Symfony Dependency injection serivces.yaml configurtion by interface - symfony

I'm looking for solution of my problem.
I wonder if i can do something like this:
App\Product\Domain\ProductSaver: '#App\Product\Infrastructure\Saver\ProductJsonSaver'
But i want to use it too:
App\Product\Domain\ProductSaver: '#App\Product\Infrastructure\Saver\ProductORMSaver'
Depends of which User Interface call it.
For example if i call /product/save/database endpoint i want to use ProductORMSaver and if i call /product/save/file endpoint i want to use ProductJsonSaver. Is it possible? I found solution which rely on parameter name but i want to avoid it.

Related

How can I register a specific singleton instance with Caliburn Micro's Simple Container?

I've registered a type in Caliburn Micro using it's Simple Container as a singleton.
_container.Singleton<MyType>("MyType");
Now I've realized I need to pass in some specific constructor parameters, so I created an instance of the registered class, configured it, then tried to use RegisterSingleton. And realized it doesn't seem to let me pass a specific instance. I've looked at the docs and don't quite understand how this was supposed to work.
How can I get it to use this specific instance for the singleton?
Call SimpleContainer.RegisterInstance with the service type, key and implementation.
For example:
_container.RegisterInstance(typeof(MyType), "MyType", new MyType());
The simple container is documented here.

Where can Symfony services be useful?

There is the example of creating and using a service in the official documentation. At start we create some class, then register it in config/services.yml an then we can use it in our code like this:
$result = $this->get('app.myservice')->myMethod($arg);
//(In the [example][1] it is little bit other code:)
//$slug = $this->get('app.slugger')->slugify($post->getTitle());
But WHAT FOR? while I can just do the SAME like this:
use MyServiceNamespace/MyService
//...
$result = (new MyService())->myMethod($arg);
Where is profit of using Services? Is this just syntax sugar?
Nope. Far from syntax sugar.
You need to have a working understanding of what dependency injection means. Perhaps start by skimming through here: http://symfony.com/doc/current/book/service_container.html
Let's suppose your service needs a doctrine repository to do it's job. Which is better?
class MyController
{...
$userManager = $this->get('user.manager');
OR
$userRepository = $this->getDoctrine()->getManager()->getRepository('MyBundle::User');
$userManager = new UserManager($userRepository);
Your choice but once you have worked through the mechanics of how to add a service then you will never look back.
I should also point out that your sluglfy example requires a use statement and ties you code directly to a specific implementation. If you ever need to adjust your slugification then you need to go back and change all the places where it is used.
// These lines make your code more difficult to maintain
use Something\Slugify;
$slugify = new Slugify();
AS Opposed to
$slugify = $this->get('slugify');
'tIn this case, it's not really relevant. But from a simple design concern, services allow to make a better dependency management.
For instance, if you declare a service relaying on another one, then you won't have to instanciate both of them. Symfony will take care of it.
And since your declaration is centralized, any modification on the way you decide to create your service (= declare it), you won't have to change all the references to the services you changed since symfony will take care of the way it's instanciated when needed.
Another point is the scope of services. This information might be checked, but I think symfony instanciate service once (Singleton) which mean a better memory usage.

How can a Tridion command extension find out the command it extends?

Tridion's user interface allows you to extend specific commands, which is a great way to modify the behavior of certain existing commands. In the configuration file of the editor this is done with a section like this:
<ext:commands>
<ext:command name="TextUnderline" extendingcommand="MyTextUnderline"/>
<ext:command name="TextStrikethrough" extendingcommand="MyTextStrikethrough"/>
I am working on a generic command extension class that can be used to modify the behavior of a number of commands:
<ext:commands>
<ext:command name="TextUnderline" extendingcommand="MyCommandExtension"/>
<ext:command name="TextStrikethrough" extendingcommand="MyCommandExtension"/>
So in this second configuration fragment, we have the same MyCommandExtension extending both TextUnderline and TextStrikethrough.
But now in the JavaScript for my MyCommandExtension, how can I determine which command was originally fired?
MyCommandExtension.prototype.isAvailable = function (selection, pipeline) {
...
console.log(this.properties.name);
...
};
In this scenario the this.properties.name will be logged as a less-than-useful-but-completely-correct:
"DisabledCommand"
I suspect that the information is available somewhere in the pipeline parameter, but haven't found it yet.
How can I find out the original command from MyCommandExtension?
Short answer: I couldn't.
I had to do something similar, and ended up having to extend various commands and set the "current" command as part of my "_execute" call (so I would now call _execute(selection, pipeline, originalCommand) for my command.
N
You cannot find out what the original command is. The assumption is that an extending command is specific to the command it extends and so would know which one it is extending. When creating generic extensions that work on different commands, I can see how it might be useful to know what the configuration would be.
Maybe you could add this as an Enhancement Request?
To work around it for now, you could create a base command with your logic - which takes the name of the command that it extends as a parameter. And then create specific classes for each command you which to extend, which just call the base command and pass in the name.
To put it differently, create a BaseExtendingCommand with all of the required methods - and then both a TextUnderlineExtendingCommand and TextStrikethroughExtendingCommand which call the methods on BaseExtendingCommand (and pass in "TextUnderline" and "TextStrikethrough", respectively, as arguments)

