Skip to content

/ symfony

Permalink
5.4

Commits on Aug 8, 2021

  1. feature #42420 [Security] Deprecate legacy signatures (wouterj) 

    This PR was merged into the 5.4 branch.

Discussion
----------

[Security] Deprecate legacy signatures

| Q             | A
| ------------- | ---
| Branch?       | 5.4
| Bug fix?      | no
| New feature?  | no
| Deprecations? | yes
| Tickets       | Ref #41613
| License       | MIT
| Doc PR        | n/a

Deprecates the left-over legacy constructor signatures in the Security system.

Commits
-------

bbc00c8 [Security] Deprecate legacy signatures
    @fabpot
    fabpot committed Aug 8, 2021
    be948ea

Commits on Aug 7, 2021

  1. [Security] Deprecate legacy signatures

    @wouterj
    wouterj committed Aug 7, 2021
    bbc00c8

  2. feature #41754 [SecurityBundle] Create a smooth upgrade path for secu… 

    …rity factories (wouterj)

This PR was merged into the 5.4 branch.

Discussion
----------

[SecurityBundle] Create a smooth upgrade path for security factories

| Q             | A
| ------------- | ---
| Branch?       | 5.4
| Bug fix?      | no
| New feature?  | yes
| Deprecations? | yes
| Tickets       | Ref #41613 (comment)
| License       | MIT
| Doc PR        | -

This change allows removing `SecurityFactoryInterface` in Symfony 6.

I've also changed the discrete ordering using "listener positions" to the much more common continuous ordering using priorities. I feel like priorities are much more self-explanatory.

Commits
-------

7385fd5 [SecurityBundle] Create a smooth upgrade path for security factories
    @chalasr
    chalasr committed Aug 7, 2021
    8e984fa

  3. [SecurityBundle] Create a smooth upgrade path for security factories

    @wouterj
    wouterj committed Aug 7, 2021
    7385fd5

  4. bug #42416 Invalid DKIM signature (metaer) 

    This PR was merged into the 5.4 branch.

Discussion
----------

Invalid DKIM signature

| Q             | A
| ------------- | ---
| Branch?       | 5.3
| Bug fix?      | yes
| New feature?  | no
| Deprecations? | no
| License       | MIT

How to reproduce:

```php
$email = (new Email())
    ->from('admin@yourdomain.com')
    ->subject('text')
    ->text('text')
;

$addresses = 'first@example.com,second@exmple.com' //could be command's or method's argument, that's why we used call_user_func_array below
\call_user_func_array([$email, 'to'], explode(',', $addresses));

$email->getHeaders()->addTextHeader('X-Transport', $transport);
$privateKeyFilePath = '/private.pem';
$filesystem = new Filesystem();

if ($filesystem->exists($privateKeyFilePath)) {
    $signer = new DkimSigner("file://$privateKeyFilePath", 'yourdomain.com', 'your-selector');
    $email = $signer->sign($email, ['headers_to_ignore' => ['x-transport']]); //#39354 (comment)
}

$this->mailer->send($email);
```

How I tested body hash (bh): https://www.appmaildev.com/site/testfile/dkim?lang=en
It shows expected body hash (`Expected-Body-Hash`) in DKIM section and received body hash (`bh=`) if fails.

![image](https://user-images.githubusercontent.com/6103208/128579511-ead9fde9-1492-4a6a-8953-2adb99eb251a.png)

My solution is based on swiftmailer signer code and DKIM specifications (see links below).
After applying patch it works correctly:
![image](https://user-images.githubusercontent.com/6103208/128579683-ba1599f4-956c-49f6-ba94-f69bb8979c8d.png)

How swiftmailer signer adds trailing line return:
```php
    protected function endOfBody()
    {
        // Add trailing Line return if last line is non empty
        if (\strlen($this->bodyCanonLine) > 0) {
            $this->addToBodyHash("\r\n");
        }
        $this->bodyHash = hash_final($this->bodyHashHandler, true);
    }
```
Swiftmailer signer works correctly with not empty body.

From DKIM signature specifications:

- simple canonicalization: https://datatracker.ietf.org/doc/html/rfc6376#section-3.4.3 `If there is no body or no trailing CRLF on the message body, a CRLF is added`
- relaxed canonicalization: https://datatracker.ietf.org/doc/html/rfc6376#section-3.4.4 `If the body is non-empty but does not end with a CRLF, a CRLF is added`

Other issues related to invalid DKIM signature: #39354, #41935, #42407. But is seems they have another problem, which is connected to templated emails.

