Wikipedia talk:WikiProject Check Wikipedia/Archive 7

Archive 1

←

Archive 5

Archive 6

Archive 7

Archive 8

Archive 9

svwiki has no errors reported

 Done

Hi, I wish you luck with Labs...

See Wikipedia talk:WPCleaner#CheckWikipedia_does_not_work_on_sv.wikipedia.org.5B....5D, svwiki has no errors reported on Labs, while some are reported on toolserver. --NicoV ^{(Talk on frwiki)} 06:55, 6 November 2013 (UTC)

NicoV, there is a problem with the svwiki dump files for the past few months. There is some sort of corruption about 1/2 way thru the dump. I don't think it is a bad dump as it always happens and only happens with svwiki. Probably some borked articles. I'll hunt down the articles in a few days. I've got to debug it on my computer and the computer is currently running the latest enwiki dump. I'll then be busy fix errors. Dump processing on my 3-year old computer is 68% done. At labs, it is 11% done and it started at the same time as my computer. Bgwhite (talk) 07:47, 6 November 2013 (UTC)

NicoV and Josve05a‎, errors are up. It's still not going thru 100% of the dump, but atleast there are errors to play with. Bgwhite (talk) 08:21, 24 November 2013 (UTC)

SCORE! Finally! Thanks! -(t) Josve05a (c) 08:31, 24 November 2013 (UTC)

Bgwhite, NicoV: error #37 is showinga lot of false errors. Since svwp now supports special characters in DEFAULTSORT. So the error what me to create a DEFAULTSORT with the exact same name as the title. Deactivate? -(t) Josve05a (c) 09:07, 24 November 2013 (UTC)

Josve05a Just mark them all done and I'll deactivate the error. Boy, marking done 2/3 of the errors is a nice feeling. Bgwhite (talk) 09:12, 24 November 2013 (UTC)

Bgwhite, THAT FEELING. Quote I said in my mind: "Is this how it feels to vandalise Wikipedia?". -(t) Josve05a (c) 09:16, 24 November 2013 (UTC)

False positive for #94

 Resolved

Hi Bgwhite, I don't understand why Jungle de Calais is detected by #94 with the notice </ref>, est une expression désignant. The reported ref tag seems to be in a valid reference <ref name="ARTFIG00250">{{Lien web|titre = Calais : les forces de l'ordre se préparent à une nouvelle nuit tendue|url = http://www.lefigaro.fr/actualite-france/2015/11/09/01016-20151109ARTFIG00250-quand-les-migrants-chargent-la-police-a-calais.php|site = Le Figaro|consulté le = 2015-11-29}}</ref>. --NicoV ^{(Talk on frwiki)} 03:55, 12 December 2015 (UTC)

NicoV There is an error, but why it shows the wrong place, I don't know. Problem was with <ref name="bfm <933162"> Per WP:REFNAME, Quote-enclosed reference names may not include a less-than sign (<) or a double straight quote symbol ("). Bgwhite (talk) 09:54, 12 December 2015 (UTC)

More useful notice by #78

 Resolved The script inserts a <br> which gets escaped and the whole notice is not useful. Matěj Suchánek (talk) 11:18, 23 December 2015 (UTC)

Matěj Suchánek Is this the web page output? If so, then fixed. Bgwhite (talk) 23:53, 23 December 2015 (UTC)

Yes, it is and it looks better now, thanks! Matěj Suchánek (talk) 09:18, 24 December 2015 (UTC)

ISSN Idea for CheckWiki

 Done

CheckWiki looks for invalid ISBNs, which helps me a lot. But, I was wondering if there was a way to look for invalid ISSNs without having to click on each ISSN to figure out which one is invalid. --MrLinkinPark333 (talk) 00:49, 24 December 2015 (UTC)

MrLinkinPark333 Very good question. I know nothing about ISSNs, so I took a look. ISSNs are similar to ISBNs in that the last digit is a checksum. It is easy to calculate the checksum. The ISSN is in a standard 8 digit format, plus hyphen. ISBNs are in a 10 or 13 digit format. It is easy to check if the ISSN is in the correct format. There are other format tests that can be done to ISBNs that are not applicable to ISSNs. For example, an ISBN contains a country code and we can see if that code is correct. ISSNs don't have this. We could check if the ISSN has been registered, but to get a copy of the file, it would cost 20,000 Euros. The file would need to be updated regularly. So, that is a no go.

This will only apply for standalone ISSNs as ISSNs in cite templates are already checked for these things. If there is an error, it gets put into Category:CS1 errors: ISSN.

tl;dr Yes, we can check for correct checksum and format for standalone ISSNs, not in cite templates. @Magioladitis: Bgwhite (talk) 05:29, 24 December 2015 (UTC)

Good suggestion, I will add it also to WPCleaner ! Error 106 ? --NicoV ^{(Talk on frwiki)} 09:13, 24 December 2015 (UTC)

Bgwhite, I can easily add that to WPCleaner (copy/paste of ISBN stuff). What will the error numbers? Suggestions: 106 for ISSN with incorrect syntax (same as #69), 107 for ISSN with wrong length (same as #70), 108 for ISSN with wrong position of X (same as #71), 109 for ISSN with wrong checksum (same as #72 and #73). --NicoV ^{(Talk on frwiki)} 16:45, 24 December 2015 (UTC)

NicoV Yes, this sounds like a good idea. Not sure about 108, could be folded into 109. Will work on coding this up. After we see how many errors there are and where they fall (#106, #107, etc), then we can set things in stone. Bgwhite (talk) 19:00, 24 December 2015 (UTC)

Ok! Almost in place for WPCleaner, I just have to code the detections themselves, which should be pretty straightforward once I know how the errors are organized. It will be working much like ISBN detections, with even a list of ISSN errors to work on. --NicoV ^{(Talk on frwiki)} 11:23, 26 December 2015 (UTC)

Yes, I like the idea. @Knife-in-the-drawer: would like it too if they ever come online again. -- Magioladitis (talk) 10:05, 24 December 2015 (UTC)

@NicoV: Thank you for making this idea a part of this WikiProject! I have a question regarding ISSNs errors: in these two articles Ojibwe writing systems and Klaus Schmidt (archaeologist), I checked the ISSNs. These ISSNS seem to be valid according to Worldcat, yet the articles are still showing up as ISSN errors. Would you happen to know why? Thank you.--MrLinkinPark333 (talk) 21:53, 27 December 2015 (UTC)

@MrLinkinPark333:, for Ojibwe writing systems, I see that ISSN 0031-5671 is used, but at worldcat, it says 0831-5671. For Klaus Schmidt (archaeologist), I don't know, it would be nice for {{ISSN}} to display an explanation on which is wrong and why... I don't know how it works on enwiki, but on frwiki the templates dealing with ISBN display an error message when needed (not yet for ISSN, I've asked about it). --NicoV ^{(Talk on frwiki)} 22:20, 27 December 2015 (UTC)

In fact, for the second article, it seems there's an inversion of 2 digits, see MIAR. --NicoV ^{(Talk on frwiki)} 22:24, 27 December 2015 (UTC)

@NicoV: Very strange. If you search 0031-5671 on worldcat, the first search result is the one with 0831-5671. I've now fixed four articles using the same inaccurate ISSN. Regarding Klaus Schmidt, if you look up 0513-9345, it gives you the same journal title as 0153-9345, but not the same reuslts. These search results shouldn't be returned since they're wrong. --MrLinkinPark333 (talk) 22:30, 27 December 2015 (UTC)

MrLinkinPark333 You can give bad ISSN and ISBN numbers to search in WorldCat and you may get the correct result. This does not mean those numbers are valid. It means a bad number was common enough to give search results to the correct media. Think of it as misspelling something in a Google search, but it returns what you were looking for. For general search engines, this is the correct choice.

The {{ISSN}} and cite templates checks for two things on ISSNs. 1) Format, is it in xxxx-xxxx? 2) Is the checksum digit correct? Bgwhite (talk) 06:25, 28 December 2015 (UTC)

NicoV #106 is coded up. It will check for:

1. ISSN (-|:|#|;) [0-9]
2. \[\[ISSN\]\] [0-9]
3. ISSN[0-9]
4. ISSN XXXXXXXX and ISSN XXXX XXXX
5. ISN [0-9]
6. issn [0-9]

There must be a space at the start of each regex. Bgwhite (talk) 10:07, 30 December 2015 (UTC)

Thanks Bgwhite, I will add it in WPCleaner probably this weekend. Will it be live by then ? --NicoV ^{(Talk on frwiki)} 11:31, 31 December 2015 (UTC)

NicoV I hope so. I'd like to add it before January's dumps start up. December's dumps won't be done for atleast a week. I've also updated the list above. My wife is off work till Monday, so she comes first, darn it. Bgwhite (talk) 00:18, 1 January 2016 (UTC)

Ok, I've released a first version of WPCleaner with #106, I will check it when the list starts to fill up. --NicoV ^{(Talk on frwiki)} 15:38, 2 January 2016 (UTC)

I've just released a version of WPCleaner with #106 (incorrect syntax), #107 (wrong length) and #108 (wrong checksum). Testing is welcome, because I've only done basic testing (MrLinkinPark333?) --NicoV ^{(Talk on frwiki)} 12:14, 3 January 2016 (UTC)

What about dashes? This works: ISSN 0028-0836; with en dash it doesn't: ISSN 0028–0836. --79.50.29.4 (talk) 03:57, 3 January 2016 (UTC)

79.50.29.4 The code only accepts a regular dash. If there is any other type of dash (en or em), it will spit out an error. Bgwhite (talk) 22:45, 5 January 2016 (UTC)

Thousands of false positives on error #43

 Done

Hello! In russian Wikipedia we insert tables into templates using something like this:

|template_param = 
{{{!}}
{{!}}1-1{{!!}}1-2
{{!-}}
{{!}}2-1{{!!}}2-2
{{!}}}

This is marked as an error - because of "{{{!}}" I suppose. I don't want to turn off this filter, because it can be really helpful, but now it's impossible to use it. There are 52 THOUSANDS of founded mistakes, and a vast majority of them are false positives. Any chance to fix? Thanks. --Facenapalm (talk) 00:32, 5 January 2016 (UTC)

Facenapalm Looks like Russian and Ukrainian are the only wikis that do this. A fix is in place and these "shouldn't" show up in the next run. Bgwhite (talk) 22:34, 5 January 2016 (UTC)

Thank you very much! --Facenapalm (talk) 04:10, 6 January 2016 (UTC)

Hmm. Seems like it not fixed yet. New dump scanned - still 54 thousands posiitives. :( Btw, what XXwiki coordinators do? I can be a ruwiki coordinator. --Facenapalm (talk) 09:48, 21 January 2016 (UTC)

Facenapalm I had an "or" instead of an "and" in an if statement. I'm rerunning ukwiki's last dump to checking things out. A coordinator handles the translation file, any questions that may arise concerning ruwiki and we know who to goto when any major changes need to be made. Tag, you are it. :) Bgwhite (talk) 05:45, 22 January 2016 (UTC)

Oh, ok. :) I'll wait next dump then. I'm watching the updates in enwiki translation page and correct the ruwiki one, and I'm ready to ask the questiions if they arise, so, I can be a ruwiki coordinator, I think. --Facenapalm (talk) 08:48, 22 January 2016 (UTC)

Web interface not showing 107 and 108

 Done

It's this bug again; see the simple.wiki web interface. --95.250.67.231 (talk) 15:18, 7 January 2016 (UTC)

Fixed Bgwhite (talk) 20:19, 7 January 2016 (UTC)

_script variables in translation page

 Resolved

Hi again. When translation page was created, there was two groups of variables for every error: with _script and _ruwiki suffixes. There was request to translate only _ruwiki variables and not to touch _script ones. Now many error filters had changed, a lot of new are created, and _script variables no longer correspond with _ruwiki ones. So, the question is: is it neccessary to still keep _script variables in every translation page? Will something breaks if I'll delete all variables, which ends with _script? As I see, all of them are already deleted in english Wikipedia. --Facenapalm (talk) 14:54, 8 January 2016 (UTC)

Hi Facenapalm, you can safely delete the _script variables I think, the configuration for frwiki has been purged a long time ago and everything is working. --NicoV ^{(Talk on frwiki)} 15:04, 8 January 2016 (UTC)

Thank you. I'll delete them too then. --Facenapalm (talk) 15:12, 8 January 2016 (UTC)

Facenapalm NicoV is correct and they are no longer used. Both the English and French translation pages are upto date thanks to NicoV (well, except where I screw something up). The translation pages are also used by WPCleaner, NicoV's wonderful tool. So, you are updating the translation page for two things. Bgwhite (talk) 21:46, 8 January 2016 (UTC)

plwiki → low priority → interwiki link saved as external link

 Done

Dear, I have been working under the list and figured out that many of those links linked to CD audio covers. Those covers are published under fair use license in enwiki while in plwiki that license is not allowed. Because of that I will be grateful if you add links like https://en.wikipedia.org/wiki/File:Joe_Bonamassa_A_New_Day_Yesterday.jpg to an exception list. Code:

\[.*?en\.wikipedia\.org\/wiki\/File:.*?\]

Best wishes, Patrol110 (talk) 23:05, 25 January 2016 (UTC)

Patrol110 I'm not sure that is possible. I don't know the copyright issues related to externally linking an image that is fair use. My first thought is it's a copyright violation, but I'm not sure. Bgwhite (talk) 00:05, 26 January 2016 (UTC)

Bgwhite Thank you for opinion. I will ask some wiser man than me in plwiki. Patrol110 (talk) 20:19, 26 January 2016 (UTC)

Bgwhite I've discussed the topic [1] and got to know that those links are allowed in plwiki. Only links which imitate being internal links are not allowed (e.g. description). So, you can add below pattern to the exception list. Patrol110 (talk) 19:00, 27 January 2016 (UTC)

Patrol110 I've coded it up. One problem, there are image links to other wiki's besides English. For example, pl:A.T.O.M. Alpha Teens On Machines has a link to https://bs.wikipedia.org/wiki/Datoteka:ATOM_grupo.jpg. I can't check for every wiki's own word for "File" like plwiki's Plik or Grafika. I could add one's for the big wikis, such as German, French or Spanish, but not to every wiki. Another solution would be to change "Datoteka" in the link to "File" or "Image". Bgwhite (talk) 21:22, 27 January 2016 (UTC)

Bgwhite Thanks for your job. I think the problem concered enwiki the most. Other wikis don't have to be taken into account. Patrol110 (talk) 10:02, 13 February 2016 (UTC)

Summary of Changes made recently

Summary of Changes made recently:

• Error 69: Now finds cases of ISBN in a wikilink ( [[ISBN]] 978-12345-6789-0) and # symbol (ISBN #978-12345-6789-0)
• Error 2: Checks for <center/>, <small/> and <br clear
• Error 85: Checks for <center></center> and <gallery></gallery>
• Error 34: Catches more cases. See Instances of 'subst:' in articles

Bgwhite (talk) 06:01, 3 February 2015 (UTC)

Bgwhite, I don't understand the rationale of grouping detection for center and small with #2. The br tag is a special tag in HTML5 (not necessarily XML compliant now), while center and small tags are more conventional tags (XML compliant). Wouldn't it be better to put them in a separate detection? --NicoV ^{(Talk on frwiki)} 10:49, 16 February 2015 (UTC)

NicoV #2 is looking for bad or malformed tags. br and small are both elements, one just has a void end tag... just like hr, img, source, meta and a host of other tags. Wikipedia no long is XML compliant nor tries to be. I really don't want to go into the intricacies of HTML tags... just what is good or bad. Bgwhite (talk) 06:23, 18 February 2015 (UTC)

I understand, it's just that </br> is invalid while </center> is not... ;-) --NicoV ^{(Talk on frwiki)} 16:19, 24 February 2015 (UTC)

Add galician wikipedia?

