This PR was merged into the 2.6 branch.
Discussion
----------
[VarDumper] fix very special vars handling
| Q | A
| ------------- | ---
| Bug fix? | yes
| New feature? | no
| BC breaks? | no
| Deprecations? | no
| Tests pass? | yes
| Fixed tickets | #13235
| License | MIT
| Doc PR | -
Commits
-------
e26dc2c [VarDumper] fix very special vars handling
* 2.5:
[TwigBundle] allowed SecurityBundle to use the latest versions of FrameworkBundle
[HttpFoundation] Make use of isEmpty() method
fix missing comma in YamlDumper
[Console] Helper\Table->addRow optimization
[Console] Helper\Table->addRow optimization
Conflicts:
src/Symfony/Bundle/SecurityBundle/composer.json
* 2.3:
[TwigBundle] allowed SecurityBundle to use the latest versions of FrameworkBundle
[HttpFoundation] Make use of isEmpty() method
fix missing comma in YamlDumper
[Console] Helper\Table->addRow optimization
Conflicts:
src/Symfony/Bundle/SecurityBundle/composer.json
src/Symfony/Component/Console/Helper/TableHelper.php
src/Symfony/Component/DependencyInjection/Tests/Fixtures/containers/container9.php
src/Symfony/Component/DependencyInjection/Tests/Fixtures/yaml/services9.yml
This PR was merged into the 2.7 branch.
Discussion
----------
lazy-load fragment renderers
| Q | A
| ------------- | ---
| Bug fix? | no
| New feature? | no
| BC breaks? | no
| Deprecations? | yes
| Tests pass? | yes
| Fixed tickets | n/a
| License | MIT
| Doc PR | n/a
Now that we have more fragment renderers (inline, hinclude, esi, and ssi), it's a waste of time to load them all for every single requests; actually, most of the time, we load them for nothing.
So, like we did for event listeners, I propose to lazy-load them.
I've also move the classes to the HttpKernel component as the corresponding classes for the lazy-loading of listeners were moved to the EventDispatcher component a while ago to make them reusable outside of the Symfony2 context.
Last, but not the least, I've named the class with a `LazyLoading` prefix instead of the usual `ContainerAware` as I think it conveys the goal much better. I'd like to rename the other ones as well when it makes sense.
Commits
-------
2be8b6e [TwigBundle] optimized the hinclude fragement renderer when only Twig is used
6148652 [FrameworkBundle] removed obsolete ContainerAwareHIncludeFragmentRenderer class
e620cbf lazy-load fragment renderers
This PR was merged into the 2.3 branch.
Discussion
----------
fix missing comma in YamlDumper
| Q | A
| ------------- | ---
| Bug fix? | yes
| New feature? | no
| BC breaks? | no
| Deprecations? | no
| Tests pass? | yes
| Fixed tickets |
| License | MIT
| Doc PR |
The YamlDumper were missing commas between tag attributes.
See https://github.com/rosstuck/TuckConverterBundle/issues/6
Commits
-------
f600d1a fix missing comma in YamlDumper
This PR was merged into the 2.7 branch.
Discussion
----------
[2.7] Update README.md to min PHP 5.3.9
| Q | A
| ------------- | ---
| Bug fix? | no
| New feature? | no
| BC breaks? | no
| Deprecations? | no
| Tests pass? | yes
| Fixed tickets |
| License | MIT
| Doc PR |
see 43c05650
Commits
-------
c8857c1 Update README.md to min PHP 5.3.9
This PR was submitted for the 2.7 branch but it was merged into the 2.3 branch instead (closes#13365).
Discussion
----------
[HttpFoundation] Make use of isEmpty() method
| Q | A
| ------------- | ---
| Bug fix? | no
| New feature? | no
| BC breaks? | no
| Deprecations? | no
| Tests pass? | yes
| Fixed tickets |
| License | MIT
| Doc PR |
Commits
-------
fc51d54 [HttpFoundation] Make use of isEmpty() method
This PR was merged into the 2.7 branch.
Discussion
----------
Twig decoupling from Templating
| Q | A
| ------------- | ---
| Bug fix? | no
| New feature? | yes
| BC breaks? | no
| Deprecations? | yes
| Tests pass? | yes
| Fixed tickets | #11748, #11876
| License | MIT
| Doc PR | n/a
The goal of this PR is to make Twig independent of the Symfony Templating system. With this PR, you can now use the Twig bundle without activating `framework.templating`. Also, even when registering the templating system in FrameworkBundle, the `php` engine is only registered when specified explicitly in `engines`.
First, the global variables has been decoupled from FrameworkBundle.
Then, the Twig bundle now tries to only rely on native Twig extensions and loaders and use the Templating sub-system only if `framework.templating` is enabled.
Commits
-------
18d4c41 [TwigBundle] added some tests
0d537c4 decoupled Twig from the Templating system
be5a208 decoupled global variables system in Twig from the Templating one
This PR was merged into the 2.7 branch.
