Say I have an route with a param like this (in Angular 2): /user1/products/:id, which could have child routes like /user1/products/3/reviews.
When I navigate to /user1/products/-1, I check with my server and if the product doesn't exist, I want to render a 404 page without affecting the browser history.
Using router.navigate(['/404']) or router.navigateByUrl('/404') doesn't seem to work because it adds both /user1/products/-1 and/404 to the browser history.
Meaning when I press the Back button in the browser, I go back to /user1/products/-1 which is immediately redirected to /404 and I'm essentially stuck at /404.
In ExpressJS, we would do something like next() to pass the request to our 404 handler. Is there a client-side equivalent in Angular 2?
Update
In the new Router V3 you can use guards as explained in https://angular.io/guide/router#canactivate-requiring-authentication
Original
I think you should use #CanActivate() to do the check. If you forward in #CanActivate() the invalid URL shouldn't be added to the history (not tried)
See also https://github.com/angular/angular/issues/4112 for how to use DI in #CanActivate()
Ok, this is already implemented, just not well-documented.
According to the docs:
router.navigateByUrl(url: string, _skipLocationChange?: boolean) : Promise<any>
Has a parameter _skipLocationChange that will not modify the history.
These will do the trick:
router.navigateByUrl('/404', true);
router.navigateByInstruction(router.generate(['/404']), true);
As of Angular 2 final this is the solution:
this._router.navigateByUrl('/404', { skipLocationChange: true })
Its really interesting question, perhaps you should report it as feature request. I would be nice to have access to router instruction inside loader callback of RouteDefinition.
You could try to emulate validation adding default route /** and using regex parameter of RouteDefinition to match only positive numbers.
Related
My controller posts a form to create a new page. After posting the form I need to redirect the user to the new page that will have the contents for that page that were entered in the previous form. If I simply do return view('mynewpageview', compact('mycontent')); where my mycontent is the object used to execute the $mycontent->save(); command, I carry the risk for someone refreshing the url thus posting the same content twice by creating a new page.
Instead I would like to redirect the user to the actual page url.
My route is
Route::get('/newpage/{id}', 'PageController#pagebyid'); and if I use return redirect()->route('/newpage/$pageid'); where $pageid = $mycontent->id; I get Route not defined error.
What would be the solution either to stop someone from resubmitting the content or a correct syntax for passing the parameter?
The correct answer that works for me is -
Give your route a name in the routes file
Then pass the parameters with an array as shown below in the controller.
return redirect()->route('newpageid', ['id' => $pageid]);
With basic (unnamed) routes, the correct syntax was return redirect('/newpage/'.$pageid);
You have already found out you can alternatively use named routes.
Last but not least, thanks for having considered the "double submit" issue! You have actually implemented the PRG pattern :)
I need someway to redirect my app to a previous url.
The problem comes when i make a submit that goes wrong, the redirect->back previous url is someway "overwrited" and i cannot get the previous real url anymore, instead the app makes the submit again.
The only thing that i tried is the redirect back, because i can't find another way to do it :S
So i´m wondering if there is a way to achieve that, redirect the app to a previous url without considering the submit fails and all this stuff.
Thank you.
Yo can try with:
return Redirect::to(URL::previous());
You can store URLs in session and then make Laravel redirect 2 or 3 pages back. Simple example of code to store URL in session:
$links = session->has('links') ? session('links') : []; // Get data from session
array_unshift($links, $_SERVER['REQUEST_URI']); // Add current URI to an array
session(compact('links')); // Save an array to session
And example of code for redirecting:
return redirect(session('links')[2]);
I have the need to pass a URL to Magento, where it should redirect the User after completing the logout. To store it, I do the following:
$BackTo = Mage::app()->getRequest()->getParam('backto');
if(!empty($BackTo)) {
Mage::getSingleton('core/session')->setBackTo($BackTo);
}
When needed, I retrieve the URL using Mage::getSingleton('core/session')->getBackTo(). The issue is that, while this works well on login, it doesn't work on logout (where it's most needed). I can store the session variable, I can also immediately retrieve it, but, when I am in logout.phtml, where the redirect JavaScript is located, such variable is set to null.
