list of project translation pages can be found at https://checkwiki.toolforge.org/ --Bamyers99 (talk) 12:26, 24 June 2022 (UTC)[reply]

Thank you. --Yining Chen (talk) 12:52, 24 June 2022 (UTC)[reply]

I'm sure this has been asked before, but why can't I sort the errors by number of To-Do's? - FakirNL (talk) 11:12, 26 June 2022 (UTC)[reply]

I've noticed a large amount of Bogus image options on the 2022 Brazilian general election and 2022 São Paulo gubernatorial election pages. I'm not quite comfortable enough to use WPCleaner to correct these errors; I'm not even sure they are even correctly flagged errors, just given the amount on each page. Any advice? Lindsey40186 (talk) 00:37, 20 July 2022 (UTC)[reply]

Neither LintHint nor Page information, the two standard ways of finding Linter errors, reports bogus image options on those pages. How are you being advised of these errors? – Jonesey95 (talk) 05:20, 20 July 2022 (UTC)[reply]

Hi Lindsey40186. I will check, but in general there are several situations where WPCleaner report a false positive for bogus image options (a few things on my todo list). --NicoV ^{(Talk on frwiki)} 06:45, 20 July 2022 (UTC)[reply]

Hi Lindsey40186. I checked, I confirm this is a false positive, currently WPCleaner doesn't parse correctly images where the image name is the result of a template call...

[[File:{{tmpv|Liberal Party (Brazil, 2006)|Infobox political party|1|logo}}|100x100px|Liberal Party (PL)]]

--NicoV ^{(Talk on frwiki)} 07:09, 20 July 2022 (UTC)[reply]

Ok, good (well, maybe not good, but understandable)! I will leave them be and just triple check when that error pops up. Lindsey40186 (talk) 14:29, 20 July 2022 (UTC)[reply]

Moin Moin together from german Wikipedia, I wanted to ask, if its possible to extend ID 1. In the german Wikipedia we have a template for Navis to show in one block (de:Vorlage:NaviBlock). Sometimes it happens that people write then:

{{NaviBlock
|Vorlage:Navi xxxx}}

When we search for this issues we use Cirrus with:

insource:/\|Vorlage\:Navigationsleiste/

Is it possible to program the ID 1 in this way, find this issues too? Regards --Crazy1880 (talk) 08:23, 5 July 2022 (UTC)[reply]

Moin Bamyers99, could you have a look at this wish, perhaps its possible? Regards --Crazy1880 (talk) 08:10, 17 July 2022 (UTC)[reply]

@Crazy1880: It would take more work than I am interested in doing to implement something to detect this special case. --Bamyers99 (talk) 15:30, 17 July 2022 (UTC)[reply]

Moin Moin Bamyers99, thanks for the answer. Then we will go in another way ;) Regards --Crazy1880 (talk) 04:42, 22 July 2022 (UTC)[reply]

Hello, could you please check if the robot is correctly running on frwiki ? We did not get any new errors since some days. Thanks by advance. --Croquemort Nestor (talk) 04:48, 4 August 2022 (UTC)[reply]

@Croquemort Nestor: Thanks for reporting. There was an issue with dump scans having priority over daily scans. This has been fixed. --Bamyers99 (talk) 21:53, 5 August 2022 (UTC)[reply]

@Bamyers99: Thanks for having fixed this issue. I am afraid there is a new one today : the list out of the daily scan this morning is the same as the one we had yesterday night, that is : same picture, showing the errors corrected yesterday, the errors not corrected yesterday, but looks like there is no new errors coming from yesterday's contributions. Thanks for your help again. --Croquemort Nestor (talk) 04:29, 8 August 2022 (UTC)[reply]

@Bamyers99: Looks fine this morning. --Croquemort Nestor (talk) 05:20, 9 August 2022 (UTC)[reply]

Hi @Bamyers99. I think #75 is producing some false positives.

For example, on frwiki, it reports Action directe (théorie politique) because of the following combination:

* '''Action directe non violente'''...

: Dans ...
:* Sensibiliser ...

I think the line starting with : should stop the search of :* after the line starting with *.

