Wikipedia talk:WPCleaner/Archive 2019

Latest comment: 4 years ago by NicoV in topic doi with ISBN


Local watchlist error

  Resolved

I recently had the application freeze up on me. After I got it restarted, I discovered that a large chunk (more than half) of my local watchlist is now missing. I checked the page here and saw that I was on an older version, so I updated to the latest, but the problem is persisting. Any thoughts? Mlaffs (talk) 17:50, 27 January 2019 (UTC)

Hello Mlaffs. Sorry to hear about this problem of losing a part of your local watchlist. Are you on Windows, Linux, Mac or any other OS? Depending on the operating system, the configuration is stored differently. On Windows, it's stored in the registry (run regedit to check it) in HKEY_CURRENT_USER\Software\JavaSoft\Prefs\org\wikipediacleaner, the local watch list being in Wikipedia\en\WatchPages. On Linux, it's probably stored in a dedicated folder named like "prefs". Changing version should have no impact, as the configuration is independent. What is stored in your configuration for the local watchlist? --NicoV (Talk on frwiki) 08:41, 28 January 2019 (UTC)
I'm on Mac OS 10.11.6. Most of the content is disambiguation pages - I use the tool to watch for incoming links. Mlaffs (talk) 23:07, 28 January 2019 (UTC)
Hello Mlaffs. Apparently, on Mac OS, preferences are stored in ~/Library/Preferences. So for WPCleaner, it would be ~/Library/Preferences/org.wikipediacleaner if I understand correctly. I don't know how it can be read to check what's in it. --NicoV (Talk on frwiki) 08:39, 29 January 2019 (UTC)
Thanks - I've checked the Preferences folder, but there's no file by that name or anything similar, or anywhere else on my system for that matter (which is weird, because there's clearly still a Local Watchlist, even if it's incomplete). Wherever the file is, it's clearly been corrupted, so I'd probably need to start over. So, I'll try solving my problem from a different angle - is there a way for me to bulk load pages to the Local Watchlist, or do they have to be loaded individually? Mlaffs (talk) 01:12, 30 January 2019 (UTC)
Hello Mlaffs. Yes, you can bulk load pages, and you have several options to do so from the main window:
  • You have a button to load a list of page from a file
  • You can create a page on Wikipedia, with internal links for each page, and load the list of internal links
  • ...
Once the list is loaded, you have the "+" button to add them to the local watch list. --NicoV (Talk on frwiki) 09:52, 30 January 2019 (UTC)
Fantastic - thanks! I'll declare bankruptcy and start again. :>) Mlaffs (talk) 22:56, 30 January 2019 (UTC)

Help - Size

  Resolved

Hi. I can barely read the text on the buttons and stuff. How do I "zoom in" or increase the font size? Thanks, --DannyS712 (talk) 02:09, 3 February 2019 (UTC)

Hi DannyS712. There's an option for the font size in the options window (button on the login windows): does that help? --NicoV (Talk on frwiki) 16:40, 3 February 2019 (UTC)
@NicoV: Thanks --DannyS712 (talk) 21:27, 10 February 2019 (UTC)

Login fails with unknown error

  Resolved

It had been a while since I ran WPCleaneer and I had a 1.x version installed via Java WS. I installed 2.0 using the installation instructions that involved the "getdown" method. Downloaded and started up fine but my login was failing with an unknown error. It seems that my password did not meet the 10 char minimum (don't know when that rule came into effect but yes been a while since I changed pw). I had to force a login in a web browser to uncover what the actual problem was associated with the "Unknown error" reported by WPCleaner. RedWolf (talk) 18:59, 15 February 2019 (UTC)

Thanks for the report RedWolf. I thought the error message was handled properly (but of course couldn't test because I first changed my password...), but it seems it's not correct. --NicoV (Talk on frwiki) 19:56, 15 February 2019 (UTC)

