Template talk:Wikidata list/Archive 1

Archive 1

How to get values at a certain level

If I want to create a list that is sectioned by p131, how can I make sure it gets a value at a certain level? I want to section this list by Bundesland, but if p131 starts with the town (which is the correct way to go up the administrative tree) then the list of course just gets that value and starts a new section d:Wikidata:WikiProject Medicine/Hospitals Austria. Is there any way to let the algorithm go up the p131 tree and grab the value only if is directly below the country level (i.e. p131 = q40)? --Tobias1984 (talk) 10:35, 16 September 2015 (UTC)

Sorting

Sorting is still a little weird. Upper case characters are sorted before lower case characters. --Tobias1984 (talk) 14:39, 16 September 2015 (UTC)

Allow multiple language outputs, to generate a multilingual wikiproject worklist

Can I suggest an additional feature for this tool? Listeria has the potential to generate worklists in all languages to create a multilingual wikiproject worklist. It would be very helpful if there was an extra option that would allow you to pick multiple languages for the tool to output. Because there would potentially be many languages it would be helpful if the lists were generated with collapsed sections for each language. John Cummings (talk) 13:32, 8 December 2015 (UTC)

WDQ/SPARQL

@Magnus Manske: I would suggest removing Find images link, if there is no WDQ provided, but only SPARQL and you don't have any plans to allow SPARQL in WDFIST or convert SPARQL to WDQ. --Edgars2007 (talk/contribs) 17:19, 17 January 2016 (UTC)

Hello! The "nolink" option in wdq is not working, it returns always a blank output (example). It is useful for generating red link lists. emijrp (talk) 12:47, 17 January 2016 (UTC)

Replying to myself: A fix for this is to use sparql. Example:
{{Wikidata list
|sparql=SELECT ?item WHERE { ?item wdt:P21 wd:Q6581072 . ?item wdt:P27 wd:Q77 . ?sitelink schema:about ?item . FILTER NOT EXISTS { ?wfr schema:about ?item . ?wfr schema:inLanguage "en" } } limit 1000
|sort=label
|columns=number:#,label:Article,description,p569,p570,p18,item
|thumb=100
|links=red
}}
{{Wikidata list end}}

See output here. emijrp (talk) 18:47, 17 January 2016 (UTC)

Hello! I have been messing around with this template in Basque Wikipedia and we find a problem with the links. In small Wikipedias most of the links are not going to work, as we don't have a very big collection of articles. For example if I try to get a list of all the films make by a director, I will get a lot of QXXXXX links. If I put to show the actors of this films the same happens. As the actors name doesn't make any variation in basque (Bud Spencer is eu:Bud Spcencer) it would be great to have a render option where the english (or another language) label appears as a red link. It will be easily readable by the people, and internally it is a red link, no more. Is there any way to achieve this? -Theklan (talk) 15:43, 15 February 2016 (UTC)

Agree, this feature is necessary. I think it is better to make QXXXX links always to Wikidata and other ones to local Wikiproject.--Zygimantus (talk) 23:02, 6 March 2016 (UTC)

No inline references - lists lacking reliability

This tool is being used more often quickly creating many articles that have no references, or no inline references. The author seems interested in generating lists and not worrying about inline references. We've been down this road before, the first years of Wikipedia no one worried about references. That is still being cleared up. Don't make the same mistake twice. References are not an afterthought or minor annoyance, they are central to what Wikipedia and Wikidata provide - reliable information. -- GreenC 00:58, 11 February 2016 (UTC)

Hi Green Cardamon, that's a very good point. Listeria is a useful tool and if it could be changed to show references that would be even better. Not all statements in Wikidata are referenced, but in principle they should be. It might be useful to stipulate which columns in the table need references, or group them in a separate column.
@Magnus Manske: Would it be possible to include references? Nev1 (talk) 17:29, 2 March 2016 (UTC)
Experimental support, see example at User:Magnus Manske/listeria test4. --Magnus Manske (talk) 18:30, 7 March 2016 (UTC)
Excellent. Is the ref name (eg. "'ref_eb48a4a2fb68214c19efb2f752b95a76'") meaningful or it can be anything? -- GreenC 02:05, 9 March 2016 (UTC)
It's just an unique identifier for the ref. --Edgars2007 (talk/contribs) 08:44, 9 March 2016 (UTC)
Ok it was so long I thought it might be coded for whatever the software is doing such as an ID somewhere else. -- GreenC 22:55, 9 March 2016 (UTC)
Great Magnus! emijrp (talk) 20:17, 9 March 2016 (UTC)

