I seem to remember that for every cobol table there is a related file with the description of the structure. This file has the same name, but a different extension. But I may be wrong...!

>I was afraid it was the case :(, the problem is to decode the number fields
>thanks
>
>>>Hi all , a customer asked me if it was possible to import acu cobol archives into vfp. Of course it is but ...
>>>how ? anybody knows where to find the odbc drivers ?
>>>
>>>thank you
>>
>>Don't those COBOL archives have a fixed-size structure, and no kind of header to tell where each field starts? If this is the case, I doubt it would be possible with an ODBC driver, which wouldn't know how to extract the data. I THINK you would have to get information about the file structure (how many bytes for every field? what is the field type? any special encoding?), and read the file with LLFF or FileToStr() and then parse it.