Summarize constructor initializers.

Summarize constructor initializers.

R=paulberry@google.com
BUG=

Committed: https://github.com/dart-lang/sdk/commit/46850c69abd5bfb2024c448749994001613f29b7

[Paul Berry, 2016-02-08 14:53:12]

lgtm

https://codereview.chromium.org/1678753002/diff/1/pkg/analyzer/lib/src/summar...
File pkg/analyzer/lib/src/summary/idl.dart (right):

https://codereview.chromium.org/1678753002/diff/1/pkg/analyzer/lib/src/summar...
pkg/analyzer/lib/src/summary/idl.dart:851: * Depending on the [kind] is the name
of the field declared in the class,
Rather than make the reader figure out which possibility relates to which value
of [kind], let's just be specific and say:

"If [kind] is `field`, the name of the field declared in the class.  If [kind]
is `thisInvocation`, the name of the constructor to redirect to.  ..."

https://codereview.chromium.org/1678753002/diff/1/pkg/analyzer/lib/src/summar...
pkg/analyzer/lib/src/summary/idl.dart:858: * If
[UnlinkedConstructorInitializerKind.field], the expression of the
Nit: how about saying "If [kind] is `field`"?  It's easier to scan and it's
grammatical.

https://codereview.chromium.org/1678753002/diff/1/pkg/analyzer/lib/src/summar...
File pkg/analyzer/lib/src/summary/resynthesize.dart (right):

https://codereview.chromium.org/1678753002/diff/1/pkg/analyzer/lib/src/summar...
pkg/analyzer/lib/src/summary/resynthesize.dart:436: throw new
UnimplementedError('$operation');
Nit: this should have a TODO comment.

https://codereview.chromium.org/1678753002/diff/1/pkg/analyzer/lib/src/summar...
File pkg/analyzer/lib/src/summary/summarize_const_expr.dart (right):

https://codereview.chromium.org/1678753002/diff/1/pkg/analyzer/lib/src/summar...
pkg/analyzer/lib/src/summary/summarize_const_expr.dart:115: //      if (name !=
'identical') {
Can you include a comment explaining why this code is commented out?

https://codereview.chromium.org/1678753002/diff/1/pkg/analyzer/test/src/summa...
File pkg/analyzer/test/src/summary/summary_common.dart (right):

https://codereview.chromium.org/1678753002/diff/1/pkg/analyzer/test/src/summa...
pkg/analyzer/test/src/summary/summary_common.dart:57: print(st);
Was this temporary code you accidentally forgot to delete, or do you want to
keep it?

If you want to keep it, please add additional print statements to show the
exception and an explanation of what went wrong, e.g.:

} catch (exception, stackTrace) {
  print('Exception occurred during evaluation of sdkPublicNamespace');
  print(exception);
  print(stackTrace);

https://codereview.chromium.org/1678753002/diff/1/pkg/analyzer/test/src/summa...
pkg/analyzer/test/src/summary/summary_common.dart:2742:
test_constructor_initializers_superInvocation() {
Also test a super invocation of an unnamed constructor.

[scheglov, 2016-02-08 16:29:51]

https://codereview.chromium.org/1678753002/diff/1/pkg/analyzer/lib/src/summar...
File pkg/analyzer/lib/src/summary/idl.dart (right):

https://codereview.chromium.org/1678753002/diff/1/pkg/analyzer/lib/src/summar...
pkg/analyzer/lib/src/summary/idl.dart:851: * Depending on the [kind] is the name
of the field declared in the class,
On 2016/02/08 14:53:12, Paul Berry wrote:
> Rather than make the reader figure out which possibility relates to which
value
> of [kind], let's just be specific and say:
> 
> "If [kind] is `field`, the name of the field declared in the class.  If [kind]
> is `thisInvocation`, the name of the constructor to redirect to.  ..."

Done.

https://codereview.chromium.org/1678753002/diff/1/pkg/analyzer/lib/src/summar...
pkg/analyzer/lib/src/summary/idl.dart:858: * If
[UnlinkedConstructorInitializerKind.field], the expression of the
On 2016/02/08 14:53:12, Paul Berry wrote:
> Nit: how about saying "If [kind] is `field`"?  It's easier to scan and it's
> grammatical.

Done.

https://codereview.chromium.org/1678753002/diff/1/pkg/analyzer/lib/src/summar...
File pkg/analyzer/lib/src/summary/resynthesize.dart (right):

https://codereview.chromium.org/1678753002/diff/1/pkg/analyzer/lib/src/summar...
pkg/analyzer/lib/src/summary/resynthesize.dart:436: throw new
UnimplementedError('$operation');
On 2016/02/08 14:53:12, Paul Berry wrote:
> Nit: this should have a TODO comment.

Done.

https://codereview.chromium.org/1678753002/diff/1/pkg/analyzer/lib/src/summar...
File pkg/analyzer/lib/src/summary/summarize_const_expr.dart (right):

https://codereview.chromium.org/1678753002/diff/1/pkg/analyzer/lib/src/summar...
pkg/analyzer/lib/src/summary/summarize_const_expr.dart:115: //      if (name !=
'identical') {
On 2016/02/08 14:53:12, Paul Berry wrote:
> Can you include a comment explaining why this code is commented out?

It should not be commented out.
I will uncomment it.

https://codereview.chromium.org/1678753002/diff/1/pkg/analyzer/test/src/summa...
File pkg/analyzer/test/src/summary/summary_common.dart (right):

https://codereview.chromium.org/1678753002/diff/1/pkg/analyzer/test/src/summa...
pkg/analyzer/test/src/summary/summary_common.dart:57: print(st);
On 2016/02/08 14:53:12, Paul Berry wrote:
> Was this temporary code you accidentally forgot to delete, or do you want to
> keep it?
> 
> If you want to keep it, please add additional print statements to show the
> exception and an explanation of what went wrong, e.g.:
> 
> } catch (exception, stackTrace) {
>   print('Exception occurred during evaluation of sdkPublicNamespace');
>   print(exception);
>   print(stackTrace);

Another debugging code that accidentally got into the CL.
Removed.

https://codereview.chromium.org/1678753002/diff/1/pkg/analyzer/test/src/summa...
pkg/analyzer/test/src/summary/summary_common.dart:2742:
test_constructor_initializers_superInvocation() {
On 2016/02/08 14:53:12, Paul Berry wrote:
> Also test a super invocation of an unnamed constructor.

Done.

[scheglov, 2016-02-08 16:31:03]

Description was changed from

==========
Summarize constructor initializers.

R=paulberry@google.com
BUG=
==========

to

==========
Summarize constructor initializers.

R=paulberry@google.com
BUG=

Committed:
https://github.com/dart-lang/sdk/commit/46850c69abd5bfb2024c448749994001613f29b7
==========

[scheglov, 2016-02-08 16:31:05]

Committed patchset #2 (id:20001) manually as
46850c69abd5bfb2024c448749994001613f29b7 (presubmit successful).