 Resolved

Hi, could be it possible add galician wikipedia to this tool? Thanks!, Elisardojm (talk) 09:58, 13 February 2015 (UTC)

Hi Elisardojm. Each wiki requires a configuration page, based on Wikipedia:WikiProject Check Wikipedia/Translation. Could you create a similar page on galician wikipedia? --NicoV ^{(Talk on frwiki)} 12:39, 13 February 2015 (UTC)

Hi Elisardojm, I saw that you've started creating the configuration. If you're interested, I've modified WPCleaner for glwiki, it can help you check the configuration. --NicoV ^{(Talk on frwiki)} 20:35, 21 February 2015 (UTC)

Yes :) NicoV, but I'm translating it too slow, I intended to notice it here when I had finished it :). How can I try the WPCleaner? --Elisardojm (talk) 17:00, 22 February 2015 (UTC)

Hi Elisardojm, see Wikipedia:WPCleaner (general info), Wikipedia:WPCleaner/Installation for installation and Wikipedia:WPCleaner/Check wiki for usage with CW: the menus in the Check Wiki window will help you check what you have configured (error labels, error activation, ...). --NicoV ^{(Talk on frwiki)} 22:24, 22 February 2015 (UTC)

Thanks for starting the Galician translation :-) I've included the current translation. --NicoV ^{(Talk on frwiki)} 17:43, 23 February 2015 (UTC)

Possible false positives in Error #47

 Resolved

I think that the template errors (#47 and the like) ignore the characters between math tags. This is good. I think, though, that the formulas between math tags dont' get filtered out if:

• Math tags are capitalized (like <Math>)
• The tag has some attributes, like <math display="inline"> in ca:Gas ideal

Could you please check it out? --Joutbis (talk) 10:59, 22 February 2015 (UTC)

Joutbis Anything between comment, math, nowiki, code, pre, source, hiero and score tags get removed before checks take place. Bgwhite (talk) 05:39, 23 February 2015 (UTC)

Yes, that's fine, but I'm afraid that if the format is <Math> or <math display="inline">, then they are not removed, and the brace counter goes wild. Is this possible? Otherwise, I can't see what's wrong in ca:Gas ideal --Joutbis (talk) 19:50, 28 February 2015 (UTC)

Joutbis Ok, two things going on here.

1. {{equació|1=<math display="block">P = \frac{N \cdot m \cdot \overline{v^2}}{3 \cdot V}</math>|2=3}} is one of the lines causing a #47 error. Checkwiki thinks there is an error because there is only one {{, while there are two }}. Math equations are a common false positive. On enwiki, we have whitelisted multiple articles with the majority being math related.
2. The code is supposed remove anything between the math tags, thus the above line shouldn't be causing a #47 error. It does remove cases including <Math> and <math display="inline">. The lower/upper case does not matter and any parameter inside the math tag does not matter. However, in order to speed up the code, I check to see if there is a math tag in the article first. I was not checking cases of <math display>. As the article only contained <math display>, the checkwiki program "saw" no math tags, thus didn't remove anything between the math tags. Therefore, #47 showed up when it shouldn't have.

In theory, there shouldn't be cases of <math display> in any article, only <math alt> and <math style>. This is especially true when used inside the {{equació}} template, as dispaly=inline is redundant and display=block can be handled by the template. I did edit ca:Gas ideal to remove 'display'. I also edited the CheckWiki program to check for more cases of <math, so it won't matter what is inside the math tags when "seeing" for math tags in the article. Bgwhite (talk) 06:43, 3 March 2015 (UTC)

Wow, thanks! --Joutbis (talk) 17:39, 3 March 2015 (UTC)

Daily scan

 Resolved Moin Moin @Bgwhite:, since yesterday, when SSL was broken, the daily scan isn't running. Can you have a look at? SSL is fixed. King Regards --Crazy1880 (talk) 04:53, 16 September 2015 (UTC)

Crazy1880 SSL only affected the web page and not any scans. dewiki did not run on the 14th or 15th, but did run on the 16th. As it didn't run for two days, the 16th processed three days and took longer. Today's run found ~10,000 errors. When I look at the errors, everyone is already listed and no new errors are listed on the web page. Very strange. Not sure what to say.

I noticed there are alot of #1, #2, #16, #22 and #64 errors. Most of these can be handled via a bot by AWB. Any interest in running a bot? Bgwhite (talk) 06:56, 16 September 2015 (UTC)

Bgwhite, Thanks for your assessments. I had only seen that two days a list was not filled, although there were always mistake expired. Interessiert bin ich immer. Regards --Crazy1880 (talk) 08:04, 16 September 2015 (UTC)

Bgwhite, was it running today? My feeling say no. Regards --Crazy1880 (talk) 04:44, 18 September 2015 (UTC)

Crazy1880 Yes, it ran. I'm running it again on any errors found in articles that were edited in the last 6 hours. Most of the errors are already listed, but I do see new errors showing up for #3. Bgwhite (talk) 06:04, 18 September 2015 (UTC)

Its running well. --Crazy1880 (talk) 18:37, 23 September 2015 (UTC)

Last scanned dumps

Resolved

It's been over 60 days since the last scanned dump for en.wp and 33 days dince the last scanned dump for sv.wp. Time for another? (t) Josve05a (c) 13:03, 4 October 2015 (UTC)

Josve05a enwiki ran on my laptop. Magioladitis and I fixed them all. They are having problems with dumps over the past several months. Only one dump of each language ran last month. Dumps just started up for this month. Bgwhite (talk) 21:17, 4 October 2015 (UTC)

Josve05a I've manually started Checkwiki's svwiki dump processing. They still are having problems. Oh joy. Bgwhite (talk) 00:02, 14 October 2015 (UTC)

Ignore <graph>...</graph>

 Done

Hi, I think we should add <graph>...</graph> in the list of tags which contents is ignored : currently, on frwiki, both fr:Liste des plus longues cavités naturelles and fr:Liste des plus profondes cavités naturelles are detected by #47 (for }} without {{). --NicoV ^{(Talk on frwiki)} 22:00, 13 October 2015 (UTC)

NicoV Done. I just manually started Checkwiki's frwiki dump processing. Entire dump creation process still isn't done because it keeps firing off errors. I had to send an email at the start of the month to ask why processing hadn't started. I've fired off two emails, with no response, on why the dumps aren't being copied to labs. It's been ~six months since they started fiddling with dump processing and it hasn't worked correctly since. Bgwhite (talk) 00:00, 14 October 2015 (UTC)

Suggestion for fixing #91 in case of google translate ?

 Resolved

Hi, would any of you have an idea on how to fix #91 errors when they're about google translate links for automatically translating a page of an other wiki ? For example, in fr:Ameesha Patel, #91 is detected for http://translate.google.fr/translate?hl=fr&sl=en&u=http://en.wikipedia.org/wiki/Rinke_Khanna&prev=/search%3Fq%3Drinke%2Bkhanna%26rlz%3D1C2GGGE_frFR509FR555%26biw%3D1366%26bih%3D642 which is a link to have a version of Rinke Khanna translated into French by google translate.

For information, I've added some parameters in WPCleaner to help fix #90 and #91, see error_090_link_templates_frwiki and error_091_link_templates_frwiki in fr:Projet:Correction syntaxique/Traduction. --NicoV ^{(Talk on frwiki)} 20:22, 24 October 2015 (UTC)

I've decided to deal with this kind of links as if it was a direct link to Wikipedia, so WPC will suggest to replace it by an interwiki link. --NicoV ^{(Talk on frwiki)} 17:17, 16 November 2015 (UTC)

Why positive?

 Resolved

Hello. Could you explain me, please, why ID 43 matches these two articles. I can see this month after month and can't understand what is the problem. Thank you in advance, IKhitron (talk) 18:21, 21 October 2015 (UTC)

IKhitron My right to left reading is horrible, but....

1. With the first one, it has {{{1|}}}}. This is a template variable and normally shouldn't be there.
2. It sees the following and thinks there are three braces together: {שגיאה לא אופיינית לאלוף העולם לשעבר. לאחר המסע 26. צד2 לא נותר לשחור אלא להיכנע}
}}

Bgwhite (talk) 18:35, 21 October 2015 (UTC)

Thanks, Bgwhite, but:

1. I know it should not be here, and ID 34 indeed finds this problem. But what is about "Template without correct end"? It should not be unbalanced parentheses, shouldn't it?
2. Indeed, this template uses "{..text..}" as code it's in hundreds of articles, but the problem is just here. Maybe it's indeed because this text is in the end' just before the "}}". But the bot should not ignore single balanced parentheses?

IKhitron (talk) 18:43, 21 October 2015 (UTC)

IKhitron The problem is Checkwiki counts in pairs. {{{1|}}} isn't a problem because it sees two pair of braces (three braces makes two sets of paired braces). So when Checkwiki sees {...text...}}}, it thinks something is off. There are pluses and minuses to this algorithm and you are seeing one of the minuses. It normally becomes a problem on math articles. Checkwiki does have a whitelist feature. For enwiki, #43's whitelist is here. You tell Checkwiki where to find the whitelist via the translation file. Bgwhite (talk) 19:01, 21 October 2015 (UTC)

Thank you very much, Bgwhite. IKhitron (talk) 12:00, 22 October 2015 (UTC)

Could you tell me, please, Bgwhite, where can I find any manual for the translation page. I'd like to find a list of per error parameters, as "error_003_templates_enwiki". Thank you, IKhitron (talk) 17:09, 23 October 2015 (UTC)

IKhitron Not sure it exists, I've compiled one but I'm not sure it's complete for CW itself and it contains a lot of parameters that are only used by WPCleaner: Wikipedia:WPCleaner/Configuration/Help#Check_Wiki_configuration. --NicoV ^{(Talk on frwiki)} 17:30, 23 October 2015 (UTC)

Thank you, Nicov. It's a (very) good start. IKhitron (talk) 17:34, 23 October 2015 (UTC)

HTML links with { or } : #43 / #47

 Resolved

Hi, on frwiki, we have several articles with HTML links containing { or } (sometimes with double { or }) and they are detected by #47 (and probably #43). Example on fr:Circonscription d'Arta: [http://ekloges-prev.singularlogic.eu/v2012b/public/index.html#{%22cls%22:%22level%22,%22params%22:{%22level%22:%22ep%22,%22id%22:19}} Επικράτεια - Εκλ. Περιφέρεια Αρτας]. Should it really be detected by CW ? If yes, how can I fix them (without adding all the articles in the white list) ? --NicoV ^{(Talk on frwiki)} 11:56, 11 November 2015 (UTC)

NicoV this Greek link is a mess. -- Magioladitis (talk) 12:02, 11 November 2015 (UTC)

Magioladitis Yes, I know, but this entire web site seems to pass some sort of JSON parameters in the URL, hence the { and }. I don't know how to rewrite the URL so that it works but isn't detected by CW. --NicoV ^{(Talk on frwiki)} 12:07, 11 November 2015 (UTC)

NicoV I have replaced many of them with elections results from a another site. -- Magioladitis (talk) 12:09, 11 November 2015 (UTC)

Ok, I found a way by replacing { by %7B and } by %7D. --NicoV ^{(Talk on frwiki)} 12:27, 11 November 2015 (UTC)

I've modified WPC to suggest the replacement by %7B and %7D when the error #47 is triggered by an URL. --NicoV ^{(Talk on frwiki)} 20:37, 12 November 2015 (UTC)

Ask for check

 Resolved

Hello. Is the cite error phab:T118391 checked? If yes, which ID? If no, could you add this, please? Thank you, IKhitron (talk) 14:12, 11 November 2015 (UTC)

The regex could be something as </?ref(^>)*$ IKhitron (talk) 14:57, 11 November 2015 (UTC)

IKhitron, yes this is checked for. It shows up in #94. AWB does fix this automatically, not sure about WPCleaner.