The existing code is:

if ( index( $line, q{*} ) == 0 or index( $line, q{#} ) == 0 ) {
    $list = 1;
}
elsif ( $list == 1
    and ( $line ne q{} and index( $line, q{:} ) != 0 ) )
{
    $list = 0;
}

if ( $list == 1
    and
    ( index( $line, q{:*} ) == 0 or index( $line, q{:#} ) == 0 ) )
{
    error_register( $error_code, substr( $line, 0, 40 ) );
}

Maybe modifying it this way would solve this problem:

if ( index( $line, q{*} ) == 0 or index( $line, q{#} ) == 0 ) {
    $list = 1;
}
elsif ( $list == 1 and $line ne q{} )
{
    if ( index( $line, q{:*} ) == 0 or index( $line, q{:#} ) == 0 ) {
        error_register( $error_code, substr( $line, 0, 40 ) );
    }
    else {
        $list = 0;
    }
}

What do you think? NicoV ^{(Talk on frwiki)} 20:36, 17 August 2022 (UTC)[reply]

@NicoV: Code has been changed. --Bamyers99 (talk) 22:00, 17 August 2022 (UTC)[reply]

 Resolved

Hi @Bamyers99. Would it be possible to modify #67 to ignore references after a ! or a ? ?

At least for frwiki, it seems that the rule should not be applied for such punctuation character (recent talk). It's often the case of citations, and in such case, the punctuation can be considered as part of the citation.

Only 3 other wikis have this rule activated:

• jawiki but no errors reported
• nlwiki but they have both #61 and #67 activated, which is strange
• plwiki

I think it's only one line to change. NicoV ^{(Talk on frwiki)} 18:08, 21 August 2022 (UTC)[reply]

@The Polish : do you know what is the rule for plwiki? NicoV ^{(Talk on frwiki)} 18:11, 21 August 2022 (UTC)[reply]

@NicoV: Just add them to error_067_abbreviations_frwiki=. Refs after the abbreviations are ignored. --Bamyers99 (talk) 20:11, 21 August 2022 (UTC)[reply]

Hi @Bamyers99. Thanks for the tip, I changed the configuration and it seems to work. --NicoV ^{(Talk on frwiki)} 17:37, 3 September 2022 (UTC)[reply]

Hi all! I did some work and updated the ISBNs in a few pages in https://en.wikipedia.org/wiki/Wikipedia:WikiProject_Check_Wikipedia/ISBN_errors , could someone make sure the list gets updated with the tool? Thanks! BhamBoi (talk) 03:59, 12 December 2022 (UTC)[reply]

How far did you get with the instructions? You can update the list yourself. (I have updated it for you this time.) – Jonesey95 (talk) 05:39, 12 December 2022 (UTC)[reply]

Hi together, in de:WP CW category 105 reports cases, where a header is integrated into an argument for template de:template:Mehrspaltige Liste as leading part of the argument liste=. In this case you have the (difficult to read for humans) syntax

{{Mehrspaltige Liste |liste==== Headertext ===
 ...}} 

where the first = of ==== (4x=) is part of declaring "This is argument 'liste'", while the trailing === (3x) define "here starts a header level 3". It is esp. confusing because it looks like a header with 4 opening = and 3 closing ===, though it is syntactically correct (at least it is interpreted by the Wiki-Software as intended).

This scenario leads to false positives in CW 105.

The reporting can be solved by splitting the 4 = (e.g. liste==== Headertext ===) into

liste=
=== Headertext ==="

, but this is considered an unwanted nulledit by some users.

An example of a change solving this false positive can be found in https://de.wikipedia.org/w/index.php?title=Kling_Klang&diff=prev&oldid=229540863 (watch out for text "liste==== Keimzeit" or at least "Keimzeit"). User @Fan-vom-Wiki: has stated, that Visual Editor will delete this separating newline (I have not checked, whether this is true), so that long term more false positives like this will occur, even when being fixed as shown above.

I'm aware that this is a very special case and that it might be difficult to have an exception for this special pattern, but because a growing amount of false positives can be expected I want to ask whether this can be fixed.

Many thanks in advance Bicycle Tourer (talk) 10:49, 8 January 2023 (UTC)[reply]

Thank you for this detailled error message. Here is another example: Here I changed some parameters in the template in the Visual-Editor-Mode. The next Edit was that another user inserted a paragraph. So in my point of view, a follwing expression is necessary: If (German Wikipedia AND first position in the parameter „Text“ in the Template „Mehrspaltige Liste“) Then exclude this error. Thanks --Fan-vom-Wiki (talk) 15:42, 8 January 2023 (UTC)[reply]

@Bicycle Tourer and Fan-vom-Wiki: I have revised the 105 check to ignore this usage pattern. --Bamyers99 (talk) 00:46, 10 January 2023 (UTC)[reply]

