Merging dart:html interfaces and implementations

Merging dart:html interfaces and implementations.

This CL leaves a lot of the infrastructure around for the interfaces (and also renames them to _IInterfaceName), but the interfaces are not in the generated files. I'll continue to clean this up in future CLs.

BUG=

Committed: https://code.google.com/p/dart/source/detail?r=14612

[vsm, 2012-11-01 18:59:50]

Pete, can you show the generated html_dart2js and html_dartium in this CL?

[Anton Muhin, 2012-11-02 12:55:46]

neat.

First pass.

If you have cycles to split this CL into smaller pieces, that would be most
appreciated.  As a starter: you add a lot of comments to templates, may go into
a separate CL.

https://codereview.chromium.org/11365019/diff/54/lib/html/scripts/generator.py
File lib/html/scripts/generator.py (right):

https://codereview.chromium.org/11365019/diff/54/lib/html/scripts/generator.p...
lib/html/scripts/generator.py:449: return interface_name
maybe add a TODO or FIXME to ease spotting those places later on

https://codereview.chromium.org/11365019/diff/7011/lib/html/scripts/generator.py
File lib/html/scripts/generator.py (right):

https://codereview.chromium.org/11365019/diff/7011/lib/html/scripts/generator...
lib/html/scripts/generator.py:703: def implementation_name(self):
implementation_name looks rather like an interface name now, correct?

https://codereview.chromium.org/11365019/diff/7011/lib/html/scripts/htmldartg...
File lib/html/scripts/htmldartgenerator.py (right):

https://codereview.chromium.org/11365019/diff/7011/lib/html/scripts/htmldartg...
lib/html/scripts/htmldartgenerator.py:9: from generator import *
maybe use names instead of globbing?

https://codereview.chromium.org/11365019/diff/7011/lib/html/scripts/systemhtm...
File lib/html/scripts/systemhtml.py (right):

https://codereview.chromium.org/11365019/diff/7011/lib/html/scripts/systemhtm...
lib/html/scripts/systemhtml.py:314: base_class =
base_type_info.implementation_name()
looks like base_class initialization should be moved out of the loop now and
look like base_class = self._backendRootClassName9) if base_type_info is None
else base_type_info.implementation_name(), or maybe with if.

https://codereview.chromium.org/11365019/diff/7011/lib/html/scripts/systemhtm...
lib/html/scripts/systemhtml.py:317: 
nit: no blank line

https://codereview.chromium.org/11365019/diff/7011/lib/html/scripts/systemhtm...
lib/html/scripts/systemhtml.py:328: implements_str += ' implements ' + ',
'.join(set(implements))
nit: no need in +=

https://codereview.chromium.org/11365019/diff/7011/lib/html/scripts/systemhtm...
lib/html/scripts/systemhtml.py:562: '\n  /**'
debugging?

https://codereview.chromium.org/11365019/diff/7011/lib/html/scripts/systemhtm...
lib/html/scripts/systemhtml.py:885: self._members_emitter.Emit('\n  /** @domName
$DOMINTERFACE.$DOMNAME */',
a helper like EmitAttributeDocumentation?

https://codereview.chromium.org/11365019/diff/7011/lib/html/scripts/systemnat...
File lib/html/scripts/systemnative.py (right):

https://codereview.chromium.org/11365019/diff/7011/lib/html/scripts/systemnat...
lib/html/scripts/systemnative.py:219: super_constructor = ': super.internal()'
no _-prefix for an internal constructor?

https://codereview.chromium.org/11365019/diff/7011/lib/html/src/dart2js_Conve...
File lib/html/src/dart2js_Conversions.dart (right):

https://codereview.chromium.org/11365019/diff/7011/lib/html/src/dart2js_Conve...
lib/html/src/dart2js_Conversions.dart:78: if (imageData is ImageData) return
imageData;
this looks sketchy: you declare imageData to be of type ImageData in the
singnature above.  Does the code behave as it should?

https://codereview.chromium.org/11365019/diff/7011/lib/html/src/dart2js_Conve...
lib/html/src/dart2js_Conversions.dart:236: if (e is _FileList) return e;
what's this _FileList?

