Reverts: flutter/flutter#152487
Initiated by: gmackall
Reason for reverting: I forgot that I need to override the compileSdkVersion in the AGP 8.0 [instance of this test](ef9cd32f5a/dev/devicelab/bin/tasks/android_java17_dependency_smoke_tests.dart (L19))
Original PR Author: gmackall
Reviewed By: {reidbaker}
This change reverts the following previous change:
Updates `compileSdk`, `targetSdk`, and `ndk` versions (former 2 to latest, latter to the version of the ndk we are hosting on CIPD).
Summary of changes:
- Updates mentioned template values
- `compileSdk` 35 requires AGP 8.0+, so updated to 8.1 in many places.
- This also necessitated Gradle upgrades in most places
- This also necessitated moving the `package` xml attribute to the AGP `namespace` field in a couple places (test + template).
- Some tests use the output of `flutter create` but then use intentionally lower AGP versions. [I downgraded the `compileSdk` in these tests.](fee34fd61a)
- [Stopped lockfile generation](82324a2570) script from hitting the `hello_world` example because it uses `.kts` gradle files.
- One test needed [some Gradle options we had already added to templates](6aa187b4b6).
Updates `compileSdk`, `targetSdk`, and `ndk` versions (former 2 to latest, latter to the version of the ndk we are hosting on CIPD).
Summary of changes:
- Updates mentioned template values
- `compileSdk` 35 requires AGP 8.0+, so updated to 8.1 in many places.
- This also necessitated Gradle upgrades in most places
- This also necessitated moving the `package` xml attribute to the AGP `namespace` field in a couple places (test + template).
- Some tests use the output of `flutter create` but then use intentionally lower AGP versions. [I downgraded the `compileSdk` in these tests.](fee34fd61a)
- [Stopped lockfile generation](82324a2570) script from hitting the `hello_world` example because it uses `.kts` gradle files.
- One test needed [some Gradle options we had already added to templates](6aa187b4b6).
The stream subscriptions in the device's VmService are used by other parts of FlutterVmService and other components throughout flutter_tools. Components that listen to streams should not call VmService.streamCancel because that will interfere with other users who still want the events.
See https://github.com/flutter/flutter/issues/153049
See https://github.com/flutter/flutter/issues/153563
Hide the `--web-renderer` option in the Flutter Tool. The defaults already cover all fully supported modes:
- `flutter build web` provides canvaskit + dart2js
- `flutter build web --wasm` provides skwasm + dart2wasm
We do not want to encourage production usage of any other permutations (e.g. `auto` or `html`), in particular those that simply do not work (e.g. `skwasm` + dart2js).
Fixes https://github.com/flutter/flutter/issues/140096
Fixes https://github.com/flutter/flutter/issues/151786
Reverts: flutter/flutter#152358
Initiated by: zanderso
Reason for reverting: Speculative revert to determine whether this PR is related to https://github.com/flutter/flutter/issues/153026
Original PR Author: andrewkolos
Reviewed By: {christopherfujino}
This change reverts the following previous change:
Contributes to fixing https://github.com/flutter/flutter/issues/137184.
Cleaned up version of earlier PR, https://github.com/flutter/flutter/pull/152187.
This PR guards all the writes to `Process::stdin` by wrapping them with `ProcessUtils.writelnToStdinUnsafe`. This way, if any writes fail, we should at least get a stacktrace in our crash reporting.
Opening a Swift package in Xcode generates `.build` and `.swiftpm` directories. These should be ignored as they contain intermediary build artifacts that aren't meant to be checked-in.
Part of https://github.com/flutter/flutter/issues/148018
Nullable types for values in map patterns require the key to be present.
Since the 'uri' key is not always present in DDS exception responses,
this was causing us to fall back to throwing a StateError.
Fixes https://github.com/flutter/flutter/issues/152684
If Swift Package Manager is enabled, the tool generates a Swift package at `<ios/macos>/Flutter/ephemeral/Packages/FlutterGeneratedPluginSwiftPackage/`. This Swift package is how the tool adds plugins to the Flutter project.
SwiftPM is strictly enforces platform versions: you cannot depend on a Swift package if its supported version is higher than your own.
On iOS, we use the project's minimum deployment version for the generated Swift package. If a plugin has a higher requirement, you'll need to update your project's minimum deployment version. The generated Swift package is automatically updated the next time you run the tool.
This updates macOS to do the same thing.
Fixes https://github.com/flutter/flutter/issues/146204
This reverts commit 7cdc23b3e1.
The failure in the `native_assets_test` integration test on Windows was caused by the DevTools process not being shutdown by the `ColdRunner` when running the profile mode portion of the test. This resulted in the test being unable to clean up the project created by the test as DevTools was still holding onto a handle within the directory. This PR adds back the mistakenly removed DevTools shutdown logic in the `ColdRunner`.
Turns out just supporting the right value for `kDebugMode` was a lot simpler than I thought. Debug builds used to never go through the build system code path when using `flutter run`, but now that we have wasm this can occur with the run command.
This should address https://github.com/flutter/flutter/issues/148850
Reverts: flutter/flutter#152049
Initiated by: cbracken
Reason for reverting: iOS builds failing in post-submit
Original PR Author: loic-sharma
Reviewed By: {jmagman}
This change reverts the following previous change:
Changes:
1. Enables Swift Package Manager by default on the main/master channel
2. Fixes tests that fail if Swift Package Manager is enabled
Corresponding docs change: https://github.com/flutter/website/pull/10938
Addresses https://github.com/flutter/flutter/issues/151567
1. Instead of getting the `FULL_PRODUCT_NAME` Xcode build setting (`Runner.app`) instead use `PRODUCT_NAME` since most places really want the product name, and the extension stripping wasn't correct when the name contained periods.
2. Don't instruct the user to open the `xcarchive` in Xcode if it doesn't exist.
Fixes https://github.com/flutter/flutter/issues/140212
For the necessary background knowledge, see the flutter.dev content on [Resolution-aware image assets](https://docs.flutter.dev/ui/assets/assets-and-images#resolution-aware) and [Conditional bundling of assets based on app flavor](https://docs.flutter.dev/ui/assets/assets-and-images#conditional-bundling-of-assets-based-on-app-flavor) if you don't have a basic understanding of these features.
