Talk:PetScan

German instructions

Hi; please see w:de:Wikipedia:Technik/Labs/PetScan. Greetings --PerfektesChaos (talk) 19:22, 29 March 2016 (UTC)

operating manual in german?

Um das Tool zu verwenden hatte ich PetScan/de besucht. Das suggeriert mir eine deutschsprachige Bedienungsanleitung. Schade - welche Sprache müsste ich wählen um deutsch zu bekommen? --2003:DE:3E1:CE01:4517:A3FB:2ADF:C616 19:59, 26 August 2017 (UTC)

Improvement suggestions

Some suggestions for things that would make this tool even more useful for some queries:

1. An option to only select pages that are (not) subpages.
2. Each line of the output could include a link to the CatCycle tool to find out how that page is in that category.
3. A "Get the categories containing the individual page" option like Quick-Intersection has.
4. If the page is a (hard) redirect then the link in the output should go to the redirect page (rather than follow the link to its target).
5. The ability to select pages (in particular namespaces) that are in no categories (e.g. any pages in the Template namespace that are not in Category:Wikipedia_templates - this always returns zero results).

DexDor (talk) 06:59, 31 March 2016 (UTC)

Feature requests

1. Generate reports on all pages linked from an arbitrary page
2. Collect and report pageview metrics

Right now PetScan presents reports based on categories. A problem with this is that this creates incentive for organizations to game categories in a way that conflicts with Wikimedia community volunteers. Organizations, including every organization that has a Wikipedian-in-residence or a partnership with a Wikimedia chapter, want reports of this sort. If reports can only be generated through categories, then that creates pressure to adapt the public category system to reflect financial pressures from organizations. This in turn positions all outreach projects including GLAM, Wikimedia chapters, and all institutional partnerships to be against Wikimedia community processes. To fix this, then as an alternative to generating reports through categories, also give an option to generate reports from pages which are private lists of links in userpages.

Going further - this tool already does a great job at listing all Wikipedia pages in categories. Given a list of articles, it would be extremely useful to be able to get metrics on those Wikipedia articles. Steiner's Wikipedia Tools for Google Chrome already does everything PetScan does but in Google Sheets, but still, that tool is data overkill and it is hard for people without good spreadsheet skills to get only the right amount of data. If this PetScan could be combined with the output of en:User:Vipul's WikipediaViews.org then this would be immensely useful to developing institutional partnerships with the Wikipedia community.

Blue Rasberry (talk) 13:56, 17 April 2016 (UTC)

new pages

Option "Only pages created during the above time window (overrides "last revision")" leads to crash, the result is not formed. Игорь Темиров (talk) 06:30, 21 April 2016 (UTC)

Manual localizations

Link "Manual" in the tool open PetScan/en, ok. But if in the tool change "Interface language" e.g to "ru", then "Manual" open dead page PetScan/ru. I think could make redirectes for all inter-subpages to main PetScan or to this english page. Or set in internalization templates, like set on top Wikidata/Development etc. --Vladis13 (talk) 10:12, 26 May 2016 (UTC)

Add free text search

Would it be possible to cross-search on a category and a piece of text on all articles in that category? For instance: have PetScan search all articles in en:Category:American racehorses for the word "California" and produce a list of those articles. I see no way to do that as it is. Gorthian (talk) 05:00, 5 June 2016 (UTC)

Does not work for me

I don't get it, and my tests fail. I see (say) three tabs "cats, props, templates", but it is unclear whether they work "and" or "sole". I cannot even replicate the regular "this template's transclusion pages" [WLH] ("do it!" result: 0; WP result: 2500). -DePiep (talk) 19:10, 7 June 2016 (UTC)

User:Magnus Manske, I may be having the same problem as User:DePiep. I'm getting "0 results" on this query, which I believe is identical to the one that worked for me last week. I want the list of pages that are in w:en:Category:Unknown-importance_medicine_articles, but not in w:en:Category:Unassessed medicine articles. There should be hundreds (maybe ~1,000) pages in the results. WhatamIdoing (talk) 01:45, 23 July 2016 (UTC)

I don't understand what DePip even tries to do; an example would be nice. As for WhatamIdoing, your query looks for articles, but the categories contain talk pages. Extend your query to talk pages (on "Page properties"), and it works as expected. --Magnus Manske (talk) 14:39, 23 July 2016 (UTC)

Thanks. I've saved a copy of this link. WhatamIdoing (talk) 06:44, 24 July 2016 (UTC)

• In regular enwiki, I can create a 'What links here' list for a template page. It lists transclusions and links for that page. However, with PetScan I can not create such a list. -DePiep (talk) 16:39, 25 July 2016 (UTC)

2 years old file returned even though max_age=96

This query returns me all pictures uploaded via the Android app in the last 4 days.

It works well, except I noticed this false positive which was uploaded in 2014 (but someone changed its categories yesterday).

Could someone add to the manual a description of how max_age works? And is there another keyword to get only files that have been modified in the last 4 days, excluding updates? Thanks! Syced (talk) 04:22, 16 June 2016 (UTC)

Add/Remove Statements on wikidata

I used Autolist 2 to add/remove Statements on wikidata items but can not find how to do the same on Petscan. Its too complicated. I could generate list of items by category but cant understand that how to add/remove statements to that list items. Please let me know.--Nizil Shah (talk) 06:36, 2 August 2016 (UTC)

+1. I cannot figure out in which cases I have editing form, and in which I have only wikified list. For example, how should I edit PSID=121595? --Infovarius (talk) 23:13, 4 August 2016 (UTC)

Hello User:Nizil Shah, User:Infovarius, I was facing the same problem, I wanted to add a wikidata property for a petscan result for already existing items. User:Vesihiisi (thanks again) found the solution: On the "Other sources" tab, select "Wikidata" in the "Use wiki" section. It's set to "Automatic" by default, but "Wikidata" will make the editing form appear. What I am still looking for is a possibility to remove items from the Petscan result which already have the property you want to add already set (e.g. if the petscan result is a long list, but only a few items actually might have to be changed). It seems, that filtering with SPARQL (...FILTER NOT EXISTS { ?item wdt:P463 wd:Q299015 }....) does not work, because this selects ALL items, not only those from the petscan result, therefore the query stops with an timeout error. --M2k~dewiki (talk) 12:40, 16 August 2016 (UTC)

User:Vesihiisi also had a solution to this problem: This one's kinda tricky -- I do these sorts of queries like this (autorun). I.e. I put the value in the "Uses items/props : None" field on the Wikidata tab, but I don't actually use P463 anywhere in the query... It means that any items that link to Q414163 from any property will be excluded, not only those that pair it with P463. It just happens to work great in this particular case --M2k~dewiki (talk) 13:47, 16 August 2016 (UTC)