SPARQL Prefixes

Can I find the used prefixes for the SPARQL query anywhere? Obviously, they differ from the "standard prefixes" in the Wikidata SPARQL query editor. What I found is "pq:" → "q:", and "ps:" → "v:". Is there more differences? Yellowcard (talk) 13:07, 16 March 2016 (UTC)

Please check in this list the URLs for Museo Casa De Gasperi, Museo Storico Archeologico di Savona, Ecomuseo del Tesino, and Ecomuseo della Valsugana; the text in the URL is the label, instead of the local link: anything wrong? Pietro (talk) 19:27, 15 July 2016 (UTC)

Pxxx/Pzzz, qualifier properties

Hi, Somebody knows if the columns parameter Pxxx/Pzzz, qualifier properties works. According to this test it returns an array. thx, --YB 20:37, 21 March 2016 (UTC)

Is there an example of the template using Pxxx/Pzzz, qualifier properties ? --YB 17:21, 15 December 2016 (UTC)

Multi-page lists

Can I split lists into groups such that an alphabetical list only has A-M on the first page and N-Z on the second page? Thanks. SharkD  Talk  21:44, 11 March 2017 (UTC)

Proposal: Possibility to control the "section level"

Would it be possible to change the "chapter level" for the sectioning from the default level (2) with an additional parameter like section_level.
section_level=3 would mean that the sections would be like this:

=== First section ===

=== Second section ===

and so on

--Larske (talk) 09:09, 7 April 2017 (UTC)

update frequency

@Magnus Manske: Hello, I think that the "freq" parameter isn't working well or the documentation needs clarification. I set the param to 30 days but the page hasn't been refreshed by bot in 2 months (see history). I don't know if the bot counts other user edits too or just those by itself. Regards. emijrp (talk) 20:30, 28 May 2017 (UTC)

Passing QID through to row template?

Hi @Magnus Manske: Is there a way to access the QID of a row in a row template? Requesting 'item' returns a formatted URL rather than just the QID. Thanks. Mike Peel (talk) 23:04, 16 July 2017 (UTC)

Is this talk page really the right place for discussing improvements to the Wikidata list template?

@Capankajsmilyo: Refering to the first line of the box you added at the top of this page on August 27. I have after six months not seen any response at all to the proposal I put on this page. Same goes for the questions from emijrp (talk · contribs) and Mike Peel (talk · contribs) from May 28 and July 16. Is there any other page that we should put the questions or proposals on to get a response? --Larske (talk) 19:41, 8 October 2017 (UTC)

Maybe Magnus's talk page? Thanks. Mike Peel (talk) 20:00, 8 October 2017 (UTC)

Double sort

I want to sort User:Capankajsmilyo/Sridevi filmography by language and year. How to do that? @Magnus Manske: would you like to help? -- Pankaj Jain Capankajsmilyo (talk · contribs · count) 04:34, 25 February 2018 (UTC)

@Capankajsmilyo: Add an "ORDER BY" into the sparql, and then use "section=" to split it into sections. E.g., see User:Mike Peel/Telescopes by country, which sorts by label and splits into countries. Thanks. Mike Peel (talk) 16:07, 15 March 2018 (UTC)

Sort descending

How can I sort a list descending? --Cavernia (talk) 10:25, 15 March 2018 (UTC)

@Cavernia: I think you can do "ORDER BY DESC" in the sparql query. Thanks. Mike Peel (talk) 16:08, 15 March 2018 (UTC)
Yes, it works, thanks! --Cavernia (talk) 20:12, 15 March 2018 (UTC)

Losing rows

When I run the above SPARQL as a query, rows like "Alexander Berkman" will repeat because both enwp and ptwp have GAs on the topic. But when run with this template, only the first instance of the row will display. Any ideas? czar 12:00, 18 June 2018 (UTC)

(I've since received help to remove the need for multiple rows, but the issue still exists in case someone wants to fix/advise for someone else. I've updated the links above to point to the historical version.)   czar 18:31, 18 June 2018 (UTC)

Multiple row with dublicate items

How can I make the possibility to specify the same item several times in single table, for example, in cases where the person held the same post several times at different times? —evs (talk) 02:41, 8 July 2018 (UTC)

Repeating multiple items

I'm writing a SPARLQ query and use this template to get a list of all hugo awards. My query can be found here but the template is deleting duplicated names (check test page): For instance years 1956, 1960, 1962 and 1967 the winner is the same, but the templates shows only the first row. How can I fix it? Mariusmm (talk) 09:21, 16 August 2018 (UTC)

