Extend and override default lighthouse directive - overriding

Is there anyway that I can override a directive like: src/Schema/Directives/WhereDirective.php for instance this doesn't support some methods on my custom builder, I know I can make another directive and extend this like #myWhere but that's dirty, would be nice to be able to override the #where itself.
I've searched around but nothing was found about this sadly!

I edit my composer.json and manipulate the class mappings. In this example, I wanted to override some cache classes.
"autoload": {
"psr-4": {
"App\\": "app/",
"Database\\Factories\\": "database/factories/",
"Database\\Seeders\\": "database/seeders/",
"Nuwave\\Lighthouse\\Cache\\": "lighthouseV6/cache/"
},
"exclude-from-classmap": [
"vendor/nuwave/lighthouse/src/Cache/CacheKeyAndTags.php",
"vendor/nuwave/lighthouse/src/Cache/CacheKeyAndTagsGenerator.php",
"vendor/nuwave/lighthouse/src/Cache/CacheDirective.php"
]
},
Then create a folder "lighthouseV6/cache" in the root of the project and copy the classes I wanted to override from "vendor/nuwave/lighthouse/src/Cache" inside it.

I found the solution. according to https://lighthouse-php.com/5/custom-directives/getting-started.html#register-directives
When Lighthouse encounters a directive within the schema, it starts looking for a matching class in the following order: 1. User-defined namespaces as configured in config/lighthouse.php, defaults to App\GraphQL\Directives 2. The RegisterDirectiveNamespaces event is dispatched to gather namespaces defined by plugins, extensions or other listeners 3. Lighthouse's built-in directive namespace.
So it did seem like override could be possible, and it was.
I haven't tried first method (App\GraphQL\Directive...) but that probably would work too, I went with the second method the RegisterDirectiveNamespaces event, since I was writing a package.
Make all your directives in the same folder under one namespace eg:
namespace SteveMoretz\Something\GraphQL\Directives;
Now in a service provider (Can be your package's service provider or AppServiceProvider or any service provider you get the idea.) register that namespace your directives are under.
use Illuminate\Contracts\Events\Dispatcher;
use Nuwave\Lighthouse\Events\RegisterDirectiveNamespaces;
class ScoutGraphQLServiceProvider {
public function register(Dispatcher $dispatcher) {
$dispatcher->listen(
RegisterDirectiveNamespaces::class,
static function (): string {
return "SteveMoretz\Something\GraphQL\Directives";
}
);
}
}
That's it so for an example I have overridden the #where directive, first I created a file named as original WhereDirective.php then put these contents in it:
<?php
namespace SteveMoretz\Something\GraphQL\Directives;
use Nuwave\Lighthouse\Scout\ScoutBuilderDirective;
use Nuwave\Lighthouse\Support\Contracts\ArgBuilderDirective;
use Nuwave\Lighthouse\Schema\Directives\BaseDirective;
use Nuwave\Lighthouse\Schema\Directives\WhereDirective as WhereDirectiveOriginal;
use Nuwave\Lighthouse\Support\Contracts\FieldResolver;
class WhereDirective extends WhereDirectiveOriginal
{
public function handleBuilder($builder, $value): object
{
$clause = $this->directiveArgValue('clause', 'where');
// do some other stuff too... my custom logic
return $builder->{$clause}(
$this->directiveArgValue('key', $this->nodeName()),
$this->directiveArgValue('operator', '='),
$value
);
}
}
Now whenever we use #where my custom directive runs instead of the original one, but be careful what you do in this directive don't alter the whole directive try to extend the original and add more options to it, otherwise you would end up confusing yourself later!

