[LUCENE-23] GermanStemFilter setting wrong values for startoffset/endoffset of stemmed tokens

• Type: Bug
• Status: Resolved
• Priority: Major
• Resolution: Fixed
• Component/s: modules/analysis

Environment

Operating System: Linux
Platform: PC

• Bugzilla Id: 7412
• Assignee: Lucene Developers
• Reporter: Rodrigo Reyes
• Created: 25/Mar/02 00:07
• Updated: 27/Aug/24 15:34
• Resolved: 27/May/06 01:35

Description

The GermanStemFilter sets wrong values to the new Token object created when the
stemmer succeeds in stemming the termText() string. Bug found in 1.2-RC5-dev

-----------------
Example, for the processing of the string "this is a simple test":
token : thi (0,3)
token : is (5,7)
token : a (8,9)
token : simpl (0,5)
token : test (17,21)

(all the stemmed tokens have wrong start/end offsets).