Check that class member names do not conflict with type parameters

Check that class member names do not conflict with type parameters

Implement check that member names do not conflict with type
parameters. Simplify name checks in class namespace.
Mark typevariable_mirror_metadata_test as failing now that we
report the proper compile time errors. (The test expects no errors.)
Some co19 tests expect a name conflict between a setter X= and type
parameter X. This is still in flux from a language perspective. This
change list does not introduce an compile error for this case.

Fix for issue 13701.

R=srdjan@google.com

Committed: https://code.google.com/p/dart/source/detail?r=28808

[srdjan, 2013-10-17 17:26:27]

lgtm

https://codereview.chromium.org/27619002/diff/7001/runtime/vm/parser.cc
File runtime/vm/parser.cc (right):

https://codereview.chromium.org/27619002/diff/7001/runtime/vm/parser.cc#newco...
runtime/vm/parser.cc:554: const char* Kind() const {
I would expect Kind() to return content of field RawFunction::Kind kind. Maybe
rename to 'ToCString()'?

https://codereview.chromium.org/27619002/diff/7001/runtime/vm/parser.cc#newco...
runtime/vm/parser.cc:3017: &String::Handle(String::Concat(*method->name,
Symbols::Equals()));
s/Handle/ZoneHandle/

https://codereview.chromium.org/27619002/diff/7001/runtime/vm/parser.cc#newco...
runtime/vm/parser.cc:3247: Field& class_field = Field::Handle();
s/Handle/ZoneHandle/ as it escapes scope via field->field_.

[hausner, 2013-10-17 17:39:03]

https://codereview.chromium.org/27619002/diff/7001/runtime/vm/parser.cc
File runtime/vm/parser.cc (right):

https://codereview.chromium.org/27619002/diff/7001/runtime/vm/parser.cc#newco...
runtime/vm/parser.cc:554: const char* Kind() const {
On 2013/10/17 17:26:27, srdjan wrote:
> I would expect Kind() to return content of field RawFunction::Kind kind. Maybe
> rename to 'ToCString()'?

Done.

https://codereview.chromium.org/27619002/diff/7001/runtime/vm/parser.cc#newco...
runtime/vm/parser.cc:3017: &String::Handle(String::Concat(*method->name,
Symbols::Equals()));
On 2013/10/17 17:26:27, srdjan wrote:
> s/Handle/ZoneHandle/

Done.

https://codereview.chromium.org/27619002/diff/7001/runtime/vm/parser.cc#newco...
runtime/vm/parser.cc:3247: Field& class_field = Field::Handle();
On 2013/10/17 17:26:27, srdjan wrote:
> s/Handle/ZoneHandle/ as it escapes scope via field->field_.

It has escaped before, too, into the list of fields in 'members'. Did we just
get lucky?

Done

[hausner, 2013-10-17 17:39:17]

Committed patchset #5 manually as r28808 (presubmit successful).

[srdjan, 2013-10-17 17:48:13]

https://codereview.chromium.org/27619002/diff/7001/runtime/vm/parser.cc
File runtime/vm/parser.cc (right):

https://codereview.chromium.org/27619002/diff/7001/runtime/vm/parser.cc#newco...
runtime/vm/parser.cc:3247: Field& class_field = Field::Handle();
On 2013/10/17 17:39:03, hausner wrote:
> On 2013/10/17 17:26:27, srdjan wrote:
> > s/Handle/ZoneHandle/ as it escapes scope via field->field_.
> 
> It has escaped before, too, into the list of fields in 'members'. Did we just
> get lucky?
> 
> Done

Right now we do not set up HandleScopes eagerly, but they have a large scope
(whole parser or more).