WPC sees it also as #94, the opening ref tag has no counterpart. --NicoV ^{(Talk on frwiki)} 04:06, 13 November 2015 (UTC)

Thank you. IKhitron (talk) 12:28, 15 November 2015 (UTC)

Can't use whitelist

 Resolved

Hello. I tried to add whitelist as explained, but it does not work. Coud you tell me, please, what's wrong? id 95 results, translation page, whitelist. Thank you in advance, IKhitron (talk) 12:34, 15 November 2015 (UTC)

IKhitron I think it is a right-left issue. I'm looking for a line to start with * [[. I haven't run into this before, so I'm not sure what to do right now. Bgwhite (talk) 06:05, 16 November 2015 (UTC)

Bgwhite, but two another whitelists I created work fine. I can't understand what's the problem with that one. IKhitron (talk) 11:49, 16 November 2015 (UTC)

IKhitron Check the name of the parameter ;-) error_034_whitelistpage_hewiki=Wikipedia:Check_Wikipedia/Error_095_whitelist --NicoV ^{(Talk on frwiki)} 15:16, 16 November 2015 (UTC)

IKhitron (talk) 16:32, 16 November 2015 (UTC)

Unbalanced attribute quote signs

 Done

See this revision, which contains: <ref name="reuterstimeline /> . Seems like something that might be in the purview of this project ? Quote signs for parser tags aren't mandatory right now in wikitext, but unbalanced is definitely problematic long term. —TheDJ (talk • contribs) 09:47, 16 November 2015 (UTC)

TheDJ you are the second person in as many days to request this. See here. It is actually causing a problem with VE, which is how the first request came to be. I'll be adding it at some point. Bgwhite (talk) 00:14, 17 November 2015 (UTC)

Ah, there was no note of that at the original VE report. —TheDJ (talk • contribs) 07:23, 17 November 2015 (UTC)

TheDJ They weren't going to fix it until your ticket (T118715), which is why it was originally requested that CheckWiki detect it. Magioladitis, NicoV and I have added 40+ tickets between us and do we have one of our issues fixed? You submit an issue that wasn't going to be fixed and its fixed within 24 hours. You are a God. Can you submit our tickets from now on? I'm going to go and bow down to the great and mighty TheDJ. Bgwhite (talk) 07:58, 17 November 2015 (UTC)

TheDJ I'm impressed too, I often even have no answer at all to the issues I'm submitting (a good example of that is about Content Translation tool : my reports are basically ignored and this tool continues to create articles with problems in 80-100% of them...). --NicoV ^{(Talk on frwiki)} 08:36, 17 November 2015 (UTC)

This one was easy for them, as it was basically a missed edge case in previous work they had done a couple of weeks ago to improve attribute parsing in references. Besides, nowhere was said that they weren't going to fix it. User:Whatamidoing (WMF) had not made a ticket for it, but that is not uncommon. I often delay creating a ticket in order to get a better grip on the stated problem, so that I can write the best, least confusing, ticket possible. Or find a pre existing ticket that matches it.

And I have some 160 tickets authored tickets currently open, 200 tickets that I 'watch' are open, and that's besides the stuff that I myself maintain, because no one else does, which is another couple of 100 open tickets. so i'm not sure if you want to take me as a benchmark :) It's all relative..... —TheDJ (talk • contribs) 09:01, 17 November 2015 (UTC)

Finding pre-existing tickets is harder in Phab than in Bugzilla. These days, I often don't write a ticket until I've personally contacted one of the devs about it, to see whether someone else believes that a ticket about it already exists. It looks like about 20% of my closed Phab tasks (not all of which are bugs) are explicitly marked as duplicates, which isn't as bad as I thought. Whatamidoing (WMF) (talk) 18:53, 17 November 2015 (UTC)

@TheDJ, NicoV, GoingBatty, XXN, Matěj Suchánek, Meno25, Josve05a, and Magioladitis: Error #104 has been added for the unbalanced quotes in <ref> tags. Error #105 will be added... it is the same as error #008, but with crap at the beginning of the section header instead of at the end. Bgwhite (talk) 00:50, 2 December 2015 (UTC)

Error #92 and special tags

 Resolved

CheckWiki detects headlines with <code>...</code> tags like

== <code>Whatever</code> ==

as

==  ==

and when there are more such headlines in one article which are not same, they are reported as same. Is there any possibility to adjust the algorithm, or should those just be put on whitelists? By the way, WPCleaner does not find this error in those articles. Matěj Suchánek (talk) 12:25, 17 November 2015 (UTC)

Matěj Suchánek The program removes anything inside <code>, <nowiki>, <source>, <math> and a few other tags. The odds of broken brackets and other things are high in these tags. A whitelisting you shall go. Bgwhite (talk) 23:05, 19 November 2015 (UTC)

Add linebreaks

 Done

I raised this at WP:BOTR, but someone sensibly suggested that I come here instead.

Perhaps someone's running a bot that already does this, but I thought I'd bring it up anyway, in case nobody was.

When text precedes a header, the header doesn't work, and the coding appears as normal text; run a Ctrl+F search for the equals sign at [2]. Fixing it is easy, because you just have to add a couple of new lines. If this isn't already being done, could someone's wikisyntax-fixing bot be given this as an additional task? Nyttend (talk) 15:31, 22 November 2015 (UTC)

Nyttend It is and isn't the right place. Before it can fixed, one needs to find the articles with the problem. There is already CheckWiki error 8 that finds cases with crap after a section header. You want to find crap before the section header. This is not only reasonable, but just an extension of error 8. I'll code it up, but I'm busy at the moment and I also have another error I need to code up.

I'm not sure if a bot can do this without seeing more articles with the error. For error 8, it is usually one of two cases: vandalism or somebody didn't press enter. Majority of the time it is vandalism. A bot can't decided between deleting vandalism or just adding a line break. Bgwhite (talk) 23:07, 24 November 2015 (UTC)

@Nyttend, NicoV, and Magioladitis: It has been added to Checkwiki. It is error #105. There are ~1,400 articles with errors in the last dump. A list of articles with problems is found at User:Bgwhite/Sandbox. There are going to be alot of false-positives. Many false-positives are going to have refs in the section heading and they used a one or multiple "new lines" in the section heading. Bgwhite (talk) 22:44, 2 December 2015 (UTC)

@Bgwhite: Thanks, I'm seeing that it detects some VE bugs... (example). --NicoV ^{(Talk on frwiki)} 20:13, 3 December 2015 (UTC)

NicoV It's also picking up VE's bug where VE adds {{anchor}} right before the section heading. There isn't a parameter in the anchor tag either. I'm thinking about not detecting it as a #105 bug if the ending is </ref>==. This will remove many false positives. Bgwhite (talk) 21:40, 3 December 2015 (UTC)

Bgwhite Yes, that would be nice to limit the number of false positives due to multiline references... I've modified my development version of WPC to deal correctly with multiline references in titles, so I'm not detecting them anymore. --NicoV ^{(Talk on frwiki)} 04:29, 5 December 2015 (UTC)

Improvement for #64

 Done

Hi, a user asked me to fix also links like [[Article|'''Article''']]. I've just added that to WPCleaner (bold and italics), maybe it could be nice to have it also in CW ? --NicoV ^{(Talk on frwiki)} 10:44, 25 July 2015 (UTC)

NicoV. New code has been updated. It also catches cases of [[Article|'''''Article''''']] or any number of apostrophe's as long as there is a minimum of two on both sides of Article. The French dump is currently processing with the new code. Bgwhite (talk) 21:36, 12 August 2015 (UTC)

update arwiki

 Resolved

Please update the arwiki Last scanned dump 2015-06-02 (70 days old).

• Last dump: 2015-06-02
• Last update: 2015-06-30

--Zaher ^talk 10:21, 11 August 2015 (UTC)

Hi Zaher, I think the problem is with the generation of the dumps, WMF hasn't been able to properly generate them for a long time... --NicoV ^{(Talk on frwiki)} 11:53, 11 August 2015 (UTC)

The new arwiki dump is available now, but if you search for "checkwiki" in the Tool Labs Grid Status you'll find that arwiki-munch and arwiki-delay are stuck on the previous dump, wasting server time for weeks. Same happened with eswiki, but in this case the new dump is still partial: it should take few more days to complete hopefully, so I suggest to wait before restarting eswiki-munch. --Vittorioo (talk) 06:58, 12 August 2015 (UTC)

Zaher kadour I deleted all the old jobs. They were all hung. Some of them were on disabled queues. They are rebooting all the machines over the next few days to upgrade the kernel, so they may get hung again. Sometimes a machine will go down, but the job remains listed with the master queue, thus it looks hung.

FYI... A dump isn't copied over to the WMF Labs side until it is entirely finished. Bgwhite (talk) 07:52, 12 August 2015 (UTC)

@Zaher kadour and Meno25: Looks like arwiki and eswiki process got hung again. A couple of times with enwiki, it wouldn't process either. It turned out to be something weird in an article causing the problem. I have a feeling this is also happening here. I can usually narrow it down to a few suspect articles. It will take a bit to narrow down as I have to run checkwiki a few times. I'll be doing this on my laptop where I can control it better. I'll need help once I narrow it down as I haven't a clue when it comes to Arabic. Bgwhite (talk) 17:22, 13 August 2015 (UTC)

@Zaher kadour and Meno25: It ran ok on my laptop. I'm at a loss on what is happening. I updated the errors from the data generated on my laptop. Bgwhite (talk) 01:01, 14 August 2015 (UTC)

I ran the bot to fix all errors on arwiki. --Meno25 (talk) 11:19, 14 August 2015 (UTC)

Vada?

Seems that the tool Vada would be useful to this project, particularly the A930913's Cleaner app. Eman235/talk 18:01, 14 August 2015 (UTC)

checkwiki error #69 (isbn) updated

Checkwiki will give off a #69 error if:

1. Finds 10-ISBN or 13-ISBN
2. Finds ISBN\d[-\d ][-\d] ie, ISBN9876543210. There are infobox parameters of isbn1 .. isbn14, so the regex tries to work around this.

From August's enwiki dump, there were 2,000 articles with the first problem and ~250 with the second. Bgwhite (talk) 09:29, 27 August 2015 (UTC)

Enable in lvwiki

 Done

@Bgwhite: Can you please enable CHECKWIKI to lv.wiki? -- Magioladitis (talk) 20:51, 11 September 2015 (UTC)

Magioladitis Yes, I'll add it. Now I need to remember how.... Bgwhite (talk) 21:30, 11 September 2015 (UTC)

Magioladitis In theory, checkwiki is processing lvwiki's dump right now. Bgwhite (talk) 23:46, 11 September 2015 (UTC)

Magioladitis Everything ran ok. Some errors are very high, but that is due to needing to tune the errors.

1. 3 is high, but I lvwiki uses there own template. Either need to remove #3 or add the template to their own translation page.
2. 37 & #6 are high. What is lvwiki's stand on what can go into defaultsort? Some wiki have various characters with umlauts or accents that are ok.
3. 61 is high, but relatively low. Is lvwiki at punctuation before or after the ref wiki? The error is relatively low in I'd think there would be more errors.

Bgwhite (talk) 06:02, 12 September 2015 (UTC)

Bgwhite I asked the Latvian guys to translate everything and disactivate what they do not like it. We need to add instructions of how things are done for other projects that are interested to participate. -- Magioladitis (talk) 06:09, 12 September 2015 (UTC)

Added lvwiki to WPCleaner also after request from Edgars2007. --NicoV ^{(Talk on frwiki)} 19:31, 29 October 2015 (UTC)

+ScoWp

@Bgwhite: Is it possible to enable Check Wikipedia on ScoWp, as well? ^Avic_ennasis @ 01:50, 11 Tishrei 5776 / 01:50, 24 September 2015 (UTC)

Avicennasis Och aye! The defaults are going to be enwiki defaults. Enwiki config file is located Here. Copy this file somewhere on scowiki. Turn off/on any errors you need and do any necessary translating. When done, tell me where you put the file. Bgwhite (talk) 22:51, 24 September 2015 (UTC)

AWB and CHECKWIKI

I think it would be good to put a link to this page somewhere on http://tools.wmflabs.org/checkwiki/ and also somewhere on WP:AWB. Personally I searched some (long) time such a page in AWB manual and on labs project pages, but I found it only now. Especially this can be useful for users with non-en.wp home wiki. --XXN, 19:35, 25 September 2015 (UTC)

Empty anchor tags - what to do?

 Resolved

See [3] (t) Josve05a (c) 11:46, 14 October 2015 (UTC)

Josve05a These are ok in this instance. It can be replaced with the {{anchor}} template. I see the same thing being used for notes or references, which I don't like. People will use <div id=> tags instead of <span id=> tags, which is incorrect on an HTML level and should be replaced with <span> or {{anchor}}. I recently ran a list of all empty span tag, minus those being used an anchor. They were located User:Bgwhite/Sandbox1. Frietjes just finished fixing them... see the history. Bgwhite (talk) 17:01, 14 October 2015 (UTC)

Missing last run

 Done

Hello. There are about 10 days from the last dump on our wiki, but there wasn't still new run. And not just there. Something's wrong? Thank you, IKhitron (talk) 12:16, 19 October 2015 (UTC)

You mean hewiki, right? Link to dumps. --Meno25 (talk) 12:29, 19 October 2015 (UTC)

Exactly. Another link to dumps, 2015-10-10. IKhitron (talk) 12:36, 19 October 2015 (UTC)

Probably because dumps are not correctly copied to labs (see Bgwhite answer 2 subjects above). --NicoV ^{(Talk on frwiki)} 13:46, 19 October 2015 (UTC)

Thanks. I can see from that answer that there should be manual start. IKhitron (talk) 14:23, 19 October 2015 (UTC)

IKhitron What one are you after? Bgwhite (talk) 21:02, 19 October 2015 (UTC)

Sorry, Bgwhite, i did not understand your question. IKhitron (talk) 12:27, 20 October 2015 (UTC)

IKhitron What language are you needing. Bgwhite (talk) 16:58, 20 October 2015 (UTC)

Do you mean which wiki? Hewiki, Bgwhite. Thank you. IKhitron (talk) 17:00, 20 October 2015 (UTC)

IKhitron It's running now. Bgwhite (talk) 21:13, 20 October 2015 (UTC)

It's done. Thanks a lot, Bgwhite. IKhitron (talk) 12:11, 21 October 2015 (UTC)

Hello, Bgwhite, it's me again. I believe it happens again. There are hundreds of ready dumps and 7 checks only, with no change in this number in the last three days. Thank you, IKhitron (talk) 12:43, 10 November 2015 (UTC)

IKhitron All the runs should be starting up in the next day or two. The dumps are almost finished. Bgwhite (talk) 09:38, 11 November 2015 (UTC)

Tיhank you, Bgwhite. IKhitron (talk) 10:49, 11 November 2015 (UTC)

IKhitron Checkwiki is starting too process the dumps. A few dumps haven't been copied over to labs yet, but most have. 23:51, 12 November 2015 (UTC)

Empty projects

Looking at the main table, there are some projects with zeros: dewikisource and svwikisource have niether project nor translation pages, commonswiki and frwikiversity (@NicoV:) do have them but no errors are reported. Maybe we could decide what to do with them.

