Symfony2 HttpKernel Before and after methods in controllers (Kohana Style) - php

I know that this problem was discussed so many times but I had found some solution and I'm not sure if it is the best and the most efficient approach.
My problem: I'm using fosuserbundle to handle user authentication and I would also like to prevent displaying of login, password resetting, etc. form for logged in users. Below I put some approaches:
The first one (which has been already implemented) based on the kernel events, there is a code
https://gist.github.com/walmen/871c13014b80c6a3d05d
The second approach which was mentioned by my colleague based on the method overloading (removing listeners and duplicate code in the each method which has some logic which shouldn't be displayed for logged in users)
Write custom annotation, i.e. #RequireAnonymous
As I mentioned before, I've already implemented the first approach but I'm not sure if it is the best and the most efficient solution (this listener would be called for each request - it is not too heavy load for application? How listeners affect the application, if they?).
The second approach is the easiest one of course but...code duplication doesn't sound really nice.
The last one might be the best but If we take a look on this example https://gist.github.com/cystbear/1391850 we will see that there is also problem with calling event during any controller call.
Any advice or other ideas with good arguments and explanation?

If I'm understanding you correctly, you can use the third approach with the JMSSecurityExtraBundle which is included with Symfony Standard.
On the action's you want to exclude from authenticated users you can do like so:
/**
* #Secure(roles="IS_AUTHENTICATED_ANONYMOUSLY")
*/
public function fooAction()
{
// stuff...
}
This ensures that the user requesting the specific path is not authenticated.

#Ramon not really because every user has the role IS_AUTHENTICATED_ANONYMOUSLY, even the authenticated ones.
What is more we don't want to throw an exception like "Access denied" or something but we want to 'hide' those pages thanks to the redirection.
What do you think about this https://github.com/FriendsOfSymfony/FOSUserBundle/issues/996#issuecomment-14812806 ?

Related

DDD and blameable

How do you handle situation with blameable in the DDD way?
Ofcourse we can ignore some things, but i think that when entity need some tracking (creator, updater, time updated / created) it should be in the class that actually performs some actions on entity.
For example we have post and user, what whould be the correct way?
$post = new Post();
$post->create(); // here we can set some created_id and
other attributes by using mixins or traits like some fw do
Or it is better like this:
$user->createPost($post);
$user->update($post);
As for me second is better, even when we need to track changes that does not apply to post directly, for example:
$post->doSomethingWithPost();
$user->updatePost($post);
Seems like blameable just throws out one important entity - user who manages some things on entities.
Ofcourse we should not overcomplicate things, but usually when blameable is implemented, entity from which you will get id is a logged in user, that is incorrect to the bounded context.
Here it is some Blogging Context, where user of this context updates post and not some authenticated user.
Whats your thoughts on this one? Is there some similar questions that i maybe missed?
All your examples seem like they are not designed with the DDD principles in mind. The first indicator to me is the usage of a $user variable. In 99% of the cases this is too generic to really capture the intent of a given Model. I think there are hidden concepts that would first have to be made explicit. I think along the lines of RegisteredAuthor and Administrator. At least that's what I understand from:
Here it is some Blogging Context, where user of this context updates post and not some authenticated user.
Another question is how can a "user of this context" not be authenticated? How do you know who he is?
In general in an application that explicitly requires User management we normally have something like an IdentityContext as a supporting Sub Domain. In the different contexts we then have other Models like Author or BlogAdministrator holding a reference to the User's identity (UserId) from the IdentityContext. The Red Book has some nice examples on how to implement this.
To answer the question on how to track who changed something and when:
This concept is also referred to as Auditability, which in most revenue relevant parts of system is actually a must once your organization is reaching a certain size. In this scenario I actually always recommend an Event Sourcing approach since it comes with auditability batteries included.
In your case it would actually be enough to either capture the executing UserId as Metadata to the commands like WritePostCommand or ChangePostContentsCommand or use the UserId in a RequestContext object that knows about the execution context (who was sending this command, when was it sent, is this user allowed to execute this use case).
You can then, as Alexander Langer pointed out in the comments, just use this metadata inside your Repositories or Handlers to pass the information to the Aggregates that need it, or could even just send them to an audit log to not pollute your Domain Model with this responsibilities.
NOTE: Generally I would not use the DoctrineExtensions like Blameable in your Domain Model. They depend heavily on Doctrine's Event system, and you do not want to tie your Model into an Infrastructure concern.
Kind regards

