-----Original Message----- From: Matthias Korn

Another idea, which comes to my mind just now, would be to have the new pages show the rendered html page rather than the wiki source. Possible? Is there a switch?

Thanks, matsch

_________________________________________________

There are a few ways to do this I think. Do you know that there is an RSS feed available from your wiki pages? I bet you could pass the RSS feed for the recent changes page to twitterfeed or something like that.

You could definitely write a new extension that made a cleaned RSS feed of new pages on your wiki available. (If one hasn't already been written.) That would be useful to anyone who wanted to subscribe with an aggregator and not just follow you on twitter.

Good luck, Courtney

On 2009-08-05 Matthias Korn wrote:

I have now settled with "{[^}]+}" which is far from perfect, because it also matches (partially) on templates. But that's ok.

s/{|.+?|}//gs would do a lot better, since wikitables start in "{|" and end in "|}" and would not touch templates.

/BP

Christensen, Courtney skrev:

Benct Philip Jonsson wrote:

...

On 2009-08-05 Matthias Korn wrote:

...

I have now settled with "{[^}]+}" which is far from perfect, because it also matches (partially) on templates. But that's ok.

s/{|.+?|}//gs would do a lot better, since wikitables start in "{|" and end in "|}" and would not touch templates.

/BP

But that would sometimes grab more than just one table. -Courtney

Not with the lazy quantifier +?: /.+?|}/ means "as few chars as possible up to an "|}"; thus the match will stop at the first "|}", and thus at most one table can participate in each match. Contrast /.+|}/ without the ? which will match everything up to the last "|}" in the string/file. (Ignoring for the sake of uncluttering the /s modifier necessary to have .+? or .+ match spans including newlines!)

/BP

Sam.Sexton@thomsonreuters.com skrev:

I believe that making the match for } non-greedy as below would match just the one table, but I've not tried it.

s/{|.+?|?}//gs

/Sam

I take it you mean s/{|.+?|}?//gs . That would not work because it would make the match stop at the first "|" inside the table, which would probably be part of a caption marker "|+" or a row marker "|-", or a cell marker "|".

The substitution s/{|.+?|}//gs (and no other AFAICT) will delete all tables. To be fool-proof against parser function and template syntax like "{{{1|}}}" it should probably be amended to

s/(?<!{){|.+?|}(?!})//gs

which makes sure that the opening/closing brace isn't preceded/followed by another brace. Incidentally and correctly any /{{+|/ or /|}}+/ *inside* a table will be included in the match and deleted.

/BP

-----Original Message----- From: mediawiki-l-bounces@lists.wikimedia.org [mailto:mediawiki-l-bounces@lists.wikimedia.org] On Behalf Of Christensen, Courtney Sent: 06 August 2009 21:05 To: MediaWiki announcements and site admin list Subject: Re: [Mediawiki-l] REGEXP for wiki tables

Benct Philip Jonsson wrote:

...

On 2009-08-05 Matthias Korn wrote:

...

I have now settled with "{[^}]+}" which is far from perfect, because it also matches (partially) on templates. But that's ok.

s/{|.+?|}//gs would do a lot better, since wikitables start in "{|"

...

and end in "|}" and would not touch templates.

/BP

But that would sometimes grab more than just one table. -Courtney

---

MediaWiki-l mailing list MediaWiki-l@lists.wikimedia.org https://lists.wikimedia.org/mailman/listinfo/mediawiki-l

This email was sent to you by Thomson Reuters, the global news and information company. Any views expressed in this message are those of the individual sender, except where the sender specifically states them to be the views of Thomson Reuters.

---

MediaWiki-l mailing list MediaWiki-l@lists.wikimedia.org https://lists.wikimedia.org/mailman/listinfo/mediawiki-l

Hi,

thanks for all the suggestions. I settled with BP's last solution. Seems to work fine. Templates are not an issue for me atm.

Best, matsch

Am Fri, 7 Aug 2009 14:00:13 +0200 schrieb Jean-Marc van Leerdam j.m.van.leerdam@gmail.com:

Hi,

I don't have anything to add to the REGEXP discussion in itself, but does the OP also have to take into account that the table-start and table-end can be provided by templates, as can the table contents?

In that case, the initial page wiki text only contains template calls, but after template expansion it will have table syntax.