[
https://issues.apache.org/jira/browse/UIMA-3115?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13723864#comment-13723864
]
Peter Klügl commented on UIMA-3115:
-----------------------------------
There are two use cases: the inner rules are only applied in the context of the
matched context, like the BLOCK construct right now, and the outer context rule
element can only match if the inner rules match. These two functionalities need
to be distinguished in the snytax of the language, maybe like:
{noformat}
Sentence{-> T1}->{CW NUM;};
// like BLOCK right now: match on each sentecne and annotate it with T1, then
apply inner rule within
// the matched annotation
Sentence{-> T1}<-{CW NUM;};
// try to match each sentence, which succeeds only if it contains a valid match
of the inner rule. Then, create a new annotation.
{noformat}
> fusion of block syntax and annotation specification syntax
> ----------------------------------------------------------
>
> Key: UIMA-3115
> URL: https://issues.apache.org/jira/browse/UIMA-3115
> Project: UIMA
> Issue Type: Improvement
> Components: ruta
> Reporter: Georg Fette
> Assignee: Peter Klügl
> Priority: Minor
>
> It would be a nice syntax improvement if block definitions and annotation
> specifications with curly brackets are written with the same syntax. I.e.:
> {noformat}
> Document { CW NUM { -> MARK(MyAnnot) } }
> {noformat}
> =
> {noformat}
> BLOCK(MyBlockName) Document { CW NUM { -> MARK(MyAnnot) } }
> {noformat}
> The style without the block syntax will be an anonymous block but that would
> be okay.
--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators
For more information on JIRA, see: http://www.atlassian.com/software/jira
