GrapeCity Forums

I have been struggling to get the xml data provided by the Spread pipeline (source schema was generated by fpoint wizard) to map properly to my destination schema.  The destination schema has only one record format defined.  The source schema has 30 sheets the 3rd-24th of which are defined.  The detail data from each sheet is writing to my destination schema successfully (via looping functoid in Biztalk map) but no matter what I try the data I am trying to pull out of the sheet headings does not map properly.  For example, when I specify the 3rd and 4th sheets (AR and AZ sheets) in my map all records in the destination schema reflect data from the AZ sheet.  I have try XSLT scripts, iteration, index, extra loops, nothing works.  Can you point me in the right direction?

Hey Michelle,

I am glad to hear that you have progressed to this point.  If I understand correctly, you are saying that your destination schema contains one record definition that includes the detail information from each sheet (3-24), and each record should also include, as part of each destination record, data from the headers of the sheet that the source detail information exists.  So you are building records for your destination schema that are detail records with header data added to each record.

What I believe is happening is that the AZ header data is overwriting the AR header data.  As a suggestion, you might want to use the Script functoid and create static variables that are set each time the loop iterates, and set those values to the header-specific fields of the schema.

The folks on www.BizTalkGurus.com are great at answering BizTalk-specif questions, and have a volume of information in their BTS2006 forum.  I have spent lots of time going through their forums.

Hope this helps.

---

Robby -
FarPoint Technologies, Inc.

A quick update with our resolution:  I have tried passing parms with both XSLT and C# - no data was ever yielded as if my script or XPATH or whatever was never specified at the right level to capture the data.  I also tried to use the index functoid to similar results (no data).  This morning our team lead looked into it with me.  She assigned a record count in the map and that count was always 1 - weird, huh?  Anyway, we ended up adding  another parent node to the destination schema and splitting the data into header and detail records.  Originally the destination schema heirarchy was:  <Schema> <Carrier_Rate_Excel> <Excel_Rate_Proposals> with all fields set as children of <Excel_Rate_Proposals>.  Now the destination schema heirarchy is: <Schema> <Carrier_Rate_Excel> <Record> with <Excel_Rate_Proposals> and <Excel_Rate_Details> set as children of <Record>.  Data from the "Header" rows in the source schema are mapped as children under <Excel_Rate_Proposals> and data from the "Record" loop in the source schema are mapped as children under <Excel_Rate_Details>.  This yielded data in the destination schema as desired because it allowed us to control (force) the looping.  Let me know if you have additional thoughts!