Re: RFA: implement ambiguous linespec proposal

[Matt Rice <ratmice@gmail.com>]

On Fri, Oct 28, 2011 at 9:28 AM, Tom Tromey <tromey@redhat.com> wrote:
> I'd appreciate comments on this patch.

In the following snippet, the initial comment doesn't mention
destroy_linespec_result,
and some of the field comments mention that the caller is responsible
for freeing.
destroy_linespec_result seems to fulfill this obligation,

I imagine that the comments just predate destroy_linespec_result, or
there is an OR situation e.g. if the caller wants to save some fields
they are responsible.  Anyhow it'd be nice to clarify that in the
comments.

thanks

>  /* An instance of this may be filled in by decode_line_1.  The caller
>    must call init_linespec_result to initialize it.  */
>
> @@ -30,22 +49,85 @@ struct linespec_result
>      display mechanism would do the wrong thing.  */
>   int special_display;
>
> -  /* If non-NULL, an array of canonical names for returned
> -     symtab_and_line objects.  The array has as many elements as the
> -     `nelts' field in the symtabs_and_line returned by decode_line_1.
> -     An element in the array may be NULL.  The array and each non-NULL
> -     element in it are allocated with xmalloc and must be freed by the
> -     caller.  */
> -  char **canonical;
> +  /* If non-zero, the linespec result should be considered to be a
> +     "pre-expanded" multi-location linespec.  A pre-expanded linespec
> +     holds all matching locations in a single linespec_sals
> +     object.  */
> +  int pre_expanded;
> +
> +  /* If PRE_EXPANDED is non-zero, this is set to the linespec entered
> +     by the user.  This is allocated with xmalloc and the caller is
> +     responsible for freeing it.  */
> +  char *addr_string;
> +
> +  /* The sals.  The vector should be freed by the caller.  */
> +  VEC (linespec_sals) *sals;
>  };
>
>  /* Initialize a linespec_result.  */
>
>  extern void init_linespec_result (struct linespec_result *);
>
> +/* Destroy a linespec_result.  */
> +
> +extern void destroy_linespec_result (struct linespec_result *);
> +
> +/* Return a cleanup that destroys a linespec_result.  */
> +
> +extern struct cleanup *
> +        make_cleanup_destroy_linespec_result (struct linespec_result *);
> +