This adds quite a lot of information though, because we send many
strings.
Before: 20686 characters.
After: 55345 characters.
Theoretically we could do better, and send information about existing
imports only once as well, so when the user continues typing in a
single file, without touching imports, we don't send any new data
about imports (maybe just a confirmation that it is still the same).
But I'm not sure if this is a worthwhile optimization.
Actually, even included suggestion sets have similar property - they
don't change for a given file, unless there are changes to other
libraries which we might want to include.
R=brianwilkerson@google.com
Change-Id: I2f55e2dc85508849146aa39eb279beabaec937c0
Reviewed-on: https://dart-review.googlesource.com/c/sdk/+/103561
Commit-Queue: Konstantin Shcheglov <scheglov@google.com>
Reviewed-by: Brian Wilkerson <brianwilkerson@google.com>
This includes:
* accepting null values like sdk-summary or .packages (which are not used when building kernel for the sdk itself)
* allow enabling language experiments
* only exclude-non-sources on the non-incremental code path if that's requested
Change-Id: I08eeb643676f1f1406f0f3030c341d68179d42a2
Reviewed-on: https://dart-review.googlesource.com/c/sdk/+/103574
Reviewed-by: Nate Bosch <nbosch@google.com>
This improves the missing class, mixin, and extension method body
error messages and lays the groundwork for improving other
missing body error messages such as switch statement, finally clause,
and others.
Change-Id: I0a4d1338fe91eee3cfe7d61652e168df1302d212
Reviewed-on: https://dart-review.googlesource.com/c/sdk/+/103570
Reviewed-by: Brian Wilkerson <brianwilkerson@google.com>
When performing type substitution, there are three types involved, the
substitution site (`this` in TypeImpl.substitute2), the type we are
replacing (`parameterType` in TypeImpl.substitute2), and the
replacement type (`argumentType` in TypeImpl.substitute2).
Previously, we only handled the nullabilities of `this` and
`argumentType`. This CL adds support for the possibility that
`parameterType` might have nullability "*". This happens because
TypeParameterElement.type returns a star type, so for instance when
subsituting `int` for `T` in `List<T>` to form `List<int>`, we are
actually substituting `int` for `T*` in `List<T*>`.
Change-Id: I0a61fdc47ec8aa205dc0c539c49ea7799ed4ac05
Reviewed-on: https://dart-review.googlesource.com/c/sdk/+/103542
Commit-Queue: Paul Berry <paulberry@google.com>
Reviewed-by: Brian Wilkerson <brianwilkerson@google.com>
Reviewed-by: Konstantin Shcheglov <scheglov@google.com>
For analyzer backend only:
* Implements `x!`
* Uses non-null type info where possible to elide null checks.
* Does not yet reify.
No hurry on this - starting to look at what's there. Some notes:
* NNBD doesn't seem to be exposed directly on DartType yet.
* x!.y doesn't seem to parse - need to write (x!).y.
* Can't run existing tests yet - `package:expect`, etc., have NNBD errors.
Change-Id: I58c24c950d00f3a40d789d4a7adf049786977c50
Reviewed-on: https://dart-review.googlesource.com/c/sdk/+/103040
Commit-Queue: Vijay Menon <vsm@google.com>
Reviewed-by: Paul Berry <paulberry@google.com>
During outline building the token stream was split after every field
initializer needed for top-level type inference. Since the token
stream is a doubly-linked list it's also necessary to split the token
stream before the field initializer in order to possibly collect the
unnecessary tokens.
Change-Id: Iafac8eee1eb87d43458f66ef08f6398634ded5df
Reviewed-on: https://dart-review.googlesource.com/c/sdk/+/103622
Auto-Submit: Kevin Millikin <kmillikin@google.com>
Commit-Queue: Dan Rubel <danrubel@google.com>
Reviewed-by: Dan Rubel <danrubel@google.com>
These builders have some unused generality: they have an unnecessary
type parameter and the implementation classes have unused fields. As
currently used, they could be replaced with an int but the class is
kept in anticipation of adding behavior (compiling annotations).
