Introduce DEFINE_AX_OBJECT_TYPE_CASTS to replace manual toFoo() in accessibility child class

Introduce DEFINE_AX_OBJECT_TYPE_CASTS to replace manual toFoo() for accessibility child class.

DEFINE_TYPE_CASTS can be used for toFoo() in accessibility child class. The DEFINE_TYPE_CASTS will generate
plenty more helper functions for accessibility class type cast.

BUG=309516
Committed: https://src.chromium.org/viewvc/blink?view=rev&revision=160154

[gyuyoung-inactive, 2013-10-21 06:35:42]

On 2013/10/21 02:50:32, gyuyoung wrote:

Kent, how do you think about this patch ? Do you know who should review this cl
?

[r.kasibhatla, 2013-10-21 12:28:30]

My thoughts on macro.

https://chromiumcodereview.appspot.com/31463002/diff/1/Source/core/accessibil...
File Source/core/accessibility/AXListBoxOption.h (right):

https://chromiumcodereview.appspot.com/31463002/diff/1/Source/core/accessibil...
Source/core/accessibility/AXListBoxOption.h:78:
DEFINE_AX_OBJECT_TYPE_CASTS(AXListBoxOption, isListBoxOption());
Wouldn't it be better to construct even isFoo from object type, something like 
DEFINE_TYPE_CASTS(thisType, AXObject, object, object->is##AXObject(),
object.is##AXObject())

1. It would really simplify the usage for devs - just define the macro passing
classname.
2. There would be consistency even in naming of isFoo() calls - just like
toFoo()
3. It would make regex matching in style-checker efficient - what to exactly
look out for.

[dmazzoni, 2013-10-21 16:16:13]

lgtm once r.kasibhatla is happy with the change too.

[tkent, 2013-10-21 22:06:11]

lgtm

https://chromiumcodereview.appspot.com/31463002/diff/1/Source/core/accessibil...
File Source/core/accessibility/AXListBoxOption.h (right):

https://chromiumcodereview.appspot.com/31463002/diff/1/Source/core/accessibil...
Source/core/accessibility/AXListBoxOption.h:78:
DEFINE_AX_OBJECT_TYPE_CASTS(AXListBoxOption, isListBoxOption());
On 2013/10/21 12:28:31, r.kasibhatla wrote:
> Wouldn't it be better to construct even isFoo from object type, something like

> DEFINE_TYPE_CASTS(thisType, AXObject, object, object->is##AXObject(),
> object.is##AXObject())
> 
> 1. It would really simplify the usage for devs - just define the macro passing
> classname.
> 2. There would be consistency even in naming of isFoo() calls - just like
> toFoo()
> 3. It would make regex matching in style-checker efficient - what to exactly
> look out for.

We should go ahead with this CL as is, and discuss this separately.
This issue affects whole the code, and we need to discuss it in blink-dev.

IMO, renaming AXObject::isFoo to AXObject::isAXFoo looks redundant. We already
have AX prefix in the class name.

[commit-bot: I haz the power, 2013-10-21 22:07:11]

CQ is trying da patch. Follow status at
https://chromium-status.appspot.com/cq/gyuyoung.kim@samsung.com/31463002/1

[commit-bot: I haz the power, 2013-10-22 08:43:59]

Change committed as 160154