[ 
https://issues.apache.org/jira/browse/LUCENE-2069?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=12782637#action_12782637
 ] 

Simon Willnauer edited comment on LUCENE-2069 at 11/25/09 9:54 PM:
-------------------------------------------------------------------

bq. damn we have to use the limit form of codePointAt, just to be sure. 
no we don't - at least not in this particular case

bq. if term text truly ends with unpaired lead surrogate, codePointAt could 
pair it with leftover trash trail surrogate from a previous token...

if this rare situation occurs the term length will still prevent the changed 
trail surrogate from being part of the token. This includes a super tiny 
overhead but I guess we can simply ignore this. The lead surrogate will not be 
changed at all in this case - if there is a situation where this could happen 
I'm not aware of it!

      was (Author: simonw):
    bq. damn we have to use the limit form of codePointAt, just to be sure. 
no we don't - at least not in this particular case

bq. if term text truly ends with unpaired lead surrogate, codePointAt could 
pair it with leftover trash trail surrogate from a previous token...

if this rare situation occurs the term length will still prevent the changed 
trail surrogate from being part of the token. This includes a super tiny 
overhead but I guess we can simply ignore this.
  
> fix LowerCaseFilter for unicode 4.0
> -----------------------------------
>
>                 Key: LUCENE-2069
>                 URL: https://issues.apache.org/jira/browse/LUCENE-2069
>             Project: Lucene - Java
>          Issue Type: Improvement
>          Components: Analysis
>            Reporter: Robert Muir
>            Assignee: Robert Muir
>            Priority: Minor
>             Fix For: 3.1
>
>         Attachments: LUCENE-2069.patch, LUCENE-2069.patch, LUCENE-2069.patch, 
> LUCENE-2069.patch, LUCENE-2069.patch
>
>
> lowercase suppl. characters correctly. 
> this only fixes the filter, the LowerCaseTokenizer is part of a more complex 
> issue (CharTokenizer)

-- 
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: java-dev-unsubscr...@lucene.apache.org
For additional commands, e-mail: java-dev-h...@lucene.apache.org

Reply via email to