Change-Id: I068cb483a19b3f7eae846c212cbdbb29ca77973d
Reviewed-on: https://dart-review.googlesource.com/c/sdk/+/103529
Auto-Submit: Kevin Millikin <kmillikin@google.com>
Reviewed-by: Dmitry Stefantsov <dmitryas@google.com>
Commit-Queue: Kevin Millikin <kmillikin@google.com>
This is important as we will soon add support for compiling the sdk as a
module and we would like to only compile it once when running a suite of
tests.
+ also enable caching in the dart2js pipeline test.
Change-Id: Ic9043f868123164f3ab425ba73f7428416b05fc0
Reviewed-on: https://dart-review.googlesource.com/c/sdk/+/103485
Commit-Queue: Sigmund Cherem <sigmund@google.com>
Reviewed-by: Johnni Winther <johnniwinther@google.com>
In DEBUG mode during AOT compilation MetadataHelper::SetMetadataMappings
may spend considerable time verifying that node offsets are sorted
in kernel metadata. The problem is that there are a lot of MetadataHelpers
created (several per each compiled function).
This change moves verification of kernel metadata mappings to
the kernel loader, so it is performed only once per kernel program.
AOT gen_snapshot time on a small test:
Before: 11.680s
After: 9.041s
Change-Id: I1c0a676b0ed28fc5cfa756ca60159f914190fac0
Reviewed-on: https://dart-review.googlesource.com/c/sdk/+/103486
Commit-Queue: Alexander Markov <alexmarkov@google.com>
Reviewed-by: Aart Bik <ajcbik@google.com>
Reviewed-by: Martin Kustermann <kustermann@google.com>
This type is equivalent to the existing type `Null`, but we need it
anyway because `Null` can only be accessed via the type provider, and
there are circumstances where we need to create this type and don't
have access to the type provider.
It also may prove beneficial to be able to distinguish between `Null`
and `Never?` in diagnostic messages.
Change-Id: I6118e87c8c4736a508a3f366f2762a96198db7ca
Reviewed-on: https://dart-review.googlesource.com/c/sdk/+/103540
Reviewed-by: Konstantin Shcheglov <scheglov@google.com>
Reviewed-by: Mike Fairhurst <mfairhurst@google.com>
Reviewed-by: Brian Wilkerson <brianwilkerson@google.com>
Rename old "Pub client" headline for consistency with older entries
Add headline for linter in CHANGELOG.md
Change-Id: I118127a6aab564f5d498441f30957e8851e69d6c
Reviewed-on: https://dart-review.googlesource.com/c/sdk/+/103527
Reviewed-by: Sigurd Meldgaard <sigurdm@google.com>
In addition, if a _ByteDataView came from a use of the ByteData
constructor, its offset is always 0.
Also add appropriate canonicalization for the following instructions for
typed data views coming from a known factory call:
* LoadField getting type arguments
* GuardFieldLength
This closes https://github.com/dart-lang/sdk/issues/36570.
Change-Id: I1c045edb2d928c3bb3340d9d12009c2afa66febb
Reviewed-on: https://dart-review.googlesource.com/c/sdk/+/102362
Reviewed-by: Vyacheslav Egorov <vegorov@google.com>
Commit-Queue: Teagan Strickland <sstrickl@google.com>
This adds a "--deflake" flag to test.dart. Deflaking will only run if
that flag is set.
Deflaking is time consuming and delays the information that users want
to see when most of the time they know that the result is not caused by
flakiness.
A typical session may look like this:
* Run test.dart with a broad test selector without deflaking.
* Re-run test.dart with a narrow test selector for suspected flakes with
"--deflake". This second step is optional.
Additionally, a new "--report-flakes" flag is added that can be
used to report test failures for tests known to be flaky.
Change-Id: I543d0b40c32065eb0a50338c55e7050b7887abce
Reviewed-on: https://dart-review.googlesource.com/c/sdk/+/102381
Reviewed-by: Jonas Termansen <sortie@google.com>
Jaime Wren