Using Xlate to create an A03

[Stuart Grimes]

I am receiving an A04 message.  When it hits the the xlate I send the A04 and create an A03.  In the testing tool it works.  The A04 gets sent and then the A03.  But when I send the message in the route,  the xlate sends the A03 first and then the A04.  Every time it does this.  I’ve tried using a delay between sends, and I have tried modifying the MSH date/time field along with the MSH control ID and the Security ID.  Why is it sending the A03 first.

Topic

[Robert Milfajt]

Post your Xlate code here.  Sounds like a CONTINUE/SUPPRESS type of issue, and how you are duplicating in Xlate, but without the code, it’s hard to tell for sure.

Robert Milfajt
Northwestern Medicine
Chicago, IL

[Stuart Grimes]

See attachment for Xlate

[Stuart Grimes]

I got it working.  I just reversed the code.

[Robert Milfajt]

Glad you got it to work.  I think the SEND might have thrown you off, I usually use CONTINUE in my Xlates to generate a second message.  I’m sure someone else on this board might have better insights on the differences between SEND and CONTINUE.

Robert Milfajt
Northwestern Medicine
Chicago, IL

[Stuart Grimes]

I think send does post proc processing and continue skips it.  No sure about that though.  It works either way on my code.

[Jim Kosloskey]

SEND sends the message directly to the destination thread bypassing post Xlate processing.

CONTINUE moves the message to the next step in the engine handling.

I ALWAYS use CONTINUE, unless the specifications and requirement of the particular integration is to bypass post xlate processing, and that is what I teach in my Xlate ITERATE class.

In 23 years of doing integrations with Cloverleaf and primarily using Xlates, I cannot recall more than one need to use SEND.

There is a ‘time-bomb’ quality with the use of SEND  this is also covered in my class.

email: jim.kosloskey@jim-kosloskey.com 29+ years Cloverleaf, 59 years IT - old fart.

[Robert Milfajt]

And I’ll bet that time bomb is what gave you the two different results, i.e., working in testing tool, failing in practice.

Thanks for the info Jim, I thought CONTINUE was better, it’s good to get the confirmation.

Robert Milfajt
Northwestern Medicine
Chicago, IL

[Author]

Replies