https://codereview.chromium.org/11365019/diff/7011/lib/html/templates/html/da...
File lib/html/templates/html/dartium/html_dartium.darttemplate (left):

https://codereview.chromium.org/11365019/diff/7011/lib/html/templates/html/da...
lib/html/templates/html/dartium/html_dartium.darttemplate:38: LocalWindow
__window;
please, add a FIXME to rename __window into _window

https://codereview.chromium.org/11365019/diff/7011/lib/html/templates/html/im...
File lib/html/templates/html/impl/impl_Element.darttemplate (right):

https://codereview.chromium.org/11365019/diff/7011/lib/html/templates/html/im...
lib/html/templates/html/impl/impl_Element.darttemplate:305: void operator
[]=(String key, value);
is that a valid Dart?

https://codereview.chromium.org/11365019/diff/7011/lib/html/templates/html/im...
lib/html/templates/html/impl/impl_Element.darttemplate:481: * Returns [:true:]
classes cannot be added or removed from this
nit: [:true:] <iff> classes

https://codereview.chromium.org/11365019/diff/7011/lib/html/templates/html/im...
lib/html/templates/html/impl/impl_Element.darttemplate:515: * Returns [:true:]
classes cannot be added or removed from this
ditto

https://codereview.chromium.org/11365019/diff/7011/lib/html/templates/html/im...
File lib/html/templates/html/impl/impl_Event.darttemplate (left):

https://codereview.chromium.org/11365019/diff/7011/lib/html/templates/html/im...
lib/html/templates/html/impl/impl_Event.darttemplate:7: part of html;
why gone?

[blois, 2012-11-02 19:25:27]

For splitting out the documentation changes- these were just merged from the
interfaces with the goal of not regressing documentation with this CL.

I'm hoping that there isn't much included in this CL that is not necessary to
maintain parity, but I'd like to keep everything in needed to maintain that
parity.

https://codereview.chromium.org/11365019/diff/7011/lib/html/scripts/generator.py
File lib/html/scripts/generator.py (right):

https://codereview.chromium.org/11365019/diff/7011/lib/html/scripts/generator...
lib/html/scripts/generator.py:703: def implementation_name(self):
On 2012/11/02 12:55:47, Anton Muhin wrote:
> implementation_name looks rather like an interface name now, correct?

This is still the implementation name, assuming that it's the interface which is
going away.

These should really change to class_name when cleaned up.

https://codereview.chromium.org/11365019/diff/7011/lib/html/scripts/htmldartg...
File lib/html/scripts/htmldartgenerator.py (right):

https://codereview.chromium.org/11365019/diff/7011/lib/html/scripts/htmldartg...
lib/html/scripts/htmldartgenerator.py:9: from generator import *
On 2012/11/02 12:55:47, Anton Muhin wrote:
> maybe use names instead of globbing?

Done.

https://codereview.chromium.org/11365019/diff/7011/lib/html/scripts/systemhtm...
File lib/html/scripts/systemhtml.py (right):

https://codereview.chromium.org/11365019/diff/7011/lib/html/scripts/systemhtm...
lib/html/scripts/systemhtml.py:314: base_class =
base_type_info.implementation_name()
On 2012/11/02 12:55:47, Anton Muhin wrote:
> looks like base_class initialization should be moved out of the loop now and
> look like base_class = self._backendRootClassName9) if base_type_info is None
> else base_type_info.implementation_name(), or maybe with if.

Done.

https://codereview.chromium.org/11365019/diff/7011/lib/html/scripts/systemhtm...
lib/html/scripts/systemhtml.py:317: 
On 2012/11/02 12:55:47, Anton Muhin wrote:
> nit: no blank line

Done.

https://codereview.chromium.org/11365019/diff/7011/lib/html/scripts/systemhtm...
lib/html/scripts/systemhtml.py:328: implements_str += ' implements ' + ',
'.join(set(implements))
On 2012/11/02 12:55:47, Anton Muhin wrote:
> nit: no need in +=

Done.

https://codereview.chromium.org/11365019/diff/7011/lib/html/scripts/systemhtm...
lib/html/scripts/systemhtml.py:562: '\n  /**'
On 2012/11/02 12:55:47, Anton Muhin wrote:
> debugging?

The dartdoc comment? I believe this used for doc generation, with this the
implementation should match the comments of the interface.

https://codereview.chromium.org/11365019/diff/7011/lib/html/scripts/systemhtm...
lib/html/scripts/systemhtml.py:885: self._members_emitter.Emit('\n  /** @domName
$DOMINTERFACE.$DOMNAME */',
On 2012/11/02 12:55:47, Anton Muhin wrote:
> a helper like EmitAttributeDocumentation?

