In using Ember Data for my models, there are some cases where I need to work around the data limitations and access other quasi-restful URLs on my server.
For example, I have a Feed object that records a stream of data. For accessing the models I have a RESTful endpoint:
/feeds/:feed_id
In order to start and stop recording a feed, I need to send a PATCH to a url like:
/feeds/:feed_id?update_action=start
Subsequently I can reload my model and see the changes reflected therein.
In this case, I need to access $.ajax and the URL is the same as the one Ember would use. However, I can't figure out how to eke this information out of Ember.
So far, the best I can do is:
DS.Model.reopen
rootForModel: Ember.computed( ->
#.store.adapterForType(#).serializer.rootForType(#.constructor)
)
pluralRootForModel: Ember.computed( ->
#.store.adapterForType(#).serializer.pluralize(#get("rootForModel"))
)
Such that for an instance of App.FeedItem I can do:
this.get("rootForModel") # feed_item
this.get("pluralRootForModel") # feed_items
And I'm guessing this would stay in sync with any settings made in the Adapter etc.
Subsequently, I can call like:
$.ajax
url: #get("pluralRootForModel") + "/" + #get("id")
data:
update_action: "start"
type: "PATCH"
Is this totally out in left field? Is there a more direct way to compose these URLs?
Another (related issue) is getting the underscored name for a given model.
App.MyModelController => my_model_controller
I've done something like:
Ember.Object.reopenClass
###*
* The underscored name for this.
* i.e. App.MyClass -> my_class
* From an instance, use this.constructor.underscored_class_name()
* #return {String} This classname, underscored.
###
underscored_class_name: ->
_.underscored("#{#}".replace(/^.*?\./g, ""))
Is this crazy? Are there any better ways?
Check out buildURL in DS.RESTAdapter.
If you want to use underscores in server paths and keys, check out DS.ActiveModelAdapter (and its default serializer, DS.ActiveModelSerializer). This adapter has its own implementation of buildURL.
Related
The models tree of my emberJS app match the tree of my API however, I have 2 different routes returning the same type of data :
/products/ and /users/:id/supported_products/ both return products data.
When I need to have the products of the app there is no problem :
this.store.query('product',params);
However I am not sure how to query products from the user path. The place to do so would be the adapter, but I need to define a secondary adapter that I would call when I need supported products,and I have no idea how to do so.
I think if it were me I would create a virtual query parameter that would instruct a custom adapter on how to change the endpoint on the fly.
For example I might have a supportedByUser flag. Then in my app/adapters/product.js do something like this:
import JSONAPIAdapter from 'ember-data/adapters/json-api';
export default JSONAPIAdapter.extend({
urlForQuery(query, modelName) {
let userId = query.supportedByUser;
delete query.supportedByUser;
return userId
? `${this.namespace || ''}/users/${userId}/supported_products`
: this._super(...arguments);
}
});
Here is an example twiddle demoing this: https://ember-twiddle.com/b406391e98ed4fda30bc227a894fa7c9
In my Ember.js Application, I am dealing with query params for list updates. I have one strange use case, in which I don’t the URL to be updated with certain query params. How can I achieve this?
I assume you want to reload your model with parameters that are different than the ones in your application route? And you keep your application route parameters synced using queryParams?
In your route's model function you can filter your model data by the same query params (that appear in the address bar) but you can add some logic that extracts additional parameters either from the controller or other place and these parameters the data fetching query. Example:
model: function(queryParams) {
var params = queryParams;
params.additional_filter = this.controllerFor('mycontroller').get('additional_filter');
return this.store.find('mymodel', params);
}
Also if you want to explicitly reload the model you will need to call Router.refresh() function.
I'm trying to get autocomplete working in my rails application using Magic Suggest.
I think this is the correct question: How can I get MagicSuggest to grab the JSON that is at the URL I give it?
This is the error that console returns when I type letters:
POST http://localhost:3000/search_foods 404 (Not Found) jquery.js:8706
Uncaught Could not reach server
Here's the magic suggest code:
input.magicSuggest({
data: "/foods/search/",
placeholder: "Search Foods...",
valueField:'idFood',
displayField:'foodName'
});
The Routes
resources :search_foods
The Controller and Action
class SearchFoodsController < ApplicationController
def index
render json: %['Crack', 'Cocain', 'Gorilla Test', 'Horse Test']
end
end
When I visit the /search_foods url directly I get
'Crack', 'Cocain', 'Gorilla Test', 'Horse Test'
as my code is designed to do.
I think the issue is in that MagicSuggest, by default, sends a POST request, although I'm not sure if that's entirely relevant:
You can pass the url from which the component will fetch its JSON data.Data will be fetched
* using a POST ajax request that will * include the entered text as 'query' parameter. The results
* fetched from the server can be:
* - an array of JSON objects (ex: [{id:...,name:...},{...}])
* - a string containing an array of JSON objects ready to be parsed (ex: "[{id:...,name:...},{...}]")
* - a JSON object whose data will be contained in the results property
* (ex: {results: [{id:...,name:...},{...}]
Try this:
input.magicSuggest({
data: "http://localhost:3000/search_foods",
placeholder: "Search Foods...",
valueField:'idFood',
displayField:'foodName'
});
The doc states that the component expects one of the following:
* - an array of JSON objects (ex: [{id:...,name:...},{...}])
* - a string containing an array of JSON objects ready to be parsed (ex: "[{id:...,name:...},{...}]")
* - a JSON object whose data will be contained in the results property
* (ex: {results: [{id:...,name:...},{...}]
When you visit /search_foods you get
'Crack', 'Cocain', 'Gorilla Test', 'Horse Test'
This does not fit any of the 3 supported cases.
