Expose API for whether an SkPicture contains text

Expose API for whether an SkPicture contains text

BUG=chromium:399728
Committed: https://skia.googlesource.com/skia/+/750ae26745d261183001a78d24251f6392e96b27

[ajuma, 2014-08-15 14:31:28]

This exposes the hasText data that was added by
https://codereview.chromium.org/435093003/

[reed1, 2014-08-15 15:32:37]

This api is not O(1) on the new backend, so lets see if its sufficient to
implement this on top of a canvas subclass.

Do you have representative skps to test against (ones used during animations I
presume)? I'd be happy to analyze our canvas performance against those.

[nduca, 2014-08-15 17:32:10]

is there a way to make the api O(1)? Presumably we set a bit during record.

[ajuma, 2014-08-18 17:49:00]

If SkPicture::suitableForGpuRasterization is still going to have an O(1)
implementation with the new backend, it's going to need to have something like
SkPictureContentInfo to gather stats during recording, and if this gathers stats
about text (as SkPictureContentInfo currently does) then we'd be able to make
hasText O(1) as well.

[mtklein, 2014-08-18 18:10:11]

On 2014/08/18 17:49:00, ajuma wrote:
> If SkPicture::suitableForGpuRasterization is still going to have an O(1)
> implementation with the new backend, it's going to need to have something like
> SkPictureContentInfo to gather stats during recording, and if this gathers
stats
> about text (as SkPictureContentInfo currently does) then we'd be able to make
> hasText O(1) as well.

Yes, I think that sounds right.  We now have and are in the process of expanding
a stats-gathering pass for the new backend.  Being able to answer hasText()
should be an easy addition.  I'm OK to land this as-is and be on the hook to do
the "fRecord TODO" and get this bit plumbed through.

[reed1, 2014-08-18 18:46:06]

with klein's comment, api lgtm

lets also add a test so we know this guy works. Should be convenient to add
something in tests/PictureTest.cpp

[ajuma, 2014-08-18 19:49:27]

PTAL, added a test.

[reed1, 2014-08-18 19:51:53]

The CQ bit was checked by reed@google.com

[commit-bot: I haz the power, 2014-08-18 19:52:13]

CQ is trying da patch. Follow status at
 https://skia-tree-status.appspot.com/cq/ajuma@chromium.org/478673002/20001

[commit-bot: I haz the power, 2014-08-18 20:00:01]

Committed patchset #2 (20001) as 750ae26745d261183001a78d24251f6392e96b27