Fixes https://github.com/flutter/flutter/issues/151813 by using unique temporary directories, per asset file, for transformations. Currently, only a single directory is used and the name of the temporary files was based only on the basename of files. This means that `assets/image.png` and `assets/2x/image.png` would share an output path (`<temp dir path>/image.png`), causing a race. If this quick and rough explanation is a bit confusing, the original issueâ#151813âprovides a full repro and correct identification of the exact cause of the failure that can occur in the asset transformation process.
Reverts: flutter/flutter#146593
Initiated by: zanderso
Reason for reverting: Consistently failing `Windows_android native_assets_android` as in https://ci.chromium.org/ui/p/flutter/builders/prod/Windows_android%20native_assets_android/2533/overview
Original PR Author: bkonyi
Reviewed By: {christopherfujino, kenzieschmoll}
This change reverts the following previous change:
This change is a major step towards moving away from shipping DDS via Pub.
The first component of this PR is the move away from importing package:dds to launch DDS. Instead, DDS is launched out of process using the `dart development-service` command shipped with the Dart SDK. This makes Flutter's handling of DDS consistent with the standalone Dart VM.
The second component of this PR is the initial work to prepare for the removal of instances of DevTools being served manually by the flutter_tool, instead relying on DDS to serve DevTools. This will be consistent with how the standalone Dart VM serves DevTools, tying the DevTools lifecycle to a live DDS instance. This will allow for the removal of much of the logic needed to properly manage the lifecycle of the DevTools server in a future PR. Also, by serving DevTools from DDS, users will no longer need to forward a secondary port in remote workflows as DevTools will be available on the DDS port.
There's two remaining circumstances that will prevent us from removing DevtoolsRunner completely:
- The daemon's `devtools.serve` endpoint
- `flutter drive`'s `--profile-memory` flag used for recording memory profiles
This PR also includes some refactoring around `DebuggingOptions` to reduce the number of debugging related arguments being passed as parameters adjacent to a `DebuggingOptions` instance.
This change is a major step towards moving away from shipping DDS via
Pub.
The first component of this PR is the move away from importing
package:dds to launch DDS. Instead, DDS is launched out of process using
the `dart development-service` command shipped with the Dart SDK. This
makes Flutter's handling of DDS consistent with the standalone Dart VM.
The second component of this PR is the initial work to prepare for the
removal of instances of DevTools being served manually by the
flutter_tool, instead relying on DDS to serve DevTools. This will be
consistent with how the standalone Dart VM serves DevTools, tying the
DevTools lifecycle to a live DDS instance. This will allow for the
removal of much of the logic needed to properly manage the lifecycle of
the DevTools server in a future PR. Also, by serving DevTools from DDS,
users will no longer need to forward a secondary port in remote
workflows as DevTools will be available on the DDS port. This code is currently
commented out and will be enabled in a future PR.
There's two remaining circumstances that will prevent us from removing
DevtoolsRunner completely:
- The daemon's `devtools.serve` endpoint
- `flutter drive`'s `--profile-memory` flag used for recording memory
profiles
This PR also includes some refactoring around `DebuggingOptions` to
reduce the number of debugging related arguments being passed as
parameters adjacent to a `DebuggingOptions` instance.
This will make
* `flutter run` have source maps enabled by default
* `flutter build` have source maps disabled by default
which mirrors what happens already today with the js compilers.
For local development this works quite well - even better than with
dart2js (see dart2js issues in [0]).
We do have some follow-up items for source maps in dart2wasm compiler,
see [1]
[0]
[flutter/flutter/issues/151641](https://github.com/flutter/flutter/issues/151641)
[1]
[dart-lang/sdk/issues/56232](https://github.com/dart-lang/sdk/issues/56232)
Stop trying to roll the dependencies for native assets automatically.
They are under active development (pre 1.0.0), and have frequent breaking changes. I'll be rolling these manually.
(We have the same setup in the Dart SDK, not auto-rolling dart-lang/native via DEPS.)
Related comment:
* https://github.com/flutter/flutter/pull/151539#issuecomment-2221134824
Stop running link hooks in debug mode.
Rationale: link hooks only get access to tree-shaking info in release builds, so they can't do anything meaningful in debug builds. Debug builds should be fast as development cycle, so running less is better.
More details:
* https://github.com/dart-lang/native/issues/1252
Also: rolls packages to latest versions.
## Implementation details
The decision whether linking is enabled is made as follows:
* For normal builds `build_info.dart::BuildMode` is used to determine whether Dart is compiled in JIT or AOT mode.
* Testers always run in JIT, so no linking.
* Native asset dry runs only run for JIT builds (e.g only when hot reload and hot restart are enabled).
## Testing
The integration test is updated to output an asset for linking if `BuildConfig.linkingEnabled` is true, and to output an asset for bundling directly if linking is not enabled.
These changes allow to override existing native endorsed (federated & inline) plugins with e.g. non-endorsed plugin implementations via direct dependencies as described in the section *Platform implementation selection* in the [design doc](https://docs.google.com/document/d/1LD7QjmzJZLCopUrFAAE98wOUQpjmguyGTN2wd_89Srs).
```yaml
# pubspec.yaml
name: example
dependencies:
my_plugin: ^0.0.1
my_plugin_android_alternative: ^0.0.1
```
Closes#80374, closes#59657
Related: Override Dart plugins (#87991, #79669)
The current URL has 2 /to, but it should only be 1. However, this url redirects to https://docs.flutter.dev/release/breaking-changes/android-java-gradle-migration-guide, so I can change it to that if it's preferred just to go to the final destination - just don't know if it's a redirect for a reason
i didn't create an issue as it's a doc change, but if that's a requirement I can do that
This PR changes the project name logic for `flutter create` to look for the name in the pubspec.yaml `name` field,
before falling back to the directory name.
Fixes https://github.com/flutter/flutter/issues/53106
*If you had to change anything in the [flutter/tests] repo, include a link to the migration guide as per the [breaking change policy].*
https://github.com/flutter/flutter/pull/150645 tries to shut down Chrome by sending a browser close command through a debug protocol. The webkit_inspection_protocol library used to send the command may throw a SocketException if Chrome has already been shut down.
This changes the mapping for the Dart SDK inside Flutter from `org-dartlang-sdk:///third_party/dart/sdk` to org-dartlang-sdk:///flutter/third_party/dart/sdk`. This URI changed in https://github.com/flutter/engine/pull/51917 but was not caught by tests because they only tested a specific set of mappings and there were no integration tests checking what URIs were actually produced by a running app.
