[
https://issues.apache.org/jira/browse/LUCENE-5472?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
Hoss Man updated LUCENE-5472:
-----------------------------
Attachment: LUCENE-5472.patch
Here's a quick pass at trying to fix this along with a test.
at the moment the test fails because i didn't see any immediately obvious way
to get the fieldname into the exception message, and that seems kind of key to
making it useful (yes a byte prefix of the term is there, but for most people
indexing text that's not going to be immediately helpful to them to understand
where to look for the long term)
I haven't dug down deeper to see if it would be safe/easy to just add the
fieldname to docState.maxTermPrefix (as a prefix on the prefix) nor have i run
any other tests to see if throwing an exception here breaks any other existing
tests that happen to depend on big ass terms being silently ignored.
> Long terms should generate a RuntimeException, not just infoStream
> ------------------------------------------------------------------
>
> Key: LUCENE-5472
> URL: https://issues.apache.org/jira/browse/LUCENE-5472
> Project: Lucene - Core
> Issue Type: Improvement
> Reporter: Hoss Man
> Attachments: LUCENE-5472.patch
>
>
> As reported on the solr-user list, when a term is greater then 2^15 bytes it
> is silently ignored at indexing time -- a message is logged in to infoStream
> if enabled, but no error is thrown.
> seems like we should change this behavior (if nothing else starting in 5.0)
> to throw an exception.
--
This message was sent by Atlassian JIRA
(v6.1.5#6160)
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
