Currently BinaryFileResponse, when configured with X-Accel-Redirect sendfile type,
will only substitute file paths specified in X-Accel-Mapping. But if the provided
file path does not have a defined prefix, then the resulting header will include
the absolute path. Nginx expects a valid URI, therefore this will result in an
issue that is very hard to detect and debug as it will not show up in error logs
and instead the request would just hang for some time and then be re-served
without query parameters(?).
Use AssertEquals for these two specific case will do a better job,
since it'll convert both '0.1' and result of `getContent()` into PHP's
internal representation of floating-point and compares them and it should be fine.
Using `AssertSame` for this tests brings floating-point serialization
into consideration which of course will be php.ini specific.
In order not missing the type assertion point that `AssertSame` does,
we also perform `assertInternalType('string'...`
Sponsored-by: Platform.sh
This PR was merged into the 3.4 branch.
Discussion
----------
FragmentListener - fix typo in annotation
| Q | A
| ------------- | ---
| Branch? | 3.4
| Bug fix? | no
| New feature? | no
| BC breaks? | no
| Deprecations? | no
| Tests pass? | yes
| Fixed tickets |
| License | MIT
| Doc PR |
Just a small typo I've found in fragment listener :)
Commits
-------
b6ff836a49 fix typo
This PR was merged into the 3.4 branch.
Discussion
----------
[Workflow] Do not trigger extra guards
| Q | A
| ------------- | ---
| Branch? | 3.4
| Bug fix? | yes
| New feature? | no
| BC breaks? | no
| Deprecations? | no
| Tests pass? | yes
| Fixed tickets | #31582
| License | MIT
| Doc PR |
---
With this patch, guards are executed only on wanted transitions
**Note for merger**: This is already fixed (in a different manner) in 4.2, So this patch should not be included in 4.2, instead take: #31585 or discard changes of Workflow class but keep tests
Commits
-------
ad0619748e [Workflow] Do not trigger extra guard
The temp-file that the test currently creates is `/tmp/log`.
This may exist on many platforms already (including `platform.sh` app containers).
With the proposed patch way the collision will be less likely.
Sponsored-by: Platform.sh
This PR was merged into the 3.4 branch.
Discussion
----------
[WebProfilerBundle] Use absolute URL for profiler links
| Q | A
| ------------- | ---
| Branch? | 3.4
| Bug fix? | yes
| New feature? | no
| BC breaks? | no
| Deprecations? | no
| Tests pass? | n/a
| Fixed tickets | n/a
| License | MIT
| Doc PR | n/a
Generate absolute URL's so you can have a different (sub)domain for the profiler and the controller that uses it.
Also uses the link which the controller generated instead of always generating a new link in Twig. The changes shouldn't inpact normal behavior.
Commits
-------
13ee1fa2b7 Use absolute URL for when the profiler's domain differs from the controller's domain which initialises the profiler.
This PR was merged into the 3.4 branch.
Discussion
----------
[DI] fix using bindings with locators of service subscribers
| 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 | -
Spotted during the workshop in Sao Paulo, with @tsantos84, @andreia and others :)
Commits
-------
7146b95303 [DI] fix using bindings with locators of service subscribers
This PR was squashed before being merged into the 3.4 branch (closes#31568).
Discussion
----------
[Process] Fix infinite waiting for stopped process
| Q | A
| ------------- | ---
| Branch? | 3.4
| Bug fix? | yes
| New feature? | no
| BC breaks? | no
| Deprecations? | no
| Tests pass? | yes
| Fixed tickets | #31548
| License | MIT
### Description
Add a regression test `Symfony\Component\Process\Tests\ProcessTest:testWaitStoppedDeadProcess` to reproduce the related bug #31548 . It consists of one file `ErrorProcessInitiator.php`, which executes as another process because otherwise if `Process::wait()` goes in an infinite loop (which is test checks) it will be impossible to handle it and stop test correctly.
The second commit contains bug fix, which is only `$this->checkTimeout();` call, that prevents infinite loop.
Commits
-------
6b9ee1e1c5 [Process] Fix infinite waiting for stopped process
This PR was merged into the 3.4 branch.