My suspicions about the POST request was correct.
My friend helped out so that's why I was able to fix this.
This is what I did..
Eliminated the FoodSearch Controller, because that's not needed at all.
Created a search action in my Food Controller like so:
def search
render json: ['cocain', 'crack', 'gorilla testosterone']
end
Edited my routes to a POST request instead of a get *This was the key:
resources :foods do
collection do
post :search
end
end
--- Another option, as karlipoppins suggests, is to simply change the type of request that magicSuggest makes by including the method attribute like so:
input.magicSuggest({
method: 'get',
data: "/foods/search/",
placeholder: "Search Foods...",
valueField:'idFood',
displayField:'foodName'
});
Then I wouldn't need to change the route to post.
Added this path to the data attribute in the js
data: "/foods/search/"
This will be a huge help to anyone trying to get magicSuggest to work in rails. It's a pretty damn easy setup to be honest. That one bit and the JSON formatting was the only thing that was tripping me up here.
Is there currently (in the latest builds) a way of specifying a URL on a model-by-model basis? in Ember Data 1.0 beta? I have found some questions on SO and issues on Github around this, but most are out-dated.
For example, I have a model that's called App.PaymentSearchResult and rather than having the request go to /payment_search_results I would like it to go to /payments/search. Where would I override the URL used for a given model (rather than overriding buildURL on the RESTAdapter)?
You can override the the find adapter
but it's kind of hackish, i think however i would take another approach. Idealy you want your Ember models to reflect your backend's models, so why would you need a PaymentSearchResult? When you probably already have a Payment model?
If you need to search in your payment records, why not handle it using query params?
http://emberjs.com/guides/models/finding-records/#toc_querying-for-records
this.store.find('payment', { total: "22" });
Then you want to answer accordingly on the server.
If you want to do a search which returns multiple models, you do this with a manual ajax request.
var self = this;
$.get( "/search", { name: "John", time: "2pm" }, function(result) {
self.store.pushMany(result);
});
PushMany assumes a sane JSON structure.
http://emberjs.com/api/data/classes/DS.Store.html#method_pushMany
Is there a nice way to force Ember Data to load the resource from server eaven if it has it already in store ?
I have a simple show user action that do store.find('users',id) the model is loaded only once at first attempt to display a page the second time i go my model is loaded from the store which is normal ember data behaviour i know. However i need to load it each time.
edit:
the only way i found is to do this :
#store.find('user',{id: params.user_id}).then (users)->
users.get('firstObject')
however it forces me to implement a "fake" show action on my index action ...
I think this... http://emberjs.com/api/data/classes/DS.Model.html#method_reload
model.reload()
Good luck
Additionally you can call getById which will return any instance of that record that exists, or null, then call unloadRecord to remove it from the cache. I like Edu's response as well though, then I wouldn't have to worry about the record existing somewhere else. Maybe I'd use getById then reload that way any references that had a reference to the user got updated. (pardon my errant coffeescript, if it's wrong).
user = #store.find('user', params.user_id)
if (user)
#store.unloadRecord(user)
Hot off the presses, thanks to machty:
There's a new method getting added as part of the query params feature going into beta this weekend called Route.refresh()...
/**
Refresh the model on this route and any child routes, firing the
`beforeModel`, `model`, and `afterModel` hooks in a similar fashion
to how routes are entered when transitioning in from other route.
The current route params (e.g. `article_id`) will be passed in
to the respective model hooks, and if a different model is returned,
`setupController` and associated route hooks will re-fire as well.
An example usage of this method is re-querying the server for the
latest information using the same parameters as when the route
was first entered.
Note that this will cause `model` hooks to fire even on routes
that were provided a model object when the route was initially
entered.
#method refresh
#return {Transition} the transition object associated with this
attempted transition
#since 1.4.0
*/
You can do this in the setupController hook, using a promise, and the reload method mentioned by Edu.
setupController: ->
#store.find('myModel', 1).then (myModel) ->
myModel.reload()
If you are sure that records to display will change after a certain action then you can call this.refresh() method in your Route. For example:
ProductsRoute = Ember.Route.extend
model: ->
#store.find 'product',
activated: true
actions:
accept: (product) ->
if not product.get('activated')
product.set 'activated', true
product.save()
.catch (err) ->
console.log err
product.rollback()
.then =>
#refresh()
ignore: (product) ->
if not product.get('ignored')
product.set 'ignored', true
product.save()
.catch (err) ->
console.log err
product.rollback()
.then =>
#refresh()
If actions are called from child route - e.g. products/proposed - models will be reloaded for parent route and also child routes.
I think that what you are looking for is DS.Store#fetchById