Implement call operator.

Implement call operator.

Committed: https://code.google.com/p/dart/source/detail?r=10495

Committed: https://code.google.com/p/dart/source/detail?r=10507

[ahe, 2012-08-10 09:11:35]

LGTM!

https://chromiumcodereview.appspot.com/10834270/diff/1/lib/compiler/implement...
File lib/compiler/implementation/resolver.dart (right):

https://chromiumcodereview.appspot.com/10834270/diff/1/lib/compiler/implement...
lib/compiler/implementation/resolver.dart:1197: (target == null ||
=== for consistency with the rest of the file.

https://chromiumcodereview.appspot.com/10834270/diff/1/lib/compiler/implement...
lib/compiler/implementation/resolver.dart:1198: target.isGetter() ||
I think you also need:

target.isInstanceMember() ||

[ahe, 2012-08-10 09:15:49]

https://chromiumcodereview.appspot.com/10834270/diff/1/lib/compiler/implement...
File lib/compiler/implementation/resolver.dart (right):

https://chromiumcodereview.appspot.com/10834270/diff/1/lib/compiler/implement...
lib/compiler/implementation/resolver.dart:1198: target.isGetter() ||
On 2012/08/10 09:11:35, ahe wrote:
> I think you also need:
> 
> target.isInstanceMember() ||

Never mind.

[kasperl, 2012-08-10 10:16:28]

LGTM.

https://chromiumcodereview.appspot.com/10834270/diff/1/lib/compiler/implement...
File lib/compiler/implementation/namer.dart (right):

https://chromiumcodereview.appspot.com/10834270/diff/1/lib/compiler/implement...
lib/compiler/implementation/namer.dart:110: if (name == const
SourceString("call")) name = CLOSURE_INVOCATION_NAME;
What's the intent here? If the name is already 'call' you want to make it
'call'? Weird. I guess you want to map the user-defined 'call' method to
whatever we've decided to call the method we use to invoke closures. Maybe
introduce a constant for the RHS of the comparison to make it clearer?

https://chromiumcodereview.appspot.com/10834270/diff/1/tests/language/call_op...
File tests/language/call_operator_test.dart (right):

https://chromiumcodereview.appspot.com/10834270/diff/1/tests/language/call_op...
tests/language/call_operator_test.dart:33: String result = "";
How about using a StringBuffer here? Seems cleaner.

https://chromiumcodereview.appspot.com/10834270/diff/1/tests/language/call_op...
tests/language/call_operator_test.dart:36: if (i < count -1) {
-1 => - 1

[floitsch, 2012-08-10 12:19:16]

https://chromiumcodereview.appspot.com/10834270/diff/1/lib/compiler/implement...
File lib/compiler/implementation/namer.dart (right):

https://chromiumcodereview.appspot.com/10834270/diff/1/lib/compiler/implement...
lib/compiler/implementation/namer.dart:110: if (name == const
SourceString("call")) name = CLOSURE_INVOCATION_NAME;
On 2012/08/10 10:16:28, kasperl wrote:
> What's the intent here? If the name is already 'call' you want to make it
> 'call'? Weird. I guess you want to map the user-defined 'call' method to
> whatever we've decided to call the method we use to invoke closures. Maybe
> introduce a constant for the RHS of the comparison to make it clearer?

Changed:
CLOSURE_INVOCATION_NAME is now initialized to Compiler.CALL_OPERATOR_NAME.
(which is set to "call").

The 'if' here and below has been removed.

https://chromiumcodereview.appspot.com/10834270/diff/1/lib/compiler/implement...
File lib/compiler/implementation/resolver.dart (right):

https://chromiumcodereview.appspot.com/10834270/diff/1/lib/compiler/implement...
lib/compiler/implementation/resolver.dart:1197: (target == null ||
On 2012/08/10 09:11:35, ahe wrote:
> === for consistency with the rest of the file.

Done.

https://chromiumcodereview.appspot.com/10834270/diff/1/tests/language/call_op...
File tests/language/call_operator_test.dart (right):

https://chromiumcodereview.appspot.com/10834270/diff/1/tests/language/call_op...
tests/language/call_operator_test.dart:33: String result = "";
On 2012/08/10 10:16:28, kasperl wrote:
> How about using a StringBuffer here? Seems cleaner.

Done.

https://chromiumcodereview.appspot.com/10834270/diff/1/tests/language/call_op...
tests/language/call_operator_test.dart:36: if (i < count -1) {
On 2012/08/10 10:16:28, kasperl wrote:
> -1 => - 1

Done.

[floitsch, 2012-08-10 14:19:09]

Fix submitted TBR=ahe.

[ahe, 2012-08-16 20:53:14]

Still LGTM.

FWIW, Kasper already got rid of the extra loop.