@Ebonetti90 Thanks for the report, do you get:

{"error":{"code":"nosuchsection","info":"Sections are not supported by Wikipedia:知识问答.","docref":"See https://zh.wikipedia.org/w/api.php for API usage. Subscribe to the mediawiki-api-announce mailing list at <https://lists.wikimedia.org/mailman/listinfo/mediawiki-api-announce> for notice of API deprecations and breaking changes."},"servedby":"mw1313"}

For the query:

https://zh.wikipedia.org/w/api.php?action=query&prop=extracts&exlimit=20&exintro=&explaintext=&exchars=512&format=json&formatversion=2&pageids=4488064

This is what I get:

curl 'https://zh.wikipedia.org/w/api.php?action=query&prop=extracts&exlimit=20&exintro=&explaintext=&exchars=512&format=json&formatversion=2&pageids=4488064'
{"batchcomplete":true,"query":{"pages":[{"pageid":4488064,"ns":0,"title":"Wikipedia:知识问答","extract":"…"}]}}

?

Could you share your request headers, that may be the difference to reproduce the problem (curl -v)? Thank you!

Odd, I get the nosuchsection error while logged in to my usual account (which has no special privileges on zhwiki, as far as I know) but not when accessing the URL while logged out. The page's content model is reported as wikitext, so it should support sections. I don't have time to debug further now, but I will look at it later on.

It was some weeks that I was aware of this issue: whatever magic you did, now it works for me too, thanks!

I'll let you know if ever anything arises again for the public API use 👍

Enrico

I do reproduce, however, T215028

curl 'https://el.wikipedia.org/w/api.php?action=query&prop=extracts&exlimit=20&exintro=&explaintext=&exchars=512&format=json&formatversion=2&pageids=298785'
{"error":{"code":"missingtitle","info":"The page you specified doesn't exist.","docref":"See https://el.wikipedia.org/w/api.php for API usage. Subscribe to the mediawiki-api-announce mailing list at <https://lists.wikimedia.org/mailman/listinfo/mediawiki-api-announce> for notice of API deprecations and breaking changes."},"servedby":"mw1277"}

That page is weird, because I can see its revision:

https://el.wikipedia.org/w/index.php?oldid=3925388

But I cannot see the page?

https://el.wikipedia.org/wiki/%CE%92%CE%A0:WDATA
https://el.wikipedia.org/w/index.php?title=%CE%92%CE%A0:WDATA&redirect=no

Yep yep, this is a different kind of error ;)

Maybe the global page prefix has been changed from "ΒΠ:" to "Βικιπαίδεια:", leaving trace of the old page in the latest dump and in the page API, so that when extracts API is called, it finds an inconsistent situation.

Enrico

In T215030#4920677, @jcrespo wrote:

That page is weird, because I can see its revision:

https://el.wikipedia.org/w/index.php?oldid=3925388

But I cannot see the page?

https://el.wikipedia.org/wiki/%CE%92%CE%A0:WDATA
https://el.wikipedia.org/w/index.php?title=%CE%92%CE%A0:WDATA&redirect=no

https://el.wikipedia.org/w/api.php?action=query&revids=3925388 provides the answer. That revision is associated with the page "ΒΠ:WDATA" in namespace 0, but since "ΒΠ" is an alias for the "Βικιπαίδεια" namespace there the title parses as page "WDATA" in namespace 4, making revision 3925388's page not accessible by title. More details on T215028#4920713.

It turns out that page ID 4488064 on zhwiki is in the same situation: that page's title is "Wikipedia:知识问答" in namespace 0, but that title actually corresponds to "知识问答" in namespace 4 (page ID 5867468), which is a Flow board, which indeed doesn't support sections. Apparently TextExtracts is trying to fetch the page by title-string even though it was requested by page ID.

As with T215028, namespaceDupes.php would clean up the inaccessible page. But probably someone should also look into TextExtracts's behavior (although I suspect the nominal maintainers' response will be along the lines of "We don't actually support the extension because we moved our specific use case to a microservice behind restbase. Feel free to submit a patch yourself.").

Thanks a lot! I have filed a task for running the script.

Enrico

See T87645: Existing pages without ability to reach and obviously wrong namespace.

As these should be solved together