@Bamyers99: Many thanks for solving this so quickly. --Bicycle Tourer (talk) 11:59, 10 January 2023 (UTC)[reply]

Well Done! Thank you very much! --Fan-vom-Wiki (talk) 14:05, 10 January 2023 (UTC)[reply]

Hi, someone who understands how it works might like to look at this edit. It has incorrectly added an additional close to a comment that resulted in the references and categories "vanishing". Probably a rare edge case the script can't handle, but worth fixing if possible. Thanks. MarcGarver (talk) 18:47, 9 February 2023 (UTC)[reply]

@MarcGarver, seems like a GIGO case. Added here. — Qwerfjkl talk 19:01, 9 February 2023 (UTC)[reply]

Hi @Bamyers99, Can you please add tewiki to the Check Wikipedia list? The Translation page is created at te:వికీపీడియా:WikiProject Check Wikipedia/Translation.

Thanks. __Chaduvari (talk) 22:42, 9 February 2023 (UTC)[reply]

@Chaduvari:  Done, see tewiki --Bamyers99 (talk) 00:58, 11 February 2023 (UTC)[reply]

Thanks @Bamyers99, for the quick action. __Chaduvari (talk) 09:20, 11 February 2023 (UTC)[reply]

Please exclude articles with [[template:under construction]] in errors results -Gunyam (talk) 13:39, 14 February 2023 (UTC)[reply]

@Gunyam: Should articles with {{in use}} be excluded instead? {{in use}} asks other users to not edit the article, while {{under construction}} invites edits from other editors. Thanks! GoingBatty (talk) 15:15, 14 February 2023 (UTC)[reply]

@Gunyam: Which error number do you want this excluded from? --Bamyers99 (talk) 15:33, 14 February 2023 (UTC)[reply]

It may not be a good idea to exclude pages in article space from any error checks. Live articles that are visible to the world should not have syntax errors in them, even if they are actively being expanded or revised. – Jonesey95 (talk) 15:50, 14 February 2023 (UTC)[reply]

@Bamyers99 @GoingBatty @Jonesey95 +1 so, lets settle then. i withdrew. i agree with Jonesey95. -Gunyam (talk) 16:22, 14 February 2023 (UTC)[reply]

The error is not limited to error 91,but also on other error numbers. This happens on bn.wikipedia. - Gunyam (talk) 17:10, 15 February 2023 (UTC)[reply]

@Gunyam: Example article? --Bamyers99 (talk) 23:13, 15 February 2023 (UTC)[reply]

Bamyers99 Examle article. -Gunyam (talk) 00:50, 16 February 2023 (UTC)[reply]

@Gunyam: "https://en.wikipedia.org/wiki/Kosovo" should be replaced with en:Kosovo --Bamyers99 (talk) 01:25, 16 February 2023 (UTC)[reply]

Many articles reflist are fixed. But it shows a lot of them as unfixed. Checked 15 articles and all of them are fixed. -03:47, 21 February 2023 (UTC) రుద్రుడు (talk) 03:47, 21 February 2023 (UTC)[reply]

There are two additional templates in Tewiki which are redirects to "Template:reflist". The number of transclusions of these templates nearly matches the CheckWiki list. These two template names were included in the "WikiProject Check Wikipedia/Translation" after the first dump was taken. I presume, the next dump will be devoid of this error (ID #003). __Chaduvari (talk) 04:48, 21 February 2023 (UTC)[reply]

@Chaduvari dump eta please? where to find time table of dumps? Comment Track_listing_deprecated_on_our_wiki? -రుద్రుడు (talk) 07:04, 21 February 2023 (UTC)[reply]

@రుద్రుడు, Please see the #Tewiki - issue with error 095 discussion above. __Chaduvari (talk) 07:26, 21 February 2023 (UTC)[reply]

 Resolved please add template - "మూలాలు" to fix reflist missing in te.wikipedia రుద్రుడు (talk) 08:00, 24 February 2023 (UTC)[reply]

please refer example te:నెల్లూరు రుద్రుడు (talk) 08:05, 24 February 2023 (UTC)[reply]

మూలాలు added with this edit. --Bamyers99 (talk) 16:14, 24 February 2023 (UTC)[reply]

Whenever i add or articles with DEFAULTSORT on telugu wikipedia, shows error: "defaultsort with special characters". Defaultsort is applicable to english language only? ping @Chaduvari and @NicoV also. I am using wpcleaner. రుద్రుడు (talk) 12:53, 24 February 2023 (UTC)[reply]