Discussion
----------
[Ldap] add a test to getResources and binding
| Q | A
| ------------- | ---
| Branch? | 3.4
| Bug fix? | no
| New feature? | no <!-- please update src/**/CHANGELOG.md files -->
| BC breaks? | no <!-- see https://symfony.com/bc -->
| Deprecations? | no <!-- please update UPGRADE-*.md and src/**/CHANGELOG.md files -->
| Tests pass? | yes <!-- please add some, will be required by reviewers -->
| Fixed tickets | none <!-- #-prefixed issue number(s), if any -->
| 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/roadmap):
- 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 the master branch.
-->
We could add more unit tests, but this is a beginning.
Commits
-------
d08f195502 minor: add some test in the ldap component
When using factories it's common best practice to use interface as class name, especially in cases
where you know impl can differ. Before this fix ProxyManager did not allow these to be lazy.
In our case this has lead several to hard to debug issues on classes we need to mark as lazyi
and often a need to add lazy on decorators if there are any or other workarounds.
As we have had this issue, and still have on both 2.8 and 3.4 this is opened against 2.8.
This PR was merged into the 3.4 branch.
Discussion
----------
[Doctrine] Respect parent class contract in ContainerAwareEventManager
| Q | A
| ------------- | ---
| Branch? | 3.4
| Bug fix? | yes
| New feature? | no
| BC breaks? | no
| Deprecations? | no
| Tests pass? | yes, failures looks unrelated
| Fixed tickets | #31051
| License | MIT
| Doc PR | -
According to method signature of [original EventManager](https://github.com/doctrine/event-manager/blob/master/lib/Doctrine/Common/EventManager.php#L50) `getListeners` method should return array of initialized objects but now it returns array of strings of listener service names.
Commits
-------
42d62721fd Respect parent class contract in ContainerAwareDoctrineEventManager
This PR was squashed before being merged into the 3.4 branch (closes#31421).
Discussion
----------
[Routing][AnnotationClassLoader] fix utf-8 encoding in default route name
| Q | A
| ------------- | ---
| Branch? | 3.4
| Bug fix? | yes
| New feature? | no
| BC breaks? | no
| Deprecations? | no
| Tests pass? | yes
| License | MIT
If controller, or one of its methods, contain a unicode character and you run:
```
./bin/console debug:router
```
you get this:
This PR fixes it into this:
Commits
-------
7ab52d3c36 [Routing][AnnotationClassLoader] fix utf-8 encoding in default route name
If the included file contains an error, it hides the real error. This
makes debugging harder.
How to reproduce:
```
composer create-project symfony/skeleton symfony-3.4 3.4
cd symfony-3.4
composer req monolog
```
Add to `monolog.yaml`:
```yaml
elasticsearch:
type: "elasticsearch"
elasticsearch:
host: 'elasticsearch'
port: '9200'
index: 'ep_php_logs_dev'
level: 'debug'
tags: 'monolog.logger'
channels: ['!event']
```
This will fail because the the \Elastica\Client class does not exist.
But this error will be hidden by the `ClassNotFoundFatalErrorHandler`
because it will try to load the `Symfony\Component\Kernel\Client` and
this class extends `Symfony\Component\BrowserKit\Client`. The last one
is a soft dependency...
---
Before
```
Fatal error: Uncaught Error: Class 'Symfony\Component\BrowserKit\Client' not found in /tmp/symfony-3.4/vendor/symfony/http-kernel/Client.php:31
```
After:
```
Fatal error: Uncaught Symfony\Component\Debug\Exception\ClassNotFoundException: Attempted to load class "Client" from namespace "Elastica".
Did you forget a "use" statement for another namespace? in /tmp/symfony-es/var/cache/dev/ContainerWXN4mS9/srcApp_KernelDevDebugContainer.php:303
```
This PR was squashed before being merged into the 3.4 branch (closes#31438).
Discussion
----------
[Serializer] Fix denormalization of object with variadic constructor typed argument
| Q | A
| ------------- | ---
| Branch? | 3.4 up to 4.2
| Bug fix? | yes
| New feature? | no
| BC breaks? | no
| Deprecations? | no
| Tests pass? | yes
| Fixed tickets | #31436
| License | MIT
This PR adds a test to demonstrate the bug, and a fix to squash it.
Commits
-------
c8c3c56cc8 [Serializer] Fix denormalization of object with variadic constructor typed argument
This PR was merged into the 3.4 branch.
Discussion
----------
[Console] Commands with an alias should not be recognized as ambiguous when using register
| Q | A
| ------------- | ---
| Branch? | 2.8
| Bug fix? | yes
| New feature? | no
| BC breaks? | no
| Deprecations? | no
| Tests pass? | yes
| Fixed tickets | #25355
| License | MIT
| Doc PR |
I think when passing an alias, it should not be treated as a ambiguous command since it's configured to response to it.