Hi User:Nizil Shah, User:Infovarius, User:M2k~dewiki & User:Vesihiisi. I figured out a way to do this using a combination of Vesihiisi's methods and the SPARQL box. For instance, for the PSID here [1], it finds people in category "American accordionists," who DO have instrument = accordion but DO NOT have gender = male. In this example I don't use the "Uses items/props" box, only the SPARQL box under "Other sources" and enter SELECT ?item WHERE { ?item wdt:P1303 wd:Q79838 . MINUS { ?item wdt:P21 wd:Q6581097 } }. So the first part of the SPARQL is the property/item pair you want, and the second is what to exclude. Sweet kate (talk) 17:03, 24 October 2016 (UTC)

To have mutliple conditions, you can string them together like this: SELECT ?item WHERE { ?item wdt:P1303 wd:Q79838 . ?item wdt:P31 wd:Q5 . MINUS { ?item wdt:P21 wd:Q6581097 } MINUS { ?item wdt:P1303 wd:Q5994 } }. Sweet kate (talk) 17:11, 24 October 2016 (UTC)

How to add qualifiers to property by Petscan? Sweet kate, User:Infovarius, User:M2k~dewiki & User:Vesihiisi, help me. I have no technical knowledge.--Nizil Shah (talk) 06:58, 9 December 2016 (UTC)

Missing pages

As I see, PetScan should be able to return a list of missing pages (red links) if checkbox «Show only redlinks to main (article) namespace» has been checked at the «Output» tab. But returned results are just the same as without checking the box at all, e.g. pages returned are existing articles. Did I do something wrong?

Second, I should enter «be_x_old» code into the «Language» field to run on be-tarask.wiki. It's ok, but after each run the text in this window is replaced by «be x old» («_» replaced by « ») which is quite uncomfortable. --Renessaince (talk) 15:33, 5 August 2016 (UTC)

@Renessaince: I just had the same problem. Most likely it is because you did not click the "Show redlinks" option first. I do have one request for redlink searching: can the output include number of missing links as a column? This is how the Missing Topics tool used to behave and it's a really key thing to include when building these kinds of links (because a missing article with 100 links is of higher priority than one with, say, 20). Thanks. Sillyfolkboy (talk) 23:59, 29 August 2016 (UTC)

Got it now, thanks.

Now there is another issue about this tool: for w:be-tarask: it works only if I enter value «be_x_old» into the «Language» field. Neither «be-x-old» nor «be-tarask» fits for this tool, and using «be_x_old» looks inappropriate because of the wrongly generated links in the output, e. g. be_x_old instead of be-x-old. --Renessaince (talk) 08:41, 31 May 2017 (UTC)

Further question. In the generated list of missing articles there are two articles which are actually exist: Пераклады Бібліі на беларускую мову and ВНУ Беларусі (second one is a redirect). What's wrong with them? --Renessaince (talk) 13:34, 5 June 2017 (UTC)

Adding coordinates from templates

I was hoping that I could use PetScan to find pages with coordinates and WD entries that don't have coordinates and fill P625 with those values easily. The only option I fund was to to extract the coordinates from the used template and manually copy/paste from the PetScan result list. See [2]

If copy/paset is the only option, it would be good to use a formatting in PetScan that is accepted in WD directly. Currently I have to reformat manually for WD to accept teh values.

I would love to do the same for Coats of Arms, location maps and other template fields of course.

Thanks. --Aeroid (talk) 06:56, 26 August 2016 (UTC)

Change statements

Is it possible to change statements using Petscan? --Epìdosis 12:59, 4 November 2016 (UTC)

has no claim

I cannot filter a list by "noclaim": https://petscan.wmflabs.org/?psid=590599. How to do it? --Infovarius (talk) 14:08, 14 November 2016 (UTC)

Default Namespaces

When calling with parameters from a link the Namespaces default of article is not switch on. I also cannot see how to add a parameter to force just articles to be listed. This is causing problems with statistics pages at Wikivoyage. --Traveler100 (talk) 19:59, 7 December 2016 (UTC)

Labels not in English

How to make labels of items to be displayed in other language? I've tried to change language of interface, language of wiki at first page but in vain... --Infovarius (talk) 10:29, 12 December 2016 (UTC)

В схожей теме выше #Manual localizations почти год нет ответа. Похоже локализация авторов не интересует. --Vladis13 (talk) 23:39, 3 January 2017 (UTC)

Can anyone write the steps to find out label which is not available in ml language. For eg : Category:American feminist writers.

I would like to get the names in english which has no label in malayalam language. --Akbarali (talk) 06:11, 30 August 2018 (UTC)

Wikidata + Sitelink - Template

I am maybe being dim here, but I can't seem to do the following. What I want is to combine:

• Wikidata item with P1367
• Sitelink to en-wiki
• No template Art UK bio

To get the first it seems I can't just use the Wikidata tab (which is only a filter?) but I need to write a query. Which is fine -- I can even specialise to extract items which have one and only one P1367.

The second seems to be done automatically, whether I want it or not. (Actually it's probably easy enough to turn off or change with the right checkboxes on the "Other sources" tab.

The third I am having more trouble with. I can generate a list of pages which *do* have the template easily enough, using the Templates tab. But I can't seem to use the tab to filter away pages which *don't* have the template.

What's the right way to do this? Jheald (talk) 12:56, 22 February 2017 (UTC)

• I now know how to do this. Some things I've learned:
  1. Template results can be excluded by saving them as a Pagepile, then using "Sparql NOT Pagepile" in 'Other sources' -- 'Combination' to exclude them. ("Sparql NOT Templates" doesn't work, because "Templates" isn't understood as an input source). So if this is what you need, first use the 'Templates' screen to produce a list of all pages that do have the template, save it as a Pagepile, then use 'Combination' to exclude it.
  2. It is important that the SPARQL query does not include the underscore character, particularly not in variable names. Such queries will successfully run once, but when PetScan stores them when you switch to another screen, it turns all the underscores into spaces, and the query then no longer works. So "?membershipStmt" is an okay name for a variable; but "?membership_Stmt" is not.
  3. It is important to use DISTINCT in the SELECT statement in the query. If there are two hits to the same item, they are not merged, but Petscan only finds a matching wiki page once -- the other hit is returned unlinked.
  4. Choose 'Use wiki' = 'From categories' to get the output as a list of Wiki pages (suitable eg to put into AWB), otherwise the list will be of Wikidata items. This works (and is needed) even if you have made no other use of the categories screen -- it still specifies the reference wiki.

-- Jheald (talk) 12:27, 20 January 2018 (UTC)

Multiple projects ?

Hi,

Is it possible to cross categories from multiple projects ? For instance s:en:Category:Authors and s:fr:Catégorie:Auteurs.

Reasoning : as PetScan is often used to import data to Wikidata, it could be useful to check the consistency of the data in different projects before, in order to importing contradictory data into Wikidata.

