camel-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Yuri Schimke (JIRA)" <j...@apache.org>
Subject [jira] Commented: (CAMEL-1857) Improvements to Flatpack DataFormat
Date Wed, 28 Oct 2009 11:05:52 GMT

    [ https://issues.apache.org/activemq/browse/CAMEL-1857?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=55050#action_55050
] 

Yuri Schimke commented on CAMEL-1857:
-------------------------------------

Yes.  It is working well for us.  I can't submit as it is work code, and not complex enough
to be worth running the legal gauntlet.

I have just implemented a DataFormat, that calls flatpack directly and knows about the special
header and footer names.

This is a low priority issue, but I think it is significant when you find the limitations
of the camel components mean you have to write your own with slight customisations.  We have
had to do that in a number of cases.

> Improvements to Flatpack DataFormat
> -----------------------------------
>
>                 Key: CAMEL-1857
>                 URL: https://issues.apache.org/activemq/browse/CAMEL-1857
>             Project: Apache Camel
>          Issue Type: Improvement
>          Components: camel-flatpack
>    Affects Versions: 2.0-M1
>            Reporter: Yuri Schimke
>
> There are some small improvements that would improve the Flatpack DataFormat.  Basically
because the header and trailer are fixed names, and are not exposed via the DataSetList, there
are some cases where we need to write our own code instead of using the FlatPack DataFormat.
> Firstly, it would be good if instead of two special cases "header" and "trailer", if
you could designate your own set of special record ids to be checked e.g. "header", "footer1",
"footer2".  This would allow us to reuse our existing flatpack config files, especially where
we have most than one footer.
> Secondly, since the FlatpackConverter already detects the record id as part of its processing,
it would be nice if it included this as a synthetic key value in the DataSetList e.g. "record_type"
-> "trailer".  
> n.b. I have raised a feature request with FlatPack, so that they expose the current record
type, but I doubt there is any point waiting for it.  If they did implement it, then you would
not need to have any special record types e.g. "header and "trailer"
> https://sourceforge.net/tracker/?func=detail&aid=2827654&group_id=127761&atid=709369

-- 
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.


Mime
View raw message