CakePHP: Filter content based on logged in user

I'm building a small CakePHP app with user authentication.
What is the best practice for filtering content by user so that users only see their own stuff?
I've seen this post but it hardly seems like an efficient option as it implies a condition is to be added to every action
CakePHP: Filter Index() with session's user ID
Is it recommended to add a beforeFilter in the AppModel which would filter by the authenticated user's ID or does Cake have a built-in way of doing this?
There's no in-built way of doing this, and there's any number of solutions that could work.
The best way all depends on your situation. In some cases, doing it on a per-action basis, like the question you linked to, would be best. In other cases a more generic solution, such as writing a beforeFind method for your Model/s, or even for your AppModel, would be better.
Which solution is best depends on your exact requirements. Eg, if users can absolutely, always, only ever view their own content, and if every table is linked to a user via a user_id field, then a more generic solution will work.
However, if it's more varied and there are only some actions, of some controllers, that need to be restricted based on the current logged in user, then a less generic approach (such as the one you linked to) is needed.

Symfony: REST web-service for bots and humans - open questions

I'm adding an API to a Symfony-application which should act as a REST web-service. But there are a few open issues.
Different URIs for bots?
I often read the "suggestion" to use URIs like /api/:id/[...], but I think they wouldn't be REST-compliant: No matter whether bot or human - the same unique resource is identified.
I'm asking, since my statement above makes sense, but I don't expect all the others to be at fault.
Modifying existing controllers?
There are several reasons why I need a separate controller-logic for both cases:
No session-login in the case of a api-requests
different Symfony-forms have to be created (For instance, no widgets are required, at all.)
JSON / XML instead of HTML output
I don't want to modify existing controllers. According to the Open-Closed Principle, classes should be open for extension but closed for modifications, and the controller classes are already in use in a "production"-environment.
My idea is to use an extra HTTP header-field (e.g. "X-UseApi"). The routing should call different actions by evaluating it. Is this possible inside routing.yml? How? Do you have other ideas?
Authentication
This is how I implemented bot-authentication:
$user = Doctrine_Core::getTable('sfGuardUser')->findOneByUsername($params['user']);
if($user->checkPassword($params['password']))
{
//...
}
But the code looks like a workaround to my eyes. Are there better solutions for the whole REST authentication issue? Is the sfGuardPlugin / sfDoctrineGuardPlugin not meeting conditions for such use cases?
Thanks in advance and cheers,
fishbone
my way of doing this would be to use sf_format in routes to distinguish between robot and human (robot will probably need to consume XML whereas human will want HTML.
I would alter my controllers in a way that I would delegate the logic to separate classes depending on what format is requested (this shouldn't be too much work and you would get the flexibility you need).
As for authentication - please provide a bit more information on how do you do it now - the example isn't enough for me to get the general idea of how your implementation works.
Different URIs for bots?
I suggest to not worry too much about URIs. There are more problems with them and thinking too much about it just results in losing time. IMHO it would be great if there would be standardized conventions how to define RESTful URIs. Here is an article about it: http://redrata.com/restful-uri-design/ . You can see that every way of designing your uris has its pros and cons.
But today I would reject the statement that 'api/...' isn't REST compliant. I would just avoid it.
Controller and authentication
Finally, my solution was to implement some sfFilters with responsibilities as follows:
ApiAccessFilter: sets request-attribute 'isApiRequest' if X-ApiKey is defined as header field.
ApiKeyAuthFilter: identifies a user by X-ApiKey, calls signIn / forwards to login-action.
SecureApiAccessFilter: Checks whether the current user has credential
'apiWriteAccess', if HTTP-method is POST, PUT or DELETE.
The first filter allows me to call $request->getAttribute('isApiRequest') later in my actions. That's similar to isXmlHttpRequest(). Finally I came to the conclusion that I have to modify existing actions, because requirements have changed due to the web-service extension.
Cheers, fishbone

PHP OOP - How to handle authorisation?

