Promote class to a reserved word

Promote class to a reserved word

Promote class as a reserved word http://code.google.com/p/dart/issues/detail?id=870
Cleaning old test that had invalid assumptions; makeing two new tests
(jUnit and shared test).  Marking Frog and VM as failing.

R=mmendez
BUG=
TEST=

Committed: https://code.google.com/p/dart/source/detail?r=2550

[mmendez, 2011-12-16 15:58:24]

lgtm with nits

http://codereview.chromium.org/8972003/diff/1/compiler/java/com/google/dart/c...
File compiler/java/com/google/dart/compiler/parser/Token.java (right):

http://codereview.chromium.org/8972003/diff/1/compiler/java/com/google/dart/c...
compiler/java/com/google/dart/compiler/parser/Token.java:93: CLASS("class",0),
Note that extends, false, is, null and true are not in this block of keywords.

http://codereview.chromium.org/8972003/diff/1/tests/language/language.status
File tests/language/language.status (right):

http://codereview.chromium.org/8972003/diff/1/tests/language/language.status#...
tests/language/language.status:46: ClassKeywordNegativeTest/01: Fail # Issue 873
Nit: Since your are fixing issue 873 you might want to create new issues for
frog and the VM.

http://codereview.chromium.org/8972003/diff/1/tests/language/src/ClassKeyword...
File tests/language/src/ClassKeywordNegativeTest.dart (right):

http://codereview.chromium.org/8972003/diff/1/tests/language/src/ClassKeyword...
tests/language/src/ClassKeywordNegativeTest.dart:9: int class = 10; /// 01:
compile-time error
Nit: I'd at least CC people from the language (gbracha?), VM (iposva?) and frog
(jimhjug?) on this change.  Also, you are checking two things here, some parsers
may stop at the first issue.  You may want to split this test.

I would have expected co19 to have a keyword suite, but it doesn't as far as I
can tell.

[codefu, 2011-12-16 16:56:29]

Adding +gbracha, +iposva, and +jimhug to CC for their opinions on this change.

http://codereview.chromium.org/8972003/diff/1/compiler/java/com/google/dart/c...
File compiler/java/com/google/dart/compiler/parser/Token.java (right):

http://codereview.chromium.org/8972003/diff/1/compiler/java/com/google/dart/c...
compiler/java/com/google/dart/compiler/parser/Token.java:93: CLASS("class",0),
On 2011/12/16 15:58:24, mmendez wrote:
> Note that extends, false, is, null and true are not in this block of keywords.

true, false, is, and null are else where in this token list (Literals & Compare
Operators).  I can add extends as part of this CL or another one.

http://codereview.chromium.org/8972003/diff/1/tests/language/language.status
File tests/language/language.status (right):

http://codereview.chromium.org/8972003/diff/1/tests/language/language.status#...
tests/language/language.status:46: ClassKeywordNegativeTest/01: Fail # Issue 873
On 2011/12/16 15:58:24, mmendez wrote:
> Nit: Since your are fixing issue 873 you might want to create new issues for
> frog and the VM.

I'm fixing issue 870. 873 and 874 are for Frog & the VM.

http://codereview.chromium.org/8972003/diff/1/tests/language/src/ClassKeyword...
File tests/language/src/ClassKeywordNegativeTest.dart (right):

http://codereview.chromium.org/8972003/diff/1/tests/language/src/ClassKeyword...
tests/language/src/ClassKeywordNegativeTest.dart:9: int class = 10; /// 01:
compile-time error
On 2011/12/16 15:58:24, mmendez wrote:
> Nit: I'd at least CC people from the language (gbracha?), VM (iposva?) and
frog
> (jimhjug?) on this change.  Also, you are checking two things here, some
parsers
> may stop at the first issue.  You may want to split this test.
> 
> I would have expected co19 to have a keyword suite, but it doesn't as far as I
> can tell.

Negative tests follow a different format and it looks for the /// 01 and /// 02
would be treated as two tests from what I could tell.  Perhaps I'm expecting to
much? Will add people to the CC.

