This PR was merged into the 3.4 branch.
Discussion
----------
[Translator] fix performance issue in MessageCatalogue and catalogue operations
| Q | A
| ------------- | ---
| Branch? | 3.4
| Bug fix? | no
| New feature? | no
| Deprecations? | no
| Tickets | -
| License | MIT
| Doc PR | -
In our project we use lots of catalogue operations during importing of translations to our system and we ran into performance issue. Code profiler showed lots or `array_replace` calls in [MessageCatalogue::add](https://github.com/symfony/symfony/blob/3.4/src/Symfony/Component/Translation/MessageCatalogue.php#L128) method. This method is actually called by [MessageCatalogue::set](https://github.com/symfony/symfony/blob/3.4/src/Symfony/Component/Translation/MessageCatalogue.php#L70), which is quite an overkill, because `MessageCatalogue::set` is meant to set only one translation at a time. Method was reworked. `MergeOperation` and `TargetOperation` was reworked as well to use this improved `MessageCatalogue::set` method instead of constructing array with only one translation and passing it to `MessageCatalogue::add` method.
Table shows execution time before and after
| | Time in seconds (avg. of 10 executions)
----------- | ------
Before | 50
After | 8
Looks like 4.* and 5.* versions can also be improved by the same changes.
<!--
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):
- 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 master.
-->
Commits
-------
5179af4796 [Translator] Performance improvement in MessageCatalogue and catalogue operations.
The used phpDocumentor library DocBlockReflection contained an BC break
that broke this component. The patch was applied in the recent released v4.3.4
version. But since it is unclear how long this issue existed it is not possible
to exclude a certain version. Therefor also `\RuntimeExpception` needs to be catched.
The BC break is possibly caused by a change in the TypeResolver library used by the
DocBlockReflection which is now supporting the more populair generics notation for arrays.
* 5.0: (24 commits)
Removing unused variable
Fixed#35084
Add missing use statement
[HttpClient] fix scheduling pending NativeResponse
do not overwrite variable value
[Profiler] wording
Use spaces correctly to display options in DebugCommand
Add supported schemes doc blocks type
X-Accel Nginx URL updated
ticket-30197 [Validator] Add the missing translations for the Chinese (Taiwan) ("zh_TW") locale
Fixed test added in #35022
Use locale_parse for computing fallback locales
[Console] Fix filtering out identical alternatives when there is a command loader
[String][UnicodeString] Remove unneeded flag in chunk regex pattern
add note about HTTP status code change
Migrate server:log command away from WebServerBundle
[DependencyInjection][CheckTypeDeclarationsPass] Handle \Closure for callable
[Security] Fix missing defaults for auto-migrating encoders
bumped Symfony version to 5.0.3
updated VERSION for 5.0.2
...
* 4.4:
Fixed#35084
Add missing use statement
[HttpClient] fix scheduling pending NativeResponse
do not overwrite variable value
[Profiler] wording
Use spaces correctly to display options in DebugCommand
X-Accel Nginx URL updated
ticket-30197 [Validator] Add the missing translations for the Chinese (Taiwan) ("zh_TW") locale
Fixed test added in #35022
Use locale_parse for computing fallback locales
[Console] Fix filtering out identical alternatives when there is a command loader
add note about HTTP status code change
Migrate server:log command away from WebServerBundle
[DependencyInjection][CheckTypeDeclarationsPass] Handle \Closure for callable
[Security] Fix missing defaults for auto-migrating encoders
bumped Symfony version to 4.4.3
updated VERSION for 4.4.2
updated CHANGELOG for 4.4.2
* 4.3:
Fixed#35084
Add missing use statement
[HttpClient] fix scheduling pending NativeResponse
do not overwrite variable value
[Profiler] wording
Use spaces correctly to display options in DebugCommand
X-Accel Nginx URL updated
ticket-30197 [Validator] Add the missing translations for the Chinese (Taiwan) ("zh_TW") locale
Fixed test added in #35022
Use locale_parse for computing fallback locales
[Console] Fix filtering out identical alternatives when there is a command loader
This PR was merged into the 4.3 branch.
Discussion
----------
[HttpClient] fix scheduling pending NativeResponse
| Q | A
| ------------- | ---
| Branch? | 4.3
| Bug fix? | yes
| New feature? | no
| Deprecations? | no
| Tickets | -
| License | MIT
| Doc PR | -
There must be one pending list per `ResponseStream` instance.
Currently, we start unrelated responses and this can lead to broken iterators when the unrelated response throws because it is a 3/4/5xx.
Commits
-------
a90a6c9c48 [HttpClient] fix scheduling pending NativeResponse
This PR was merged into the 4.3 branch.
Discussion
----------
[Cache] do not overwrite variable value
| Q | A
| ------------- | ---
| Branch? | 4.3
| Bug fix? | yes
| New feature? | no
| Deprecations? | no
| Tickets | Fix#35030
| License | MIT
| Doc PR |
Commits
-------
e379dbbf21 do not overwrite variable value
* 3.4:
Add missing use statement
[Profiler] wording
X-Accel Nginx URL updated
ticket-30197 [Validator] Add the missing translations for the Chinese (Taiwan) ("zh_TW") locale
Fixed test added in #35022
Use locale_parse for computing fallback locales
[Console] Fix filtering out identical alternatives when there is a command loader
This PR was merged into the 3.4 branch.
