You are viewing a plain text version of this content. The canonical link for it is here.
Posted to dev@lucene.apache.org by "Adrien Grand (JIRA)" <ji...@apache.org> on 2013/01/15 15:00:15 UTC
[jira] [Closed] (LUCENE-4685) IndexedField (is StringField) set to
tokenized when getting the Document
[ https://issues.apache.org/jira/browse/LUCENE-4685?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]
Adrien Grand closed LUCENE-4685.
--------------------------------
Resolution: Invalid
Hi Rob
bq. It seems the tokenized vs. non tokenized is not stored in the Lucene40StoredFieldsWriter
Right, the StoredFieldsWriter only stores the type of the data (binary|String|int|long|float|double).
bq. Maybe the bits can be used to determine just that? (it seems there is room left?)
While we could add a bit to store this information, I think that the best solution to your problem is to manage a schema on top of Lucene, similarly to Solr and ElasticSearch. Storing every information required to re-index the field would waste a lot of space (precisionStep for numeric types, analyzer for text fields, ...).
I know the current behavior is a little confusing, but the future major release of Lucene (5.0, not scheduled yet) will use a different return type for IndexReader.document(int): StoredDocument (instead of Document) whose fields don't expose any indexing option.
> IndexedField (is StringField) set to tokenized when getting the Document
> -------------------------------------------------------------------------
>
> Key: LUCENE-4685
> URL: https://issues.apache.org/jira/browse/LUCENE-4685
> Project: Lucene - Core
> Issue Type: Bug
> Components: core/index
> Affects Versions: 4.0
> Reporter: Rob Audenaerde
> Attachments: TestLuceneTokenizedFieldDocument.java
>
>
> I have added a Document with StringField to the index. When I search for it, I retrieve the Document from the searcher and iterate the IndexedFields. The field I have added is reported to be a 'tokenized' field.
> This should not be the case, as it is a StringField and these are not tokenized.
> It is very unconvenient when updating a Document with an extra field. The original fields are all set to tokenized.
> I will add a TestCase that shows the unwanted behaviour
--
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