This PR was squashed before being merged into the 3.2-dev branch (closes#18533).
Discussion
----------
[FrameworkBundle] Wire PhpArrayAdapter with a new cache warmer for annotations
| Q | A
| ------------- | ---
| Branch? | master
| Bug fix? | no
| New feature? | yes
| BC breaks? | no
| Deprecations? | no
| Tests pass? | WIP
| Fixed tickets | -
| License | MIT
| Doc PR | -
Depends on https://github.com/symfony/symfony/pull/18825 and https://github.com/symfony/symfony/pull/18823
This PR implements the usage of the new OpCacheAdapter in the annotations caching system. The idea to use this adapter as much as possible in Symfony (validator, serializer, ...). These other implementations will be the object of different PRs.
Commits
-------
f950a2b [FrameworkBundle] Wire PhpArrayAdapter with a new cache warmer for annotations
This PR was merged into the 3.2-dev branch.
Discussion
----------
[Filesystem] Add a cross-platform readlink method
| Q | A
| ------------- | ---
| Bug fix? | no
| New feature? | yes
| BC breaks? | no
| Deprecations? | no
| Tests pass? | yes
| Fixed tickets | -
| License | MIT
| Doc PR | -
`readlink()` and `realpath()` have a completely different behavior under Windows and Unix:
- `realpath()` resolves recursively the children links of a link until a final target is found on Unix and resolves only the next link on Windows ;
- `readlink()` resolves recursively the children links of a link until a final target is found on Windows and resolves only the next link on Unix ;
I propose to solve this by implementing a helper method in the Filesystem component that would behave always the same way under all platforms.
Commits
-------
c36507e [Filesystem] Add a cross-platform readlink/realpath methods for nested links
This PR was squashed before being merged into the 2.7 branch (closes#19446).
Discussion
----------
[Console] Overcomplete argument exception message tweak.
| Q | A
| ------------- | ---
| Branch? | 2.7
| Bug fix? | no
| New feature? | no
| BC breaks? | no
| Deprecations? | no
| Tests pass? | yes
| License | MIT
Updates the exception message when to many arguments are passed.
From;
```php
'Too many arguments.'
```
To:
```php
'No argument expected, got "foo".'
// or
'Too many arguments, expected arguments "foo".'
// or
'Too many arguments, expected arguments "foo, bar".'
// ... turtles all the way down
```
Commits
-------
7af59cd [Console] Overcomplete argument exception message tweak.
* 3.1:
[TwigBundle] Removed redundant return statement.
enable property info
[Cache] Fix default lifetime being ignored
[DependencyInjection] Fixed deprecated default message template with XML
Reference the actual location of the documentation
[TwigBridge] Removed extra arguments in 2 places.
[Cache] Fix incorrect timestamps generated by FilesystemAdapter
[Process] Fix write access check for pipes on Windows
[HttpKernel] Use flock() for HttpCache's lock files
Conflicts:
src/Symfony/Component/Cache/Adapter/FilesystemAdapter.php
* 3.0:
[TwigBundle] Removed redundant return statement.
[DependencyInjection] Fixed deprecated default message template with XML
[TwigBridge] Removed extra arguments in 2 places.
[Process] Fix write access check for pipes on Windows
[HttpKernel] Use flock() for HttpCache's lock files
* 2.8:
[TwigBundle] Removed redundant return statement.
[DependencyInjection] Fixed deprecated default message template with XML
[TwigBridge] Removed extra arguments in 2 places.
[Process] Fix write access check for pipes on Windows
[HttpKernel] Use flock() for HttpCache's lock files
* 2.7:
[TwigBundle] Removed redundant return statement.
[TwigBridge] Removed extra arguments in 2 places.
[Process] Fix write access check for pipes on Windows
[HttpKernel] Use flock() for HttpCache's lock files
This PR was merged into the 3.2-dev branch.
Discussion
----------
[VarDumper] Dumping exceptions is now more compact
| Q | A
| ------------- | ---
| Branch? | master
| Bug fix? | no
| New feature? | yes
| BC breaks? | no
| Deprecations? | no
| Tests pass? | no
| Fixed tickets | -
| License | MIT
| Doc PR | -
Before:
![before](https://cloud.githubusercontent.com/assets/243674/16578686/2c7285b0-429c-11e6-9139-293e0c899d43.png)
Commits
-------
19e9cbe [VarDumper] Dumping exceptions is now more compact
This PR was merged into the 2.7 branch.
