No, .hiverc only works for CLI. UDFs are tricky. The only way I can think of is to add them to the function registry (https://github.com/apache/hive/blob/trunk/ql/src/java/org/apache/hadoop/hive/ql/exec/FunctionRegistry.java) and recompile Hive.
On Mon, Dec 10, 2012 at 8:01 AM, John Omernik <j...@omernik.com> wrote: > Will that work for my thrift server connections? > > > On Sun, Dec 9, 2012 at 7:56 PM, विनोद सिंह <vi...@vinodsingh.com> wrote: >> >> Put a .hiverc file in your home directory containing commands, Hive CLI >> will execute all of them at startup. >> >> Thanks, >> Vinod >> >> On Sun, Dec 9, 2012 at 10:25 PM, John Omernik <j...@omernik.com> wrote: >>> >>> I am looking for ways to streamline some of my analytics. One thing I >>> notice is that when I use hive cli, or connect to my hive thrift server, >>> there are a some commands I always end up running for my session. If I have >>> multiple CLIs or connections to Thrift, then I have to run it each time. If >>> I lose a connection to hive thrift, I have to run them. Etc etc. >>> >>> My thought was, is there a way that upon opening a hive cli or connection >>> to a hive thrift server, could I have certain commands be executed? >>> >>> These commands include a use command to get me to a specific database >>> (perhaps there is a default database config variable?) or loading up all the >>> temporary functions I use (UDFs) . >>> >>> For example, I have a UDF to do URL decoding: >>> >>> CREATE TEMPORARY FUNCTION uridecode AS 'org.domain.analytics.URIDECODE; >>> >>> Can I get this to run auto magically at hive cli start or thrift server >>> connection? >>> >>> If not, could we build it in that we can add UDFs to hive without doing a >>> recompile that stay in permanently? >>> >>> I would welcome discussion on this! >>> >>> >> >