Cdlt, VIGNERON * ^discut. 15:49, 28 February 2017 (UTC)

Anniversaries

I want to make a list of people, connected with Ukraine, who celebrate anniversaries.

For that I have to make a lot of PetScan requests (10 for every century) from the Ukrainian Wikipedia like:

Українці Ukrainian people

Народились 6 березня March 6 births

Народились 1907 1907 births

Українці Ukrainian people

Народились 6 березня March 6 births

Народились 1917 1917 births

Українці Ukrainian people

Народились 6 березня March 6 births

Народились 1927 1927 births

Українці Ukrainian people

Народились 6 березня March 6 births

Народились 1937 1937 births

…

Instead of many requests I would like to make one like

Українці Ukrainian people

Народились 6 березня March 6 births

Народились 1907 .or. Народились 1917 .or. Народились 1927 .or. Народились 1937 .or. Народились 1947 … 1907 births .or. 1917 births .or. 1927 births .or. 1937 births .or. 1947 births …

What can be recommended? Probably data from Wikidata could be used? If Yes, then How?

--Perohanych (talk) 07:57, 6 March 2017 (UTC) P.S. I am aware that in English and in German Wikipedias there are no categories like March 6 births, but in the Ukrainian Wikipedia we do have such categories.

You can get this directly from Wikidata, provided there items for the people, they have a birthday and nationality set. Query is here, it will automatically use the current year, let me know if you need other years. I tried a mixed Wikipedia/Wikidata PetScan query but is does not return any results. I guess the Wikidata birthdays are incomplete. This is a list of Ukrainian people without birthdays on Wikidata. --Magnus Manske (talk) 10:01, 7 June 2017 (UTC)

Categories

Hi,

I am having problem, when I set more categories into Categories:Categories getting zero results. I follow the manual, which say one per line (eg. Towns in Kladno County). Am I doing something wrong, that it doesnt work? When I place just one category it works, while I set more it doesnt.--Juandev (talk) 19:10, 18 April 2017 (UTC)

Perhaps you selected Combination: Subset instead of Combination: Union? --FriedhelmW (talk) 20:08, 18 April 2017 (UTC)

Cool, thx. Now it works.--Juandev (talk) 12:47, 19 April 2017 (UTC)

I tried several time to get one list by using category. But it does not work. Anyone can help me. I need to get "List of schools in the United Arab Emirates". I have given this (List of schools in the United Arab Emirates) in Categories box. Wikipedia link is given as follows. https://en.wikipedia.org/wiki/List_of_schools_in_the_United_Arab_Emirates --Akbarali (talk) 02:57, 7 September 2017 (UTC)

The term you entered into the Categories box is not a category, it is a Wikipedia article. Go to the "Templates&links" tab and put it into the field "Linked from / All of these pages", and it will work. OMHalck (talk) 11:43, 28 November 2017 (UTC)

Not updating

Results do not appear to be updating. Values of some enquires are returning same results for last 2 days even though should have changed. --Traveler100 (talk) 08:22, 29 April 2017 (UTC)

As bug reports go, this is indeed better than "stuff doesn't work", but not significantly so. --Magnus Manske (talk) 10:02, 7 June 2017 (UTC)

Creator

When creating new items, PetScan doesn't fill labels... It's a step back from http://tools.wmflabs.org/wikidata-todo/creator.html. --Infovarius (talk) 08:58, 21 June 2017 (UTC)

Orphans

Is it possible to find orphan articles in a category? For example, articles from Spanish Wikipedia in the "Matemáticas" category with level 4 whithout articles linking to them. The problem is that in Spanish Wikipedia is going to be deleted the Spanish template for "orphan" and I want (if it is possible) a similar method for doing this. Thanks, Juan Mayordomo (talk) 17:11, 10 July 2017 (UTC)

Modules used

At the moment when you do a search on petscan, you can check for the use of a template. Would be great to implement this same behavior with Modules. --Zackmann08 (talk) 21:24, 5 September 2017 (UTC)

Article and talk page

How could I look for articles belonging to a category, whose talk page don't belong to another categorie? I would need this query to list all articles belonging to a Portal, whose talk page doesn't have any associated wikiproject assessments. Any help will be appreciated. Djiboun (talk) 22:04, 26 September 2017 (UTC)

Suggested feature

The ability to tell PetScan to ignore the contents of certain templates e.g links in NavBoxes would be a good feature to have. As you can probably guess, my results are being polluted by NavBox links. The reverse would be good to, to only scan inside a specified template and ignore the contents of the rest of the article. - X201 (talk) 08:14, 20 October 2017 (UTC)

Gallery output

Where can i find the gallery output from Catscan? Do i have to use a different tool now? --Ailura (talk) 18:53, 26 November 2017 (UTC)

I was wondering the same thing. While it doesn't seem to have this. If you choose as output option "Plain text" that you can just copy paste into a gallery page and have them show up technically Thibaultmol (talk) 09:10, 5 August 2022 (UTC)

Wishlist item: Sorting by number of languages

Thanks for a truly awesome tool! If I were to have one wish for further functionality, it would be the ability to sort the results by the number of language links from each article, ie the size of the language list in the sidebar when viewing Wikipedia on a desktop. Like the number of incoming links and the size of the article (which are among the current sort criteria), this would be a useful proxy for the importance of a topic, but with an added weight on how internationally known a subject is. OMHalck (talk) 11:51, 28 November 2017 (UTC)

I'd add sorting by number of categories to this wish. That would be a much easier way to identify possible undercategorization in articles. StefenTower (talk) 23:40, 23 January 2024 (UTC)

Parameters to call

So this is almost what I want

call PetScan

But how do I get Combination to be Union and page property Namespace tick switch on (only articiel)? --Traveler100 (talk) 20:37, 17 June 2018 (UTC)

Using magic words?

I have found more than a few pages for different organisms where the talk page is a redirect and the article is not. I thought to run a petscan to search for articles including various taxonomy templates which have talk pages that include #REDIRECT. However, I can't seem to add magic words anywhere. is there a way to do this? --NessieVL (talk) 18:53, 16 August 2018 (UTC)

How do I find new pages only?

Can anybody tell how I can find new pages only with this tool? Tagging "Only pages created during the above time window" does not work for that matter, in fact. Ymnes (talk) 19:00, 18 August 2018 (UTC)

out of order

Petscan doesn't work since at least one week. There's only a white page without any content. Does anybody know why? Many thanks, best regards, Aspiriniks (talk) 11:01, 10 November 2018 (UTC)

It's working fine for me. Nihlus 11:09, 10 November 2018 (UTC)

Maybe depending on the browser? In my case: Iceweasel, which is nearly identical with Firefox. -- Aspiriniks (talk) 11:26, 10 November 2018 (UTC)

I'm trying it for the first time, in Firefox, and either I'm doing something wrong or it isn't working for me. To take a simple example: https://petscan.wmflabs.org/?psid=6710931 isn't finding Commons:Alaska_Theatre_of_Sensations,_A-Y-P,_1909.jpg (or anything else).

