Re: [PATCH 3/5] ref-filter.c: add trailer options to used_atom

[Jeff King <peff@peff.net>]

On Fri, Sep 29, 2017 at 11:22:36PM -0700, Taylor Blau wrote:

> In preparation for "%(trailers)" to take trailer parsing arguments, use Jeff's
> convenience structure for trailer processing options introduced in 8abc89800c.
> 
> We will later populate this field from the arguments given to %(trailers), and
> then use the trailer_opts instance to format ref trailers correctly using
> `format_trailer_from_commit`.

I think this should probably just be squashed in with the next patch,
since this does nothing at all without adding a caller that uses it.

> diff --git a/ref-filter.c b/ref-filter.c
> index 467c0279c..84f14093c 100644
> --- a/ref-filter.c
> +++ b/ref-filter.c
> @@ -82,6 +82,7 @@ static struct used_atom {
>  		} remote_ref;
>  		struct {
>  			enum { C_BARE, C_BODY, C_BODY_DEP, C_LINES, C_SIG, C_SUB, C_TRAILERS } option;
> +			struct process_trailer_options trailer_opts;
>  			unsigned int nlines;
>  		} contents;

This contents struct is odd. The used_atom struct can have many types,
and has has a big union of type-specific data.  And then we break down
the "contents" type with a further enum, but don't actually put the
type-specific data into a union (not just your patch, but already
"nlines" is specific only to C_LINES).

It's probably not worth caring about, though. The point of a union is to
reduce the overall struct size, and here our type-specific data is
fairly small. It would only change the overall struct size if it were
larger than other parts of the union (and remote_ref, for example, is
pretty big).

-Peff

PS As an aside, I find the whole %(contents:...) thing a bit
   unfortunate. I understand why the _implementation_ wants to group
   similar options together so that it can avoid parsing too much. But
   the user doesn't care about that. Just "%(trailers)" should be
   sufficient (as evidenced by the fact that we added it a separate
   alias). But none of that is new to your series.