In a message dated 8/19/2006 1:23:40 A.M. Central Daylight Time, [EMAIL PROTECTED] writes:
>Does anyone have any sample code that does I/O (read only) by >manually building a DEB? (Bypassing OPEN and CLOSE.) I have only general comments and not specific code samples to offer. Of course your program must be authorized to do what you intend. As another poster pointed out, you can read a PDS's directory blocks with BSAM, and with many other non-authorized techniques, but only if you have a DD statement pointing to the data set and do an official allocate, enqueue, and OPEN. Assuming your program is authorized, there are many ways to do what you ask, which is to bypass all system security for accessing data to which you are not supposed to have access. E.g., you can OPEN a DCB for EXCP that points to a real DD statement with a zero-track temporary data set name, then get into key 0, then modify the DEB that was built by your OPEN process so that the DEB now points to the data set that you wish to read, then use EXCP to read the data. When you CLOSE such a DCB, if you don't restore the original DEB contents before the CLOSE you will see ugly looking but irrelevant error messages. Long ago I saw in the microfiche for OBTAIN code that does exactly what you seem to be asking for; namely, build a partial DEB that looks as if it has been built by a legitimate OPEN. This was probably before the days of the DEB check service, which, I would think, would make that technique no longer workable, but I could be wrong. I will leave it to other posters to question your motives. Bill Fairchild ---------------------------------------------------------------------- For IBM-MAIN subscribe / signoff / archive access instructions, send email to [EMAIL PROTECTED] with the message: GET IBM-MAIN INFO Search the archives at http://bama.ua.edu/archives/ibm-main.html
