I'm working on an AREV 3.12 application. Is there any way to save as it's own SAVELIST a partially processed active select list?

I won't bore you with the details but I've created an new column in the target table. This new column will hold the foreign key to another table. I've created a program to fill that new column with the proper key. So when the process is done, it'll look as if that data has been written to that column all along. I'll also have to modify a few programs to write new data to that column going forward but that doesn't pertain to my question.

In my initial testing, the creation and sorting of the active select list will take somewhere between 2 and 3 hours. Then another 6 to 8 hours to process the resulting data in the active select list. Even tho this is a one time program, it'll need to be run on several of our applications.

I know how to halt the processing of the program in the middle of things. What I can't find is how to save the remaining of the active select list into a saved list. My testing just saves the original active select list. What I'm hoping to find is a way to save the unprocessed portion of the active select list into a saved list. So when the process is started again, it essentially can start where it left off.

Can anyone point me in the right direction?

Thank you for your time.

Michael Slack

If you track your progress through the initial list.

And, you understand the structure of saved lists in the LISTS table.

Then when you halt the process you can use the original savelist structure to create a new savelist that starts where you halted. Then use that new savelist when you resume.

hints: savelist rows are @fm with a trailing @fm in the LISTS table.

the name of the overflow row of a saved list is "*" and overflow number, starting with 2.

Dave

Hello Dave:

From you explanation, the part I'm fuzzy on is about tracking my progress thru the list.

In at least one example, I've written code to create the rows in the LISTS table so I could later do a GETLIST and process the active select list from that. So, I'm familiar with the LISTS row structure.

Are you suggesting that I process the unused part of the active select list into another SAVELIST (set of LISTS rows) to be used later by way of a GETLIST? I had thought of that after writing my original question but was hoping for something a bit easier. But what would be the fun in that.

Somewhere in my poor little brain, I seem to remember something about a pointer that the system uses as it goes thru an active select list to keep track of where it's at. In my reading so far, I haven't found anything mentioned about that. Is there a system variable that I can use or set? I'm thinking that I SAVELIST my original list. Then at the time I have to halt the process, that I can take the pointer and save it off somewhere. When I restart the process, I can take that value and set the system pointer so things will start up right were they left off.

Thanks,

Michael Slack

Yes, that was my suggestion. I assumed you weren't running in the Browse mode of a Window and thus the variables in Window Common aren't available.

You keep your own counter, I always initialize @RECCOUNT to 0 and increment with each read.

There may be a system common variable showing progress, but you probably don't have access to it, and to the system common variable(s) where the list of original keys are stored. So my suggestion of making a new saved list will always work.

Dave

Hello Dave:

Thanks for the information. I've already started designing that part of my program using your approach.

Thanks,

Michael Slack

The information is all in @Cursors and was documented in the poster giveaway we did for the last conference. Regretfully I'm on vacation and don't have one to hand. If you send an email to support@sprezzatura.com with your snail mail address we'll send you a copy.

The Sprezzatura Group

World leaders in all things RevSoft

View this thread on the forum...