I have tested dkim signature manually (with gmail) with these cases:

canonicalization: simple
body: ''

canonicalization: simple
body: "\r\n"

canonicalization: simple
body: 'text'

canonicalization: relaxed
body: ''

canonicalization: relaxed
body: "\r\n"

canonicalization: relaxed
body: 'text'

![image](https://user-images.githubusercontent.com/6103208/128577659-1e765fe3-19f7-4ef2-bd59-22064e0d9b73.png)

Commits
-------

0cb61f4 Add trailing Line return if last line is non empty
    @fabpot
    fabpot committed Aug 7, 2021
    65e1463

Commits on Aug 6, 2021

  1. Add trailing Line return if last line is non empty

    @metaer
    metaer committed Aug 6, 2021
    0cb61f4

  2. feature #42198 [Security] Deprecate PassportInterface (chalasr) 

    This PR was merged into the 5.4 branch.

Discussion
----------

[Security] Deprecate `PassportInterface`

| Q             | A
| ------------- | ---
| Branch?       | 5.4
| Bug fix?      | no
| New feature?  | no
| Deprecations? | yes
| Tickets       | -
| License       | MIT
| Doc PR        | -

As explained in #42181, the right extension point is badges, not passports.

Also renames `AuthenticatorInterface::createAuthenticatedToken()` to `createToken()` because of the signature change and the recent abandon of the `authenticated` state for tokens.

Commits
-------

a446030 [Security] Deprecate `PassportInterface`
    @fabpot
    fabpot committed Aug 6, 2021
    810599d

  3. [Security] Deprecate PassportInterface

    @chalasr
    chalasr committed Aug 6, 2021
    a446030

  4. feature #42332 [HttpFoundation] Add litespeed_finish_request to `Re… 

    …sponse` (thomas2411)

This PR was squashed before being merged into the 5.4 branch.

Discussion
----------

[HttpFoundation] Add `litespeed_finish_request` to `Response`

| Q             | A
| ------------- | ---
| Branch?       | 5.4
| Bug fix?      | no
| New feature?  | yes
| Deprecations? | no
| Tickets       | #42293
| License       | MIT

For now Litespeed has dropped support for `fastcgi_finish_request` function due to some problems: php/php-src@ccf051c so when Litespeed is being used on a server instead of Apache, then there is no possibility to use `fastcgi_finish_request` function as the alias for `litespeed_finish_request` is turned off.

Commits
-------

5fb91a5 [HttpFoundation] Add `litespeed_finish_request` to `Response`
    @fabpot
    fabpot committed Aug 6, 2021
    bb68baf

  5. [HttpFoundation] Add litespeed_finish_request to Response

    @thomas2411 @fabpot
    thomas2411 authored and fabpot committed Aug 6, 2021
    5fb91a5

  6. Fix markup

    @fabpot
    fabpot committed Aug 6, 2021
    bf7470b

  7. feature #42286 [HttpFoundation] Add SessionFactoryInterface (kbond) 

    This PR was merged into the 5.4 branch.

Discussion
----------

[HttpFoundation] Add `SessionFactoryInterface`

| Q             | A
| ------------- | ---
| Branch?       | 5.4
| Bug fix?      | no
| New feature?  | yes
| Deprecations? | no
| Tickets       | Fix #41443
| License       | MIT
| Doc PR        | -

Assuming `@jderusse`'s [solution](#41443 (comment)) in #41443 is the best way to add custom session bags by a 3rd party bundle, there should really be a `SessionFactoryInterface` to make decorating the `SessionFactory` easier.

Commits
-------

2079eb5 add SessionFactoryInterface
    @fabpot
    fabpot committed Aug 6, 2021
    fcd3a0c

  8. minor #42352 [Finder] Add type to Finder iterator (simPod) 

    This PR was merged into the 5.4 branch.

Discussion
----------

[Finder] Add type to Finder iterator

| Q             | A
| ------------- | ---
| Bug fix?      |no
| New feature?  | yes <!-- please update src/**/CHANGELOG.md files -->
| Deprecations? | no <!-- please update UPGRADE-*.md and src/**/CHANGELOG.md files -->
| License       | MIT

This improves static analysis so consumers code can know that iterating over `Finder` returns `SplFileInfo` and not `mixed`

Commits
-------

1463c1f Add type to Finder iterator
    @fabpot
    fabpot committed Aug 6, 2021
    d2e2fd4

  9. bug #42382 [Console] Add return type to OutputFormatterInterface::for… 

    …mat() (derrabus)

This PR was merged into the 5.4 branch.

Discussion
----------

[Console] Add return type to OutputFormatterInterface::format()

| Q             | A
| ------------- | ---
| Branch?       | 5.4
| Bug fix?      | yes
| New feature?  | no
| Deprecations? | no
| Tickets       | N/A
| License       | MIT
| Doc PR        | N/A

Spotted while working on #42381. If we look at how `OutputFormatterInterface::format()` is used in the codebase, we can see that an implementation of that method is supposed to return something. Yet the interface does not declare a return value and the `NullOutputFormatter` implementation even has a `void` return type which does not make sense at all, imho.