Discussion
----------
[Translation] Use `locale_parse` for computing fallback locales
| Q | A
| ------------- | ---
| Branch? | 3.4
| Bug fix? | yes
| New feature? | no
| Deprecations? | no
| Tickets | N/A
| License | MIT
| Doc PR | N/A
As done in this PR https://github.com/symfony/symfony/pull/24157 for the `Intl` component, the `Translation` component should use `locale_parse` as well when available.
It will allow to manage [BCP 47](https://tools.ietf.org/html/bcp47) locales, which is why it is considered a bugfix ([locale_set_default](https://www.php.net/manual/en/locale.setdefault.php) is using BCP 47 compliant locale).
As done with the forementioned PR, there is also a fallback to make it work with `-`.
Sadly, I think it will create some conflicts when merging it upstream since the modified code has changed little by little.
Commits
-------
3657c0e664 Use locale_parse for computing fallback locales
This PR was squashed before being merged into the 5.1-dev branch.
Discussion
----------
[Form] Added default `inputmode` attribute to Search, Email and Tel form types
| Q | A
| ------------- | ---
| Branch? | master
| Bug fix? | no
| New feature? | yes
| Deprecations? | no
| Tickets | no
| License | MIT
| Doc PR | no
There is an HTML5 attribute `inputmode`. See https://developer.mozilla.org/en-US/docs/Web/HTML/Global_attributes/inputmode It is supported by most of mobile browsers.
There are such supported values for `inputmode`:
* `none`
* `text` (default value)
* `decimal`
* `numeric `
* `tel`
* `search`
* `email`
* `url`
The `url` inputmode is already implemented in UrlType https://github.com/symfony/symfony/blob/master/src/Symfony/Component/Form/Extension/Core/Type/UrlType.php#L39
I propose to add `tel`, `search`, `email` as default form view attribute to the relevant form types. Why only these three? `url` is already implemented, `none` makes no sense as default value for any form type, `text` is a default input type for any browser.
`decimal` and `numeric` has different behaviour on iOS. iOS doesn't show `-` (minus) sign on keyboard for these input modes. Of course in cases, when only positive numbers are expected, it is normal. But it is not suitable as default value in Symfony form type. Developers can add this attribute manually in their forms, if they need only positive numbers.
But `search`, `tel` and `email` input modes don't have problems and can be added as default attributes to Symfony form types. This will improve user experience, while using web-sites developed on Symfony on mobile devices. I add it into the *Type classes inside `buildView` method, so it will be possible to override this parameter if needed.
If you are interested in how it looks like in mobile browsers, you can open this link https://inputmodes.com/ on you mobile device.
Here are screenshots, how mobile keyboard looks like on Android and iOS with using `inputmode` attribute.
## `tel` inputmode
![image](https://user-images.githubusercontent.com/815865/70866507-45256480-1f73-11ea-9e82-320ef0b978ab.png)
## `email` inputmode
![image](https://user-images.githubusercontent.com/815865/70866502-3a6acf80-1f73-11ea-8f4c-3e7faca47f54.png)
## `search` inputmode
![image](https://user-images.githubusercontent.com/815865/70866498-2cb54a00-1f73-11ea-9fdc-2bb8abed107d.png)
Commits
-------
dbc500feb7 [Form] Added default `inputmode` attribute to Search, Email and Tel form types
This PR was merged into the 5.0 branch.
Discussion
----------
[String][UnicodeString] Remove unneeded flag in chunk regex pattern
| Q | A
| ------------- | ---
| Branch? | 5.0
| Bug fix? | no
| New feature? | no
| Deprecations? | no
| Tickets | -
| License | MIT
| Doc PR | -
The pattern cannot contain a `.`.
Commits
-------
dcb7c9a484 [String][UnicodeString] Remove unneeded flag in chunk regex pattern
This PR was merged into the 3.4 branch.
Discussion
----------
Fixed test added in #35022
| Q | A
| ------------- | ---
| Branch? | master for features / 3.4, 4.3, 4.4 or 5.0 for bug fixes <!-- see below -->
| Bug fix? | yes/no
| New feature? | yes/no <!-- please update src/**/CHANGELOG.md files -->
| Deprecations? | yes/no <!-- please update UPGRADE-*.md and src/**/CHANGELOG.md files -->
| Tickets | Fix #... <!-- prefix each issue number with "Fix #", if any -->
| License | MIT
| Doc PR | symfony/symfony-docs#... <!-- 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):
- 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 master.
-->
Commits
-------
6eeec7c270 Fixed test added in #35022
* 5.0:
[DependencyInjection] Fix binding tagged services to containers
[ProxyManager] fix generating proxies for root-namespaced classes
[DoctrineBridge] Cleanup 3.4 legacy
[DI] skip looking for config class when the extension class is anonymous
Fix typo
Docs - Update debug section of UPGRADE guides for 4.4 and 5.0 versions.
Fix invalid typehint for subject in is_granted Twig function
[Dotenv] FIX missing getenv
[HttpFoundation] fix pdo session handler for sqlsrv
[HttpClient][Psr18Client] Remove Psr18ExceptionTrait
[HttpKernel] ignore failuresgenerated by opcache.restrict_api
* 4.4:
[DependencyInjection] Fix binding tagged services to containers
[ProxyManager] fix generating proxies for root-namespaced classes
[DI] skip looking for config class when the extension class is anonymous
Fix typo
Docs - Update debug section of UPGRADE guides for 4.4 and 5.0 versions.
