From: <Mar...@gm...> - 2003-07-29 20:32:52
|
Hi, I am adapting the library-example in order to search quite large=20 xml-documents for key-words. I want to display the paragraph, where the=20 keyword(s) have been found plus meta-information like the name of the=20 document, the chapter number and the chapter title where the keyword(s)=20 have been found. In order to be able to do so I cannot only retrieve the=20 paragraph which contains the keywords, but I have to retrieve the whole=20 document. In the adapted biblio=5Frdf.xsl style-sheet I am filtering the th= e=20 paragraph that contains the keywords using the exist:match tag. If I find=20 a paragraph that contains the tag I display its content plus the=20 meta-information. In the last stable release 0.9.1 that was working fine.=20 I wanted to upgrade to the unstable release 0.9.2 due to new=20 xpath-functionality. Unfortunately it seems that the tagging with the=20 exist:match tag is not working anymore. It is not contained in the=20 xmldb:result. The query I am using is=20 "document(*)//part[//par[match-all(., 'keyword1', 'keyword2')]]". The part = tag contains the document and the par tag the paragraphs. I have two questions and would be happy about any help: 1. Why does the exist:match tag not work as reliably as in version 0.9.1? 2. How can I access the search criteria from within the biblio=5Frdf.xsl=20 style-sheet? With best regards Marco H=F6dtke |