Add --draw-view-bounds-rects to draw bounds rects for all views.

Add --draw-view-bounds-rects to draw bounds rects for all views.

This is useful for debugging and measuring layouts.

Some details:

* Simple on/off command-line switch.  Filtering to a subset of views seems
theoretically useful.  I thought of comparing the accessible name of each view
against a wildcarded command-line argument, but this seemed more trouble than it
was worth, when people were unlikely to know accessible names anyway.

* I tried filled (rather than stroked) rects, which was better for visualizing
overdraw, but worse for easily measuring distances and such.  I suppose we could
have flags to do both if people really cared.

* Uses ARGB 0x30ff0000 rects, so the intensity of the red indicates the number
of views sharing that border.

* Borders are drawn with 1 px (not DIP) rects inset from the bounds, so a view
adjacent to another view should not appear to share a border line; you should
get two adjacent lines.

* For non-integral DSFs, snaps to integral pixel widths by flooring the view's
scaled size.

* Implemented as a separate painting pass so children/siblings don't draw atop
the rects, but this only goes so far; separate layers can still potentially
obscure these rects.  I don't know how to solve this.

BUG=none
TEST=none

Review-Url: https://codereview.chromium.org/2770603002
Cr-Commit-Position: refs/heads/master@{#458925}
Committed: https://chromium.googlesource.com/chromium/src/+/be50046d5ab84d12920bfab89131506f002aab11

[Peter Kasting, 2017-03-22 05:27:28]

pkasting@chromium.org changed reviewers:
+ sky@chromium.org

[sky, 2017-03-22 15:26:05]

Nice!
LGTM - I would like a better name for PaintRoot() (see comments below), but I
readily admit my suggestion is not the greatest.

https://codereview.chromium.org/2770603002/diff/1/ui/views/view.h
File ui/views/view.h (right):

https://codereview.chromium.org/2770603002/diff/1/ui/views/view.h#newcode1400
ui/views/view.h:1400: // Painting
------------------------------------------------------------------
There's already a section with this name on line 1278. Move there?

https://codereview.chromium.org/2770603002/diff/1/ui/views/view.h#newcode1402
ui/views/view.h:1402: // Returns true if this view should be painted, i.e. is
visible and nonempty.
Can you make this comment more specific. It doesn't check drawn state, which is
intentional, but given the name it's easy to assume if this returns true the
view will be painted, but that isn't necessarily the case.

https://codereview.chromium.org/2770603002/diff/1/ui/views/view.h#newcode1421
ui/views/view.h:1421: void PaintRoot(const ui::PaintContext& parent_context);
Generally a 'root' view refers to RootView. That isn't the case for this
function. We don't really have a good name for a view with a layer, so maybe
PaintFromPaintRoot?

[Peter Kasting, 2017-03-22 21:39:11]

https://codereview.chromium.org/2770603002/diff/1/ui/views/view.h
File ui/views/view.h (right):

https://codereview.chromium.org/2770603002/diff/1/ui/views/view.h#newcode1400
ui/views/view.h:1400: // Painting
------------------------------------------------------------------
On 2017/03/22 15:26:05, sky wrote:
> There's already a section with this name on line 1278. Move there?

Oh hey!  I looked for that and missed it.  Thanks.

https://codereview.chromium.org/2770603002/diff/1/ui/views/view.h#newcode1402
ui/views/view.h:1402: // Returns true if this view should be painted, i.e. is
visible and nonempty.
On 2017/03/22 15:26:05, sky wrote:
> Can you make this comment more specific. It doesn't check drawn state, which
is
> intentional, but given the name it's easy to assume if this returns true the
> view will be painted, but that isn't necessarily the case.

Done.

https://codereview.chromium.org/2770603002/diff/1/ui/views/view.h#newcode1421
ui/views/view.h:1421: void PaintRoot(const ui::PaintContext& parent_context);
On 2017/03/22 15:26:05, sky wrote:
> Generally a 'root' view refers to RootView. That isn't the case for this
> function. We don't really have a good name for a view with a layer, so maybe
> PaintFromPaintRoot?

Done.

I had the same concerns as you but couldn't think of a better name.  Yours is
better.

