0

We are converting a legacy system to a Java Platform. In the interim, we are sending files from legacy to a distributed server running Java. We are encountering an issue with packed fields. The COBOL (EBCDIC) COMP-3 fields have to be manipulated to BINARY then ASCII. However, when send the file back up to mainframe or transferring the file to a legacy system for further processing, the conversion back to packed becomes cumbersome and repetitive.

Is anyone aware of a process or function that can help with this type of manipulation more seamlessly. We have all ready looked into Velocidata and it has it's challenges as well.

Kevin McFadden
  • 337
  • 1
  • 5
  • 23
  • 1
    You're asking for a "process or product", but _"Questions asking us to recommend or find a book, tool, software library, tutorial or other off-site resource are off-topic for Stack Overflow as they tend to attract opinionated answers and spam. Instead, describe the problem and what has been done so far to solve it."_ – Richard Erickson Oct 21 '15 at 19:55
  • Thanks for the comment. Disregarding the "product", is there a simpler process to do these types of conversion without forward/backward re-engineering the file. i guess I would be asking/looking for a function or set of functions that could make this conversion easier and we could "can" it in a process or routine to turn it into a service. – Kevin McFadden Oct 21 '15 at 20:33
  • 2
    JRecord allows you work directly with EBCDIC Cobol Data. It also has Cobol2Csv / Csv2Cobol, Cobol2Xml /Xml2Cobol programs. To do the transfer back to the Mainframe, use Fixed-Length Record – Bruce Martin Oct 21 '15 at 21:16
  • 1
    Seems I never tire of the same advice. Don't take any binary or decimal fields, just take character fields. Separate sign, explicit decimal point or scaling factor. Whatever is convenient for you. If the Mainframe team aren't professional enough to give you that for two-way communication, go with Bruce's suggestion. It's easy-peasy to give and receive that on the Mainframe. – Bill Woodger Oct 21 '15 at 22:51
  • Thanks for your comments. We will try to complete a quick POC using JRecord. Unfortunately, we are dealing with literally hundreds and maybe even thousands of files so having the mainframe team accommodate would be really time consuming. Thanks, -Kevin – Kevin McFadden Oct 22 '15 at 14:08
  • 1
    Either The JRecords csv2cobol / cobol2csv or Data2xml / Xml2data (Cobol <--> xml) could fit with your needs. Using them means Java programmers do not need to learn about Cobol. But wath out for VB files (difficult to Transport) and duplicate field names in Cobol. – Bruce Martin Oct 26 '15 at 02:30

0 Answers0