[LUCENE-4821] AnalyzingSuggester should use end() offset to decide if last token ended or not

• Type: Improvement
• Status: Closed
• Priority: Major
• Resolution: Fixed
• Fix Version/s: 4.3, 6.0
• Component/s: modules/spellchecker
• Lucene Fields: New
• Assignee: Michael McCandless
• Reporter: Michael McCandless
• Created: 10/Mar/13 16:51
• Updated: 28/Aug/22 13:40
• Resolved: 11/Mar/13 21:26

Description

For example, today if you index "i love lucy" and "isla de muerta", and then you ask for suggestions for "i" and for "i " (space after the i) you'll get the same results.

But if we use the ending offset, we can determine (I think?) that there were non-token characters after the last token, so that "i " would only suggest "i love lucy".