Re: Using -Wshadow

Peter Eisentraut <peter_e(at)gmx(dot)net> writes:
> I'm not sure what the point is anyway. Shadowing is perfectly
> well-defined and I've never heard of a real problem because of it.

Well, shadowing a formal parameter with a local variable is most likely
a mistake, and shadowing a local with a more-tightly-nested local is,
if not an outright mistake, certain to confuse future maintainers.
So I'd be in favor of getting rid of cases like that.

I can't get excited about forbidding shadowing of globals by locals,
though ... seems like that's practically giving up one of the
advantages of having a block-structured language in the first place.

BTW, what I find by experiment with gcc 2.95.3 is that
local-shadowing-formal is warned of just with -Wall, if the local is
declared at the function's outermost brace level, whether or not you
say -Wshadow. So we already know we have none of those.

regards, tom lane