By the way, do you, Bgwhite, have any possibility to quickly check if some others project have many CSW errors. I would be interested in cswikisource, cswikiquote etc. Thanks. Matěj Suchánek (talk) 09:56, 24 December 2015 (UTC)

Matěj Suchánek I never 100% ported over commonswiki from toolserver. Josve05a asked me to add svwikisource The three projects (minus commons) probably show 0 because an article is causing problems with Checkwiki. Currently have that problem with dewiki and its something on dewiki's end. I can hunt down the error and get them working. However, nobody has complained they are down.

It is fairly easy to check other projects. I can run it on my laptop and give you how many errors were detected and log files for which error. If it something you want to make permanent, I can add it to the regular processing. Bgwhite (talk) 19:09, 24 December 2015 (UTC)

@Bgwhite: Sv.wp has 0 errors, either someone has done a great job while I've been inactive, or something has broken...I'm guessing the last option. (t) Josve05a (c) 16:42, 8 January 2016 (UTC)

Josve05a You forgot another and mostly likely cause, I screwed up somewhere. I'm downloading svwiki's dump file now and I'm looking into it. Bgwhite (talk) 21:52, 8 January 2016 (UTC)

Bgwhite, its still empty for SVWP although i think the time from latest dump has changed. /Hangsna (talk) 15:07, 22 January 2016 (UTC)

False positives in new error 105

This new suggestion is catching some interesting mistakes, but, unfortunately, it is also flagging some false positives, when the heading has a reference with a multi-line template inside. Something like:

=== header <ref>{{cite web|url=blabla
|title = http://bla.bla
|date = today}} ===

which gives perfectly valid pages. Could you please modify your script so this is not flagged? --Joutbis (talk) 23:31, 6 December 2015 (UTC)

Joutbis Already saw the problem and mentioned above at #Add linebreaks. Bgwhite (talk) 05:21, 7 December 2015 (UTC)

You should not close the reference tag? IKhitron (talk) 11:14, 7 December 2015 (UTC)

True, that's error #94. :-) --Joutbis (talk) 17:04, 7 December 2015 (UTC)

Dumps everywhere

Hello. I'm just interested. There was a dump on November 27, and another one on December 2. This means the new one will be from now once a week instead of once a month? IKhitron (talk) 12:17, 3 December 2015 (UTC)

IKhitron The dumps have been messed up for over 6 months. They were doing dumps twice a month except for enwiki. However, it's been sporadic ever since they started making "improvements" The dumps labeled November 27 were started late. It should have started around the 16th. The ones labeled December 2 didn't complete all the way, that's why they are early. In theory, they are still only doing dumps twice a month, except for enwiki. Bgwhite (talk) 21:46, 3 December 2015 (UTC)

Thank you. IKhitron (talk) 21:49, 3 December 2015 (UTC)

ISBN-check

It would be very helpful if the check could recognize and ignore

• ISBNistFormalFalsch=J

Example: de:Erich Burgener - {{Literatur | Autor=Bertrand Zimmermann | Titel=Erich Burgener | Verlag= Editions de la Thèle| Ort=Yverdon-les-Bains | Jahr=1987 | ISBN=2-8283-0024 | ISBNistFormalFalsch=J }}

• http://xxxxx/isbn/282830024

--Tsor (talk) 09:09, 2 March 2014 (UTC)

Tsor, as usual, I'm confused. Why give a bad ISBN in the first place? I did a Google search and only two non-Wikipedia derived websites give this number and one of them is Wikipedia. Bgwhite (talk) 23:43, 2 March 2014 (UTC)

Hello Bgwhite, this ist just a (bad) example. Sometimes we find in a book an ISBN which is formal wrong. Some guys use the template Vorlage:Literatur where they can mark such invalid ISBNs by "ISBNistFormalFalsch=J". There is another template Vorlage:Falsche ISBN which can mark such invalid ISBNs: {{Falsche ISBN|3-123-45678-9}} leads to "ISBN 3-123-45678-9 (formal falsche ISBN)". This template is used very often: https://de.wikipedia.org/wiki/Spezial:Linkliste/Vorlage:Falsche_ISBN

I will look for a better example for an invalid ISBN. --Tsor (talk) 10:10, 3 March 2014 (UTC)

PS: An additional column in the error-list "marked as invalid" would help. --Tsor (talk) 10:18, 3 March 2014 (UTC)

Tsor, I'm slow, but I still fail to see what is wrong. It would be best to use a correct ISBN? A better example would help me understand. TMg, could you help me out.

There are whitelists in which articles can be added so they won't be raised as an error again. To many things can go wrong with "marked as invalid" button... Already a problem of vandalism by people clicking done when they have no intention of fixing errors. Bgwhite (talk) 3 March 2014 (UTC)

Here are 349 examples. --Tsor (talk) 11:10, 3 March 2014 (UTC)

I just looked at the first one in the list, de:Charles de Melun and I don't understand why the ISBN is qualified as bad: the checksum is correct. Is it normal to have "ISBNistFormalFalsch=J" with an ISBN that seems correct? Edit: idem for second example de:Bussard (Einheit). --NicoV ^{(Talk on frwiki)} 12:26, 3 March 2014 (UTC)

Hmm, you are right, in de:Charles de Melun ISBN is marked as bad but ist is ok. Same at your second example. I will have a closer look. --Tsor (talk) 13:26, 3 March 2014 (UTC)

Please repeat your calculation. The checksum digit is false, if the first 9 digits are corect the checksum digit in the end should be a 1, so the ISBN should be 2902091311 and not 2902091312. --Cepheiden (talk) 19:15, 5 March 2014 (UTC)

Well, you're just not looking at the version as was looking at, the page was modified since my comment and changed completely about the ISBN: a ISBN-13 with a coherent checksum was replaced by a ISBN-10 with a non-coherent checksum. --NicoV ^{(Talk on frwiki)} 20:21, 5 March 2014 (UTC)

I'm sorry, you are right i didn't notice the edit. --Cepheiden (talk) 17:48, 8 March 2014 (UTC)

I also looked at other, a lot seem in the same situation. There's also cases where the ISBN has indeed a wrong checksum, but the book can be found with the correct ISBN on the internet: de:Mare Imbrium and the corresponding book on google. I've spent quite some time on frwiki to fix ISBN reported by CW (still quite some work to do), but I've found very few situations where the ISBN with the incorrect checksum was confirmed as being the ISBN (it's usually fixed at some point). --NicoV ^{(Talk on frwiki)} 15:51, 3 March 2014 (UTC)

Yes, there are cases of ISBN's with false checksum digits used as the original ISBN (printed in book and listed in databases of libraries etc.). If someone cites this book with this ISBN we mark them as "formally false" like some libraries do. So what's the point here? --Cepheiden (talk) 19:15, 5 March 2014 (UTC)

My point was that I was surprised by the size of the list (349 pages), because as I said, I fixed a lot of ISBN on frwiki, and didn't find so much situations where the ISBN with the non-coherent checksum had to be kept. Given that the first hits in the search seemed to be mistakes, I was wondering if it was normal that you have so many page with ISBN tagged as formally false. --NicoV ^{(Talk on frwiki)} 20:26, 5 March 2014 (UTC)

This was more a reply to Bgwhite (like Tsor already did). --Cepheiden (talk) 17:48, 8 March 2014 (UTC)

Just an example for the second point: http://www.randomhouse.ca/catalog/display.pperl?isbn=9780676978223 found in de:28 Stories über Aids in Afrika. --Tsor (talk) 22:08, 3 March 2014 (UTC)

It links to "Page not found", the correct link seems to be at http://www.randomhouse.ca/catalog/display.pperl?isbn=9780676978230 (different last 2 digits ISBN). --NicoV ^{(Talk on frwiki)} 22:29, 3 March 2014 (UTC)

#69 additions

@Magioladitis and NicoV: Checkwiki *should* find cases of ISBN Pound-sign.... ISBN # Bgwhite (talk) 23:26, 28 January 2015 (UTC)

@Magioladitis and NicoV: {{Infobox comics character and title}} contains ISBN# as a parameter name. I need to put a fix in to avoid this, don't know about your programs. Bgwhite (talk) 05:18, 30 January 2015 (UTC)

ISBN# is not a valid parameter name. ISBN1, ISBN2, etc. are. ISBN# need to be replaced with ISBN1 is not empty, otherwise removed. -- Magioladitis (talk)

@Bgwhite and NicoV: I removed any instances of ISBN# from the Infobox and all other similar infoboxes. -- Magioladitis (talk) 23:12, 30 January 2015 (UTC)

@Magioladitis and NicoV: Checkwiki *should* find cases of [[ISBN]] now.... [[ISBN]] 978-3948-3838-33, [[ISBN]]: 978-3949-3838-33, etc... Bgwhite (talk) 21:10, 2 February 2015 (UTC)

Not everything of the cases above will be fixed by AWB. I am afraid of false positives. I do knot know whether Rjwilmsi could help us here. -- Magioladitis (talk) 09:07, 3 February 2015 (UTC)

Category:Articles with links needing disambiguation from June 2011 is getting close

Category:Articles with links needing disambiguation from June 2011, containing the oldest dated links tagged as needing disambiguation, is now under a thousand. I am sure that with some teamwork, we can wipe it out this month. Cheers! bd2412 T 00:31, 1 February 2015 (UTC)

Less than 600 pages! -- Magioladitis (talk) 16:26, 1 August 2015 (UTC)

CHECKWIKI #81

I know that #81 was turned off on enwp due to the was amount of these errors, but is it possible to turn it on, even if only for one database scan or something, for me? (t) Josve05a (c) 04:55, 3 March 2015 (UTC)

Josve05a Yup, I can run it. The next enwiki dump should be out by the end of the week. I'll run it, which is when I run the regular dump scan. The big problem will be me remembering. Bgwhite (talk) 05:32, 3 March 2015 (UTC)

(t) Josve05a (c) 05:34, 3 March 2015 (UTC)

Josve05a The list is at User:Bgwhite/Sandbox. It only contains the first 49,000 articles. The entire list (89,000) was too big to save. Bgwhite (talk) 17:13, 18 March 2015 (UTC)

Template programming element

I don't understand why this is classified as an error – I have never seen any rule that parser functions are restricted to templates, not, for instance in Help:Magic words. Is that (another) unwritten Law of Wiki? --Unbuttered parsnip (talk) mytime= Mon 08:56, wikitime= 00:56, 16 March 2015 (UTC)

Unbuttered Parsnip, an example would be good. Bgwhite (talk) 04:59, 16 March 2015 (UTC)

List of errors > #500

As you probably know, WPCleaner can detect some errors that are not listed by Check Wiki, using error numbers > #500, without any link to a list of pages to fix.

I've modified WPCleaner to be able to manage a list for some of these errors:

• #518 (nowiki tags) can use an existing abuse filter to get a list of pages where nowiki tags were added recently (requires adding a parameter to Check Wiki configuration)
• #520 (weird characters like snowmen) can use an existing abuse filter to get a list of pages where weird characters were added recently (requires adding a parameter to Check Wiki configuration)
• #524 (Template calls with duplicate arguments) can use the category defined in MediaWiki:Duplicate-args-category to get a list of pages with template calls with duplicate arguments
• #526 (Invalid link on a year) can use an existing abuse filter to get a list of pages where links like [[1980|1985]] were created recently (requires adding a parameter to Check Wiki configuration)

If you know some way of getting a list of pages for other errors > #500, I can add it to WPCleaner. --NicoV ^{(Talk on frwiki)} 19:22, 30 March 2015 (UTC)

The abuse filter extraneous markup is one. Checkwiki will catch some of these, such as this, but won't others. Bgwhite (talk) 07:05, 31 March 2015 (UTC)

Error #81 - what to do?

