You are viewing a plain text version of this content. The canonical link for it is here.
Posted to dev@uima.apache.org by "Marshall Schor (JIRA)" <de...@uima.apache.org> on 2018/07/18 15:41:00 UTC

[jira] [Commented] (UIMA-5830) Unable to select from the end

    [ https://issues.apache.org/jira/browse/UIMA-5830?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16547975#comment-16547975 ] 

Marshall Schor commented on UIMA-5830:
--------------------------------------

yes this is a bug, fixing.  It was broken for a bunch of cases; moving the backwards wrapping logic so it applies in all cases.

Also, for cases where there is no "index" specified but the "Type" is an Annotation-subtype, force the use of the Annotation Index.

> Unable to select from the end
> -----------------------------
>
>                 Key: UIMA-5830
>                 URL: https://issues.apache.org/jira/browse/UIMA-5830
>             Project: UIMA
>          Issue Type: Bug
>          Components: UIMA
>    Affects Versions: 3.0.0SDK
>            Reporter: Richard Eckart de Castilho
>            Priority: Major
>
> I don't find a way to implement selection from the end by index using SelectFS - maybe it is a bug or maybe it is not supported:
> {code}
>     String text = "Rot wood cheeses dew?";
>     tokenBuilder.buildTokens(jCas, text);
>     // uimaFIT: JCasUtil.selectByIndex(jCas, Token.class, -1).getCoveredText()
>     assertEquals("dew?", jCas.select(Token.class).backwards().get(0).getCoveredText());
> {code}
> The `uimaFIT` line is how it works in uimaFIT. Below my attempt in doing it using SelectFS. my understanding is that `backwards()` should change the iteration order (so starting at the end of the index instead of at the beginning in this case) and then `get(0)` should get the final token `dew?`. But instead, the statement returns the first token `Rot`.



--
This message was sent by Atlassian JIRA
(v7.6.3#76005)