Re: [rfa] linespec.c, part 3

[David Carlton <carlton@math.stanford.edu>]

On Mon, 11 Nov 2002 17:28:45 -0500, Elena Zannoni <ezannoni@redhat.com> said:
> David Carlton writes:

>> This part of the decode_line_1 refactoring moves some code
>> initializing some flags into a separate function set_flags.  Some
>> comments:

> Looks ok, but I have a question. Wouldn't the code that sets up
> has_comma and is_quote_enclosed also belong to this set_flags
> function?

It seems entirely plausible to me that that is the case.  For now,
that will come in my next patch, which won't quite do what you ask:
the problem is that that part of the code is also updating 'p' to
figure out where a compound linespec separator is (for filenames or
C++/Java objects).  Teasing this apart is a bit of a mess; my current
"final" version of these changes has 'has_comma' buried in an internal
function, and 'is_quote_enclosed' doesn't last for very long.  But
it's messy, so I think that, once I'm done submitting patches that
just break off code into separate functions, we should revisit this
issue.

> (Ideally a struct of flags could be set up).

Yes, that would be a much better idea than passing the flag variables
by reference.

> But this is for a separate patch, you are just moving code around
> now.

Right.

> go ahead, thanks!

Committed.

(Now you've caught up to me!  Eek!  I'll have to get my grading done
quickly so I can keep up my end of things. :-) )

David Carlton
carlton@math.stanford.edu