This PR was merged into the 3.4 branch.
Discussion
----------
[BrowserKit vs HttpFoundation][TestSuite]Run test in separate process
| Q | A
| ------------- | ---
| Branch? | 3.4
| Bug fix? | yes
| New feature? | no
| BC breaks? | no
| Deprecations? | no
| Tests pass? | yes
| Fixed tickets | n/a
| License | MIT
| Doc PR | n/a
This test calls code that defines some environment variables, which in
turn trigger the registration of a the deprecation error handler, which
causes unexpected issues when testing other components.
Commits
-------
6d51a04b11 Run test in separate process
This test calls code that defines some environment variables, which in
turn trigger the registration of a the deprecation error handler, which
causes unexpected issues when testing other components.
This PR was merged into the 3.4 branch.
Discussion
----------
[Debug vs DI][TestSuite] Use a class name that does not actually exist
| Q | A
| ------------- | ---
| Branch? | 3.4
| Bug fix? | yes
| New feature? | no
| BC breaks? | no
| Deprecations? | no
| Tests pass? | yes
| Fixed tickets | n/a
| License | MIT
| Doc PR | n/a
Using "Foo", a class name that corresponds to no less than 22 fixture
classes, results in the first found "Foo" being loaded when one is
found by the ClassNotFoundFatalErrorHandler error handler, I am not sure
exactly why, but it is not really a big issue because this is a fatal
error handler and execution is not supposed to continue after that.
Except that is very much the case when running the whole test suite
sequentially with ./phpunit . Then we arrive to the DI component test
suite, and a failure happens because \\foo was not supposed to be defined:
> Failed asserting that exception message 'The definition for "\foo" has
> no class attribute, and appears to reference a class or interface in the
> global namespace. Leaving out the "class" attribute is only allowed for
> namespaced classes. Please specify the class attribute explicitly to get
> rid of this error.' contains 'The definition for "\foo" has no class.'.
Commits
-------
a53e0fe1cd Use a class name that does not actually exist
Using "Foo", a class name that corresponds to no less than 22 fixture
classes, results in the first found "Foo" being loaded when one is
found by the ClassNotFoundFatalErrorHandler error handler, I am not sure
exactly why, but it is not really a big issue because this is a fatal
error handler and execution is not supposed to continue after that.
Except that is very much the case when running the whole test suite
sequentially with ./phpunit . Then we arrive to the DI component test
suite, and a failure happens because \\foo was not supposed to be defined:
> Failed asserting that exception message 'The definition for "\foo" has
> no class attribute, and appears to reference a class or interface in the
> global namespace. Leaving out the "class" attribute is only allowed for
> namespaced classes. Please specify the class attribute explicitly to get
> rid of this error.' contains 'The definition for "\foo" has no class.'.
This PR was merged into the 3.4 branch.
Discussion
----------
[Translator] Warm up the translations cache in dev
| Q | A
| ------------- | ---
| Branch? | 3.4
| Bug fix? | yes
| New feature? | no
| BC breaks? | no
| Deprecations? | no
| Tests pass? | yes
| Fixed tickets | -
| License | MIT
| Doc PR | -
This PR fixes a bug in development when using the DataCollectorTranslator: because it's not implementing WarmableInterface, the translations cache is not built during cache:clear during development.
Commits
-------
a5f1afca15 [Translator] Warm up the translations cache in dev
This PR was merged into the 3.4 branch.
Discussion
----------
[Translator] Fix wrong dump for PO files
| Q | A
| ------------- | ---
| Branch? | 3.4
| Bug fix? | yes
| New feature? | no
| BC breaks? | no
| Deprecations? | no
| Tests pass? | yes
| Fixed tickets |
| License | MIT
| Doc PR |
When working on #30909 I encountered this bug.
Currently the tests were passing, because the po fixture file contained only one translation.
Commits
-------
b54abfc643 Fix wrong dump for PO files
This PR was merged into the 3.4 branch.
Discussion
----------
[DependencyInjection] Fix a wrong error when using a factory
…d a call
| Q | A
| ------------- | ---
| Branch? | master
| Bug fix? | no
| New feature? | no <!-- don't forget to update src/**/CHANGELOG.md files -->
| BC breaks? | no <!-- see https://symfony.com/bc -->
| Deprecations? | no <!-- don't forget to update UPGRADE-*.md and src/**/CHANGELOG.md files -->
| Tests pass? | yes <!-- please add some, will be required by reviewers -->
| Fixed tickets | #30885 <!-- #-prefixed issue number(s), if any -->
| License | MIT
| Doc PR |
<!--
Write a short README entry for your feature/bugfix here (replace this comment block.)