"Not permitted" in mainspace

@Fram: I'm puzzled about how you can in good faith get from "no consensus regarding this has been reached in this discussion" to "not allowed". No consensus means no consensus, not consensus against. Similarly, the multiple AfDs and talk page discussions about the use of ListeriaBot in mainspace have failed to reach a consensus. They didn't, as you assert, all result in its removal: many articles were using it until you went around unilaterally disabling it last January, citing, bizarrely, Wikipedia:Wikidata#Lua module (which also does not say it's "not permitted"). That isn't consensus, it's getting your way through attrition. – Joe (talk) 09:34, 4 December 2018 (UTC)

Do you have links to discussions which actually lead to a consensus that inclusion was allowed? You linked to an AfD, Wikipedia:Articles for deletion/List of female Egyptologists, where you argued for the use of Listeriabot, but the close was "consensus to stop auto-generation from Wikidata". You used this very AfD to argue, at Wikipedia talk:New pages patrol/Reviewers#Lists whose content is entirely WikiData Generated to claim " In short, there is no consensus", which is bizarre. You also link to Talk:List of women linguists, but not to Wikipedia:Articles for deletion/List of women linguists, which again ends with "To the extent that the issue of auto-generating the list from Wikidata has been discussed, people are mostly opposed to it, so I see consensus here to stop that" despite your opposition to the removal of listeriabot. So we have an RfC against the use of Wikidata in the body of articles, while leaving open the option of discussing the use in tables ("it might be worth discussing use in tables specifically – but not consensus regarding this has been reached in this discussion."); and subsequent specific discussions about the use in tables have ended with a consensus for removal (not just of listeriabot: even the use of the Qnumber is no longer allowed in the body of an article). So basiczally, in good faith, it is up to you to try to find a consensus for use of this bot in the mainspace before claiming at discussions that it is allowed to be used, and changing the template and doc page to remove the clear consensus against its use in the mainspace. Fram (talk) 09:43, 4 December 2018 (UTC)
WP:BOLD. If you're going to state that something is absolutely prohibited, the burden is on you to show consensus for that. – Joe (talk) 11:05, 4 December 2018 (UTC)
You are aware that I didn't make that change in the template (I didn't edit this template before today) and that this was the stable version for nearly two years, until you removed it? I did add it to the documentation, in response to the AfDs, and there it has been the stable version for nearly two years as well.
To reiterate: RfC concluded "don't use it in articles" but "discussion for use in table may happen". Some articles used it anyway in tables, but when challenged, these were found to have no consensus for them. So it is clear that the consensus from all discussions so far is that this is not allowed, and that it is up to you to show (by e.g. a new RfC) that consensus has changed and use of this is now acceptable in the main space. Fram (talk) 12:28, 4 December 2018 (UTC)
Let's not misquote Coren who closed Wikipedia:Requests for comment/Wikidata Phase 2. Here is his close:

It is, on the other hand, not appropriate to use Wikidata in article text on English Wikipedia at this time ... There is a valid point raised that while running text is clearly not suitable for Wikidata use, it might be worth discussing use in tables specifically – but not consensus regarding this has been reached in this discussion.

