This PR was merged into the 4.1 branch.
Discussion
----------
[FrameworkBundle] cleanup generated test container
| Q | A
| ------------- | ---
| Branch? | 4.1
| Bug fix? | yes
| New feature? | no
| BC breaks? | no
| Deprecations? | no
| Tests pass? | yes
| Fixed tickets | -
| License | MIT
| Doc PR | -
Cleans up removed and hidden services, fixes an issue when the private container locator is inlined.
Commits
-------
d8cbec00ad [FrameworkBundle] cleanup generated test container
* 4.0:
do not mock the session in token storage tests
[DependencyInjection] resolve array env vars
Add Occitan plural rule
Fix security/* cross-dependencies
[Lock] Skip test if posix extension is not installed
[DI] Allow defining bindings on ChildDefinition
use strict compare in url validator
Disallow illegal characters like "." in session.name
[HttpKernel] do file_exists() check instead of silent notice
fix rounding from string
* 3.4:
do not mock the session in token storage tests
[DependencyInjection] resolve array env vars
Add Occitan plural rule
Fix security/* cross-dependencies
[Lock] Skip test if posix extension is not installed
[DI] Allow defining bindings on ChildDefinition
use strict compare in url validator
Disallow illegal characters like "." in session.name
[HttpKernel] do file_exists() check instead of silent notice
fix rounding from string
* 2.8:
do not mock the session in token storage tests
Add Occitan plural rule
Fix security/* cross-dependencies
Disallow illegal characters like "." in session.name
fix rounding from string
* 2.7:
do not mock the session in token storage tests
Add Occitan plural rule
Disallow illegal characters like "." in session.name
fix rounding from string
This PR was merged into the 2.7 branch.
Discussion
----------
Disallow invalid characters in session.name
| Q | A
| ------------- | ---
| Branch? | 2.7
| Bug fix? | yes
| New feature? | no
| BC breaks? | no
| Deprecations? | no
| Tests pass? | yes
| Fixed tickets | #27023
| License | MIT
| Doc PR |
PHP saves cookie with correct name, but upon deserialization to
`$_COOKIE`, it replaces "." characters with "_".
This is probably also reason why \SessionHandler is not able to find
a session.
https://harrybailey.com/2009/04/dots-arent-allowed-in-php-cookie-names/https://bugs.php.net/bug.php?id=75883
Commits
-------
16ebb43bd4 Disallow illegal characters like "." in session.name
This PR was merged into the 4.1 branch.
Discussion
----------
[Messenger] implement several senders using a ChainSender
| Q | A
| ------------- | ---
| Branch? | master
| Bug fix? | no
| New feature? | yes
| BC breaks? | no
| Deprecations? | no <!-- don't forget to update UPGRADE-*.md and src/**/CHANGELOG.md files -->
| Tests pass? | yes <!-- please add some, will be required by reviewers -->
| Fixed tickets |
| License | MIT
| Doc PR |
Commits
-------
198925ee4e [Messenger] implement several senders using a ChainSender
This PR was squashed before being merged into the 4.1 branch (closes#27128).
Discussion
----------
[Messenger] Middleware factories support in config
| Q | A
| ------------- | ---
| Branch? | master <!-- see below -->
| Bug fix? | no
| New feature? | yes <!-- don't forget to update src/**/CHANGELOG.md files -->
| BC breaks? | no <!-- see https://symfony.com/bc -->
| Deprecations? | no <!-- don't forget to update UPGRADE-*.md and src/**/CHANGELOG.md files -->
| Tests pass? | yes <!-- please add some, will be required by reviewers -->
| Fixed tickets | N/A <!-- #-prefixed issue number(s), if any -->
| License | MIT
| Doc PR | todo
Following https://github.com/symfony/symfony/pull/26864, this would allow to configure easily the middlewares by using an abstract factory definition to which are provided simple arguments (just scalars, no services references).
