Highlighter picks wrong offset for fragment boundaries
------------------------------------------------------
Key: LUCENE-2587
URL: https://issues.apache.org/jira/browse/LUCENE-2587
Project: Lucene - Java
Issue Type: Bug
Components: contrib/highlighter
Affects Versions: 3.0.2
Environment: Java 6 + Lucene 3.0.2
Reporter: Terje Eggestad
Priority: Trivial
I have written a new Fragmenter since we need fragments for hitlines to be on
sentence boundaries and not cross paragraphs.
When using it with org.apache.lucene.search.highlight.Highlighter, I get
hitlines that starts with ". ", "? ", "! "...
Consider the text "A b c d e. F g h i j! K l m n o. "
which become the tokenstream : (A) (b) (c) (d) (e) (F) (g) (h) (i) (j) (K) (l)
(m) (n) (o)
If the fragmenter return isNewFragment() = true on F and K and Highlighter
pick the middle fragment, lets say we search on "g" the hitline becomes:
". F <B>g</B> h i j"
The reason, it seems, is that the offset to the fragment boundaries found by
taking the endOffset of the last token in a fragment ,
not the startOffset of the first.
TJ
--
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]
