This PR was merged into the 4.4 branch.
Discussion
----------
[Intl] Update the ICU data to 68.1
| Q | A
| ------------- | ---
| Branch? | 4.4
| Bug fix? | no
| New feature? | no
| Deprecations? | no
| Tickets | -
| License | MIT
| Doc PR | -
Ideally, this should be merged after #39370.
Commits
-------
f3eda44058 [Intl] Update the ICU data to 68.1
This PR was merged into the 4.4 branch.
Discussion
----------
[FrameworkBundle] fix preserving some special chars in the query string
| Q | A
| ------------- | ---
| Branch? | 4.4
| Bug fix? | yes
| New feature? | no
| Deprecations? | no
| Tickets | Fix#39307
| License | MIT
| Doc PR | -
Commits
-------
6ecae5719c [FrameworkBundle] fix preserving some special chars in the query string when redirecting
This PR was merged into the 4.4 branch.
Discussion
----------
Run intl-data tests on resources change
| Q | A
| ------------- | ---
| Branch? | 4.4
| Bug fix? | no
| New feature? | no
| Deprecations? |no
| Tickets | -
| License | MIT
| Doc PR | -
`intl-data` tests are run whenever we update ICU data as they're rather time-consuming.
This proposal automates intl-data test runs with Github Actions. Tests will only be run on pull requests or pushes that modifyy ICU data. Full run on Github Actions takes 6-7mins. Example run: https://github.com/jakzal/symfony/actions/runs/406526765
<img width="1204" alt="image" src="https://user-images.githubusercontent.com/190447/101393171-92a28c00-38be-11eb-9efa-8f0a75ff04b0.png">
Commits
-------
745020892f Run intl-data tests on resources change
This PR was squashed before being merged into the 4.4 branch.
Discussion
----------
[HttpFoundation] Fix TypeError: Argument 1 passed to JsonResponse::setJson() must be of the type string, object given
| Q | A
| ------------- | ---
| Branch? | 4.4
| Bug fix? | yes
| New feature? | no
| Deprecations? | no
| Tickets |
| License | MIT
| Doc PR |
This PR fixes the type error when user set `null` as data in the `new JsonResponse(null, 200, [], true)` and true in the fourth argument to mark that value is json.
```
TypeError: Argument 1 passed to Symfony\Component\HttpFoundation\JsonResponse::setJson() must be of the type string, object given, called in /home/projects/symfony/src/Symfony/Component/HttpFoundation/JsonResponse.php on line 52
```
Commits
-------
7fcba9611d [HttpFoundation] Fix TypeError: Argument 1 passed to JsonResponse::setJson() must be of the type string, object given
This PR was merged into the 4.4 branch.
Discussion
----------
[Messenger] Test generated SQL
| Q | A
| ------------- | ---
| Branch? | 4.4
| Bug fix? | no
| New feature? | no
| Deprecations? | no
| Tickets | N/A
| License | MIT
| Doc PR | N/A
Follow-up to #39166.
This is an attempt to the SQL generated by the `Symfony\Component\Messenger\Transport\Doctrine\Connection::get()` method. The challange was to create a locking SELECT statement, taking into account the different ways to create a lock in different SQL dialects.
This test verifies if the correct lock statements are generated for MySQL and SQL Server.
Commits
-------
1f1b62afb6 [Messenger] Test generated SQL
This PR was merged into the 4.4 branch.
Discussion
----------
[Config] YamlReferenceDumper: No default value required for VariableNode with array example
| Q | A
| ------------- | ---
| Branch? | 4.4
| Bug fix? | yes
| New feature? | no
| Deprecations? |
| Tickets |
| License | MIT
| Doc PR |
This will fix#39328 in a better way.
A `VariableNode` may have an array as value and also a default value as "null". (Like the workflow component).
When we dump the config, we generate invalid yaml:
```
workflows:
enabled: false
workflows:
# Prototype
name:
# Select which Transition events should be dispatched for this Workflow
events_to_dispatch: null
# Examples:
- workflow.enter
- workflow.transition
```
With this PR, we will remove the `null` default value from the dumped config.
#SymfonyHackday
Commits
-------
9104fd4539 [Config] YamlReferenceDumper: No default value required for VariableNode with array example
This PR was merged into the 4.4 branch.
