This PR was submitted for the master branch but it was merged into the 2.2 branch instead (closes#8081).
Discussion
----------
Use strstr instead of strpos in ClassLoader (4% perf improvement)
| Q | A
| ------------- | ---
| Bug fix? | no
| New feature? | no
| BC breaks? | no
| Deprecations? | no
| Tests pass? | yes
| Fixed tickets | [
| License | MIT
| Doc PR |
Using ClassLoader in Drupal 8, using strstr instead of strpos nets 4% perf improvement.
XHPROF diff https://dl.dropboxusercontent.com/u/10201421/diff.html
Commits
-------
25d7b90 [ClassLoader] Use strstr instead of strpos
This PR was submitted for the master branch but it was merged into the 2.3 branch instead (closes#8688).
Discussion
----------
[Debug] Make sure ContextErrorException is loaded during compile time errors
| Q | A
| ------------- | ---
| Bug fix? | yes
| New feature? | no
| BC breaks? | no
| Deprecations? | no
| Tests pass? | no
| Fixed tickets | n/a
| License | MIT
| Doc PR | n/a
### Description
When the `ErrorHandler->handle()` attempts to throw a `ContextErrorException` the script might die with a Fatal Error because the class is not found.
FatalErrorException: Error: Class 'Symfony\Component\Debug\Exception\ContextErrorException' not found in ...
That effectively hides the original error and causes confusion.
### Cause
The cause is a bug/limitation of PHP that makes class autoloading completely inactive during "compiling".
Relevant PHP source code in `zend_execute_API.c`:
/* The compiler is not-reentrant. Make sure we __autoload() only during run-time
* (doesn't impact functionality of __autoload()
*/
if (!use_autoload || zend_is_compiling(TSRMLS_C)) {
#### More information about the PHP bug/limitation:
- https://bugs.php.net/bug.php?id=65322
- many duplicates/relevant: https://www.google.com/search?btnG=1&pws=0&q=site%3Abugs.php.net+autoload+error+handling
### The fix
The most simple way to fix this is to manually ensure that `ContextErrorException` is loaded and if not, load it. This is what I did.
Another way is to detect if autoloading is available (possible by prepending a temporary autoloader) and act accordingly (how?).
### The test
I also added a test case that would fail with a fatal error if the error handler does not successfully throw an exception in case of a compile time error. But I'm not sure if this is the best way or if there should be a test case at all.
###
Commits
-------
b0082ab [Debug] Make sure ContextErrorException is loaded during compile time errors
This PR was merged into the 2.3 branch.
Discussion
----------
Validators
| Q | A
| ------------- | ---
| Bug fix? | yes
| New feature? | no
| BC breaks? | no
| Deprecations? | no
| Tests pass? | yes
| License | MIT
This PR avoid comparison validators to be executed if the compared value is null.
Commits
-------
48338fc Ignore null value in comparison validators
* 2.2:
added missing support for the new output API in PHP 5.4+
Fixed bug introduced in #8675
made the filesystem loader compatible with Twig 2.0
bumped Symfony version to 2.2.6
updated VERSION for 2.2.5
update CONTRIBUTORS for 2.2.5
updated CHANGELOG for 2.2.5
replaced deprecated Twig features
Conflicts:
src/Symfony/Bridge/Twig/Extension/FormExtension.php
src/Symfony/Bridge/Twig/Extension/RoutingExtension.php
src/Symfony/Component/HttpKernel/Kernel.php
This PR was merged into the 2.2 branch.
Discussion
----------
[2.2][TwigBridge] Fix form_enctype bug
| Q | A
| ------------- | ---
| Bug fix? | yes
| New feature? | no
| BC breaks? | no
| Deprecations? | no
| Tests pass? | yes
| Fixed tickets | -
| License | MIT
| Doc PR | -
After a copy/past error in #8675, `{{ form_enctype() }}` was not working because it results in a `Class "Symfony\Bridge\Twig\Node\FormEnctypeNode" not found` error.
This should only be merged in 2.2, as of 2.3 this is moved to the `Symfony\Bridge\Twig\Node\FormEnctypeNode` class
Commits
-------
e0c7d3d Fixed bug introduced in #8675
This PR was submitted for the master branch but it was merged into the 2.2 branch instead (closes#8677).
Discussion
----------
[TwigBundle] made the filesystem loader compatible with Twig 2.0
| Q | A
| ------------- | ---
| Bug fix? | yes (for Twig 2.x)
| New feature? | no
| BC breaks? | no
| Deprecations? | no
| Tests pass? | yes
| Fixed tickets | n/a
| License | MIT
| Doc PR | n/a
Without adding the exists() method, the code happens to work by chance,
just because the current implementation of Twig exits() method calls
findTemplate().
But we know that it won't be the case anymore as of Twig 2.0.
Commits
-------
d00548b [TwigBundle] made the filesystem loader compatible with Twig 2.0
Without adding the exists() method, the code happens to work by chance,
just because the current implementation of Twig exits() method calls
findTemplate().
But we know that it won't be the case anymore as of Twig 2.0.
This PR was merged into the 2.2 branch.
Discussion
----------
replaced deprecated Twig features
| Q | A
| ------------- | ---
| Bug fix? | no
| New feature? | no
| BC breaks? | no
| Deprecations? | no
| Tests pass? | yes
| Fixed tickets | n/a
| License | MIT
| Doc PR | n/a
As Symfony depends on Twig 1.11+, removing the usage of deprecated features will allow Symfony 2.2+ to work with Twig 2.0.
