Datastage custom stage to read "BAI file format"

december7 · Post by **december7** » Wed Jul 23, 2014 10:38 am

Hello,

We are looking to build a datastage job to read BAI file and load into DB.
Just a quick info on BAI file
"BAI, or the BAI file format, is a file format for performing electronic cash management balance reporting. The BAI format was developed and previously maintained by the Bank Administration Institute (BAI). One common application of the BAI format is for use by banks to transmit returned item data to customers (for example, checks which have been marked insufficient funds (NSF))"

Complexity with BAI: This file is not simple flat file, its structure is a bit complex, one single transaction is spilled in multiple records.

Do any one has worked on this earlier, if so can you please give some idea of your approach.

Thanks

chulett · Post by **chulett** » Wed Jul 23, 2014 10:51 am

Doesn't look all that bad and no need for a 'custom stage' for the multiple record types with grouping and detail records. The spec is freely available out there now, here is one such manual documenting it. I also see there are converters for sale (either direction) which might be a more cost effective way to go.

Do you already have the target table(s) defined so you know which bits need to go where?

rameshrr3 · Post by **rameshrr3** » Wed Jul 30, 2014 7:09 pm

I think reading such a file in Datastage must be easier than creating it. But we have generated pretty complex DTCC PVF File and IFT File using a pure datastage approach by designing a tricky sort algorithm, and we may be set out to do the even more complex FAR file for DTCC - all using server jobs.

It would have been easier if the server job had a wave generator stage and an ordered link collector - but we did it in spite of that .
Such a file is designed to be consumed by a transaction handlng system typically on a Mainframe platform, but Websphere TX may support it 'out of the box'