So, this change also adds an integration tests that ensures that a real running app produces URIs that are then correctly mapped.
Fixes https://github.com/Dart-Code/Dart-Code/issues/5164.
Reverts: flutter/flutter#150969
Initiated by: goderbauer
Reason for reverting: Failing test in https://logs.chromium.org/logs/flutter/buildbucket/cr-buildbucket/8743574743030691569/+/u/run_android_obfuscate_test/stdout
Original PR Author: gmackall
Reviewed By: {christopherfujino, reidbaker}
This change reverts the following previous change:
After the land of https://github.com/flutter/engine/pull/53592, there is some log spam:
```
e: /Users/mackall/.gradle/caches/transforms-3/c1e137371ec1afe9bc9bd7b05823752d/transformed/fragment-1.7.1/jars/classes.jar!/META-INF/fragment_release.kotlin_module: Module was compiled with an incompatible version of Kotlin. The binary version of its metadata is 1.8.0, expected version is 1.6.0.
e: /Users/mackall/.gradle/caches/transforms-3/d86c7cb1c556fe1655fa56db671c649c/transformed/jetified-activity-1.8.1/jars/classes.jar!/META-INF/activity_release.kotlin_module: Module was compiled with an incompatible version of Kotlin. The binary version of its metadata is 1.8.0, expected version is 1.6.0.
...
```
I think this is harmless, but still annoying. Upgrading the AGP version fixes it. To be honest, I don't know why (I expected the Kotlin version would do it). But after https://github.com/flutter/flutter/pull/146307, our tests have been running on AGP/Gradle 8.1/8.3 for a while, so it makes sense to upgrade anyways.
In a follow up PR:
1. Also upgrade the tests that were left behind in https://github.com/flutter/flutter/pull/146307, as I think removal of discontinued plugins paved the way here.
After the land of https://github.com/flutter/engine/pull/53592, there is some log spam:
```
e: /Users/mackall/.gradle/caches/transforms-3/c1e137371ec1afe9bc9bd7b05823752d/transformed/fragment-1.7.1/jars/classes.jar!/META-INF/fragment_release.kotlin_module: Module was compiled with an incompatible version of Kotlin. The binary version of its metadata is 1.8.0, expected version is 1.6.0.
e: /Users/mackall/.gradle/caches/transforms-3/d86c7cb1c556fe1655fa56db671c649c/transformed/jetified-activity-1.8.1/jars/classes.jar!/META-INF/activity_release.kotlin_module: Module was compiled with an incompatible version of Kotlin. The binary version of its metadata is 1.8.0, expected version is 1.6.0.
...
```
I think this is harmless, but still annoying. Upgrading the AGP version fixes it. To be honest, I don't know why (I expected the Kotlin version would do it). But after https://github.com/flutter/flutter/pull/146307, our tests have been running on AGP/Gradle 8.1/8.3 for a while, so it makes sense to upgrade anyways.
In a follow up PR:
1. Also upgrade the tests that were left behind in https://github.com/flutter/flutter/pull/146307, as I think removal of discontinued plugins paved the way here.
The Gradle Kotlin DSL also allows for type-safe application of the Flutter Gradle plugin, which is currently undetected by the CLI
```kotlin
plugins {
dev.flutter.`flutter-gradle-plugin`
}
```
Please note that the added test case isn't ideal, since the example gradle isn't actually valid kotlin DSL, however the `kotlin host app language with Gradle Kotlin DSL` is identical
Fixes#149859
If the user specifies the `--no-web-resources-cdn` or `--local-web-sdk`, we should use the local version of CanvasKit. `flutter.js` now has a flag that can be specified in the build configuration that tells it to load locally instead.
Also, added a link to the relevant docs in the web template warnings.
This addresses https://github.com/flutter/flutter/issues/148713
Also fixes https://github.com/flutter/flutter/issues/145559
This is a small refactor.
Using `loadPackageConfigWithLogging` seems wrong. We should be able to
assume that the flutter_tool package_config is in place.
Warm-up for https://github.com/flutter/flutter/issues/150196
Work towards https://github.com/flutter/flutter/issues/132712.
After this PR, after a completed `flutter build apk` command, we:
- Emit a `manifest-impeller-disabled` command if `io.flutter.embedding.android.EnableImpeller` is `'false'`.
- Emit a `manifest-impeller-disabled` command if `io.flutter.embedding.android.EnableImpeller` is _missing_.
- Emit a `manifest-impeller-enabled` command if `io.flutter.embedding.android.EnableImpeller` is `'true'`.
We will need to change the default (see `_impellerEnabledByDefault` in `project.dart`) before releasing, otherwise we will misreport `manifest-impeller-disabled` at a much higher rate than actual. If there is a way to instead compute the default instead of hard-coding, that would have been good.
See <https://docs.flutter.dev/perf/impeller#android> for details on the key-value pair.
---
I also did a tad of TLC, by removing the (now-defunct) `Usage` events for `flutter build ios`, so they are consistent.
/cc @zanderso, @chinmaygarde, @jonahwilliams
The previous approach of killing the Chromium parent process sometimes caused leaks of child processes on Windows. The Browser.close command in the debug protocol will tell Chromium to shut down all of its processes.
Two issues I noticed when I hit the issue at https://github.com/flutter/flutter/issues/149258
1. When the an app.stop event arrives from Flutter with an error, DAP does not pass that error back to the client so there's no visibility of the error.
2. If app.stop occurs but no app.start ever did, we leave the progress notification hanging around
This fixes both by handling app.stop and closing any open launch progress as well as passing any error to the client.
Fixes https://github.com/Dart-Code/Dart-Code/issues/5124
Even though this does not fix the below issue lets land this anyway as not logging to stderr when clearing logs makes sense to me.
related https://github.com/flutter/flutter/issues/150093
The test added is bad. It does not verify the behavior changed.
To verify the behavior changed correctly I would need to modify the generic device class to have clearLogs be an async function like many of the other calls. That would mean modifying every other device type and their implementations and their tests. Then I would need to update android_device to expose its logger. That is more than I have time for to validate a 2% flake error.
Feel free to disagree in the comments on this pr.
This adds the 'fail-fast' argument to flutter test, since dart test already supports this feature. Tests can now be stopped after first failure.
