Re: [Ecm] Howto configure JCR repository in nuxeo 5.2 to support blobs indexing

[Stefan Dimov]

Thank you very much about this explanation and wiki post. You spent me a lot of debugging nights. I just tested and everything is working fine. I had to add some textExtractors to configuration and indexing/searching are perfect for attached documents. My working configuration for workspace.xml is:         <SearchIndex class="org.nuxeo.ecm.core.repository.jcr.jackrabbit.SearchIndex">             <param name="path" value="${wsp.name}/index"/>             <param name="indexingConfiguration" value="${wsp.home}/indexing_configuration.xml"/>             <param name="textFilterClasses" value="org.apache.jackrabbit.extractor.MsWordTextExtractor,                                 org.apache.jackrabbit.extractor.MsExcelTextExtractor,                                 org.apache.jackrabbit.extractor.MsPowerPointTextExtractor,                                 org.apache.jackrabbit.extractor.PdfTextExtractor,                                 org.apache.jackrabbit.extractor.OpenOfficeTextExtractor,                                 org.apache.jackrabbit.extractor.RTFTextExtractor,                                 org.apache.jackrabbit.extractor.XMLTextExtractor"/>         <param name="extractorPoolSize " value="2"/>         </SearchIndex> Where can I find documentation about schema mappings to JCR, i.e. when I contribute schema and new document to nuxeo, how can I understand their mappings to JCR, also about built-in core types? Thank you very much again. Regards, Stefan jiveforums-...@nuxeo.com wrote: If you use the JCR backend, then fulltext configuration is left entirely to Jackrabbit. You may want to read the documentation at http://wiki.apache.org/jackrabbit/IndexingConfiguration Read especially the parts about index aggregates, which is what has to be used to make the fulltext stored in Nuxeo's complex properties (which are JCR subnodes) be indexed with the main document. You should get something like: <aggregate primaryType="ecmdt:File"> <include primaryType="ecmft:content">*</include> <include primaryType="ecmft:content">*/*/*</include> </aggregate> Note that Jackrabbit doesn't use subtyping when matching the aggregate primaryType, so you'll have to repeat the rule for all the document types you have. -- Posted by "fguillaume" at Nuxeo Discussions <http://nuxeo.org/discussions> View the complete thread: <http://www.nuxeo.org/discussions/thread.jspa?threadID=2397#6510> _______________________________________________ ECM mailing list ECM@lists.nuxeo.com http://lists.nuxeo.com/mailman/listinfo/ecm To unsubscribe, go to http://lists.nuxeo.com/mailman/options/ecm -- Stefan Dimov tel/fax : +359 2 987 08 15 mobile: +359 888 52 54 88 web : http://www.setelis.com

_______________________________________________
ECM mailing list
ECM@lists.nuxeo.com
http://lists.nuxeo.com/mailman/listinfo/ecm
To unsubscribe, go to http://lists.nuxeo.com/mailman/options/ecm