[Dotenv] FIX missing getenv
[HttpFoundation] fix pdo session handler for sqlsrv
[HttpClient][Psr18Client] Remove Psr18ExceptionTrait
[HttpKernel] ignore failuresgenerated by opcache.restrict_api
This PR was merged into the 4.4 branch.
Discussion
----------
[DependencyInjection] Fix binding tagged services to containers
| Q | A
| ------------- | ---
| Branch? | 4.4
| Bug fix? | yes
| New feature? | no
| Deprecations? | no
| Tickets | Fix#34913
| License | MIT
| Doc PR | -
We missed this part in #33623
Commits
-------
9d48ba4f11 [DependencyInjection] Fix binding tagged services to containers
* 4.3:
[ProxyManager] fix generating proxies for root-namespaced classes
[DI] skip looking for config class when the extension class is anonymous
Fix typo
[Dotenv] FIX missing getenv
[HttpClient][Psr18Client] Remove Psr18ExceptionTrait
* 3.4:
[ProxyManager] fix generating proxies for root-namespaced classes
[DI] skip looking for config class when the extension class is anonymous
Fix typo
[Dotenv] FIX missing getenv
This PR was merged into the 3.4 branch.
Discussion
----------
[DI] skip looking for config class when the extension class is anonymous
| Q | A
| ------------- | ---
| Branch? | 3.4
| Bug fix? | yes
| New feature? | no
| Deprecations? | no
| Tickets | Fix#34613
| License | MIT
| Doc PR | -
Commits
-------
1c7eda4649 [DI] skip looking for config class when the extension class is anonymous
This PR was merged into the 5.1-dev branch.
Discussion
----------
[DI] allow "." and "-" in env processor lines
| Q | A
| ------------- | ---
| Branch? | master
| Bug fix? | no
| New feature? | yes
| Deprecations? | no
| Tickets | Fix#34864
| License | MIT
| Doc PR | -
As explained in the linked issue, this is especially usefull with the `key` processor.
Commits
-------
231c505a47 [DI] allow "." and "-" in env processor lines
* 5.0:
[Config] improve perf of glob discovery when GLOB_BRACE is not available
use utf8mb4_bin to align code with documentation
[HttpClient] make pushed responses retry-able
[VarDumper] ignore failing __debugInfo()
* 4.4:
[Config] improve perf of glob discovery when GLOB_BRACE is not available
use utf8mb4_bin to align code with documentation
[HttpClient] make pushed responses retry-able
[VarDumper] ignore failing __debugInfo()
This PR was merged into the 4.3 branch.
Discussion
----------
[HttpClient] make pushed responses retry-able
| Q | A
| ------------- | ---
| Branch? | 4.3
| Bug fix? | yes
| New feature? | no
| Deprecations? | no
| Tickets | https://github.com/orgs/symfony/projects/1#card-30499375
| License | MIT
| Doc PR | -
This moves the PUSH matching logic down so that the curl handle of pushed responses can be properly configured. This should make pushed requests retry-able when they fail just after the push-promise frame.
Commits
-------
c2864f65ab [HttpClient] make pushed responses retry-able
This PR was squashed before being merged into the 5.1-dev branch (closes#34548).
Discussion
----------
Added access decision strategy to respect voter priority
| Q | A
| ------------- | ---
| Branch? | master
| Bug fix? | no
| New feature? | yes
| Deprecations? | no
| Tickets | -
| License | MIT
| Doc PR | _will happily do if this is of interest/to be merged 🙃_
The priority-based access decision strategy will decide based on the first voter that does not abstain from the decision. Security voters can be registered with priority (`PriorityTaggedServiceTrait`), so a voter with higher priority can overrule other voters.
In [Contao CMS](https://github.com/contao/contao), the core system should provide security voters that provide the "default permissions", but extensions/bundles can override almost anything and therefore need to be able to override the core decision. None of the existing strategies allow for something like that.
/ping @chalasr @Toflar @leofeyer @ausi
#SymfonyHackday
Commits
-------
0b8028a0ec Added access decision strategy to respect voter priority
* 5.0:
[DI] Service locators can't be decorated
[HttpClient] force HTTP/1.1 when NTLM auth is used
[Validation][FrameworkBundle] Allow EnableAutoMapping to work without auto-mapping namespaces
[Console][SymfonyQuestionHelper] Handle multibytes question choices keys and custom prompt
[DI] fix auto-binding service providers to their service subscribers
[Mailer] fixed undefined index when sending mail
* 4.4:
[DI] Service locators can't be decorated
[HttpClient] force HTTP/1.1 when NTLM auth is used
[Validation][FrameworkBundle] Allow EnableAutoMapping to work without auto-mapping namespaces
[Console][SymfonyQuestionHelper] Handle multibytes question choices keys and custom prompt
[DI] fix auto-binding service providers to their service subscribers
[Mailer] fixed undefined index when sending mail
* 4.3:
[DI] Service locators can't be decorated
[Console][SymfonyQuestionHelper] Handle multibytes question choices keys and custom prompt
[DI] fix auto-binding service providers to their service subscribers
This PR was merged into the 4.3 branch.