Hi రుద్రుడు. By default, only a very limited set of characters are considered as ok (see list). For a given wiki, I can add a list of authorized characters in WPCleaner. You just need to tell me what characters are valid. --NicoV ^{(Talk on frwiki)} 15:15, 25 February 2023 (UTC)[reply]

@Bamyers99, Tewiki checkwiki report shows there are more than 110000 pages with error_095. I checked a few of those pages, but did not notice any issue there. Even redirect pages are included in the list. Am I doing anything wrong? Or do I need to add something to the whitelist? Appreciate help. Thanks. __Chaduvari (talk) 03:52, 12 February 2023 (UTC)[reply]

@Chaduvari: I forgot to do a step in the new project setup. I have now done this step. I see that someone is fixing other issues, so if I rerun the 01 February wiki dump, the fixed issues will reappear. The other option is wait for the 20 February wiki dump to run for the 095 errors to disappear. --Bamyers99 (talk) 18:56, 12 February 2023 (UTC)[reply]

@Bamyers99: Lets wait until 20th. Thank you. __ Chaduvari (talk) 23:37, 12 February 2023 (UTC)[reply]

@Bamyers99: The 21 Feb dump is ready. Can you find time to run now? Thanks. __ Chaduvari (talk) 02:28, 25 February 2023 (UTC)[reply]

@Chaduvari: The dump processing is automated. It takes days to process all of the different language dumps. --Bamyers99 (talk) 15:18, 25 February 2023 (UTC)[reply]

Okay. Thanks.__ Chaduvari (talk) 00:39, 26 February 2023 (UTC)[reply]

 Resolved

Per suggestion deleted </nowiki>. It seems wrong? Correct? రుద్రుడు (talk) 08:36, 4 March 2023 (UTC)[reply]

@రుద్రుడు: The fix is correct. The closing /nowiki tag was missing an opening nowiki tag. --Bamyers99 (talk) 15:09, 4 March 2023 (UTC)[reply]

Moin Moin together from Germany, hello Bamyers99, the last daily scan from the german CheckWikipedia is two day ago, can you check if there is a task is not running? King Regards --Crazy1880 (talk) 08:55, 3 March 2023 (UTC)[reply]

@Crazy1880: Thanks for the report. The program that dispatches the dump scans and daily scans uses an api (url) to submit the jobs, that api was blocked without notice. I have recoded the job submitter to use a different api. The daily scans are running, however, only 3 can run simultaneously (currently dewiki, enwiki, eswiki). The rest are queued and waiting. The dump scans are running, however, only 2 can run simultaneously (currently alswiki, afwiki). --Bamyers99 (talk) 21:51, 3 March 2023 (UTC)[reply]

Works, thanks Bamyers99 ;) Regards --Crazy1880 (talk) 09:17, 9 March 2023 (UTC)[reply]

I was thinking that there could be an error relating to redirects without rcats (redirects which don’t have Template:Rcat shell, which should have things like Template:R from chemical formula for example). This would increase redirect maintenance categorisation, which, as said here, should happen. greyzxq ^talk 19:14, 9 March 2023 (UTC)[reply]

 Resolved

For error 10 on bn:আকৃতি it is suggesting [[উপবৃত্ত]] ] or [[উপবৃত্ত]] instead of [[উপবৃত্ত]]. Correct? రుద్రుడు (talk) 02:35, 5 March 2023 (UTC)[reply]

Hi రుద్రుడు. This is an issue with WPCleaner, not Check Wikipedia (which only detects problems). I will look into it. --NicoV ^{(Talk on frwiki)} 19:05, 5 March 2023 (UTC)[reply]

Hi రుద్రుడు. It should be fixed in WPCleaner. --NicoV ^{(Talk on frwiki)} 14:18, 11 March 2023 (UTC)[reply]

 Resolved

@Bamyers99 please add kn wiki for checkwiki errors. Translation page https://kn.m.wikipedia.org/wiki/%E0%B2%B5%E0%B2%BF%E0%B2%95%E0%B2%BF%E0%B2%AA%E0%B3%80%E0%B2%A1%E0%B2%BF%E0%B2%AF:WikiProject_Check_Wikipedia/Translation రుద్రుడు (talk) 12:36, 10 March 2023 (UTC)[reply]

@రుద్రుడు: kn has been added here. --Bamyers99 (talk) 14:42, 11 March 2023 (UTC)[reply]

