This PR was submitted for the master branch but it was merged into the 2.3 branch instead (closes#11635).
Discussion
----------
[Finder] Fix findertest readability
| Q | A
| ------------- | ---
| Bug fix? | yes
| New feature? | no
| BC breaks? | no
| Deprecations? | no
| Tests pass? | yes
| Fixed tickets | part of #11631, #11588
| License | MIT
When running on Ubuntu it is hard to setup an account that allows the `chmod` to downgrade the rights on a directory but is not allow to read from the downgraded directory.
Commits
-------
8a47b62 [Finder] Fix findertest readability
This PR was merged into the 2.3 branch.
Discussion
----------
[Security] Made optimization on constant-time algorithm removing modulus operator
| Q | A
| ------------- | ---
| Bug fix? | no
| New feature? | no
| BC breaks? | no
| Deprecations? | no
| Tests pass? | yes
| Fixed tickets | -
| License | MIT
| Doc PR | -
This fix improves the constant-time algorithm used to compare strings, as it removes the `%` operator inside the loop.
Commits
-------
000bd0d Made optimization deprecating modulus operator
This PR was submitted for the master branch but it was merged into the 2.3 branch instead (closes#11772).
Discussion
----------
[Filesystem] Add FTP stream wrapper context option to enable overwrite
| Q | A
| ------------- | ---
| Bug fix? | yes
| New feature? | no
| BC breaks? | no
| Deprecations? | no
| Tests pass? | yes
| License | MIT
Without this change it's not possible to override a file on FTP by calling Filesystem::copy($originFile, $targetFile, true) as PHP/FTP server responds with error like:
fopen(ftp://...@ftp/file.txt): failed to open stream: Remote file already exists and overwrite context option not specified FTP server reports 213 166440 []
TODO: Write an integration tests? How? Use some real FTP server?
Commits
-------
c056a9c [Filesystem] Add FTP stream wrapper context option to enable overwrite (override)
This PR was submitted for the master branch but it was merged into the 2.3 branch instead (closes#11802).
Discussion
----------
[HttpKernel] add use statement for phpdoc
| Q | A
| ------------- | ---
| Fixed tickets | https://github.com/symfony/symfony/issues/11801
| License | MIT
Added use statement for the Phpdoc
Commits
-------
0010fba [HttpKernel] add use statement for phpdoc
This PR was merged into the 2.3 branch.
Discussion
----------
[Yaml] fixed mapping keys containing a quoted #
| Q | A
| ------------- | ---
| Bug fix? | yes
| New feature? | no
| BC breaks? | no
| Deprecations? | no
| Tests pass? | yes
| Fixed tickets | #11700, #11723
| License | MIT
| Doc PR | n/a
Commits
-------
110f999 [Yaml] fixed mapping keys containing a quoted #
8ba3b28 Added fixture to test parsing of hash keys ending with a space and #
This PR was merged into the 2.3 branch.
Discussion
----------
Disabled the PHPUnit self-update on Travis
The 4.2.2 release available on Travis currently ships with a broken self-update command. Given that it is new enough to contain the fixes for the other bugs affecting the build, it is fine to keep using the available phar without updating it.
This avoids making all builds except 5.3.3 fail before running tests.
FYI, the bug itself is already fixed in PHPUnit so 4.2.3 will have a working command.
Commits
-------
0fca352 Disabled the PHPUnit self-update on Travis
The 4.2.2 release available on Travis currently ships with a broken
self-update command. Given that it is new enough to contain the fixes for
the other bugs affecting the build, it is fine to keep using the available
phar without updating it.
This PR was merged into the 2.3 branch.
Discussion
----------
[DoctrineBridge] Abstract Doctrine Subscribers with tags
| Q | A
| ------------- | ---
| Bug fix? | yes
| New feature? | no
| BC breaks? | no
| Deprecations? | no
| Tests pass? | yes
| Fixed tickets | this one
| License | MIT
| Doc PR | N/A
I've hit a problem with some doctrine listeners, built by decorating an abstract definition.
I want the abstract definition to hold the tag, however because the RegisterEventListenersAndSubscribersPass runs before abstract definitions are removed, they get added as method calls to the EventManager definition, which once the abstract service is removed, we end up with a method call that breaks the container.
I don't know if this is the best approach, it might be better not to return abstract services when calling `findTaggedServiceIds` instead?
Commits
-------
cbcf513 Disallow abstract definitions from doctrine event listener registration
This PR was squashed before being merged into the 2.3 branch (closes#11768).
Discussion
----------
[ClassLoader] Add a __call() method to XcacheClassLoader
| Q | A
| ------------- | ---
| Bug fix? | yes
| New feature? | no
| BC breaks? | no
| Deprecations? | no
| Tests pass? | yes
| Fixed tickets | #11733
| License | MIT
| Doc PR |
Commits
-------
dd0d6af [ClassLoader] Add a __call() method to XcacheClassLoader
This PR was merged into the 2.3 branch.
