This PR was submitted for the 2.8 branch but it was merged into the 3.3-dev branch instead (closes#20161).
Discussion
----------
add toolbar & profiler SVG style classes
| Q | A |
| --- | --- |
| Bug fix? | no |
| New feature? | Technically, yes - but actually not |
| BC breaks? | no |
| Deprecations? | no |
| Tests pass? | yes |
| License | MIT |
This allows the usage of SVG not only containing `path` elements. I opted for a generic solution using the two classes to apply to any SVG, one would like to use within the toolbar (`sf-toolbar-path`) and/or profiler (`sf-profiler-path`).
Commits
-------
155a1fc add toolbar & profiler SVG style classes
This PR was merged into the 3.3-dev branch.
Discussion
----------
[DomCrawler] Add support for formaction and formmethod attributes
| 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
This adds supports for the ``formaction`` and ``formmethod`` of submit elements, which override the values defined on the ``<form>`` element.
This works only when you call ``$crawler->form()`` on a Crawler containing a button, not when it contains the ``<form>`` itself of course (as the button override is applied only when using this button to submit, not when using another way).
Other button-level overrides are not implemented:
- ``formtarget`` is useless as we don't implement ``target`` either (the Crawler does not deal with frame-based pages anyway)
- ``formnovalidate`` is ignored, as we don't automatically disable the form validation on ``novalidate`` either, but we require an explicit disabling instead (this might be subject to a separate PR though, as it could make sense)
- ``formenctype`` is ignored as we also ignore ``enctype`` (we always submit file fields, even when missing the proper enctype)
Commits
-------
717cf8a [DomCrawler] Add support for formaction and formmethod attributes
This PR was submitted for the 2.8 branch but it was merged into the 3.3-dev branch instead (closes#20452).
Discussion
----------
debug service with one line by tag
| Q | A
| ------------- | ---
| Branch? | 2.8, 3.*
| Bug fix? | no
| New feature? | no
| BC breaks? | no
| Deprecations? | no
| Tests pass? | yes
| License | MIT
Since the 2.8 branch, the `debug:container` command output service tags in one single line.
So if the debugged service have many tags, this print an unreadable large table.
This PR solve this issue by printing one line by tag.
Commits
-------
dbf4850 debug service with one line by tag
This PR was squashed before being merged into the 3.3-dev branch (closes#20509).
Discussion
----------
[Serializer] Allow to specify a single value in @Groups
| Q | A
| ------------- | ---
| Branch? | master
| Bug fix? | no
| New feature? | yes
| BC breaks? | no
| Deprecations? | no
| Tests pass? | yes
| Fixed tickets | https://github.com/symfony/symfony/pull/19374#issuecomment-256688002
| License | MIT
| Doc PR | todo
Tiny DX improvement:
Before:
```php
use Symfony\Component\Serializer\Annotation\Groups;
class Product
{
/**
* @Groups({"admins"})
*/
public $itemsSold;
}
```
Now allowed:
```php
use Symfony\Component\Serializer\Annotation\Groups;
class Product
{
/**
* @Groups("admins")
*/
public $itemsSold;
}
```
Commits
-------
926aa48 [Serializer] Allow to specify a single value in @Groups
This PR was merged into the 3.2 branch.
Discussion
----------
[Serializer] Add deprecation missing from UPGRADE files
| Q | A
| ------------- | ---
| Branch? | 3.2
| Bug fix? | no
| New feature? | no
| BC breaks? | no
| Deprecations? | no, simply mention one in upgrade files
| Tests pass? | yes
| Fixed tickets | N/A
| License | MIT
| Doc PR | N/A
Commits
-------
d681c1a Lines length should be 80 or less characters
597bbdc Lines length should be 80 or less characters
4514c1b [Serializer] Add deprecation missing from UPGRADE files
This PR was merged into the 3.2 branch.
Discussion
----------
[Form] Remove unused var cloner property
| Q | A
| ------------- | ---
| Branch? | 3.2
| Bug fix? | no
| New feature? | no
| BC breaks? | no
| Deprecations? | no
| Tests pass? | yes
| Fixed tickets | N/A
| License | MIT
| Doc PR | N/A
Not used anymore after https://github.com/symfony/symfony/pull/19986
EDIT: add missing `use` too.
Commits
-------
0708003 [Form] Remove unused var cloner property
This PR was merged into the 2.7 branch.