Would whoever responds to this please ping me, since I don't keep a watchlist on Meta? Thanks in advance. - Jmabel (talk) 05:51, 7 December 2018 (UTC)

be-tarask

"be-tarask" does not work --Чаховіч Уладзіслаў (talk) 19:39, 9 December 2018 (UTC)

Limit output?

Hi, is there a way to limit /paginate the output? If i try to search for images in certain categories and select "Thumbnail" output, i encounter performance problems with big result sets (file count > 1000). Paginating the output may be a solution, but i didn't find a way to do this. --Fl.schmitt (talk) 11:25, 17 December 2018 (UTC)

@Fl.schmitt:, Yes. I think there is a way. In the tab "output" select "limit"=100. It will show the 100 first images. If you want the 100 last, select "Sort order"=descending. So you can only select 200 images. It would be nice if there were an "offset" parameter to be able to select 100 images starting on a given images. (ex: limit=100, offset=100 , would select the images n. 101 to 201). Perhaps @Magnus Manske:can develop this functionality? --JotaCartas (talk) 19:59, 31 May 2019 (UTC)

API

Powerful tool. Does it have any API with documentation, so I can pass from my tool a list of 90 000 articles and filter it?--Alex Blokha (talk) 21:55, 13 March 2019 (UTC)

Variables

It would be nice that it works with variables if possible. Like the Google's * (asterix) which stays for whatever string.--Juandev (talk) 16:40, 16 April 2019 (UTC)

Previously working query now returning 0 results

It seems like starting ~10 minutes ago, all my queries have been (promptly) returning 0 results. For example here is a simple query that keeps all settings at their default and just searches for pages in the category "Individual eagles". It should return 6 results (and was doing so earlier today), but now it's returning 0. I'm wondering if this is some temporary server issue? Are others seeing the same thing? Colin M (talk) 17:47, 3 June 2019 (UTC)

@Colin M:,Yes, I have having the same problem, but after 3 or 4 attempts I get the correct result. --JotaCartas (talk) 17:58, 3 June 2019 (UTC)

Fixed Ah, cool, it's working again for me too. I guess it was just gremlins. Colin M (talk) 18:05, 3 June 2019 (UTC)

Zero results

PetScan used to work flawlessly, but now returns no results for me, similar to the situation described in the comments directly above me. I thought the situation might resolve itself like it did for Colin M, but it has not. Any insight into why this might be happening?--MainlyTwelve (talk) 16:03, 13 June 2019 (UTC)

Yes, the problem has been getting worse from day to day. At the moment I launch PetScan ... I make some attempts and I will go treat another subject for 5 minutes ... I return ... some more tries ... and I leave again. Sometimes it's only after 30 minutes that I get the result. --JotaCartas (talk) 17:14, 13 June 2019 (UTC)

@MainlyTwelve: , please read ... Some tools on Toolforge may break on or after 3 June because of database changes. Maintainers should update their tools to use the new schema ... in Commons:Commons:Village pump/Technical#Tech News: 2019-24 --JotaCartas (talk) 19:36, 13 June 2019 (UTC)

@JotaCartas: Thank you! I will read it now.--MainlyTwelve (talk) 19:41, 13 June 2019 (UTC)

@JotaCartas:Forgive my ignorance, does that mean we're waiting on the Maintainers? Is there anything I can do to help?--MainlyTwelve (talk) 19:43, 13 June 2019 (UTC)

@MainlyTwelve: The maintainers are expert software developers that maintain sites like Toolforge that allocate some tools (like PetScan) used in all the wikipedias, so .. we have to wait. --JotaCartas (talk) 20:41, 13 June 2019 (UTC)

@JotaCartas: Ah, I see. I will wait. Thanks again!--MainlyTwelve (talk) 20:44, 13 June 2019 (UTC)

@MainlyTwelve:, not at all, here is the correct link to the Techinal News where I read the report of the problem - Commons:Commons:Village pump/Technical#Tech News: 2019-23, regards--JotaCartas (talk) 20:57, 13 June 2019 (UTC)

Any idea when this will be fixed? --Traveler100 (talk) 08:26, 12 July 2019 (UTC)

Some questions about PetScan

I'm linking zhwikisource articles to wikidata these days, and I've encountered some problems. Is it able to import the original pagename (like, zhwikisource) to the label of wikidata page above via PetScan automatically? I've created some items with no lable defined, and the number of them is too large for me to fix them manually. Can anyone help me? --TongcyDai (talk) 15:27, 14 October 2019 (UTC)

Also, I would like to add statements with multiple qualifiers, is it possible to do it via PetScan? --TongcyDai (talk) 15:29, 14 October 2019 (UTC)

Showing zero results, 8 months on

Has PetScan officially died? I'm using this scan which should show this page at a minimum, so I do know it should not come up with an empty set. I therefore can only presume that it's the above-mentioned issue from June of last year. Has this officially died or is there very slow work being done? Jerodlycett (talk) 20:44, 5 February 2020 (UTC)

Syntax of Uses items/props

Hi there, what is the syntax for listing multiple items/props in the Wikidata tab? In other cases like for categories or templates the related textbox contains multiple lines. --Arnd (talk) 13:28, 10 February 2020 (UTC)

Down

petscan.wmflabs.org returns a 504 gateway time-out (nginx/1.13.6) after 60 seconds, for at least a few days that I've been trying it. —[AlanM1(talk)]— 08:27, 8 April 2020 (UTC)

+1 Yesterday and today. Juandev (talk) 10:26, 13 April 2020 (UTC)

+1 Same experience here and now. --MarcoSwart (talk) 10:58, 16 April 2020 (UTC)

And again today. --MarcoSwart (talk) 21:23, 20 April 2020 (UTC)

+1 Same situation at the moment... — WikiLucas (🖋️) 23:12, 23 April 2020 (UTC)

+1 Here. Ahmad^talk 02:27, 24 April 2020 (UTC)

+1 Works one time out of ten, isn't there anyone who can tell us what's happening ? --Cortomaltais (talk) 15:30, 25 April 2020 (UTC)

The page to notify of issues is here : https://bitbucket.org/magnusmanske/petscan/issues?status=new&status=open, but no recent reaction from the owner of the tool. --Cortomaltais (talk) 15:47, 25 April 2020 (UTC)

Also found this notice from around 2015 about a similar tool, CatScan2. It might shed some light on our current problem (or not) : https://tools.wmflabs.org/catscan2/notice.html --Cortomaltais (talk) 16:17, 25 April 2020 (UTC)

+1, also note this tool worked for a while when I checked around 00:00 UTC today. (Although HTTP error 5XX is related to server in general), is it somehow related to heavy traffic? Regards. -- Tito Dutta (talk) 07:28, 26 April 2020 (UTC) Nor working today around the same time. Regards. -- Tito Dutta (talk) 00:51, 27 April 2020 (UTC)

