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

[Joel Brobecker <brobecker@adacore.com>]

Hi Keith,

I tested your patch locally, and unfortunately, it regresses in
the following situation:

    (gdb) break ops.adb:"+":10
    FUNCTION:OFFSET is unimplemented

It seems to be doing it for all operators, for instance:

    (gdb) break ops.adb:"/":25
    FUNCTION:OFFSET is unimplemented

It's a series of tests I meant to add after you check in your patch,
because I believe the output changes (that's something we were
discussing on Archer). Perhaps it would be better if I wrote a patch
to be applied after we apply yours. That way, you can include it
in your tree and immediately notice this regression...

I'll probably have some time to look at this this afternoon, if you're
tied up.

Git also notices the following little details when applying your
patch:

> /home/brobecke/linespec-rewrite-parser-2.patch:810: trailing whitespace.
> /* Throw an appropriate error when an unexpected token is encountered
> /home/brobecke/linespec-rewrite-parser-2.patch:3180: trailing whitespace.
> /home/brobecke/linespec-rewrite-parser-2.patch:3960: trailing whitespace.
> {
> /home/brobecke/linespec-rewrite-parser-2.patch:4106: trailing whitespace.
>     add the_tests "$srcfile:main:$x" invalid_label [string trim $x] "main"
> /home/brobecke/linespec-rewrite-parser-2.patch:4246: trailing whitespace.
>   return f.baz (foo::bar () + 3);
> warning: squelched 1 whitespace error
> warning: 6 lines add whitespace errors.

Not terribly important, but easy to fix...

-- 
Joel