Discussion
----------
[DI] fix auto-binding service providers to their service subscribers
| Q | A
| ------------- | ---
| Branch? | 4.3
| Bug fix? | yes
| New feature? | no
| Deprecations? | no
| Tickets | Fix https://github.com/orgs/symfony/projects/1#card-30503621
| License | MIT
| Doc PR | -
Spotted during a workshop at SymfonyLive Sao Paulo if I recall well :)
Commits
-------
6c2ceb0c44 [DI] fix auto-binding service providers to their service subscribers
This PR was merged into the 4.4 branch.
Discussion
----------
[Mailer] Fixed undefined index when sending via Mandrill API
| Q | A
| ------------- | ---
| Branch? | 4.4
| Bug fix? | yes
| New feature? | no
| Deprecations? | no
| Tickets | Fix#34892
| License | MIT
When a [Mandrill API send request](https://www.mandrillapp.com/api/docs/messages.JSON.html#method=send) is succesful, it returns an array of results, one for each recipient. To get rid of the undefined index error, we grab the message ID from the first recipient in the array.
Commits
-------
ef0aa4dc87 [Mailer] fixed undefined index when sending mail
This PR was squashed before being merged into the 4.3 branch.
Discussion
----------
[DI] Service locators can't be decorated
| Q | A
| ------------- | ---
| Branch? | 4.3
| Bug fix? | yes
| New feature? | no
| Deprecations? | no
| Tickets | n/a
| License | MIT
| Doc PR | n/a
This popped up while I was trying to update my work project as we have decorated the `messenger.receiver_locator` service. Not sure if this is a regression in DI or a change in messenger that caused the issue thus I'm not marking this as a BC break.
Exception while trying to compile the container:
```
Invalid definition for service "Symfony\Component\DependencyInjection\Tests\Compiler\DecoratedServiceLocator": an array of references is expected as first argument when the "container.service_locator" tag is set.
```
Expected result: service locator can be decorated.
Commits
-------
343282b9d4 [DI] Service locators can't be decorated
This PR was merged into the 4.4 branch.
Discussion
----------
[HttpClient] force HTTP/1.1 when NTLM auth is used
| Q | A
| ------------- | ---
| Branch? | 4.4
| Bug fix? | yes
| New feature? | no
| Deprecations? | no
| Tickets | -
| License | MIT
| Doc PR | -
NTLM requires HTTP/1.1.
Commits
-------
0e87e9b960 [HttpClient] force HTTP/1.1 when NTLM auth is used
This PR was merged into the 5.1-dev branch.
Discussion
----------
[FrameworkBundle] Allow using the kernel as a registry of controllers and service factories
| Q | A
| ------------- | ---
| Branch? | master
| Bug fix? | no
| New feature? | yes
| Deprecations? | no
| Tickets | Fix#28992, fix#29997
| License | MIT
| Doc PR | -
This PR builds on #34873 and #34872 and allows using the `Kernel` as a registry of autowired controllers and service factories. The `ContainerConfigurator` passed to `configureContainer()` defaults to declaring autowired and autoconfigured services.
TL;DR: Silex is back but in a much more powerful way \o/
Here is a Kernel that just works and displays `Hello App\Foo` on the `/` route:
```php
class Kernel extends BaseKernel
{
use MicroKernelTrait;
protected function configureContainer(ContainerConfigurator $container): void
{
$container->services()
->load('App\\', '../src')
->set(Foo::class)
->factory([$this, 'createFoo']);
}
public function createFoo(Bar $bar)
{
return new Foo($bar);
}
protected function configureRoutes(RoutingConfigurator $routes): void
{
$routes->add('home', '/')->controller([$this, 'helloAction']);
}
public function helloAction(Foo $foo)
{
return new Response('Hello '.get_class($foo));
}
}
```
Commits
-------
9c9b99cc65 [FrameworkBundle] Allow using the kernel as a registry of controllers and service factories
This PR was merged into the 5.1-dev branch.
Discussion
----------
[EventDispatcher] Deprecate LegacyEventDispatcherProxy
| Q | A
| ------------- | ---
| Branch? | master
| Bug fix? | no
| New feature? | no
| Deprecations? | yes
| Tickets | Cleanup of #28920
| License | MIT
| Doc PR | N/A
> This class should be deprecated in Symfony 5.1
Well, here you go. 😃
Commits
-------
c7e612d4ad [EventDispatcher] Deprecate LegacyEventDispatcherProxy.
This PR was merged into the 4.3 branch.
Discussion
----------
[Validator] skip not applicable tests
| Q | A
| ------------- | ---
| Branch? | 4.3
| Bug fix? | no
| New feature? | no
| Deprecations? | no
| Tickets |
| License | MIT
| Doc PR |
Commits
-------
227c38180f skip not applicable tests
This PR was merged into the 3.4 branch.
