I have a form in my React/Relay app that I am using to modify some fields. I don't want to send a server update every time a new letter is typed into an input. How can I use Relay to support the application state without always pushing to the server? Having read through most of the Relay docs it seems to me that I have to basically copy the Relay state either to the local state of my form or to some other Flux store, deal with changing it in there, and then commit it into Relay. That seems like a lot of extra work though just to keep a local state.
Relay current version is a glue between graphQL and React, it only handles the data from server.
...it seems to me that I have to basically copy the Relay state either
to the local state of my form...
I fail to see the problem. The way React works, requires you to store state for the whole form or for each input separately anyway.
If you add edit functionality (render form and fetch data to populate inputs for user to change them), all the Relay data is available as this.props.RelayFragmentName on form level anyway.
You pass it down to inputs where you assign it to state and circle is complete. You could also feed inputs directly from props (without assigning data to input state) and change form state from inputs via methods.
Now when user makes the changes and confirms it, you'll just collect all the state again to make a mutation.
I see it as a complete circle that is basically unbreakable if all the defaults are set.
Relay has imperative API to get, add or update cache directly but it's relatively dirty and rarely used (especially for local state). It's used for features like WebSockets where you want to push updates from server to client and update cache manually.
Current Relay is meant to work with server data only. Relay 2 has a local cache but it's not out yet.
If we're talking about different things or you could use some code samples, let me know.
In relay modern you can use commitLocalUpdate to do that.
For example
onEmailChange = (email) => {
commitLocalUpdate(environment, (proxy) => {
const userProxy = proxy.get(this.props.user.__id);
userProxy.setValue(email, 'email');
});
}
Related
Hopefully the title is clear, I couldn't find a better name but if someone can improve it please update it, thanks.
I would like the Firebase database to write on a node if a certain condition is met. For example, if one node receives an input from a client (say an angular app) then another node in the database should write certain data, something like a callback that is fired when a node receives some data.
I know there are 4 rule types (.read .write .validate .indexOn), what I am thinking of is some kind of .callback rule that is fired and writes on a node after some other node has received an input.
Obviously this can be achieved via a server side script but Firebase is about a server-less approach so I am trying to understand what are its current limits and what I can do with it.
Thanks for your responses
firebaser here
Running the multi-location update client-side or on a server-side process that you control are currently the only ways to accomplish this.
There is currently no way to trigger updates based on modifications to the database on Firebase servers. It is no big secret that we've been working on such functionality for a while now, but we have made no announcement as to when that will be available.
Also see Can I host a listener on Firebase?, which (I realize now) is probably a duplicate.
I have a object that i want to store for a moment. The object is in a controller for now, the controller will generate a view. A AJAX request is made from the view to next controller. For that moment i need the object previously stored. Previously, i used session and it worked well. But not sure it is the right thing to do. Is session the answer for this or is there anything else?
I have used cache also.but as per the cache concept.It will access for all the users.So one user data will be override to another.So the cached object data will be change for the same user.I need to handle the data storage for an particular user(Independent).
How is it possible? anyother approach is there please share me.
In Controller I have used Httpcontext.cache["key"]=dataset;
but some one suggested like this.but its not displaying
Explain:
In Controller: httpcontext.current.cache is not coming.
HttpContext.Currenthandler and HttpContext.Currentnotification properties only coming.So How can we handle the temp data storage in MVC.
Please help me.
You could use TempData if you want to store data for the next request only. If data should be accessible between multiple requests, then use Session. Here is short explanation of each one with examples.
As Alex said you could use TempData but if you want to use the data in multiple request, you could use TempData.Keep("YourKey") after reading the value to retain the data for the next request too. For your Information TempData internally uses Session to store your data (temporarily)
I would recommend URL parameters for a HTTP Get, or hidden form fields for a HTTP Post, if this is short lived. This is highly about avoiding the session.
But if it should really persist, then a database might be a reasonable location. Imagine a shopping cart that you don't want to dump just because a session timed out; because you'd like to remind the user next time about items they still haven't purchased.
