[
https://issues.apache.org/activemq/browse/CAMEL-2287?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=56422#action_56422
]
Roman Kalukiewicz commented on CAMEL-2287:
------------------------------------------
There is also a test that could be discussed called
{{SplitShouldSkipFilteredExchanges}}.
It has flow like:
{code}
from("direct:split")
.split(body(List.class), new MyAggregationStrategy())
.filter(goodWord)
.to("mock:filtered");
{code}
But it should be equivalent to
{code}
from("direct:split")
.split(body(List.class), new MyAggregationStrategy())
.filter(goodWord)
.to("mock:filtered")
.end() ;
{code}
but then {{filter()}} reads like *if* statement that is simply not matched.
Does it mean we shouldn't aggregate it back? And definitely following code
should be aggregated back even if it contains unmatched filter:
{code}
from("direct:split")
.split(body(List.class), new MyAggregationStrategy())
.to("direct:otherprocessing")
.filter(goodWord)
.to("mock:filtered")
.end();
{code}
> Using filter before split causes filtered messages to not be aggregated as
> they where marked as filtered
> --------------------------------------------------------------------------------------------------------
>
> Key: CAMEL-2287
> URL: https://issues.apache.org/activemq/browse/CAMEL-2287
> Project: Apache Camel
> Issue Type: Bug
> Components: camel-core
> Affects Versions: 2.0.0, 2.1.0
> Reporter: Claus Ibsen
> Assignee: Claus Ibsen
> Priority: Minor
> Fix For: 2.2.0
>
>
> Splitter supports filtering exchanges when a filter is used inside it.
> However it should clear the filtered flag for exchanges arriving in, before
> the splitting so Camel has no memory in case filter has been used before the
> splitter
> {code}
> Predicate goodWord = body().contains("World");
> from("direct:start")
> .to("mock:before")
> .filter(goodWord)
> .to("mock:good")
> .end()
> .split(body().tokenize(" "), new MyAggregationStrategy())
> .to("mock:split")
> .end()
> .to("mock:result");
> {code}
--
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.
