[Info-vax] Returning data from Cobol AST routine.

[Jan-Erik Söderholm]

Den 2021-09-21 kl. 15:22, skrev abrsvc:
> On Tuesday, September 21, 2021 at 9:06:18 AM UTC-4, Jan-Erik Söderholm wrote:
>> Den 2021-09-21 kl. 15:00, skrev Arne Vajhøj:
>>> On 9/21/2021 3:15 AM, Jan-Erik Söderholm wrote:
>>>> Den 2021-09-21 kl. 03:27, skrev Arne Vajhøj:
>>>>> On 9/20/2021 6:55 PM, Jan-Erik Söderholm wrote:
>>>>>> I have been looking at the Cobol AST example here:
>>>>>> http://computer-programming-forum.com/48-cobol/b75d8c5fdd43048e.htm
>>>>>>
>>>>>> This does work fine as an example. But what are the options to get
>>>>>> some data back to the main program (called "x" in the example)
>>>>>> from the ast program (called "ast_routine" in the example)?
>>>>>>
>>>>>> I have tried different versions of global, external and so on, but
>>>>>> no luck so far. How to get "x" and "ast_routine" to share some data?
>>>>>> Anything similar to an COMMON area in Fortran?
>>>>>>
>>>>>> The idea is to have an AST routine that will read from a mailbox when
>>>>>> something is written to it. My idea was that the read of the mailbox
>>>>>> would be using an AST to avoid polling the mailbox.
>>>>>>
>>>>>> This is to have a command input to detached processes to get them to
>>>>>> reload the config, repoen the log file, close and exit and so on.
>>>>>>
>>>>>> Are there more options if not everything is done in Cobol?
>>>>>> We can easily add some C if that helps...
>>>>>
>>>>> external works for me.
>>>>>
>>>>> identification division.
>>>>> program-id. x.
>>>>> environment division.
>>>>> data division.
>>>>> working-storage section.
>>>>> 01 magic-value pic 9(10) display external.
>>> ...
>>>>> end program x.
>>>>> identification division.
>>>>> program-id. ast_routine.
>>>>> environment division.
>>>>> data division.
>>>>> working-storage section.
>>>>> 01 magic-value pic s9(10) display external.
>>> ...
>>>>> end program ast_routine.
>>>>>
>>>>> $ cobol x
>>>>> $ link x
>>>>> $ run x
>>>>> sleeping for 5 seconds
>>>>> 000000012C
>>>>
>>>> Ah, OK! that will be tested. Hm, is this limited to numeric variables?
>>>> My test was using a x(3) variable...
>>>
>>> Try it!
>>>
>>> Per documentation then external just moves the variable to a PSECT
>>> with the same name as the variable to tell the linker to overlay
>>> them.
>>>
>>> I would expect that to work with any data type.
>>>
>>>> Hm, if this would had been a read-AST on a mailbos instead of a timer,
>>>> maybe the ast_routine can just awake the main code and the mailbox
>>>> been read in the main code. The ast_routine could set a variable that
>>>> just signals the reason fo the wakeup.
>>>
>>> I don't understand the problem.
>>>
>>> If you want the main flow to wait until something is in the mailbox,
>>> then why not do a blocking read?
>>>
>>> Arne
>>>
>> No, sorry... The main code does other things (the regular business
>> processing) all the time. I just do not want it to have to poll the
>> mailbox with some fixed interval. I expect an AST to be faster also
>> with a more "on-demand" response time.
>>
>> Will do some more testing this evening after regular business hours...
> 
> Can't you just post an asynchronous QIO read to the mailbox with an AST routine to service the read within the main program?  In this fashion, the writing to the mailbox will trigger the action.  In this fashion, the main program will continue to proceed as normal only getting interrupted when the mailbox read copletes.
> 

Yes, that is my plan. What I had issues with was how to get the data
from the mailbox message back from the AST routine to the main code.
Since the AST routine has to be a separate PROGAM section. And it
cannot be a sub-program within the main code, the POINTER EXTERNAL,
to get the address to the AST routine, does not work in that case.
At least not if I understood my tests correctly.

But I think I understand the basic AST processing as such...