Mad Installer

I am reading a simple flat file with \r\n delimited records. The schema validates fine, but when I try to use the orchestration to read it I always get the "Unexpected end of stream while lookng for \r\n" message. I've chaged the lookahead_depth to "0" and parser_optimization to "complexity". I'm new to BizTalk, but it seems to me this should be a 10 minute setup, but I'm burning hours trying to debug this error with no luck. Example of file:

Date: 01-25-2005
Start Time: 11:36:25
Stop Time: 13:09:41
Location: Not Specified

I just want to read in each line, but get the \r\n error.




Re: BizTalk R2 General Flat File Dissassembler

Sultan Abdul Kader - MSFT

Can you attach the sample instance file



Re: BizTalk R2 General Flat File Dissassembler

Sultan Abdul Kader - MSFT

If the schema seems to match the data. Several things you can check are as follows:

  1. Make sure the schema was re-deployed and the BizTalk service stopped/started to reset the schema cache.
  2. Make sure the instance file your testing against is postfix (has the Ħ°\r\nĦħ at the end).

If it is not fixed yet then can you provide the instance file





Re: BizTalk R2 General Flat File Dissassembler

AlanJamesSmith

Hi,

Good points from Sultan, another few things to try:

Use the "Validate Instance" option for the schema, with the test input file and "Native" type. This may give you some clues.

Use the "Generate Insance" with the output set to native, then open the generated file. View the generated and the input file and check for the CR and LF characters, see of they match.

Experiment with Infix and Post Fix delimeter pattern if there is a missmatch at the end of the file.

Also, you can try running through the Flat File Generator wizard starting with a fresh schema, i find it's usually quicker to do this than to fix a broken schema.

Hope this helps,

Regards,

Alan Smith