The Question is:
I am using the F$SEARCH command to define a symbol for a input file name that
will be used in a Cobol program as an input file
load_product_file == f$search("TELE$DATA:ACCOUNTPRODUCT_*SMS.TXT")
A new file can be posted (using the same name) several times before my command
procedure runs in the batch queue. So the target directory will contain
multiple versions of the same file name but with different version numbers.
Is there anyway I can make the F$SEARCH smart enough to capture the oldest
file?? Currently, the F$SEARCH command captures the files from the most recent
to the oldest (3,2,1). I would like F$SEARCH to find the oldest file in the
directory. Thereby allowi
ng the command file to process the files in order of creation(1,2,3).
Can it be done?
The Answer is :
Version number ";-0" is defined to be the lowest version number. So the
$ load_product_file == f$search("TELE$DATA:ACCOUNTPRODUCT_*SMS.TXT;-0")
will find the oldest version of the first file that matches the filename
There is no syntax for finding the *second* oldest file. You can find the
second and subsequent *youngest* files using version numbers ";-1", ";-2",...
To work through a list of files in creation order by version number, the
simplest method would be to RENAME or DELETE the files after they've been
processed, and use the ";-0" syntax to find the next file. For example:
$ IF nextfile.EQS."" THEN GOTO NoMore
$ PROCESS 'nextfile'
$ RENAME 'nextfile' COMPLETED_DIRECTORY:
$ GOTO FileLoop