[jira] Created: (LUCENE-2587) Highlighter picks wrong offset for fragment boundaries

["Terje Eggestad (JIRA)" <ji...@apache.org>]

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: dev-unsubscribe@lucene.apache.org
For additional commands, e-mail: dev-help@lucene.apache.org