[RFA 1/2] Linespec rewrite (update 2)

[Joel Brobecker]

> As you can see, the ":10" and ":25" were simply ignored. I've
> stepped through the code, and decode_variable will see "\"+\":10\n",
> but ada_name_for_lookup will return "+". From there on out, the
> ":10" is lost.

Hah! I thought that the line numbers were because GDB tries to find
the first line that matches, starting from the given one...

> Unless I'm missing something, this appears to be another special
> case of maintaining bug-for-bug compatibility.

I think that the syntax itself should be accepted, and that it's not
a bug (operators are functions). But we probably do not want to
perpetuate the bug where the line number is ignored :-).

> How would you like me to fix this?

I was hoping that we'd be able to handle this in the parser?
I am not completely familiar with the new code yet, so it's just
a wild guess.

I am almost done with the extension of operator_bp.exp; I'll send it
when I've removed all the typos and thinkos...

Thanks!
-- 
Joel