ISSN 0003-2700

  Resolved

This issn appears in Hollow cathode effect. Why is it on Wikipedia:WikiProject Check Wikipedia/ISSN errors? What should be done to correct this problem? Comfr (talk) 20:06, 17 February 2019 (UTC)

Hi Comfr. I don't see this one in the list of errors, only the one with the missing "-" which I have just fixed. --NicoV (Talk on frwiki) 06:52, 18 February 2019 (UTC)
The ISSN was generated by http://reftag.appspot.com/doiweb.py? I will ask Apoc2400 if he/she can fix it. Thanks for correcting the article and telling me what was wrong. Comfr (talk) 08:01, 18 February 2019 (UTC)
Hi . I think you already asked him the same question some time ago, see User_talk:Apoc2400#ISSN_03009572. Unfortunately, Apoc2400 doesn't seem to be contributing anymore... I'm thinking of adding automatic fixing in WPCleaner for such cases of incorrectly formatted ISSN, but I don't know when I will have the time to do so... --NicoV (Talk on frwiki) 10:39, 18 February 2019 (UTC)
I have a short memory and you are a good detective. Comfr (talk) 16:50, 18 February 2019 (UTC)

Test installation program

Hello everybody.

Version 2.0 is more difficult to install than the previous version, so I created an installer for WPCleaner. I hope it will be easy and intuitive enough for everybody to use. I'm looking for volunteers to try it and give me some feedback. I only tested it on Windows, but the idea is for it to work on most operating systems (the desktop shortcut works on Windows, and I hope on Linux).

Prerequisite: having Java (minimum 7, as before) installed (and associated with .jar files, which should be the case with a classical installation). Then, retrieve WPCleanerInstaller.jar and execute it. Fill in the various informations and click on Install.

Tell me if you encounter problems or if you think of enhancements that would be useful to add! --NicoV (Talk on frwiki) 18:57, 20 February 2019 (UTC)

I will try installing it on my new iMac once it arrives. I will let you know. Johnny Au (talk/contributions) 01:07, 1 May 2019 (UTC)
It is installed beautifully. However, it does not create a desktop icon. It would be great if the Mac version creates a shortcut in the Applications folder. Johnny Au (talk/contributions) 01:34, 1 June 2019 (UTC)
Hi Johnny Au. Thank you for the report. For the shortcut, my problem is that I don't have a MacOS to test it, and that the documentation on the format used by Apple is difficult to use. Can you help? For example, can you manually create a shortcut in the Applications folder and send me details on what has actually been created (apparently, in the Finder, you have an option to see "Package content" for an application). I need to know where the application package is put, what is the directory hierarchy in it, what the eventual files contain... --NicoV (Talk on frwiki) 10:16, 3 June 2019 (UTC)
I tried making it an application, but it didn't work. Johnny Au (talk/contributions) 12:40, 3 June 2019 (UTC)

ListISBNWarnings can not handle citation template ISBN parameters

On Czech Wikipedia we noticed ISBN errors in citation templates are not handled by WPCleaner. Therefore our list is short (cca 80 results), but the monitoring category of non-digit and non-x first or last letter in ISBN parameters has got over 700 broken ISBNs. The citation template code looks like:

{{Citace monografie|autor = Dave Sperry|titul = Example book title|isbn = 123-456-broken|isbn2 = 987-654-321}}

Would it be possible to fix this in WPCleaner? Also I'm not sure, if it can handle ISBN template as well {{ISBN|1234-5678-9012}}, but the citation templates are much more used than the ISBN template on our wiki. --Dvorapa (talk) 10:48, 12 March 2019 (UTC)