Is error 81 worth keeping around? I thought it was but upon being reverted Bgwhite says "it's deactivated for a reason" due to it being one of those errors that's very frequent and I don't disagree, only #70 or #72 get close to the level of backlog and while I've long seen them as prime examples of errors that'll never be fixed they're not the issue here. This dump of articles from March that Bgwhite provided shows it is a very big issue, something I wasn't aware of prior but that's my fault for not asking about it here. I'm going to assume the length is just cumulative from when it was deactivated until now but it was frequent at the time it was still active. I hear that AWB has an issue with parsing it automatically so it wouldn't be any good as a bot task and it's hard to see even a dedicated group of people going through the backlog even semi-manually. From what Bgwhite has said, this is a perennial issue and the rate of new entries is much higher than the fixing rate so I think that with the error list at a few hundred shy of 90,000 would it be wise to simply not have it as an error? Looking for consensus or simply input as to what we could do because I'm not convinced it's insurmountable but the tools we have to process them quickly have too high a margin of error. tutterMouse (talk) 09:22, 16 April 2015 (UTC)

tutterMouse, as Rjwilmsi may recall too, we had editors complaining of us merging references. AWB will merge references only if there is a merged reference in a given page. -- Magioladitis (talk) 09:43, 16 April 2015 (UTC)

I'm a little confused, was it that there were complaints of either the fact AWB/WPC was merging references at all or that AWB only merged them when there were other merged references? tutterMouse (talk) 10:02, 16 April 2015 (UTC)

On frwiki, it's still active, but I'm not sure detections are correct because the list contains only 74 instances right now, which is clearly not the reality. WPCleaner can detect them in articles if the error is activated or if parameter error_081_bot_enwiki=true is added in the configuration (like #82 on frwiki) --NicoV ^{(Talk on frwiki)} 09:52, 16 April 2015 (UTC)

tutterMouse not everybody likes merged references. We found a workaround to merge only if merged references exist in page. Check Wikipedia:AutoWikiBrowser/General_fixes#Duplicate_Unnamed_References_.28DuplicateUnnamedReferences.29 for more. -- Magioladitis (talk) 11:45, 16 April 2015 (UTC)

I see so why do we have this if we've got editors complaining about it? I know we have some old hands who tend to complain about everything if it's not catering to them and oppose everything new but if it's only merging refs where merged refs already exist then why have this error if the solution is an error to someone else? It's made redundant by being hobbled. tutterMouse (talk) 13:04, 16 April 2015 (UTC)

Reason is simple: complaints usually come after things being implemented, rather than before, and when the complaints come, we change configuration or behavior... And enwiki is not the only wiki where CW is running, other wikis are not necessarily complaining. --NicoV ^{(Talk on frwiki)} 14:54, 16 April 2015 (UTC)

Oh no, I know the reason and I find it a pretty stupid complaint. I do know CW runs on other wikis too as you did mention the error is active on fr.wiki but en.wiki has a giant backlog no doubt hindered by complainers and we're resistant to change in a way that makes you think we'd be ten years back technologically if some of us were able to keep it that way seeing as the most prolific editors only consider their use case. I still think it's a worthwhile task but we're here to work out what we do about it, not what we did. tutterMouse (talk) 17:32, 16 April 2015 (UTC)

New error in WPCleaner, #527 : reference tags with same name but different content

Hi, I've just added detection #527 to WPCleaner to detect reference tags with the same name but different content, like <ref name="A">A</ref> and <ref name="A">B</ref>. I've not activated it by default (just need to uncomment the error_527_bot_enwiki line in Wikipedia:WikiProject Check Wikipedia/Translation).

I've also prepared the configuration for #526 which detects incorrect links to years, like [[1985|2005]] (classic VE error..., still not fixed by the development team after so many time...). Same, it's not activated here.

A list of additional detection available in WPCleaner is described at Wikipedia:WPCleaner/Configuration/Help#Check Wiki configuration. --NicoV ^{(Talk on frwiki)} 17:25, 29 October 2015 (UTC)

@NicoV: For those who want to fix this error, on enwiki, Category:Pages with duplicate reference names is the tracking category. Bgwhite (talk) 20:12, 29 October 2015 (UTC)

@Bgwhite: Thanks ! I've included it in WPCleaner's configuration, so the error #527 will look like a regular CW error (with list available).

Do you know how the category name is configured in Wikipedia ? On frwiki, we only have fr:Catégorie:Page avec des erreurs de référence (equivalent to Category:Pages with reference errors) which has all cite errors, not only the same name with different content. --NicoV ^{(Talk on frwiki)} 20:28, 29 October 2015 (UTC)

NicoV, I'm not sure how all this works, but...

1. MediaWiki:Cite error references duplicate key is what labels the error on MediaWiki's side.
2. The default on enwiki is to place ref errors into Category:Pages with incorrect ref formatting.
3. On enwiki Template:Broken ref/cat controls what goes into a specific category or just passes into the default. It lists 3 specific errors not going into default.

The main reason this ref error does not going into the default category is because of bots. A bot works on trying to fix the ref errors in the default category, but it can't fix the duplicate reference name error. Bgwhite (talk) 20:44, 29 October 2015 (UTC)

Ok, thanks for the explanation. I've created a phab:T117099 to request that Cite extension can use specific categories depending on the error. --NicoV ^{(Talk on frwiki)} 20:49, 29 October 2015 (UTC)

Wrong localization

Hi. Romanian description for this page is wrong, unintelligible. Please put default English description instead. --XXN, 23:09, 18 November 2015 (UTC)

XXN The page that controls the descriptions is at ro:Wikipedia:WikiProject Check Wikipedia/Translation. Feel free to put in the correct translation. The web pages are updated at 0 UTC with the contents of the translation file. Bgwhite (talk) 23:08, 19 November 2015 (UTC)

Thank you, I'll do some corrections. --XXN, 00:34, 20 November 2015 (UTC)

Manually-generated tagging

 Resolved

See here, where someone writes ^[when?] instead of placing {{when}}. I've occasionally found the same thing in place of other templates, e.g. ^{[citation needed]} instead of {{fact}}. Do all of you have some way of catching and fixing these? Nyttend (talk) 12:15, 27 January 2016 (UTC)

Nyttend, in January's dumpfile, the only articles with ^[when?] are Drosophila melanogaster, Newville, Ohio and St. Joseph's Morrow Park Catholic Secondary School. Bgwhite (talk) 21:10, 27 January 2016 (UTC)

Is there a way of periodically checking the dumpfiles for superscripted versions of this and other tags? I don't expect that Special:Search would help. Nyttend (talk) 04:04, 28 January 2016 (UTC)

Nyttend, not that I'm aware of. I download the dump every month for CheckWiki and other searches. I don't mind doing searches for you. Just tell me what you need. Bgwhite (talk) 05:31, 28 January 2016 (UTC)

Nyttend: a search like this is quite fast and finds 103 results at the moment. It's also easy to expand. Bye. --CX42 (talk) 09:36, 28 January 2016 (UTC)

I never knew that it was possible to search the source, distinct from searching the rendered text. Very useful! Nyttend (talk) 12:20, 28 January 2016 (UTC)

Parameter named isbn10 : is it possible to prevent its detection as error #69 ?

 Done

@Bgwhite: On frwiki, the template fr:Modèle:Ouvrage defines a parameter named "isbn10" to store the ISBN-10 equivalent to an ISBN-13 (just to keep it, it's not displayed). Is there a way to prevent its detection as error #69 ? --NicoV ^{(Talk on frwiki)} 13:15, 8 February 2016 (UTC)

@NicoV: In theory, it should already be skipping it. The parameter "isbn10" is found in infoboxes. The check for isbn10 in infobox should keep the isbn10 in the template from being caught. I skip \|\s*ISBN(10|13)\s*=. Could you tell me some articles CheckWiki is barfing on? Bgwhite (talk) 20:16, 8 February 2016 (UTC)

@Bgwhite: Oh, is it a recent modification ? I saw this a few weeks ago (I even added this detection to WPCleaner on January 17th before becoming aware of the way this parameter is used in frwiki), but I didn't check if it was still detecting it recently (too busy finishing to fix the ISSN on frwiki). My bad if it's already working like that, I will remove it from WPCleaner. --NicoV ^{(Talk on frwiki)} 21:02, 8 February 2016 (UTC)

NicoV It is not a recent modification. Of course, having the skip in there doesn't mean it is working correctly or I did it correctly. Bgwhite (talk) 02:59, 9 February 2016 (UTC)

Ok, I will report it here if I see an article reported for an isbn10 parameter. --NicoV ^{(Talk on frwiki)} 10:12, 9 February 2016 (UTC)

@Bgwhite: Currently, there are 2 articles reported on frwiki due to an isbn10 parameter, so there are some situations where it's still detected.

• fr:Abbaye Notre-Dame de Boscodon for | isbn10 = | | passage = | lire en lign
• fr:Oblast autonome juif for | isbn10 = 2-86260-882-3 | passage = | l

Is it due to "isbn" being in lowercase (as your regexp only seems to skip "ISBN" in uppercase) ?

--NicoV ^{(Talk on frwiki)} 15:59, 24 February 2016 (UTC)

NicoV Yes, it is a lowercase/uppercase issue. Now fixed. Bgwhite (talk) 23:15, 24 February 2016 (UTC)

@Bgwhite: Apparently, it's still happening fr:Formation territoriale de la France métropolitaine, isbn10 = 2-262-02282-8 }}</ref>.[[Je, 2016-02-28 00:05:12. --NicoV ^{(Talk on frwiki)} 09:25, 28 February 2016 (UTC)

NicoV It would help if I put /i on the right regex. It really should be "fixed" now. Bgwhite (talk) 21:43, 1 March 2016 (UTC)

Template list for error #78?

 Done

Evening. As you know, reference lists can be used as via tag <references />, as via template, for example, {{reflist}}. Seems like script doesn't understand it while scanning error #78. I added a template list in translation page by analogy with error #3 (error_078_templates_ruwiki), but nothing has changed. Situations with two used templates are real (at least on ruwiki) - for example, here (both of Примечание and Примечания templates are used to insert reflist). Can you add parsing of template list in this error too? Thanks. --Facenapalm (talk) 23:53, 21 January 2016 (UTC)

Facenapalm #78 isn't currently set up to read in other templates like #3. I only have it looking for <references>. #78 should do the same as #3, so I'll at it into the code. Bgwhite (talk) 05:55, 22 January 2016 (UTC)

@Facenapalm, NicoV, Edgars2007, Josve05a, Magioladitis, Matěj Suchánek, and Meno25:

Error #78 (two or more <references>, {{reflist}}, ...) now has support for adding templates on the translation page. As there can be more than one template on the page for valid reasons (enwiki... <nowiki>{{reflist|group=}}), you will need to specify the exact parameters used with a }} at the end. For example... reflist}}, reflist|30em}}, reflist|2}}. This should get the vast majority of cases. At the moment, the notice window will show blank. Will fix that in a few days. Bgwhite (talk) 00:47, 29 January 2016 (UTC)

Maybe it's better just to ignore templates with group= parameter? There are 11 reflist templates in ruwiki (I really don't know why), and writing all possible parameters for every template is impossible. I can specify the most common cases, but it's not the best solution. Facenapalm (talk) 13:12, 29 January 2016 (UTC)

Probably similar for frwiki. Suggestion : being able to specify which templates to check for, with an exclusion when some parameters are present with a value : take {{reflist}} into account only if parameter group is missing or empty. --NicoV ^{(Talk on frwiki)} 13:36, 29 January 2016 (UTC)

@Facenapalm and NicoV: I struggled on what to do with this when coding it up. I need it general enough so it works with all languages and easy enough that people can add it to the translation file. I don't have the code set up to specify exclusions. This can get messy... may have five templates each with a different exclusion parameter. This is the first writeup (beta). Thinking was to get this setup with the most common reference templates. Won't get all errors, but most cases. At the moment, I don't have any good ideas on how to handle exclusion parameters. Bgwhite (talk) 21:19, 29 January 2016 (UTC)

Ok, is it possible to allow to write an regular expression instead of the template name? Regexp like reflist(?![^}]*group) should works. Maybe, other wiki's users will have to write other regexp, but it's better than list all possible cases. Facenapalm (talk) 01:10, 30 January 2016 (UTC)

@Facenapalm and NicoV: Sorry, I forgot about this. Yes, it should be possible. I'll need to test if first. Bgwhite (talk) 20:09, 8 February 2016 (UTC)

@Facenapalm and NicoV: It has been coded up and run on today's daily run. Bgwhite (talk) 21:24, 17 February 2016 (UTC)

Sounds good, thank you. Templates in the list are read as regular expressions by default, isn't it? No backward compatibility? :) Facenapalm (talk) 22:00, 17 February 2016 (UTC)

Facenapalm In this case, they are read in like a regular expression. No backward compatibility. Just put everything onto one line with some "|". Man, these ungrateful @#*!( just want it all.   :)  Bgwhite (talk) 00:29, 18 February 2016 (UTC)

@Bgwhite and Facenapalm: WPCleaner should now handle the templates parameter for #78. Not much testing done... --NicoV ^{(Talk on frwiki)} 18:32, 24 February 2016 (UTC)

Generating lists of errors with WPCleaner

I've added the possibility to generate list of errors using MediaWiki dump files directly: WPCleaner tests all articles in the dump file and when it detects an error, it checks if the error is still present in the current version of the article. You can see the result for checking #106 on frwiki using the dump from from 11st of January at fr:User:NicoV/Test#Test 106. It took me 6h, so it's not fast, but it can work also for errors above #500. It may help to improve the detections in both CW and WPC: with the test for #106, WPC detected pages that were missed by CW. --NicoV ^{(Talk on frwiki)} 20:13, 24 January 2016 (UTC)

I've used this for a few errors on frwiki, like #526 (incorrect link to a date). I plan to do this also for enwiki once I managed to use my account on Tools Lab. Any error for which you would like a list ? --NicoV ^{(Talk on frwiki)} 10:58, 25 February 2016 (UTC)

@Bgwhite and Magioladitis: I've generated the first lists for enwiki : Wikipedia:CHECKWIKI/WPC 106 dump, Wikipedia:CHECKWIKI/WPC 107 dump, Wikipedia:CHECKWIKI/WPC 108 dump. I don't think that the scan was full completed, I got an error at some point. A lot of extra work that will show up in Wikipedia:WikiProject Check Wikipedia/ISSN errors‎‎ on the next update. --NicoV ^{(Talk on frwiki)} 19:20, 5 March 2016 (UTC)

Wikipedia:WikiProject Check Wikipedia/ISSN errors‎‎ is updated : a lot of ISSN with incorrect syntax are due to a missing "-" between the 2 groups of 4 digits. I think that running a bot through them would be very helpful (that's what was done on frwiki). --NicoV ^{(Talk on frwiki)} 19:58, 5 March 2016 (UTC)

NicoV I am already doing it. -- Magioladitis (talk) 08:49, 10 March 2016 (UTC)

False positives for #1

 Done

@Bgwhite: Dozens of articles are appearing for #1 on frwiki (list), and for most of them the notice doesn't seem to be remotely related to the #1 error (Template: prefix) : each notice is about a <ref> with a name parameter. --NicoV ^{(Talk on frwiki)} 21:19, 10 March 2016 (UTC)

@NicoV: Everyone one of them are #104 errors. It's fixed and should report as #104 now. Bgwhite (talk) 22:05, 10 March 2016 (UTC)

Ignore tags <ce>...</ce>

 Done

Hi Bgwhite, according to Help:Displaying_a_formula#Chemistry, <ce> is a shortcut for <math chem>. I think we should ignore it the same way we ignore <math>. A user reported on frwiki that several articles are reported for several errors due to this tag. --NicoV ^{(Talk on frwiki)} 08:55, 3 April 2016 (UTC)

NicoV Done. Bgwhite (talk) 21:47, 4 April 2016 (UTC)

Error 16

CHECKWIKI now will detect

• U+2004
• U+2005
• U+2006
• U+2007
• U+2008
• U+007F

too.-- Magioladitis (talk) 05:46, 8 April 2016 (UTC)

 Done WPCleaner. Should be avalaible very soon. --NicoV ^{(Talk on frwiki)} 09:09, 11 April 2016 (UTC)

Pages missed by error #101

Hi, it seems that some cases of #101 (Ordinal number found inside <sup> tags) are currently missed (see Wikipedia:CHECKWIKI/WPC 101 dump):

• when there's a whitespace character in the sup tag (1999–2000 Cuban National Series: 39^th)
• when there's an attribute in the sup tag (Anton Cermak: 33<sup style="font-variant: small-caps; font-weight: bold;">rd</sup>)

--NicoV ^{(Talk on frwiki)} 07:48, 18 April 2016 (UTC)

rev 12006 fixes the ones with whitepace. -- Magioladitis (talk) 07:05, 19 April 2016 (UTC)

CHECKWIKI now detects this case. -- Magioladitis (talk) 21:25, 27 April 2016 (UTC)

Something new?