For instance, here is how the DoctrineBundle would benefit from such a feature (also solving the wiring of the `DoctrineTransactionMiddleware` reverted in https://github.com/symfony/symfony/pull/26684):
```yaml
framework:
messenger:
buses:
default:
middleware:
- logger
- doctrine_transaction_middleware: ['entity_manager_name']
```
where `doctrine_transaction_middleware` would be an abstract factory definition provided by the doctrine bundle:
```yml
services:
doctrine.orm.messenger.middleware_factory.transaction:
class: Symfony\Bridge\Doctrine\Messenger\DoctrineTransactionMiddlewareFactory
arguments: ['@doctrine']
doctrine_transaction_middleware:
class: Symfony\Bridge\Doctrine\Messenger\DoctrineTransactionMiddleware
factory: ['@doctrine.orm.messenger.middleware_factory.transaction', 'createMiddleware']
abstract: true
# the default arguments to use when none provided from config.
# i.e:
# middlewares:
# - doctrine_transaction_middleware: ~
arguments: ['default']
```
and is interpreted as:
```yml
buses:
default:
middleware:
-
id: logger
arguments: { }
-
id: doctrine_transaction_middleware
arguments:
- entity_manager_name
default_middleware: true
```
---
<details>
<summary>Here is the whole config reference with these changes: </summary>
```yaml
# Messenger configuration
messenger:
enabled: true
routing:
# Prototype
message_class:
senders: []
serializer:
enabled: true
format: json
context:
# Prototype
name: ~
encoder: messenger.transport.serializer
decoder: messenger.transport.serializer
adapters:
# Prototype
name:
dsn: ~
options: []
default_bus: null
buses:
# Prototype
name:
default_middleware: true
middleware:
# Prototype
-
id: ~ # Required
arguments: []
```
</details>
Commits
-------
f5ef421474 [Messenger] Middleware factories support in config
This PR was merged into the 4.1 branch.
Discussion
----------
[Messenger] Add more tests around the AMQP transport
| Q | A
| ------------- | ---
| Branch? | 4.1
| Bug fix? | no
| New feature? | no
| BC breaks? | no
| Deprecations? | no
| Tests pass? | yes
| Fixed tickets | ø
| License | MIT
| Doc PR | ø
Adding more tests to the AMQP transport/factory. These should have captured the following 3 bugs: #27198, #27197, #27196.
Commits
-------
faf9382223 Add more tests around the AMQP transport
This PR was merged into the 4.1-dev branch.
Discussion
----------
[FrameworkBundle][Messenger] Remove convention-based service id generation for buses
| Q | A
| ------------- | ---
| Branch? | master
| Bug fix? | yes (DX)
| New feature? | no
| BC breaks? | no
| Deprecations? | no
| Tests pass? | yes
| Fixed tickets | -
| License | MIT
| Doc PR | -
Service ids for buses are user facing ids: ppl *will* have to deal with them when wiring their services.
Like for cache pools, we should ask ppl to declare them using full identifiers, instead of having a convention-based way to turn a name into and id (which is something nobody will be able to remember at some point.)
Commits
-------
c7a0045cbd [FrameworkBundle][Messenger] Remove convention-based service id generation for buses
This PR was merged into the 4.1-dev branch.
Discussion
----------
[Messenger] Relax messenger config and fix some bugs
| Q | A
| ------------- | ---
| Branch? | master
| Bug fix? | yes
| New feature? | no
| BC breaks? | no
| Deprecations? | no
| Tests pass? | yes
| Fixed tickets | -
| License | MIT
| Doc PR | -
After playing a little with the config of this component I found some bugs around.
Reproducer:
1. Install a fresh Symfony flex project with `^4.0@dev` dependencies
2. Add the `symfony/messenger` requirement
3. Add the following configuration separately:
Note that `symfony/serializer` has not been installed yet. ATM it's not required.
----------------------
Configuring my custom transport (not amqp):
```yaml
framework:
messenger:
transports:
custom: 'my_transport://localhost/msgs'
routing:
'*': custom
```
**Before** (Current behaviour):
Threw a logic exception, IMO unrelated at this point:
> Using the default encoder/decoder, Symfony Messenger requires the Serializer. Enable it or install it by running "composer require symfony/serializer-pack".
