Hello,
I am writing a Java program to extract the abstract of the wikipedia page
given the title of the wikipedia page. I have done some research and found
out that the abstract with be in rvsection=0
So for example if I want the abstract of 'Eiffel Tower" wiki page then I am
querying using the api in the following way.
http://en.wikipedia.org/w/api.php?action=query&prop=revisions&titles=Eiffel…
and parse the XML data which we get and take the wikitext in the tag <rev
xml:space="preserve"> which represents the abstract of the wikipedia page.
But this wiki text also contains the infobox data which I do not need. I
would like to know if there is anyway in which I can remove the infobox data
and get only the wikitext related to the page's abstract Or if there is any
alternative method by which I can get the abstract of the page directly.
Looking forward to your help.
Thanks in Advance
Aditya Uppu
When list=allusers is used with auactiveusers, a property 'recenteditcount'
is returned in the result. In bug 67301[1] it was pointed out that this
property is including various other logged actions, and so should really be
named something like "recentactions".
Gerrit change 130093,[2] merged today, adds the "recentactions" result
property. "recenteditcount" is also returned for backwards compatability,
but will be removed at some point during the MediaWiki 1.25 development
cycle.
Any clients using this property should be updated to use the new property
name. The new property will be available on WMF wikis with 1.24wmf12, see
https://www.mediawiki.org/wiki/MediaWiki_1.24/Roadmap for the schedule.
[1]: https://bugzilla.wikimedia.org/show_bug.cgi?id=67301
[2]: https://gerrit.wikimedia.org/r/#/c/130093/
--
Brad Jorsch (Anomie)
Software Engineer
Wikimedia Foundation
_______________________________________________
Mediawiki-api-announce mailing list
Mediawiki-api-announce(a)lists.wikimedia.org
https://lists.wikimedia.org/mailman/listinfo/mediawiki-api-announce
Hello,
not sure if this is the proper mail list to notify - I am experiencing
problems with data corrupted on wiki dump.
I tried three times to download the last english dump (multistream)
If you're also on wikitech-l, you may have seen the recent announcement[1]
of some work by WMF's Discovery team to improve the search results popped
up by the search box in wiki pages. Assuming it continues to receive
positive feedback, it's rather likely that the code behind their temporary
"action=cirrus-suggest" API module will become the backing engine behind
action=opensearch and list=prefixsearch.
This brought to my attention that "list=prefixsearch" has never been
particularly well-documented, so Gerrit change 260383[2] was submitted and
merged to clarify the issue. Note this isn't a change in behavior, it's a
clarification of the existing behavior to avoid future confusion when that
behavior is improved.
ApiQueryPrefixSearch was added in Gerrit change 123118[3] to be like
action=opensearch but usable as a generator. If your client code is using
list=prefixsearch in the expectation that it would work like the
similarly-named web UI page Special:PrefixIndex, you should review your
usage to make sure that it is actually behaving as you intend. If you are
really wanting a list of titles with a prefix rather than a *search* for
pages "matching" a prefix in some loosely-defined sense, you should be
using list=allpages with the apprefix parameter instead.
[1]:
https://lists.wikimedia.org/pipermail/wikitech-l/2015-December/084356.html
[2]: https://gerrit.wikimedia.org/r/#/c/260383/
[3]: https://gerrit.wikimedia.org/r/#/c/123118/
--
Brad Jorsch (Anomie)
Senior Software Engineer
Wikimedia Foundation
Long ago, the only mechanism for session management in MediaWiki was
certain cookies set by the User class. When ApiLogin was written, in
addition to setting these cookies as usual it also returned some of the
values needed to construct these cookies on the client. Presumably this was
to make things easier for clients that somehow couldn't handle the standard
cookie headers.
Then CentralAuth came along. Now, constructing the cookies manually would
log you in to the local wiki only, without taking advantage of the SUL
mechanism.
Then T55032[1] happened, and clients that were using the
manual-construction mechanism had to update their code because one of the
cookie names changed and that wasn't part of the data being returned.
And soon, we'll have SessionManager and AuthManager, which will make it
possible for login to easily happen in ways that don't involve cookies at
all.
So it's time to eliminate the pretense that clients can manually construct
the cookies instead of handing the standard HTTP cookie headers. Tentative
plan is to deprecate them now and then remove them sometime during 1.28; if
anyone objects to this schedule, please raise your concerns in
https://phabricator.wikimedia.org/T121527.
[1]: https://phabricator.wikimedia.org/T55032
--
Brad Jorsch (Anomie)
Senior Software Engineer
Wikimedia Foundation
_______________________________________________
Mediawiki-api-announce mailing list
Mediawiki-api-announce(a)lists.wikimedia.org
https://lists.wikimedia.org/mailman/listinfo/mediawiki-api-announce
Dear Mediawiki-api,
My web+hybrid mobile app uses the Mediawiki-api to search for Wikipedia
articles associated with a particular latitude/longitude. The searched
coordinates are determined at run-time. In a session, the user may view
several lists of article summaries and several Wikipedia articles.
Article summaries of nearby articles are displayed in a list and, when
one of the summaries is clicked, the associated Wikipedia article is
displayed as-is in a frame that is in my app.
I am not sure if this is acceptable to Wikipedia. Is it?
Also, assuming it is, what sort of credit/attribution is appropriate.
Thanks,
Jim Andrews