IBM Mainframe Forum

HI all,
I have just a smattering of experience with Easytrieve, and all of it gleaned from my co-workers. I do have the "Easytrieve Plus student guide" though! My coworkers are willing to help when I hit a bump, but I can't seem to get a satisfactory answer from them on this one. I've been reading through the forums here and elsewhere trying to find an answer for my problem, but haven't come across a working solution.

Basically, We have a CICS screen where a user keys in a number of transaction codes for the night. That screen can only accommodate 54 codes though, so when we have more than that, I have to hold a batch job and manually add the missing codes to a file. These codes are listed three times within the file though, each time with different info after the code number.

Input File1 example, just transaction codes:


Input File2 example


Expected Output


The program reads FILE2 and writes it out with no changes until it finds the end-of-section markers (The two line with % and the line with ZZZZZZZZ), It then reads FILE1 with a get until EOF loop. It then formats the transaction codes into the style for the section it is adding it to (Pull, Drop) before adding it to the output.
This works great for the first marker, the output contains all the old & new transactions, but it fails on a A010 INVALID FILE REFERENCE since FILE1 is now at EOF. Is there a way to close the file, or start the read from the top of FILE1 again?



I've tried multiple JOB statements, but that resets the output too. I eventually got it to work using three separate DD's for FILE1, all pointing to the same DSN, and each having a separate PROC, but that looks sloppy.

Hello and welcome to the forum,

It may help if you post the jcl for this step.

If you use vsam for file one, you can re-read the data and not use multiple dd statements. I'm not aware of a way to close and re-open a file with Easytrieve.

You might also consider reading file1 into an array within the program.

Thanks for the reply, dick scherrer.

An Array seems to be exactly what I needed. I had been looking at using one, but thought it wouldn't work out based on the examples in the Easytrieve Student Guide. The examples I saw had an array as one line, for example "JanFebMarAprMayJunJulAugSepOctNovDec" I couldn't figure out how to make that idea work for me. After showing your answer to a co-worker, he showed me the way to use an array as I needed, one record at a time. I swear, they'll fight tooth and nail to resist imparting more knowledge on others.

Interestingly enough, I had been asking if Easytrieve supported just such a feature, from my barely remembered days of BASIC coding on the old C64 computer, but I couldn't recall the name for it. And I guess my poor communication of "You would load a variable with a counter for a reference" didn't quite explain what I barely remembered. The easytrieve implementation seems nearly identical from what I recall of tables in Basic. Below is my testing of the code. It is not full featured yet, but it does display the loaded array.

So now I have this code to load the array



And for testing purposes, this display of the values:

Good to hear you have it working - thank you for letting us know

d