Re: Calling a CLLE program from the command line with parms

On 10/30/13 10:38 AM, Charles Wilt wrote:

The system has no way of knowing how big the parm is defined in the
CL. That's the job of the *CMD object.

But default, character parms are passed as char(32) unless the value
has more than 32 characters; in which case the parm is passed at it's
actual size.

However, trailing blanks are trimmed.

If you wanted to pass 'HI' as a char(1024), you'd need to pass: HI
followed by 1022 blanks followed by some non-blank character, for a
total of 1025 characters. The trailing non-blank character will never
be seen by your program.

Alternatively, the the "best" solution in IMHO is to define a *CMD
front end to your program.

A command-line invocation, from the Command-Entry panel [having
pressed the F11="Display full" for the expanded-view, thus allowing
~1300 characters], simply requires the full 1024-bytes of data delimited
by the apostrophe as string-delimiter. There is no need to pass an
extra non-blank character in that case. However when a *CHAR _variable_
has been passed as an element of the PARM in the CALL statement [not
allowed in the *EXEC command-line environment], having been passed
through an additional iteration of the command interpreter, the extra
non-blank character would be required; e.g. for a request such as:
SBMJOB CMD(CALL PGM(pgmname) PARM(&C1024))