Drop [LegacyInterfaceTypeChecking] for most of Selection

Drop [LegacyInterfaceTypeChecking] for most of Selection

For both addRange() and collapse(), Firefox and IE11 already throw when
the node argument is null:
http://software.hixie.ch/utilities/js/live-dom-viewer/saved/3781

removeAllRanges() and empty() still provide the functionality of
collapse(null).

DOMSelection::collapse() is only called from bindings, so the ASSERT
will hold.

As for setPosition(), that isn't implemented by Firefox or IE, but it's
an alias of collapse() and usage of setPosition() itself is very low:
https://www.chromestatus.com/metrics/feature/timeline/popularity/327

Leave setBaseAndExtend() for now, as usage is slightly higher and it
also requires making the arguments non-optional:
https://www.chromestatus.com/metrics/feature/timeline/popularity/406

BUG=391673, 561338

[philipj_slow, 2015-12-08 11:33:20]

The CQ bit was checked by philipj@opera.com to run a CQ dry run

[commit-bot: I haz the power, 2015-12-08 11:33:37]

Dry run: CQ is trying da patch. Follow status at
 https://chromium-cq-status.appspot.com/patch-status/1498253002/20001
View timeline at
 https://chromium-cq-status.appspot.com/patch-timeline/1498253002/20001

[commit-bot: I haz the power, 2015-12-08 12:03:30]

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

[commit-bot: I haz the power, 2015-12-08 12:03:30]

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

[philipj_slow, 2015-12-08 12:30:08]

philipj@opera.com changed reviewers:
+ yoichio@chromium.org

[philipj_slow, 2015-12-08 12:30:08]

PTAL. The tree is a mess right now so the dry run is all red, but the two tests
I'ev updated were the only two failing on PS1, so I think this could be landed
once the tree is in order.

[philipj_slow, 2015-12-08 22:34:54]

The CQ bit was checked by philipj@opera.com to run a CQ dry run

[commit-bot: I haz the power, 2015-12-08 22:43:06]

Dry run: CQ is trying da patch. Follow status at
 https://chromium-cq-status.appspot.com/patch-status/1498253002/20001
View timeline at
 https://chromium-cq-status.appspot.com/patch-timeline/1498253002/20001

[yoichio, 2015-12-08 23:55:48]

Once I changed collapse to throw exception,
 there were regression in the 3rd party library.
https://codereview.chromium.org/546703003
https://code.google.com/p/chromium/issues/detail?id=395318
Did you consider that and also other Selection functions?

[commit-bot: I haz the power, 2015-12-09 00:20:45]

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

[commit-bot: I haz the power, 2015-12-09 00:20:45]

Dry run: This issue passed the CQ dry run.

[philipj_slow, 2015-12-09 07:02:33]

philipj@opera.com changed reviewers:
+ haraken@chromium.org

[philipj_slow, 2015-12-09 07:02:33]

haraken, can you PTAL? yoichi is OOO, it turns out.

[haraken, 2015-12-09 07:26:21]

LGTM

[philipj_slow, 2015-12-09 09:12:35]

On 2015/12/08 23:55:48, yoichio_ooo_until_Jun_5 wrote:
> Once I changed collapse to throw exception,
>  there were regression in the 3rd party library.
> https://codereview.chromium.org/546703003
> https://code.google.com/p/chromium/issues/detail?id=395318
> Did you consider that and also other Selection functions?

Oh, sorry, I didn't see this before I asked haraken to review just now. I was
not aware of this problem. I have filed a spec bug
<https://github.com/w3c/selection-api/issues/64> and will implement that
behavior instead. By removing [LegacyInterfaceTypeChecking] but making the node
argument nullable, only trying to pass something other than null or a Node (like
a Range) will throw a TypeError.

[philipj_slow, 2015-12-09 12:55:16]

On 2015/12/09 09:12:35, philipj wrote:
> On 2015/12/08 23:55:48, yoichio_ooo_until_Jun_5 wrote:
> > Once I changed collapse to throw exception,
> >  there were regression in the 3rd party library.
> > https://codereview.chromium.org/546703003
> > https://code.google.com/p/chromium/issues/detail?id=395318
> > Did you consider that and also other Selection functions?
> 
> Oh, sorry, I didn't see this before I asked haraken to review just now. I was
> not aware of this problem. I have filed a spec bug
> <https://github.com/w3c/selection-api/issues/64> and will implement that
> behavior instead. By removing [LegacyInterfaceTypeChecking] but making the
node
> argument nullable, only trying to pass something other than null or a Node
(like
> a Range) will throw a TypeError.

I will close this in favor of two separate CLs to move this along:
https://codereview.chromium.org/1511913002/
https://codereview.chromium.org/1509353004/

[philipj_slow, 2015-12-09 12:55:22]

Description was changed from

==========
Drop [LegacyInterfaceTypeChecking] for most of Selection

For both addRange() and collapse(), Firefox and IE11 already throw when
the node argument is null:
http://software.hixie.ch/utilities/js/live-dom-viewer/saved/3781

removeAllRanges() and empty() still provide the functionality of
collapse(null).

DOMSelection::collapse() is only called from bindings, so the ASSERT
will hold.

As for setPosition(), that isn't implemented by Firefox or IE, but it's
an alias of collapse() and usage of setPosition() itself is very low:
https://www.chromestatus.com/metrics/feature/timeline/popularity/327

Leave setBaseAndExtend() for now, as usage is slightly higher and it
also requires making the arguments non-optional:
https://www.chromestatus.com/metrics/feature/timeline/popularity/406

BUG=391673,561338
==========

to

==========
Drop [LegacyInterfaceTypeChecking] for most of Selection

For both addRange() and collapse(), Firefox and IE11 already throw when
the node argument is null:
http://software.hixie.ch/utilities/js/live-dom-viewer/saved/3781

removeAllRanges() and empty() still provide the functionality of
collapse(null).

DOMSelection::collapse() is only called from bindings, so the ASSERT
will hold.

As for setPosition(), that isn't implemented by Firefox or IE, but it's
an alias of collapse() and usage of setPosition() itself is very low:
https://www.chromestatus.com/metrics/feature/timeline/popularity/327

Leave setBaseAndExtend() for now, as usage is slightly higher and it
also requires making the arguments non-optional:
https://www.chromestatus.com/metrics/feature/timeline/popularity/406

BUG=391673,561338
==========