From:	HENRY::IN%"info-vax-request%kl.sri.com%kl.sri.com%csnet-relay.CSNET%relay.cs.net@RCA.COM" 19-JUN-1987 09:37
To:	info-vax@KL.SRI.COM
Subj:	shell$to_vms calling sequence


Someone asked about the shell$to_vms (and other) functions:

Carl J Lydick wrote:
>	SHELL$TO_VMS
>
>	num_files_found = 
>	shell$to_vms( vms_filespec, action_routine, wild_card )
>
>	char *vms_filespec;
>	int wild_card;
>	int action_routine();
>	int shell$to_vms();

This interested me, because I need just a routine myself. I did some testing
and came up with the following (note; this is from my observations and
may not be 100% accurate).

1) The action routine is called with two arguments (actually: according
   to the stack frame two longwords). The first I assume to be a 
   pointer to the vms translated name (it printf's okay).
   The second always seems to be the value 1. 
   Whats the scoop? Anybody know?

2) The action routine can return 0 or 1. A return code of 0 tells 
   shell$to_vms not to call it again and return to the caller, useful
   in wildcard matching. A return code of 1 will cause shell$to_vms
   to call the action routine again with the next file found -  in the
   case of wildcard matching.

3) shell$to_vms calls sys$parse() and sys$search(). (I stepped into 
   shell$to_vms with the debugger...).

4) shell$to_vms requires that each directory exists in order to convert
   the UNIX file spec. I assume that this is a side affect of calling
   sys$parse which also require that the directories exist. 

5) shell$to_vms is written in BLISS (according to the debugger).

Item 4 above is a royal pain. I would like a routine that converted the
UNIX file spec to VMS regardless of whether it exists or not. I guess
I shall have to write my own...

Can anybody fill in the missing holes in the above???

BTW we are running uVMS 4.5 with VAXC V2.3. I have the C code for the
above testing if anybody wants it.

		Cheers!
			John Weald
-- 
UUCP: {sdencore,cbosgd,amdahl,ptsfa,dana}!aussjo!john
UUCP: {styx,imagen,dlb,gould,sci,altnet}!auspyr!john