Adding an option to render only the shadow in SkDropShadowImageFilter

Adding an option to render only the shadow in SkDropShadowImageFilter

This is basically how blink uses the filter. Currently, I can't use it for "ShadowOnly" mode with the filter at all, but instead of copying the code and risking to have the codepaths diverge, I'm simply going to add the option here.

BUG=skia:

Committed: https://skia.googlesource.com/skia/+/234f036b3e731e06e616c5291157d3bb4fbfdee2

[sugoi1, 2014-10-23 18:01:47]

sugoi@chromium.org changed reviewers:
+ bsalomon@google.com, junov@google.com, senorblanco@google.com

[sugoi1, 2014-10-23 18:01:48]

No worries, I'll add a test for the new shadowOnly path and I'll add
"dropshadowimagefilter" to expectations/gm/ignored-tests.txt (although it's
already there for another reason, so I might get lucky if this gets committed
quickly enough).

I just want to know if this code is ok for skia.

[sugoi1, 2014-10-23 18:28:26]

Added the new test case in the dropshadowimagefilter gm.

https://codereview.chromium.org/646213004/diff/20001/gm/dropshadowimagefilter...
File gm/dropshadowimagefilter.cpp (right):

https://codereview.chromium.org/646213004/diff/20001/gm/dropshadowimagefilter...
gm/dropshadowimagefilter.cpp:105: virtual SkISize onISize() { return
SkISize::Make(400, 656); }
Made the gm a bit smaller. We now have 8 filters, each 64x64, with a 16 pixel
margin, so we have 8 * (64 + 16) + 16 = 656 pixels.

[reed1, 2014-10-23 19:01:52]

reed@google.com changed reviewers:
+ reed@google.com

[reed1, 2014-10-23 19:01:53]

do we have a need for the draw-both version at all?

https://codereview.chromium.org/646213004/diff/20001/include/effects/SkDropSh...
File include/effects/SkDropShadowImageFilter.h (right):

https://codereview.chromium.org/646213004/diff/20001/include/effects/SkDropSh...
include/effects/SkDropShadowImageFilter.h:32: SkImageFilter* input = NULL,
nit: while we're adding a new entry-point, can we arrange to not have so many
default parameters?

[Stephen White, 2014-10-23 19:23:03]

senorblanco@chromium.org changed reviewers:
+ senorblanco@chromium.org

[Stephen White, 2014-10-23 19:23:04]

https://codereview.chromium.org/646213004/diff/20001/include/effects/SkDropSh...
File include/effects/SkDropShadowImageFilter.h (right):

https://codereview.chromium.org/646213004/diff/20001/include/effects/SkDropSh...
include/effects/SkDropShadowImageFilter.h:32: SkImageFilter* input = NULL,
On 2014/10/23 19:01:53, reed1 wrote:
> nit: while we're adding a new entry-point, can we arrange to not have so many
> default parameters?

The uniqueID param can be removed once the non-SkRecord path has been removed
from Skia (has it?)

The rest I'd prefer to keep, since from a Skia API POV I think it's nice to
avoid having to think about DAGs and crop rects if all you care about is adding
a correct drop shadow to your primitive.

[reed1, 2014-10-23 19:30:11]

On 2014/10/23 19:23:04, Stephen White wrote:
>
https://codereview.chromium.org/646213004/diff/20001/include/effects/SkDropSh...
> File include/effects/SkDropShadowImageFilter.h (right):
> 
>
https://codereview.chromium.org/646213004/diff/20001/include/effects/SkDropSh...
> include/effects/SkDropShadowImageFilter.h:32: SkImageFilter* input = NULL,
> On 2014/10/23 19:01:53, reed1 wrote:
> > nit: while we're adding a new entry-point, can we arrange to not have so
many
> > default parameters?
> 
> The uniqueID param can be removed once the non-SkRecord path has been removed
> from Skia (has it?)
> 
> The rest I'd prefer to keep, since from a Skia API POV I think it's nice to
> avoid having to think about DAGs and crop rects if all you care about is
adding
> a correct drop shadow to your primitive.

3 default params? Lets create a variant function that just takes 3 fewer
parameters, as a helper. (as you already have for the method that doesn't take
the new enum).

[sugoi1, 2014-10-23 19:42:42]

On 2014/10/23 19:30:11, reed1 wrote:
> 3 default params? Lets create a variant function that just takes 3 fewer
> parameters, as a helper. (as you already have for the method that doesn't take
> the new enum).

Done.

[reed1, 2014-10-23 19:44:13]

api lgtm

[Stephen White, 2014-10-23 19:53:21]

LGTM

[sugoi1, 2014-10-23 20:51:12]

The CQ bit was checked by sugoi@chromium.org

[commit-bot: I haz the power, 2014-10-23 20:51:26]

CQ is trying da patch. Follow status at
 https://chromium-cq-status.appspot.com/patch-status/646213004/60001

[commit-bot: I haz the power, 2014-10-23 20:59:57]

Committed patchset #4 (id:60001) as 234f036b3e731e06e616c5291157d3bb4fbfdee2