This PR was merged into the 2.7 branch.
Discussion
----------
[Debug] Always decorate existing exception handlers to deal with fatal errors
| Q | A
| ------------- | ---
| Branch? | 2.7
| Bug fix? | yes
| New feature? | no
| BC breaks? | no
| Deprecations? | no
| Tests pass? | yes
| Fixed tickets | #25827
| License | MIT
| Doc PR | -
Decorating the exception is seamless, let's always do it and fix handling of fatal errors.
Related to #25408 also.
Commits
-------
205d7ae3f7 [Debug] Always decorate existing exception handlers to deal with fatal errors
This PR was merged into the 2.7 branch.
Discussion
----------
Fixing a bug where the dump() function depended on bundle ordering
| Q | A
| ------------- | ---
| Branch? | 2.7
| Bug fix? | yes
| New feature? | no
| BC breaks? | no
| Deprecations? | no
| Tests pass? | yes
| Fixed tickets | none
| License | MIT
| Doc PR | not needed
If DebugBundle is registered *before* TwigBundle, then the simpler `dump()` function wins over the fancy, var-dumper one from DebugBundle. In other words, you get different functionality based on the *order* in which you install libraries. To get the "bad" way, you can:
```
composer create-project symfony/skeleton show_bad_dump
cd show_bad_dump
composer require symfony/debug-bundle
composer require twig
```
Then create a Twig template with a `dump()` inside. It will use the less-fancy XDebug version.
I'm not sure if there's a more elegant fix for this or not... I have verified locally that this DOES fix the issue.
Thanks!
Commits
-------
717663aac1 Fixing a bug where the dump() function depended on bundle ordering
This PR was squashed before being merged into the 2.7 branch (closes#25789).
Discussion
----------
Enableable ArrayNodeDefinition is disabled for empty configuration
| Q | A
| ------------- | ---
| Branch? | 2.7+
| Bug fix? | yes
| New feature? | no
| BC breaks? | no
| Deprecations? | no
| Tests pass? | yes
| Fixed tickets | #25760
| License | MIT
Fixes#25760.
Currently, documented behavior is not true:
70c8c2d47b/src/Symfony/Component/Config/Definition/Builder/ArrayNodeDefinition.php (L207-L208)
Commits
-------
a6a330dcd9 Enableable ArrayNodeDefinition is disabled for empty configuration
This PR was squashed before being merged into the 2.7 branch (closes#25816).
Discussion
----------
Problem in phar see mergerequest #25579
| Q | A
| ------------- | ---
| Branch? | 2.7
| Bug fix? | yes
| New feature? | no
| BC breaks? | no
| Deprecations? | no
| Tests pass? | yes
| Fixed tickets | #... <!-- #-prefixed issue number(s), if any -->
| License | MIT
| Doc PR | symfony/symfony-docs#... <!--highly recommended for new features-->
If packed into phar the old version creates path like 'file:///phar%3A/'. see https://github.com/symfony/symfony/pull/25579
Commits
-------
3e0c8ffb43 Problem in phar see mergerequest #25579
This PR was squashed before being merged into the 2.7 branch (closes#25781).
Discussion
----------
[Form] Disallow transform dates beyond the year 9999
Fixes#14727
| Q | A
| ------------- | ---
| Branch? | 2.7
| Bug fix? | yes
| New feature? | no
| BC breaks? | not really
| Deprecations? | no
| Tests pass? | yes
| Fixed tickets | #14727
| License | MIT
Explicitly locked out submission of dates beyond December 31st 9999 in forms as PHP is highly incapable of consistently handling such dates. Before this patch dates were randomly transformed or mangled.
Technically there is a BC break as this will now cause validation to fail on input that was *accepted* before, but it was mangled. Not my call but I prefer the rejection over data corruption:
```
// Old behavior
$transformer = new DateTimeToLocalizedStringTransformer('UTC', 'UTC', null, null, \IntlDateFormatter::GREGORIAN, 'yyyy-MM-dd');
$result = $transformer->reverseTransform('20107-03-21');
// $result is now 2007-03-21
```
Commits
-------
70cc969537 [Form] Disallow transform dates beyond the year 9999
This PR was merged into the 2.7 branch.