[Peter Kasting, 2017-03-22 21:39:16]

The CQ bit was checked by pkasting@chromium.org

[Peter Kasting, 2017-03-22 21:39:16]

The patchset sent to the CQ was uploaded after l-g-t-m from sky@chromium.org
Link to the patchset: https://codereview.chromium.org/2770603002/#ps20001
(title: "Review comments")

[commit-bot: I haz the power, 2017-03-22 21:40:09]

CQ is trying da patch. Follow status at
 
https://chromium-cq-status.appspot.com/v2/patch-status/codereview.chromium.or...

[commit-bot: I haz the power, 2017-03-22 22:37:29]

The CQ bit was unchecked by commit-bot@chromium.org

[commit-bot: I haz the power, 2017-03-22 22:37:30]

Try jobs failed on following builders:
  linux_chromium_chromeos_ozone_rel_ng on master.tryserver.chromium.linux
(JOB_FAILED,
http://build.chromium.org/p/tryserver.chromium.linux/builders/linux_chromium_...)

[Peter Kasting, 2017-03-22 22:46:14]

The CQ bit was checked by pkasting@chromium.org

[commit-bot: I haz the power, 2017-03-22 22:46:55]

CQ is trying da patch. Follow status at
 
https://chromium-cq-status.appspot.com/v2/patch-status/codereview.chromium.or...

[commit-bot: I haz the power, 2017-03-22 23:21:15]

CQ is committing da patch.

Bot data: {"patchset_id": 20001, "attempt_start_ts": 1490222774336650,
"parent_rev": "8d7f56980ecec9a434c81c5d0b4915cc6ef069bd", "commit_rev":
"be50046d5ab84d12920bfab89131506f002aab11"}

[commit-bot: I haz the power, 2017-03-22 23:21:58]

Description was changed from

==========
Add --draw-view-bounds-rects to draw bounds rects for all views.

This is useful for debugging and measuring layouts.

Some details:

* Simple on/off command-line switch.  Filtering to a subset of views seems
theoretically useful.  I thought of comparing the accessible name of each view
against a wildcarded command-line argument, but this seemed more trouble than it
was worth, when people were unlikely to know accessible names anyway.

* I tried filled (rather than stroked) rects, which was better for visualizing
overdraw, but worse for easily measuring distances and such.  I suppose we could
have flags to do both if people really cared.

* Uses ARGB 0x30ff0000 rects, so the intensity of the red indicates the number
of views sharing that border.

* Borders are drawn with 1 px (not DIP) rects inset from the bounds, so a view
adjacent to another view should not appear to share a border line; you should
get two adjacent lines.

* For non-integral DSFs, snaps to integral pixel widths by flooring the view's
scaled size.

* Implemented as a separate painting pass so children/siblings don't draw atop
the rects, but this only goes so far; separate layers can still potentially
obscure these rects.  I don't know how to solve this.

BUG=none
TEST=none
==========

to

==========
Add --draw-view-bounds-rects to draw bounds rects for all views.

This is useful for debugging and measuring layouts.

Some details:

* Simple on/off command-line switch.  Filtering to a subset of views seems
theoretically useful.  I thought of comparing the accessible name of each view
against a wildcarded command-line argument, but this seemed more trouble than it
was worth, when people were unlikely to know accessible names anyway.

* I tried filled (rather than stroked) rects, which was better for visualizing
overdraw, but worse for easily measuring distances and such.  I suppose we could
have flags to do both if people really cared.

* Uses ARGB 0x30ff0000 rects, so the intensity of the red indicates the number
of views sharing that border.

* Borders are drawn with 1 px (not DIP) rects inset from the bounds, so a view
adjacent to another view should not appear to share a border line; you should
get two adjacent lines.

* For non-integral DSFs, snaps to integral pixel widths by flooring the view's
scaled size.

* Implemented as a separate painting pass so children/siblings don't draw atop
the rects, but this only goes so far; separate layers can still potentially
obscure these rects.  I don't know how to solve this.

BUG=none
TEST=none

Review-Url: https://codereview.chromium.org/2770603002
Cr-Commit-Position: refs/heads/master@{#458925}
Committed:
https://chromium.googlesource.com/chromium/src/+/be50046d5ab84d12920bfab89131...
==========

[commit-bot: I haz the power, 2017-03-22 23:21:59]

Committed patchset #2 (id:20001) as
https://chromium.googlesource.com/chromium/src/+/be50046d5ab84d12920bfab89131...