Done, but because of how the dartium code is currently factored it's not
straightforward to use there.

https://codereview.chromium.org/11365019/diff/7011/lib/html/scripts/systemnat...
File lib/html/scripts/systemnative.py (right):

https://codereview.chromium.org/11365019/diff/7011/lib/html/scripts/systemnat...
lib/html/scripts/systemnative.py:219: super_constructor = ': super.internal()'
On 2012/11/02 12:55:47, Anton Muhin wrote:
> no _-prefix for an internal constructor?

Yeah, I'm not thrilled about this part (and don't have a good suggestion here).
Basically once the library gets split then the constructor needs to be public
for the subclasses in other libs to call their super constructors.

It gets even more complicated with Element subclassing, where you essentially
have two constructors- a 'create a new X' and a 'wrap an instance of X'.

I was considering naming it 'wrap' rather than 'internal', but decided to call
it 'internal' for now to keep it public but send the message that it should not
be used.

I believe we'll need to revisit this with element subclassing.

https://codereview.chromium.org/11365019/diff/7011/lib/html/src/dart2js_Conve...
File lib/html/src/dart2js_Conversions.dart (right):

https://codereview.chromium.org/11365019/diff/7011/lib/html/src/dart2js_Conve...
lib/html/src/dart2js_Conversions.dart:78: if (imageData is ImageData) return
imageData;
On 2012/11/02 12:55:47, Anton Muhin wrote:
> this looks sketchy: you declare imageData to be of type ImageData in the
> singnature above.  Does the code behave as it should?
Good catch.

https://codereview.chromium.org/11365019/diff/7011/lib/html/src/dart2js_Conve...
lib/html/src/dart2js_Conversions.dart:236: if (e is _FileList) return e;
On 2012/11/02 12:55:47, Anton Muhin wrote:
> what's this _FileList?

Some types specify that they should not have interfaces, in the merged scheme
this translates to being a private type (it's only exposed as List<File>). So
this is essentially the impl of List<File>.

https://codereview.chromium.org/11365019/diff/7011/lib/html/templates/html/da...
File lib/html/templates/html/dartium/html_dartium.darttemplate (left):

https://codereview.chromium.org/11365019/diff/7011/lib/html/templates/html/da...
lib/html/templates/html/dartium/html_dartium.darttemplate:38: LocalWindow
__window;
On 2012/11/02 12:55:47, Anton Muhin wrote:
> please, add a FIXME to rename __window into _window

Done.

https://codereview.chromium.org/11365019/diff/7011/lib/html/templates/html/im...
File lib/html/templates/html/impl/impl_Element.darttemplate (right):

https://codereview.chromium.org/11365019/diff/7011/lib/html/templates/html/im...
lib/html/templates/html/impl/impl_Element.darttemplate:305: void operator
[]=(String key, value);
On 2012/11/02 12:55:47, Anton Muhin wrote:
> is that a valid Dart?

I assume so- this is a straight copy from Element interface, since without the
interfaces being generated this was getting dropped.

https://codereview.chromium.org/11365019/diff/7011/lib/html/templates/html/im...
lib/html/templates/html/impl/impl_Element.darttemplate:481: * Returns [:true:]
classes cannot be added or removed from this
On 2012/11/02 12:55:47, Anton Muhin wrote:
> nit: [:true:] <iff> classes

Done.

https://codereview.chromium.org/11365019/diff/7011/lib/html/templates/html/im...
lib/html/templates/html/impl/impl_Element.darttemplate:515: * Returns [:true:]
classes cannot be added or removed from this
On 2012/11/02 12:55:47, Anton Muhin wrote:
> ditto

Done.

https://codereview.chromium.org/11365019/diff/7011/lib/html/templates/html/im...
File lib/html/templates/html/impl/impl_Event.darttemplate (left):

https://codereview.chromium.org/11365019/diff/7011/lib/html/templates/html/im...
lib/html/templates/html/impl/impl_Event.darttemplate:7: part of html;
On 2012/11/02 12:55:47, Anton Muhin wrote:
> why gone?

This was moved from being an interface to an implementation. None of the
implementations currently have this. Don't know why. Should they?

Would prefer a separate CL if so.

[sra1, 2012-11-02 20:14:10]

https://codereview.chromium.org/11365019/diff/7011/lib/html/src/dart2js_Conve...
File lib/html/src/dart2js_Conversions.dart (right):

https://codereview.chromium.org/11365019/diff/7011/lib/html/src/dart2js_Conve...
lib/html/src/dart2js_Conversions.dart:78: if (imageData is ImageData) return
imageData;
On 2012/11/02 12:55:47, Anton Muhin wrote:
> this looks sketchy: you declare imageData to be of type ImageData in the
> singnature above.  Does the code behave as it should?

