lucene-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Hoss Man (JIRA)" <j...@apache.org>
Subject [jira] [Commented] (SOLR-4842) Field faceting with local params affects successive field faceting parameters
Date Tue, 21 May 2013 03:55:20 GMT

    [ https://issues.apache.org/jira/browse/SOLR-4842?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13662646#comment-13662646
] 

Hoss Man commented on SOLR-4842:
--------------------------------

Erik: ok, now looking at your test, i think it's just flawed.

Ignore for a minute the issue of faceting multiple ways, ignore the "foo" key in the assertQ
your patch modifies, ignore everything about it, delete it from the test, and just consider
a query using only the "bar" key like so...

{noformat}
      assertQ("ignore foo, look at bar",
              req("q", "id:[42 TO 47]"
                  ,"facet", "true"
                  ,"facet.zeros", "false"
                  ,"fq", "id:[42 TO 45]"
                  ,"facet.field", "{!key=bar " +
                     "facet.missing=true "+
                  "}"+fname
                  )
              ,"*[count(//doc)=4]"
              ,"*[count(//lst[@name='bar']/int)=5]"
              ,"//lst[@name='bar']/int[not(@name)][.='1']"
              );
{noformat}

That test is still going to fail because facet.zeros=false but you are asserting that there
should be 5 terms for "bar".  the only way there could be 5 terms is if you include the terms
with a  zero.

I don't think the docs have never really specified what happens if you mix and match "facet.mincount"
with the deprecated "facet.zeros" (ie: "facet.mincount=1&facet.zeros=true&facet.field=XXX"),
let alone in the case of per-field overrides (ie: "facet.mincount=1&f.XXX.facet.zeros=true&facet.field=XXX")
-- i think it's fair game to say all bets are off in the new situation of localparams.  but
in this specific case, there's no way it makes sense to think that the "bar" key should have
a mincount of "0".

                
> Field faceting with local params affects successive field faceting parameters
> -----------------------------------------------------------------------------
>
>                 Key: SOLR-4842
>                 URL: https://issues.apache.org/jira/browse/SOLR-4842
>             Project: Solr
>          Issue Type: Bug
>          Components: search, SearchComponents - other
>    Affects Versions: 4.3
>            Reporter: Erik Hatcher
>            Assignee: Erik Hatcher
>            Priority: Critical
>         Attachments: SOLR-4842__hoss_tests.patch, SOLR-4842.patch
>
>
> SOLR-4717 introduced local param support for per-field faceting, allowing the same field
to be faceted in various ways.  There's a problem such that one fields local param setting
will override the defaults of the next field.  For example: {code}facet.field={!key=foo facet.mincount=1
facet.missing=false}field&facet.field={!key=bar facet.missing=true}field{code} causes
the "bar" faceting to use facet.mincount=1 when it should use the default of 0.

--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators
For more information on JIRA, see: http://www.atlassian.com/software/jira

---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@lucene.apache.org
For additional commands, e-mail: dev-help@lucene.apache.org


Mime
View raw message