Why not use the session? I don't generally recommend using the session, as you could find yourself with a global variable that two different browser windows are manipulating. Imagine a glass. One window is trying to fill it with Ice Tea. Another window is trying to fill it with Lemonade. But what do you have? Is it Lemonade? Is it Ice Tea? Or is it an Arnold-Palmer? If you try to put too much stuff on the session, and overly expect it to just be there, you might create an application that is non-deterministic if heaven forbid a user opens a second window or tab, and switches back and forth between the windows.
I'm more ok with Temp Data, if you truly have no other options. But this is not for persisting data for more than a second. Temp data will disappear after the first request reads it, as in, it's meant for a very temporary usage.
I personally only use TempData if I have to do a redirect where I can't otherwise keep it with me, or if I need to have that data for say generating a PDF or image that is going to be called via a HTTP Get by a viewer on the actual page, and then only if the model data is too large for the GET url ( many browsers only support just over 2000 characters, which long description or many fields could fill up.)
But again, pushing items around in hidden form variables, or in url parameters can be safe, because you have no multiple window use conflicts (each carries around its own data for peace of mind.)
I'm using Shippinglogic to gather tracking information for submitted tracking numbers.
I'm handling a number of things behind the scenes of the UI, but I'm not sure how to properly organize this.
So here's the flow:
User submits tracking number either via form input or URL (example.com/track/1234567890). If the number doesn't already exist in the database, then the next step happens...
After a number is submitted, I run the number through some logic to determine who the carrier is (UPS, FedEx, USPS, DHL, etc). The user never specifies...it's all done automatically.
After the carrier is determined, then I need to make the actual call to the carrier API (via Shippinglogic) to get tracking information.
After I get the tracking details, I need to save it to the database.
Then, the tracking details are finally returned to the user.
Since users can submit either via form or via a URL (without any sort of POST action), I'm trying to run it all through my show method in the controller where I check if the number exists and if not, submit it via Number.create(:tracking_number => '1234567890') but once I get into the model, I just kinda get lost on what to do next.
Well I would have the users directed to the new or create actions where you can handle creation and detect if the record already exists. Once that's handled you most likely want to send them off to the show page where you can display the tracking information from your data source and any information you have saved in your database. This way you are preserving the nature of the application and other developers would be able to work with the application if they need to.
Edit:
I had a project like this and I move my detection code out into a separate function inside the model so I could make changes to it and abstract it from a specific call on the model. I performed my API requests in the background on the model so I could cache data in the database and refresh the records that were deemed active once an hour.
Basically if it needed to use the data from the record or save some data as part of the record I made a function in the model. This enabled me to split a bunch of functions out from specific modifications to controller actions and the like.
I can set javascript variables.. but can I retrieve them? if so. how?
I think what you really asking how you can retrieve client-side state, right? The process of "setting" a variable is actually downloading code into the browser which runs and has that side-effect. To get state back from the browser, be that Javascript variables or DOM element values, requires the opposite, i.e. to upload something.
What you most likely want is do XHR requests within your Javascript back to the server to transmit the required state. Make sense?
I want to trigger an API call to fetch some supplemental data and display it alongside my react final form. What's the best way to do this?
I've started by using OnChange from react-final-form-listeners to listen for a field change and make an API call, but I don't really know how to store this data somewhere and display it within the react-final-form framework. Using a functional component and a State hook, it would be relatively trivial, because I would just set the state of a variable and then display it somewhere in the same component. Is this a sign that I should be storing these kinds of things in redux? Can I use hooks instead?
I forked one of the examples and added what I am trying to do:
https://codesandbox.io/embed/react-final-form-field-arrays-k9pqm
You're talking a little too abstractly about your goal here, but sure, you could keep the loaded data in state with a useState() hook. If I were you, I'd want to debounce it, and keep track of whether or not the Promise from your API call was the Promise from the last call or not.