Hi, Ben Saller would know better, but there are already plugins for extracting text from pdf, odt and doc.
Alternatively, all metadata properties whose key finishes with :text should be indexed and be searchable with fulltext. In python it would be something like: self.metadata['summary:text'] = 'text I want to be indexed' Cheers, Tomeu On Thu, 2007-07-12 at 15:47 +0200, Bert Freudenberg wrote: > Then there need to be datastore plugins for different document types. > Like, Open Office documents are zipped, so it needs to look inside them. > > - Bert - > > On Jul 12, 2007, at 15:17 , Marco Pesenti Gritti wrote: > > > The datastore is doing fulltext indexing of the documents. It's not > > quite working yet I think but that's the plan. > > > > On 7/12/07, Bert Freudenberg < [EMAIL PROTECTED]> wrote:Is it > > anticipated to have the journal search by full text (which > > seems useful to me if the kid does not bother to provide explicit > > keywords when sav^H^H^Hkeeping)? If so, should an activity store a > > text representation of the document as metadata, or will there be a > > way to extract text/metadata from known document types? > > > > - Bert - > > > > > _______________________________________________ > Sugar mailing list > [email protected] > http://lists.laptop.org/listinfo/sugar _______________________________________________ Sugar mailing list [email protected] http://lists.laptop.org/listinfo/sugar