This PR attempts to fix that.

Commits
-------

d75f5e6 Add return type to OutputFormatterInterface::format()
    @fabpot
    fabpot committed Aug 6, 2021
    f041e03

  10. minor #42371 [DependencyInjection] Remove unused code in `DecoratorSe… 

    …rvicePass` (lyrixx)

This PR was merged into the 5.4 branch.

Discussion
----------

[DependencyInjection] Remove unused code in `DecoratorServicePass`

| Q             | A
| ------------- | ---
| Branch?       | 5.4
| Bug fix?      | no
| New feature?  | no
| Deprecations? | no
| Tickets       |
| License       | MIT
| Doc PR        |

this is also present on 4.4, but I'm not sure about the target branch according to the new merge policy

Commits
-------

a4f3dcc [DependencyInjection] Remove dead code in DecoratorServicePass
    @fabpot
    fabpot committed Aug 6, 2021
    c9f821a

  11. minor #42390 [FrameworkBundle] remove dead conditions in Translation … 

    …Commands (acran)

This PR was merged into the 5.4 branch.

Discussion
----------

[FrameworkBundle] remove dead conditions in Translation Commands

| Q             | A
| ------------- | ---
| Branch?       | 5.4
| Bug fix?      | no
| New feature?  | no
| Deprecations? | no
| Tickets       | -
| License       | MIT
| Doc PR        | -

This is just a trivial removal of unused code I stumbled upon while debugging #42361. In the [original code](https://github.com/symfony/symfony/blob/e617a9b/src/Symfony/Bundle/FrameworkBundle/Command/TranslationDebugCommand.php#L165-L170):

~~~php
$transPaths = [$path.'/translations'];
$codePaths = [$path.'/templates'];

if (!is_dir($transPaths[0]) && !isset($transPaths[1])) {
		throw new InvalidArgumentException(sprintf('"%s" is neither an enabled bundle nor a directory.', $transPaths[0]));
}
~~~

The second part of the condition `isset($transPaths[1])` will **always** evaluate to true, since `$targetPath` is just set 3 lines above but only has a single element.

This check was originally to support legacy paths which was removed in b6eb1f4:
* in [`TranslationDebugCommand.php`](b6eb1f4#diff-67afa5b8860d0df4e44f1e1fc89f444b7ac77de515b698a6824dd5403a0acdbcL187-L194)
* in [`TranslationUpdateCommand.php `](b6eb1f4#diff-a01c7858e84f1868a427634740511da7c8c73e56772baa78bdcd98200d7125c0L180-L187)

Rebased from 5.3 to 5.4, see #42362
/cc `@fabpot`

Commits
-------

22db5ad [FrameworkBundle] remove dead conditions in Translation Commands
    @fabpot
    fabpot committed Aug 6, 2021
    2f8ac04

  12. feature #42392 [HttpFoundation] Mark Request::get() internal (ro0NL) 

    This PR was merged into the 5.4 branch.

Discussion
----------

[HttpFoundation] Mark Request::get() internal

| Q             | A
| ------------- | ---
| Branch?       | 5.4
| Bug fix?      | no
| New feature?  | no
| Deprecations? | yes
| Tickets       | Fix #40984
| License       | MIT
| Doc PR        | symfony/symfony-docs#... <!-- required for new features -->

we should not forget about this :)

Commits
-------

e84efc4 [HttpFoundation] Mark Request::get() internal
    @fabpot
    fabpot committed Aug 6, 2021
    2d42a8a

Commits on Aug 5, 2021

  1. (minor) remove author tags from test classes

    @chalasr
    chalasr committed Aug 5, 2021
    83bdd2e

  2. feature #39601 [Notifier] add SentMessageEvent and `FailedMessageEv… 

    …ent` (ismail1432)

This PR was squashed before being merged into the 5.4 branch.

Discussion
----------

[Notifier] add `SentMessageEvent` and `FailedMessageEvent`

| Q             | A
| ------------- | ---
| Branch?       | 5.4
| Bug fix?      | no
| New feature?  | yes
| License       | MIT
| Doc PR        |

Dispatch a new event  `SentMessageEvent` **which is dispatched once the notification is sent**.  The `Symfony\Component\Notifier\Transport\AbstractTransport`  return an instance of `SentMessage` that contains the original message + an id that [can be returned by the API](https://github.com/symfony/linked-in-notifier/blob/5.x/LinkedInTransport.php#L96) it can be helpful to pass this object to the event.

