Multi format file

[jerome_rajan]

Hi,

We have a Multi format source file that has a header followed by details and finally a summary record. Now this file isn't restricted to just one header but multiple header/detail/summary groups. A typical file would look like

H10 A B C 20140827
D10 1 2 3 4 5 6 Ab CDE FG
D20 A B
D20 X Y
D20 M N
S10 3
H10 X Y Z 20140827
(and so on)
S10 5

Each H10 will have one associated D10 record. And each D10 will have multiple associated D20 records. And each H10 will have a corresponding S10 record. These 4 record types together form one group of data.

Our objective is to read this data (which I'm trying to do using CFF) and load into an Oracle table which has 'nested tables; in them.

What is the best way to read such a file? If CFF, can you please guide or point me to a good example that I can use as reference?

Thank you

[FranklinE]

CFF does not look like a good option for what you describe. It looks more like a repeating tag/sub-tag pattern. CFF will split your incoming data into separate links based on the record type. If someone here has better understanding of CFF, I'd like to learn it too.

I had a similar situation in the old 7.x days. I used the repeating header designation as the "delimiter" for a full record, then parsed the rest in a subsequent transformer.

[jerome_rajan]

Thank you Franklin. The first line of your post is a part of my solution for sure. Helps me channelize the next few hours in a better direction. :D