Minimize use of fields in lucene index
--------------------------------------
Key: JCR-106
URL: http://issues.apache.org/jira/browse/JCR-106
Project: Jackrabbit
Type: Improvement
Components: query
Environment: svn revision: 161184
Reporter: Marcel Reutegger
Priority: Minor
Currently every property name creates a field in the lucene index, bloating the
size of the index because of the norm files created for each field.
When values are indexed as is (not tokenized for fulltext indexing), then the
property name may be part of the term text. That way lucene must only maintain
one field for all property names. With this approach the search terms are
always a combination of property name and literal value. e.g. instead of using
TermQuery(new Term("prop", "foo")) the query must be TermQuery(new
TermQuery("common-field", "prop:foo")). this works for general comparison /
value comparison operators and also for the like function. the contains
function uses the fulltext index which uses a different field anyway.
Using the property name as part of the indexed term text, requires a custom
SortComparator which is aware of the property name.
This change will not be backward compatible with earlier indexes created by
jackrabbit.
--
This message is automatically generated by JIRA.
-
If you think it was sent incorrectly contact one of the administrators:
http://issues.apache.org/jira/secure/Administrators.jspa
-
If you want more information on JIRA, or have a bug to report see:
http://www.atlassian.com/software/jira