It is perfectly clear that Coren distinguished between "running text" where consensus was against using Wikidata, and "in tables specifically" where nobody could disagree that he stated that no consensus had been reached.
Now, if you can cite another RfC that has changed that no consensus to consensus against, then please give a link to it, otherwise quit edit-warring and accept the sensible compromise that I've offered. --RexxS (talk) 15:27, 4 December 2018 (UTC)
Apart from the fact that in those cases where it was used and then discussed, the consensus was clearly against, no I can't cite yet other cases where the consensus was against. If the RfC says "discuss before using it", and in the cases were it then gets discussed (after it was already used) the consensus is against them, then there is actual consensus against the use, until people wanting to use it can show that the consensus has changed.
The RfC instructs people to discuss before using it, actual practice makes it clear that it shouldn't be used, but still you try to put the burden of evidence on the people arguing against the use of it, and not those wanting to use it? That is rather backwards. Fram (talk) 15:44, 4 December 2018 (UTC)
No, that's exactly how we do things on Wikipedia where there is no prevailing consensus. Editors are free to edit without asking prior permission. I fully accept that in the handful of specific instances where a discussion has taken place, it has gone against using this template in mainspace. What I completely refute is your assertion that you can generalise those instances into a blanket ban on the use of this template in articles. I'm not prepared to stand by and see you attempt to forbid editing in particular articles where there is no existing general consensus against the edit. Be satisfied with an accurate statement of the result of the RfC. You always have the choice of debating the issue on a article-by-article basis (as we do whenever no broader consensus exists), or starting a new RfC to establish that broader consensus. --RexxS (talk) 18:10, 4 December 2018 (UTC)
I'm not surprised. I'll follow BRD then, I guess? Someone adds it (B), I (or, like in the recent cases, someone else) removes (R) the Listeriabot auto-update / overrule local editing aspect, and whoever wants to have it anyway can then discuss it on the talk page or start an RfC. Fram (talk) 05:28, 5 December 2018 (UTC)
Yes! following BRD would be a great improvement. The template had no edits between June and December, and had existed without an error message from its inception until 2 December (three days ago). That's when JMMC89 boldly added the error message, which Joe Roe reverted the following day. What it needed then was discussion here to try to find an acceptable compromise that could gain consensus. What it didn't need was your re-revert to JMMC89's version. That added nothing and was not helpful. If you don't like my suggested compromise, why not try to find another form of words that more closely fits how you view it, by discussion, not by trying to force an already rejected version into the template. The most helpful thing would be to attract more editors into the discussion to get a broader range of views. But Fram, you know all of this already. You'd find life on Wikipedia much more agreeable if you were looking for ways of accommodating others instead of a "my way or the highway" approach. --RexxS (talk) 13:00, 5 December 2018 (UTC)
Hmm, I misread the history on this page (I thought Pppery had added in February 2017, but what was reverted was the edit by JJMC89 from December 2018). So you are right, for this page the BRD was done by Joe Roe, not by me. As excuse I offer that I came to this from Template:Wikidata list/doc, where the edit had stood since January 2017 and was now boldly removed by Joe Roe, and reverted by me. But you're right, I was wrong here (on the history and the BRD issue, not necessarily on the actual issue). Fram (talk) 14:25, 5 December 2018 (UTC)
@Fram: That's okay – errare humanum est. My apologies for misunderstanding where you were coming from.
Seriously, I'm no fan of using a bot to update content from Wikidata, either, because of the lack of filtering and the complications of trying to correct errors. I can see you want a stronger form of words to discourage editors from using it in mainspace, so can you think of better wording than what I suggested that might be acceptable to all of us: Joe, JJMC89 included? --RexxS (talk) 14:37, 5 December 2018 (UTC)
No apology needed, but thanks anyway. Let's leave it at the current wording for now, and just wait and see. If it becomes a problem, then we can always reconsider and change the wording. Fram (talk) 14:46, 5 December 2018 (UTC)
Thanks for the effort to reach a consensus. I am not sure about the error—will it always be visible, or only when editing a page?—but if we have to have one the current wording is an improvement. – Joe (talk) 15:57, 5 December 2018 (UTC)

Escaping characters

I would like to add '?vervaardiger wdt:P485|wdt:P6379 ?item.' as a Query. Is there a way to escape '|'?--Hannolans (talk) 10:56, 20 March 2019 (UTC)

@Hannolans: For use in templates we use {{!}} to encode a | character. That may or may not work in other situations.
It's also worth noting that, in general, the WDQS interface is quite lenient about what it will accept as a query; and that other tools may not be so accommodating. It's often worth clicking the diamond {'Format query') to put the query into a standard format, and then see if it works in Listeriabot, or whatever other place you're trying to use the query. In this case, it changes your query to read ?vervaardiger (wdt:P485|wdt:P6379) ?item. which may be more compatible in other situations. --RexxS (talk) 19:49, 20 March 2019 (UTC)

Use with Pagepile

Is there an option to specify Pagepile when using wikidata query in Wikidata list. For example select all women from needed pagepile? I need such option :). --Alex Blokha (talk) 22:36, 16 June 2019 (UTC)

"Misc." section and final summary "∑ 309 items. " only in english

Is it possible to change the language for the label of the "Misc." section and for the final summary (es:"∑ 309 items.")? For example here, a french page, it would be better to show "Divers" (and not "Misc.") and "∑ 309 entrées." (and not "items")--Arosio Stefano (talk) 07:06, 5 September 2019 (UTC)

How to compile list of multiple values (e.g. a list of X or Y)?

Please excuse my ignorance with coding, but how does one set the template to generate a list of X or Y, or a list of "X and all subclasses of X"? Case in point, Wikipedia:WikiProject Women in Red/Missing articles by occupation/Actresses - US seems to only draw from occupation = actor (Q33999), and ignores subclasses of actors (e.g. film actor (Q10800557) or stage actor (Q2259451)) when actor (Q33999) is not present. Is there a way to specify that all subclasses of actor should also be included. I suppose another workaround is to add "actor" to all film actors, TV actors, etc., but redundancy seems silly, and often results in cluttered infoboxes ("Jane Doe is an actor and film actor and TV actor"). Thanks. --Animalparty! (talk) 17:22, 25 October 2019 (UTC)