Commits
-------
322f880 replaced deprecated Twig features
This PR was merged into the 2.3 branch.
Discussion
----------
[Intl] Updated stubs to reflect ICU 51.2
| Q | A
| ------------- | ---
| Bug fix? | no
| New feature? | no
| BC breaks? | no
| Deprecations? | no
| Tests pass? | yes
| Fixed tickets | -
| License | MIT
| Doc PR | -
In addition to the upgrade of the ICU component to ICU 51.2, this PR upgrades the stub classes to behave like their C counterparts do with ICU 51.2.
Commits
-------
8fa0453 [Intl] Updated stubs to reflect ICU 51.2
This PR was merged into the 2.3 branch.
Discussion
----------
[HttpKernel] Added a missing use statement
| Q | A
| ------------- | ---
| Bug fix? | yes
| New feature? | no
| BC breaks? | no
| Deprecations? | no
| Tests pass? | yes
| Fixed tickets | -
| License | MIT
| Doc PR | -
Build was broken by #8572 merged into 2.3 and master.
Commits
-------
ccb497a [HttpKernel] Added a missing use statement.
This PR was merged into the 2.3 branch.
Discussion
----------
[Process] Increased the timeout in a test to prevent random failures
This test fails quite often when travis is under a heavy load.
| Q | A
| ------------- | ---
| Bug fix? | no
| New feature? | no
| BC breaks? | no
| Deprecations? | no
| Tests pass? | yes
| Fixed tickets | -
| License | MIT
| Doc PR | -
Commits
-------
9f3b1e1 [Process] Increased the timeout in a test to prevent random failures when travis is under a heavy load.
* 2.2:
[HttpKernel] added a missing dep for dev
[Form] fixed wrong call to setTimeZone() (closes#8644)
Fix issue with \DateTimeZone::UTC / 'UTC' for PHP 5.4
[Form] Removed the "disabled" attribute from the placeholder option in select fields due to problems with the BlackBerry 10 browser
[routing] added ability for apache matcher to handle array values
removed dead code and fixed CS
[Validator] fixed StaticMethodLoader trying to invoke methods of abstract classes (closes#8589)
Conflicts:
src/Symfony/Bundle/TwigBundle/TokenParser/RenderTokenParser.php
src/Symfony/Component/Form/FormConfigBuilder.php
src/Symfony/Component/HttpKernel/composer.json
src/Symfony/Component/Validator/Tests/GraphWalkerTest.php
This PR was merged into the 2.3 branch.
Discussion
----------
moved some fixed dep versions from 2.2.* to ~2.2
| Q | A
| ------------- | ---
| Bug fix? | yes
| New feature? | no
| BC breaks? | no
| Deprecations? | no
| Tests pass? | yes
| Fixed tickets | #8613
| License | MIT
| Doc PR | -
Commits
-------
e18f3ed moved some fixed dep versions from 2.2.* to ~2.2 (refs #8613)
This PR was merged into the 2.2 branch.
Discussion
----------
Added a missing dev dep for Httpkernel
| Q | A
| ------------- | ---
| Bug fix? | yes
| New feature? | no
| BC breaks? | no
| Deprecations? | no
| Tests pass? | yes
| Fixed tickets | #8613
| License | MIT
| Doc PR | -
Commits
-------
a741414 [HttpKernel] added a missing dep for dev
This PR was merged into the 2.3 branch.
Discussion
----------
[Form] Fixed patched forms to be valid even if children are not submitted
| Q | A
| ------------- | ---
| Bug fix? | yes
| New feature? | no
| BC breaks? | no
| Deprecations? | no
| Tests pass? | yes
| Fixed tickets | #8111
| License | MIT
| Doc PR | -
#8362 was reverted because it introduces a failing test that is caused by a regression. This PR takes the alternative approach that
* unsubmitted fields in the PATCH request remain unsubmitted
* `isValid()` ignores unsubmitted children
* `mapFormsToData()` ignores unsubmitted children
In my opinion this is a more proper solution than #8362.
Commits
-------
85330a6 [Form] Fixed patched forms to be valid even if children are not submitted
50f201e Revert "[Form] Fix of "PATCH'ed forms are never valid""
This PR was merged into the 2.3 branch.
Discussion
----------
[Form] Fixed: If a form is not present in a request, it is not automatically submitted
| Q | A
| ------------- | ---
| Bug fix? | yes
| New feature? | no
| BC breaks? | no
| Deprecations? | no
| Tests pass? | yes
| Fixed tickets | #8385
| License | MIT
| Doc PR | -
This PR changes the following behavior of `handleRequest()`:
Current behavior:
```php
$_POST = array('foo' => 'bar');
$form = $factory->createNamed('myform', 'mytype');
// "myform" is not present in the request
$form->handleRequest();
assert(true === $form->isSubmitted());
```
Behavior after PR:
```php
$_POST = array('foo' => 'bar');
$form = $factory->createNamed('myform', 'mytype');
// "myform" is not present in the request
$form->handleRequest();
assert(false === $form->isSubmitted());
```
As #8385 pointed out, the latter behavior is expected, so I consider the current behavior a bug.
Commits
-------
cb5e765 [Form] Fixed: If a form is not present in a request, it is not automatically submitted