Hello, Bgwhite. A long time ago you told me there will be an id for noinclude and includeonly tags, as I asked. Is there something new? Thank you, IKhitron (talk) 15:36, 20 April 2016 (UTC)

IKhitron I've forgotten about it. I do need to add it. Also, I'm not ignoring your request just above this. My main laptop's screen died, so it will take a bit. Bgwhite (talk) 18:19, 20 April 2016 (UTC)

Thank you very much, Bgwhite. IKhitron (talk) 18:21, 20 April 2016 (UTC)

@IKhitron, Magioladitis, and NicoV: Error #109 has been added. At the moment, I only have it checking for missing open and close <includeonly> tags. If all goes well, I'll add <onlyinclude> and <noinclude> tags. Bgwhite (talk) 22:11, 14 June 2016 (UTC)

Sorry, Bgwhite, I don't understand. What does this have with the issue we are talking about? IKhitron (talk) 22:45, 14 June 2016 (UTC)

IKhitron This has nothing todo with the #60 error way down below on this page. Bgwhite (talk) 23:03, 14 June 2016 (UTC)

? I meant the issue about tags like includeonly, Bgwhite IKhitron (talk) 23:24, 14 June 2016 (UTC)

IKhitron This is part 1. Error #110 will be the second part. Bgwhite (talk) 04:28, 15 June 2016 (UTC)

Thank you! IKhitron (talk) 11:08, 15 June 2016 (UTC)

IKhitron There was an error in #109 that has been corrected. #110 has been added and the default is off, so you will need to turn it on in your translation file.

Thanks a lot! IKhitron (talk) 22:18, 15 June 2016 (UTC)

I've added #109 to WPC for the 3 tags. --NicoV ^{(Talk on frwiki)} 17:08, 15 June 2016 (UTC)

Errors 98 and 99

Now CHECKWIKI detects sup/sub tags with attributes too. This catches more cases and hopefully it matches WPCleaner's method. -- Magioladitis (talk) 21:50, 27 April 2016 (UTC)

Serious sanitizing bug

 Done

Hi, Bgwhite. There is a problem. If the article name includes double quote, the html connects only to the part before the quotes. For example, if the problem was in some Qwert"y and you're clicking on history in done page, you'll see Qwert. IKhitron (talk) 12:26, 6 May 2016 (UTC)

IKhitron I already sanitize double quotes, along with & < > ' \. What was the article's name you were having problems with? Also, was it with clicking something on the webpage or the name on the webpage was borked? Bgwhite (talk) 21:12, 6 May 2016 (UTC)

I don't sure how long it will be avialable, Bgwhite. It is on hewiki id10 done page. The page (named properly in the list) is he:אריה רלב"ג. If I open the page link or the history link, it tries to find some unexisting 'אריה רלב'. IKhitron (talk) 22:46, 6 May 2016 (UTC)

IKhitron Thank you, this helped me narrow down what program to look at. I think I fixed it. Not exactly sure it's fixed because the article is no longer listed. Bgwhite (talk) 00:48, 7 May 2016 (UTC)

Thank you. I hope you're right. If something like this happens again, I'll be back ;-) IKhitron (talk) 00:59, 7 May 2016 (UTC)

Syntaxhighlight and error #14

 Done

Hello! There're two false positives of 14th error in ruwiki: ru:APML and ru:HTML5 video. In both of these articles the <syntaxhighlight lang=""> tag is contained in <syntaxhighlight> one, but the article still matched as an error. Seems like it's easy to fix by changing the order of function calls in check_article() function. Facenapalm (talk) 23:04, 12 May 2016 (UTC)

Facenapalm Fixed Bgwhite (talk) 05:42, 13 May 2016 (UTC)

Thank you! Facenapalm (talk) 12:22, 13 May 2016 (UTC)

Thousands of false positives on error #34

 Done

Hello. Several month ago I wrote about false positives оn 43th error, which are occures when somebody inserts tables into templates using {{!}} magicword. Seems like we have the same problem with 34th error. @Bgwhite: could you fix it too? Thanks! Facenapalm (talk) 22:48, 14 May 2016 (UTC)

Btw, there are also some false positives on 60th error with similar causes. When somebody needs to insert this code:

{| style="background: transparent"

into template parameter, he may wrote something like that:

|param = {{{!}} style="background: transparent"

and it's matched as an error. I'm not sure, why. The example is here. Facenapalm (talk) 23:03, 14 May 2016 (UTC)

@Bgwhite: I don't want to seem annoying, but could you watch this, please? I think the best way to fix it is just to replace "{{{" with something like "{{{[^}]+}}}" in this regular expression. New regular expression will search unsolved parameters entirely, for example, "{{{1}}}" instead of "{{{". Cases like "{{{1}}" will be found by either 47th or 43rd errors, isn't it? Facenapalm (talk) 21:30, 18 June 2016 (UTC)

Facenapalm I just added a fix. Checkwiki is running on the dump right now. Bgwhite (talk) 23:55, 21 June 2016 (UTC)

False positives for #99

 Resolved

Hi, on frwiki, there are 3 articles reported has having an error for #99, but I can't find it and the notice shows a correct <sup> tag:

• fr:Chaoui : <sup>w</sup>||align="center" | '<span st
• fr:Kabyle : <sup>w</sup>||align="center"|ۥ<span styl
• fr:Somme de Minkowski : <sup>p</sup>.(δ + 1). C'est-à-dire que l

--NicoV ^{(Talk on frwiki)} 05:52, 7 June 2016 (UTC)

NicoV In the first one I think it's because of a tag called super. -- Magioladitis (talk) 06:23, 7 June 2016 (UTC)

The second one also has a <super> tag. – Jonesey95 (talk) 13:02, 7 June 2016 (UTC)

And the third one probably due to strange things like <sup>2</sub>, <sup3</sup>, ...

At least for the third one, I need to modify WPC to detect this...

@Bgwhite: Could CW be modified also to:

• Maybe report the problem is "sup" is followed by something different than a letter ? (to prevent "super" to be detected)
• Display the correct part in the notice ?

Thanks. --NicoV ^{(Talk on frwiki)} 16:44, 7 June 2016 (UTC)

NicoV Done Bgwhite (talk) 21:42, 7 June 2016 (UTC)

Thanks ! I've also modified WPC which was able to detect a #98 in fr:Somme de Minkowski (the closing sub tag in <sup>2</sub>), and when I fixed the tag it detected a #99 (the closing sup tag in <sup3</sup>). --NicoV ^{(Talk on frwiki)} 04:38, 8 June 2016 (UTC)

@Bgwhite: By modifying WPC to detect the problem in fr:Somme de Minkowski, I'm finding other problems that could be included in #98/#99, a standalone closing sub/sup tag, like in fr:Diazote. What do you think ? --NicoV ^{(Talk on frwiki)} 06:02, 8 June 2016 (UTC)

NicoV I'm only checking for open tags. A standalone closing tag doesn't harm anything. I could make the change easily... just replace a > with a != in an if statement. Bgwhite (talk) 07:30, 8 June 2016 (UTC)

@Bgwhite: Thanks for the answer. In principle, a standalone closing doesn't harm anything, except that it may mask an actual error like here : things like <sup3</sup> are missed if standalone closing tags are ok. --NicoV ^{(Talk on frwiki)} 07:57, 8 June 2016 (UTC)

Tags <br \> missed by #2

 Done

Hi, I've started a dump analysis for #2 with WPC, and I'm finding a few cases of <br \>, <br :> not reported by CW:

• Résidu vert
• Loup (constellation)
• Ratatouille
• ...

--NicoV ^{(Talk on frwiki)} 19:19, 9 June 2016 (UTC)

NicoV Done Bgwhite (talk) 23:08, 13 June 2016 (UTC)

False positives for #67 with whitespace between abbreviation and ref tag

 Resolved

Hi @Bgwhite:, it seems that #67 is giving false positives for a lot of pages on frwiki. It may be when there's a whitespace between the abbreviation and the ref tag, like in fr:NGC 236 :

• CW reports the problem as . <ref>On obtient le diamètre d'une gala
• The corresponding line is 88 000 a.l. <ref>On obtient le diamètre d'une galaxie par le produit de la distance qui nous en sépare et de l'angle, exprimé en [[radian]], de sa plus grande dimension.</ref>
• The configuration contains a.l. in the abbreviations.

I tested my theory on fr:NGC 237 :

• CW reported a similar problem on this version
• CW stopped reporting a problem after this modification

--NicoV ^{(Talk on frwiki)} 17:36, 13 June 2016 (UTC)

NicoV The regex for this is

/[ ]{0,2}(\.|,|\?|:|!|;)[ ]{0,2}<ref[ >]/

Tell me how it should be changed. Bgwhite (talk) 23:07, 13 June 2016 (UTC)

@Bgwhite: I don't think this one needs to be changed, rather the other one

for my $temp (@ack) {
    $test_text =~ s/($temp)<ref[ >]//sg;
}

to include possible whitespace characters between ($temp) and <ref. --NicoV ^{(Talk on frwiki)} 06:21, 14 June 2016 (UTC)

@Bgwhite: Do you think it's possible to include optional whitespace characters between ($temp) and <ref ? --NicoV ^{(Talk on frwiki)} 17:12, 27 June 2016 (UTC)

NicoV Done. Bgwhite (talk) 20:40, 28 June 2016 (UTC)

Add scowiki?

 Resolved

Would it be possible to add scowiki to the list of supported wikis? --AmaryllisGardener ^talk 19:08, 14 June 2016 (UTC)

AmaryllisGardener Aye. Could you create a translation file for scowiki. The enwiki is found at Wikipedia:WikiProject Check Wikipedia/Translation. I guess in this case your don't have to translate everything into Scottish. Just copy the English file to somewhere on scowiki, turn off/on what errors you desire and then tell me where it is located. Other option is just to use enwiki's translation file as default. Bgwhite (talk) 22:19, 14 June 2016 (UTC)

@Bgwhite: Ok. I'll start translating it and I'll notify you when I finish. (It might take a few days) --AmaryllisGardener ^talk 23:14, 14 June 2016 (UTC)

You can finish it later, every change in the translation page will be shown on the start of the next day. Only location is necessary. Facenapalm (talk) 10:22, 15 June 2016 (UTC)

AmaryllisGardener I see you have started creating the translation file, so I've configured WPCleaner so that it can also be used. --NicoV ^{(Talk on frwiki)} 17:22, 15 June 2016 (UTC)

AmaryllisGardener scowiki has been added and is using your translation file. In theory, a scan of the latest dump should happen tonight. Bgwhite (talk) 19:36, 15 June 2016 (UTC)

Add bewiki

 Resolved

@NicoV: @Bgwhite: Please add be:Вікіпедыя:WikiProject Check Wikipedia to the code. -- Magioladitis (talk) 08:25, 23 June 2016 (UTC)

 Done for WPC, its configuration is available at Удзельнік:NicoV/WikiCleanerConfiguration. --NicoV ^{(Talk on frwiki)} 15:40, 23 June 2016 (UTC)

@Magioladitis: Added. A dumpscan has been fired off. Bgwhite (talk) 21:03, 23 June 2016 (UTC)

Can we change the scan frequency of zhwiki?

I find that the last check is 2015-12-09, but it's too old. May we scan zhwiki more frequently? Thank you! --Kanashimi (talk) 13:21, 29 February 2016 (UTC)

Kanashimi The zhwiki dump file is currently corrupted. I'm unable to make a scan. Bgwhite (talk) 20:47, 29 February 2016 (UTC)

I find the dump page says "Dump complete". Is it also corrupted?--Kanashimi (talk) 01:54, 1 March 2016 (UTC)

Kanashimi "Complete" doesn't say anything about corruption. I get an error when using the zhwiki file. I don't get that with any other dump file. The enwiki file was corrupted this past December. For some reason, I could use it, but most people couldn't. The dumps have been highly unstable the past ~9 months as they have changed things around. There is supposed to be 2 dumps a month for all wiki's except enwiki, but that hasn't been the case for 2016 and 1/2 of 2015. Bgwhite (talk) 09:01, 1 March 2016 (UTC)

Thank you. --Kanashimi (talk) 10:28, 1 March 2016 (UTC)

New run, same dump?