**After** (Proposal):
Pass! The Messenger's serializer config is disabled by definition because the Serializer component is not installed yet, then the related (default) encoder/decoder aliases are invalid, so the amqp transport factory service is removed altogether.
----------------------
According to the previous exception I configured my custom encoder/decoder services:
```yaml
framework:
messenger:
encoder: App\Serializer\Serializer
decoder: App\Serializer\Serializer
transports:
custom: 'my_transport://localhost/msgs'
routing:
'*': custom
```
**Before**:
The same exception has been thrown, now a bit vague according to the config:
> Using the default encoder/decoder, Symfony Messenger requires the Serializer. Enable it or install it by running "composer require symfony/serializer-pack".
**After**:
Pass! the serializer is disabled by definition but there is custom encoder/decoder services, so let's keep the amqp transport factory with their custom encoder/decoder deps.
----------------------
Just enabling the serializer option:
```yaml
framework:
messenger:
serializer: true
```
**Before**:
Pass! Unexpected, as there is no transport configuration the exception wasn't thrown and still keeps the amqp transport factory service with invalid encoder/decoder (Serializer) dependencies.
**After**:
The Serializer config & support is verified if it's enabled regardless of the transport configuration and this exception is thrown for this case:
> The default Messenger serializer cannot be enabled as the Serializer support is not enabled.
I've removed the "install" part because at this point only we're checking whether the `framework.serializer` is enabled or not, so the next step after that should be enable the Serializer support in `framework.serializer`, which already verify whether the Serializer component is installed or not. IMO "composer require symfony/serializer-pack" should be there and not here. Also because `symfony/serializer` is not a hard dependency of this component.
----------------------
By last, I disabled the serializer option manually:
```yaml
framework:
messenger:
serializer: false
transports:
custom: 'my_transport://localhost/msgs'
routing:
'*': custom
```
**Before**:
I received this DI exception:
> The service "messenger.transport.amqp.factory" has a dependency on a non-existent service "messenger.transport.serializer".
**After**:
Pass! (same behaviour that the first example)
----------------------
As I explained earlier, this PR enables or disables the Messenger's serializer config based on the current Symfony platform and whether the Serializer component is installed or not, like other config with similar behaviour.
Tests included :)
Cheers!
Commits
-------
a05e2e2c3b Relax Messenger config and fix some bugs
* 4.0:
use brace-style regex delimiters
Fixed typo RecursiveIterator -> RecursiveIteratorIterator
[Cache] fix logic for fetching tag versions on TagAwareAdapter
[FrameworkBundle] Remove dead code
[FrameworkBundle] Use the correct service id for CachePoolPruneCommand in its compiler pass
Hide short exception trace by default
[Doctrine Bridge] fix priority for doctrine event listeners
[Validator] make phpdoc of ObjectInitializerInterface interface more accurate
[Validator] fixes phpdoc reference to an interface that was removed in Symfony 3.0
* 3.4:
use brace-style regex delimiters
Fixed typo RecursiveIterator -> RecursiveIteratorIterator
[Cache] fix logic for fetching tag versions on TagAwareAdapter
[FrameworkBundle] Use the correct service id for CachePoolPruneCommand in its compiler pass
Hide short exception trace by default
[Doctrine Bridge] fix priority for doctrine event listeners
[Validator] make phpdoc of ObjectInitializerInterface interface more accurate
[Validator] fixes phpdoc reference to an interface that was removed in Symfony 3.0
This PR was squashed before being merged into the 4.1-dev branch (closes#26864).
