INPUT__FILE__NAME, BLOCK__OFFSET__INSIDE__FILE Both virtual column names use TWO underscores .
On Mon, Sep 20, 2010 at 9:26 AM, Edward Capriolo <edlinuxg...@gmail.com> wrote: > On Mon, Sep 20, 2010 at 6:29 AM, lei liu <liulei...@gmail.com> wrote: >> I use INPUT_FILENAME and BLOCKOFFSETINSIDE_FILE, but there is same error. >> hive> select INPUT_FILENAME, BLOCKOFFSETINSIDE_FILE from person1; >> FAILED: Error in semantic analysis: line 1:7 Invalid Table Alias or Column >> Reference INPUT_FILENAME >> >> 2010/9/20 Thiruvel Thirumoolan <thiru...@yahoo-inc.com> >>> >>> I dont think https://issues.apache.org/jira/browse/HIVE-417 which added >>> virtual columns was committed to 0.6. >>> >>> On Sep 20, 2010, at 3:47 PM, Thiruvel Thirumoolan wrote: >>> >>> It should be INPUT__FILE__NAME and BLOCK__OFFSET__INSIDE__FILE. >>> >>> On Sep 20, 2010, at 3:15 PM, lei liu wrote: >>> >>> I use hive0.6 version and execute 'select INPUT_FILE_NAME, >>> BLOCK_OFFSET_INSIDE_FILE from person1' statement, hive0.6 throws below >>> error: >>> >>> FAILED: Error in semantic analysis: line 1:7 Invalid Table Alias or Column >>> Reference INPUT_FILE_NAME error. >>> >>> Don't hive0.6 support virtual columns? >>> >>> >>> >>> >> >> > You have to be careful here. The wiki represents trunk, not a > particular release. > > I started the process of moving the wiki to XDOC documentation for > just this reason, so we can have accurate concise documentation for a > release. version/feature confusion is very prevalent to those outside > of hive. > > Currently the culture is in place for jira and wiki. I do not see it, > because I often spend 10 hours working on a feature. While writing the > xdoc for that feature probably takes about 8 minutes, and updating the > wiki takes about 3. > > I imagine the majority of committers (at Facebook) they have an > internal wiki and their users are less confused. The very astute users > dig deep and figure out how to use the less documented features, but > (here) some people just stop and ask at the first wiki inaccuracy. The > long delay to crank out hive-6 has not helped the issue. If i had to > hazard a guess right now I would say 80% of deployments are running a > trunk between 5 branch and now. > > Edward >