@Animalparty: the contents of the list generated by Listeriabot depend on the SPARQL query embedded in the page. You will need to familiarise yourself with the Wikidata Query Service at https://query.wikidata.org/ in order to create the lists you want. In the example you give, take a look at the wikitext for the page and you'll see the SPARQL query used, which looks for entries where occupation (P106) contains the value actor (Q33999). I tried adding film actor (Q10800557), but it timed out for me. You could, of course, just make another list that looks for "occupation=film actor" by simply copying the template used in that page and changing Q33999 to Q10800557. Here's a link to that query on WDQS: https://w.wiki/AmL so you can see it yourself – I got 636 results for "film actor". HTH, --RexxS (talk) 20:09, 25 October 2019 (UTC)
Postscript: Here's the url for a query that combines film actor (Q10800557) with stage actor (Q2259451): https://w.wiki/AmS – it's short enough not to time out and gives 733 results. --RexxS (talk) 20:14, 25 October 2019 (UTC)

Transclude query

Is it possible to transclude query text from somewhere else? Rather than, e.g., editing the same code in multiple queries czar 07:32, 1 January 2020 (UTC)

@Czar: Yes. I used the example in the documentation to create a test in a subpage of this template at Template:Wikidata list/Tests/DutchLighthouses, which transcludes the SPARQL query that I put in Template:Wikidata list/Tests/DutchLighthouses/query. Listeriabot happily updated the first page. --RexxS (talk) 09:17, 2 January 2020 (UTC)
@RexxS, beautiful! Thank you. I'll have to test whether it can combine two query fragments across two transclusions next. czar 02:41, 3 January 2020 (UTC)

Mainspace

@Nikkimaria: Re [1], what discussion? I think this should be removed as per my edit summary: this error message hides the update link; is also shown to readers (although most of the other text should also be hidden for readers - is there a good way to do this?); plus it's inaccurate as there is no consensus not to use it in mainspace, it's just that some people WP:DONTLIKE. Thanks. Mike Peel (talk) 17:19, 8 April 2020 (UTC)

@Mike and Nikkimaria: I've done some searching (User talk:ListeriaBot and ListeriaBot in WP:Village pump (technical), WP:ANI archives) and they point back to the same place, Wikipedia:Requests for comment/Wikidata Phase 2: "It is appropriate to modify existing infoboxes to permit Wikidata inclusion ... It is, on the other hand, not appropriate to use Wikidata in article text on English Wikipedia. Given Fram's "state of Wikidata 2017/18" and your own RfC, I haven't seen any change in consensus from 2013. The SPARQL queries used by {{Wikidata list}} don't filter out unsourced statements, so they don't meet the consensus from your RfC to ensure verifiability. We shouldn't be using this template in articles on enwiki. -- RexxS (talk) 17:58, 8 April 2020 (UTC)
This is for lists, not article text. A modification to only show referenced information would be interesting, though. Thanks. Mike Peel (talk) 18:22, 8 April 2020 (UTC)
Lists are part of article text and are not infoboxes. Nevertheless the closer of the Wikidata Phase 2 RfC did add this rider "There is a valid point raised that while running text is clearly not suitable for Wikidata use, it might be worth discussing use in tables specifically – but not [sic] consensus regarding this has been reached in this discussion. Maybe it's worth exploring that along with filtering the SPARQL results on referencing. The problem then would be the production of incomplete lists that editors could not manually augment. See Wikipedia:Village pump (technical)/Archive 152 #ListeriaBot is making destructive updates - why, and how to remedy this? for an example. --RexxS (talk) 18:42, 8 April 2020 (UTC)
@Pppery: Thanks for [2], that's a step forward. Any chance of hiding the messages for readers, though? I think it's possible (via css perhaps?), but I can't remember how/find the right google search. Thanks. Mike Peel (talk) 18:37, 8 April 2020 (UTC)
Mike Peel, it shouldn't be hidden for readers. The discussion I was referencing was Template_talk:Wikidata_list/Archive_1#"Not_permitted"_in_mainspace which was when the current wording was arrived at. Nikkimaria (talk) 21:34, 8 April 2020 (UTC)
Although the Wikipedia:Wikidata/2018 Infobox RfC discusses use of Wikidata in infoboxes, I think it is worth bearing in mind the arguments raised in the discussion, because many of them were general in nature. The point I would consider most worthy of note is part of the closers' commentary:

... there is a consensus on one point: if Wikipedia wants to use data from Wikidata, there needs to be clear assurances on the reliability of this data. Of the many options, 3A, "Each statement imported from Wikidata must comply with Wikipedia policies", was the only option in the entire matrix in the poll to gain a majority: 52% percent of those expressing a preference in question 3 supported this option. If we define this to mean that the data drawn from Wikidata must be what is generally considered "reliable", then when we combine the share of votes from two other related options, 3C ("Require source") & 3D ("BLPs sourced"), this results in 78% of participants in the poll expressing some form of assurance about the reliability of the material drawn from Wikidata.