Also see:

• phab:T251567
• d:Wikidata:Contact_the_development_team#PetScan_problem:_504_Gateway_Time-out
  • d:Wikidata:Project_chat/Archive/2020/04#PetScan_problem:_504_Gateway_Time-out
  • de:Wikipedia_Diskussion:Technik/Labs/PetScan#Defekt_2. --M2k~dewiki (talk) 20:17, 30 April 2020 (UTC)

Names of wikis containing hyphen are treated incorrectly

@Renessaince, Sillyfolkboy, and Чаховіч Уладзіслаў: Thread in bitbucket.org: https://bitbucket.org/magnusmanske/petscan/issues/165/names-of-wikis-containing-hyphen-are --Estopedist1 (talk) 06:43, 17 April 2020 (UTC)

Thanks. I asked about this earlier, but probably ticket will do better. --Renessaince (talk) 07:30, 17 April 2020 (UTC)

symbol plus

How to find articles with template's name containing symbol plus ("+")? Help, please [[User:|Игорь Темиров]] (talk) 09:12, 22 May 2020 (UTC)

@Игорь Темиров: maybe this one do the job: https://quarry.wmflabs.org/query/45154 --Estopedist1 (talk) 07:39, 25 May 2020 (UTC)

@Estopedist1: Thank you! Игорь Темиров (talk) 07:47, 25 May 2020 (UTC)

Petscan wikidata creation down?

The last few days when I wanted to use Petscans ability to create wikidata items, when I click "Process commands" the first item is created and then the process just stops and hangs there. Can this please be fixed?--Hjart (talk) 16:13, 2 June 2020 (UTC)

Help requested

I have been unable to get a list of articles that have (or don't have, either will do) a short description (identifiable by en:Template:Short description or en:Category:Articles with short description in the articles) that are in the en:Category:WikiProject SCUBA articles, which is tagged on the talk page. Can someone more skilled then me either show me how to do it or explain why it is not possible? Cheers, · · · Peter (Southwood) ^(talk): 07:01, 27 August 2020 (UTC)

@Pbsouthwood: hmm, the description is on the articles themself while the category is on the talk page of the articles, so technically not the same page. It still should be possible but I'm not entirely sure how... (I only know how to do it the other way around: when the category is on the article and the template on the talk page). Maybe in two times, first you get the list of all articles from the category and then you re-inject the list to filter thoe with/without description? Cheers, VIGNERON * ^discut. 07:13, 27 August 2020 (UTC)

@Pbsouthwood and VIGNERON: with AutoWikiBrowser, I transformed talk page (NS1) to page (NS0). Result in Petscan, see: https://petscan.wmflabs.org/?psid=17236625 --Estopedist1 (talk) 06:45, 1 September 2020 (UTC)

Thanks, the results are helpful, but it looks like it was necessary to use a manual listing so not a simple procedure to repeat periodically. · · · Peter (Southwood) ^(talk): 07:06, 1 September 2020 (UTC)

@Pbsouthwood: The manual (PetScan/en#Articles in a WikiProject) says this is possible with an option called "Use talk pages instead", but I can't find that anywhere in the current interface. Maybe it was removed? @Magnus Manske: Can you offer any insight? Joe Roe (talk) 06:38, 22 October 2020 (UTC)

Ah, I misunderstood the instructions. The "Use talk pages instead" option is for searches by templates, not categories. So @Pbsouthwood: I think you can achieve what you want by searching for pages in Category:Articles with short description that "Have any of these templates:" WikiProject SCUBA. Joe Roe (talk) 09:17, 22 October 2020 (UTC)

HTTP 502

Queries have been failing with "502 Bad Gateway" today. 𝟙𝟤𝟯𝟺𝐪𝑤𝒆𝓇𝟷𝟮𝟥𝟜𝓺𝔴𝕖𝖗𝟰 (𝗍𝗮𝘭𝙠) 12:07, 4 October 2020 (UTC)

unrelated articles in category depth search

https://petscan.wmflabs.org/?psid=18023194 I was essentially searching for short articles in subcategories of the category Philosophers by nationality to find stubs (with a Depth of 3) and noticed multiple articles that didnt belong to any category that would show up as a subcategory of "[country] philosophers"(for example Kristian Rønneberg). I assume its because some category he belongs to was mistakenly added as a subcategory to a philosophy-related category, but I dont know a way to either find that or prevent it from showing up in my search results. Especially since without the article size limit get close to 10000 results and it becomes impossible to sort through manually.

Jonas1015119 (talk) 00:55, 15 December 2020 (UTC)

@Jonas1015119: Category:Philosophers by nationality → Category:Political philosophers by nationality → Category:Agrarianists by nationality → Category:Norwegian agrarianists → Category:Centre Party (Norway) politicians → Kristian Rønneberg. 𝟙𝟤𝟯𝟺𝐪𝑤𝒆𝓇𝟷𝟮𝟥𝟜𝓺𝔴𝕖𝖗𝟰 (𝗍𝗮𝘭𝙠) 12:03, 24 December 2020 (UTC)

Thanks! Though why does that turn up with a depth a 3? I always assumed that would mean it would stop at Agrarianists by nationality, und not go any deeper into subcategories. I really wish Wikipedia wouldnt subcompartmentalize everything so much that you end up with garbage data, you could just put 30000 people into the Philospher category and make it easier. That way I wouldnt end up with an asteroid named after someone in my dataset. Jonas1015119 (talk) 12:59, 24 December 2020 (UTC)

@Jonas1015119: Your PetScan link has depth 4 configured, and this is exactly the fourth level of subcategories. 𝟙𝟤𝟯𝟺𝐪𝑤𝒆𝓇𝟷𝟮𝟥𝟜𝓺𝔴𝕖𝖗𝟰 (𝗍𝗮𝘭𝙠) 13:02, 24 December 2020 (UTC)

didnt realize it started at 0, so 3 is of course 4, whoops. Jonas1015119 (talk) 13:04, 24 December 2020 (UTC)

@Jonas1015119: Not sure what you mean. Depth 0 means no subcategories, depth 1 means the category and its subcategories, and depth 4 (this is the number you have configured in the link) means up to subsubsubsubcategories. 𝟙𝟤𝟯𝟺𝐪𝑤𝒆𝓇𝟷𝟮𝟥𝟜𝓺𝔴𝕖𝖗𝟰 (𝗍𝗮𝘭𝙠) 13:06, 24 December 2020 (UTC)

I was looking at the Minimum links field in the Page Properties header, nevermind Jonas1015119 (talk) 13:10, 24 December 2020 (UTC)

