Well , thats Amazing, i would like to thanq for effort and spending some time in looking into my problem..
sorry to say , the solution works absolutely fine for the above data, but rather i have 2 more points as below.

* The file's shown is only an extract of large file, actual file has got 1 million records of which A9000 is one entry.
* The file1 and file2 and header contains different tables data, i have Appended one more set of data ..
How do i extend the same solution, Can we use Hash to store and then loop on differences?

i have one more requirement, the date field i should skip based on some position in another file, i can handle this one..
can i have some idea on the updated set of data, i will try my best to extend your solution.