Discussion
----------
Use `::class` constants instead of `__NAMESPACE__` when possible
| Q | A
| ------------- | ---
| Branch? | 3.4
| Bug fix? | no
| New feature? | no
| Deprecations? | no
| Tickets | Related to #34987
| License | MIT
| Doc PR | no
Form component has a lot of built-in form types. Some of them were implemented from the very beginning. In most of them there is a such method
```php
/**
* {@inheritdoc}
*/
public function getParent()
{
return __NAMESPACE__.'\TextType';
}
```
This `getParent()` method was refactored in Symfony 2.8. The upgrade instructions are given here https://github.com/symfony/symfony/blob/2.8/UPGRADE-2.8.md#form
I think the `__NAMESPACE__.'\TextType';` expression was used because Symfony 2.8 was using `"php": ">=5.3.9"`, and the constant `::class` was added only in PHP 5.5
Now this line can be refactored into
```php
/**
* {@inheritdoc}
*/
public function getParent()
{
return TextType::class;
}
```
For example new form types, that were added later, already using the `::class` constant.
https://github.com/symfony/symfony/blob/master/src/Symfony/Component/Form/Extension/Core/Type/ColorType.php#L23https://github.com/symfony/symfony/blob/master/src/Symfony/Component/Form/Extension/Core/Type/TelType.php#L23
So, in this pull request I propose to refactor all old form types to use `::class` constant. It will give a benefit during the future refactoring, because IDE or static analysers will find all usages of parent class. Unlike the `__NAMESPACE__.'\TextType';` line, which doesn't show the real link to the class for IDE or static analysers, and it could complicate finding all usages of parent class.
Commits
-------
32bf50abca Use `::class` constants instead of `__NAMESPACE__` when possible
* 5.0: (21 commits)
fix merge
CS
[FrameworkBundle][ContainerLintCommand] Improve messages when the kernel or the container is not supported
[Serializer] Skip uninitialized (PHP 7.4) properties in PropertyNormalizer and ObjectNormalizer
stop using deprecated Doctrine persistence classes
[Cache] Fix wrong classname in deprecation message
Fix regex lookahead syntax in ApplicationTest
Fixed syntax in comment
[SecurityBundle][FirewallMap] Remove unused property
[Messenger][AMQP] Use delivery_mode=2 by default
[FrameworkBundle][DependencyInjection] Skip removed ids in the lint container command and its associated pass
[SECURITY] Revert "AbstractAuthenticationListener.php error instead info. Rebase of #28462"
[FrameworkBundle][Secrets] Hook configured local dotenv file
[DI] Improve performance of processDefinition
fix redis multi host dsn not recognized
fix constructor argument type declaration
Fix invalid Windows path normalization
[Validator][ConstraintValidator] Safe fail on invalid timezones
[DoctrineBridge] Fixed submitting invalid ids when using queries with limit
[FrameworkBundle] Add info & example to auto_mapping config
...
* 4.4: (21 commits)
fix merge
CS
[FrameworkBundle][ContainerLintCommand] Improve messages when the kernel or the container is not supported
[Serializer] Skip uninitialized (PHP 7.4) properties in PropertyNormalizer and ObjectNormalizer
stop using deprecated Doctrine persistence classes
[Cache] Fix wrong classname in deprecation message
Fix regex lookahead syntax in ApplicationTest
Fixed syntax in comment
[SecurityBundle][FirewallMap] Remove unused property
[Messenger][AMQP] Use delivery_mode=2 by default
[FrameworkBundle][DependencyInjection] Skip removed ids in the lint container command and its associated pass
[SECURITY] Revert "AbstractAuthenticationListener.php error instead info. Rebase of #28462"
[FrameworkBundle][Secrets] Hook configured local dotenv file
[DI] Improve performance of processDefinition
fix redis multi host dsn not recognized
fix constructor argument type declaration
Fix invalid Windows path normalization
[Validator][ConstraintValidator] Safe fail on invalid timezones
[DoctrineBridge] Fixed submitting invalid ids when using queries with limit
[FrameworkBundle] Add info & example to auto_mapping config
...
* 4.3:
fix merge
CS
[Serializer] Skip uninitialized (PHP 7.4) properties in PropertyNormalizer and ObjectNormalizer
stop using deprecated Doctrine persistence classes
[Cache] Fix wrong classname in deprecation message
Fix regex lookahead syntax in ApplicationTest
Fixed syntax in comment
[SecurityBundle][FirewallMap] Remove unused property
[Messenger][AMQP] Use delivery_mode=2 by default
[DI] Improve performance of processDefinition
Fix invalid Windows path normalization
[Validator][ConstraintValidator] Safe fail on invalid timezones
[DoctrineBridge] Fixed submitting invalid ids when using queries with limit
[FrameworkBundle] Add info & example to auto_mapping config
fix comparisons with null values at property paths
This PR was merged into the 3.4 branch.
Discussion
----------
CS for AccessDecisionManager
| Q | A
| ------------- | ---
| Branch? | 3.4
| Bug fix? | no
| New feature? | no
| Deprecations? | no
| Tickets | #34548
| License | MIT
| Doc PR | -
As discussed in #34548 with @nicolas-grekas here's a CS change for the `AccessDecisionManager`
Commits
-------
b3742ec493 CS
This PR was squashed before being merged into the 3.4 branch (closes#34791).
Discussion
----------
[Serializer] Skip uninitialized (PHP 7.4) properties in PropertyNormalizer and ObjectNormalizer
| Q | A
| ------------- | ---
| Branch? | 3.4
| Bug fix? | yes
| New feature? | no
| Deprecations? | no
| Tickets | n/a
| License | MIT
| Doc PR | n/a
When trying to read from an uninitialized property in PHP 7.4, a `TypeError` is generated, see https://wiki.php.net/rfc/typed_properties_v2#uninitialized_and_unset_properties. This PR fixes the issue.