This will help people understand your PR and can be used as a start of the Doc PR.
Additionally:
- Bug fixes must be submitted against the lowest branch where they apply
(lowest branches are regularly merged to upper ones so they get the fixes too).
- Features and deprecations must be submitted against the master branch.
-->
This a work in progress, it needs tests, firstly I wanted to ask the author to test and tell if it's ok.
Commits
-------
5d4e3a2d57 [WIP] [DependencyInjection] Fix a wrong error when using a factory and a call
This PR was submitted for the master branch but it was merged into the 4.2 branch instead (closes#30912).
Discussion
----------
Restore previous state for libxml option
| Q | A
| ------------- | ---
| Branch? | 4.2
| Bug fix? | yes
| New feature? | yes
| BC breaks? | no
| Deprecations? | no
| Tests pass? | yes
| Fixed tickets | n/a
| License | MIT
| Doc PR | n/a
Whenever libxml_use_internal_errors() is called, the previous value for
it should be restored. This used to be the case in this piece of code,
but it was wrongly removed in e53bf5839b ,
which has the nasty side effect of making the Validator component test
suite break with this message:
Validation failed: No DTD found!
Commits
-------
dde74aa8c9 Restore previous state for libxml option
Whenever libxml_use_internal_errors() is called, the previous value for
it should be restored. This used to be the case in this piece of code,
but it was wrongly removed in e53bf5839b ,
which has the nasty side effect of making the Validator component test
suite break with this message:
Validation failed: No DTD found!
This PR was merged into the 4.2 branch.
Discussion
----------
[Messenger] Uses the `SerializerStamp` when deserializing the envelope
| Q | A
| ------------- | ---
| Branch? | 4.2
| Bug fix? | yes
| New feature? | no
| BC breaks? | no
| Deprecations? | no
| Tests pass? | yes
| Fixed tickets | #29970
| License | MIT
| Doc PR | ø
Uses the `SerializerStamp` when decoding the message.
Commits
-------
ab55e8e7ef Uses the SerializerStamp when deserializing the envelope
* 3.4:
[Form] various minor fixes
bugfix: the terminal state was wrong and not reseted
[Console] Fix inconsistent result for choice questions in non-interactive mode
Define null return type for Constraint::getDefaultOption()
[HttpKernel] Fix DebugHandlersListener constructor docblock
Skip Glob brace test when GLOB_BRACE is unavailable
bumped Symfony version to 3.4.25
updated VERSION for 3.4.24
update CONTRIBUTORS for 3.4.24
updated CHANGELOG for 3.4.24
[EventDispatcher] cleanup
This PR was merged into the 3.4 branch.
Discussion
----------
[Form] Php doc fixes and cs + optimizations
| Q | A
| ------------- | ---
| Branch? | 3.4
| Bug fix? | no
| New feature? | no <!-- don't forget to update src/**/CHANGELOG.md files -->
| BC breaks? | no <!-- see https://symfony.com/bc -->
| Deprecations? | no <!-- don't forget to update UPGRADE-*.md and src/**/CHANGELOG.md files -->
| Tests pass? | yes <!-- please add some, will be required by reviewers -->
| Fixed tickets | ~
| License | MIT
| Doc PR | ~
<!--
Write a short README entry for your feature/bugfix here (replace this comment block.)
This will help people understand your PR and can be used as a start of the Doc PR.
Additionally:
- Bug fixes must be submitted against the lowest branch where they apply
(lowest branches are regularly merged to upper ones so they get the fixes too).
- Features and deprecations must be submitted against the master branch.
-->
Some micro optimizations may target master with some profiling but I would like some reviews first. Thanks!
Commits
-------
56429a6f08 [Form] various minor fixes
This PR was merged into the 3.4 branch.