Fixes#124406
Fixes https://github.com/flutter/flutter/issues/145158.
In an ideal world, the `--machine` flag would be strictly a global flag which sub-commands can choose to use (or perhaps just to report a `toolExit` that they don't have a `--machine` supported-mode if not. However currently, there is both a global flag, and command-specific flags.
This leads to the confusing scenario where:
```sh
flutter devices --machine
```
... still checks for a Flutter update, printing a banner and breaking the JSON output.
This PR "fixes" that by allowing `--machine` _anywhere_ in the command-line arguments to suppress the check.
/cc @johnmccutchan.
Currently, the error message displayed to regenerate the lockfiles gives a Unix-like command ./gradlew, which will be incorrect for Windows environments. This PR uses globals.platform.isWindows to give the appropriate command.
closes#136763
- When `--web-renderer` is omitted, keep the value `null` until it later materializes to either `canvaskit` or `skwasm`.
- No more hardcoded defaults anywhere. We use `WebRendererMode.defaultForJs/defaultForWasm` instead.
- When in `--wasm` mode, the JS fallback is now `canvaskit` instead of `auto`.
- Add test for defaulting to `skwasm` when `--wasm` is enabled.
Fixes https://github.com/flutter/flutter/issues/149826
If the service is disposed when we make calls on it, we need to bubble up an exception that is understood by the callers.
This is another speculative fix for https://github.com/flutter/flutter/issues/149238.
Unpin `archive` version and run `flutter update-packages --force-upgrade` to show there are no additional dependencies pulled in (url_launcher was updated, but unrelated).
Fixes https://github.com/flutter/flutter/issues/149427
In service of #146879. In summary, a bunch of `printTrace` calls were added to try to troubleshoot this flake, but the flake has stopped happening.
This PR more-or-less reverts https://github.com/flutter/flutter/pull/148596, since this was the only change that could have _theoretically_ (somehow maybe) stopped the flake. I'll consider reverting the rest of the `printTrace` calls if the flake does not reappear after this PR lands.
Fixes https://github.com/flutter/flutter/issues/149386. Fixes https://github.com/flutter/flutter/issues/106150.
The stocks test app includes Dart files containing localized messages generated by `package:flutter_localizations`. However, these files appear to have become out of date. Running `pub get` in the project will regenerate these files and generate a diff, which can be annoying when working on the repo.
This PR generates the files. ~~It also updates the templates for these files to be compliant with flutter/flutter repo lint rules, including `noop_primitive_operations` and `use_super_parameters`.~~ It also adds `// ignore_for_file: type=lint` to these files to disable linting for these files. This avoids issues like https://github.com/flutter/flutter/issues/106150 and [this](https://github.com/flutter/flutter/pull/148741#issuecomment-2141161753).
Dart2wasm only needs a platform file, which contains the compiled
`dart:*` libraries. There's no need to specify a seperate `--dart-sdk`
option (anymore).
(See also https://dart-review.googlesource.com/c/sdk/+/366821)
This PR modifies the `flutter create --empty` command to not delete the `test/` folder when run on an existing app project.
Before:
```bash
flutter create my_app --empty
mkdir my_app/test
if test -d my_app/test; then echo "test exists"; else echo "test does not exist"; fi # test exists
flutter create my_app --empty
if test -d my_app/test; then echo "test exists"; else echo "test does not exist"; fi # test does not exist
```
After:
```bash
flutter create my_app --empty
mkdir my_app/test
if test -d my_app/test; then echo "test exists"; else echo "test does not exist"; fi # test exists
flutter create my_app --empty
if test -d my_app/test; then echo "test exists"; else echo "test does not exist"; fi # test exists
```
Fixes https://github.com/flutter/flutter/issues/134928
This PR adds support invoking `link.dart` hooks.
Link hooks can add new assets. Link hooks can transform assets sent to link hook from build hooks.
This PR does not yet add support for getting tree-shake information in the link hooks. This is pending on defining the `resources.json` format (https://github.com/dart-lang/sdk/issues/55494).
Issue:
* https://github.com/flutter/flutter/issues/146263
## Implementation considerations
The build hooks could be run before Dart compilation and the link hooks after Dart compilation. (This is how it's done in Dart standalone.) However, due to the way the `Target`s are set up, this would require two targets and serializing and deserializing the `BuildResult` in between these. This would lead to more code but no benefits. Currently there is nothing that mandates running build hooks before Dart compilation.
## Testing
* The unit tests verify that the native_assets_builder `link` and `linkDryRun` would be invoked with help of the existing fake.
* The native assets integration test now also invokes an FFI call of a package that adds the asset during the link hook instead of the build hook.
* In order to keep coverage of the `flutter create --template=package_ffi`, `flutter create` is still run and the extra dependency is added and an extra ffi call is added. (Open to alternative suggestions.)
This PR adds a new flag `default-flavor` in the `flutter` section of `pubspec.yaml`. It allows developers of multi-flavor android apps to specify a default flavor to be used for `flutter run`, `flutter build` etc.
Using `flutter run` on flavored apps already works without specifying `--flavor` already works on iOS (it defaults to the `runner` schema), so I (and others in #22856) figured this would be nice to have.
fixes#22856
Fixes https://github.com/flutter/flutter/issues/148354
Fixes https://github.com/flutter/flutter/issues/147142
Closes https://github.com/flutter/flutter/pull/147144
## Pre-launch Checklist
- [x] I read the [Contributor Guide] and followed the process outlined
there for submitting PRs.
- [x] I read the [Tree Hygiene] wiki page, which explains my
responsibilities.
- [x] I read and followed the [Flutter Style Guide], including [Features
we expect every widget to implement].
- [x] I signed the [CLA].
- [x] I listed at least one issue that this PR fixes in the description
above.
- [x] I updated/added relevant documentation (doc comments with `///`).
- [x] I added new tests to check the change I am making, or this PR is
[test-exempt].
- [x] I followed the [breaking change policy] and added [Data Driven
Fixes] where supported.
- [x] All existing and new tests are passing.
This PR is step 12 in the journey to solve issue #136139 and make the
entire Flutter repo more readable.
Most of it involves implementing switch expressions, and there's also a
few other random things that I wanted to clean up a bit.
DDS was temporarily pinned to 4.1.0 because 4.2.0 triggered some test
failures (see https://github.com/flutter/flutter/pull/147250). Those
failures should be fixed by vm_service 14.2.2, so this unpins DDS and
rolls both of these packages (along with devtools_shared, which is a DDS
dependency).
