bug#14422: 24.3; Eager Macro Expansion

[Stefan Monnier <monnier@iro.umontreal.ca>]

retitle 14422 Apply eager-macroexpansion everywhere (eval-region, ...)
thanks

> The remaining point is that the ERT test still fails in exactly the same
> way when it is _not_ byte-compiled and batch-tested (like the original
> case in Org), but it produces the correct result when testing in
> interactive mode or in the debugger.  I guess I'm asking for a warning
> for recursive macro expansions that manipulate the same variable both at
> expansion and at runtime in separate macros.  Alternatively if the buggy
> code would always fail in the same way that would at least ensure it can
> be found more easily.

Indeed, that's part of the reason why I introduced this "eager
macroexpansion", which makes it that my previous foo.el test now behaves
identically whether it's byte-compiled or not.

The behavior is still different in a few other cases (where eager macro
expansion is not performed, typically M-C-x and things like that), but
I hope to reduce/eliminate them at some point.


        Stefan