I suspect that the redirect performed by Magento upon logout has something to do with this "disappearing" session variable, but I can't say for sure.
For completeness, here is the relevant code (there's more code than this, but they are mainly auxiliary functions, which don't get called on logout).
Account Controller
class MyPackage_Redirectplugin_AccountController extends Mage_Customer_AccountController {
/**
* #see AccountController:logoutAction()
*/
public function logoutAction() {
$this->_getSession()
->logout()
->setBeforeAuthUrl(Mage::getUrl());
// Store the "back to" URL in a session variable
$this->StoreBackToURL();
$this->_redirect('*/*/logoutSuccess');
}
protected function StoreBackToURL() {
// Store the value of the "backto" argument, if it was passed
$BackTo = Mage::app()->getRequest()->getParam('backto');
if(!empty($BackTo)) {
Mage::getSingleton('core/session')->setBackTo($BackTo);
// At this point I can see the correct value stored in the session variable
}
}
}
Logout.phtml
// The following command returns null
$redirectURL = Mage::getSingleton('core/session')->getBackTo();
Thanks in advance for the help.
Update 12/09/25 - Found a workaround
Since I couldn't find a way to pass a session variable to the logout page opened by redirect, I chose an alternative way: I'm passing it via the URL. Specifically, I implemented a logoutAction() which calls $this->_redirect('*/*/logoutSuccess', array('myvar' => $MyValue));.
In the template, where I have to do the redirect, I simply read such parameter using Mage::helper('core')->urlDecode(Mage::app()->getRequest()->getParam('myvar'));.
I'm aware that there might have been better ways to implement the whole thing, but I needed a bug fix solution and this does the job. Thanks to all people who answered.
The cleanest option which I see (no controller rewrite necessary!) is to observe the dynamically-dispatched controller_action_postdispatch_customer_account_logout event - see the relevant line fromMage_Core_Controller_Varien_Action::postDispatch(). The postDispatch() method is called after the controller action completes.
Example observer method:
public function logoutRedirect($obs)
{
$redirectUrl = Mage::getUrl(/* url args */);
$obs->getControllerAction()->getResponse()->setRedirect($redirectUrl);
}
Doing this will redirect the user to the desired URL directly upon logout meaning that the logoutSuccess page with the JS redirect will not be accessed.
If the desire is to have the logoutSuccess page render as normal, but redirect to a different URL, this can be achieved a couple of ways:
1. By creating a custom template
2. By creating a custom template block class, overriding the getUrl() method to retrieve the URL of your choice, and assigning that as block to render in the content area (by removing or displacing the customer_logout block) in a custom layout XML update file.
just observe this event customer_logout, and when event call method you save the session key on database.
Learn more: http://www.magentocommerce.com/wiki/5_-_modules_and_development/reference/events
http://www.magentocommerce.com/wiki/5_-_modules_and_development/0_-_module_development_in_magento/customizing_magento_using_event-observer_method
GL.
Update 12/09/25 - Found a workaround
This workaround has also been posted in the question itself.
Since I couldn't find a way to pass a session variable to the logout page opened by redirect, I chose an alternative way: I'm passing it via the URL. Specifically, I implemented a logoutAction() which calls $this->_redirect('*/*/logoutSuccess', array('myvar' => $MyValue));.
In the template, where I have to do the redirect, I simply read such parameter using Mage::helper('core')->urlDecode(Mage::app()->getRequest()->getParam('myvar'));.
I'm aware that there might have been better ways to implement the whole thing, but I needed a bug fix solution and this does the job. Thanks to all people who answered.
I have a tiny application in MVC 3.
In this tiny application, I want my URLs very clear and consistent.
There's just one controller with one action with one parameter.
If no value is provided (that is, / is requested by the browser), then a form is displayed to collect that single value. If a value is provided, a page is rendered.