I'm building a management system for an idea I have. I'm well versed in PHP (at least enough to do everything I need to do) but I'm not that experienced with using OOP. I use it as much as I can but a lot of the best practices I'm not familiar with so when I do things I worry I'm doing them in the wrong order.
For this project I have a class for the thing the user is managing, I need to check whether or not the user has permissions to manage it. I know how to check the permissions, my question is: where should I be doing it?
Should I be doing it outside the class, like so:
if user permissions are valid
initialize class
else return error
or should I be doing
initialize class
class checks permissions
class returns error if permissions are invalid
I'm unsure which is the correct approach. On the one hand checking within the class seems the best based on what I know of OOP methodology, but then I also have the feeling that letting it get as far as initializing the class when permissions are unknown might be bad.
How should I be doing it? If there's any sort of article that covers this sort of thing a link would be greatly appreciated (I can't find anything through searches but I'm not 100% sure if I'm searching for the right thing as I know little of OOP)
It depends on what is your permissions model, and there is no "one correct way" to do it. It's a matter of approach. The important thing, is that whatever you choose, you use it consistently.
In my latest projects, I came across several different models. One of the most straightforward is a page-based permission, which is good if you do page-based flow, and use objects for support: you define at the top of the page who is supposed to access it and in case you can redirect. This is the simplest one, but can be very useful on specific applications.
If you, on the contrary, use objects to do your main flow, you should secure your object methods (rather than class instantiation). If you have a "save()" method, which can be called by specific users only, first thing when you enter that method, do your permissions check.
I am currently using an MVC pattern, and I have a Controller, which dispatches the actions to its children. Its only public method is execAction($params) and it will call actionAction($params) on itself, but first it will check permissions.
One important thing to remember is: never present actions on the UI that the user is not allowed to do (unless you are trying to force him to buy your "PRO version", that is) ;-)
I have written a pretty solid and robust CMS system. Here's how I do it and I hope you can extrapolate some information on making your own solution.
I have an index file, which loads an Admin class. Functionality in my CMS is modular (so containing in its own file and class).
A module is loaded based on a $_GET parameter.
Because the function to check the $_GET parameter and load the corresponding function is in my Admin class ($admin->pick_method()) and I also have a User object in my Admin class, I can first check the requested module is in the currently logged in user's permissions array.
If the permissions check returns true, I load the module. If false, I display a friendly "unauthorized access" page.
Hope this helps.
I think best way to do is to have class of permissions.
And then you can check before creating object or in object.
create permission class
if access then create class and set permission object
else error
// do action
if have permissions show something
else do not show something
Look how done in zend acl component
Validating within the class generates a dependency between the class and the permissions authorisation which isn't good because you are tying together two potential disparate items. You can solve this by Inversion of Control (eg. dependency injection) or as I do by authorisation notifications using Emesary.
Validating outside of the class is possibly worse because there is a real danger that the authorisation check will be wrong or missed completely. It also creates the wrong sort of linking between objects as the object is not in control of itself.
If you are going to do it outside of the object when it is created then it is probably better to require the object to provide an interface such as IAuthorisable which can be verified by a seperate object.
e.g.
interface IAuthorisable
{
public function getRequirements();
}
class Authorisation
{
public static createObject($newObj)
{
if ( canDo( $newObj->getRequirements()) )
return $newObj;
return null;
}
}
class Something implements IAuthorisable
{
public function getRequirements()
{
return SomeSortOfIdentifierOrSomething;
}
}
$mySomething = Authorisation::createObject(new Something($p1, $p2), "
If $mySomething is null it isn't allowed. Obviously this needs extending and designing properly which is left as an exercise for the reader.
OO basics
If it makes sense to include it in the class. You could do it.
Does a Book class have an authenticate function? No, functions like download() and convertToPDF() make more sense.
My approach
I always try to find the route of the least resistance.
If there are 10 scripts/pages that talk to 1 class and 1 or 2 scripts needs authentication for certain actions i would build the authentication into those 1 or 2 scripts. (or put them in a subfolder with a .htpasswd)
But when you're using a MVC structure, everything is a class, so its become a question of deciding which class.
I tend to put the authentication rules in the Controllers classes and the authentication to database-logic in a User class.