Discussion
----------
[HttpKernel] Use flock() for HttpCache's lock files
| Q | A
| ------------- | ---
| Branch? | 2.7
| Bug fix? | yes
| New feature? | no
| BC breaks? | no
| Deprecations? | no
| Tests pass? | no
| Fixed tickets | #16777, #15813 and #16312 are also related
| License | MIT
| Doc PR |
When a PHP process crashes or terminates (maybe the OOM killer kicks in or other bad things ™️ happen) while the `HttpCache` holds a `.lck` file, that lock file may not get `unlink()`ed.
The result is that other requests trying to access this cache entry will see a few seconds delay while waiting for the lock; they will eventually continue but send 503 status codes along with the response. The sudden buildup of PHP processes caused by the additional delay may cause further problems (sudden load increase).
As `LockHandler` is using `flock()`-based locking, locks should be released by the OS when the PHP process terminates.
I wrote this as bugfix against 2.7 because every once in a while I encounter situations (not always reproducible) where `.lock` files are left over and keep the cache locked.
Commits
-------
2668edd [HttpKernel] Use flock() for HttpCache's lock files
This PR was merged into the 2.7 branch.
Discussion
----------
[Process] Fix write access check for pipes on Windows
| Q | A
| ------------- | ---
| Branch? | 2.7
| Bug fix? | yes
| New feature? | no
| BC breaks? | no
| Deprecations? | no
| Tests pass? | yes
| Fixed tickets | #19336, #19416
| License | MIT
| Doc PR | -
Commits
-------
66e694e [Process] Fix write access check for pipes on Windows
* 3.1:
[Console] Application update PHPDoc of add and register methods
[Config] Extra tests for Config component
Fixed bugs in names of classes and methods.
[DoctrineBridge] Fixed php doc
[FrameworkBundle] Fixed parameters number mismatch declaration
[BrowserKit] Added test for followRedirect method (POST method)
Fix the money form type render with Bootstrap3
[BrowserKit] Uppercase the "GET" method in redirects
[DomCrawler] Inherit the namespace cache in subcrawlers
[WebProfilerBundle] Fixed JSDoc parameter definition
[HttpFoundation] HttpCache refresh stale responses containing an ETag
Conflicts:
src/Symfony/Component/Console/Application.php
* 3.0:
[Console] Application update PHPDoc of add and register methods
[Config] Extra tests for Config component
Fixed bugs in names of classes and methods.
[DoctrineBridge] Fixed php doc
[FrameworkBundle] Fixed parameters number mismatch declaration
[BrowserKit] Added test for followRedirect method (POST method)
Fix the money form type render with Bootstrap3
[BrowserKit] Uppercase the "GET" method in redirects
[DomCrawler] Inherit the namespace cache in subcrawlers
[WebProfilerBundle] Fixed JSDoc parameter definition
[HttpFoundation] HttpCache refresh stale responses containing an ETag
* 2.8:
[Console] Application update PHPDoc of add and register methods
[Config] Extra tests for Config component
Fixed bugs in names of classes and methods.
[DoctrineBridge] Fixed php doc
[FrameworkBundle] Fixed parameters number mismatch declaration
[BrowserKit] Added test for followRedirect method (POST method)
Fix the money form type render with Bootstrap3
[BrowserKit] Uppercase the "GET" method in redirects
[DomCrawler] Inherit the namespace cache in subcrawlers
[WebProfilerBundle] Fixed JSDoc parameter definition
[HttpFoundation] HttpCache refresh stale responses containing an ETag
Conflicts:
src/Symfony/Component/Finder/Tests/Shell/CommandTest.php
* 2.7:
[Console] Application update PHPDoc of add and register methods
[Config] Extra tests for Config component
Fixed bugs in names of classes and methods.
[DoctrineBridge] Fixed php doc
[FrameworkBundle] Fixed parameters number mismatch declaration
[BrowserKit] Added test for followRedirect method (POST method)
Fix the money form type render with Bootstrap3
[BrowserKit] Uppercase the "GET" method in redirects
[WebProfilerBundle] Fixed JSDoc parameter definition
[HttpFoundation] HttpCache refresh stale responses containing an ETag
Conflicts:
src/Symfony/Component/BrowserKit/Tests/ClientTest.php
src/Symfony/Component/Security/Acl/Resources/bin/generateSql.php
This PR was squashed before being merged into the 2.7 branch (closes#19389).
