you might want to look here: http://www.teamweaver.org/wiki/index.php/Woogle4MediaWiki
An integration with Semantic MediaWiki is on their todo list, maybe you can help out...
----- "don undeen" <donundeen@...> schrieb:
> so do you think I'll need to write my own search extension? That's
> fine by me; the features I want will be useful in many of my
> Any advice on where to get started?
> I'm looking at the SphinxSearch extension now, just to see what kind
> of function hooks I need.
> But any development documentation would be great too.
> All advice, hints, criticism, beer, pointers, hugs, etc appreciated!
> From: Yaron Koren <yaron57@...>
> To: don undeen <donundeen@...>
> Cc: smw list <semediawiki-user@...>
> Sent: Tue, May 4, 2010 6:09:29 PM
> Subject: Re: [Semediawiki-user] fulltext /displayed text indexer for
> Well, if you store the queried data semantically, there are at least
> two ways to allow for text search of that data: Semantic Forms'
> RunQuery page, and Semantic Drilldown, which includes a free-form
> "combobox" input. But those require the user to know what field to
> search in; is that what you're talking about, or do you want some sort
> of interface where the user can search on every field at the same
> time, like a standard text search?
> On Tue, May 4, 2010 at 3:18 PM, don undeen <donundeen@...>
> >Sorry, I haven't followed up on this thread in some time, but I'm
> coming back to it now.
> >>My initial query was regarding a search engine for SMW that did page
> spidering/indexing, as opposed to searching/indexing the wiki db
> >>This is because I'm doing lots of external data calls, and want that
> information to be searchable as well.
> >>I took Antonio's advice (Thanks Antonio), and starting playing with
> iSearch. It works ok, and was a useful experiment to see what I would
> really get out of a spiderer. However, it seemed kind of non-mature to
> me, and I found myself playing a lot with word definitions and the
> like (I have lots of strings like 123.45.67 that are very important to
> index). Plus, it's not "integrated" with SMW (or MW) in any way; it's
> not an extension. Plus there's limited non-fee support (forum's
> >>Ideally I'd like to be able to do a search that returns matches from
> ExternalData calls, and then lets me use the semantic information on
> the matched pages to do filtering/faceting/drilldown/etc.
> >>I've also seen the htdig MW extension, listed here:
> >>but that hasn't been updated since 2008-11. , and lists support for
> MW up to 1.13 (I'm on 1.15).
> >>looking here,
> >>I don't see any extensions that fit the bill.
> >>It would seem to me that a robust wiki spidering/indexing extensions
> would be something with real utlility, but maybe that's only since the
> advent of these externalData extensions.
> >>anyone have any suggestions? Even a beta project I could help work
> on would be great.
> >>thanks everyone!
> >>Hi, you could try with this one:
> >>"The iSearch PHP Search Engine"
> >>Let me know how it goes, if you do that.
> >>> Hi All,
> >>> I'm using a lot of templates, and ExternalData calls,
> >>> and a problem I'm having is that the search tool seems to search
> >>> the page tables in the database, not the displayed text.
> >>> So, if I've got a work used in a template, and I search on that
> word, I
> >>> won't get a match for pages that use that template.
> >>> And more importantly, if I've got text coming from an ExternalData
> >>> I can't search on those terms.
> >>> However, if I had a spider crawling the wiki pages and indexing
> >>> displayed text, like google or something, then all that text would
> >>> found in a search.
> >>> Are there any existing mediawiki solutions that could help me with
> >>> I guess it doesn't even have to be a "wiki" tool; any good web
> >>> spidering/indexing tool would probably do the trick, right?
> >>> any ideas?
> >>> Thanks!!!
> >>Semediawiki-user mailing list
> WikiWorks · MediaWiki Consulting · http://wikiworks.com
> Semediawiki-user mailing list