Hi Dvorapa. I changed the configuration to handle your monitoring category. I also changed the configuration to handle the ISBN template. The modifications should be taken into account in the list at the next update of the page (my bot will do it Friday, or someone else can launch it before): we'll see if everything is handled properly. --NicoV (Talk on frwiki) 12:57, 12 March 2019 (UTC)
Hi, thank you for your help. The category is only temporary, make sure WPCleaner does not depend on it. The category is filled from citation templates (| isbn[0-9]? = something) and ISBN template ({{ISBN|something}}) and was created just to be sure WPCleaner works correctly (and we'd found it wasn't). Perhaps we could list the citation templates containing isbn parameter/parameters in the preferences as well instead of the category? (I'm not familiar with the format) --Dvorapa (talk) 13:05, 12 March 2019 (UTC)
PS: citation templates use ISBN template to format ISBN correctly. I don't know how WPCleaner works with ISBNs, but it can be a useful information. --Dvorapa (talk) 13:08, 12 March 2019 (UTC)
Hi Dvorapa. If the category is only temporary, how WPCleaner will find the problems? On frwiki and enwiki, I also run a dump analysis which provides the list of pages with errors in complement of the monitoring category. I will try to run the update to see the result. --NicoV (Talk on frwiki) 15:14, 12 March 2019 (UTC)
Hi Dvorapa. Update done, does it seem good? --NicoV (Talk on frwiki) 15:28, 12 March 2019 (UTC)
It seems ok to me. Well the category really was quick-coded as it just tests if the first and last character is a digit (or x). I can make it proper maintenance category, but I thought there is some WPCleaner's own mechanism to find the broken ISBNs. I'm not really familiar with the way how WPCleaner collects items for this maintenance list. In fact I'm not really much familiar with WPCleaner at all, I'm commiting to/working with Pywikibot most of the time. But this maintenance list is really a good feature. --Dvorapa (talk) 20:29, 12 March 2019 (UTC)
Hi Dvorapa. To update the list, WPCleaner does the following: consolidate a list of pages to check from various sources, check each page in the list, report the results. To have the most complete results, there are several parts: configuring WPCleaner to detect as many cases as possible in a page (for example, adding the ISBN template in the configuration), and use a consolidated list as complete as possible. The sources for the consolidated list can be CheckWiki (#69 to 73, but it doesn't check in templates), categories (monitoring category) and list of links inside a page. So to have the more complete list, I suggest:
  • Setting up a complete monitoring category. You can use some code examples in fr:Module:Biblio/Références (report any ISBN that doesn't match a normal ISBN), or in the enwiki modules (I think they try to automatically fix common mistakes, and only report the ones not fixed). This will only provide you with incorrect ISBN inside templates.
  • Regularly analyzing dump files with WPCleaner for errors #69 to #73, like I do for frwiki or enwiki. I can help in configuring this and explaining how to do it, but I don't think I will be able to run it regularly myself (needs to download the dump file and run the analysis, preferably for each new dump, so twice a month). This will provide all errors that can be detected by WPCleaner so more thorough, but with a delay in finding them.
--NicoV (Talk on frwiki) 06:51, 13 March 2019 (UTC)
Okay, I made a module, that checks the control digit and length (and format partially) for ISBNs in templates. It fills the same category and it works nicely. Perhaps I'll make the category permanent. I don't have time for dump processing, but anyway I ope the combination of Check Wiki and check module should be sufficient. Thank you for your help! --Dvorapa (talk) 13:30, 29 March 2019 (UTC)
Hi Dvorapa. Ok. This combination will already give good results, the dump analysis provides just a few more situations that are missed by Check Wiki outside of templates. It can be added at any point in the future if someone has time for running the dump analysis on a regular basis. --NicoV (Talk on frwiki) 15:19, 29 March 2019 (UTC)

Error occurring when try to run 'Disambiguation'

  Resolved

Hi. I just started getting an error when I run the Disambiguation command on a page e.g. Madonna. It is:

"An error occurred: org/apache/commons/lang3/tuple/ImmutablePair - java.lang.NoClassDefFoundError"

If I run 'Full analysis' it just seems to be stuck on 'Analyzing data'.

I tried reinstalling (removing directory WPCleaner and rerunning WPCleanerInstaller.jar) but that didn't help. I'm running openjdk version "1.8.0_202" on Windows 10.

