With no DDF files, you'll have to guess at the format. The paper will make this make some sense. However, without app source code, all you know is this:
1) The record length is 739 bytes.
2) The first 12 bytes are a unique string field.
3) The next 35 bytes are a duplicate string field.
4) The next 20 bytes are a duplicate string field.
5) The next 35 bytes are a duplicate string field.
Beyond that, we know nothing about the file's datya or structure. Luckily, if ALL fields are strings, then you should be able to browse the records (with WBEXEC32) and see a pattern for where each field starts and stops. This should help you to identify how to interpret the bytes.