I've [pushed a commit](2f5209a687) that reproduce the bug and with this patch it does work.
Commits
-------
ae7ee46465 [Console] Commands with an alias should not be recognized as ambiguous
This PR was squashed before being merged into the 3.4 branch (closes#31371).
Discussion
----------
[DI] Removes number of elements information in debug mode
| Q | A
| ------------- | ---
| Branch? | 3.4
| Bug fix? | yes
| New feature? | no
| BC breaks? | no
| Deprecations? | no
| Tests pass? | yes
| Fixed tickets | #31340
| License | MIT
| Doc PR | -
With this services config:
```yaml
my_service:
class: stdClass
arguments: [!tagged my_tag]
my_tagged_service_1:
class: stdClass
tags: [my_tag]
my_tagged_service_2:
class: stdClass
tags: [my_tag]
```
Executing `./bin/console debug:container my_service --show-arguments --env=dev` resulted in
```bash
Information for Service "my_service"
====================================
---------------- -------------------------
Option Value
---------------- -------------------------
Service ID my_service
Class stdClass
Tags -
Public no
Synthetic no
Lazy no
Shared yes
Abstract no
Autowired yes
Autoconfigured yes
Arguments Iterator (0 element(s))
---------------- -------------------------
```
With this fix the output changed to:
```bash
Information for Service "my_service"
====================================
---------------- ------------
Option Value
---------------- ------------
Service ID my_service
Class stdClass
Tags -
Public no
Synthetic no
Lazy no
Shared yes
Abstract no
Autowired yes
Autoconfigured yes
Arguments Tagged Iterator for "my_tag"
---------------- ------------
```
and with `./bin/console debug:container my_service --show-arguments --env=prod`
```bash
Information for Service "my_service_tagged_iterator"
====================================================
---------------- ---------------------------------------------
Option Value
---------------- ---------------------------------------------
Service ID my_service
Class stdClass
Tags -
Public no
Synthetic no
Lazy no
Shared yes
Abstract no
Autowired yes
Autoconfigured yes
Arguments Tagged Iterator for "my_tag" (2 element(s))
---------------- ---------------------------------------------
```
Commits
-------
0da4b83197 [DI] Removes number of elements information in debug mode
This PR was merged into the 3.4 branch.
Discussion
----------
[FrameworkBundle] clarify the possible class/interface of the cache
| 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 the fallback cache pool is returned (on PHP 5.6, HHVM, or when
Opcache is disabled), the configured service can be any implementation
of the CacheItemPoolInterface.
Commits
-------
40273745ce clarify the possible class/interface of the cache
This PR was merged into the 3.4 branch.
Discussion
----------
[Intl] Add FallbackTrait for data generation
| Q | A
| ------------- | ---
| Branch? | 3.4
| Bug fix? | yes
| New feature? | no
| BC breaks? | no <!-- see https://symfony.com/bc -->
| Deprecations? | no
| Tests pass? | yes <!-- please add some, will be required by reviewers -->
| Fixed tickets | #... <!-- #-prefixed issue number(s), if any -->
| License | MIT
| Doc PR | symfony/symfony-docs#... <!-- required for new features -->
This is the last architectural change for the Intl data compilation. Promised.
It fixes de-duplicating a locale from its fallback locale. The problem is it uses a while-loop, comparing the locale to each fallback locale.
Given
- `root` (val=A)
- `ur` (val=B)
- `ur_IN` (val=A)
We have an edge case where a locale (ur_IN) override its fallback locale (ur), setting/restoring the value back to the root locale. This happens for the GMT format in the timezone bundle i know of ... in this case the `ur_IN` locale needs to write its own value.
The current approach is a while-loop comparing each fallback locale (ur, root) to the current locale (ur_IN). Eventually comparing `ur_IN <> root`, which causes a wrong diff, as such `ur_IN` falls back to `ur` providing the wrong value (val=B, where val=A is expected).
The new approach uses recursion so we only compare `ur <> ur_IN`, where `ur_IN` on itself is compared to `root`.
4.2) https://github.com/ro0NL/symfony/commit/e24d8e6
4.3) https://github.com/ro0NL/symfony/commit/31591d0
Commits
-------
36ddfd58b9 [Intl] Add FallbackTrait for data generation
