Skip to content

[Messenger] Add a "in-memory://" transport #29097

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Jump to bottom
Merged
merged 2 commits into from
Apr 6, 2019
Merged

[Messenger] Add a "in-memory://" transport #29097

merged 2 commits into from
Apr 6, 2019

Conversation

GaryPEGEOT
Copy link
Contributor

@GaryPEGEOT GaryPEGEOT commented Nov 5, 2018 •
edited
Loading

Q A
Branch? master
Bug fix? no
New feature? yes
BC breaks? no
Deprecations? no
Tests pass? yes
Fixed tickets #29040
License MIT
Doc PR Todo

Add a new InMemoryTransport for test purpose, usable by starting your DSN by in-memory://

@ogizanagi
Copy link
Contributor

ogizanagi commented Nov 5, 2018 •
edited
Loading

Just a random thought: Isn't a null transport just a special case of a memory transport without any listener?

@ogizanagi
Copy link
Contributor

Note regarding #29097 (comment): that would mean making the MemoryTransport MessageBusInterface $bus constructor argument optional or extract dispatch in a dedicated listener.
No strong opinion yet.

@nicolas-grekas nicolas-grekas added this to the next milestone Nov 6, 2018
@javiereguiluz
Copy link
Member

@GaryPEGEOT thanks for this contribution! If this is finally merged, please review the list of Messenger issues and list of pending Messenger pull requests to see if there are issues/PRs conflicting with this change. If there are not, please open an issue in https://github.com/symfony/symfony-docs/issues to track this change. You don't have to provide the docs yourself if you don't want, but this would help us track the change. Thanks!

@ogizanagi ogizanagi mentioned this pull request Nov 10, 2018
Closed
@weaverryan weaverryan mentioned this pull request Mar 12, 2019
Closed
36 tasks
sroze
sroze approved these changes Mar 19, 2019
View reviewed changes
Copy link
Contributor

@sroze sroze left a comment

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Except these stylish details to be changed, looks good to me 👍

@GaryPEGEOT GaryPEGEOT force-pushed the feature/null-transporter branch from 52f0a2c to 09512a5 Compare March 19, 2019 09:48
@stof
Copy link
Member

stof commented Mar 20, 2019

Just a random thought: Isn't a null transport just a special case of a memory transport without any listener?

If you never flush your MemoryTransport, it would leak memory. So using it instead of a NullTransport looks wrong (having to flush something when you expect it to be a blackhole is not intuitive)

fabpot
fabpot reviewed Mar 20, 2019
View reviewed changes
src/Symfony/Component/Messenger/Transport/NullTransport.php Outdated
*/
public function send(Envelope $envelope): Envelope
{
$this->sent[] = $envelope;
Copy link
Member

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

As we accumulate envelopes over time, do we need to implement ResetableInterface to flush the envelopes from one request to the next?

Copy link
Contributor Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

@sroze, @fabpot I've implemented Symfony\Contracts\Service\ResetInterface and I now see that there is an ack an reject method in the Transport interface. Do you think we need to store which message has been ack / nack?

@ogizanagi
Copy link
Contributor

ogizanagi commented Mar 21, 2019 •
edited
Loading

having to flush something when you expect it to be a blackhole is not intuitive)

But it's not just a blackhole (null transport might be misleading in this way): this transport is storing the sent envelopes (to be used in tests for instance). So I would have implemented ResetableInterface on the MemoryTransport and just use a different factory registering it.

@sroze
Copy link
Contributor

sroze commented Mar 23, 2019

@GaryPEGEOT can you implement the ResetableInterface and test that the messages are properly removed?

@GaryPEGEOT GaryPEGEOT force-pushed the feature/null-transporter branch from 1ade719 to 1711e48 Compare March 23, 2019 15:42
@sroze
Copy link
Contributor

sroze commented Mar 23, 2019 via email

@GaryPEGEOT
Copy link
Contributor Author

@sroze I added the getAcknowledged and getRejected methods

ogizanagi
ogizanagi reviewed Mar 23, 2019
View reviewed changes
src/Symfony/Component/Messenger/Transport/NullTransport.php Outdated
*
* @author Gary PEGEOT <garypegeot@gmail.com>
*/
class NullTransport implements TransportInterface, ResetInterface
Copy link
Contributor

@ogizanagi ogizanagi Mar 23, 2019 •
edited
Loading

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Note implementing ResetInterface isn't enough for this to be properly integrated with the framework and stored envelopes to be cleared from one request to the next. The service definition must be tagged with kernel.reset.

