30+4 has made some changes to the remoting protocol that are incompatbile
with a test harness running 30+3 (e.g. adding a type='initial' field to the
first message)
* Use engine-built dart sdk
* Download dart-sdk from engine
* Move up deps to fix dart sdk constraint problem
* Update update_dart_sdk.ps1 for Windows
* Fix tests so they pass analysis
* More types for tests
* Roll engine
* Update dart sdk stamp location in flutter.bat
* Add newline
There were some problems I introduced with the last PR for this. It passed the test, but failed in practice.
This adds tests for those failure cases, adds a "--help" and fixes the test so that it doesn't try and actually download MinGit as part of the Windows test.
I added package:platform as a dependency, so I did a force upgrade on the packages.
Also, re-enabling 'create package' in the cache warming code, now that #14448 is fixed.
* Unpin package:test and upgrade packages
* Update packages/flutter/test/foundation/stack_trace_test.dart
* Also add packages/flutter_tools/test/data/asset_test/font/.dartignore to ensure that update-packages --force-upgrade does not crash.
Generate the "version" file from git tags.
Remove the old VERSION file and mentions of versions in pubspec.yaml files.
Replace the old update_versions.dart script with a new roll_dev.dart script.
Update "flutter channel".
Update "flutter upgrade", including making it transition from alpha to dev.
Update "flutter --version" and "flutter doctor".
* Revert "Reverting package changes until I can figure out how to fix Windows. (#14007)"
This reverts commit 6fda8ee821.
* Make prepare_package run on Windows
* Revert "Fixed output validation. (#14005)"
This reverts commit d84398db72.
* Revert "Update package prep script to do async process execution and emit output as it happens. (#13918)"
This reverts commit b7169c1d95.
- Switches to async process execution, which now shows output as it happens instead of in chunks when the process completes
- Now uses ProcessManager so that it may be mocked for the test.
- Adds in the download and install of mingit on Windows.
- Updated package dependencies because of added dependency on process package.
This simply updates the package dependencies by running flutter update-packages --force-upgrade.
I'm doing this with no other changes, because the last time I tried that, redness occurred. I want to isolate the problem to a "clean" update of the packages.
It looks like the plugins device_info, connectivity, and url_launcher haven't yet had their gradle configurations updated, so they fail when trying to build with the new gradle. I did not upgrade for those three packages only (in flutter_gallery) until we are ready to fix them (fixing them for master will break them for alpha users, so we need to do an alpha roll to do that).
This adds our self-compiled copy of the MinGit executable (built from the flutter/git repo) to the archive when building an archive for Windows.
I also tweaked the internal API for prepare_package.dart so that there's a single entry point to build an archive.
* fix updrade script; upgrade to the latest package versions
* exclude special dependencies from transitive closure
* fix stack trace handling in flutter_test due to stack_trace change
* change type on _emptyStackTrace
* Pin all dependencies ONCE AND FOR ALL
This replaces the secret `flutter update-packages --upgrade` with a destructive `flutter update-packages --force-upgrade` that actually goes and pins every dependency and transitive dependency in every flutter package to the same version.
* Add comments.
* Bump to test `0.12.20`.
Some test `0.12.20` highlights:
* introduces `expectLater()` that returns a `Future` that completes when the matcher has finished running
* deprecates the `verbose` parameter to `expect()` and the `formatFailure()` (to be removed in `0.13.0`)
Otherwise:
* to keep up w/ the deprecation of `verbose`, removes `widget_tester` API to pass `verbose` flag (alternatively we could suppress the warning for now)
* Update stack manipulation.
* Fix framecount.
* Remove the workaround that pinned args to v0.13.6
This reverts most of the changes in commit 6331b6c8b5
* throw exception if exit code is not an integer
* rework command infrastructure to throw ToolExit when non-zero exitCode
* convert commands to return Future<Null>
* cleanup remaining commands to use throwToolExit for non-zero exit code
* remove isUnusual exception message
* add type annotations for updated args package
Switch our pubspec.yamls to using SDK sources so that we can have consistent
source types when we depend on these packages from external packages using SDK
sources.
* Update tools to use `analyzer` from vended Dart SDK.
* updates `flutter_tools` and `flutter_test` to use the SDK-vended `analyzer` package
* tweaks dependency tracking logic to only record the SDK-vended `analyzer` so as not to crash on spurious conflicts (due to transitive dependencies)
* Review fixes.
* brings in analyzer version (`0.27.4-alpha.14`) corresponding to current Dart SDK (`1.18.0-dev.2.0`).
* updates analysis to use prefered API for embedder URI resolution
* adds trampolines to `State` and `StatelessWidget` to allow for warning-free within-library @protected access (needed since we closed off access to @protected closures from outside subclasses).
* turns off cache dependency tracking for analysis (in DDC this amounted to a 10% speed improvement).
Introduces a new Dart analysis wrapper that works directly with the analyzer API (in favor of shelling out to a separate process).
Some consequences:
* we no longer need to fear parts (simplifying our dart file gathering)
* we can filter by error code (when needed), rather than by error strings
* no more IO scraping
* no need to generate `main()` or to run with `--package-warnings`
* we now specify an analyzer (and linter) version in the pubspec (we’ll want to make sure this doesn’t diverge too far from the analyzer shipped with the SDK but it does give us some room to play with experimental builds)
* no more (re)scanning of error source files (and so no more source cache)
* should generally be a bit simpler and easier to maintain
* runs a bit faster :)
We use a number of non-public APIs in the test package, which makes our
dependency quite fragile. This patch pins a specific, known-good version. We
should update to the lastest version in a follow-up patch.