(If the bot updates vm_service before this is done, I can rebase over
that will reduce the size of this PR to just a few files)
### Some background:
* Flutter registers the hot-restart service extension for all devices **except** web devices. For web devices, DWDS registers the service extensions.
* When a user is debugging their code via VS Code, the VS Code Dart extension [sends a hot-restart](94cb81c552/src/debug/run_daemon_base.ts (L100)) request via stdin/out to flutter_tools
* flutter_tools then [calls the "hotRestart" service extension](f3978c7a46/packages/flutter_tools/lib/src/isolated/resident_web_runner.dart (L447)) (which, again, has been registered by DWDS)
### Why is this change necessary?
In DWDS, we are changing how we register the "hotRestart" service extension (here is the PR for that: https://github.com/dart-lang/webdev/pull/2388). Previously, we registered the "hotRestart" service extension on a client that was directly connected to the VmService. With https://github.com/dart-lang/webdev/pull/2388, we will be registering the "hotRestart" service extension on a client that is connected to DDS.
When a service extension is registered against DDS, DDS adds a prefix to the service extension method name (e.g. "hotRestart" becomes "s0.hotRestart"). It informs clients of the service extension name via `kServiceRegistered` events sent to the `Service` stream.
Therefore, this change simply listens to those service extension registered events, and uses them to determine the "hotRestart" service extension's method name.
Part of #137040 and #80374
- Extracted getting plugin implementation candidates and the default implementation to its own method
- Extracted resolving the plugin implementation to its own method
- Simplify candidate selection algorithm
- Support overriding inline dart implementation for an app-facing plugin
- Throw error, if a federated plugin implements an app-facing plugin, but also references a default implementation
- Throw error, if a plugin provides an inline implementation, but also references a default implementation
Fix dependency tracking in the build.
* https://github.com/flutter/flutter/issues/147643
## Testing
It's not clear to me where to test the caching behavior of specific targets.
* test/general.shard/build_system/targets/common_test.dart
* This doesn't test the caching behavior of any targets
* test/general.shard/build_system/build_system_test.dart
* test/general.shard/cache_test.dart
* Both of these don't test specific `Target`s, these have `TestTarget`s.
* test/integration.shard/isolated/native_assets_test.dart
* This could work, but it's an integration test that already takes long to run.
This pull request aims for improved readability, based on issue #146600.
```dart
// before
List<SupportedPlatform> getSupportedPlatforms({bool includeRoot = false}) {
final List<SupportedPlatform> platforms = includeRoot
? <SupportedPlatform>[SupportedPlatform.root]
: <SupportedPlatform>[];
if (android.existsSync()) {
platforms.add(SupportedPlatform.android);
}
if (ios.exists) {
platforms.add(SupportedPlatform.ios);
}
if (web.existsSync()) {
platforms.add(SupportedPlatform.web);
}
if (macos.existsSync()) {
platforms.add(SupportedPlatform.macos);
}
if (linux.existsSync()) {
platforms.add(SupportedPlatform.linux);
}
if (windows.existsSync()) {
platforms.add(SupportedPlatform.windows);
}
if (fuchsia.existsSync()) {
platforms.add(SupportedPlatform.fuchsia);
}
return platforms;
}
// after
List<SupportedPlatform> getSupportedPlatforms({bool includeRoot = false}) {
return <SupportedPlatform>[
if (includeRoot) SupportedPlatform.root,
if (android.existsSync()) SupportedPlatform.android,
if (ios.exists) SupportedPlatform.ios,
if (web.existsSync()) SupportedPlatform.web,
if (macos.existsSync()) SupportedPlatform.macos,
if (linux.existsSync()) SupportedPlatform.linux,
if (windows.existsSync()) SupportedPlatform.windows,
if (fuchsia.existsSync()) SupportedPlatform.fuchsia,
];
}
```
In service of https://github.com/flutter/flutter/issues/146879 and https://github.com/flutter/flutter/issues/145812. In these issues, we see what appears to be the flutter tool getting stuck somewhere during hot reload. It may help if we knew were exactly where we are getting stuck (preparing assets, writing them to device, etc.).
This PR adds a new parameter to `FlutterTestDriver::run`, `verbose`. When verbose is set, `FlutterTestDriver` will run `flutter` with `--verbose` in its tests. Keep in mind that `FlutterTestDriver` only prints logs from `flutter` when a test fails, so this shouldn't spam the logs of passing tests.
This PR sets the parameter when invoking the flaky tests from https://github.com/flutter/flutter/issues/146879 and #145812, so we should see more detailed logs in future flakes.
While this is a low risk PR, you can verify the change by intentionally breaking hot reload code, clearing the cached tool binaries, and then running either of these tests.
## Description
This introduces a list of packages that we will explicitly not pin. It is to be used for things where the package isn't actually published, but is a transitive dependency of another package included in the SDK. This happens with the `macros` package, for instance, which depends on the private, unpublished, `_macros` package where the SDK does some tricky things to depend on it (it depends on "any", but ships it as part of the SDK).
Also ran `flutter update-packages --force-update` to update all of the pubspec files.
## Related Issues
- Fixes#147656
## Tests
- Added a test that makes sure that explicitly unpinned packages don't show up in the pinned list.
- added language for all code blocks
- replaced `bash` or `shell` with `sh` for consistency.
- added `sh` and `console` in the GitHub template link generator.
- updated test for GitHub template.
Uses kernel concatenation to concatenate the `program.dill` and `native_assets.dill`.
This will enable compiling `native_assets.dill` after `program.dill` so that tree-shaking information can be used.
Issue:
* https://github.com/flutter/flutter/issues/146270
## Implementation choices
* We can either run the frontend_server twice as a process invocation (current PR), or keep the process running after the dart compilation and then shut it down after native assets compilation to kernel. Keeping the process running would require more coordination between different `Target`s.
* We can either chose to only do separate kernel file compilation in release (AOT) builds, or do it both for JIT and AOT (current PR). Doing it differently in JIT/AOT means more conditionals in lib/src/build_system/targets/, doing it single-shot in JIT might be more performant though.
Update: Both of these are mitigated by not running the kernel compiler process if there are no native assets at all.
## Implementation notes
This only updates `flutter assemble`.
