hive-user mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Devopam Mittra <>
Subject need input for heavy duty string manipulation
Date Mon, 16 Mar 2015 05:42:08 GMT
hi there,
Need help / suggestion to formulate a hive-only best feasible/optimized
strategy for:

   - I have a table with list of reserved words - one word per row
   - I have a string wherein I want to search the presence of each reserve
   word and take a count of it .e.g: If there are 5 matches from my reserved
   keyword list , then I expect the query output to be 5.
   - I need to repeat this exercise recursively for all the 'sentences' in
   another table, that has one sentence per row.

How should I do it to get the best (read quickest) performance without
getting too much into recursion , using Hive.

Advise/ suggestions requested.

Environment details: Hive 0.14 / HDP 2.2 , CentOS 6.5 , 2 NN + 6 DN


View raw message