Discussion
----------
[TwigBundle] Fix twig loader registered twice
| Q | A
| ------------- | ---
| Branch? | 2.7
| Bug fix? | yes
| New feature? | no
| BC breaks? | no
| Deprecations? | no
| Tests pass? | yes
| Fixed tickets | https://github.com/symfony/symfony/issues/20665
| License | MIT
| Doc PR | N/A
Generated code:
### Before
```php
protected function getTwig_LoaderService()
{
$a = new \Twig_Loader_Filesystem(array(), $this->targetDirs[3]);
$a->addPath(...);
// ...
$this->services['twig.loader'] = $instance = new \Twig_Loader_Chain();
$instance->addLoader($a);
$instance->addLoader($a);
return $instance;
}
```
### After
```php
protected function getTwig_LoaderService()
{
$this->services['twig.loader'] = $instance = new \Twig_Loader_Filesystem(array(), $this->targetDirs[3]);
$instance->addPath(...);
// ...
return $instance;
}
```
~~Another solution is to simply create a private alias. But I don't know if we should care or not about the case people may rely on the fact both services exist as definition, and not as an alias, in a compiler pass.~~ (Has been preferred over of using a child definition)
For reference, this issue was introduced in https://github.com/symfony/symfony/pull/13354.
Commits
-------
2c81819 [TwigBundle] Fix twig loader registered twice
This PR was merged into the 2.8 branch.
Discussion
----------
[WebProfilerBundle] Fix dump block is unfairly restrained
| Q | A
| ------------- | ---
| Branch? | 2.8
| Bug fix? | yes
| New feature? | no
| BC breaks? | no
| Deprecations? | no
| Tests pass? | yes
| Fixed tickets | N/A
| License | MIT
| Doc PR | N/A
The `display: table-row` rule does not suit well for such blocks and prevent from containing them properly in the parent container (thus the reason why the width was set to `200px` I guess).
### Before
<img width="539" alt="screenshot 2016-12-01 a 20 21 49" src="https://cloud.githubusercontent.com/assets/2211145/20808878/8af6faa2-b804-11e6-8656-8ebd710b4acb.PNG">
### After
<img width="524" alt="screenshot 2016-12-01 a 20 21 14" src="https://cloud.githubusercontent.com/assets/2211145/20808885/9106b0cc-b804-11e6-9ddc-0cc09a546274.PNG">
(max width is still fixed to `480px` by `.sf-toolbar-block:hover .sf-toolbar-info`)
Commits
-------
997beb2 [WebProfilerBundle] Fix dump block is unfairly restrained
This PR was merged into the 3.2 branch.
Discussion
----------
[Validator] Bring egulias/email-validator ~2.0 to parity with ~1.2
| Q | A
| ------------- | ---
| Branch? | 3.2
| Bug fix? | could be ?
| New feature? | no
| BC breaks? | no
| Deprecations? | no
| Tests pass? | yes
| Fixed tickets |
| License | MIT
| Doc PR |
When using egulias/email-validator ~1.2, we make a strict check of the email, which means that emails with RFC warnings will fail validation.
Currently with egulias/email-validator ~2.0, emails with warnings are considerate valids.
This pull request make EmailValidator with egulias/email-validator ~2.0 behave as with egulias/email-validator ~1.2.
Commits
-------
01b2f66 Switch to NoRFCWarningsValidation
* 3.2:
[Config] ConfigCache::isFresh() should return false on __PHP_Incomplete_Class
[VarDumper] Use default color for ellipsed namespaces/paths
mark alias as private during creation
[Serializer] Remove unused GetSetMethodNormalizer::denormalize
* 3.1:
[Config] ConfigCache::isFresh() should return false on __PHP_Incomplete_Class
mark alias as private during creation
[Serializer] Remove unused GetSetMethodNormalizer::denormalize
This PR was merged into the 2.7 branch.
Discussion
----------
[Config] ConfigCache::isFresh() should return false when unserialize() fails
| Q | A
| ------------- | ---
| Branch? | 2.7
| Bug fix? | yes
| New feature? | no
| BC breaks? | no
| Deprecations? | no
| Tests pass? | yes
| Fixed tickets | #20654
| License | MIT
| Doc PR | -
Removes some `Warning: Class __PHP_Incomplete_Class has no unserializer` failures when clearing the cache.
Commits
-------
609245e [Config] ConfigCache::isFresh() should return false on __PHP_Incomplete_Class