[Tcsh] [PATCH] Fix: bug in "which" builtin

[Jamie Landeg-Jones]

(Re-sent with more appropriate subject line)

Damn! I just last night discovered a bug I was going to mention today!

Basically, I noticed my aliases were out of wack. I discovered that "which"
is no longer returning error 1 if a command is not found.

It seems that when "setcopy" was changed in some cases to "setstatus", a mistake
was made.

This is the fix:

--- tc.func.c.orig	2021-12-09 11:32:52.000000000 +0000
+++ tc.func.c	2022-02-02 07:23:26.065413000 +0000
@@ -503,7 +503,7 @@
 	rv &= cmd_expand(*v, NULL);
 
     if (!rv)
-	setstatus(0);
+	setstatus(1);
 }
 
 static int

I didn't check when this change occurred, but I plucked out a tc.func.c from 2016 and
did the following:

"tcsh" jamie at thompson% sdiff -s /tmp/tc.func.c tc.func.c | egrep 'setstatus|setcopy.STRstatus'
        setcopy(STRstatus, STR1, VAR_READWRITE);                 |      setstatus(0);
        setcopy(STRstatus, STR1, VAR_READWRITE);                 |      setstatus(1);
        setcopy(STRstatus, STR0, VAR_READWRITE);                 |      setstatus(0);
        setcopy(STRstatus, STR1, VAR_READWRITE);                 |      setstatus(1);

The inconsistency can be easily seen there.

Cheers, Jamie