Discussion
----------
Copied NO language files to the new NB locale
| Q | A
| ------------- | ---
| Branch? | 2.7
| Bug fix? | yes
| New feature? | no
| BC breaks? | no
| Deprecations? | no
| Tests pass? | yes
| Fixed tickets | #25792
| License | MIT
| Doc PR | N/A
This PR copies all `NO` language files to a new locale `NB`. It also adds unit tests to ensure that `NB` and `NO` will always contain the same translations. This way, we allow application developers to either use the generic `NO` language code or the more precise `NB` (e.g. if they need to distinguish between the `NB` and `NN` variants of the Norwegian language).
For further details, please have a look at the discussion in #25792.
Commits
-------
aee9b1ea3e Copied NO language files to the new NB locale.
This PR was merged into the 2.7 branch.
Discussion
----------
[Console] Improve phpdoc on StyleInterface::ask()
| Q | A
| ------------- | ---
| Branch? | 2.7
| Bug fix? | yes
| New feature? | no
| BC breaks? | no
| Deprecations? | no
| Tests pass? | yes
| Fixed tickets | #... <!-- #-prefixed issue number(s), if any -->
| License | MIT
| Doc PR | symfony/symfony-docs#... <!--highly recommended for new features-->
In a CLI command i keep asking an optional question until i get an answer. So interactively it's a required question. It looks like..
```php
do {
$value = $io->ask('Value', null, function ($value) { return $value; });
} while (null === $value);
```
Which works nice.. but SA is complaining about
```
Strict comparison using === between null and string will always evaluate to false.
```
This should fix it. The mixed API goes back to https://github.com/symfony/symfony/blob/master/src/Symfony/Component/Console/Helper/QuestionHelper.php#L38
Commits
-------
499d04f [Console] Improve phpdoc on StyleInterface::ask()
This PR was merged into the 2.7 branch.
Discussion
----------
[appveyor] set memory_limit=-1
| 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
-------
10e33ac [appveyor] set memory_limit=-1
This PR was squashed before being merged into the 2.7 branch (closes#25801).
Discussion
----------
[Router] Skip anonymous classes when loading annotated routes
| Q | A
| ------------- | ---
| Branch? | 2.7
| Bug fix? | yes
| New feature? | no
| BC breaks? | no
| Deprecations? | no
| Tests pass? | yes
| Fixed tickets | #25796
| License | MIT
| Doc PR |
Skip any usage of anonymous classes when parsing files in `AnnotationFileLoader`
Commits
-------
d76a545 [Router] Skip anonymous classes when loading annotated routes
This PR was merged into the 2.7 branch.
Discussion
----------
[Security] Fix fatal error on non string username
| 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/25612
| License | MIT
| Doc PR | n/a
That's consistent with what #22569 did for the `json_login` listener.
Commits
-------
8f095683d0 [Security] Fix fatal error on non string username
This PR was submitted for the 3.4 branch but it was squashed and merged into the 3.3 branch instead (closes#25752).
Discussion
----------
Don't right trim the deprecation message
| Q | A
| ------------- | ---
| Branch? | 3.4
| Bug fix? | maybe yes
| New feature? | no
| BC breaks? | no
| Deprecations? | no <!-- don't forget to update UPGRADE-*.md files -->
| Tests pass? | yes
| Fixed tickets | #... <!-- #-prefixed issue number(s), if any -->
| License | MIT
| Doc PR | symfony/symfony-docs#... <!--highly recommended for new features-->
<!--
- Bug fixes must be submitted against the lowest branch where they apply
(lowest branches are regularly merged to upper ones so they get the fixes too).
- Features and deprecations must be submitted against the master branch.
- Replace this comment by a description of what your PR is solving.
-->
The PhpUnit bridge lists deprecation messages after a test. In order to do this it outputs the message but it right trims the message - removing any fullstops. This is unexpected. It does this to add the number of time the message appears but this is not really necessary because the number of the times a deprecation message is triggered and from where is added below.
Commits
-------
0b03631 Don't right trim the deprecation message
This PR was merged into the 2.7 branch.