The only route is this one:
routes.MapRoute(
"Default",
"{account}",
new { controller = "Main", action = "Index", account = UrlParameter.Optional }
);
This all works fine, but the account parameter never appears in the address line as a part of the URL. I can manually type test.com/some_account and it will work, but other than that, the account goes as a post parameter and therefore does not appear. And if I use FormMethods.Get in my form, I get ?account=whatever appended to the URL, which is also not what I want and which goes against my understanding. My understanding was that the MVC framework would try to use parameters set in the route, and only if not found, it would append them after the ?.
I've tried various flavours of setting the routes -- one route with a default parameter, or one route with a required parameter, or two routes (one with a required parameter and one without parameters); I've tried mixing HttpGet/HttpPost in all possible ways; I've tried using single action method with optional parameter string account = null and using two action methods (one with parameter, one without), but I simply can't get the thing appear in the URL.
I have also consulted the Steven Sanderson's book on MVC 3, but on the screenshots there are no parameters either (a details page for Kayak is displayed, but the URL in the address bar is htpp://localhost:XXXX/).
The only thing that definitely works and does what I want is
return RedirectToAction("Index", new { account = "whatever" });
But in order to do it, I have to first check the raw incoming URL and do not redirect if it already contains an account in it, otherwise it is an infinite loop. This seems way too strange and unnecessary.
What is the correct way to make account always appear as a part of the URL?
My understanding was that the MVC framework would try to use
parameters set in the route, and only if not found, it would append
them after the ?
Your understanding is not correct. ASP.NET MVC doesn't append anything. It's the client browser sending the form submission as defined in the HTML specification:
The method attribute of the FORM element specifies the HTTP method used
to send the form to the processing agent. This attribute may take two
values:
get: With the HTTP "get" method, the form data set is appended to the URI specified by the action attribute (with a question-mark ("?")
as separator) and this new URI is sent to the processing agent.
post: With the HTTP "post" method, the form data set is included in the body of the form and sent to the processing agent.
ASP.NET MVC routes are used to parse an incoming client HTTP request and redispatch it to the corresponding controller actions. They are also used by HTML helpers such as Html.ActionLink or Html.BeginForm to generate correct routes. It's just that for your specific scenario where you need to submit a user entered value as part of the url path (not query string) the HTML specification has nothing to offer you.
So, if you want to fight against the HTML specification you will have to use other tools: javascript. So you could use GET method and subscribe to the submit handler of the form and inside it manipulate the url so the value that was appended after the ? satisfy your requirements.
Don't think of this as ASP.NET MVC and routes and stuff. Think of it as a simple HTML page (which is what the browser sees of course) and start tackling the problem from that side. How would you in a simple HTML page achieve this?
Is it possible to prevent direct access to an action in symfony. The action is only accessible by "forward" only. So basically, a way to see if the request is coming from another action.
I'm trying to achieve this because the first action handles plenty of verifications then if it fails, it stays on that action. If it succeed, it will forward to an appropriate action; this action needs to have safe inputs (validated from the first action). In order to keep the code DRY, the second action doesn't need to re-verify all the inputs again.
Then why not doing simply a private method? The second action is sort of a plugin, it's decided on the fly where it's going from the first one, that action has its own set of other future action/template. It makes more sense to simply forward instead of trying to handle plenty of cases that Symfony already takes care of.
There are multiple ways to achieve this.
1) Make sure your action isn't accessible by the routing. If you have wildcard routes this will be harder, but you can always add a route which would point the url for your action to a 404 page. Add something like this to your routing.yml:
disabled_action:
url: /disabledController/disabledAction
params: { module: default, action: error404 }
2) Check the action stack upon executing your action. The action stack let's you know from which action you were redirected. You can access it within your action using $this->getController()->getActionStack(). If the getSize() is bigger than 1 (in a default configuration) you we're forwarded.
Use referrer parameter available in request
$request->getReferer() will give you full url of previous action
I'm curious why you're trying to achieve this. Are you looking to have multiple points of access that forward to this action? What if you simply defined a private method (which by default aren't web-accessible) and called it directly from another action?