Related

Symfony 5 switch cache adapter on condition

I need to switch the Symfony cache adapter depending on ENV conditions. Like if some variable is set, use "cache.adapter.apcu" or use "cache.adapter.filesystem" otherwise.
Is it possible somehow? The documentation is not really helpful with it.
P.S.: It is not possible for us to do this via the creation of a whole new environment
Here is a basic example for a CacheAdapter which has adapters fed into it and then picking one based on a parameter (or alternatively envvar):
<?php
namespace App\Cache;
use Psr\Cache\CacheItemInterface;
use Psr\Cache\InvalidArgumentException;
use Psr\Container\ContainerInterface;
use Symfony\Component\Cache\Adapter\AdapterInterface;
use Symfony\Component\Cache\CacheItem;
use Symfony\Contracts\Service\ServiceSubscriberInterface;
use Symfony\Contracts\Service\ServiceSubscriberTrait;
class EnvironmentAwareCacheAdapter implements AdapterInterface, ServiceSubscriberInterface
{
use ServiceSubscriberTrait;
private string $environment;
public function __construct(string $environment)
{
$this->environment = $environment;
}
public function getItem($key)
{
return $this->container->get($this->environment)->getItem($key);
}
public function getItems(array $keys = [])
{
return $this->container->get($this->environment)->getItems($key);
}
// ...
}
This is how you would configure it:
services:
App\Cache\EnvironmentAwareCacheAdapter:
arguments:
$environment: '%kernel.environment%'
tags:
- { name: 'container.service_subscriber', key: 'dev', id: 'cache.app' }
- { name: 'container.service_subscriber', key: 'prod', id: 'cache.system' }
It's not the most elegant solution and is missing error handling and possibly a fallback. Basically, by adding tags with an appropriately named key and the alias to an existing cache as id, you can then refer to that cache with the key in your own adapter. So, depending on your environment you will pick either one. You can replace the key and the constructor argument with anything else you like. I hope that helps.
It seems like you can not set up your cache configuration to use a environment variable like so:
framework:
cache:
app: %env(resolve:CACHE_ADAPTER)%
It is the constraint of FrameworkBundle that provides the cache service. And this constraint will not be "fixed" (Using environment variables at compile time #25173).
To make it possible you need to make your own cache provider that can just pass all arguments to the needed cache provider. You will have access to environment variables at runtime and so you can use it as a proxy that knows what provider to use.

How to release or distribute an application that uses mikro-orm?

In the configuration I have to specify the paths to .js and .ts files defining entities:
MikroORM.init({
...
entitiesDirs: ["build/entities"],
entitiesDirsTs: ["src/entities"],
});
So, when I will go to release or distribute the application. Will I need distribute the typescript code too? or will I need distribute only the cache generated? or will I need distribute both? or... none?
As of MikroORM v2.2
Now you can work with default metadata provider, it will require entity source files only if you do not provide entity or type options in your decorators (you can use entity callback to use reference to entity class instead of using string name in type, handle for refactoring via IDE like webstorm).
Original answer:
You should ship the typescript code too, and let the cache regenerate on the server - cache would be rebuilt anyway as it checks absolute path to cached entity for invalidation.
You could implement your own cache adapter or metadata provider to get around this, if you don't want to ship the typescript code.
This is how you could implement custom metadata provider that simply throws error when the type option is missing:
import { MetadataProvider, Utils } from 'mikro-orm';
import { EntityMetadata } from 'mikro-orm/dist/decorators';
export class SimpleMetadataProvider extends MetadataProvider {
async loadEntityMetadata(meta: EntityMetadata, name: string): Promise<void> {
// init types and column names
Object.values(meta.properties).forEach(prop => {
if (prop.entity) {
prop.type = Utils.className(prop.entity());
} else if (!prop.type) {
throw new Error(`type is missing for ${meta.name}.${prop.name}`)
}
});
}
}
Then provide this class when initializing:
const orm = await MikroORM.init({
// ...
metadataProvider: SimpleMetadataProvider,
});
The value of type should be JS types, like string/number/Date... You can observe your cached metadata to be sure what values should be there.
Also keep in mind that without TS metadata provider, you will need to specify entity type in #ManyToOne decorator too (either via entity callback, or as a string via type).

Getting a list of tagged services in my controller

What i want is to add services to the service container that i want to use later in my controller or service.
So i created two services with my custom tag fbeen.admin
here they are:
services:
app.test:
class: AppBundle\Admin\TestAdmin
tags:
- { name: fbeen.admin }
fbeen.admin.test:
class: Fbeen\AdminBundle\Admin\TestAdmin
tags:
- { name: fbeen.admin }
Now i want to use all the services with the tag fbeen.admin in my controller but i dont know how.
I followed the How to work with service tags tutorial but i get stuck on this rule:
$definition->addMethodCall('addTransport', array(new Reference($id)));
On some way the addTransport method of the TransportChain class should be called but it seems that it isn't been called.
And even if it would be called then i still do not have a list of services with the fbeen.admin tag into my controller.
I am sure that i am missing something but who can explain me what it is?
p.s. I know that compilerPass runs at buildtime but for example sonata admin knows all admin classes and twig knows all twig extensions. How do they know?
Thank you for reading this :-)
Symfony 3.3
Container gets compiled once (in debug more often, but in production only once). What you manage with addMethodCall... is that once you request your service from container, which you are storing in $definition (that in this case is controller). Then container will call method addMethodCall('method'.. during initialising your service.
What it will look in container:
// This is pseudo content of compiled container
$service = new MyController();
// This is what compiler pass addMethodCall will add, now its your
// responsibility to implement method addAdmin to store admins in for
// example class variable. This is as well way which sonata is using
$service->addAdmin(new AppBundle\Admin\TestAdmin());
$service->addAdmin(new AppBundle\Admin\TestAdmin());
return $service; // So you get fully initialized service
Symfony 3.4+
What you can do is this:
// Your services.yaml
services:
App/MyController/WantToInjectSerivcesController:
arguments:
$admins: !tagged fbeen.admin
// Your controller
class WantToInjectSerivcesController {
public function __construct(iterable $admins) {
foreach ($admins as $admin) {
// you hot your services here
}
}
}
Bonus autotagging of your services. Lets say all your controllers implements interface AdminInterface.
// In your extension where you building container or your kernel build method
$container->registerForAutoconfiguration(AdminInterface::class)->addTag('fbeen.admin');
This will tag automatically all services which implement your interface with tag. So you don't need to set tag explicitly.
The thing to note here is this: The CompilerPass doesn't run the 'addTransport' (or whatever you may call it) in the compiler-pass itself - just says 'when the time is right - run $definition->addTransport(...) class, with this data'. The place to look for where that happens is in your cache directory (grep -R TransportChain var/cache/), where it sets up the $transportChain->addTransport(...).
When you come to use that service for the first time - only then is the data filled in as the class is being instantiated from the container.
This worked for me:
extend the TransportChain class with a getTransports method:
public function getTransports()
{
return $this->transports;
}
and use the TransportChain service in my controller:
use AppBundle\Mail\TransportChain;
$transportChain = $this->get(TransportChain::class);
$transports = $transportChain->getTransports();
// $transports is now an array with all the tagged services
Thank you Alister Bulman for pushing me forwards :-)