Dispatch a new event `FailedMessageEvent`  **which is dispatched if sending the notification fails** it can be helpful for a retry strategy

Commits
-------

025256d [Notifier] add `SentMessageEvent` and `FailedMessageEvent`
    @chalasr
    chalasr committed Aug 5, 2021
    7d370f1

  3. [Notifier] add SentMessageEvent and FailedMessageEvent

    @ismail1432 @chalasr
    ismail1432 authored and chalasr committed Aug 5, 2021
    025256d

  4. [HttpFoundation] Mark Request::get() internal

    @ro0NL
    ro0NL committed Aug 5, 2021
    e84efc4

  5. bug #42391 [ErrorHandler] Add missing to semi-colon to exception.js (… 

    …stevemarvell)

This PR was merged into the 5.4 branch.

Discussion
----------

[ErrorHandler] Add missing to semi-colon to exception.js

Without this, the js is reported as in error using at least Firefox on Linux, which ultimately prevents the debug bar appearing.

| Q             | A
| ------------- | ---
| Branch?       | 5.4
| Bug fix?      | yes
| New feature?  | no
| Deprecations? | no
| Tickets       |
| License       | MIT
| Doc PR        |

Commits
-------

ca988fb Add missing to semi-colon to exception.js
    @derrabus
    derrabus committed Aug 5, 2021
    e0a08b5

  6. Add missing to semi-colon to exception.js 

    Without this, the js is reported as in error using at least Firefox on Linux, which ultimately prevents the debug bar appearing.
    @stevemarvell
    stevemarvell committed Aug 5, 2021
    ca988fb

  7. [FrameworkBundle] remove dead conditions in Translation Commands 

    Since $targetPath is directly set just 3 lines above
!isset($targetPath[1]) will *always* evaluate to true.

This check was originally to support legacy paths which was removed in
b6eb1f4
    @acran
    acran committed Aug 5, 2021
    22db5ad

  8. Merge branch '5.3' into 5.4

    @Tobion
    Tobion committed Aug 5, 2021
    ff174c6

  9. Merge branch '4.4' into 5.3

    @Tobion
    Tobion committed Aug 5, 2021
    567f007

  10. minor #42384 [HttpFoundation] Fix broken mock (derrabus) 

    This PR was merged into the 4.4 branch.

Discussion
----------

[HttpFoundation] Fix broken mock

| Q             | A
| ------------- | ---
| Branch?       | 4.4
| Bug fix?      | yes
| New feature?  | no
| Deprecations? | no
| Tickets       | N/A
| License       | MIT
| Doc PR        | N/A

Backport from #42378.

This PR fixes a couple of mocked methods that currently return `null` although a real implementation would never do that.

Commits
-------

af37c64 Fix broken mock
    @Tobion
    Tobion committed Aug 5, 2021
    e6c9155

  11. Fix broken mock 

    Signed-off-by: Alexander M. Turek <me@derrabus.de>
    @derrabus
    derrabus committed Aug 5, 2021
    af37c64

  12. Add return type to OutputFormatterInterface::format() 

    Signed-off-by: Alexander M. Turek <me@derrabus.de>
    @derrabus
    derrabus committed Aug 5, 2021
    d75f5e6

Commits on Aug 4, 2021

  1. Merge branch '5.3' into 5.4

    @Tobion
    Tobion committed Aug 4, 2021
    49a076a

  2. minor #42379 Fix return types for PHP 8.1 (derrabus) 

    This PR was merged into the 5.3 branch.

Discussion
----------

Fix return types for PHP 8.1

| Q             | A
| ------------- | ---
| Branch?       | 5.3
| Bug fix?      | yes
| New feature?  | no
| Deprecations? | no
| Tickets       | #41552
| License       | MIT
| Doc PR        | N/A

Follow-up of #42260.

Commits
-------

ab3c43f Fix return types for PHP 8.1
    @Tobion
    Tobion committed Aug 4, 2021
    e9b764b

  3. Fix return types for PHP 8.1 

    Signed-off-by: Alexander M. Turek <me@derrabus.de>
    @derrabus
    derrabus committed Aug 4, 2021
    ab3c43f

  4. Merge branch '4.4' into 5.3 

    * 4.4:
  Fix test
    @derrabus
    derrabus committed Aug 4, 2021
    bc1b5be

  5. Fix test 

    Signed-off-by: Alexander M. Turek <me@derrabus.de>
    @derrabus
    derrabus committed Aug 4, 2021
    6d0e144

  6. Merge branch '5.3' into 5.4 

    * 5.3:
  Add test for non-callable arrays as controllers
    @derrabus
    derrabus committed Aug 4, 2021
    53c7ab7
Older