Discussion
----------
[Process] fix some minor typos in tests
| Q | A
| ------------- | ---
| Bug fix? | yes
| New feature? | no
| BC breaks? | no
| Deprecations? | no
| Tests pass? | yes
| Fixed tickets |
| License | MIT
| Doc PR |
Commits
-------
24d2c9a fix some minor typos in tests
This PR was merged into the 2.3 branch.
Discussion
----------
[YAML] resolve variables in inlined YAML
| Q | A
| ------------- | ---
| Bug fix? | yes
| New feature? | no
| BC breaks? | no
| Deprecations? | no
| Tests pass? | yes
| Fixed tickets | #11665
| License | MIT
| Doc PR |
#11569 does not resolve variables in inline YAML.
Commits
-------
45a5863 [YAML] resolve variables in inlined YAML
This PR was submitted for the master branch but it was merged into the 2.3 branch instead (closes#11639).
Discussion
----------
[DependencyInjection] Fixed factory service not within the ServiceReferenceGraph.
| Q | A
| ------------- | ---
| Bug fix? | yes
| New feature? | no
| BC breaks? | no
| Deprecations? | no
| Tests pass? | yes
| Fixed tickets |
| License | MIT
| Doc PR |
Fixed a problem where Factory services are not added to the ServiceReferenceGraph.
Commits
-------
e992f8e Fixed Factory services not within the ServiceReferenceGraph.
This PR was merged into the 2.3 branch.
Discussion
----------
[Validator] Fixed wrong translations for Collection constraints
| Q | A
| ------------- | ---
| Bug fix? | yes
| New feature? | no
| BC breaks? | no
| Deprecations? | no
| Tests pass? | yes
| Fixed tickets | #11630
| License | MIT
| Doc PR |
The error messages for a missing field and an unexpected field did not match the Constraint class.
Commits
-------
808de2b [Validator] Fixed wrong translation keys/messages for Collection constraint. The error messages for a missing field and an unexpected field did not match the Contraint class
This PR was submitted for the master branch but it was merged into the 2.3 branch instead (closes#11779).
Discussion
----------
[CssSelector] Fix URL to SimonSapin/cssselect repo in README
This fixes a broken link in the CssSelector README spotted by @philsturgeon in https://github.com/symfony/CssSelector/pull/2.
| Q | A
| ------------- | ---
| Fixed tickets | none
| License | MIT
Commits
-------
b9a237c [CssSelector] Fix URL to SimonSapin/cssselect repo
This PR was submitted for the master branch but it was merged into the 2.3 branch instead (closes#11756).
Discussion
----------
[DependencyInjection] fix @return anno created by PhpDumper
| Q | A
| ------------- | ---
| Fixed tickets | ---
| License | MIT
If a generated container is inside namespace, generated @return annos are not valid.
Commits
-------
39c4cc3 [DependencyInjection] fix @return anno created by PhpDumper
This PR was merged into the 2.3 branch.
Discussion
----------
Fixed the phpdoc of the VoterInterface
| Q | A
| ------------- | ---
| Bug fix? | no
| New feature? | no
| BC breaks? | no
| Deprecations? | no
| Tests pass? | yes
| Fixed tickets |
| License | MIT
| Doc PR | n/a
``null`` is an accepted value for voters (it is passed when you are not voting on a specific object, for instance for roles.
Commits
-------
b591404 Fixed the phpdoc of the VoterInterface
This PR was squashed before being merged into the 2.3 branch (closes#11711).
Discussion
----------
[DoctrineBridge] Fix empty parameter logging in the dbal logger
| Q | A
| ------------- | ---
| Bug fix? | yes
| New feature? | no
| BC breaks? | no
| Deprecations? | no
| Tests pass? | yes
| Fixed tickets | #11710
| License | MIT
| Doc PR | -
Commits
-------
ef23f02 [DoctrineBridge] Fix empty parameter logging in the dbal logger
This PR was merged into the 2.3 branch.
Discussion
----------
[DomCrawler] check for the correct field type
| Q | A
| ------------- | ---
| Bug fix? | yes
| New feature? | no
| BC breaks? | no
| Deprecations? | no
| Tests pass? | yes
| Fixed tickets | #11689
| License | MIT
| Doc PR |
HTML allow to define different form fields with the same name. Imagine the following form:
```html
<html>
<body>
<form action="/">
<input type="hidden" name="option" value="default">
<input type="radio" name="option" value="A">
<input type="radio" name="option" value="B">
<input type="hidden" name="settings[1]" value="0">
<input type="checkbox" name="settings[1]" value="1" id="setting-1">
<button>klickme</button>
</form>
</body>
</html>
```
Since the `FormFieldRegistry` can only handle one field per name, the hidden field option is registered first before the radio field with the same name is evaluated. Thus, the `FormFieldRegistry` returns an `InputFormField` instance on which the `addChoices()` method can not be called.