Override a symfony service tag with a compiler pass

I'm trying to override a tag in a symfony service definition with a compiler pass. The service as an example would be data_collector.translation.
The goal is to deactivate the data collector service to disable the element in the symfony web developer toolbar. To do this, I have to set the priority of the data_collector tag to 0.
I could also override it in my own service definition:
services:
data_collector.translation:
class: 'Symfony\Component\Translation\DataCollector\TranslationDataCollector'
tags:
- {name: 'data_collector', priority: '0'}
arguments: [#translator.data_collector]
But as I want to do this for a few of the data collectors, I would need to know the mandatory arguments for the data collector definition. The priority works the same for all collectors and therefore I would only need the name of the collector to disable it.
So I wrote the following compiler pass:
class DataCollectorCompilerPass implements CompilerPassInterface
{
public function process(ContainerBuilder $container)
{
if (!$container->hasDefinition('data_collector.translation')) {
return;
}
$definition = $container->getDefinition('data_collector.translation');
$tags = $definition->getTags();
$tags['data_collector'][0]['priority'] = 0;
$definition->setTags($tags);
$container->setDefinition('data_collector.translation', $definition);
}
}
To make things more wired: When I run this command:
$ php app/console container:debug --show-private --tag='data_collector'
I get the following output:
data_collector.translation #WebProfiler/Collector/translation.html.twig translation 0 Symfony\Component\Translation\DataCollector\TranslationDataCollector
So the priority even in the debugger is set to 0.
But for which reason ever the element is still shown in the toolbar.
What did I do wrong here? Is there another mechanism for overwriting a tag within a compiler pass?
The compiler pass does run (tested it with printing out stuff)
I'm using Symfony 2.7.1
Turns out the code does work, the only problem is, that the CompilerPass is run after the ProfilerPass which is part of the FrameworkBundle. Putting my bundle with the CompilerPass before the FrameworkBundle in the AppKernel solves the problem (more information here). For not even initiating the data collectors it's better to remove all tags instead of just setting the priority to 0.
That's what the final solution looks like:
class DataCollectorCompilerPass implements CompilerPassInterface
{
public function process(ContainerBuilder $container)
{
$collectorsToRemove = [
'data_collector.form',
'data_collector.translation',
'data_collector.logger',
'data_collector.ajax',
'data_collector.twig'
];
foreach($collectorsToRemove as $dataCollector) {
if (!$container->hasDefinition($dataCollector)) {
continue;
}
$definition = $container->getDefinition($dataCollector);
$definition->clearTags();
}
}
}
Can you try this?
if (!$container->hasDefinition('data_collector.form')) {
return;
}
$definition = $container->getDefinition('data_collector.form');
$definition->clearTags();
$container->setDefinition('data_collector.form', $definition);
Why not use your compiler pass to manipulate directly the service Definition of the service holding all these collectors ?
If I look at the compiler pass responsible for loading the data collector, it seems that they are all injected using a method call injection.
You could use your compiler pass to rewrite the method call array using methods like setMethodCalls, removeMethodCall, ... of the Definition entity.
The method call manipulation documentation : link