Commits
-------
1ed8e42d15 [Serializer] Skip uninitialized (PHP 7.4) properties in PropertyNormalizer and ObjectNormalizer
This PR was merged into the 4.3 branch.
Discussion
----------
[Messenger][AMQP] Use delivery_mode=2 by default
| Q | A
| ------------- | ---
| Branch? | master
| Bug fix? | no
| New feature? | no
| Deprecations? | no
| Tickets | Fix#34685
| License | MIT
| Doc PR |
Commits
-------
be2eb6fcc7 [Messenger][AMQP] Use delivery_mode=2 by default
This PR was merged into the 3.4 branch.
Discussion
----------
[DI] Improve performance of processDefinition
| Q | A
| ------------- | ---
| Branch? | 3.4
| Bug fix? | kind of
| New feature? | no
| Deprecations? | no
| Tickets | no
| License | MIT
| Doc PR |
Saves some time during container compilation by instantiating the reflection class only once. In my case this speeds up container compilation in dev mode by ~10% (saves almost 100k calls to `getReflectionClass`).
Tests still run locally and my compiled container was identical pre and post change, but I found this improvement by Blackfire profiling and am not familiar with the surrounding code, so it would be great if someone could doublecheck if the change causes problems.
Commits
-------
41b56eac29 [DI] Improve performance of processDefinition
This PR was merged into the 5.1-dev branch.
Discussion
----------
[FrameworkBundle] Added flex-compatible default implementations for `MicroKernelTrait::registerBundles()` and `getProjectDir()`
| Q | A
| ------------- | ---
| Branch? | master
| Bug fix? | no
| New feature? | yes
| Deprecations? | no
| Tickets | -
| License | MIT
| Doc PR | -
I think it's safe now to migrate some flex-based defaults from the recipe to the bundle.
No BC break here as all existing kernels already redefine this method.
Commits
-------
a689807387 [FrameworkBundle] Added flex-compatible default implementations for `MicroKernelTrait::registerBundles()` and `getProjectDir()`
This PR was merged into the 3.4 branch.
Discussion
----------
[Validator][ConstraintValidator] Safe fail on invalid timezones
Co-authored-by: Scott Dawson <scott@loyaltycorp.com.au>
| Q | A
| ------------- | ---
| Branch? | 3.4
| Bug fix? | yes
| New feature? | no
| Deprecations? | no
| Tickets | https://github.com/symfony/symfony/issues/33901
| License | MIT
| Doc PR |
Alternative to https://github.com/symfony/symfony/pull/33902.
I will explain why I think it is better this way:
1. We set the timezone with the setter because it's 100% safe, it never fails. It fall backs to the default timezone if the provided timezone is not supported (as if we passed null, so the same behavior that always existed). We are therefore compatible with all edge cases.
2. We don't validate the timezone with `\DateTimeZone::listIdentifiers()`. It only returns full identifiers like "Europe/Paris" but it doesn't take into account "numeric" identifiers such as "+08:00" which are perfectly valid. I added a test case to ensure we stay valid with this case. + some invalid identifiers for the native `\IntlDateFormatter` are valid with the polyfill that uses `\DateTimeZone` (eg : `X`). I don't think we can validate anything safely that will work reliably on both implementations.
Commits
-------
3b1b994cb3 [Validator][ConstraintValidator] Safe fail on invalid timezones
This PR was merged into the 4.4 branch.
Discussion
----------
[FrameworkBundle][DependencyInjection] Skip removed ids in the lint container command and its associated pass
| Q | A
| ------------- | ---
| Branch? | 4.4
| Bug fix? | yes
| New feature? | no
| Deprecations? | no
| Tickets | https://github.com/symfony/symfony/issues/34858
| License | MIT
| Doc PR | -
We remove the "removing" passes again and to avoid what https://github.com/symfony/symfony/pull/34502 fixed, we skip validating the "live" container removed ids in the pass (the "live" container is supposed to have the same definitions than the "debug container" one).
Logically, an errored service cannot pass the "live" container compilation without being removed. Consequently, it also skips the errored services that ended up being removed in the "live" container.
Commits
-------
a0f581ba9d [FrameworkBundle][DependencyInjection] Skip removed ids in the lint container command and its associated pass
This PR was squashed before being merged into the 5.1-dev branch.
Discussion
----------
[DI] Add support for defining method calls in InlineServiceConfigurator
| Q | A
| ------------- | ---
| Branch? | master
| Bug fix? | no
| New feature? | yes
| Deprecations? | no
| Tickets |
| License | MIT
| Doc PR | N/A
Commits
-------
5af6e21529 [DI] Add support for defining method calls in InlineServiceConfigurator
This PR was merged into the 4.4 branch.
Discussion
----------
[HttpFoundation] fix redis multi host dsn not recognized
| Q | A
| ------------- | ---
| Branch? | 4.4
| Bug fix? | yes
| New feature? | no
| Deprecations? | no
| Tickets | Fix#34477
| License | MIT
#34177 added support configurating session handlers with DSNs. It was no possible to pass a redis-DSN like
`redis:?host[localhost]&host[localhost:6379]&host[/var/run/redis.sock:]&auth=my-password&redis_cluster=1'`
since the check was
`case 0 === strpos($connection, 'redis://'):`
Commits
-------
81ba07aa26 fix redis multi host dsn not recognized
When a Mandrill API request is succesful, it returns an array of
results, one for each recipient. To get rid of the undefined index
error, we grab the message ID from the first recipient in the array.