My bad. The function is passed a plain JS object on FF, so the parameter should
be dynamic.
I guess we need to run --checked on all browsers.

[blois, 2012-11-02 20:30:15]

https://codereview.chromium.org/11365019/diff/7011/lib/html/src/dart2js_Conve...
File lib/html/src/dart2js_Conversions.dart (right):

https://codereview.chromium.org/11365019/diff/7011/lib/html/src/dart2js_Conve...
lib/html/src/dart2js_Conversions.dart:78: if (imageData is ImageData) return
imageData;
On 2012/11/02 20:14:10, sra1 wrote:
> On 2012/11/02 12:55:47, Anton Muhin wrote:
> > this looks sketchy: you declare imageData to be of type ImageData in the
> > singnature above.  Does the code behave as it should?
> 
> My bad. The function is passed a plain JS object on FF, so the parameter
should
> be dynamic.
> I guess we need to run --checked on all browsers.

That's the convertNativeToDart function. Anton is correct, but I somehow lost my
fix. Re-adding it. I have been checking with --checked on chrome & ff.

[sra1, 2012-11-05 20:21:40]

On 2012/11/02 20:30:15, blois wrote:
>
https://codereview.chromium.org/11365019/diff/7011/lib/html/src/dart2js_Conve...
> File lib/html/src/dart2js_Conversions.dart (right):
> 
>
https://codereview.chromium.org/11365019/diff/7011/lib/html/src/dart2js_Conve...
> lib/html/src/dart2js_Conversions.dart:78: if (imageData is ImageData) return
> imageData;
> On 2012/11/02 20:14:10, sra1 wrote:
> > On 2012/11/02 12:55:47, Anton Muhin wrote:
> > > this looks sketchy: you declare imageData to be of type ImageData in the
> > > singnature above.  Does the code behave as it should?
> > 
> > My bad. The function is passed a plain JS object on FF, so the parameter
> should
> > be dynamic.
> > I guess we need to run --checked on all browsers.
> 
> That's the convertNativeToDart function. Anton is correct, but I somehow lost
my
> fix. Re-adding it. I have been checking with --checked on chrome & ff.

Please add the generated html file(s).

[sra1, 2012-11-06 02:42:49]

lgtm

[Anton Muhin, 2012-11-06 12:28:43]

m

https://chromiumcodereview.appspot.com/11365019/diff/7011/lib/html/scripts/sy...
File lib/html/scripts/systemnative.py (right):

https://chromiumcodereview.appspot.com/11365019/diff/7011/lib/html/scripts/sy...
lib/html/scripts/systemnative.py:219: super_constructor = ': super.internal()'
I'd propose to have a private name for it until we're starting to split the
libs---it's a real security threat and I'd rather not commit it in even
temporary.

On 2012/11/02 19:25:28, blois wrote:
> On 2012/11/02 12:55:47, Anton Muhin wrote:
> > no _-prefix for an internal constructor?
> 
> Yeah, I'm not thrilled about this part (and don't have a good suggestion
here).
> Basically once the library gets split then the constructor needs to be public
> for the subclasses in other libs to call their super constructors.
> 
> It gets even more complicated with Element subclassing, where you essentially
> have two constructors- a 'create a new X' and a 'wrap an instance of X'.
> 
> I was considering naming it 'wrap' rather than 'internal', but decided to call
> it 'internal' for now to keep it public but send the message that it should
not
> be used.
> 
> I believe we'll need to revisit this with element subclassing.

https://chromiumcodereview.appspot.com/11365019/diff/2083/sdk/lib/html/script...
File sdk/lib/html/scripts/generator.py (right):

https://chromiumcodereview.appspot.com/11365019/diff/2083/sdk/lib/html/script...
sdk/lib/html/scripts/generator.py:449: return interface_name
there are only 4 usages of this method and you already touch two of those, maybe
just inline it?

