Re: [RFC PATCH 2/2] Allow and evaluate expressions in command arguments

[Keith Seitz <keiths@redhat.com>]

On 2/3/23 09:22, Andrew Burgess wrote:
> Keith Seitz via Gdb-patches <gdb-patches@sourceware.org> writes:
>
>> In order to mark the bounds of expressions to be evaluated, I've
>> introduced the (arbitrary) marker "$()". Everything contained inside
>> this marker will be passed to the expression parser, evaluated, printed
>> as a string, and then substituted for the original expression.
> 
> I really like the goal of this series, and think something like this
> would be awesome to add.
> 
> I do worry about the extra $(..) wrapper though, that feels a little
> clunky, if the common case turns out to be just doing simple things like
> environment variables, I wonder if we could make things like:
> 
>    (gdb) file $_env("HOME")/$_env("BLAH")
> 
> just work?

I agree. That $() marker *is* cumbersome.

> We could still retain $(...) for full expression evaluation, because I
> guess, in theory this would allow things like:
> 
>    (gdb) some_gdb_command $(some_inferior_function())
> 
> thought I suspect things like this will be far less common ... maybe?

That's another good idea. We could explicitly permit:

1. $(<any valid expression>): Evaluate ANY arbitrary expression
2. $VAR_OR_FUNC: Evaluate /only/ a convenience variable or expression.
    Convenience funcs are easier than vars, but I think this can be made
    to work.

As you note, if using just a convenience variable or function is insufficient,
users could fallback to the more clumsy $() notation.

Thank you for the input! I will work on completing this patch
for formal submission.

Keith