Looking for a good technique for storing email templates

I am building a site in which we are making moderate use of email templates. As in, HTML templates which we pass tokens into like {UserName}, {Email}, {NameFirst}, etc.
I am struggling with where to store these, as far as best practice goes. I'll first show the approach I took, and I'd be really excited to hear some expert perspective as a far as alternate approaches.
I created HTML templates in a folder called /Templates/.
I call a static method in my service layer, which takes in the following arguments:
UserName
UserID
Email
TemplatePath ("~/Templates")
Email Subject
Within the service layer I have my static method SendUserEmail() which makes use of a Template class - which takes a path, loads it as a string, and has a AddToken() Method.
Within my static SendUserEmail(), I build the token list off of the method signature, and send the email.
This makes for a quite long method call in my actual usage, especially since I am calling from the web.config the "TemplatePath", and "Email Subject". I could create a utility that has a shorter method call than the ConfigurationManager.AppSettings, but my concern is more that I don't usually see method signatures this long and I feel like it's because I'm doing something wrong.
This technique works great for the emails I have now, which at the most are using the first 3 tokens. However in the future I will have more tokens to pass in, and I'm just wondering what approach to take.
Do I create methods specific to the email needing to be sent? ie. SendNewUserRegistration(), SendMarketingMaterial(), and each has a different signature for the parameters?
I am using ASP.NET Membership, which contains probably the extend of all the fields I'll ever need. There are three main objects, aspnet_User, aspnet_Mebership and aspnet_profile. If it was all contained in one object, I would have just passed that in. Is there performance concerns with passing in all 3, to get all the fields I need? That is versus just passing in aspnet_User.UserID, aspnet_User.Email, etc?
I could see passing in a dictionary with the token entries, but I'm just wondering if that is too much to ask the calling page?
Is there a way to stick these in a config file of it's own called Templates.config, which has tags like -
<Templates>
<EmailTemplate Name="New User Registration">
<Tokens>
<UserName>
<UserID>
<Email>
</Tokens>
<Message Subject="Hi welcome...">
Hi {UserName}...
</Message>
</EmailTemplate>
</Templates>
I guess the main reason I'm asking, is because I'm having a hard time determining where the responsibility should be as far as determining what template to use, and how to pass in parameters. Is it OK if the calling page has to build the dictionary of TokenName, TokenValue? Or should the method take each in as a defined parameter? This looks out of place in the web.config, because I have 2 entries for and , and it feels like it should look more nested.
Thank you. Any techniques or suggestions of an objective approach I can use to ask whether my approach is OK.
First of all I would like to suggest you to use NVelocity as a template engine. As for main problem I think you can create an abstract class MailMessage and derive each one for every needed message (with unique template). So you will use this like following:
MailMessage message = new UserRegistrationMessage(tokens);
//some code that sends this message
Going this way you force each concrete XXXMessage class to be responsible for storing a template and filling it with the given tokens. How to deal with tokens? The simpliest way is to create a dictionary before passing it to the message, so each concrete message class will know how to deal with passed dictionary and what tokens it should contain, but you also need to remember what tokens it should contain. Another way (I like it more) is to create a general abstract type TokenSet and a derived one for every needed unique set of tokens. For example you can create a UserMessageTokenSet : TokenSet and several properties in it:
UserNameToken
SomeUserProfileDataToken
etc. So using this way you will always know, what data you should set for each token set and
UserRegistrationMessage will know what to take from this tokenSet.
There are a lot of ways to go. If you will describe you task better I think I will try suggest you something more concrete. But general idea is listed above. Hope it helps =)

Updating/Intercepting HttpContext.Current.Request.QueryString

Here's a wierd one. I'm reusing a code base that unfortunately must not be updated. This code makes a call to HttpContext.Current.Request.QueryString. Ideally, I need to push a value into this collection with every request that is made. Is this possible - perhaps in an HTTP Module?
Without using reflection, the simplest way to do it would be to use the RewritePath function on the current HttpContext object in order to modify the querystring.
Using an IHttpModule, it might look something like:
context.RewritePath(context.Request.Path, context.Request.PathInfo, newQueryStringHere!);
Hope this helps!
Ditto Espo's answer and I would like to add that usually in medium trust (specific to many shared hostings) you will not have access to reflection so ... RewritePath will remain your probably only choice.

Resources