https://chromiumcodereview.appspot.com/11365019/diff/2083/sdk/lib/html/script...
File sdk/lib/html/scripts/htmldartgenerator.py (right):

https://chromiumcodereview.appspot.com/11365019/diff/2083/sdk/lib/html/script...
sdk/lib/html/scripts/htmldartgenerator.py:7: Dart:html APIs from the IDL
database."""
nit: dart:html

https://chromiumcodereview.appspot.com/11365019/diff/2083/sdk/lib/html/script...
File sdk/lib/html/scripts/systemhtml.py (right):

https://chromiumcodereview.appspot.com/11365019/diff/2083/sdk/lib/html/script...
sdk/lib/html/scripts/systemhtml.py:306: 
nit: no need in blank line

https://chromiumcodereview.appspot.com/11365019/diff/2083/sdk/lib/html/script...
sdk/lib/html/scripts/systemhtml.py:357: if typed_array_type:
this code has been moved into the helper you introduced, correct?

https://chromiumcodereview.appspot.com/11365019/diff/2083/sdk/lib/html/templa...
File sdk/lib/html/templates/html/impl/impl_CustomEvent.darttemplate (right):

https://chromiumcodereview.appspot.com/11365019/diff/2083/sdk/lib/html/templa...
sdk/lib/html/templates/html/impl/impl_CustomEvent.darttemplate:9: Object detail
= null]) => _$(CLASSNAME)FactoryProvider.create$CLASSNAME(
no need in = null

https://chromiumcodereview.appspot.com/11365019/diff/2083/sdk/lib/html/templa...
File sdk/lib/html/templates/html/impl/impl_Element.darttemplate (right):

https://chromiumcodereview.appspot.com/11365019/diff/2083/sdk/lib/html/templa...
sdk/lib/html/templates/html/impl/impl_Element.darttemplate:658: // Relative to
offsetParent
nit: missing trailing .

https://chromiumcodereview.appspot.com/11365019/diff/2083/sdk/lib/html/templa...
sdk/lib/html/templates/html/impl/impl_Element.darttemplate:684: // In global
coords
ditto

https://chromiumcodereview.appspot.com/11365019/diff/2083/sdk/lib/html/templa...
sdk/lib/html/templates/html/impl/impl_Element.darttemplate:687: // In global
coords
ditto

https://chromiumcodereview.appspot.com/11365019/diff/2083/sdk/lib/html/templa...
File sdk/lib/html/templates/html/impl/impl_IDBKeyRange.darttemplate (left):

https://chromiumcodereview.appspot.com/11365019/diff/2083/sdk/lib/html/templa...
sdk/lib/html/templates/html/impl/impl_IDBKeyRange.darttemplate:7: part of html;
why part of html is gone?

https://chromiumcodereview.appspot.com/11365019/diff/2083/sdk/lib/html/templa...
File sdk/lib/html/templates/html/impl/pure_interface.darttemplate (right):

https://chromiumcodereview.appspot.com/11365019/diff/2083/sdk/lib/html/templa...
sdk/lib/html/templates/html/impl/pure_interface.darttemplate:2: abstract class
$CLASSNAME$EXTENDS$IMPLEMENTS {
why no part of html?

Why it should be different from interface.darttemplate?

[blois, 2012-11-06 22:11:44]

https://chromiumcodereview.appspot.com/11365019/diff/2083/sdk/lib/html/script...
File sdk/lib/html/scripts/generator.py (right):

https://chromiumcodereview.appspot.com/11365019/diff/2083/sdk/lib/html/script...
sdk/lib/html/scripts/generator.py:449: return interface_name
On 2012/11/06 12:28:43, Anton Muhin wrote:
> there are only 4 usages of this method and you already touch two of those,
maybe
> just inline it?

Done.

https://chromiumcodereview.appspot.com/11365019/diff/2083/sdk/lib/html/script...
File sdk/lib/html/scripts/htmldartgenerator.py (right):

https://chromiumcodereview.appspot.com/11365019/diff/2083/sdk/lib/html/script...
sdk/lib/html/scripts/htmldartgenerator.py:7: Dart:html APIs from the IDL
database."""
On 2012/11/06 12:28:43, Anton Muhin wrote:
> nit: dart:html

Done.

https://chromiumcodereview.appspot.com/11365019/diff/2083/sdk/lib/html/script...
File sdk/lib/html/scripts/systemhtml.py (right):

https://chromiumcodereview.appspot.com/11365019/diff/2083/sdk/lib/html/script...
sdk/lib/html/scripts/systemhtml.py:306: 
On 2012/11/06 12:28:43, Anton Muhin wrote:
> nit: no need in blank line

Done.

https://chromiumcodereview.appspot.com/11365019/diff/2083/sdk/lib/html/script...
sdk/lib/html/scripts/systemhtml.py:357: if typed_array_type:
On 2012/11/06 12:28:43, Anton Muhin wrote:
> this code has been moved into the helper you introduced, correct?

