I too am interested in something like this, but not necessarily for ranking. My use case is for selective(search based)-aggregates like
"SELECT SUM(col1) WHERE col2='foo' and col3='bar'. The data structure lucene uses ( per-field-indexing and posting-stream based) seems to be ideal for these kind of use case which no other open-source-data-structure has. I would be more than happy to stand corrected. -Prasen On Fri, Mar 12, 2010 at 10:08 AM, Kay Kay <kaykay.uni...@gmail.com> wrote: > On 03/11/2010 07:56 AM, Stack wrote: >> >> Whats this look like Kay Kay? Implementations of IndexReader and >> IndexWriter? >> > > Yes, true to begin with, primarily applicable for the indexing phase, and an > hbase uri that can be exported to katta , eventually ( like the hdfs uri, > that is being currently supported ). > > > > > > > > >> St.Ack >> >> On Wed, Mar 10, 2010 at 10:05 PM, Kay Kay<kaykay.uni...@gmail.com> wrote: >> >>> >>> Hi - >>> I had initially forked out the component , of indexing a specific column >>> (family) contents in HBase, that was already present and moved it to >>> github. >>> Currently - I am working on an indexer on top of HBase, that I have not >>> yet >>> pushed to github yet , but will do soon, (hopefully with mvn artifacts as >>> well). >>> If it relates to your needs, please feel free to hop onto the same. >>> >>> -- >>> K K. >>> >>> >>> >>> On 03/09/2010 08:48 AM, Otis Gospodnetic wrote: >>> >>>> >>>> Thomas, >>>> >>>> I think it's a fine idea. :) >>>> >>>> See also: >>>> http://github.com/akkumar/lucene-hbase >>>> http://www.simpy.com/user/otis/search/hbase%2520lucene >>>> >>>> >>>> Otis---- >>>> Sematext :: http://sematext.com/ :: Solr - Lucene - Nutch >>>> Hadoop ecosystem search :: http://search-hadoop.com/ >>>> >>>> >>>> >>>> ----- Original Message ---- >>>> >>>> >>>>> >>>>> From: Thomas Koch<tho...@koch.ro> >>>>> To: hbase-user@hadoop.apache.org >>>>> Sent: Tue, March 9, 2010 9:46:39 AM >>>>> Subject: lucene index{reader|writer} on hbase (GSOC idea?) >>>>> >>>>> Hi, >>>>> >>>>> this blogpost >>>>> >>>>> >>>>> backend> >>>>> >>>>> describes lucandra[1], a project to use cassandra as a persistence >>>>> backend for >>>>> lucene indices. From the description in the blogpost and my limited >>>>> knowledge >>>>> of hbase, it seems, that the same thing could easily be written for >>>>> hbase. - >>>>> Is that so? Has anybody done this or want's do to this? >>>>> >>>>> [1] http://github.com/tjake/Lucandra >>>>> >>>>> Thanks for your insights! >>>>> >>>>> Thomas Koch, http://www.koch.ro >>>>> >>>>> >>>> >>>> >>> >>> > >
