SOLR-1316 uses a much faster data structure (Ternary Search Tree), not a Lucene index. Using Ngram-based tools like the spellchecker, or your implementation is inherently slower.
Netflix, for example, uses a dedicated TST server farm (their own implementation of TST) to do auto-complete. On Fri, Mar 26, 2010 at 3:32 AM, stockii <st...@shopgate.com> wrote: > > hey thx. > > i think the component runs so far, but i don´t see what it brings me. > > my first autocompletion-solution was with EdgeNGram ... and its exactly the > same result ... > > can anyone, plese show me the advantages of the Issue-1316 ?! > -- > View this message in context: > http://n3.nabble.com/SOLR-1316-How-To-Implement-this-patch-autoComplete-tp506492p661787.html > Sent from the Solr - User mailing list archive at Nabble.com. > -- Lance Norskog goks...@gmail.com