Discussion
----------
[Console] Application update PHPDoc of add and register methods
| Q | A
| ------------- | ---
| Branch? | 2.7
| Bug fix? | yes
| New feature? | no
| BC breaks? | no
| Deprecations? | no
| Tests pass? | yes
| License | MIT
The [PHPDoc](https://github.com/symfony/symfony/blob/2.7/src/Symfony/Component/Console/Application.php#L354) states the method will always return a command, but it doesn't. Since [Application::register](https://github.com/symfony/symfony/blob/2.7/src/Symfony/Component/Console/Application.php#L328) returns the result of `add` directly is also doesn't always return the command (as its PHPDoc states).
Commits
-------
6f0474f [Console] Application update PHPDoc of add and register methods
This PR was squashed before being merged into the 2.7 branch (closes#19399).
Discussion
----------
[Config] Extra tests for Config component
| Q | A
| ------------- | ---
| Branch? | 2.7
| Bug fix? | no
| New feature? | no
| BC breaks? | no
| Deprecations? | no
| Tests pass? | yes
| Fixed tickets | ~
| License | MIT
| Doc PR | ~
Commits
-------
d0f4434 [Config] Extra tests for Config component
This PR was squashed before being merged into the 2.7 branch (closes#19392).
Discussion
----------
[BrowserKit] Added test for followRedirect method (POST method)
| Q | A
| ------------- | ---
| Branch? | 2.7
| Bug fix? | no
| New feature? | no
| BC breaks? | no
| Deprecations? | no
| Tests pass? | yes
| Fixed tickets | ~
| License | MIT
| Doc PR | ~
Test covers 'else' branches in these conditions:
```php
if (in_array($this->internalResponse->getStatus(), array(302, 303))) {
$method = 'get';
$files = array();
$content = null;
} else {
$method = $request->getMethod();
$files = $request->getFiles();
$content = $request->getContent();
}
if ('get' === strtolower($method)) {
// Don't forward parameters for GET request as it should reach the redirection URI
$parameters = array();
} else {
$parameters = $request->getParameters();
}
```
Commits
-------
2ace5d8 [BrowserKit] Added test for followRedirect method (POST method)
This PR was merged into the 2.7 branch.
Discussion
----------
[2.7] [HttpFoundation] HttpCache refresh stale responses containing an ETag
| Q | A
| ------------- | ---
| Branch? | 2.7
| Bug fix? | yes
| New feature? |no
| BC breaks? |no
| Deprecations? |no
| Tests pass? | yes
| Fixed tickets | #19390, #6746
| License | MIT
| Doc PR |
This PR is the replacement of #19391, which points at the wrong branch.
Commits
-------
96df6b9 [HttpFoundation] HttpCache refresh stale responses containing an ETag
This PR was merged into the 2.8 branch.
Discussion
----------
[DomCrawler] Inherit the namespace cache in subcrawlers
| Q | A
| ------------- | ---
| Branch? | 2.8
| Bug fix? | no
| New feature? | no
| BC breaks? | no
| Deprecations? | no
| Tests pass? | yes
| Fixed tickets | #12298
| License | MIT
| Doc PR | n/a
This inherits any already discovered/registered namespace with subcrawlers, improving performance when using namespaces.
I submitted to 2.8 rather than 2.7, because the namespace mapping feature was actually buggy in 2.x, because of the fact that nodes could belong to different documents in the same Crawler while the namespace map was shared. The fact that the map was not inherited in subcrawler mitigated this issue (by reducing changes to have multiple documents in the same subcrawler). 2.8 deprecated this possibility to have multiple documents, so I'm fine with applying this here.
Note that the subcrawler inherits the namespace cache at the time it is created, but the cache is not shared between instance (so if a subcrawler discovers an additional namespace of the document, it will not be available for the parent crawler of other subcrawlers of the parent). Sharing the cache would be totally possible (as they share the same document anyway) and would make the experience even better (removing the need to ensure that the root crawler discovers namespace before filtering). But it would require moving from an array to an object. I'm not sure we want to do this in a patch release. What do you think @symfony/deciders ?