Bug or what: randomly adding "®exp_filter" into fields

Sometimes PetScan adds ®exp_filter into text fields after I hit the "Do it!" button. Those can be found in random fields, like in Linked from or Site links text fields, sometimes in multiple fields and other times just in one field. Any idea what's causing this odd behaviour? Everytime this happens, I'll have to reload a fresh PetScan page to make PetScan work properly because those "®exp_filter"'s break the search. Others have also noticed this: [3] 87.95.206.253 22:05, 20 January 2021 (UTC)

For some odd reason, Petscan sometimes randomly changes regexp_filter to ®exp_filter in Petscan URLs so the "reg" part in URLs is changed to the registered trademark symbol ®, which messes up some Petscan searches when the ®exp_filter is then added to fields. In this Petscan URL this has happened (the %C2%AE is ®):

• https://petscan.wmflabs.org/?cb_labels_any_l=1&project=wikipedia&interface_language=en&search_max_results=500%C2%AEexp_filter=&active_tab=tab_other_sources&edits%5Banons%5D=both&edits%5Bbots%5D=both&language=fi&depth=100&manual_list_wiki=fiwiki&manual_list=Suomi%0ARuotsi%0AVen%C3%A4j%C3%A4&edits%5Bflagged%5D=both&categories=Valtiot&cb_labels_no_l=1&ns%5B0%5D=1&since_rev0=&cb_labels_yes_l=1

and decoded:

• https://petscan.wmflabs.org/?cb_labels_any_l=1&project=wikipedia&interface_language=en&search_max_results=500®exp_filter= *snip*

85.23.79.231 14:57, 3 November 2021 (UTC)

Ah, this seems to be an old known issue: [4] 85.23.79.231 15:05, 3 November 2021 (UTC)

Has been closed, though. ~~~~
User:1234qwer1234qwer4 (talk) 15:34, 3 November 2021 (UTC)

Other discussions about this longstanding annoying bug:

• https://github.com/magnusmanske/petscan_rs/issues/120 (most recent discussion on GitHub)
• en:Wikipedia talk:PetScan#Bug or what: randomly adding "®exp_filter" into fields
• https://github.com/search?q=repo%3Amagnusmanske%2Fpetscan_rs+%C2%AEexp_filter&type=issues (proves that this issue dates at least back to 2019)

2001:14BA:9C35:6600:193F:16A0:BF8A:43A1 18:57, 8 July 2023 (UTC)

Thank you for cross-referencing these discussions. I suggested a simple solution at the second of those three links (not knowing about the parallel discussions). Certes (talk) 19:30, 8 July 2023 (UTC)

I have replaced 'regexp_filter' with 'rxp_filter', which should prevent the issue. 'regexp_filter' is still supported for backwards compatibility. --Magnus Manske (talk) 10:55, 22 December 2023 (UTC)

Adding descriptions

Is it possible to add descriptions using the QS command box?--So9q (talk) 06:14, 18 April 2021 (UTC)

@So9q The QS manual explains it: You use "D" + language code instead of the property identifier. 𝟙𝟤𝟯𝟺𝐪𝑤𝒆𝓇𝟷𝟮𝟥𝟜𝓺𝔴𝕖𝖗𝟰 (𝗍𝗮𝘭𝙠) 12:53, 19 April 2021 (UTC)

Feature request: sort results by Wikidata item ID number

It would be nice to have an option to sort results by their Wikidata item's ID number. For example:

Spain (Q29)

Italy (Q38)

France (Q142)

Russia (Q159)

Germany (Q183)

This sorting option would be really helpful e.g. if you have two separate Petscan lists compiled from two different wikis and you want those two lists to be in same exact order so that you can then easily compare them. You can already do this by choosing Pages with items: Only pages with item and Format: Wiki, and then edit the results' class='wikitable' -> class='wikitable sortable' but that's not very convenient. 85.23.79.231 11:38, 1 September 2021 (UTC)

Is it ok to edit this page?

I wanted to mention some stability issues ("No result for source categories", 502 bad gateway errors) and that petscan is still useful. And hints what triggers them (very large categories, large exclusion categories).

Is it OK to edit this page? I am confused due to weird translation setup Mateusz Konieczny (talk) 14:15, 14 October 2021 (UTC)

Or maybe no categories. Right now I am trying to search files which links my username and I am receiving this error message. Juandev (talk) 10:22, 20 December 2023 (UTC)

Searching for Wikisource works with transcluded content

I want to search for works on English Wikisource which transclude content from the Page: namespace and have a text quality indicator. I've successfully used PetScan to search for pages which use the Page template (link), but I haven't yet figured out how to search for pages using other methods like the <pages /> tag or labeled section transclusion. Can I do this, and if so, how? —CalendulaAsteraceae (discuss • contribs) 06:43, 3 December 2021 (UTC)

Another transclusion method I'd like to search for is direct transclusion using the {{NS:PAGENAME}} format, as seen in The Emancipation Proclamation and Musical Structure as Narrative in Rock. —CalendulaAsteraceae (discuss • contribs) 03:07, 9 December 2021 (UTC)

Negative categories

How can I exclude categories in an URL? Like, I'm using https://petscan.wmflabs.org/?language=de&project=wikipedia&depth=15&categories=Wikipedia%3ADefekte%20Weblinks%2FUngeprüfte%20Archivlinks%0D%0AAdventure&show_redirects=no&sortby=title&interface_language=en&&doit= and I would like to exclude a category but I don't know the name of the variable that matches "negative categories". Thanks and kind geragrds, Grueslayer (talk) 19:01, 1 February 2022 (UTC)

Would anyone know? Kind regards, Grueslayer (talk) 15:29, 25 February 2023 (UTC)

@Grueslayer

Short Answer: &negcats=

Example:

Heads of state of Tuvalu,

but NOT:

Elizabeth II

https://petscan.wmflabs.org/?search_max_results=500&wikidata_item=no&wikidata_source_sites=&regexp_filter=&search_query=&negcats=Elizabeth+II&larger=&show_disambiguation_pages=both&after=&manual_list_wiki=&wpiu=any&labels_any=&cb_labels_no_l=1&subpage_filter=either&maxlinks=&min_sitelink_count=&minlinks=&format=html&referrer_name=&common_wiki=auto&language=en&links_to_no=&outlinks_any=&templates_any=&langs_labels_yes=&pagepile=&categories=Heads+of+state+of+Tuvalu&ns%5B0%5D=1&ores_prob_to=&active_tab=tab_categories&templates_yes=&output_limit=&interface_language=en&project=wikipedia&sortby=none&show_redirects=both&cb_labels_yes_l=1&manual_list=&links_to_all=&sortorder=ascending&cb_labels_any_l=1&search_wiki=

--CmdrDan (talk) 21:31, 23 June 2023 (UTC)

Thanks, CmdrDan! Kind regards, Grueslayer (talk) 05:16, 24 October 2023 (UTC)

Queries for uncategorized media error

For more than a year I've been doing almost daily queries for uncategorized media uploaded from dawiki, nowiki and svwiki. Since a few days I've been getting PageList::run_batch_query: SQL query error[2]: Io(Io(Custom { kind: Other, error: "packet too large" })) errors while attempting this. Any ideas?--Hjart (talk) 16:03, 9 April 2022 (UTC)

Different update frequencies?

