[jira] [Commented] (SOLR-6015) managed synonyms don't gracefully handle lowercasing

["Steve Rowe (JIRA)" <ji...@apache.org>]

    [ https://issues.apache.org/jira/browse/SOLR-6015?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13993017#comment-13993017 ] 

Steve Rowe commented on SOLR-6015:
----------------------------------

Tim,

The approach looks great - cool that all ignoreCase changes are now supported.

Two things I noticed:

* One optimization you might consider in {{CasePreservingSynonymMappings.getMappings()}}: in the (very likely overwhelmingly ordinary) case of only one cased version of a key, you could just return the mappings from {{CasePreservingSynonymMappings}} rather than copying them to a new map, since no merging will be required.
* You can drop the overridden {{ManagedSynonymFilterFactory.updateInitArgs()}}, since it just calls the superclass method of the same name (after validating params, which the superclass method also does).



> managed synonyms don't gracefully handle lowercasing
> ----------------------------------------------------
>
>                 Key: SOLR-6015
>                 URL: https://issues.apache.org/jira/browse/SOLR-6015
>             Project: Solr
>          Issue Type: Bug
>            Reporter: Yonik Seeley
>            Assignee: Timothy Potter
>         Attachments: SOLR-6015.patch
>
>
> I've been having bad luck testing new functionallity lately - the first thing I try never works ;-)
> {code}
> /opt/code/lusolr48/solr/example/solr/collection1/conf$ curl -XPUT "http://localhost:8983/solrsis/synonyms/english"   -H 'Content-type:application/json'   --data-binary '{"MB":["MiB","Megabyte"]}'
> {
>   "responseHeader":{
>     "status":500,
>     "QTime":3},
>   "error":{
>     "msg":"Bad Request",
>     "trace":"Bad Request (400) - Unsupported value null for mb; expected single value or a JSON array!\n
> [...]
> {code}
> I finally figured out that if I lowercased "MB" to "mb", then it works as expected.
> Also, it looks like because ignoreCase is true in the initArgs, everything is *stored* as lower case in the managed map (losing information).  Ideally case would be preserved in the synonym file (esp if one wanted to change ignoreCase either way).  I imagine stopwords may have the same issue, but I haven't checked.



--
This message was sent by Atlassian JIRA
(v6.2#6252)

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