Commits
-------
e89c758 [DomCrawler] Inherit the namespace cache in subcrawlers
* 3.1:
fixed bad merge
Fix PHP 7.1 related failures
[VarDumper] Fix for 7.1
fixed CS
Added class existence check if is_subclass_of() fails in compiler passes
Fix the DBAL session handler version check for Postgresql
* 3.0:
Fix PHP 7.1 related failures
[VarDumper] Fix for 7.1
fixed CS
Added class existence check if is_subclass_of() fails in compiler passes
Fix the DBAL session handler version check for Postgresql
* 2.8:
Fix PHP 7.1 related failures
[VarDumper] Fix for 7.1
fixed CS
Added class existence check if is_subclass_of() fails in compiler passes
Fix the DBAL session handler version check for Postgresql
This PR was squashed before being merged into the 3.2-dev branch (closes#19276).
Discussion
----------
[ClassLoader] Add ClassCollectionLoader::inline() to generate inlined-classes files
| Q | A
| ------------- | ---
| Branch? | master
| Bug fix? | no
| New feature? | yes
| BC breaks? | no
| Deprecations? | no
| Tests pass? | yes
| Fixed tickets | -
| License | MIT
| Doc PR | -
Unfortunately, can't be tested because the method relies too much on side effects.
Coupled with https://github.com/sensiolabs/SensioDistributionBundle/pull/272, allows inlining `ClassCollectionLoader` itself into the `bootstrap.php.cache` file.
Commits
-------
88fdcea [ClassLoader] Add ClassCollectionLoader::inline() to generate inlined-classes files
This PR was merged into the 2.8 branch.
Discussion
----------
Added class existence check if is_subclass_of() fails in compiler passes
| Q | A
| ------------- | ---
| Branch? | 2.8-3.1
| Bug fix? | no
| New feature? | yes
| BC breaks? | no
| Deprecations? | no
| Tests pass? | yes
| Fixed tickets | -
| License | MIT
| Doc PR | -
If you create an event subscriber and make typo in file name it will cause next error:
```
[InvalidArgumentException]
Service "event.notification_subscriber" must implement interface "Symfony\Component\EventDispatcher\EventSubscriberInterface".
```
That's because of `is_subclass_of()` fails on class absentee. I made error message more clear.
Commits
-------
72db6e7 Added class existence check if is_subclass_of() fails in compiler passes
* 3.1:
[VarDumper] Fix dumping jsons casted as arrays
PassConfig::getMergePass is not an array
Revert "bug #19114 [HttpKernel] Dont close the reponse stream in debug (nicolas-grekas)"
[Serializer] Include the format in the cache key
Fix the retrieval of the last username when using forwarding
[Yaml] Fix PHPDoc of the Yaml class
[HttpFoundation] Add OPTIONS and TRACE to the list of safe methods
Update getAbsoluteUri() for query string uris
Conflicts:
src/Symfony/Component/DependencyInjection/Compiler/PassConfig.php
src/Symfony/Component/HttpFoundation/Tests/RequestTest.php
* 3.0:
[VarDumper] Fix dumping jsons casted as arrays
PassConfig::getMergePass is not an array
Revert "bug #19114 [HttpKernel] Dont close the reponse stream in debug (nicolas-grekas)"
Fix the retrieval of the last username when using forwarding
[Yaml] Fix PHPDoc of the Yaml class
[HttpFoundation] Add OPTIONS and TRACE to the list of safe methods
Update getAbsoluteUri() for query string uris
Conflicts:
src/Symfony/Component/Yaml/Yaml.php
* 2.8:
[VarDumper] Fix dumping jsons casted as arrays
PassConfig::getMergePass is not an array
Revert "bug #19114 [HttpKernel] Dont close the reponse stream in debug (nicolas-grekas)"
Fix the retrieval of the last username when using forwarding
[Yaml] Fix PHPDoc of the Yaml class
[HttpFoundation] Add OPTIONS and TRACE to the list of safe methods
Update getAbsoluteUri() for query string uris
* 2.7:
[VarDumper] Fix dumping jsons casted as arrays
PassConfig::getMergePass is not an array
Revert "bug #19114 [HttpKernel] Dont close the reponse stream in debug (nicolas-grekas)"
Fix the retrieval of the last username when using forwarding
[Yaml] Fix PHPDoc of the Yaml class
[HttpFoundation] Add OPTIONS and TRACE to the list of safe methods
Update getAbsoluteUri() for query string uris
This PR was squashed before being merged into the 2.7 branch (closes#19343).