[jimhug, 2011-12-16 17:47:47]

http://codereview.chromium.org/8972003/diff/1/tests/language/language.status
File tests/language/language.status (right):

http://codereview.chromium.org/8972003/diff/1/tests/language/language.status#...
tests/language/language.status:189: ClassKeywordNegativeTest/01: Fail # Issue
874
I'm a little bit surprised that you don't have a ClassKeywordNegativeTest/02:
Fail as well here.  Is that because frog and the VM are handling the second
error correctly?

http://codereview.chromium.org/8972003/diff/1/tests/language/src/ClassKeyword...
File tests/language/src/ClassKeywordNegativeTest.dart (right):

http://codereview.chromium.org/8972003/diff/1/tests/language/src/ClassKeyword...
tests/language/src/ClassKeywordNegativeTest.dart:9: int class = 10; /// 01:
compile-time error
This should behave as two negative tests - plus one positive test that the code
will compile with both of these lines removed.  That looks good to me and as you
intended.

BTW - As a future TODO, I'd really like to see our testing infrastructure verify
that tests occur on the correct line number as it is really easy to pass a
negative test with a horrible compiler failure - but that's not at all related
to your current change.

On 2011/12/16 16:56:30, codefu wrote:
> On 2011/12/16 15:58:24, mmendez wrote:
> > Nit: I'd at least CC people from the language (gbracha?), VM (iposva?) and
> frog
> > (jimhjug?) on this change.  Also, you are checking two things here, some
> parsers
> > may stop at the first issue.  You may want to split this test.
> > 
> > I would have expected co19 to have a keyword suite, but it doesn't as far as
I
> > can tell.
> 
> Negative tests follow a different format and it looks for the /// 01 and ///
02
> would be treated as two tests from what I could tell.  Perhaps I'm expecting
to
> much? Will add people to the CC.

[codefu, 2011-12-16 19:11:59]

http://codereview.chromium.org/8972003/diff/1/tests/language/language.status
File tests/language/language.status (right):

http://codereview.chromium.org/8972003/diff/1/tests/language/language.status#...
tests/language/language.status:189: ClassKeywordNegativeTest/01: Fail # Issue
874
On 2011/12/16 17:47:47, jimhug wrote:
> I'm a little bit surprised that you don't have a ClassKeywordNegativeTest/02:
> Fail as well here.  Is that because frog and the VM are handling the second
> error correctly?

I'm not really sure why there's not a /02 on this one. Will the negative test
remove the first /// 01 line and see if the second line compiles?  It should
still be an error in that case, but undefined perhaps since there isn't a valid
variable named class.

http://codereview.chromium.org/8972003/diff/1/tests/language/src/ClassKeyword...
File tests/language/src/ClassKeywordNegativeTest.dart (right):

http://codereview.chromium.org/8972003/diff/1/tests/language/src/ClassKeyword...
tests/language/src/ClassKeywordNegativeTest.dart:9: int class = 10; /// 01:
compile-time error
On 2011/12/16 17:47:47, jimhug wrote:
> This should behave as two negative tests - plus one positive test that the
code
> will compile with both of these lines removed.  That looks good to me and as
you
> intended.
> 
> BTW - As a future TODO, I'd really like to see our testing infrastructure
verify
> that tests occur on the correct line number as it is really easy to pass a
> negative test with a horrible compiler failure - but that's not at all related
> to your current change.
> 
> On 2011/12/16 16:56:30, codefu wrote:
> > On 2011/12/16 15:58:24, mmendez wrote:
> > > Nit: I'd at least CC people from the language (gbracha?), VM (iposva?) and
> > frog
> > > (jimhjug?) on this change.  Also, you are checking two things here, some
> > parsers
> > > may stop at the first issue.  You may want to split this test.
> > > 
> > > I would have expected co19 to have a keyword suite, but it doesn't as far
as
> I
> > > can tell.
> > 
> > Negative tests follow a different format and it looks for the /// 01 and ///
> 02
> > would be treated as two tests from what I could tell.  Perhaps I'm expecting
> to
> > much? Will add people to the CC.
> 

Eric and  Konstantin really done a great job on our junit tests.  We test for
line, column & length, and error code.

[jimhug, 2011-12-16 22:36:12]

> Eric and  Konstantin really done a great job on our junit tests.  We test for
> line, column & length, and error code.

Yep, I saw those and was feeling a little jealous.  We're going to have to
figure out how to get the benefits of those for frog.

In case you were waiting, this lgtm so long as the tests are "passing" in frog
and the VM - i.e. this won't break the build.