Discussion
----------
[Console] Fix stty not reset when aborting in QuestionHelper::autocomplete()
| Q | A
| ------------- | ---
| Branch? | 3.4
| Bug fix? | yes
| New feature? | no <!-- don't forget to update src/**/CHANGELOG.md files -->
| BC breaks? | no <!-- see https://symfony.com/bc -->
| Deprecations? | no <!-- don't forget to update UPGRADE-*.md and src/**/CHANGELOG.md files -->
| Tests pass? | yes <!-- please add some, will be required by reviewers -->
| Fixed tickets | #30861 <!-- #-prefixed issue number(s), if any -->
| License | MIT
| Doc PR | none
<!--
Write a short README entry for your feature/bugfix here (replace this comment block.)
This will help people understand your PR and can be used as a start of the Doc PR.
Additionally:
- Bug fixes must be submitted against the lowest branch where they apply
(lowest branches are regularly merged to upper ones so they get the fixes too).
- Features and deprecations must be submitted against the master branch.
-->
Commits
-------
f54389b95c bugfix: the terminal state was wrong and not reseted
This PR was merged into the 4.2 branch.
Discussion
----------
[Lock][TestSuite]Ensure the parent process is always killed
| Q | A
| ------------- | ---
| Branch? | 4.2
| Bug fix? | yes, but for the test suite
| New feature? | no
| BC breaks? | no
| Deprecations? | no
| Tests pass? | yes
| Fixed tickets |
| License | MIT
| Doc PR | n/a
If you try to run the test suite but do not have a redis instance
running, the parent process that was supposed to be killed will never be
as the test is marked as skipped and the store never returned.
This results in the test suite hanging forever at the end.
In this patch, the exception is thrown again, and then caught in the trait, and
the parent gets killed as it should.
Commits
-------
94d41825fd Ensure the parent process is always killed
If you try to run the test suite but do not have a redis instance
running, the parent process that was supposed to be killed will never be
as the thing being thrown is not an exception.
This results in the test suite hanging forever at the end.
In this patch, the exception is thrown again, and then caught in the trait, and
the parent gets killed as it should.
This PR was submitted for the 4.2 branch but it was merged into the 3.4 branch instead (closes#30869).
Discussion
----------
[Validator] Define null return type for Constraint::getDefaultOption()
| Q | A
| ------------- | ---
| Branch? | 4.2
| Bug fix? | yes
| New feature? | no
| BC breaks? | no
| Deprecations? | no
| Tests pass? | yes
| Fixed tickets |
| License | MIT
| Doc PR |
The `Constraint::getDefaultOption()` method is documented with a return type of `string`, but has no `return` statement. This behaviour is expected, as the return value is compared against `null` elsewhere in the code, but the docblock causes problems with static analysis when extending this class.
This PR corrects the documented return type to `string|null` and adds an explicit `return null`.
Commits
-------
03987f2ba5 Define null return type for Constraint::getDefaultOption()
This PR was merged into the 4.2 branch.
Discussion
----------
[Routing] Fix: annotation loader ignores method's default values
| Q | A
| ------------- | ---
| Branch? | 4.1
| Bug fix? | yes
| New feature? | no
| BC breaks? | no
| Deprecations? |no
| Tests pass? | yes
| Fixed tickets | no
| License | MIT
| Doc PR | no
In some cases annotation loader ignores method param's default values.
For example this code won't work as expected:
```php
/**
* @Route("/hello/{name<\w++>}", methods="GET", name="hello")
*/
public function hello(Request $request, string $name = 'World'): Response
{
// If you try to open "/hello" path an exception (No route found for "GET /hello") will be thrown.
return $this->json([
'hello' => \sprintf('Hello, %s!', $name),
]);
}
```
Commits
-------
9b37793cbe [Routing] Fix: annotation loader ignores method's default values
This PR was merged into the 3.4 branch.
Discussion
----------
Skip Glob brace test when GLOB_BRACE is unavailable
| Q | A
| ------------- | ---
| Branch? | 3.4
| Bug fix? | yes
| New feature? | no
| BC breaks? | no
| Deprecations? | no
| Tests pass? | yes
| Fixed tickets | -
| License | MIT
| Doc PR | n/a
From [PHP.net](https://www.php.net/glob):
> Note: The GLOB_BRACE flag is not available on some non GNU systems, like Solaris.
This means running the testsuite in e.g. a php-alpine container fails atm. This test should be skipped in these environments.
Commits
-------
ce38fe3e58 Skip Glob brace test when GLOB_BRACE is unavailable