Discussion
----------
PassConfig::getMergePass is not an array
| Q | A
| ------------- | ---
| Branch? | 2.7
| Bug fix? | no
| New feature? | no
| BC breaks? | no
| Deprecations? | no
| Tests pass? | yes
| Fixed tickets | -
| License | MIT
| Doc PR | -
Just a minor glitch my IDE noticed :-)
Commits
-------
edbefac PassConfig::getMergePass is not an array
This PR was merged into the 3.2-dev branch.
Discussion
----------
[Serializer] Argument objects
| Q | A
| ------------- | ---
| Branch? | 3.1
| Bug fix? | no
| New feature? | yes
| BC breaks? | no
| Deprecations? | no
| Tests pass? | TODO
| Fixed tickets | none
| License | MIT
| Doc PR | TODO
Assuming with have the two following entities:
```php
namespace AppBundle\Entity;
class Dummy
{
public function __construct(int $id, string $name, string $email, AnotherDummy $anotherDummy)
{
$this->id = $id;
$this->name = $name;
$this->email = $email;
$this->anotherDummy = $anotherDummy;
}
}
class AnotherDummy
{
public function __construct(int $id, string $uuid, bool $isEnabled)
{
$this->id = $id;
$this->uuid = $uuid;
$this->isEnabled = $isEnabled;
}
}
```
Doing the following will fail:
```php
$serializer->denormalize(
[
'id' => $i,
'name' => 'dummy',
'email' => 'du@ex.com',
'another_dummy' => [
'id' => 1000 + $i,
'uuid' => 'azerty',
'is_enabled' => true,
],
],
\AppBundle\Entity\Dummy::class
);
```
with a type error, because the 4th argument passed to `Dummy::__construct()` will be an array. The following patch checks if the type of the argument is an object, and if it is tries to denormalize that object as well.
I'm not sure if it's me missing something or this is a use case that has been omitted (willingly or not), but if it's a valuable patch I would be happy to work on finishing it.
Commits
-------
988eba1 fix tests
98bcb91 Merge pull request #1 from dunglas/theofidry-feature/param-object
7b5d55d Prevent BC in instantiateObject
e437e04 fix reflection type
3fe9802 revert CS
5556fa5 fix
d4cdb00 fix CS
93608dc Add deprecation message
f46a176 Apply patch
f361e52 fix tests
4884a2e f1
e64e999 Address comments
e99a90b Add tests
7bd4ac5 Test
This PR was squashed before being merged into the 2.7 branch (closes#19321).
Discussion
----------
[HttpFoundation] Add OPTIONS and TRACE to the list of safe methods
| Q | A
| ------------- | ---
| Branch? | 2.7
| Bug fix? | yes
| New feature? | no
| BC breaks? | no
| Deprecations? | no
| Tests pass? | yes
| Fixed tickets | n/a
| License | MIT
| Doc PR | n/a
According to [RFC 7231](https://tools.ietf.org/html/rfc7231#section-8.1.3) `OPTIONS` and `TRACE` are safe methods.
Commits
-------
1404607 [HttpFoundation] Add OPTIONS and TRACE to the list of safe methods
This PR was merged into the 2.7 branch.
Discussion
----------
[BrowserKit] Update Client::getAbsoluteUri() for query string only URIs
| Q | A
| ------------- | ---
| Branch? | 2.7
| Bug fix? | yes
| New feature? | no
| BC breaks? | no
| Deprecations? | no
| Tests pass? | yes
| Fixed tickets | #19303
| License | MIT
| Doc PR | -
This PR allows BrowserKit to treat a value containing only query string parameters the same way anchor/hash values are treated when passed as a URI argument to the getAbsoluteUri() method. Helps when encountering sites that force a redirect with a location header value containing only a query string.
Commits
-------
965408f Update getAbsoluteUri() for query string uris
This PR was squashed before being merged into the 3.2-dev branch (closes#19322).
