From: Chris D. <ce...@ui...> - 2008-07-02 15:25:25
|
On Tue, Jul 01, 2008 at 02:59:59PM -0400, Wayne Graham wrote: > Hi Chris, > > Actually, this is taken care of in Solr (that's included in Vufind). > Grab the most recent revision with the solr stuff (the actual parsers > are in the $VUFIND_HOME/solr/lib folder. BTW: adding "marc.unicode_normalize" to the import.properties file seemed to "fix" the decomposed unicode characters. On a separate matter, I still can't seem to get my facets to work on these records. If I run solr/admin/analysis using the following URL: http://localhost:8080/solr/admin/analysis.jsp?nt=name&name=author&verbose=on&highlight=on&val=P%C3%A9rez&qverbose=on&qval=P%C3%A9rez I get unexpected results. I will provide URLs to screen shots since the web page will not save correctly. BTW, I am testing VuFind out of the box with only the single "Perez" record. Any hints as to what I should try next? VuFind demo site: http://vufind-devel.ilcso.uiuc.edu/vendor SOLR analyzer page 1: http://vufind-devel.ilcso.uiuc.edu/vendor/images/solranalyze1.jpg SOLR analyzer page 2: http://vufind-devel.ilcso.uiuc.edu/vendor/images/solranalyze2.jpg SOLR analyzer page 3: http://vufind-devel.ilcso.uiuc.edu/vendor/images/solranalyze3.jpg Thanks! Chris > > Let me know how it goes! > > Wayne > > Delis, Christopher wrote: > >I tried the latest solrmarc build > >(http://solrmarc.googlecode.com/svn/trunk rev.38) and ran a test load for > >a record containing decomposed unicode (pe'rez), but it didn't appear to > >put them into composed form. In luke, I see references to ? (ampersand > >pound-sign seven six nine semi-colon, which is a COMBINING ACUTE ACCENT). > >Shouldn't I see a á (ampersand pound-sign two two five semi-colon, which > >is a LATIN SMALL LETTER A WITH ACUTE) instead? Am I checking out the > >correct code? > > > >Thanks, > >Chris > > > > > >-----Original Message----- > >From: vuf...@li... on behalf of Wayne Graham > >Sent: Fri 6/27/2008 11:25 AM > >To: vuf...@li... > >Subject: [VuFind-Tech] Solr 1.3 SVN > > > >Hi folks, > > > >I've made some changes to the Solr instance. First, I just committed the > >nightly build of Solr 1.3 into the trunk. If you want to do the > >indexing, please use SolrMarc (see > >http://code.google.com/p/solrmarc/source/checkout), until I can get it > >merged into the project properly. > > > >You can use solrmarc's index_file.sh like so: > > > >./index_file.sh /usr/local/vufind/import/catalog.mrc > >./importVufindSample.properties > > > >Also, I added some code that Robert Haschart had written that addresses > >the issues of decomposed unicode (Pe'rez) to its composed form (Pérez). > > > > Wayne > > > > > > -- > /** > * Wayne Graham > * Earl Gregg Swem Library > * PO Box 8794 > * Williamsburg, VA 23188 > * 757.221.3112 > * http://swem.wm.edu/blogs/waynegraham/ > */ > > |