Do not split attributes on getters and setters in databasebuilder.

Do not split attributes on getters and setters in databasebuilder.

So we don't have to combine them again in generator.

R=antonm@google.com,sra@google.com

Committed: https://code.google.com/p/dart/source/detail?r=9414

[Anton Muhin, 2012-07-04 15:41:25]

https://chromiumcodereview.appspot.com/10704100/diff/1/lib/dom/scripts/genera...
File lib/dom/scripts/generator.py (right):

https://chromiumcodereview.appspot.com/10704100/diff/1/lib/dom/scripts/genera...
lib/dom/scripts/generator.py:324: if attr1.id == attr2.id]
should you check readonly'ness here?

https://chromiumcodereview.appspot.com/10704100/diff/1/lib/dom/scripts/genera...
lib/dom/scripts/generator.py:462: if a.is_read_only < b.is_read_only: return -1
shouldn't you have a.is_read_only > b.is_read_only: return 1 her?

https://chromiumcodereview.appspot.com/10704100/diff/1/lib/dom/scripts/genera...
lib/dom/scripts/generator.py:463: return 0
overall, I'd rather code that as:

def AOO(a, b):
  return cmp([a.id, a.is_read_only], [b.id, b.is_read_only])

https://chromiumcodereview.appspot.com/10704100/diff/1/lib/dom/scripts/idlnod...
File lib/dom/scripts/idlnode.py (right):

https://chromiumcodereview.appspot.com/10704100/diff/1/lib/dom/scripts/idlnod...
lib/dom/scripts/idlnode.py:384: if (attribute.id == candidate.id and
if AOO(attribute, candidate) == 0?

And, btw, maybe it's time to have this AOO as __cmp__ method?

[podivilov, 2012-07-04 16:32:47]

https://chromiumcodereview.appspot.com/10704100/diff/1/lib/dom/scripts/genera...
File lib/dom/scripts/generator.py (right):

https://chromiumcodereview.appspot.com/10704100/diff/1/lib/dom/scripts/genera...
lib/dom/scripts/generator.py:324: if attr1.id == attr2.id]
On 2012/07/04 15:41:25, antonmuhin wrote:
> should you check readonly'ness here?

Nope, if attribute was declared with different "readonly'ness" in parent, we
still need to shadow it.

https://chromiumcodereview.appspot.com/10704100/diff/1/lib/dom/scripts/genera...
lib/dom/scripts/generator.py:462: if a.is_read_only < b.is_read_only: return -1
On 2012/07/04 15:41:25, antonmuhin wrote:
> shouldn't you have a.is_read_only > b.is_read_only: return 1 her?

Turned out this function isn't used at all :)

https://chromiumcodereview.appspot.com/10704100/diff/1/lib/dom/scripts/idlnod...
File lib/dom/scripts/idlnode.py (right):

https://chromiumcodereview.appspot.com/10704100/diff/1/lib/dom/scripts/idlnod...
lib/dom/scripts/idlnode.py:384: if (attribute.id == candidate.id and
On 2012/07/04 15:41:25, antonmuhin wrote:
> if AOO(attribute, candidate) == 0?
> 
> And, btw, maybe it's time to have this AOO as __cmp__ method?

This is the only place where we compare nodes.

[Anton Muhin, 2012-07-04 16:44:39]

lgtm, but one question: do you want to protect us from the case when
getters/setters will show up in IDLs?

[podivilov, 2012-07-04 17:30:27]

On 2012/07/04 16:44:39, antonmuhin wrote:
> lgtm, but one question: do you want to protect us from the case when
> getters/setters will show up in IDLs?

Just in case WebKit will adopt FremontCut idl syntax some day? :) 
Why do you think it would ever happen?

[Anton Muhin, 2012-07-04 20:04:36]

Just to have a safety net (and that should be cheap I believe).  But if you
don't want to mess with it, so it be, all consequences are yours though :)

On 2012/07/04 17:30:27, podivilov wrote:
> On 2012/07/04 16:44:39, antonmuhin wrote:
> > lgtm, but one question: do you want to protect us from the case when
> > getters/setters will show up in IDLs?
> 
> Just in case WebKit will adopt FremontCut idl syntax some day? :) 
> Why do you think it would ever happen?