Discussion
----------
[HttpFoundation] Add Request::isMethodIdempotent method
| Q | A
| ------------- | ---
| Branch? | master
| Bug fix? | no
| New feature? | yes
| BC breaks? | no
| Deprecations? | no
| Tests pass? | yes
| Fixed tickets | n/a
| License | MIT
| Doc PR | n/a
Addd a new method in the spirit of `isMethodSafe` to know if the current method is idempotent according to RFCs.
Commits
-------
44df6a4 [HttpFoundation] Add Request::isMethodIdempotent method
* 3.1:
[HttpKernel] fixed internal subrequests having an if-modified-since-header
[Security] Fix deprecated usage of DigestAuthenticationEntryPoint::getKey() in DigestAuthenticationListener
[Validator] Added additional MasterCard range to the CardSchemeValidator
Make the exception message more clear.
[Form] fixed bug - name in ButtonBuilder
[DoctrineBridge] added missing error code for constraint.
[ClassLoader] Fix declared classes being computed when not needed
[varDumper] Fix missing usage of ExceptionCaster::$traceArgs
* 3.0:
[HttpKernel] fixed internal subrequests having an if-modified-since-header
[Security] Fix deprecated usage of DigestAuthenticationEntryPoint::getKey() in DigestAuthenticationListener
[Validator] Added additional MasterCard range to the CardSchemeValidator
Make the exception message more clear.
[Form] fixed bug - name in ButtonBuilder
[DoctrineBridge] added missing error code for constraint.
[ClassLoader] Fix declared classes being computed when not needed
[varDumper] Fix missing usage of ExceptionCaster::$traceArgs
* 2.8:
[HttpKernel] fixed internal subrequests having an if-modified-since-header
[Security] Fix deprecated usage of DigestAuthenticationEntryPoint::getKey() in DigestAuthenticationListener
[Validator] Added additional MasterCard range to the CardSchemeValidator
Make the exception message more clear.
[Form] fixed bug - name in ButtonBuilder
[DoctrineBridge] added missing error code for constraint.
[ClassLoader] Fix declared classes being computed when not needed
[varDumper] Fix missing usage of ExceptionCaster::$traceArgs
Conflicts:
src/Symfony/Bridge/Doctrine/Validator/Constraints/UniqueEntityValidator.php
src/Symfony/Component/ClassLoader/ClassCollectionLoader.php
* 2.7:
[HttpKernel] fixed internal subrequests having an if-modified-since-header
[Validator] Added additional MasterCard range to the CardSchemeValidator
Make the exception message more clear.
[Form] fixed bug - name in ButtonBuilder
[ClassLoader] Fix declared classes being computed when not needed
This PR was merged into the 2.7 branch.
Discussion
----------
[ClassLoader] Fix declared classes being computed when not needed
| Q | A
| ------------- | ---
| Branch? | 2.7
| Bug fix? | yes
| New feature? | no
| BC breaks? | no
| Deprecations? | no
| Tests pass? | no
| Fixed tickets | -
| License | MIT
| Doc PR | -
Commits
-------
d513eae [ClassLoader] Fix declared classes being computed when not needed
This PR was merged into the 2.7 branch.
Discussion
----------
[Validator] Added additional MasterCard range to the CardSchemeValidator
From October 2016 MasterCard will introduce a new card range, 222100 through 272099.
See: https://www.mctestcards.com/ (click the help in top right)
This implements the additional regex for validation to succeed, and some additional unit tests for this new range.
| Q | A
| ------------- | ---
| Branch? | 2.7
| Bug fix? | yes
| New feature? | no
| BC breaks? | no
| Deprecations? | no
| Tests pass? | yes
| Fixed tickets |
| License | MIT
| Doc PR |
Commits
-------
4d68f56 [Validator] Added additional MasterCard range to the CardSchemeValidator
This PR was squashed before being merged into the 2.7 branch (closes#19290).
Discussion
----------
[HttpKernel] fixed internal subrequests having an if-modified-since-header
| Q | A
| ------------- | ---
| Branch? | 2.7
| Bug fix? | yes
| New feature? | no
| BC breaks? | no
| Deprecations? | no
| Tests pass? | yes
| Fixed tickets | -
| License | MIT
| Doc PR | -
As the InlineFragmentRenderer has no access to a cached copy of a subrequest's target and hence couldn't handle a response with a HTTP status code of 304 (not modified), it makes no sense to send an if-not-modified-since header.
Commits
-------
e90038c [HttpKernel] fixed internal subrequests having an if-modified-since-header