Commits
-------
169b397 check for the correct field type
This PR was merged into the 2.3 branch.
Discussion
----------
[Routing] fix handling of nullable XML attributes
| Q | A
| ------------- | ---
| Bug fix? | yes
| New feature? | no
| BC breaks? | no
| Deprecations? | no
| Tests pass? | yes
| Fixed tickets |
| License | MIT
| Doc PR |
As @Tobion pointed out in #11394, ``true`` and ``1`` are valid values in boolean XML attributes. The XmlFileLoader didn't handle ``1`` values properly.
Commits
-------
7b4d4b6 fix handling of nullable XML attributes
This PR was merged into the 2.3 branch.
Discussion
----------
[DomCrawler] fix the axes handling in a bc way
| Q | A
| ------------- | ---
| Bug fix? | yes
| New feature? | no
| BC breaks? | no
| Deprecations? | no
| Tests pass? | yes
| Fixed tickets | #11503
| License | MIT
| Doc PR |
The previous fix in #11548 for handling XPath axes was not backward compatible. In previous Symfony versions the Crawler handled nodes by holding a "fake root node". This must be taken into account when evaluating (relativizing) XPath expressions.
Commits
-------
d26040f [DomCrawler] fix the axes handling in a bc way
This PR was submitted for the master branch but it was merged into the 2.3 branch instead (closes#11676).
Discussion
----------
[Form] Fixed#11675 ValueToDuplicatesTransformer accept "0" value
| Q | A
| ------------- | ---
| Bug fix? | [yes]
| New feature? | [no]
| BC breaks? | [no]
| Deprecations? | [no]
| Tests pass? | [yes]
| Fixed tickets | #11675
| License | MIT
Commits
-------
31d48abFixed#11675 ValueToDuplicatesTransformer accept "0" value
This PR was submitted for the master branch but it was merged into the 2.3 branch instead (closes#11695).
Discussion
----------
[Validators] Fixed failing tests requiring ICU 52.1 which are skipped otherwise
| Q | A
| ------------- | ---
| Bug fix? | yes
| New feature? | no
| BC breaks? | no
| Deprecations? | no
| Tests pass? | yes
| Fixed tickets | -
| License | MIT
| Doc PR | -
I just ran the tests with ICU 52.1 installed and noticed that a few of them failed. This PR fixes these tests.
Commits
-------
5440ed5 [Validators] Fixed failing tests requiring ICU 52.1 which are skipped otherwise
This PR was merged into the 2.3 branch.
Discussion
----------
[Validator] return empty metadata collection if none do exist
| Q | A
| ------------- | ---
| Bug fix? | yes
| New feature? | no
| BC breaks? | no
| Deprecations? | no
| Tests pass? | yes
| Fixed tickets |
| License | MIT
| Doc PR | [The reference to the documentation PR if any]
Backport of #11614 for Symfony 2.3 and 2.4.
Commits
-------
f5bc18d return empty metadata collection if none do exist
HTML allow to define different form fields with the same name.
Imagine the following form:
<html>
<body>
<form action="/">
<input type="hidden" name="option" value="default">
<input type="radio" name="option" value="A">
<input type="radio" name="option" value="B">
<input type="hidden" name="settings[1]" value="0">
<input type="checkbox" name="settings[1]" value="1" id="setting-1">
<button>klickme</button>
</form>
</body>
</html>
Since the `FormFieldRegistry` can only handle one field per name, the
hidden field option is registered first before the radio field with
the same name is evaluated. Thus, the `FormFieldRegistry` returns an
`InputFormField` instance on which the `addChoices()` method can not
be called.
This PR was submitted for the master branch but it was merged into the 2.3 branch instead (closes#11674).
Discussion
----------
[Validator] Add missing translation NL
| Q | A
| ------------- | ---
| Bug fix? | yes
| New feature? | no
| BC breaks? | no
| Deprecations? | no
| Tests pass? | yes
| Fixed tickets | part of #11489
| License | MIT
Commits
-------
2d5ab91 add missing options
Fabien Potencier
1emming
Damian Sromek
Miliooo
Christophe Coevoet
Tobias Stöckler
Christian Flothmann
Harm van Tilborg
Ka
Warnar Boekkooi
Johnson Page
Sam Malone
Jakub Kulhan
Damien Tournoud
Jakub Zalas
Lee Rowlands
Nicolas Grekas
Tobias Schultze
Bernhard Schussek
Nek (Maxime Veber)