Thanks a ton, Robert. I checked out the latest nightly and changed the following in my solrconfig.xml:
<luceneMatchVersion>LUCENE_33</luceneMatchVersion> to <luceneMatchVersion>LUCENE_40</luceneMatchVersion> The new SynonymFilter loaded all the 1.9 million lines of synonyms in less than 5 minutes! Awesome! Thanks to all who developed this "huper duper" fast synonym filter! On Thu, Aug 4, 2011 at 5:01 PM, Robert Muir <rcm...@gmail.com> wrote: > https://issues.apache.org/jira/browse/LUCENE-3233 > > On Thu, Aug 4, 2011 at 7:24 PM, Arun Atreya <my.2.pai...@gmail.com> wrote: > > Hello, > > > > I would like to know the best way to load a huge synonym list into Solr. > > > > I would like to do concept indexing (a.k.a category indexing) with Solr. > For > > example, I want to be able to index all cities and be able to search for > all > > of them using a special keyword, say 'CONCEPTcity', where 'CONCEPTcity' > will > > match anything that IS-A city, as specified in the index_synonyms.txt > file. I > > believe the best way to do this is via the SynonymFilterFactory and do > > index-time synonym expansion. Or is there a better alternative? > > > > I would still like to keep the original city names and do not want to > > replace them with 'CONCEPTcity', so if someone searches for 'Lake', the > city > > name 'Salt Lake City' still matches. Also, obviously, I do not want two > > different city names to be synonyms of each other. > > > > Is the correct way to specify the index_synonyms.txt file like this? > > > > ----------------------------------------------------- > > CONCEPTcity, Salt Lake City > > CONCEPTcity, New York > > CONCEPTcity, San Jose > > . > > . > > . > > ----------------------------------------------------- > > > > and then keep > > expand="true" > > for SynonymFilterFactory? > > > > I tried to load a synonym file with 10K entries like this, and Solr/Jetty > > took a few seconds to start, but if I try to load a synonym file with 1M+ > > entries, then it is taking a long time. What is the best way to do this? > > > > Thanks, > > Arun. > > > > > > -- > lucidimagination.com >