This perceived consensus is in agreement with Wikipedia community beliefs. In general, we want the contents of Wikipedia articles to be as accurate as possible, which we ensure by basing articles on reliable sources.

In infoboxes (where there is specific consensus for drawing information from Wikidata), we are required to clearly assure the community that we are importing reliable data. In tables/lists (where there is an acknowledged lack of consensus on drawing information from Wikidata), we should not think that the information imported is free from the constraint that it must be reliable. That raises considerable issues for using this template in mainspace. --RexxS (talk) 01:02, 9 April 2020 (UTC)
Meanwhile, "a list with over a thousand refs is ridiculous"?? @RexxS: I still think the best way to handle this is to show the references, or mark imported information as missing a reference, rather than just not showing it. Thanks. Mike Peel (talk) 18:51, 9 April 2020 (UTC)
Just to be clear: I don't think a list with a thousand refs is ridiculous. I like the idea of showing the references and marking [citation needed] automatically, but ListeriaBot alone is unlikely to be able to provide templated citations. We really need, in addition, an automated tool to do the job of reconciling duplicate references and then transforming them into templated versions as reFill does. Then make sure that it is called every time ListeriaBot does an update. There's quite a bit of work there. --RexxS (talk) 19:10, 9 April 2020 (UTC)
@RexxS: You're not going to like this, but it is straightforward to get ListeriaBot to return a set of calls along the lines of {{RowTemplate|QID}}, which could then be routed into WikidataIB calls. I think I tried this out here a while back, but I can't remember where - but this is how the pages in commons:Category:Uses of Wikidata list as gallery work. So, if the functionality can be added to WikidataIB...? Sorry... Thanks. Mike Peel (talk) 20:00, 9 April 2020 (UTC)
That's fine, Mike, but does that give templated cites? As you might recall, I won't use WikidataIB to return references because I'm not willing to cope with the inevitable blowback from the large number of wikidata-haters who will carp about changing the existing reference style in any number of articles to whichever style I pick to use for rendering citations generated from WikidataIB. It's okay to use a fixed citation format where almost the whole article is a list generated from {{Wikidata list}}, but that could be done by calling a something like {{CiteQ}} for each cite. --RexxS (talk) 20:45, 9 April 2020 (UTC)
@RexxS: Every step taken here seems to upset someone somehow. It's quite quaint, it reminds me of the early 2000's when websites were transitioning from static content to databases. As you know, it's possible to import references from Wikidata using {{Wd}}, but I still think it would be much better if WikidataIB could return them in at least one format. {{CiteQ}} won't work here as the individual references aren't stored on Wikidata as separate items. Thanks. Mike Peel (talk) 21:31, 9 April 2020 (UTC)
@Mike: as I was developing asp-based sites from around 1998, and php-based sites with a MySql backend a little later, I've seen more of the problems that I care to remember. Quaint would not be the word I'd use. It wouldn't be better for me for WikidataIB to return references in a single citation format as I'd get the blame for any misuse, and I don't need ARBINFOBOX3. When our citation modules gain the ability to mimic an arbitrary citation style via parameter, I'll be happy to call them from within WikidataIB to supply references. I have no intention of writing code to duplicate the functionality of Module:Citation plus the ability to mimic arbitrary styles. As for CiteQ, why wouldn't Wikidata store sources as separate items? --RexxS (talk) 21:57, 9 April 2020 (UTC)
CiteQ is for thing like journal articles, where there's a notable reference. In the cases I'm worried about here, it's website URLs that probably aren't notable in their own right, so should be stored as references in individual Wikidata items rather than having their own Wikidata item. But I completely agree with you about citation formats. :-( Thanks. Mike Peel (talk) 22:18, 9 April 2020 (UTC)

I'm not sure what's the point of this warning, but there's no consensus prohibiting every possible use of the template in articles, as shown by the comments above which had to refer to a discussion which is not directly related. to The language "There is no consensus to use Template:Wikidata list in articles" is therefore misleading and Nikkimaria's addition should be reverted until there is a consensus on what to tell users about this template. Nemo 15:05, 10 April 2020 (UTC)

That notice is the stable version. If you'd like to establish consensus for something to replace it you're welcome to try to do so. Nikkimaria (talk) 15:09, 10 April 2020 (UTC)
Mike Peel, if you want to let a bot overrule article edits on enwiki, and force the editors to change Wikidata instead of Wikipedia to get information corrected, then you should start an RfC to get consensus for this. The instances were this was tried with this bot clearly showed that there was no consensus to continue using this, because this caused many problems (with many of the lists the bot created of rather low quality) and confused many editors. Previous Wikidata RfCs have shown that, apart from the infobox, there is usually a consensus against using Wikidata in articles (or even showing links to Wikidata items). Even with infoboxes, it is decided on an individual basis and some have been deleted or restricted. The warning has stood for three years or so, and all that time no Listeriabot articles have existed. If you want to revert this, start a discussion at some general forum (VPP or so), with an RfC if you want to allow the use of Listeriabot (or a similar bot) in the mainspace again. See for example Wikipedia:Articles for deletion/List of female Egyptologists to get a reminder of why this feature was turned off. Fram (talk) 16:19, 10 April 2020 (UTC)