Discussion
----------
[Serializer] Fixed throwing exception with option JSON_PARTIAL_OUTPUT_ON_ERROR
| Q | A
| ------------- | ---
| Branch? | 2.7
| Bug fix? | yes
| New feature? | no <!-- don't forget to update src/**/CHANGELOG.md files -->
| BC breaks? | no
| Deprecations? | no <!-- don't forget to update UPGRADE-*.md files -->
| Tests pass? | yes
| Fixed tickets | no
| License | MIT
| Doc PR | no
<!--
- Bug fixes must be submitted against the lowest branch where they apply
(lowest branches are regularly merged to upper ones so they get the fixes too).
- Features and deprecations must be submitted against the master branch.
- Replace this comment by a description of what your PR is solving.
-->
Php function json_encode/decode with option JSON_PARTIAL_OUTPUT_ON_ERROR return result on error, but if have is error json_last_error() always return error code even if there is a result and it is not false. Because of this is impossible set JSON_PARTIAL_OUTPUT_ON_ERROR option across variable $context.
Current fix solves this problem.
Verification on the false is completely correct, since json_encode / decode returns false only on error if not set JSON_PARTIAL_OUTPUT_ON_ERROR option.
Such have a problem e.g when encoding data is not utf-8 (emoji from facebook).
Commits
-------
e7e410b [Serializer] Fixed throwing exception with option JSON_PARTIAL_OUTPUT_ON_ERROR
This PR was merged into the 2.7 branch.
Discussion
----------
[Security] Fail gracefully if the security token cannot be unserialized from the session
| 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 |
If the security token in the session can't be unserialized, an `E_NOTICE` is issued. This prevents it (and provides a better log message if it's not even a `__PHP_Incomplete_Class`).
This is similar to #24731, but I saw it triggered when changing OAuth library (https://github.com/elifesciences/journal/pull/824), so the token class itself no longer exists. (I want to avoid having to manually invalidate all sessions, as not all sessions use that token class.)
Commits
-------
053fa43add [Security] Fail gracefully if the security token cannot be unserialized from the session
This PR was merged into the 2.7 branch.
Discussion
----------
[Routing] Use the default host even if context is empty
| Q | A
| ------------- | ---
| Branch? | 2.7
| Bug fix? | yes
| New feature? | no
| BC breaks? | no
| Deprecations? | no
| Tests pass? | yes
| Fixed tickets | #25464
| License | MIT
| Doc PR | ø
When the host in the context is empty... we should still use the default host.
**Note:** it seems like a lot of changes but all I did was to remove the `if` and de-indent the code that was inside.
Commits
-------
8f357df75b Use the default host even if context is empty and fallback to relative URL if empty host
This PR was squashed before being merged into the 2.7 branch (closes#25653).
Discussion
----------
PHP CS Fixer: clean up repo and adjust config
| Q | A
| ------------- | ---
| Branch? | 2.7
| Bug fix? | no
| New feature? | no
| BC breaks? | no
| Deprecations? | no
| Tests pass? | n/a
| Fixed tickets | n/a
| License | MIT
| Doc PR | n/a
Reason for this PR is that one want to have `php-cs-fixer fix -v` command executed without changes that shall not be applied for this repo. To achieve that, we need to groom config to exclude files that violate CS willingly, fix files that are violating CS unwillingly, and deliver missing case handling at PHP CS Fixer itself (https://github.com/FriendsOfPHP/PHP-CS-Fixer/pull/3359) (already merged!).
Commits
-------
b14cbc1 PHP CS Fixer: clean up repo and adjust config
Christian Flothmann
Edi Modrić
Fabien Potencier
Nicolas Grekas
Robin Chalas
Mateusz Sip
Ryan Weaver
Gunnstein Lye
Boris Betzholz
Niels Keurentjes
Alexander M. Turek
Pierre du Plessis
Yanick Witschi
Robin Chalas
Grégoire Pineau
Samuel ROZE
Vacheslav Silyutin
Chris Wilkinson
Tomáš Votruba
Roland Franssen
Gabriel Ostrolucký
Dariusz