[ 
https://issues.apache.org/jira/browse/SOLR-2002?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=12922102#action_12922102
 ] 

Andrzej Bialecki  commented on SOLR-2002:
-----------------------------------------

bq. Does reload() imply build()? If so, it's a suggester problem, if not, it's 
a SpellCheckerListener problem.
The problem is in Suggester, and it's related to a non-persistent 
implementation of Lookup in Suggester, as opposed to a persistent spellchecking 
index. With spellchecker, when firstSearcher event is processed we want to load 
an existing spellchecking index (that's why it doesn't call build(), but 
reload()). On the other hand, if we want to rebuild the spellchecking index on 
commit, and we process a newSearcher event, then we need to use build(), not 
reload().

So far the code was correct. However, current implementations of Lookup don't 
persist their data, so it's lost after commit and it needs to be rebuilt both 
on firstSearcher and on newSearcher. I created SOLR-2173 to track this issue.

> improve build/tests
> -------------------
>
>                 Key: SOLR-2002
>                 URL: https://issues.apache.org/jira/browse/SOLR-2002
>             Project: Solr
>          Issue Type: Task
>          Components: Build
>            Reporter: Robert Muir
>            Assignee: Robert Muir
>            Priority: Minor
>             Fix For: 3.1, 4.0
>
>         Attachments: SOLR-2002.junit.xml.formatter.patch, SOLR-2002.patch, 
> SOLR-2002_buildfailed.patch, SOLR-2002_core_contrib.patch, 
> SOLR-2002_dih.patch, SOLR-2002_lbhttpsolrserver.patch, 
> SOLR-2002_localization.patch, SOLR-2002_lucenetestcase.patch, 
> SOLR-2002_merged.patch, SOLR-2002_merged.patch, SOLR-2002_merged.patch, 
> SOLR-2002_merged.patch, SOLR-2002_replication.patch, 
> SOLR-2002_testiter.patch, SOLR-2002_testmethod.patch, 
> SOLR-2002_timeout.patch, SOLR-2002setupteardown.patch, 
> TEST-org.apache.solr.spelling.suggest.SuggesterTest.txt
>
>
> we are working on improving some functionality in lucene's build/tests, it 
> would be good to improve the solr side to take advantage of it.
> currently its only sorta-kinda integrated and a bit messy.
> i'd like to do some incremental improvements piece-by-piece on this issue.

-- 
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.


---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]

Reply via email to