@Mike Peel: @Pppery: I just tried to use this in List of active coal-fired power stations in Turkey. How can I request that only referenced results are returned please? Chidgk1 (talk) 13:22, 20 July 2020 (UTC)

@Chidgk1: It's possible, but complicated to set up. You need to use a query that returns QIDs only, then feed them into Module:WikidataIB. You can see an example at commons:Dytaster - that commons gallery is built using Listeria, and passes the results into commons:Template:Wikidata list/Gallery item that then formats each entry/table row (this is where you can add onlysourced=yes). Thanks. Mike Peel (talk) 18:20, 24 July 2020 (UTC)

Showing local artcile name instead of Wikidata label

In species:List of three-letter genera, on Wikispecies, I have:

|columns=label:Article,item

which results in some entries like:

| [[Uca|Fiddler crab]]
| [[:d:Q1329937|Q1329937]]

whereas I want to display the Wikispecies article title, not the Wikidata label:

| [[Uca]]
| [[:d:Q1329937|Q1329937]]

How should I code this? Andy Mabbett (Pigsonthewing); Talk to Andy; Andy's edits 13:35, 19 July 2020 (UTC)

@Pigsonthewing: I think you want the SPARQL query to return the sitelink as a variable, which you can then use for the link instead of 'label'. I suggest asking this at d:Wikidata:Request a query. Thanks. Mike Peel (talk) 18:24, 24 July 2020 (UTC)
@Mike Peel: Will do; thank you. Andy Mabbett (Pigsonthewing); Talk to Andy; Andy's edits 15:25, 25 July 2020 (UTC)

Adding property qualifiers

I'm trying to produce a list of US Senators in the 116th Congress (https://www.wikidata.org/wiki/Wikidata:WikiProject_every_politician/United_States_of_America/data/Senate/116th_United_States_Congress) but I'm getting multiple start times because there are multiple entries for my column specifier: "P39/Q4416090/P580:Start", but I'd like to add another test for a property qualifier to choose one. Something like the following: "P39/Q4416090/P580;P2937/Q28227688:Start". Is there anything available to do this? Thanks, Gettinwikiwidit (talk) 23:44, 4 August 2020 (UTC)


You need to have a look at the data on Wikidata to see what is going to be returned. For example, Lamar Alexander (Q419976) has multiple values for position held (P39), some are for United States senator (Q4416090) (among others) but one is for (Q98077491). You won't catch Lamar Alexander's start term for the 116th Congress if you're searching for (P39 = Q4416090) with qualifier (parliamentary term (P2937) = 116th United States Congress (Q28227688). I think you need to examine the data set and decide on whether you're going to use (P39 = Q98077491) and simply read the start time from its qualifier, or whether you're going to use (P39 = Q4416090) where qualifier P2937 = Q28227688, or attempt both. There's no way of knowing before time how that data is organised in Wikidata, so perhaps it needs some sanitising before you can use it. --RexxS (talk) 00:52, 5 August 2020 (UTC)
User:RexxS Thanks. I think I understood what was happening. I was asking if it's possible to add extra qualifiers for the column field as in my proposed syntax. I'm guessing the answer is no. The (Q98077491) is new. I added it to work around this issue, though I understand it could be argued that it's the preferred solution to my problem. But given that Wikidata allows multiple triples with the same property, I don't think it's an outrageous request to be able to qualify which of those statements you want. Gettinwikiwidit (talk) 08:54, 5 August 2020 (UTC)
@Gettinwikiwidit: I suggest that the best place to start finding out about what sort of queries are possible is the Wikidata Query Service. You can try out examples and test your search - the one you're using is at https://w.wiki/Yng (please excuse me if you already know this) and you'll find it easier to tinker with and run the WDQS queries directly than via editing a page and running the bot. You may find wikibooks:SPARQL/WIKIDATA Qualifiers, References and Ranks helpful. Cheers --RexxS (talk) 14:51, 5 August 2020 (UTC)
@RexxS: Thanks. I'm not having any trouble querying for the data. This question is specifically about defining columns for Wikidata lists. Would you be able to address this specific question?
@Magnus Manske: Looking at the source code, it looks like there isn't support for qualifying statements as suggested above. Does the suggestion sound reasonable though?