Petscan works great for me on German Wikipedia, results always current. However, on Thai Wikipedia it takes many days until a new article shows up in the scan. Are smaller Wikipedias cached less frequently? Ahoerstemeier (talk) 20:33, 13 July 2022 (UTC)

Problems and questions

• Every time I run a search, the language jumps back to English
• Filling in something in the tab Templates&links (at least in German) doesn't work since some time - I always get "No result for source categories" error although it used to work
• How can I search for articles without images? What do the "Lead image" options mean?

Lupe (talk) 01:42, 3 September 2022 (UTC)

"with" property

Hi! The "Link to a pre-filled form for the query you just ran with and without auto-run" when you press with it gives you an error message stares that "No possible data source found in parameters". It was working properly before. We need technical support to fix it soon because I depend on it on monthly basis for wikipedia education program. Best--Avicenno (talk) 17:29, 17 September 2022 (UTC)

Further project "regiowiki.at"

Perhaps it would be possible to implement the project https://regiowiki.at - it's a project about Austria from Wikimedia Austria. Who han can help me ion this question. thx and regards Karl Gruber (talk) 09:46, 24 September 2022 (UTC)

Also regiowiki:Hauptseite --Karl Gruber (talk) 18:03, 24 September 2022 (UTC)

WikiProject example not working

The WikiProject tag example[5] from the Examples section returns no results ("No result for source categories"). Did something change? I'm looking to find the intersection of two WikiProjects when this is resolved. czar 03:52, 4 February 2023 (UTC)

Hi, did you find any solution to this? I'm trying to follow the example to extract a list of all the articles of a certain WP (articles, not talk pages), but same result ("No result for source categories"). Thanks -- Gor1995 (talk) 16:20, 17 January 2024 (UTC)

@Gor1995: Check out my example based on the project articles category. I exclude disambig and redirect categories, I check Talk only on "Page Properties", and then check "Change to page/topic" next to Namespaces on the "Other sources" tab. This gives me a pure article list from the project. StefenTower (talk) 21:42, 20 January 2024 (UTC)

Suggestion to resolve manual restart issue

@Magnus Manske: Hi Magnus, hope you're well — I recently saw some information about the manual restart required for PetScan and wondered if you'd considered using supervisord? I mocked up a possible configuration on GitLab — let me know if you'd like a hand implementing it, or if you have any questions or concerns — TheresNoTime (talk • they/them) 13:06, 10 March 2023 (UTC)

Getting consistent results with SPARQL queries

I learned that to get consistent results with SPARQL queries the query needs to output just QID and no other information such as labels. In use wiki option only Automatic or manually specifying wiki works. "From categories" does not work as stated in the documentation. In the past the query used in WDQS used to work without any changes if I remember correctly. Arjunaraoc (talk) 01:20, 20 April 2023 (UTC)

Page loads but no content is displayed

For the past day or so, I have attempted to access PetScan; however, while the page loads fine, the content consists of only a white screen. I have tried 'zooming in' on mobile, and the scroll bar adapts to this behaviour. However, the normal fields are not present. What should I do? --TheGraciousPedian

PetScan does not work

Dear colleagues, for about two days the PetScan tool seems not to work. If you open the page https://petscan.wmflabs.org/, it remains empty. I noticed this because I wanted to make selections for categorizations, such as this one, based on a fixed PSID: https://petscan.wmflabs.org/?psid=25775450 Thank you very much in advance for help and a fix. Best wishes, -- Radomianin (talk) 07:15, 27 September 2023 (UTC)

• Update: It seems that the application has been working again for a few minutes now. Whoever fixed it, thanks a lot :) -- Radomianin (talk) 08:44, 27 September 2023 (UTC)

Doesn't work

This tool doesn't work. I encountered the following error //This site can’t be reached// Kindly fix it. Thanks Sridhar G (talk) 03:46, 6 October 2023 (UTC)

It'sworking. Thanks Sridhar G (talk) 12:05, 6 October 2023 (UTC)

Creation of items

Hi all, I am happy that PetScan is up again. But I don't understand why creation of items for Wikinews articles doesn't work - nothing happens for me when I click "Start QS". Strangely, if I "uncheck all", QS is started with an empty list of jobs. Infovarius (talk) 08:32, 17 October 2023 (UTC)

sha1 hashing

What hashing algorithm is used in here. ChatGPT 4 say that the hash key is not valid for sha1 nor sha2, because it includes other characters than a-f and does not necesary have 40 chars. Juandev (talk) 23:05, 15 December 2023 (UTC)

So I figured out, that the format differes because SHA1 hashes produced by PetScan are Base36 encoded. But another question might be, weather it is the last version of the file, if the file is changes on Commons. Juandev (talk) 09:39, 27 December 2023 (UTC)

Need Help, for Missing Infobox items

Hi Supports, I was Working All Indian Live Constituencies, I have doubt about the Query to find the missing Infobox parameters, for e-g check this, i want only missing this parameters "| constituency_no" and "| constituency_no = " is blank/empty. - IJohnKennady (talk) 08:10, 15 January 2024 (UTC)

Ability to display wiki-formatted results in a Wikipedia template

I realize I can already cut/paste wiki-formatted results from a PetScan query into, for example, a WikiProject page. For greater utility, I thought it might make sense to have a template that pulls in and displays the results of a PetScan query if provided a PSID. I imagine queries for such a thing might have to be restricted for performance purposes, but at least you already have the built-in page caching. Is this a workable idea, or out of the question for any particular reason? This would be nice for displaying project-related data without ongoing manual efforts. StefenTower (talk) 23:45, 20 January 2024 (UTC)

I have come up with a workaround for this. It's not very easy, but if you can replicate the query used for the PetScan run in Quarry, that query can be used in a database report (available as a template on English Wikipedia). StefenTower (talk) 21:30, 1 July 2024 (UTC)

"By incoming links" sort acting like "randomly"

In this PetScan result, when I sort "by incoming links (ns0)", it acts exactly like the "randomly" choice. Is the former implemented as of yet? I could really use that. StefenTower (talk) 23:53, 23 January 2024 (UTC)

I have an idea of why this may be happening. When I push the above result into a PagePile, and take that into a second query, I can then sort by incoming links. I believe what's happening is the kind of PetScan query I first put together somehow doesn't connect to the data for incoming links, because they don't even show in their column. And so without that data, the sort has nothing to sort by, thereby producing a randomized order. StefenTower (talk) 18:30, 25 January 2024 (UTC)

I have filed a GitHub issue for this. StefenTower (talk) 18:49, 25 January 2024 (UTC)

PetScan - "Wikimedia Cloud Services Error"