But as the service is built using a factory and might use an env var for the DSN, it may not be possible to add the tag conditionally from the extension (compile time). Instead, you could implement ResetInterface on the factory too and keep track of created services to clear them from here (so only the factory will be tagged with kernel.reset).

onEXHovia
onEXHovia reviewed Mar 24, 2019
View reviewed changes
src/Symfony/Component/Messenger/Tests/Transport/NullTransportFactoryTest.php Outdated
*
* @author Gary PEGEOT <garypegeot@gmail.com>
*/
class NullTransportFactoryTest extends \PHPUnit\Framework\TestCase
Copy link
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

maybe worth importing TestCase?
use PHPUnit\Framework\TestCase;

@fabpot
Copy link
Member

fabpot commented Mar 31, 2019

I agree with @ogizanagi that I would not expect a null transport to store things.

@GaryPEGEOT
Copy link
Contributor Author

to avoid duplication with memory listener

What do you mean? thinking

memory transport, my bad!, both this transport and the one in #28746 store message in memory, so are kind of duplicate. Should this one do nothing instead, ie not storing any message?

@weaverryan
Copy link
Member

I think the idea is this: This one SHOULD store in memory, and be called in-memory. And then #28746 could actually use this transport behind the scenes to read the stored messages and send them.

@sroze
Copy link
Contributor

sroze commented Mar 31, 2019

Exactly 👍

@GaryPEGEOT GaryPEGEOT force-pushed the feature/null-transporter branch from 53de888 to 38e1b51 Compare March 31, 2019 18:08
@GaryPEGEOT
Copy link
Contributor Author

I think the idea is this: This one SHOULD store in memory, and be called in-memory. And then #28746 could actually use this transport behind the scenes to read the stored messages and send them.

Renamed to InMemoryTransport, as well as the dsn

@GaryPEGEOT GaryPEGEOT changed the title [Messenger] Add a "null://" transport [Messenger] Add a "in-memory://" transport Mar 31, 2019
sroze
sroze suggested changes Mar 31, 2019
View reviewed changes
@sroze sroze force-pushed the feature/null-transporter branch from c1b34d7 to 83267d2 Compare April 6, 2019 09:17
@sroze sroze force-pushed the feature/null-transporter branch from 83267d2 to 8e01179 Compare April 6, 2019 09:18
@sroze sroze force-pushed the feature/null-transporter branch from 8e01179 to 8f8c82e Compare April 6, 2019 09:32
fabpot
fabpot approved these changes Apr 6, 2019
View reviewed changes
src/Symfony/Component/Messenger/Transport/InMemoryTransport.php
use Symfony\Contracts\Service\ResetInterface;

/**
* Transport that stay in memory. Useful for testing purpose.
Copy link
Member

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

stays

@fabpot
Copy link
Member

fabpot commented Apr 6, 2019

Thank you @GaryPEGEOT.

@fabpot fabpot merged commit 8f8c82e into symfony:master Apr 6, 2019
fabpot added a commit that referenced this pull request Apr 6, 2019
@fabpot
feature #29097 [Messenger] Add a "in-memory://" transport (GaryPEGEOT…
ec18af4 
…, sroze)

This PR was merged into the 4.3-dev branch.

Discussion
----------

[Messenger] Add a "in-memory://" transport

| Q             | A
| ------------- | ---
| Branch?       | master
| Bug fix?      | no
| New feature?  | yes
| BC breaks?    | no
| Deprecations? | no
| Tests pass?   | yes
| Fixed tickets | #29040
| License       | MIT
| Doc PR        | Todo

Add a new `InMemoryTransport` for test purpose, usable by starting your DSN by `in-memory://`

Commits
-------

8f8c82e Make the in-memory transport resettable
fe75920 Add a "null://" transport
Nyholm
Nyholm reviewed Apr 6, 2019
View reviewed changes
Copy link
Member

@Nyholm Nyholm left a comment

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Thank you

#eu-fossa

@nicolas-grekas nicolas-grekas modified the milestones: next, 4.3 Apr 30, 2019
@fabpot fabpot mentioned this pull request May 9, 2019
Merged
@tienvx tienvx mentioned this pull request May 23, 2019
Closed
@onEXHovia onEXHovia mentioned this pull request Mar 16, 2024
Closed
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Reviewers

@fabpot fabpot fabpot approved these changes

@webignition webignition webignition left review comments

@Ocramius Ocramius Ocramius left review comments

@sroze sroze sroze approved these changes

@Nyholm Nyholm Nyholm left review comments

@ogizanagi ogizanagi ogizanagi left review comments

@onEXHovia onEXHovia onEXHovia left review comments

@chalasr chalasr chalasr approved these changes

Assignees
No one assigned
Labels
Feature Messenger Status: Reviewed
Projects
None yet
Milestone
4.3
Development

Successfully merging this pull request may close these issues.