Spellchecker doesn't need to store ngrams
-----------------------------------------
Key: LUCENE-882
URL: https://issues.apache.org/jira/browse/LUCENE-882
Project: Lucene - Java
Issue Type: Improvement
Components: Other
Affects Versions: 2.1
Reporter: Daniel Naber
Attachments: lucene-spellchecker.diff
The spellchecker in contrib stores the ngrams although this doesn't seem to be
necessary. This patch changes that, I will commit it unless someone objects.
This improves indexing speed and index size. Some numbers on a small test I did:
Input of the original index: 2200 text files, index size 5.3 MB, indexing took
17 seconds
Spell index before patch: about 60.000 documents, index size 13 MB, indexing
took 62 seconds
Spell index after patch: about 60.000 documents, index size 6.3 MB, indexing
took 52 seconds
BTW, the test case fails even before this patch. I'll probaby submit another
issue about how to fix that.
--
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]