* 5.0:
Fix merge
[DoctrineBridge] try to fix deprecations from doctrine/persistence
[DI] Add support for immutable setters in CallTrait
[Cache] Propagate expiry when syncing items in ChainAdapter
Removed request header "Content-Type" from the preferred format guessing mechanism
[Routing] fix memoryleak when loading compiled routes
[Translation] fix memoryleak in PhpFileLoader
fix triggering deprecation in file locator
bug #34877 [TwigBundle] fix findTemplate() to return `null`
* 4.4:
Fix merge
[DoctrineBridge] try to fix deprecations from doctrine/persistence
[DI] Add support for immutable setters in CallTrait
[Cache] Propagate expiry when syncing items in ChainAdapter
Removed request header "Content-Type" from the preferred format guessing mechanism
[Routing] fix memoryleak when loading compiled routes
[Translation] fix memoryleak in PhpFileLoader
fix triggering deprecation in file locator
bug #34877 [TwigBundle] fix findTemplate() to return `null`
* 4.3:
Fix merge
[DoctrineBridge] try to fix deprecations from doctrine/persistence
[DI] Add support for immutable setters in CallTrait
[Cache] Propagate expiry when syncing items in ChainAdapter
[Routing] fix memoryleak when loading compiled routes
[Translation] fix memoryleak in PhpFileLoader
This PR was merged into the 4.3 branch.
Discussion
----------
[DI] Fix support for immutable setters in CallTrait
| Q | A
| ------------- | ---
| Branch? | master
| Bug fix? | no
| New feature? | yes
| Deprecations? | no
| Tickets |
| License | MIT
| Doc PR | N/A
Commits
-------
90ace51328 [DI] Add support for immutable setters in CallTrait
This PR was merged into the 4.4 branch.
Discussion
----------
[HttpFoundation] Removed "Content-Type" from the preferred format guessing mechanism
| Q | A
| ------------- | ---
| Branch? | 4.4
| Bug fix? | yes
| New feature? | no
| Deprecations? | no
| Tickets | Fix#34906, Fix#34857
| License | MIT
| Doc PR | -
Confirmed, inferring the `Content-Type` of the response using the `Content-Type` provided for the request body is NOT a good idea. The HTTP RFC explicitly states that `Accept` must be used to hint a preferred response format (`Content-Type` on the request indicates the type of associated its the body).
Use `Accept` if provided (a best practice anyway), and fallback to the default value (HTML by default) otherwise.
Commits
-------
776523e56a Removed request header "Content-Type" from the preferred format guessing mechanism
This PR was merged into the 3.4 branch.
Discussion
----------
[Translation] fix memoryleak in PhpFileLoader
| Q | A
| ------------- | ---
| Branch? | 3.4
| Bug fix? | yes
| New feature? | no
| Deprecations? | no
| Tickets | -
| License | MIT
| Doc PR | -
This happens when running the test suite with opcache disabled (as it is the case by default since `opcache.enable_cli=0`).
Doing this "require" in a loop (for each test case) compiles the file at each iteration and doesn't reclaim memory (there is no garbage collector for opcodes).
Commits
-------
5c9e3bac96 [Translation] fix memoryleak in PhpFileLoader
This PR was merged into the 4.3 branch.
Discussion
----------
[Routing] fix memoryleak when loading compiled routes
| Q | A
| ------------- | ---
| Branch? | 4.3
| Bug fix? | yes
| New feature? | no
| Deprecations? | no
| Tickets | -
| License | MIT
| Doc PR | -
Same as #34918 for Routing. That's the last significant memory hog I've identified.
Commits
-------
85371a174e [Routing] fix memoryleak when loading compiled routes
* 5.0:
[Validator] Fix auto-mapping constraints should not be validated
[Debug] Updated the README to deprecate the component
[Cache] fix memory leak when using PhpFilesAdapter
[Yaml] Implement multiline string as scalar block for tagged values
[HttpFoundation] Use `Cache-Control: must-revalidate` only if explicit lifetime has been given
[FrameworkBundle] Use UserInterface to @return in getUser method
[CI] Replace php7.4snapshot with php7.4 in Travis configuration
[ExpressionLanguage][Node][BinaryNode] Process division by zero
Fixing bad order of operations with null coalescing operator
forward caught exception
[Validator][ConstraintValidator] Stop passing unnecessary timezone argument to \DateTime
add tags before processing them
[FrameworkBundle][ContainerLintCommand] Reinitialize bundles when the container is reprepared
[Process] change the syntax of portable prepared command lines
[MonologBridge] Fix debug processor datetime type
* 4.4:
[Validator] Fix auto-mapping constraints should not be validated
[Debug] Updated the README to deprecate the component
[Cache] fix memory leak when using PhpFilesAdapter
[Yaml] Implement multiline string as scalar block for tagged values
[HttpFoundation] Use `Cache-Control: must-revalidate` only if explicit lifetime has been given
[FrameworkBundle] Use UserInterface to @return in getUser method
[CI] Replace php7.4snapshot with php7.4 in Travis configuration
[ExpressionLanguage][Node][BinaryNode] Process division by zero
Fixing bad order of operations with null coalescing operator
forward caught exception
[Validator][ConstraintValidator] Stop passing unnecessary timezone argument to \DateTime
add tags before processing them
[FrameworkBundle][ContainerLintCommand] Reinitialize bundles when the container is reprepared
[Process] change the syntax of portable prepared command lines
[MonologBridge] Fix debug processor datetime type
This PR was merged into the 4.4 branch.
