Hi Claus

I updated the unit tests of SplitterPOJOTest and added some comments for
this feature, I think the wiki will be updated when the confluence
exports the static page.

Willem

Claus Ibsen wrote:
> On Mon, May 11, 2009 at 2:34 AM, Willem Jiang <willem.ji...@gmail.com> wrote:
>> How about return a List<Message> from the POJO bean method?
>> Then we can check the List Object in the Splitter's
>> createProcessorExchangePairsList() and
>> createProcessorExchangePairsIterable(), if the object in the list is
>> Message, we can set the Exchange's InMessage
>> with the Message object that we get from the list.
> Willem
> 
> A very good idea. Glad you thought of this solution. Very concise and clean.
> And leveraging the existing Camel API.
> 
> Do you mind updating the Splitter EIP wiki page with this new feature?
> 
> 
>> Willem
>>
>> Claus Ibsen wrote:
>>> On Fri, May 8, 2009 at 3:28 PM, rohitbrai <rohitb...@gmail.com> wrote:
>>>> Can't I achieve something similar using bean in the splitter as explained 
>>>> in
>>>> Using a "Pojo to do the splitting" on http://camel.apache.org/splitter.html
>>> Yeah I have thought of that one too, but I cannot see how you should
>>> be able to alter the message for each of the individual splitted new
>>> message.
>>> As we only return the body as result.
>>>
>>> And in java we cannot return 2 types, 1 for the body, 1 for the headers
>>>
>>> We might be able to introduce some convention and let you return some
>>> object holder the body and the header.
>>> List<BodyAndHeaderHolder>
>>>
>>> But then I might get a bit ugly?
>>>
>>> Any thoughts?
>>>
>>>
>>>
>>>
>>>> I was trying to understand the code, but am wondering how will I get back
>>>> the whole message. The example there sends a arraylist of string, but I
>>>> would want messages with the existing headers intact.
>>>>
>>>> Another solution I am thinking of is making the message a object and not a
>>>> string and passing the to address string as part of it.
>>> Yeah that works perfect. Or you can temporary return the To header in
>>> the 1st line of the body response
>>> and then fix it in a POJO afterwards
>>>
>>> from(x).split(MySplitterBean).bean(MyFixUpBean.class).to(z)
>>>
>>> And in your MyFixUpBean
>>> you read the first line of the Body and set it back as To header
>>> and remove it from the body.
>>>
>>>
>>>
>>>> I am making a reliable mass mailing solution with spam control based on
>>>> ActiveMQ and camel, is it the right approach?
>>> Yeah I see why not. However I have not personally build spam
>>> protection software.
>>>
>>>
>>>>
>>>> Claus Ibsen-2 wrote:
>>>>> On Fri, May 8, 2009 at 2:58 PM, rohitbrai <rohitb...@gmail.com> wrote:
>>>>>> I have a message which has -
>>>>>> Header
>>>>>>    "To" - "a...@sdf.com,x...@dsfsdf.com,s...@serr.com"
>>>>>> Body
>>>>>>    Hello
>>>>>>
>>>>>> onthis message I tried -
>>>>>>
>>>>>> from("jms:queue:new.test1").splitter(header("To").tokenize(",")).to("jms:queue:new.test2");
>>>>>>
>>>>>> and I was expecting 3 entries on test2 queue
>>>>>> Header
>>>>>>    "To" - "a...@sdf.com"
>>>>>> Body
>>>>>>    Hello
>>>>>>
>>>>>> Header
>>>>>>    "To" - "x...@dsfsdf.com"
>>>>>> Body
>>>>>>    Hello
>>>>>>
>>>>>> Header
>>>>>>    "To" - "s...@serr.com"
>>>>>> Body
>>>>>>    Hello
>>>>>>
>>>>>>
>>>>>> But instead I got 3 messages on test2 queue like
>>>>>> Header
>>>>>>    "To" - "a...@sdf.com,x...@dsfsdf.com,s...@serr.com"
>>>>>> Body
>>>>>>    a...@sdf.com
>>>>>>
>>>>>> Header
>>>>>>    "To" - "a...@sdf.com,x...@dsfsdf.com,s...@serr.com"
>>>>>> Body
>>>>>>    x...@dsfsdf.com
>>>>>>
>>>>>> Header
>>>>>>    "To" - "a...@sdf.com,x...@dsfsdf.com,s...@serr.com"
>>>>>> Body
>>>>>>    s...@serr.com
>>>>>>
>>>>>> So I guess, I am doing it and even understanding it wrong.
>>>>>>
>>>>>> Can anyone here guide me how to handle this situation.
>>>>> Hi
>>>>>
>>>>> Welcome on the Camel ride.
>>>>>
>>>>> The EIP patterns is about message routing where the message relies in
>>>>> the BODY payload.
>>>>> The header is just meta data about the message.
>>>>>
>>>>> So the splitter operates on splitting the BODY and not the headers,
>>>>> hence why you get the email address in the body.
>>>>>
>>>>> So by default there EIP patterns dont really support your use case out
>>>>> of the box, unless you do some manual fixup in Java code.
>>>>>
>>>>> You could use a POJO or the like where you create new messages to send
>>>>> along.
>>>>>
>>>>> private ProducerTemplate producer
>>>>>
>>>>> public void sendSplittedMessages(String body, @Headers Map headers) {
>>>>>   // loop the headers for each email adr
>>>>>   for (...) {
>>>>>      String email = ...
>>>>>      producer.sendBodyAndHeader("jms:queue:new:test02", body, "To",
>>>>> email);
>>>>>    }
>>>>>
>>>>> And then have a route that is like
>>>>> from("jms:queue:new.test1").bean(MySplitMessageClass.class,
>>>>> "sendSplittedMessages");
>>>>>
>>>>> where we route to our bean that, will do the "split" manually and send
>>>>> a new message to the JMS queue.
>>>>>
>>>>>
>>>>>> Thanks and Regards,
>>>>>>
>>>>>>
>>>>>>
>>>>>> --
>>>>>> View this message in context:
>>>>>> http://www.nabble.com/Split-using-tokenize-on-header-tp23445496p23445496.html
>>>>>> Sent from the Camel - Users mailing list archive at Nabble.com.
>>>>>>
>>>>>>
>>>>>
>>>>> --
>>>>> Claus Ibsen
>>>>> Apache Camel Committer
>>>>>
>>>>> Open Source Integration: http://fusesource.com
>>>>> Blog: http://davsclaus.blogspot.com/
>>>>> Twitter: http://twitter.com/davsclaus
>>>>> Apache Camel Reference Card:
>>>>> http://refcardz.dzone.com/refcardz/enterprise-integration
>>>>> Interview with me:
>>>>> http://architects.dzone.com/articles/interview-claus-ibsen-about?mz=7893-progress
>>>>>
>>>>>
>>>> --
>>>> View this message in context: 
>>>> http://www.nabble.com/Split-using-tokenize-on-header-tp23445496p23445999.html
>>>> Sent from the Camel - Users mailing list archive at Nabble.com.
>>>>
>>>>
>>>
>>>
>>
> 
> 
> 

Reply via email to