Discussion
----------
[Messenger] Define multiple buses from the `framework.messenger.buses` configuration
| Q | A
| ------------- | ---
| Branch? | master
| Bug fix? | no
| New feature? | yes
| BC breaks? | no
| Deprecations? | no
| Tests pass? | yes
| Fixed tickets | #26652
| License | MIT
| Doc PR | https://github.com/symfony/symfony-docs/issues/9617
Not everybody will benefit from having only one bus, especially with the CQRS-like usages. While keeping the extremely use of use of the single bus, this PR has the following:
- Create multiple buses from the YAML configuration
- Tag middleware only a specific buses
- Register middlewares from the YAML configuration
Even if it's visible in the PR's tests, here's how it will look like, for a completely full-customised version:
```yaml
framework:
messenger:
default_bus: commands
buses:
commands: ~
events:
middlewares:
- validation
- route_messages
- "Your\\Middleware\\Service"
- call_message_handler
```
A few things to note:
1. The YAML configuration creates `messenger.bus.[name]` services for the buses.
2. The YAML configuration for middleware just adds tags to the corresponding middlewares.
3. If the middleware definition does not exists, it creates it. (without any magic on the arguments though, if it isn't auto-wirable, well... "your problem")
4. In the PR, there is this "TolerateNoHandler" middleware that is a great example for event buses
Commits
-------
e5deb8499b [Messenger] Define multiple buses from the `framework.messenger.buses` configuration
Before this fix, only services explicitly configured with public=false were exposed via the test service container:
```
$container->register(PrivateService::class, PrivateService::class)->setPublic(false);
```
but not those explicitly configured as private:
```
$container->register(PrivateService::class, PrivateService::class)->setPrivate(true);
```
nor those implicitly configured as private:
```
$container->register(PrivateService::class, PrivateService::class);
```
This PR was merged into the 4.1-dev branch.
Discussion
----------
[DI][FrameworkBundle] Hide service ids that start with a dot
| Q | A
| ------------- | ---
| Branch? | master
| Bug fix? | no
| New feature? | yes
| BC breaks? | no
| Deprecations? | yes
| Tests pass? | yes
| Fixed tickets | #26630
| License | MIT
| Doc PR | -
Now that services are private by default, `debug:container` should display them by default.
In fact, the public/private concept should not trigger a difference in visibility for this command.
Instead, I propose to handle service ids that start with a dot as hidden services.
PR should be ready.
(For reference, I tried using a tag instead in #26891, but that doesn't work in practice when working on the implementation.)
Commits
-------
cea051ee5e [DI][FrameworkBundle] Hide service ids that start with a dot
This PR was squashed before being merged into the 4.1-dev branch (closes#26941).
Discussion
----------
[Messenger] Allow to configure the transport
| Q | A
| ------------- | ---
| Branch? | master
| Bug fix? | no
| New feature? | ish
| BC breaks? | no
| Deprecations? | no
| Tests pass? | yes
| Fixed tickets | #26900, #26908, #26935
| License | MIT
| Doc PR | ø
We allow users to configure the encoder/decoder used by the built-in adapter(s). This also adds the support of configuring the default's encoder/decoder format and context.
Commits
-------
1a3f0bbb14 [Messenger] Allow to configure the transport
This PR was squashed before being merged into the 4.1-dev branch (closes#26909).
Discussion
----------
[Messenger][DX] Uses the adapter name instead of the service name
| Q | A
| ------------- | ---
| Branch? | master
| Bug fix? | no
| New feature? | no
| BC breaks? | no
| Deprecations? | no
| Tests pass? | yes
| Fixed tickets | ø
| License | MIT
| Doc PR | ø
As a developer, I don't get why I would need to use the service name while I give a name to the adapters. This is basically what this PR means for me in terms of configuration:
```patch
framework:
messenger:
routing:
# We want the command below to be handled async with the "messenger.default_sender" (AMQP producer).
- 'App\Message\Command\CreateNumber': messenger.sender.amqp
+ 'App\Message\Command\CreateNumber': amqp
adapters:
amqp: "%env(AMQP_DSN)%"
```
And in terms of pulling the messages:
```patch
- bin/console messenger:consume-messages messenger.receiver.amqp
+ bin/console messenger:consume-messages amqp
```
Commits
-------
b5afb40441 [Messenger][DX] Uses the adapter name instead of the service name