Hi. I saw now many new results for hewiki. Opened some of them, all are already fixed by me in about 10th March. Is there a possibility that there was accidentally a new run on old dump? Thank you, IKhitron (talk) 10:52, 28 March 2016 (UTC)

IKhitron For some reason, about five wikis were reprocessed on March 19, 20 and 21. All five were reprocessed twice on consecutive days. Very strange. There were no new dumps available. No reason for them start up. Bgwhite (talk) 22:21, 28 March 2016 (UTC)

Well, I gave you a problem to think about. :-) IKhitron (talk) 22:42, 28 March 2016 (UTC)

Reflist

Your list of errors includes a missing reflist where there is a ref tag. Why is that still an error since Wikipedia will now automatically list references? For example see this version of a page which contains a reference list but no reflist or references tag. Quasihuman (talk • contribs) 10:18, 30 March 2016 (UTC)

Quasihuman Yes and no. WikiMedia does add a list of references, but it's done at the end of the page... after all the navboxes, external link section, bibliography section and further reading section. It also doesn't add a "Reference" section header nor add it to the Table of Contents. See this. Bgwhite (talk) 19:01, 30 March 2016 (UTC)

And if it's just after bulleted sections, as bibliography, it's bulleted itself, in place of numbering list. IKhitron (talk) 20:18, 30 March 2016 (UTC)

Error 104

Error 104 will now catch cases where the ending quote marks OR beginning quote marks are missing. -- Magioladitis (talk) 05:37, 6 May 2016 (UTC)

Links to mobile version and #90/#91

 Done

Hi @Bgwhite:, I'm starting to work on #91 in WPC, and I see that all links to mobile version of Wikipedia are reported as #91 errors even when they are linking to the same edition of Wikipedia (for example, links to fr.m.wikipedia.org are detected as #91 for frwiki). I think they should be detected as #90 instead (that's what I've code in WPC). What do you think ? -- NicoV ^{(Talk on frwiki)} 16:26, 29 June 2016 (UTC)

@NicoV and Magioladitis: It's coded up. I want to double check it with tonight's run. If everything is ok, will upload it tomorrow. Bgwhite (talk) 23:32, 25 July 2016 (UTC)

Maybe a bug

 Resolved

Hi. See here. Some ids, as 103, 78 or 50, have 1 to do and 1 done, when there should be 0 done. If it is gone, try here IKhitron (talk) 22:03, 7 July 2016 (UTC)

IKhitron My first guess.... Did you fix these articles in the previous 15 minutes of the screenshot? If so... Program runs every 15 minutes to update what you see on the webpage. Program first updates the "Done" for every error and for every language. Then, the program updates the "To-do" for every error and for every language. There is a gap between these two of 20 to 60 seconds (depends on how slow/fast the computer and database are). So, my guess is you saw the webpage at this gap point. Program runs every 15 minutes at 14, 29, 44 and 59 minutes past the hour. The odd minutes are because everybody was running at 0 and 30 minutes past the hour and bogging down the system. Bgwhite (talk) 22:56, 7 July 2016 (UTC)

This was exactly the case, thank you. IKhitron (talk) 23:08, 7 July 2016 (UTC)

#110 false negatives

 Resolved

Hello. Thank you again for #110. But there are some (maybe a lot of) articles that were not recognized. I know this because AWB list comparer gives 7 pages unique for #109, which should be a subgroup of #110. Could you check it, please? IKhitron (talk) 12:47, 8 July 2016 (UTC)

IKhitron Could you give me some example articles? Bgwhite (talk) 21:31, 22 July 2016 (UTC)

Well, [4], [5], [6] can be good. Thank you, IKhitron (talk) 23:56, 22 July 2016 (UTC)

IKhitron What's going on is... in #110, I only look for opening tags because this means a possible transclusion is happening. The examples you gave only had one closing tag, it's not possible to have a transclusion and therefore not a #110, but a #109. Bgwhite (talk) 22:30, 26 July 2016 (UTC)

I see, Bgwhite. But does this mean that any *include* tag, opening, closing, or self, is recognized by at least one of these ids? IKhitron (talk) 22:59, 26 July 2016 (UTC)

IKhitron Yes, it will be recognized by one of the ids. If there is atleast one opening tag, it's on #110. If there is a closing tag(s), but no opening, it's on #109. If there is a self closing tag, it will be on #110. Bgwhite (talk) 23:34, 26 July 2016 (UTC)

Great. Thank you very much, Bgwhite. IKhitron (talk) 18:47, 27 July 2016 (UTC)

Manual update needed?

 Resolved

@Bgwhite: The English Wikipedia hasn't gotten a new scanned dump in 69 days (1 May). Anything gone wrong in the system? (Also reminder that I would like to see error 81's this run, thanks!) (t) Josve05a (c) 22:02, 9 July 2016 (UTC)

Josve05a English Wikipedia's dump scan is done differently than others. The dump is done manually on my laptop and then uploaded to User:Magioladitis/AWB and CHECKWIKI. Keeping daily/monthly lists separate makes things easier for Magioladitis and I to do things. Magioladitis has been on vacation for a week and should come home today/tomorrow, so I'll update the Magioladitis' Wiki pages tomorrow. Bgwhite (talk) 05:12, 11 July 2016 (UTC)

Whitelist formatting

 Resolved

Hello. How does the whitelist parser works? Parser just tries to find something like /\*\s*\[\[(.*?)\]\]/ and ignores other lines, or not? I want to add some comments to whitelists and I'm not sure which methods will work correctly.

First of all, I want to write comments to some difficult cases. Which methods are guaranteed to work?

* [[Article 1]] — comment 1
* [[Article 2]] — comment 2

* [[Article 1]] <!-- comment 1 -->
* [[Article 2]] <!-- comment 2 -->

* [[Article 1]]
(comment 1)
* [[Article 2]]
(comment 2)

Next, I want to group some cases to one section. Will this code parsed correcly?

== Group 1 ==
* [[Article 1]]
* [[Article 2]]

== Group 2 ==
* [[Article 3]]
* [[Article 4]]
* [[Article 5]]

Thanks. Facenapalm (talk) 12:40, 17 July 2016 (UTC)

Facenapalm The whitelist file can only contain lines of * [[Article name]], so no section headers or lines that are just comments. I haven't tried it, but I think you can add a comment after the last ]] on each line... atleast looking at the code it should be just fine. The translation file get updated just after 0z everyday. Bgwhite (talk) 04:49, 18 July 2016 (UTC)

@Facenapalm and Bgwhite: On frwiki, I tried adding more information to the list and it seems to work, see for example fr:Projet:Correction syntaxique/Erreur syntaxique 090/Liste blanche. --NicoV ^{(Talk on frwiki)} 07:45, 18 July 2016 (UTC)

Hm. Interesting. I'll try to do the same. Thanks. Facenapalm (talk) 10:19, 18 July 2016 (UTC)

#6

 Done

I added two patchs #5 and #6 to define farsi characters now report #6 is useless Yamaha5 (talk) 15:25, 24 July 2016 (UTC)

Yamaha5 I'll be adding #5 on the next update. I've some some other patches in there too. As for #6, it shouldn't be needed as they are changing MediaWiki's sorting. In theory, #6 and #37 will only check for symbols and not letters anymore. That reminds me to ask when that is going to happen. They announced it several months back. Bgwhite (talk) 06:43, 25 July 2016 (UTC)

Yamaha5 The translation file and main checkwiki page on fawiki (#5) has been updated. Bgwhite (talk) 17:53, 26 July 2016 (UTC)

bug on #29

 Resolved

why here it reported these pages? they had gallery ending tagYamaha5 (talk) 15:35, 24 July 2016 (UTC)

Yamaha5 In both cases, there were two <gallery> tags in a row with no </gallery> in between them. Bgwhite (talk) 06:33, 25 July 2016 (UTC)

False positive #91

 Resolved

Hello. The link [7] recognized as #91. Thank you. IKhitron (talk) 19:31, 25 July 2016 (UTC)

IKhitron what is the use of this link? It is totally unexpected. -- Magioladitis (talk) 19:45, 25 July 2016 (UTC)

It's used in an article about foreign wikipedia - two or three links "How this wikipedia looked in some special day". IKhitron (talk) 20:28, 25 July 2016 (UTC)

IKhitron we can whitelist the page but I wonder why the did not use {{diff}} or something else that does not require external site access. -- Magioladitis (talk) 20:31, 25 July 2016 (UTC)

Because it's main page - it transcludes a lot of templates that also changed from that day. What do you mean in "whitelist the day"? IKhitron (talk) 20:34, 25 July 2016 (UTC)

IKhitron whitelist the page. Digital dyslexia. -- Magioladitis (talk) 20:47, 25 July 2016 (UTC)

Very well. Thank you. — Preceding unsigned comment added by IKhitron (talk • contribs) 21:33, 25 July 2016 (UTC)

We're moved

 Done

Hello! Ruwiki project page has moved. New translation page lies here:

Проект:Check Wikipedia/Перевод

@Bgwhite: can you change its destination in your scripts? Thanks! Facenapalm (talk) 11:26, 26 July 2016 (UTC)

Facenapalm Done. Bgwhite (talk) 17:28, 26 July 2016 (UTC)

Facenapalm Done also for WPCleaner --NicoV ^{(Talk on frwiki)} 21:48, 26 July 2016 (UTC)

#3

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)

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)

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)

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)

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)

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

New false positives for #22

Hi Bgwhite, new false positives are appearing on frwiki when the category name itself contains a colon with whitespace characters around it, like [[Catégorie:Acteur de Lost : Les Disparus]] in fr:Terry O'Quinn. --NicoV ^{(Talk on frwiki)} 19:21, 28 July 2016 (UTC)

NicoV Should be fixed for the run that starts in an hour. enwiki doesn't have two colons in a cat. No good #*$(@ nothing &(*! French. Problem was caused by the update that catches the #22s WPC found. Bgwhite (talk) 23:11, 28 July 2016 (UTC)

#6 and #37 mostly obsolete.

@NicoV, Magioladitis, Yamaha5, Josve05a, Edgars2007, and Facenapalm: MediaWiki is moving to a new collation scheme called Unicode collation algorithm (UCA). Letters with diacritics will be sorted the same as with the non-diacritic version. I still don't know the timetable, but I did find the phab ticket (T136150) on moving enwiki to UCA. They have already moved several other wikis to UCA, including Russian, French, Latvian, Farsi and Swedish wikis. The listing of wikis can be found here; I'm thinking, #6 and #37 will only check for punctuation at some point for all wikis. I'll work on getting the wikis already on UCA to only check punctuation. Bgwhite (talk) 02:14, 29 July 2016 (UTC)

@Bgwhite: keep in your mined we have T139110 bug. is it makes problom for #6 and #37? Yamaha5 (talk) 03:49, 29 July 2016 (UTC)

lvwiki has disabled those ones, so I'm fine. --Edgars2007 (talk/contribs) 06:44, 29 July 2016 (UTC)

Same on ruwiki. In ruwiki, the only allowed letter with diacritic in titles is ё, but it's sorted correctly. Facenapalm (talk) 10:29, 29 July 2016 (UTC)

Reference localization

Hello. Is there a possibility to recognize a template as footnote? Thank you. IKhitron (talk) 15:35, 29 July 2016 (UTC)

• You're talking about this?

 error_003_templates_ruwiki=
   Примечания
   Список примечаний
   Reflist
   Reflist+ END
# ...
 error_078_templates_ruwiki=
   (Примечания|Список примечаний|Reflist\+?)(?![^}]*group) END

Facenapalm (talk) 16:16, 29 July 2016 (UTC)

Not at all, Facenapalm, thank you, I'm talking about a footnote (ref), bot references. IKhitron (talk) 16:28, 29 July 2016 (UTC)

Facenapalm I'm also unclear what you are asking. Remember, I'm slow. Could you put what your asking in different words?

Is there any possibility that you wanted to ask me this question, Bgwhite? IKhitron (talk) 23:59, 29 July 2016 (UTC)

IKhitron Yes. Like I said, I'm slow. Bgwhite (talk) 00:39, 30 July 2016 (UTC)

Well, Bgwhite, when you want to add a footnote you use <ref name=somename...>some text</ref>. I can't do this in rtl, so I use {{reftemplate|name=somename|...|some text}}, which is transcluded to the previous form. I asked if there is a possibility to add local name of footnote template, that will be recognized as ref tag. IKhitron (talk) 00:47, 30 July 2016 (UTC)

@Bgwhite: IKhitron (talk) 10:33, 9 August 2016 (UTC)

Syntax highlight

 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)

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)

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)

More errors / more bots

If we manage to have more bots running daily we can reduce the time required to fix errors drastically. This means we have more free time to detect more errors and and add to our list. What could these errors be? In an ideal world, we could check all of WP:GENFIXES and see what is worth to be done even as a sole task. -- Magioladitis (talk) 09:25, 30 July 2016 (UTC)

Help with translation page

Hello. I hope somebody who read this can find 5 minutes to help me. I'll be very glad if it's possible, so if I know it's not your "duty". I made a lot of changes in our translation page, because most of it was there from the time when checkwiki was a beta on dewiki. But it doesn't work any more! I tryed to find some variable without END or some another syntax error, but could not. What could be the problem? Thank you very very much in advance, IKhitron (talk) 11:55, 31 July 2016 (UTC)

Isn't "description_text_hewiki" the one, that screws up everything? --Edgars2007 (talk/contribs) 13:54, 31 July 2016 (UTC)

Everything is possible. Why do you think it's there, there is some problem in the description? Thank you very much, IKhitron (talk) 15:09, 31 July 2016 (UTC)

As I don't know, how those translation files are getting parsed to Checkwiki system, I'm just guessing. </syntaxhighlight> looked suspicous (and other non-HTML stuff), but I may be wrong. --Edgars2007 (talk/contribs) 16:05, 31 July 2016 (UTC)

I see. I created this part as in frwiki, and it works there. IKhitron (talk) 21:04, 31 July 2016 (UTC)

Article that doesn't exist appears in the database and in maintenance categories

The page USA:S inrikessäkerhetsdepartement has appeared on sv.wp's list of #2-errors for ~1 year now (or longer), at least when processing with WPCleaner. That page does not exist (the page USA:s inrikessäkerhetsdepartement however does exists). Yet this page appears on the CHECKWIKI list, and in the automated maintenece category Pages using invalid self-closed HTML tags on sv.wp. Why is this? (t) Josve05a (c) 10:02, 1 August 2016 (UTC)

It looks like parsers think USA is a namespace and automaticaly uppercase the first letter of the rest. IKhitron (talk) 15:07, 1 August 2016 (UTC)

Another unexpected run

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)

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)

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

#88 has false positive

At here most of the reported items are false positive. the {{DEFAULTSORT:}} on fa.wikipedia is {{ترتیب‌پیش‌فرض:}}. checkwiki shows any texts which is started with ترتیب: it doesn't care that it should have {{ at the first. for example fa:آرایه‌های ادبی doesn't have blank at first position.Yamaha5 (talk) 11:48, 9 August 2016 (UTC)

Request: Report for wrong dictation

There are some pages on wikipedia's like below which shows common wrong dictation. please add this to the reports to show which pages have these words.