The kernel compilation calls in `flutter run` do not require kernel concatenation. The native-assets-kernel-mapping in `flutter run` contains results from `--dry-run` invocations of `hook/build.dart` (and in the future `hook/link.dart`). These `--dry-run` invocations do not get access to tree-shaking information, so the `link` hook to be introduced later can be run before Dart compilation.
## Tests
* Updated the `Target`s tests to reflect the new implementation.
## Related CLs
* frontend_server support: https://dart-review.googlesource.com/c/sdk/+/363567
* Same PR for Dart standalone: https://dart-review.googlesource.com/c/sdk/+/362384
When Swift Package Manager feature is enabled, create app and create plugin will have Swift Package Manager integration already added and will not need to undergo a migration.
Fixes https://github.com/flutter/flutter/issues/146371.
```
flutter config --enable-swift-package-manager
flutter create --ios-language swift --platforms ios,macos swift_app_name
flutter create --ios-language objc --platforms ios objc_app_name
flutter create --template=plugin --ios-language swift --platforms ios,macos swift_plugin_name
flutter create --template=plugin --ios-language objc --platforms ios objc_plugin_name
```
Here's another PR with a couple of typos fixed. As you can see there was a typo in _fileReferenceI**n**dentifiers_, in class _ParsedProjectInfo._ Maybe we should do some check on that since I'm not sure if that property is used somewhere outside Flutter?
From https://github.com/flutter/flutter/issues/143348#issuecomment-2016047148:
> before we ship, we should add a printTrace to the tool about each asset transformer we're invoking and the path/arguments we called it with
I think this is a good idea since asset transformers can be arbitrary Dart programsâmeaning that a lot can go wrong when running them. For example, they can hang indefinitely or perform some sort of I/O that later results in a tool crash. Knowing that asset transformation was involved when debugging a crash (or a slow/stuck `flutter build`) could be useful, so I think adding a `printTrace` or two is a good idea (or at least not a bad one).
This PR will opt out users from legacy analytics if they have already been opted out from package:unified_analytics.
After successfully merging into main, this will be CP'd into beta and stable channels
This PR adds initial support for Swift Package Manager (SPM). Users must opt in. Only compatible with Xcode 15+.
Fixes https://github.com/flutter/flutter/issues/146369.
## Included Features
This PR includes the following features:
* Enabling SPM via config
`flutter config --enable-swift-package-manager`
* Disabling SPM via config (will disable for all projects)
`flutter config --no-enable-swift-package-manager`
* Disabling SPM via pubspec.yaml (will disable for the specific project)
```
flutter:
disable-swift-package-manager: true
```
* Migrating existing apps to add SPM integration if using a Flutter plugin with a Package.swift
* Generates a Swift Package (named `FlutterGeneratedPluginSwiftPackage`) that handles Flutter SPM-compatible plugin dependencies. Generated package is added to the Xcode project.
* Error parsing of common errors that may occur due to using CocoaPods and Swift Package Manager together
* Tool will print warnings when using all Swift Package plugins and encourage you to remove CocoaPods
This PR also converts `integration_test` and `integration_test_macos` plugins to be both Swift Packages and CocoaPod Pods.
## How it Works
The Flutter CLI will generate a Swift Package called `FlutterGeneratedPluginSwiftPackage`, which will have local dependencies on all Swift Package compatible Flutter plugins.
The `FlutterGeneratedPluginSwiftPackage` package will be added to the Xcode project via altering of the `project.pbxproj`.
In addition, a "Pre-action" script will be added via altering of the `Runner.xcscheme`. This script will invoke the flutter tool to copy the Flutter/FlutterMacOS framework to the `BUILT_PRODUCTS_DIR` directory before the build starts. This is needed because plugins need to be linked to the Flutter framework and fortunately Swift Package Manager automatically uses `BUILT_PRODUCTS_DIR` as a framework search path.
CocoaPods will continue to run and be used to support non-Swift Package compatible Flutter plugins.
## Not Included Features
It does not include the following (will be added in future PRs):
* Create plugin template
* Create app template
* Add-to-App integration
Changing the web renderer doesn't directly modify the environment's dart defines, and so doesn't do a full build invalidation. We need to include the web renderer in the build key for the compiler configuration. This information is used directly by the web targets to modify the dart defines that are passed into the compiler, so we need to rebuild if this information changes.
When performing artifact lookups for `Artifact.genSnapshot` for macOS desktop builds, a `TargetPlatform` is used to determine the name of the tool, typically `gen_snapshot_$TARGET_ARCH`. Formerly, this tool was always named `gen_snapshot`.
The astute reader may ask "but Chris, didn't we support TWO target architectures on iOS and therefore need TWO `gen_snapshot` binaries?" Yes, we did support both armv7 and arm64 target architectures on iOS. But no, we didn't initially have two `gen_snapshot` binaries. We did *build* two `gen_snapshots`:
* A 32-bit x86 binary that emitted armv7 AOT code
* A 64-bit x64 binary that emitted arm64 AOT code
At the time, the bitness of the `gen_snapshot` tool needed to match the bitness of the target architecture, and to avoid having to do a lot of work plumbing through suffixed `gen_snapshot` names, the author of that work (who, as evidenced by this patch, is still paying for his code crimes) elected to "cleverly" lipo the two together into a single multi-architecture macOS binary still named `gen_snapshot`. See: https://github.com/flutter/engine/pull/4948
This was later remediated over the course of several patches, including:
* https://github.com/flutter/engine/pull/10430
* https://github.com/flutter/engine/pull/22818
* https://github.com/flutter/flutter/pull/37445
However, there were still cases (notably `--local-engine` workflows in the tool) where we weren't computing the target platform and thus referenced the generic `gen_snapshot` tool.
See: https://github.com/flutter/flutter/issues/38933
Fixed in: https://github.com/flutter/engine/pull/28345
The test removed in this PR, which ensured that null `SnapshotType.platform` was supported was introduced in https://github.com/flutter/flutter/pull/11924 as a followup to https://github.com/flutter/flutter/pull/11820 when the snapshotting logic was originally extracted to the `GenSnapshot` class, and most invocations still passed a null target platform.
Since there are no longer any cases where `TargetPlatform` isn't passed when looking up `Artifact.genSnapshot`, we can safely make the platform non-nullable and remove the test.