Discussion
----------
URL manipulations as a Twig extension
| Q | A
| ------------- | ---
| Bug fix? | no
| New feature? | yes
| BC breaks? | no
| Deprecations? | no
| Tests pass? | yes
| Fixed tickets | n/a
| License | MIT
| Doc PR | symfony/symfony-docs#4805
While working on the new asset component, I realized that the "absolute URL" feature was misplaced and would benefit from being exposed as a Twig function (composition is always a good thing). Then, I wondered if having a Twig function to generate a relative path (like done by the Routing component would also make sense). And here is the corresponding PR.
```jinja
{# generate an absolute URL for the given absolute path #}
{{ absolute_url('/me.png') }}
{# generate a relative path for the given absolute path (based on the current Request) #}
{{ relative_path('/foo/me.png') }}
{# compose as you see fit #}
{{ absolute_url(asset('me.png')) }}
```
As you can see, we require an absolute path for both functions (and we even add the leading slash if it is omitted), not sure if we want to do otherwise.
ping @tobion
Commits
-------
0ec852d added a relative_path Twig function
ee27ed8 added an absolute_url() Twig function
This PR was merged into the 2.7 branch.
Discussion
----------
[2.7] Fix deprecations on TwigBundle and FrameworkBundle semantic configurations
| Q | A
| ------------- | ---
| Bug fix? | no
| New feature? | no
| BC breaks? | no
| Deprecations? | no
| Tests pass? | yes
| Fixed tickets | ~
| License | MIT
| Doc PR | ~
Commits
-------
8d60396 [FrameworkBundle|TwigBundle] update functional tests configuration files to not use deprecated config keys anymore.
* 2.6:
[PropertyAccessor] Added test to allow null value for a array
[Yaml] Fixed#10597: Improved Yaml directive parsing
[Form] Fixed check of violation constraint #12792
[Form] Set a child type to text if added to the form without a type.
* 2.5:
[PropertyAccessor] Added test to allow null value for a array
[Yaml] Fixed#10597: Improved Yaml directive parsing
[Form] Set a child type to text if added to the form without a type.
* 2.3:
[PropertyAccessor] Added test to allow null value for a array
[Yaml] Fixed#10597: Improved Yaml directive parsing
[Form] Set a child type to text if added to the form without a type.
Conflicts:
src/Symfony/Component/PropertyAccess/PropertyAccessor.php
src/Symfony/Component/PropertyAccess/Tests/PropertyAccessorTest.php
This PR was merged into the 2.3 branch.
Discussion
----------
[PropertyAccessor] Allow null value for a array (2.3)
| Q | A
| ------------- | ---
| Bug fix? | yes
| New feature? | no
| BC breaks? | no
| Deprecations? | no
| Tests pass? | yes
| Fixed tickets | #12482
| License | MIT
| Doc PR | -
Original PR is #12511
A rebase on 2.3 was requested by @fabpot
Commits
-------
9706b09 [PropertyAccessor] Added test to allow null value for a array
This PR was merged into the 2.7 branch.
Discussion
----------
Validator deprecation fixes
| 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
Commits
-------
a34220e [Validator] always use the lazy loading metadata factory
818ca58 [Validator] removed usage of deprecated getMessageParameters() and getMessagePluralization() in unit tests
c6f1f69 [Validator] fixed deprecation notices for BuildViolation() calls in constraints
fb3f9d2 [Validator] fixed usage of deprecate Validator features
91606b5 [Validator] removed obsolete code
5aa44ee removed the Validator BC layer for PHP < 5.3.9
a4139c0 removed code for PHP < 5.3.9
6dc8979 bumped min PHP version to 5.3.9
This PR was merged into the 2.3 branch.
Discussion
----------
[Form] Set a child type to text if added to the form without a type.
| Q | A
| ------------- | ---
| Bug fix? | yes
| New feature? | no
| BC breaks? | no
| Deprecations? | no
| Tests pass? | yes
| Fixed tickets | #13167
| License | MIT
| Doc PR | -
This copies the behaviour of `FormBuilder::create()` to `Form::add()`.
ping @webmozart
Commits
-------
57070a2 [Form] Set a child type to text if added to the form without a type.
This PR was submitted for the 2.7 branch but it was merged into the 2.3 branch instead (closes#13334).
Discussion
----------
[Yaml] Fixed#10597: Improved Yaml directive parsing
| Q | A
| ------------- | ---
| Bug fix? | yes
| New feature? | no
| BC breaks? | no
| Deprecations? | no
| Tests pass? | yes
| Fixed tickets | #10597
| License | MIT
| Doc PR | N/A
There was a problem with the regular expression used to process the YAML directive. The modifier `s` made the regex too greedy, replacing everything but the last line of the YAML. Existing tests using the YAML directive contained only one line, that's why they were passing.
Commits
-------
95d8ce3 [Yaml] Fixed#10597: Improved Yaml directive parsing