code review 10379049: go.tools/go/types: simplified lookup

go.tools/go/types: simplified lookup

- much simpler lookup
- more result information
- will make tracking of pointer-ness easier

TODO: apply the same changes to method set computation

[gri, 2013-06-19 04:31:42]

Hello adonovan@google.com (cc: golang-dev@googlegroups.com),

I'd like you to review this change to
https://code.google.com/p/go.tools

[adonovan, 2013-06-19 17:52:15]

LGTM

https://codereview.appspot.com/10379049/diff/11001/go/types/lookup.go
File go/types/lookup.go (right):

https://codereview.appspot.com/10379049/diff/11001/go/types/lookup.go#newcode13
go/types/lookup.go:13: // The last index entry is the field or method index in
the (possible embedded)
s/possible/possibly/

https://codereview.appspot.com/10379049/diff/11001/go/types/lookup.go#newcode15
go/types/lookup.go:15: // the embedded fields traversed to get to the entry,
starting at depth 0.
The indices can be into three different lists:
1 the list of declared methods of a named type
2 the list of all methods (method set) of an interface type
3 the list of fields of a struct type.

Could you clarify this, and state the invariants for which indices may be of
which kinds?

https://codereview.appspot.com/10379049/diff/11001/go/types/lookup.go#newcode45
go/types/lookup.go:45: // typ must be an underlying type in this case.
Isn't it tautological to state that a non-named type is an underlying type?

[gri, 2013-06-19 19:28:47]

https://codereview.appspot.com/10379049/diff/11001/go/types/lookup.go
File go/types/lookup.go (right):

https://codereview.appspot.com/10379049/diff/11001/go/types/lookup.go#newcode13
go/types/lookup.go:13: // The last index entry is the field or method index in
the (possible embedded)
On 2013/06/19 17:52:15, adonovan wrote:
> s/possible/possibly/

Done.

https://codereview.appspot.com/10379049/diff/11001/go/types/lookup.go#newcode15
go/types/lookup.go:15: // the embedded fields traversed to get to the entry,
starting at depth 0.
On 2013/06/19 17:52:15, adonovan wrote:
> The indices can be into three different lists:
> 1 the list of declared methods of a named type
> 2 the list of all methods (method set) of an interface type
> 3 the list of fields of a struct type.
> 
> Could you clarify this, and state the invariants for which indices may be of
> which kinds?

Done.

https://codereview.appspot.com/10379049/diff/11001/go/types/lookup.go#newcode45
go/types/lookup.go:45: // typ must be an underlying type in this case.
On 2013/06/19 17:52:15, adonovan wrote:
> Isn't it tautological to state that a non-named type is an underlying type?

yes. fixed.

[gri, 2013-06-19 19:29:20]

*** Submitted as
https://code.google.com/p/go/source/detail?r=f40ac11d013b&repo=tools ***

go.tools/go/types: simplified lookup

- much simpler lookup
- more result information
- will make tracking of pointer-ness easier

TODO: apply the same changes to method set computation

R=adonovan
CC=golang-dev
https://codereview.appspot.com/10379049