This is a script to access data in a Microfocus COBOL database and export it. It runs on Python 3 and requires the argparse library (needs to be installed on Python < 3.2).
I'm not at all familiar with COBOL, and built this based on information I was able to gather on the web, and looking at hex dumps. It's not a proper library to work with these files, but a cludge to get the data out. It works for the files I need to process, and might need furter improvements for yours.
COBOL seems to have dealing with these types of databases built into the language itself. Unfortuately, the files itself do not contain a schema; the schema is written in COBOL code, in copybooks or FDD files.
As a result, to actually export the data, we need to define the fields that it contains. Lacking the original source code or definition files, this needs to be reverse-engeneered. This exporter uses an ad-hoc format to define the fields, rather than actual COBOL .FDD files.
Print the headers:
$ ./mfcobol-export FILE1.DAT
FILE1.DAT: Indexed File with Variable Length Records
Organization: index
Recording Mode: variable
Index Type: (not set)
Long Records: False
Maximum Record Length: 155
Minimum Record Length: 155
Creation Date: 1996-09-06 09:29:31
Data Compression: none
DB Sequence Number: 0
Indexed Handler Version: 1090527224
Integrity Flag: 0
Print the raw bytes of every database item (you do not have field layout yet):
$ ./mfcobol-export --bytes FILE1.DAT
Export the data with a field layout defined:
$ ./mfcobol-export --fields=datadef.cfg FILE1.DAT $ ./mfcobol-export --fields=datadef.cfg --json FILE1.DAT
For more information on how to declare fields in the .cfg file, see the example file that is included.
Microfocus COBOL supports different types of databases:
- Line Sequential (no header)
- Printer Sequential (no header)
- Record Sequential Files with Fixed Length Records (no header)
- Record Sequential Files with Variable Length Records
- Relative Files with Fixed Length Records (no header)
- Relative Files with Variable Length Records
- Indexed Files
Of those, files containing no header are generally not supported by this script. In addition, "Relative Files with Variable Length Records" are also not supported. This leaves 4. and 7.
Personally tested I have only indexed files (7). Variable Length Record Sequential should work as well, though, since it's similar.
- http://supportline.microfocus.com/documentation/books/nx40/fhorgs.htm
- Official documentation on the file format.
- http://www.tek-tips.com/viewthread.cfm?qid=1510638&page=1
- Script to parse database records via ad-hoc field definitions. Used by this exporter.
- http://g4u0419c.houston.hp.com/en/64/hppg/hpmixc.htm
- Explains the datatypes of Microfocus COBOL.
- http://cb2java.sourceforge.net/
- Requires the COBOL copybook files, then can probably export the data.
- http://www.cobolproducts.com/
- Silber Systems DataViewer and related products; Very expensive, not a great UI. Can guess the record layout, but this doesn't really work particularily well.
- http://www.alchemysolutions.com/products/Fujitsu-Data-Converter-for-Windows/overview
- Fujitsu Data Converter for Windows, part of their NetCOBOL product. No demo version, prices only on request.
- http://www.connx.com/products/microfocus.html
- CONNX for Microfocus