Using modules in Meteor.js with Typescript

Folks, I'm trying to do something that I thought ought to be simple, but I must be doing something wrong. I'm trying to simply have a clear structure in my meteor application which uses Typescript.
Here are my requirements:
All interfaces are available in both client and server
Some class implementations are only available on the server
I don't want to rely on file load order for my application to work properly
I need my own module to not clash with global objects (such as the Position class for example)
I need to have one monolithic include file for server, one for both client and server and one for client (don't want to have 10s of includes on top of my files)
The setup that I have right now is this
server
server-book.ts
client
shared
collections.ts
definitions
server
include.d.ts (includes all .d.ts files in this folder)
server-book.d.ts (server specific implementation of book)
client
shared
include.d.ts (includes all .d.ts files here)
book.d.ts (book interface definition)
collections.d.ts
In each .d.ts file I have
module MyModule {
interface Bla {}
};
In each .ts file that defines a class I have:
module MyModule {
export class MyBla implements Bla {};
}
All .d.ts files generated for classes are generated by tsc -d.
No .ts files are being included via ///<reference> rather only .d.ts files.
Now, when I run this, I get an error that MyModule is undefined:
/// <reference path="shared/include.d.ts"/>
/// <reference path="server/include.d.ts"/>
Meteor.startup(() => {
var temp = new MyModule.ServerBook();
});
The error occurs right on MyModule.
What am I doing wrong? What should be the proper setup here?
Thanks!
I have dealt with this issue on my blog. I decided to use the evil eval command, since it gave me the easiest possibility of using modules till something more sophisticated appears.
File /lib/foo.ts is position in the subdirectory since it has to be loaded before Bar.
eval('var Hugo = (this.Hugo || (this.Hugo = {})'); // this will override the automatically emitted var Hugo and assigns it with globally defined Hugo module
module Hugo {
export class Foo {
foo():string {
return 'foo'
}
}
}
File /bar.ts
/// <reference path="lib/foo.ts"/>
eval('var Hugo = (this.Hugo || (this.Hugo = {})'); // this will override the automatically emitted var Hugo and assigns it with globally defined Hugo module
module Hugo {
export class Bar extends Foo {
bar () : string {
return 'bar';
}
}
}
File /test.ts
/// <reference path="lib/foo.ts"/>
/// <reference path="bar.ts"/>
var m = new Hugo.Bar();
console.log(m.bar());
console.log(m.foo());
As mentioned here, for classes, the solution is even simpler:
class ExportedClass {
variable : int;
}
this.ExportedClass = ExportedClass;
Definition files should use the declare keyword. You would normally get an error if you didn't use this keyword.
declare module MyModule {
export interface Bla {}
}
And
declare module MyModule {
export class MyBla implements Bla {
}
}
It is also worth checking that the ServerBook class has the export keyword (just like MyBla in your examples).
After lot of trial and errors, here are my findings so far :
Using typescript "module" keyword doesn't get well with Meteor. I think at the moment you cannot use it (or the workarounds are too complicated for me).
However, here is what you can do :
Let say that you have package A where you want to define a class ClassToExport which you want to make public.
class ClassToExport {
getFoo(){
return "foo";
}
}
Please note that you can't write this.ClassToExport = ClassToExport and
api.export('ClassToExport') or else ClassToExport won't be available in the global scope of package A, hence the need for a module/namespace for exporting your class, which we will see next.
Now, for the class to be available for the consumers of your package, you have to create a namespace, which will be the equivalent of the "module" typescript keyword for internal module.
So let's write :
declare var packageA; //so that the compiler doesn't complain about undeclared var
packageA = packageA || {}; //so that this namespace can be reused for the entire package
packageA.ClassToExport = ClassToExport; //the actual export
Now, don't forget to write
api.export('packageA') in the package.js of package A
If you have a package B where you want to use ClassToExport, you write in package B:
var cte = new packageA.ClassToExport();
without forgetting to api.use package A in package B's package.js
If you don't want to write the namespace each time you use the class, you can also write var ClassToExport = packageA.ClassToExport; at the top of your using file.
If you need a global class for you package only, without exporting it, then you can do instead just :
this.ClassToExport = ClassToExport
and again don't write api.export('ClassToExport'), or it won't be available in the package anymore.
This way, i think the features (export/ import) of internal typescript modules are there.
If you are not afraid of gulp build, I have prepared a typescript boilerplate project which allows you to comfortably use typescript from within your app, not depending on packages.
https://github.com/tomitrescak/meteor-boilerplate-typescript
Random idea, what about extend Meteor instead of Window.
Meteor.yournamespace = Meteor.yournamespace || {};
Meteor.yournamespace.myclass = new MyClass();
or
Meteor.yournamespace.MyClass = MyClass();
I think this is less invasive than go directly to the window object IMHO. my two cents.
now you can do Meteor.yournamespace.MyClass :P
--EDIT
Then you could create a meteor-extend.d.ts file and do something like:
/// <reference path="main.d.ts" />
declare module Meteor {
var yournamespace: any;
}
Now you can remove the <any> before Meteor and Typescript will not complaint.

Resources