This is pre-factoring towards the removal of the generic `gen_snapshot` artifact from the macOS host binaries (which are currently unused since we never pass a null `TargetPlatform`), which is pre-factoring towards the goal of building `gen_snapshot` binaries with an arm64 host architecture, and eliminate the need to use Rosetta during iOS and macOS Flutter builds.
Part of: https://github.com/flutter/flutter/issues/101138
Umbrella issue: https://github.com/flutter/flutter/issues/103386
Umbrella issue: https://github.com/flutter/flutter/issues/69157
No new tests since the behaviour is enforced by the compiler.
This tweaks the Flutter doctor messages for CocoaPods.
This also switches the "unknown version" error to link to the update instructions instead of the installation instructions; the user has already installed CocoaPods in this scenario.
Example error before:
```
â CocoaPods not installed.
CocoaPods is used to retrieve the iOS and macOS platform side's plugin code that responds to your plugin usage on the Dart side.
Without CocoaPods, plugins will not work on iOS or macOS.
For more info, see https://flutter.dev/platform-plugins
To install see https://guides.cocoapods.org/using/getting-started.html#installation for instructions.
```
Example error after:
```
â CocoaPods not installed.
CocoaPods is a package manager for iOS or macOS platform code.
Without CocoaPods, plugins will not work on iOS or macOS.
For more info, see https://flutter.dev/platform-plugins
For installation instructions, see https://guides.cocoapods.org/using/getting-started.html#installation
```
In a ProxiedDevicePortForwarder, there might be a race condition where the local socket has been disconnected, but the remote end was still sending new data. In this case, avoid forwarding new data to the socket.
This adds support for adding the `--wasm` flag to `flutter run` and `flutter drive`
* Emits errors if you attempt to use the skwasm renderer without the `--wasm` flag
* Emits errors if you try to use `--wasm` when not using a web device
* Uses the skwasm renderer by default if you pass `--wasm` and no `--web-renderer`
Convert `ProjectMigration.run()` and `ProjectMigrator.migrate()` to be async.
Needed for Swift Package Manager migration, which requires some async processes: https://github.com/flutter/flutter/pull/146256
Also move the vm service discovery logic into platform-specific implementation of `Device`s. This is to avoid having platform-specific code in attach.dart.
Xcode 15 will be required for iOS App Store submission
> Please note that as of April 2024 all iOS and iPadOS apps submitted to the App Store must be built with a minimum of Xcode 15 and the iOS 17 SDK.
https://developer.apple.com/ios/submit/
And will also be required for Swift Package Manager support https://github.com/flutter/flutter/pull/146256.
We could swap to "required" but macOS developers don't technically need to upgrade. We can let the Store itself enforce its policies. And we can swap to Xcode 15 "required" when SPM adoption is further along.
Part of https://github.com/flutter/flutter/issues/144582
Part of #137040 and #80374
`possibleResolutions` and `possibleResolutions` used a key containing the platform and the plugin name via `getResolutionKey`. The global list is not necessary and also confusing. It can be avoided by handling the resolution of plugins per platform. This helps to have a more self-contained plugin resolution. This also is necessary to reuse the plugin resolution logic (per platform) for #137040.
This PR is to update material_color_utilities package version to the latest. `material_color_utilities/scheme/scheme_fruit_salad.dart` and `material_color_utilities/scheme/scheme_rainbow.dart` are exported after version 0.9.0.
Once this PR is merged, we don't need to explicitly import these two files like the change in PR #144805, which breaks some dependencies in `Google testing`.
Fixes https://github.com/flutter/flutter/issues/144454
For reference, after formatting it looks like:
```
ââ Flutter Fix âââââââââââââââââââââââââââââââââââââââââââââââââââââââââââââââââââââââââââââââââââââââââââ
â [!] Your project requires a newer version of the Kotlin Gradle plugin. â
â Find the latest version on https://kotlinlang.org/docs/releases.html#release-details, then update the â
â version number of the plugin with id "org.jetbrains.kotlin.android" in the plugins block of â
â /Users/mackall/development/BugTesting/ELIMINATEME/blah/android/settings.gradle. â
â â
â Alternatively (if your project was created before Flutter 3.19), update â
â /Users/mackall/development/BugTesting/ELIMINATEME/blah/android/build.gradle â
â ext.kotlin_version = '<latest-version>' â
ââââââââââââââââââââââââââââââââââââââââââââââââââââââââââââââââââââââââââââââââââââââââââââââââââââââââââ
```
In service of https://github.com/flutter/flutter/issues/143348
This PR makes hot reloads reflect changes to transformer configurations under the `assets` section in pubspec.yaml.
This PR is optimized for ease of implementation, and should not be merged as-is. If it is merged as-is, seriously consider creating a tech debt issue and assigning it to someone to make sure it gets resolved.
There are two issues in the previous implementation:
1. `_populateDevices` will return the devices from `deviceNotifier` if it had been initialized, assuming that once it's initialized, it has been properly populated. That assumption is not true because calling getters like `onAdded` would initialize `deviceNotifier` without populating it.
2. `deviceNotifier` instance might be replaced in some cases, causing `onAdded` subscribers to lose any future updates.
To fix (1), this commit added the `isPopulated` field in `deviceNotifier` as a more accurate flag to determine if we need to populate it.
To fix (2), this commit made `deviceNotifier` a final member in `PolingDeviceDiscovery`.
* Adds support for `flutter test --wasm`.
* The test compilation flow is a bit different now, so that it supports compilers other than DDC. Specifically, when we run a set of unit tests, we generate a "switchboard" main function that imports each unit test and runs the main function for a specific one based off of a value set by the JS bootstrapping code. This way, there is one compile step and the same compile output is invoked for each unit test file.
* Also, removes all references to `dart:html` from flutter/flutter.
* Adds CI steps for running the framework unit tests with dart2wasm+skwasm
* These steps are marked as `bringup: true`, so we don't know what kind of failures they will result in. Any failures they have will not block the tree at all yet while we're still in `bringup: true`. Once this PR is merged, I plan on looking at any failures and either fixing them or disabling them so we can get these CI steps running on presubmit.
This fixes https://github.com/flutter/flutter/issues/126692
Pre work for https://github.com/flutter/engine/pull/51229. Removes a lot of code referencing v1 of the android embedding, though not necessarily all of it (I may have missed some, it is hard to know).
Will hopefully make landing that PR less painful (or maybe painless?)