Discussion
----------
[Form] do not apply the Valid constraint on scalar form data
| Q | A
| ------------- | ---
| Branch? | 4.4
| Bug fix? | yes
| New feature? | no
| Deprecations? | no
| Tickets | Fix#38039
| License | MIT
| Doc PR |
#SymfonyHackday
Commits
-------
453bb3ebde do not apply the Valid constraint on scalar form data
41b9457887 [Test] Reproduce issue with cascading validation
This PR was merged into the 4.4 branch.
Discussion
----------
[PhpUnitBridge] Fixed PHPunit 9.5 compatibility
| Q | A
| ------------- | ---
| Branch? | 4.4
| Bug fix? | yes
| New feature? | no
| Deprecations? | no
| Tickets | Fix#39329
| License | MIT
| Doc PR | -
Let's see if this works 🍀
#SymfonyHackday
Commits
-------
5134de52f1 Added compatibility with PHPunit 9.5
This PR was squashed before being merged into the 4.4 branch.
Discussion
----------
[Messenger] StopWorkersCommand improve doc helper
| Q | A
| ------------- | ---
| Branch? | 4.4
| Bug fix? | no
| New feature? | no
| Deprecations? | no
| Tickets | Fix
| License | MIT
Small improvement :)
Commits
-------
e136068743 [Messenger] StopWorkersCommand improve doc helper
This PR was merged into the 4.4 branch.
Discussion
----------
[SecurityBundle] Don't use the container as resource type in fixtures
| Q | A
| ------------- | ---
| Branch? | 4.4
| Bug fix? | yes
| New feature? | no
| Deprecations? | no
| Tickets | N/A
| License | MIT
| Doc PR | N/A
Discovered while working on #39292: Some fixtures of SecurityBundle call config loaders and pass the whole container as resource type. This does not really make sense.
Commits
-------
e7e38b7213 [SecurityBundle] Don't use the container as resource type in fixtures.
This PR was merged into the 4.4 branch.
Discussion
----------
[HttpKernel] Fix bug with whitespace in Kernel::stripComments()
| Q | A
| ------------- | ---
| Branch? | 4.4
| Bug fix? | yes
| New feature? | no
| Deprecations? | no
| Tickets |
| License | MIT
| Doc PR |
`Kernel::stripComments()` removes significant whitespace in some cases.
I noticed this in one of the generated classes, the code `<?php include_once \dirname(__DIR__).'/file.php';` got replaced with `<?php include_once\dirname(__DIR__).'/file.php';` which is a syntax error.
Commits
-------
8d368e1fe3 Fix bug with whitespace in Kernel::stripComments()
This PR was merged into the 4.4 branch.
Discussion
----------
[HttpFoundation] Make test pass without Xdebug
| Q | A
| ------------- | ---
| Branch? | 4.4
| Bug fix? | yes
| New feature? | no
| Deprecations? | no
| Tickets | N/A
| License | MIT
| Doc PR | N/A
Fixup of #39293. This change should make the tests pass without Xdebug. Sorry for the messup. 🙈
Commits
-------
cf94f2a591 [HttpFoundation] Make test pass without Xdebug.
This PR was merged into the 4.4 branch.
Discussion
----------
[HttpFoundation] Make sure we reuse the current PHP binary for the webserver process in functional tests.
| Q | A
| ------------- | ---
| Branch? | 4.4
| Bug fix? | yes
| New feature? | no
| Deprecations? | no
| Tickets | N/A
| License | MIT
| Doc PR | N/A
While debugging HttpFoundation's functional tests with different PHP binaries, I noticed that the webserver process that is started there would always use the `php` binary from my `$PATH` instead of the binary I'm running PHPUnit with.
This PR attempts to fix this by using the `PHP_BINARY` constant.
Commits
-------
9208c69b21 [HttpFoundation] Make sure we reuse the current PHP binary for the webserver process in functional tests.
This PR was squashed before being merged into the 4.4 branch.
Discussion
----------
[FrameworkBundle] TextDescriptor::formatControllerLink checked method…
…_exists method before it ensures that $controller is not null
| Q | A
| ------------- | ---
| Branch? | 4.4
| Bug fix? | yes
| New feature? | no
| Deprecations? | no
| Tickets | -
| License | MIT
| Doc PR | -
Today I tried PHP8 with Symfony 5.2. - by checking my routes with
```
bin/console debug:router
```
It throws
```
method_exists(): Argument #1 ($object_or_class) must be of type object|string, null given
```
The reason is the configured auth route for the login API defined in routes.yaml:
```
authorization::login:
path: '/login'
```
This route has no controller configured and the value of $controller in `TextDescriptor::formatControllerLink` in the FrameworkBundle is `null`. This method has a elseif condition with `method_exists` without a check if $controller is a string or object.