PHP framework URL conventions

A lot of frameworks use URL conventions like /controller/action/{id} which is great, but if you need any configuration beyond that, it's up to you to write your own routes.
How would you handle URLs like /users/{id}/friends on the backend? (to list all of a user's friends)
I'm thinking that in the controller, something like this would be appropriate:
class User {
function index() {
echo 'user index';
}
}
class Friend extends User {
function index($user_id) {
echo 'friend index';
}
}
Then you would have the following map:
/users -> User::index()
/users/{id} -> User::view($id)
/users/{id}/friends -> Friend::index($user_id)
I wanted to put the Friend class inside the User class but apparently you can't do that in PHP so this is the best I could come up with. Thoughts?
What URL would use for editing your list of friends? /users/{id}/friends/edit could work, but it doesn't seem appropriate, since you should never be editing someone else's friend list. Would /account/friends/edit be a better choice? Where would you put the corresponding code for that? In a friend controller, or a user controller, or a specialized account controller?
Bonus question: which do you prefer? /photos/delete/{id} or /photos/{id}/delete
The answers:
So, what I've gathered from the answers is that if the "thing" is complicated (like "friends") but doesn't have its own controller, you can give it one without a model, or if it's not, you should stuff it in with whatever it's most closely related to. Your URLs should not influence where you put your code. Most people seem to think you should stick to /controller/action/{id} whever possible, because it's what people are familiar with.
No one really commented on the extended class aside from saying it's "awkward". Perhaps FriendList would have been a more appropriate class in that case if I really wanted to separate it out.
Thanks for all the answers :)
The routes you're talking about, and the way you're using subclasses to achieve this structure, seems a bit awkward to me. The standard convention of /controller/action/{id} works great for simple actions, but if you're creating a complex application you will always need to create custom routes. There are probably some good guidelines to use when creating these routes, but it really boils down to staying consistent across your application and keeping things as simple as possible.
I don't see any good reason to have /user/{id}/friends mapping to a "Friend" controller. Why not just have "friends" be an action on the User controller? Once you actually drill down to view a specific friend's page, you could use a Friend controller (/friends/view/123) or you could repurpose your User controller so that it works for a friend or the currently logged in user (/user/view/123).
Re: the bonus question, I'd stick with /photos/delete/{id} (/controller/action/{id}) as that's the most widely accepted mechanism.
I would prefer /photos/{id}/delete. My reasoning is that if you take one component off the end of an URL, it should still make sense.
It's pretty easy to assume what /photos/{id} should do: view the set of photos for that {id}.
But what should /photos/delete do? That's really unclear.
I know that there's kind of a default convention of /controller/action/id, but that organization is for the sake of mapping to the class/method architecture of controllers. I don't think it's a good idea to organize the UI to accommodate the code (the URL is in a way part of the UI).
Re comments: Yes, /photos/{id} maybe makes more sense to view a given photo by its id. /users/{id}/photos perhaps to view a collection. It's up to you.
The point is that you should think of the UI in terms of users, not in terms of code organization.
You can do either or. The problem is when you mix the two. /users/{id}/friends and /users/friends/{id} When someone has the id of "friends" this will fail. This may seem like a trivial case but it's very popular to use usernames for ids. You will have to limit user names for every action.
Sometimes you can't do /{controller}/{action}/{id}
I did a indie music site a while back and we did
/artist/{username}
/artist/{username}/albums
/artist/{username}/albums/{album}
We didn't want to test for conditionals so we didn't do
/artist/{username}/{album}
Since we didn't want to check for anyone with an album named "albums"
We could have done it
/artist/{username}
/artist/{username}/albums
/albums/{album}
but then we would lose the SEO advantage of having both the artist name and the album name in the URL. Also in this case we would be forcing album names to be unique which would be bad since it's common for artist to have album names the same as other artist.
You could do pure /{controller}/{action}/{id} but then you would lose some SEO and you can't do URL shortening.
/artist/view/{username}
/artist/albums/{username}
/album/view/{album}
Getting back to your example.
/users/{id}/friends/edit could work,
but it doesn't seem appropriate, since
you should never be editing someone
else's friend list.
In this case it should be /friends/edit since your user id is duplicate information assuming your in a session somehow. In general you want to support URL shortening not URL expansion.
(Bonus question)
Neither, i'd use REST. DELETE /photo?id={id}
It also depends on how you are storing your data. I could imagine in some cases you need a 'friend-list' to be a entity in your model. A logical approach would then be to specify a unique identifier for each friend-list, a primary key.
This would logically result in the following route, as you only need a primary key of the friend-list to edit or delete it...
/friends/edit/{friendListId}
It's up to you to decide. As pix0r stated: convention for small applications is /{controller}/{action}/{id} where {id} should be optional to match with most of your websites actions. In some cases applications get big and you want to define specific routes with more than 3 elements. In some cases certain entities just get a bigger meaning (above example) and you could decide to define a custom controller for it (which makes the default route perfect again...).
I'd stick with the default route /controller/action/id but just don't start making controllers for everything (like friends) in the beginning. The Model-View-Controller pattern makes it very easy for you to change routes later on, as long as all your route-links and actions (forms etc.) are generated based on routes and actions. So you don't really have to bother that much :)
The URLs themselves don't really matter too much. What is more important is what goes in each of your controllers. In your example you had your friend list extend the User class. If your list of friends is really just a list of users, maybe it should extend the Users controller so that you deal with lists of users in one place.
class Users {
public function index() {
$users = $this->findUsers();
}
protected function findUsers($userId=null) { ... }
}
class Friends extends Users {
public function index($userId) {
$users = $this->findUsers($userId);
}
}
If you have a hard time figuring out which class to extend write out what you need from each of the classes and pick the one with the longest list.

Categories