Yes, this is the interface generation code. I just created a junk emitter to
remove the interfaces, otherwise all of the code to generate them is still here.
I was going to clean up all of this as a separate CL.

https://chromiumcodereview.appspot.com/11365019/diff/2083/sdk/lib/html/templa...
File sdk/lib/html/templates/html/impl/impl_CustomEvent.darttemplate (right):

https://chromiumcodereview.appspot.com/11365019/diff/2083/sdk/lib/html/templa...
sdk/lib/html/templates/html/impl/impl_CustomEvent.darttemplate:9: Object detail
= null]) => _$(CLASSNAME)FactoryProvider.create$CLASSNAME(
On 2012/11/06 12:28:43, Anton Muhin wrote:
> no need in = null

Done.

https://chromiumcodereview.appspot.com/11365019/diff/2083/sdk/lib/html/templa...
File sdk/lib/html/templates/html/impl/impl_Element.darttemplate (right):

https://chromiumcodereview.appspot.com/11365019/diff/2083/sdk/lib/html/templa...
sdk/lib/html/templates/html/impl/impl_Element.darttemplate:658: // Relative to
offsetParent
On 2012/11/06 12:28:43, Anton Muhin wrote:
> nit: missing trailing .

Done.

https://chromiumcodereview.appspot.com/11365019/diff/2083/sdk/lib/html/templa...
sdk/lib/html/templates/html/impl/impl_Element.darttemplate:687: // In global
coords
On 2012/11/06 12:28:43, Anton Muhin wrote:
> ditto

Done.

https://chromiumcodereview.appspot.com/11365019/diff/2083/sdk/lib/html/templa...
File sdk/lib/html/templates/html/impl/impl_IDBKeyRange.darttemplate (left):

https://chromiumcodereview.appspot.com/11365019/diff/2083/sdk/lib/html/templa...
sdk/lib/html/templates/html/impl/impl_IDBKeyRange.darttemplate:7: part of html;
On 2012/11/06 12:28:43, Anton Muhin wrote:
> why part of html is gone?

This line is not present in any of the implementation files, only the interface
files.

Does it need to be added to all of the implementation files?

https://chromiumcodereview.appspot.com/11365019/diff/2083/sdk/lib/html/templa...
File sdk/lib/html/templates/html/impl/pure_interface.darttemplate (right):

https://chromiumcodereview.appspot.com/11365019/diff/2083/sdk/lib/html/templa...
sdk/lib/html/templates/html/impl/pure_interface.darttemplate:2: abstract class
$CLASSNAME$EXTENDS$IMPLEMENTS {
On 2012/11/06 12:28:43, Anton Muhin wrote:
> why no part of html?
> 
> Why it should be different from interface.darttemplate?
None of the impls currently have this, I'll add a bug to put these in.

[Anton Muhin, 2012-11-07 08:50:55]

https://chromiumcodereview.appspot.com/11365019/diff/2083/sdk/lib/html/script...
File sdk/lib/html/scripts/systemhtml.py (right):

https://chromiumcodereview.appspot.com/11365019/diff/2083/sdk/lib/html/script...
sdk/lib/html/scripts/systemhtml.py:357: if typed_array_type:
Sorry, I meant you have pretty much the same code in the helper you've
introduced in this CL.  Can we reuse it here?

On 2012/11/06 22:11:44, blois wrote:
> On 2012/11/06 12:28:43, Anton Muhin wrote:
> > this code has been moved into the helper you introduced, correct?
> 
> Yes, this is the interface generation code. I just created a junk emitter to
> remove the interfaces, otherwise all of the code to generate them is still
here.
> I was going to clean up all of this as a separate CL.

https://chromiumcodereview.appspot.com/11365019/diff/2083/sdk/lib/html/templa...
File sdk/lib/html/templates/html/impl/pure_interface.darttemplate (right):

https://chromiumcodereview.appspot.com/11365019/diff/2083/sdk/lib/html/templa...
sdk/lib/html/templates/html/impl/pure_interface.darttemplate:2: abstract class
$CLASSNAME$EXTENDS$IMPLEMENTS {
Thanks!

On 2012/11/06 22:11:44, blois wrote:
> On 2012/11/06 12:28:43, Anton Muhin wrote:
> > why no part of html?
> > 
> > Why it should be different from interface.darttemplate?
> None of the impls currently have this, I'll add a bug to put these in.