[jira] [Commented] (DAFFODIL-2171) Delimiter scanning takes into account too many delimiters

["Steve Lawrence (Jira)" <ji...@apache.org>]

    [ https://issues.apache.org/jira/browse/DAFFODIL-2171?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17279172#comment-17279172 ] 

Steve Lawrence commented on DAFFODIL-2171:
------------------------------------------

Need to determine IBM DFDL's behavior and ensure we are consistent with their behavior.

> Delimiter scanning takes into account too many delimiters
> ---------------------------------------------------------
>
>                 Key: DAFFODIL-2171
>                 URL: https://issues.apache.org/jira/browse/DAFFODIL-2171
>             Project: Daffodil
>          Issue Type: Bug
>          Components: Back End
>    Affects Versions: 2.4.0
>            Reporter: Steve Lawrence
>            Priority: Major
>
> When scanning for delimiters, we can sometimes scan for delimiters that we shouldn't. See test {{delmiterScanning_03}} in {{SequenceGroupDelimiters.tdml}} as an example where we scan for comma separator that, while it is in scope, should be ignored because the element is the last in the sequence with an infix separator.
> I've confirmed that the DelimitedStringParser scans for and finds the wrong delimiter, but it's possible that this is a problem in the delimiter stack. Perhaps we need to manipulate the delimiter stack when we determine that the current scope of delimiters should be ignored due to being the last in the sequence? With the correct stack, all the delimiter scanning and delimite consuming sould just work.



--
This message was sent by Atlassian Jira
(v8.3.4#803005)