Reporting that PetScan is not running - URL = https://petscan.wmflabs.org. JoeNMLC (talk) 00:26, 26 January 2024 (UTC)

This has been logged on Github, too. https://github.com/magnusmanske/petscan_rs/issues/155 William Avery (talk) 07:47, 26 January 2024 (UTC)

"Search filter"

How does the "Search filter" in the "Output" tab work exactly? It's not mentioned in the manual, and on PetScan it only says "Search query, eg "insource"" when it's empty and also that "A search query that will be used as a filter on every single result. Expensive, please use common sense. I don't seem to get it working with regex patterns like -insource:/regex/, so I assume it's just basic text search and I can't include punctuation like colon in it? Maybe someone can write a section for "Search filter" in the manual? 2001:14BA:9CD6:4200:D43C:5ABA:9AD8:104 14:43, 8 April 2024 (UTC)

Pages using a module

Templates&links allows to find pages using a template. Is there any way to search by a module? Some modules don't a wrapping template (en:wikt:Module:Armn-translit for example) so I can't use the template instead. ԱշոտՏՆՂ (talk) 03:52, 19 April 2024 (UTC)

I agree that "Templates&links" should also work with modules; like MediaWiki's search's hastemplate:"Module:Armn-translit" does (example). Someone needs to add a request for this feature to User:Magnus Manske.
But, if you know the code for en.wiktionary for PetScan's "Search" field (because I can't figure it out), you can set the code in that field in this following pre-filled search and get full results if the module is used on less than 10,000 pages (10,000 results is MediaWiki's search's max limit, so you can't get more than that unfortunately, meaning that you'll only get partial results): pre-filled search
I.e. enwiktionarywiki gives: SiteMatrix::get_server_url_for_wiki: Cannot find server for wiki 'enwiktionarywiki', so it's not that code. 2001:14BA:9C40:0:12C:E20D:95CC:ED2 15:24, 17 May 2024 (UTC)

Find uncategorized photo contributions in Commons in a given language

The example given in PetScan#Find uncategorized photo contributions in Commons in a given language unfortunately stopped working a while ago. For a couple of years I had been happily using it to find uncategorized media uploaded primarily through da-, no- and svwiki almost daily. Now it returns completely empty lists. I miss using it. Hjart (talk) 17:18, 27 April 2024 (UTC)

That query looks for "Category:Media needing categories" which has been redirected/renamed to "Category:Files needing categories."

If you change that, does it work as desired? For example, this one works for me, to find 400+ files [6] - Fuzheado (talk) 17:26, 27 April 2024 (UTC)

@Hjart, i think the correct category now is c:Category:Uncategorized files. This is the query for dawiki. ZebaX2010 (talk) 01:39, 3 June 2024 (UTC)

@ZebaX2010 Thanks a lot for pinging me, because I forgot where I posed this question. Also appears that I had a bit of catching up to do. Unexperienced users often fail to correctly add their uploads to wikipedia pages. Hjart (talk) 13:20, 3 June 2024 (UTC)

Media linked from Commons:Files used on OpenStreetMap not in Pages with maps

Hi all, is there a way to search for Commons Files that are linked on Commons:Files used on OpenStreetMap or one of its subpages AND which are not in the Pages with maps category? In other words: images used on OpenStreetMap but lacking coordinates on Commons? Fl.schmitt (talk) 06:12, 17 May 2024 (UTC)

It seems that if a file is linked like this: [[:File:Example.jpg]], i.e. with ":" in the beginning, PetScan's "Linked from" finds them: e.g. PetScan finds only one file on the page commons:Calendar even though there are multiple files. So, if files use the <gallery> tag or just simply [[File:Example.jpg]], PetScan's "Linked from" doesn't find them. Someone needs to report this to User:Magnus Manske. 2001:14BA:9C40:0:12C:E20D:95CC:ED2 13:42, 17 May 2024 (UTC)

Thanks a lot, seems you're right - i've created a new issue for this. Fl.schmitt (talk) 14:41, 17 May 2024 (UTC)

I've just found that a Quarry query together with the PagePile tool could be used as a workaround. Here's a Quarry query to get the list of image links (limit 10 for test reasons) which can be uses as source for a PagePile: #58370. This allows for a negative category search with PetScan (PSID 28323802). Maybe it's useful for anybody else, too. Fl.schmitt (talk) 16:03, 17 May 2024 (UTC)

IT STOPPED WORKING

It did. Whenever I try to run queries it doesn't do anything anymore. So I guess I just have to wait. QuantumFoam66 (talk) 07:06, 28 June 2024 (UTC)

PSID queries stopped working?

Lately, whenever I try to run a https://petscan.wmflabs.org/?psid=28967* I just get a "301 Moved Permanently". Something was moved serverside and can't currently be accessed? Hjart (talk) 16:27, 23 July 2024 (UTC)

@Hjart: It seems that PetScan moved from wmflabs.org to wmcloud.org. Your link still yields an error, but this seems to be due to the trailing asterisk. I'm not sure if PetScan allows for such placeholders instead of pure int psids. Removing the asterisk makes the link work. Fl.schmitt (talk) 09:20, 21 August 2024 (UTC)

@Fl.schmitt I didn't post the full ID, so the above link isn't supposed to work. Right now calling https://petscan.wmcloud.org/?psid=29128308 from a script on my computer returns the same error as above. Hjart (talk) 09:34, 21 August 2024 (UTC)

OK - strange - but your link works, so the psid call seems to be ok. I just tried to curl that URL from my local pc which works fine, too. So I think this isn't a PetScan issue. Fl.schmitt (talk) 09:43, 21 August 2024 (UTC)

@Hjart - sorry, forgot to ping... Fl.schmitt (talk) 10:01, 21 August 2024 (UTC)

"Unstable" result sets for search-based PetScans?

As alread mentioned at the Common's Village Pump, currently there seem to be some disturbing problems with search-based PetScan results:

1. There's a discrepancy between search results on Commons and PetScan search results based on the very same search. For example, a search on Commons with search string Tschubby map incategory:"Media_missing_infobox_template" currently yields 1,087 hits. Executing exactly the same search on PetScan yields between (!) 1,039 and 1,065 results (I did multiple search runs with same psid).
2. Even worse, there's a discrepancy between search results for one and the same PetScan - the tool yields different results, while the commons search result is stable. This makes it hard to use PetScan results (via PagePile) as Input sets for other operations.

Fl.schmitt (talk) 09:27, 21 August 2024 (UTC)

See also issue #179 on github. Fl.schmitt (talk) 09:28, 21 August 2024 (UTC)

Missing "difference" as combination

Hello, I am looking in vain for the "difference" function in the combination selection (subset). Only "intersection" and "union" are displayed. Do I have to make a setting somewhere that I have overlooked? Аныл Озташ (talk) 19:01, 1 September 2024 (UTC)