 Resolved

bn:.নেট ফ্রেমওয়ার্ক has [[ไฟล์:DotNET3.0Stack.png|থাম্ব|333x333পিক্সেল|ডটনেট ফ্রেমওয়ার্ক ৩.০ আর্কিটেকচার]]. Its nothing but a variant of file [Image] in this case. It shows error: "Article contains a link like [[text|text2|text3]] (displays as text2|text3) or [[text||text3]] (displays as |text3)". advice! రుద్రుడు (talk) 09:47, 4 March 2023 (UTC)[reply]

@రుద్రుడు: I don't see this article in the error 32 list. A check of the article does not show the error either. The : in the link causes the multiple |s to be ignored by error 32 and not reported. --Bamyers99 (talk) 16:45, 4 March 2023 (UTC)[reply]

@రుద్రుడు: I see the error with WPCleaner, but it's normal because there's indeed a problem. Look at the page there's a red line displaying থাম্ব|333x333পিক্সেল|ডটনেট ফ্রেমওয়ার্ক ৩.০ আর্কিটেকচার at the beginning of the article. --NicoV ^{(Talk on frwiki)} 21:50, 13 March 2023 (UTC)[reply]

It seems bn:২০২২ সাফ অনূর্ধ্ব-১৮ মহিলা চ্যাম্পিয়নশিপ flagging <ref name="IND2022" /> as dupes? రుద్రుడు (talk) 13:59, 2 March 2023 (UTC)[reply]

@రుద్రుడు: What error number are you referring to? --Bamyers99 (talk) 15:29, 2 March 2023 (UTC)[reply]

I did not save. If i find more, i will note and reply. రుద్రుడు (talk) 08:22, 4 March 2023 (UTC)[reply]

@రుద్రుడు: Are you sure it was flagged as dupes? Because on the version before you edited it with WPCleaner, the ref was flagged because there are no references tag (or equivalent template) after. --NicoV ^{(Talk on frwiki)} 21:55, 13 March 2023 (UTC)[reply]

 Declined

It still shows err 3 on kn:೧೩ ಉಪನಿಷತ್ತುಗಳ ಸಾರಾಂಶ. Article contains reflist template. Did i incorrectly added? రుద్రుడు (talk) 09:26, 14 March 2023 (UTC)[reply]

@రుద్రుడు: I don't see template ಉಲ್ಲೇಖಗಳು used on that page. After the ಉಲ್ಲೇಖ section is just categories. --Bamyers99 (talk) 17:02, 14 March 2023 (UTC)[reply]

@Bamyers99 Sorry for noise. Next time i take screen shot or some kind of backup. రుద్రుడు (talk) 18:31, 14 March 2023 (UTC)[reply]

 Resolved

Err 3 is reported on kn:ಅರ್ಥಶಾಸ್ತ್ರ with ಉಲ್ಲೇಖ ಪಟ್ಟಿ template. ಉಲ್ಲೇಖ ಪಟ್ಟಿ is included on kn:ವಿಕಿಪೀಡಿಯ:WikiProject Check Wikipedia/Translation. Showing as done _ok? రుద్రుడు (talk) 02:32, 27 March 2023 (UTC)[reply]

Hi రుద్రుడు. This template doesn't seem to be configured for error 3, only for error 78. So, it's normal. --NicoV ^{(Talk on frwiki)} 13:49, 28 March 2023 (UTC)[reply]

Due to a filesystem migration, the daily and dump scans will be delayed by up to 19 hours. See Two toolforge outages coming next week, Monday and Thursday (lists.wikimedia.org) for details. --Bamyers99 (talk) 00:25, 3 April 2023 (UTC)[reply]

 Resolved

Something strange is happening on te wiki. te:ఆప్టికల్ మౌస్ has మూలాలు సమీక్షించండి on top of article. This article is not listed on error 3. See also somewhat related error? Keeping @NicoV in loop. రుద్రుడు (talk) 01:44, 1 April 2023 (UTC)[reply]

Hi రుద్రుడు. I think that for both CW#3 and CW#78, the parameter templates is not a list of templates, but a list of regular expressions for templates. You currently have మూలాలు in the configuration, so any template that starts with మూలాలు (like మూలాలు సమీక్షించండి) is counted. If you want to clearly limit to the template with this exact name, you should add \s*[|}] (any number of whitespace characters, then a pipe or a closing bracket), like on some other lines (RE\s*[|}]). --NicoV ^{(Talk on frwiki)} 09:12, 8 April 2023 (UTC)[reply]