limiting lines in FF pipeline

Home Page › Forums › BizTalk 2004 – BizTalk 2010 › limiting lines in FF pipeline

Viewing 1 reply thread
  • Author
    Posts
    • October 7, 2006 at 2:39 AM #15981
      community-content

        Hi,

        I'm a little bit new in BizTalk, and I found my self facing a problem that is bigger than my knowledge.

        I have a send port that uses a flat file pipeline.
        My problem is that I have to make sure that the structure of every line in the output FF will not exceed above 100 characters.
        The fields' structure inside each line is not important, because we use comas for delimiters, and I can trim the end of the line as much as I need.

        Where can I limit the numbers of characters in each line in my flat file? (it's have to done it in the pipeline.)
        Or if I want to put a simple code (C#) that counts the number of characters in each line, where can I place it in the pipeline?

        Thank you very much

        Bizon

      • October 9, 2006 at 12:01 PM #15989
        community-content

          You could do something like this inside the map, assuming you have all your data in one field.  You could concatenate all the fields together and map them forward into a single node on the output side.  Then just use the Trim functoid. 

          • October 9, 2006 at 3:20 PM #15992
            community-content

              Thank you very much Stephen.

              I have done something similar to your suggestion on Sunday.

              I build a FF schema with one field node but I done this node as a positional and set it to start from 1 to 100.

              In the map, I concatenate all the fields from incoming schema with coma fore delimiter and I connect it to the one field in the FF schema.

              Work perfect.

              I have another question: How is the performance of this solution?

              Thanx

              • October 9, 2006 at 5:55 PM #15994
                community-content

                  Well, if you have to map the data anyway this is minimal performance impact.

                   

                  If you are mapping only to trip the fields, it a hit but this solution is much easer than build a custom streaming pipeline component to accomplish the task. 

                   

                  Unless you have “HUGE” messages or a “LOT” of them you should be fine. 

                   

                  Huge and Lot or of course relative to your solution and hardware J

                  • October 10, 2006 at 12:53 PM #16005
                    community-content

                      Let me see if my message can be defined as "HUGE" (I really don't know what is A big message in BizTalk it is because I just started 2 mouths ago)

                      The message structure is about 20 records, and in each record there are 10 fields.

                      And for the "a LOT" of them? The answer is YES. It's an end of day check list (about few hundred, once a day); to make sure all the messages was passing thru.

                      Thank you for your answers.

                      Bizon

            • Author
              Posts
            Viewing 1 reply thread
            • The forum ‘BizTalk 2004 – BizTalk 2010’ is closed to new topics and replies.