Thanks, Tassedethe (talk) 21:00, 4 May 2019 (UTC)

Hi Tassedethe. It seems I released a buggy version with a missing dependency. I'm fixing it right now, should be available in a few minutes (fingers crossed). --NicoV (Talk on frwiki) 21:23, 4 May 2019 (UTC)
Hi Tassedethe. It should work now. --NicoV (Talk on frwiki) 21:30, 4 May 2019 (UTC)
Thanks NicoV! And as always thanks for the great software. Tassedethe (talk) 21:57, 4 May 2019 (UTC)

  Resolved

It would be great if WPCleaner could check dead external links as well. Johnny Au (talk/contributions) 03:00, 1 March 2018 (UTC)

You mean a tool to check every external link in a page and report the ones that do not work? It can be included in a later version, when I find time... --NicoV (Talk on frwiki) 14:35, 1 March 2018 (UTC)
I will look forward to that. Johnny Au (talk/contributions) 02:02, 1 April 2019 (UTC)
Hi Johnny Au. First version of dead links checker is now available, look for the   icon (in page list and in full analysis). It will probably require some more work to accurately report all dead links, do not hesitate to give me some examples of incorrect reports. --NicoV (Talk on frwiki) 08:19, 12 August 2019 (UTC)

Track listings with deprecated parameters

  Resolved

Does WPcleaner fix articles that have the category "Track listings with deprecated parameters"?--Mjs1991 (talk) 23:05, 17 August 2019 (UTC)

Hi Mjs1991. Currently, there's nothing in WPCleaner to detect/fix use of templates with deprecated parameters, like the articles in Track listings with deprecated parameters.This is something that could be added, if I find the time to do it... --NicoV (Talk on frwiki) 17:31, 22 August 2019 (UTC)
Hi Mjs1991. I've added error #545 to WPCleaner to report deprecated parameters, and I've configurer it for Track listings with deprecated parameters in Wikipedia:WikiProject Check Wikipedia/Translation. Other templates can be added to the configuration if need be. --NicoV (Talk on frwiki) 10:14, 7 September 2019 (UTC)

Auto Login to WPCleaner

  Resolved

It would be nice to have a setting where upon loading the WPCleaner.bat file I would be auto-logged in. Jimj wpg (talk) 21:48, 19 August 2019 (UTC)

Hi Jimj wpg. It's already possible, you just need to have a file named credentials.txt in the same folder, with 2 lines in it
user=...
password=...
This file can be created manually or by the installation program. -NicoV (Talk on frwiki) 17:28, 22 August 2019 (UTC)
Jimj wpg The file credentials.txt only allows to automatically fill up the username and password in WPCleaner main screen. If you create a shortcut to WPCleaner.bat, you can simply add the following parameters: -credentials credentials.txt en to be automatically logged in on enwiki. The installation program creates such a shortcut if you give all the information when installing. --NicoV (Talk on frwiki) 18:50, 22 August 2019 (UTC)

WPCleaner Apostrophe 's'

  Declined This problem is due to rules in Wikipedia:AutoWikiBrowser/Typos.

I was just checking some pages in WPCleaner and noticed that it regularly flags plural or apostrophe'd words as false positives. An example is the Eaton's article, as the department store officially had an apostrophe in its name, yet WPCleaner turned the text red. Jimj wpg (talk) 00:42, 12 August 2019 (UTC)

Hello Jimj wpg. WPCleaner applies the rules from Wikipedia:AutoWikiBrowser/Typos, maybe there's a rule that needs some fixing? --NicoV (Talk on frwiki) 06:17, 12 August 2019 (UTC)
Hello Jimj wpg. Maybe the first rule in Wikipedia:AutoWikiBrowser/Typos#Punctuation? --NicoV (Talk on frwiki) 12:30, 12 August 2019 (UTC)
Is it this discussion about the same problem? https://en.wikipedia.org/wiki/Wikipedia_talk:AutoWikiBrowser/Typos#Move_%22's%22_rule_to_WP:GENFIXES? Jimj wpg (talk) 15:22, 12 August 2019 (UTC)
Hi Jimj wpg. Yes, it seems to be about the same rule. --NicoV (Talk on frwiki) 07:14, 13 August 2019 (UTC)
Until I can better figure out how to handle this issue on my end, I have disabled spelling & grammar setting on my copy of WPCleaner. Jimj wpg (talk) 21:46, 19 August 2019 (UTC)