Discussion
----------
[Debug] Updated the README to deprecate the component
| Q | A
| ------------- | ---
| Branch? | 4.4
| Bug fix? | no
| New feature? | no <!-- please update src/**/CHANGELOG.md files -->
| Deprecations? | no <!-- please update UPGRADE-*.md and src/**/CHANGELOG.md files -->
| Tickets | https://github.com/symfony/symfony-docs/issues/12775
| License | MIT
| Doc PR | -
Commits
-------
ab9e0eb2f5 [Debug] Updated the README to deprecate the component
This PR was squashed before being merged into the 4.4 branch.
Discussion
----------
[Validator] Fix auto-mapping constraints should not be validated
| Q | A
| ------------- | ---
| Branch? | 4.4 <!-- see below -->
| Bug fix? | yes
| New feature? | no <!-- please update src/**/CHANGELOG.md files -->
| Deprecations? | no <!-- please update UPGRADE-*.md and src/**/CHANGELOG.md files -->
| Tickets | #34672 <!-- prefix each issue number with "Fix #", if any -->
| License | MIT
| Doc PR | N/A
As for `Traverse`, I don't think we should add these "constraints" to the list.
I'm also wondering if it really makes sense to have these annotations as constraints. I think it should rather behave like the `GroupSequence` annotation to add the info the generic metadata at loading time, but we don't need to rely on the constraints behavior at all.
Commits
-------
bc53e4bca0 [Validator] Fix auto-mapping constraints should not be validated
This PR was merged into the 4.4 branch.
Discussion
----------
[Process] change the syntax of portable command lines
| Q | A
| ------------- | ---
| Branch? | 4.4
| Bug fix? | yes
| New feature? | no
| Deprecations? | no
| Tickets | Fix#34838
| License | MIT
| Doc PR | https://github.com/symfony/symfony-docs/pull/12772
An alternative to #34845
Right now, portable command lines use `"$FOO"` for placeholders.
But because we validate that a corresponding variable exists before running the command, this fails with `Command line is missing a value for key "$FOO"` when `FOO` is not defined.
This PR proposes to use `"${:FOO}"` instead. The difference with the previous syntax is that this cannot collide with existing shell scripts as it is invalid for them.
When this is merged, we'll have to update https://symfony.com/blog/new-in-symfony-4-1-prepared-commands too.
Commits
-------
3c7b775b3e [Process] change the syntax of portable prepared command lines
* 4.3:
[Cache] fix memory leak when using PhpFilesAdapter
[Yaml] Implement multiline string as scalar block for tagged values
[HttpFoundation] Use `Cache-Control: must-revalidate` only if explicit lifetime has been given
[FrameworkBundle] Use UserInterface to @return in getUser method
[CI] Replace php7.4snapshot with php7.4 in Travis configuration
[ExpressionLanguage][Node][BinaryNode] Process division by zero
forward caught exception
[Validator][ConstraintValidator] Stop passing unnecessary timezone argument to \DateTime
add tags before processing them
[MonologBridge] Fix debug processor datetime type
* 3.4:
[Yaml] Implement multiline string as scalar block for tagged values
[HttpFoundation] Use `Cache-Control: must-revalidate` only if explicit lifetime has been given
[FrameworkBundle] Use UserInterface to @return in getUser method
[CI] Replace php7.4snapshot with php7.4 in Travis configuration
[ExpressionLanguage][Node][BinaryNode] Process division by zero
forward caught exception
[Validator][ConstraintValidator] Stop passing unnecessary timezone argument to \DateTime
add tags before processing them
[MonologBridge] Fix debug processor datetime type
This PR was merged into the 4.3 branch.
Discussion
----------
[Cache] fix memory leak when using PhpFilesAdapter
| Q | A
| ------------- | ---
| Branch? | 4.3
| Bug fix? | yes
| New feature? | no
| Deprecations? | no
| Tickets | Fix#34687
| License | MIT
| Doc PR | -
Similar to #34839 but for `PhpFilesAdapter`, as the "appendOnly" mode is a v4-only feature.
Commits
-------
0b46226648 [Cache] fix memory leak when using PhpFilesAdapter
This PR was merged into the 3.4 branch.
Discussion
----------
[HttpFoundation] Use `Cache-Control: must-revalidate` only if explicit lifetime has been given
| Q | A
| ------------- | ---
| Branch? | 3.4
| Bug fix? | yes
| New feature? | no
| Deprecations? | no
| Tickets |
| License | MIT
| Doc PR |
This is really nit-picking: The conservative, safe default for `Cache-Control` is `private, no-cache` which means the response must not be served from cache unless it has been validated.
If `Last-Modified` or `Expires` are present, we can relax `no-cache` to be `must-revalidate`, which means that _once the response has become stale_, it must be revalidated.
An `ETag` alone does not give the response a lifetime, so IMO sticking with `no-cache` in this case would be more consistent.
Commits
-------
1b1002b426 [HttpFoundation] Use `Cache-Control: must-revalidate` only if explicit lifetime has been given