In PHP8 this throws an exception/error and did not work. This PR checks that `$controller` is not null before the method_exists check to prevent this failure.
I'm not sure how to test this in the existing Testsuite
Commits
-------
67bd779ef4 [FrameworkBundle] TextDescriptor::formatControllerLink checked method…
This PR was merged into the 4.4 branch.
Discussion
----------
[HttpFoundation] Ignore stack trace printed by Xdebug 3
| Q | A
| ------------- | ---
| Branch? | 4.4
| Bug fix? | yes
| New feature? | no
| Deprecations? | no
| Tickets | N/A
| License | MIT
| Doc PR | N/A
If we're running HttpFoundation's functional tests with Xdebug 3, a stack trace will be printed when a warning is emitted. This confuses our `cookie_max_age` test currently where we perform a `setcookie()` call that will emit a warning.
I have patched the corresponding fixture so the printed stack trace is ignored.
A failed test can be seen here: https://travis-ci.com/github/symfony/symfony/jobs/452077515#L11078
Commits
-------
b67baa4e7f [HttpFoundation] Ignore stack trace printed by Xdebug 3.
This PR was squashed before being merged into the 4.4 branch.
Discussion
----------
[HttpClient] throw clearer error when no scheme is provided
| Q | A
| ------------- | ---
| Branch? | 4.4
| Bug fix? | yes
| New feature? | no
| Deprecations? | no
| Tickets | Fix#39285
| License | MIT
| Doc PR | N/A
This could be considred a BC break, as previously this would've a `TransportExcepiton`, instead of an `InvalidArgumentException`. But i see no reason to catch this specific error, as it would generally be a configuration error.
Commits
-------
4d821d6c34 [HttpClient] throw clearer error when no scheme is provided
This PR was merged into the 4.4 branch.
Discussion
----------
Fix github pr template and include 5.2 for bugfixes
| Q | A
| ------------- | ---
| Branch? | 5.x
| Bug fix? | no
| New feature? | no <!-- please update src/**/CHANGELOG.md files -->
| Deprecations? | no <!-- please update UPGRADE-*.md and src/**/CHANGELOG.md files -->
| Tickets | - <!-- prefix each issue number with "Fix #", no need to create an issue if none exist, explain below instead -->
| License | MIT
| Doc PR | - <!-- required for new features -->
<!--
Replace this notice by a short README for your feature/bugfix. This will help people
understand your PR and can be used as a start for the documentation.
Additionally (see https://symfony.com/releases):
- Always add tests and ensure they pass.
- Never break backward compatibility (see https://symfony.com/bc).
- Bug fixes must be submitted against the lowest maintained 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 branch 5.x.
-->
Commits
-------
c53a6d04c8 Fix github pr template and include 5.2 for bugfixes
This PR was merged into the 4.4 branch.
Discussion
----------
Remove Xdebug from php-extra runs
| Q | A
| ------------- | ---
| Branch? | 4.4
| Bug fix? | yes
| New feature? | no
| Deprecations? | no
| Tickets | N/A
| License | MIT
| Doc PR | N/A
Apparently, Travis already bundles Xdebug 3 with their PHP 7.4 binaries. This causes a lot of segfaults in our CI at the moment. This PR removes the Xdebug configuration files from all PHP tarballs that we download for our Travis build.
Commits
-------
1bb0fefbd5 Remove Xdebug from php-extra runs.
This PR was squashed before being merged into the 4.4 branch.
Discussion
----------
[DependencyInjection] Fixed incorrect report for private services if required service does not exist
…does not exist
| Q | A
| ------------- | ---
| Branch? | 4.4
| Bug fix? | yes
| New feature? | no
| Deprecations? | no
| Tickets | Fix#29359
| License | MIT
| Doc PR | no
looks like `CheckExceptionOnInvalidReferenceBehaviorPass` can be moved to `beforeRemovingPasses` section without any consequences. this solves the problem and all tests still pass
Commits
-------
39bd05c5bc [DependencyInjection] Fixed incorrect report for private services if required service does not exist