Introduce initial plumbing for type inference in fasta.

Introduce initial plumbing for type inference in fasta.

R=ahe@google.com

Committed: https://github.com/dart-lang/sdk/commit/6d737cb68459ebd53010ecded7fa1d3ea04cebea

[Paul Berry, 2017-04-18 15:07:56]

paulberry@google.com changed reviewers:
+ ahe@google.com

[ahe, 2017-04-18 15:50:00]

lgtm

Only nits below.

https://codereview.chromium.org/2823243002/diff/1/pkg/front_end/lib/src/fasta...
File pkg/front_end/lib/src/fasta/kernel/body_builder.dart (right):

https://codereview.chromium.org/2823243002/diff/1/pkg/front_end/lib/src/fasta...
pkg/front_end/lib/src/fasta/kernel/body_builder.dart:14: import
'package:front_end/src/fasta/kernel/kernel_shadow_ast.dart'
Make relative?

https://codereview.chromium.org/2823243002/diff/1/pkg/front_end/lib/src/fasta...
pkg/front_end/lib/src/fasta/kernel/body_builder.dart:89: _typeInferrer;
I prefer to keep things public unless implementing public API.

https://codereview.chromium.org/2823243002/diff/1/pkg/front_end/lib/src/fasta...
File pkg/front_end/lib/src/fasta/kernel/kernel_shadow_ast.dart (right):

https://codereview.chromium.org/2823243002/diff/1/pkg/front_end/lib/src/fasta...
pkg/front_end/lib/src/fasta/kernel/kernel_shadow_ast.dart:20: import
'package:front_end/src/fasta/type_inference/type_inferrer.dart';
Relative?

https://codereview.chromium.org/2823243002/diff/1/pkg/front_end/lib/src/fasta...
File pkg/front_end/lib/src/fasta/source/diet_listener.dart (right):

https://codereview.chromium.org/2823243002/diff/1/pkg/front_end/lib/src/fasta...
pkg/front_end/lib/src/fasta/source/diet_listener.dart:7: import
'package:front_end/src/fasta/type_inference/type_inferrer.dart'
Relative?

https://codereview.chromium.org/2823243002/diff/1/pkg/front_end/lib/src/fasta...
File pkg/front_end/lib/src/fasta/source/source_loader.dart (right):

https://codereview.chromium.org/2823243002/diff/1/pkg/front_end/lib/src/fasta...
pkg/front_end/lib/src/fasta/source/source_loader.dart:11: import
'package:front_end/src/fasta/kernel/kernel_shadow_ast.dart'
Relative?

https://codereview.chromium.org/2823243002/diff/1/pkg/front_end/lib/src/fasta...
File pkg/front_end/lib/src/fasta/type_inference/type_inferrer.dart (right):

https://codereview.chromium.org/2823243002/diff/1/pkg/front_end/lib/src/fasta...
pkg/front_end/lib/src/fasta/type_inference/type_inferrer.dart:5: import
'package:kernel/ast.dart';
May want to use show on this import to be explicit about being independent of
the AST.

https://codereview.chromium.org/2823243002/diff/1/pkg/front_end/lib/src/fasta...
pkg/front_end/lib/src/fasta/type_inference/type_inferrer.dart:10: /// underlying
AST representation.
But still using DartType from kernel?

[Paul Berry, 2017-04-18 16:33:17]

https://codereview.chromium.org/2823243002/diff/1/pkg/front_end/lib/src/fasta...
File pkg/front_end/lib/src/fasta/kernel/body_builder.dart (right):

https://codereview.chromium.org/2823243002/diff/1/pkg/front_end/lib/src/fasta...
pkg/front_end/lib/src/fasta/kernel/body_builder.dart:14: import
'package:front_end/src/fasta/kernel/kernel_shadow_ast.dart'
On 2017/04/18 15:49:59, ahe wrote:
> Make relative?

I actually prefer absolute imports, and I was hoping that we could transition
all imports in fasta to absolute ones--let's talk about this in person.

https://codereview.chromium.org/2823243002/diff/1/pkg/front_end/lib/src/fasta...
pkg/front_end/lib/src/fasta/kernel/body_builder.dart:89: _typeInferrer;
On 2017/04/18 15:49:59, ahe wrote:
> I prefer to keep things public unless implementing public API.

Heh, I prefer the reverse.  Let's talk about this too :)

https://codereview.chromium.org/2823243002/diff/1/pkg/front_end/lib/src/fasta...
File pkg/front_end/lib/src/fasta/type_inference/type_inferrer.dart (right):

https://codereview.chromium.org/2823243002/diff/1/pkg/front_end/lib/src/fasta...
pkg/front_end/lib/src/fasta/type_inference/type_inferrer.dart:5: import
'package:kernel/ast.dart';
On 2017/04/18 15:49:59, ahe wrote:
> May want to use show on this import to be explicit about being independent of
> the AST.

Done.

https://codereview.chromium.org/2823243002/diff/1/pkg/front_end/lib/src/fasta...
pkg/front_end/lib/src/fasta/type_inference/type_inferrer.dart:10: /// underlying
AST representation.
On 2017/04/18 15:49:59, ahe wrote:
> But still using DartType from kernel?

Yes, for now at least.  I don't know if this will change or not.  I've added
documentation to this effect.

[Paul Berry, 2017-04-18 16:37:58]

Description was changed from

==========
Introduce initial plumbing for type inference in fasta.
==========

to

==========
Introduce initial plumbing for type inference in fasta.

R=ahe@google.com

Committed:
https://github.com/dart-lang/sdk/commit/6d737cb68459ebd53010ecded7fa1d3ea04cebea
==========

[Paul Berry, 2017-04-18 16:38:00]

Committed patchset #1 (id:1) manually as
6d737cb68459ebd53010ecded7fa1d3ea04cebea (presubmit successful).

[scheglov, 2017-04-18 20:09:48]

scheglov@google.com changed reviewers:
+ scheglov@google.com

[scheglov, 2017-04-18 20:09:49]

lgtm

https://codereview.chromium.org/2823243002/diff/1/pkg/front_end/lib/src/fasta...
File pkg/front_end/lib/src/fasta/kernel/kernel_shadow_ast.dart (right):

https://codereview.chromium.org/2823243002/diff/1/pkg/front_end/lib/src/fasta...
pkg/front_end/lib/src/fasta/kernel/kernel_shadow_ast.dart:38: /// Calls back to
[inferrer] to perform type inference for whatever concrete
The reference [inferrer] is not resolved here.

[Paul Berry, 2017-04-18 22:49:14]

https://codereview.chromium.org/2823243002/diff/1/pkg/front_end/lib/src/fasta...
File pkg/front_end/lib/src/fasta/kernel/kernel_shadow_ast.dart (right):

https://codereview.chromium.org/2823243002/diff/1/pkg/front_end/lib/src/fasta...
pkg/front_end/lib/src/fasta/kernel/kernel_shadow_ast.dart:38: /// Calls back to
[inferrer] to perform type inference for whatever concrete
On 2017/04/18 20:09:48, scheglov wrote:
> The reference [inferrer] is not resolved here.

Good catch, thanks.  I meant the signature to be:

DartType _inferExpression(KernelTypeInferrer inferrer, DartType typeContext,
bool typeNeeded);

I'll fix in a follow-up CL.