I like the idea. Write up a more complete proposal and we'll all
review it. A few thoughts:
- The same mechanism could be used to implement Perl 6 sub wrapping,
so keep that in mind as you're thinking through the implementation
options. Names like "trace_sub" may be too specific to one particular
use.
- Runtime addition of traces/wrappers will be more important than
adding them at compile-time, considering that they're likely to be
mostly used for layering debugging/profiling/etc checks over existing
code, and in the debugger. So much more important, that I'd consider
skipping the compile-time syntax.
- I like the Perl 6 way of specifying the point in the wrapping sub
where the wrapped sub should be called better than setting up
conditionals based on an argument for the operation type. So,
something like:
.sub joe
.param pmc wrapped_sub
.param pmc orig_args
.local string proc_name
proc_name = wrapped_sub.name()
print "entering "
say proc_name
invoke_wrapped wrapped_sub, orig_args
print "leaving "
say proc_name
.end
.sub bob :wrap joe
say "whoa."
.end
(That also means you can monkey around with the args before passing
them on to the wrapped sub.)
- It should be possible both to put multiple wrappers/traces on a
subroutine, and to put wrappers around other wrappers. So these are
all possible:
.sub joe
#...
.end
.sub bob :wrap joe
#...
.end
.sub phil :wrap joe
#...
.end
.sub dave :wrap phil
#...
.end
Allison
FWIW, it might be easier to put the :wrap flag onto the wrapping sub,
IMHO. If there is some sub you want to wrap, according to the syntax
above you would need to change that sub, which might not be desirable or
even possible. The wrapped sub could be only available as a library, for
instance.
so, the idea would then be:
.sub joe :wrap('bob')
.param pmc args
say "calling bob..."
invoke_wrapped args
say "returned from bob"
.end
.sub bob
say "this is bob"
.end
which outputs:
calling bob....
this is bob
returned from bob
This way, the original sub is not touched, and you just 'add' the
wrapper function.
my 2c.