• de:Wikipedia:Helferlein/Rechtschreibprüfung/Wortliste
• es:Wikipedia:Corrector_ortográfico/Listado
• gl:Wikipedia:Revisor_ortográfico/Listaxe
• he:ויקיפדיה:סקריפטים/בודק_איות/מילון
• ur:ویکیپیڈیا:املا_پڑتالگر/فہرست_الفاظ
• fa:ویکی‌پدیا:اشتباه‌یاب/فهرست

The first word before || is the wrong oneYamaha5 (talk) 09:30, 11 August 2016 (UTC)

#28 possible false positives

Hi. I started to fix #28, and found he:(Miss)understood and he:Anastacia at start of the list. It doesn't look like there are problems there. Maybe there are some more, didn't check yet. Thank you, IKhitron (talk) 18:06, 11 August 2016 (UTC)

IKhitron It was fixed a few days ago. The problem happens when a table is the very last thing in an article... no categories, defaultsort or other templates. I made a change to catch more cases of #28. It was thinking |}} was a table ending when it's most likely a template ending. As a result of the change, #28 will pick up cases of {{|, such as {{|url=http... , where "cite web" is missing. This is an error, but not related to tables. Bgwhite (talk) 21:47, 11 August 2016 (UTC)

Checkwiki updated errors #85, #90 and #91

Checkwiki has been updated to check for more cases of empty <gallery> tags in #85. Will also check for alot more cases of #90 and #91, especially when Wikipedia is used as a reference. Bgwhite (talk) 22:55, 24 September 2015 (UTC)

Bgwhite, NicoV Checkwiki updated #91 to also catch mobile links. This includes mobiles link to English Wikipedia too. (I am OK with that.) -- Magioladitis (talk) 06:05, 9 April 2016 (UTC)

Update scanned dump for nowiki?

Resolved

The last scanned dump was from 2016-05-01 (115 days ago).[8] Most other projects seem to have had updates since then.[9] Anybody know why it hasn't updated in this long for nowiki?--Telaneo (User talk page) 21:39, 24 August 2016 (UTC)

Telaneo Checkwiki was barfing on the article no:Afroaves. I've fixed the article. I'm rerunning nowiki (on my laptop) to see if there are any more problems. If not, I'll startup nowiki's processing on labs today. Bgwhite (talk) 21:05, 25 August 2016 (UTC)

Sweet! Thanks a lot!--Telaneo (User talk page) 21:10, 25 August 2016 (UTC)

HTML element whitelisting

Resolved

I don't know if we need to just ignore html in {{code}} or what, but every error on there as of now is wrapped in that template. Jerod Lycett (talk) 18:53, 29 August 2016 (UTC)

Jerodlycett Yea, that is a problem child. It's already on alot of whitelists. Besides error #40, what other whitelists does it need to be on? Oh... thank you for fixing alot of the ISBN errors. Bgwhite (talk) 22:26, 29 August 2016 (UTC)

I think it needs to be on the whitelist for #002. See This dump analysis. – Jonesey95 (talk) 22:41, 29 August 2016 (UTC)

Jonesey95 HTML element is listed, but I didn't know #002 had been rerun. The listing picked up a problem CheckWiki was getting. Bgwhite (talk) 04:38, 30 August 2016 (UTC)

Bgwhite It's out of WPCleaner so I can't look at it again, but there was issues with ul, li, and I think an a, and the one heading was upsetting it too I think. I've been away from Wikipedia for a while, but I'm back for now and this is something I've enjoyed doing, and the ISBNs are easy enough. Jerod Lycett (talk) 23:56, 29 August 2016 (UTC)

Jerodlycett Hmmm, it's on whitelists and WPCleaner isn't accessing the whitelists(?). It's on the #04, #39 and #100 whitelists. I need to add it to #40. Checkwiki isn't picking up #49 or #12 errors, but WPCleaner is. @NicoV:. Bgwhite (talk) 04:38, 30 August 2016 (UTC)

WPCleaner is accessing the whitelists: #04 and #39 are shown but in green because of the whitelist. No whitelist has been configured for #100 in Wikipedia:WikiProject Check Wikipedia/Translation, so WPCleaner doesn't use a whitelist for it. I don't know what to do for #49 and #12. --NicoV ^{(Talk on frwiki)} 06:25, 30 August 2016 (UTC)

I had the whitelist 100 set up, but didn't add it to the translation file for some reason. Both #40 and #100 whitelists have been added. Bgwhite (talk) 20:12, 30 August 2016 (UTC)

Improvements for #67

 Done

Hi @Bgwhite:, it seems that #67 is taking into accounts extra spaces put after the abbreviations in the error_067_abbreviations parameter. I had to search for them and remove them from frwiki configuration to remove some false positives. Would it be possible for CW to trim() each abbreviation by itself ? --NicoV ^{(Talk on frwiki)} 16:40, 31 August 2016 (UTC)

NicoV I like these very simple requests. It was already trimming the whitespaces at the beginning, so I added 5 characters to the code. It now trims the end of whitespaces. Bgwhite (talk) 18:32, 31 August 2016 (UTC)

Latvia(n Wikipedia) calling

Hi all! Have some stupid questions, so sorry if the answer is obvious :)

1. Does Checkwiki catch up such problematic headings: "== Foo ==="?
2. Heading "INFORMĀCIJA" isn't here. Bug or I'm missing something?
3. Wikilinks... I usually mess up them :) Will [{Foo]] and {[Foo]] be catched? And what about something like this (OK, this is quite hypothetical):
   • Some text {[link]] and some more text and [[link]}

--Edgars2007 (talk/contribs) 19:31, 30 October 2015 (UTC)

Edgars2007 Now I have London Calling playing in my head.

1. Only in a few cases. If there is only one "=" at the start of the heading or if it happens to be the first heading.
2. Yes and no. It should catch it, but I don't have it catching all the different types of characters. Ā is not on that list. Another thing I need to fix.
3. It should catch the first set ([{Foo]]), but will not catch the second {[link]].

Bgwhite (talk) 20:13, 30 October 2015 (UTC)

@Bgwhite: Isn't it the error #8 what Edgars2007 is referring to in the first question and the error #46 in the third question? (Anyway, you know the code better than me...) Matěj Suchánek (talk) 20:22, 30 October 2015 (UTC)

(edit conflict) OK, thanks. Then will try to do DB scan for 1st and 3rd. it would be LOL if I would listen to that song now - have all my music in computer in random order get played :) --Edgars2007 (talk/contribs) 20:25, 30 October 2015 (UTC)

Matěj Suchánek Edgars2007 #8 only catches headings that don't end in "=". So #8 doesn't apply to the first question. #19 and #83 are the exceptions I mentioned. For the third question, I check for how many [[ and ]] there are. The problem is there are the right amount of opening and closing brackets. I have to look at the code to remember what I did... why I didn't often eludes me :) Bgwhite (talk) 20:39, 30 October 2015 (UTC)

Edgars2007 Checkwiki should now pick up cases of INFORMĀCIJA and any other cases of whacked out Latvian or other non-standard Latin letters. Bgwhite (talk) 00:32, 2 November 2015 (UTC)

Thanks. --Edgars2007 (talk/contribs) 10:00, 2 November 2015 (UTC)

Oh, one more question. Does Checkwiki catch such URL: [url link text [[linked text]] some more]. It's quite worse, when wikilink is the first thing after URL. --Edgars2007 (talk/contribs) 18:28, 2 November 2015 (UTC)

Edgars2007 No, CW doesn't catch it. WPCleaner can if you configure error #513 (but WPC doesn't generate any list of pages with errors). --NicoV ^{(Talk on frwiki)} 08:02, 7 November 2015 (UTC)

OK, thanks. --Edgars2007 (talk/contribs) 08:10, 7 November 2015 (UTC)

Please fix substing of modules

[10] has this been fixed? It's not appropriate that these are being expanded. --Izno (talk) 22:16, 28 November 2020 (UTC)

quarry:query/50137 has a list of some 50 revisions starting in 2017 that are suspect. There may be more out there but I'm probably stretching quarry there with a 10 minute query. --Izno (talk) 05:19, 29 November 2020 (UTC)

Hi Izno. Have you contacted the contributors who did the modifications ? Substing is done manually with WPCleaner, so the decision was made by the editor. I checked 2 edits from my bot, WikiCleanerBot in your query, and they're not related with this, they consist of adding missing end of bold/italic in many table cells which results in several thousands bytes added. Your request doesn't match the description ("10,000+ bytes larger" in the description, only 2000 in the actual query). --NicoV ^{(Talk on frwiki)} 09:41, 29 November 2020 (UTC)

NicoV, I don't really understand what you say when it was done manually, not having used the tool. Are you suggesting they added "subst" themselves? I doubt that, given the standard edit summary that I saw. So maybe it is a choice in the interface? (I do not know, not having used the tool.)

Yes, I know there are false positives in that query, but size (the original query was for 10k but I shifted it to 2) was the best pattern I thought to look for. It does look like your bot has no issues, but it may also be coincidental; there are not many pages using modules in the conflicting way. --Izno (talk) 15:14, 29 November 2020 (UTC)

Hi Izno. By manually, I meant that WPCleaner in this case is used as an editing interface by a human, not a bot. WPCleaner detects the template programming element (which is regarded as bad practice to use them directly in an article, and is reported as error #34 for project Check Wikipedia) and marks it as an error #34. The human editor can then go to the error, and decide to do something about it or not. The substitution is something suggested by WPCleaner, but it's the human editor that decides what to do (apply the suggestion, do something else, do nothing...). The edit summary is computed by WPCleaner (there was an error #34 when the article was loaded, there isn't anymore when it is saved, so it's normal that it's added to the edit summary), and can be modified by the human editor. --NicoV ^{(Talk on frwiki)} 17:18, 29 November 2020 (UTC)

@NicoV: While I agree with which is regarded as bad practice to use them directly in an article, the users of these particular modules seem to have decided otherwise. What do you want to do about it and/or are you the right person to bug? :) --Izno (talk) 17:20, 29 November 2020 (UTC)

@NicoV: Please review this case. I would like to avoid future instances of this. If we need to get you a list of modules that may be found in mainspace wikitext so that some exceptions can be added, let me know. --Izno (talk) 05:48, 3 March 2021 (UTC)

@Izno: Do you have any example of such problem in the last months? As I said, this problem is due to a mix of human decision when applying the subst and bad practice of using modules directly in the main namespace (it's usually easy to create a one line template that simply calls the module so I don't understand why it's not done). The correct solution would be to tell the human editor who made the mistake (have you told him?) and to create the template to avoid using the module directly in main namespace (using modules in main namespace results for example in a poor editing experience on Visual Editor). BTW, WPCleaner doesn't suggest substing modules for several months.... --NicoV ^{(Talk on frwiki)} 06:53, 3 March 2021 (UTC)

@NicoV: The particular templates of interest are very heavy and are usually called multiple times on a page, causing issues with WP:PEIS. Calling them through a template doubles the expansion cost of the module (yes, doubles). That is why these are not called by a passthrough template. You may reasonably disagree that using modules directly in the mainspace is a good practice (and I would also reasonably disagree), but WCW should respect that that's what those editors are doing to try to fit the contents of interest onto the pages on which those modules are used. (Though, I think now that it should never have expanded modules anyway; it will always cause a mess. Would you appreciate having to clean up a post-expansion Module:Navbox? Or one of these tables? Me neither.)

No, I haven't told the editors who pressed the button. At best that is a N-person solution; the query indicates some 10 people have done it in the past. The fundamental issue is WCW recommending the change in the first place. That is a 1-person solution. Thanks for indicating that you attempted to fix this 3 months ago, I guess as a result of #WCW or User crap? (@Jonesey95:). It would have been nice to let me know. I'll rerun the query above when I get a free (literal) 10 minutes to see if any have reoccurred since. --Izno (talk) 07:08, 3 March 2021 (UTC)

Configuration for alswiki

Hi. It seems that a configuration exists for alswiki (als:Wikipedia:WikiProjekt Syntaxkorrektur/Übersetzung) and a project page also (als:Wikipedia:WikiProjekt Syntaxkorrektur), but on the project interface links to the project page and translation are not linking to them and it uses the default configuration rather than the one on the wiki. Is it possible to change that? --NicoV ^{(Talk on frwiki)} 08:33, 20 March 2021 (UTC)

@NicoV:  Done --Bamyers99 (talk) 14:34, 20 March 2021 (UTC)