A ProxiedDevice has two port forwarder, one to forward remote host port and another to forward remote device port. During an attach workflow, the port was forwarded by the latter. Update ProxiedDDS to make sure that it works in such case.
This pull request is part of the effort to solve issue #144903.
In the past, my efforts to reduce line length involved refactoring away from switch statements, but unlike [yesterday's PR](https://github.com/flutter/flutter/pull/144905), this one is full of switch statements that make things more concise!
This makes several changes to flutter web app bootstrapping.
* The build now produces a `flutter_bootstrap.js` file.
* By default, this file does the basic streamlined startup of a flutter app with the service worker settings and no user configuration.
* The user can also put a `flutter_bootstrap.js` file in the `web` subdirectory in the project directory which can have whatever custom bootstrapping logic they'd like to write instead. This file is also templated, and can use any of the tokens that can be used with the `index.html` (with the exception of `{{flutter_bootstrap_js}}`, see below).
* Introduced a few new templating tokens for `index.html`:
* `{{flutter_js}}` => inlines the entirety of `flutter.js`
* `{{flutter_service_worker_version}}` => replaced directly by the service worker version. This can be used instead of the script that sets the `serviceWorkerVersion` local variable that we used to have by default.
* `{{flutter_bootstrap_js}}` => inlines the entirety of `flutter_bootstrap.js` (this token obviously doesn't apply to `flutter_bootstrap.js` itself).
* Changed `IndexHtml` to be called `WebTemplate` instead, since it is used for more than just the index.html now.
* We now emit warnings at build time for certain deprecated flows:
* Warn on the old service worker version pattern (i.e.`(const|var) serviceWorkerVersion = null`) and recommends using `{{flutter_service_worker_version}}` token instead
* Warn on use of `FlutterLoader.loadEntrypoint` and recommend using `FlutterLoader.load` instead
* Warn on manual loading of `flutter_service_worker.js`.
* The default `index.html` on `flutter create` now uses an async script tag with `flutter_bootstrap.js`.
Reverts: flutter/flutter#144752
Initiated by: andrewkolos
Reason for reverting: compilation issue has turned the tree red
Original PR Author: andrewkolos
Reviewed By: {christopherfujino}
This change reverts the following previous change:
In service of https://github.com/flutter/flutter/issues/143348
When invoking a package to transform an asset, we set `FLUTTER_BUILD_MODE` to the CLI name of the build mode being used. Inspired by https://github.com/flutter/flutter/issues/101077#issuecomment-1890379501:
> Do transformers know whether they get executed in debug or release mode? I kinda imagine that being useful. Ex: There's a transformer that optimizes the file size of images. Depending on the amount and size of the images, that could take a significant amount of time. Therefore, I might want to only execute it in release builds.
Note for the reviewer: the interesting part of this change can be found in the commit [set environment variable to build mode when running asset transformer…](579912d470). The rest of the change is updating call sites with a new argument.
In service of https://github.com/flutter/flutter/issues/143348
When invoking a package to transform an asset, we set `FLUTTER_BUILD_MODE` to the CLI name of the build mode being used. Inspired by https://github.com/flutter/flutter/issues/101077#issuecomment-1890379501:
> Do transformers know whether they get executed in debug or release mode? I kinda imagine that being useful. Ex: There's a transformer that optimizes the file size of images. Depending on the amount and size of the images, that could take a significant amount of time. Therefore, I might want to only execute it in release builds.
Note for the reviewer: the interesting part of this change can be found in the commit [set environment variable to build mode when running asset transformerâ¦](579912d470). The rest of the change is updating call sites with a new argument.
This fixes https://github.com/flutter/flutter/issues/143703
We need to make sure that when source maps are enabled for the dart2js target, it advertises the sourcemap file as one of its outputs so that the web release bundle can copy it over.
The flutter engine & framework can opt out of this optimization for
individual classes / class hierarchies via:
* `@pragma(flutter:keep-to-string)`
* `@pragma(flutter:keep-to-string-in-subtypes)`
Or by using the convenience constant `@keepToString` from `dart:ui`.
=> This aligns the build process more with g3 (which already does this)
Closes https://github.com/flutter/flutter/issues/52759
Part of #137040 and #80374
The flag `throwOnPluginPubspecError` never actually was enabled during production in #79669, but only in some dart plugin tests. And in the tests the case of the error when enabling the flag was not explicitly tested. The only thing tested was, that it is not thrown when disabled.
As explained [here](https://github.com/flutter/flutter/pull/142035#discussion_r1484237904) the only case, where this error could be thrown is, when a dart implementation and a native inline implementation are provided simultaneously. But throwing an exception there is a wrong behavior, as both can coexist in a plugin package, thus in the pubspec file.
Disabling the flag means, that the error is not thrown and not shown to the user. This is the case in production (contrary to the dart plugin tests), which acts like these plugin cases of implementations are just skipped. And this is what actually should be done.
In conclusion, I think the case of coexisting dart and native implementation in pubspec was just overlooked and therefore this error validation was introduced, which is not necessary or even valid.
For more discussion, see: https://discord.com/channels/608014603317936148/608022056616853515/1200194937791205436
- This is tricky: I already added a test in #142035, which finally complies with the other tests, by removing the flag. So I think it falls in the category of "remove dead code".
- Theoretically this is a breaking change, as removing / altering some tests. But the flag actually was never valid or used, so IDK. But this may not does fall in the category of "contributed tests".
*Replace this paragraph with a description of what this PR is changing or adding, and why. Consider including before/after screenshots.*
*List which issues are fixed by this PR. You must list at least one issue. An issue is not required if the PR fixes something trivial like a typo.*
*If you had to change anything in the [flutter/tests] repo, include a link to the migration guide as per the [breaking change policy].*
`DevFSBytesContent` (and it's descendant `DevFSStringContent`) have setters that change the underlying content. These are unused outside of tests, so this PR removes them. Amongst other things, this could help me refactor https://github.com/flutter/flutter/pull/144660 into something that has fewer pitfalls.
This is purely a refactoring.
We should always target the newest, and 34 is the newest. This isn't a requirement yet (like it is for 33+) but presumably it will be made required in the nearish future.
Partial implementation of https://github.com/flutter/flutter/issues/143348
This enables asset transformation during hot reload (except for web, because that has its own implementation of `DevFS` ð). Asset transformers will be reapplied after changing any asset and performing a hot reload during `flutter run`.