Linter errors positions

  Resolved

Hello,
When you check Linter errors in WPC, a pop-up shows some information as the position of the error (ie beginning and ending). Do you know if there is a way to go directly to a specified position (or to position displayed by Linter errors pop-up) ? It would be very helpfull for long articles, it's sometimes very hard just to find the place of the error (and not the error itself). Please ask clarifications if needed.
Many thanks in advance, --Tomo8 5 (talk) 15:34, 5 October 2019 (UTC)

Edit : Previously it was a button looking as a small magnifying glass (petite loupe in french) in the Linter errors pop-up but it seems to have disappeared. --Tomo8 5 (talk) 15:38, 5 October 2019 (UTC)
Hi Tomo8 5. Thanks for the report. It's a bug I introduced when I refactored the integration with Linter and added the checks for dead links. It should be fixed now. --NicoV (Talk on frwiki) 13:33, 6 October 2019 (UTC)
Yes it is. Many thanks NicoV ! --Tomo8 5 (talk) 13:42, 6 October 2019 (UTC)

Unrecognized option: -jar /home/jmax/WPCleaner/libs/getdown-launcher.jar . client -credentials credentials.txt

  Resolved

After a new installation, I have this new error message

$ bash -x WPCleaner.sh -wiki fr +++ dirname WPCleaner.sh ++ cd . ++ pwd -P + JAVA_APP_DIR=/home/jmax/WPCleaner + JAVA_LIB_DIR=/home/jmax/WPCleaner/libs + cd /home/jmax/WPCleaner + JAVA_PARAMS='-jar /home/jmax/WPCleaner/libs/getdown-launcher.jar . client' + '[' -f credentials.txt ']' + JAVA_PARAMS='-jar /home/jmax/WPCleaner/libs/getdown-launcher.jar . client -credentials credentials.txt' + case $# in + java '-jar /home/jmax/WPCleaner/libs/getdown-launcher.jar . client -credentials credentials.txt' -wiki fr Unrecognized option: -jar /home/jmax/WPCleaner/libs/getdown-launcher.jar . client -credentials credentials.txt Error: Could not create the Java Virtual Machine. Error: A fatal exception has occurred. Program will exit.

% ls libs/get* libs/getdown-launcher-1.8.2.jar libs/gettext-commons.jar libs/getdown-launcher-1.8.2.jarv libs/gettext-commons.jarv

so, libs/getdown-launcher.jar is missing Jmax (talk) 10:42, 10 November 2019 (UTC)

Thanks for the report Jmax. It should be fixed now, it was a small side effect of a modification proposed by RogueScholar (the build process was supposed to replace the getdown-launcher.jar in WPCleaner.sh by getdown-launcher-1.8.2.jar, but it didn't work due to a modification in WPCleaner.sh). Tell me if it works now. --NicoV (Talk on frwiki) 11:30, 10 November 2019 (UTC)
I have to adapt WPCleaner.sh
% diff WPCleaner.sh WPCleaner.sh.orig
57c57
< 0) java ${JAVA_PARAMS}
---
> 0) java "${JAVA_PARAMS}"
59c59
< *) java ${JAVA_PARAMS} $@
---
> *) java "${JAVA_PARAMS}" "$@"
Jmax (talk) 05:52, 16 November 2019 (UTC)
Thanks Jmax. I tried with this modification, hope it works... --NicoV (Talk on frwiki) 12:34, 23 November 2019 (UTC)

Invalid ISBN

  Resolved

Hello, some books are published with invalid ISBNs. Therefore the templates have a parameter for marking invalid ISBNs which hides the error message. WPCleaner should check for such parameters before reporting an invalid ISBN.

