[cairo] [PATCH 03/51] core: allow NULL parameter for _cairo_path_fixed_is_box()

[Bryce Harrington]

On Fri, Dec 18, 2015 at 02:28:42PM +0100, Enrico Weigelt, metux IT consult wrote:
> Allow that function to be called with NULL parameter for the
> return buffer, so we can use it for only checking whether
> a path is a box, w/o actually returning the box dimensions.
> 
> This feature will be used in subsequent patches.
> 
> Signed-off-by: Enrico Weigelt, metux IT consult <enrico.weigelt at gr13.net>
> ---
>  src/cairo-path-fixed.c | 3 ++-
>  1 file changed, 2 insertions(+), 1 deletion(-)
> 
> diff --git a/src/cairo-path-fixed.c b/src/cairo-path-fixed.c
> index ac44bb7..df72b47 100644
> --- a/src/cairo-path-fixed.c
> +++ b/src/cairo-path-fixed.c
> @@ -1295,7 +1295,8 @@ _cairo_path_fixed_is_box (const cairo_path_fixed_t *path,
>  
>      buf = cairo_path_head (path);
>      if (_points_form_rect (buf->points)) {
> -	_canonical_box (box, &buf->points[0], &buf->points[2]);
> +	if (box != NULL)
> +	    _canonical_box (box, &buf->points[0], &buf->points[2]);
>  	return TRUE;
>      }

Fwiw, _canonical_box is also used in _cairo_path_fixed_is_stroke_box.
If a check should be added to _cairo_path_fixed_is_box then for
consistency it should be added there too.

Although from the function name I would not expect passing box=NULL
would be acceptable.  Perhaps the check needs to be further up by
whatever is calling in here, since it appears to be an internal routine
anyway?

Frankly, _cairo_path_fixed_is_box() seems odd even aside from this fix,
since it's checking to see if a path is a rectangle and then has a side
effect of maybe giving it back as box.  Seems like this side-effect
ought to be done as a separate activity by the caller.

In any case, a little more code doc explaining what these routines
are intended to return would likely help.

Bryce
  
> -- 
> 2.6.4.442.g545299f
> 
> -- 
> cairo mailing list
> cairo at cairographics.org
> http://lists.cairographics.org/mailman/listinfo/cairo