Templatizing Wikidata list items?

Hello,

I created a Wikidata list (Wikidata:WikiProject every politician/United States of America/data/Senate/116th United States Congress) and it seems to be working great. I'm going to need to make a list for each of the 116 United States Congresses. It would be great if I could templatize a single description of the list to ensure that each of these were identical modulo some parameters. I experimented with parametizing the the sparql argument and that seems to work, but it didn't work when I tried using Wikidata:WikiProject_every_politician/United_States_of_America/Columns/Senate_Members_by_Congress for the columns argument. You can see my attempts in the history of my Wikidata list. It's possible that having columns as the last argument causes problems? I'll play around some more, but it would be nice to know if there was a preferred way to do this.

I would have made the whole list a template, but I'm guessing that that might trigger bot runs on the template itself. Any thoughts on how best to achieve this?

Thanks, Gettinwikiwidit (talk) 09:03, 5 August 2020 (UTC)


@Magnus Manske: From the source code it looks like templates are only expanded explicitly on the sparql argument. Might it make sense to transclude the entire file before subjecting it to argument parsing? Gettinwikiwidit (talk) 11:13, 6 August 2020 (UTC)

"Spam/abuse filter" in Listeria bot status

Someone know what kind of pages are reported in Listeria bot status, section Spam/abuse filter? --ValterVB (talk) 09:07, 28 November 2020 (UTC)

Found it: spamblacklist --ValterVB (talk) 18:01, 29 November 2020 (UTC)

Change text style based on statement rank

Currently, when using a property as a column all values of the property are rendered regardless of rank. Especially with deprecated values this can be problematic. I propose rendering deprecated values with strikethrough and preferred values in bold. Of course simply not rendering deprecated statements would be another option. ChristianKl15:57, 8 December 2020 (UTC)

Sorting by a qualifier property

Question originally asked at wikidata:Template_talk:Wikidata_list#Sorting_by_a_qualifier_property

Any ideas on how to sort by a qualifier property (e.g. |sort=P106/Q3809586/P582)? The example below still seems to just sort by ascending QID (see my sandbox). T.Shafee(Evo&Evo)talk 09:58, 3 March 2021 (UTC)

@Evolution and evolvability: As far as I know, Listeria can't do it, but if you order the query (using ORDER BY and so), Listeria uses that order. With SPARQL you can make ordering as complex as you want.--Pere prlpz (talk) 09:43, 11 April 2021 (UTC)
@Pere prlpz: Thanks. I tried using ORDER BY DESC(?end), but listeria seems to ignore that order when building the output here. Would you mind taking a quick look to see if I've missed anything obvious! T.Shafee(Evo&Evo)talk 01:48, 12 April 2021 (UTC)
@Evolution and evolvability: Please check if it works as you wish. Problems fixed:
  • The query must work (in query.wikidata.org) to work for Listeria. This one didn't order by ?end because ?end wasn't selected.
  • Template sort parameters may override query order.--Pere prlpz (talk) 10:03, 12 April 2021 (UTC)
@Pere prlpz: Perfect - thank you. I've got it working as intended now! T.Shafee(Evo&Evo)talk 11:40, 12 April 2021 (UTC)

Freq parameter stopped working

Hello Magnus Manske (talk · contribs). The nice freq parameter does not seem to work correctly since the end of November 2020 (since Wikidata list updated [V2] was introduced). The update interval is much shorter than "freq" days, sometimes only one day, if there is something to update. (Besides, interval would have been a better parameter name, since frequencies usually are measured in events per time unit.) Tomastvivlaren (talk) 12:03, 26 May 2021 (UTC)

@Magnus Manske:. Tomastvivlaren (talk) 18:09, 12 June 2021 (UTC)

Infoboxes in descriptions

In recent updates, I'm getting infoboxes in the description field. See, for example, [3]. MarioGom (talk) 13:15, 18 December 2021 (UTC)

For some items in the gallery on the Commons page Beeldentuin Middelheim Museum, thumbnail images aren't shown. Instead, the parameter code {{{p18}}} is displayed. Senator2029 【talk】 07:28, 23 April 2022 (UTC)

References

If a reference has P248 "stated in" defined, and it is pointing at an object with a website, it will generate the following error in the reference list for that reference: "{{cite web}}: Italic or bold markup not allowed in: |website=", with help link. A fix would be appreciated! robertsilen (talk) 07:28, 5 August 2022 (UTC)

Separator in cells with multiple values

If a table cell contains multiple values (for example several authors of one publication), it would make sense to be able to separate the values with a comma "," to make things clearer. I suggest this to be implemented. robertsilen (talk) 09:03, 5 August 2022 (UTC)