Templates:

  • {{ISBN|0-0000-0000-0|invalid1=yes}}
  • {{ISBNT|0-0000-0000-0|invalid1=yes}}
  • {{cite book | ... |ignore-isbn-error=true}}
  • {{cite encyclopedia | ... |ignore-isbn-error=true}}

The first two templates can be used with many ISBNs, so the parameteres are invalid1=yes to invalid9=yes.
Articles with marked invalid ISBN (for testing WPCleaner):

Regards --GünniX (talk) 09:15, 24 November 2019 (UTC)

Hi GünniX. It was already working for Coaches of the London and North Eastern Railway, and it should work now for Northwest Germanic. I will look into the other ones which require some development. --NicoV (Talk on frwiki) 18:28, 26 November 2019 (UTC)
Hi GünniX. WPCleaner should now be able to handle such cases. I've done the configuration for parameter invalid1 for ISBNT, others can be added to the configuration. Each line contains the following fields separated by pipes : name of the template, optional name of the parameter that can disable checking ISBN (example: ignore-isbn-error), optional value of the parameter that can disable checking ISBN (example: true), optional parameter which would be ignored (example: 1). --NicoV (Talk on frwiki) 20:29, 26 November 2019 (UTC)
Hi NicoV, I have added {{Citation | ... |ignore-isbn-error=true}} to the configuration. --GünniX (talk) 18:39, 30 November 2019 (UTC)
Thanks GünniX. It removed a few ISBN from the list. --NicoV (Talk on frwiki) 18:48, 30 November 2019 (UTC)
Hi NicoV, I've added even {{cite book | ... |ignore-isbn-error=yes}} to the configuration. --GünniX (talk) 10:33, 7 December 2019 (UTC)

False positives in #71

  Resolved

There are false positives in Wikipedia:CHECKWIKI/WPC 071 dump.

Example: Hiyokoi: isbn=X78-4-08-867044-7

The string occurs within a URL: http://books.shueisha.co.jp/CGI/search/zen_list.cgi?siries_isbn=X78-4-08-867044-7&siries_kanren_isbn=&mode=2

Can the code that generates this report be modified to ignore patterns within URLs? I believe that this problem also occurs in the other ISBN reports (69, 70, 72, and Wikipedia:WikiProject Check Wikipedia/ISBN errors). I can look for other examples if that is helpful. – Jonesey95 (talk) 15:05, 9 December 2019 (UTC)

Hi Jonesey95. It should be in the next release. --NicoV (Talk on frwiki) 12:43, 26 December 2019 (UTC)

doi with ISBN

  Resolved

Hi @NicoV:, the page ISBN errors shows the article Physiology of dinosaurs, which has the keyword ISBN in a doi inside a template:

{{cite journal | doi=10.1159/isbn.978-3-318-01866-0 }}

How to handle this situation in the configuration?
The parameter isbn_ignore_templates= checks for values (without wildcard?)
The parameter isbn_ignore_incorrect_templates= isn't explained at the Help page. Is this parameter for ignoring a whole template or only for a field in a template? Would this line Cite journal|doi in the configuration ignore just the field doi of the template cite journal? Best regards --GünniX (talk) 17:19, 29 December 2019 (UTC)

All text within |doi= or {{doi}} should be ignored when compiling this report. The templates have their own error-checking code. – Jonesey95 (talk) 00:57, 30 December 2019 (UTC)
Hi GünniX, Jonesey95.
I've added the description for isbn_ignore_templates= and isbn_ignore_incorrect_templates=. In your case, yes, it's the second one that can be used for doi. I'm adding Cite journal, but other templates should probably be added (but I didn't find other similar problems when looking at Wikipedia:CHECKWIKI/WPC 069 dump). --NicoV (Talk on frwiki) 08:30, 30 December 2019 (UTC)