[ ruby-Bugs-1396 ] Inconsistent use of st_foreach

Bugs item #1396, was opened at 2005-01-26 21:10
You can respond by visiting: 
http://rubyforge.org/tracker/?func=detail&atid=1698&aid=1396&group_id=426

Category: Core
Group: None
Status: Open
Resolution: None
Priority: 5
Submitted By: Lyle Johnson (lyle)
Assigned to: Nobody (None)
Summary: Inconsistent use of st_foreach

Initial Comment:
This has not caused any specific problems that I know of, but it worries me. It seems like the kind of thing that, if it ever were to cause a problem, it would be awfully hard to track down.

As of Ruby 1.8.2, the second argument to st_foreach() is supposed to be a callback function with a declaration like this:

int (*func)(st_data_t, st_data_t, st_data_t, int);

We can deduce this because of how the "func" argument gets used inside st_foreach(). In each of the two times this callback "func" is used, it's called with four arguments; in one case, the last argument is a one, in the other case it's a zero.

But if you look at a number of places in the code where st_foreach() is used (and there are examples in class.c, dln.c, eval.c, gc.c, hash.c, process.c, st.c and variable.c), many callers are passing in an old-style callback function that only expects three arguments. By "old-style", I'm referring to the expected argument list for this callback function in Ruby versions 1.8.1 and earlier (where I think that last "int" argument was omitted).

Just wanted to point this out in case it had gone unnoticed.

----------------------------------------------------------------------

You can respond by visiting: 
http://rubyforge.org/tracker/?func=detail&atid=1698&aid=1396&group_id=426