I want to validate the entity data separately from making a POST request and display errors in the form. I'm using class-transformer. What's the best way to achieve this with Nest?
Sounds exactly like what the ValidationPipe is for. You decorate your DTO with class-validator decorators, set the #Body() type in your endpoint, and bind the pipe and viola, the request gets validated when sent to the server and returns errors that you can parse if any exist
Related
Currently building a RESTful API with express on my web server, and some routes like the delete route for a document with mongoose ex. await Note.findByIdAndRemove(request.params.id) response.status(204).end() send response statuses with end()
Why do I need to add the .end()? What in these cases, and why cant one just send response.status(204)
With some responses that return json, the response.status(201).json works fine
Only certain methods with Express or the http interface will send the response. Some methods such as .status() or .append() or .cookie() only set state on the outgoing response that will be used when the response is actually sent - they don't actually send the response itself. So, when using those methods, you have to follow them with some method that actually sends the response such as .end().
In your specific example of:
response.status(204)
You can use the Express version that actually sends the response:
response.sendStatus(204)
If you choose to use .status() instead, then from the Express documentation, you have to follow it with some other method that causes the response to be sent. Here are examples from the Express documentation for .status():
res.status(403).end()
res.status(400).send('Bad Request')
res.status(404).sendFile('/absolute/path/to/404.png')
Since all three of these other methods will cause the response to be sent and when the response goes out, it will pick up the previously set status.
I can't understand why do we add properties to request object?
As I know, the request object comes from the front end side and if any information is needed or they have, the frontend side should attach it to the request object and send it to backend side like a cookie or JWT token, etc.
And after we got that information in the backend side we can add/update more properties to the response object, like new JWT token, new session ID, etc.
But I see in backend codes, they also add properties to request object and this is vague to me. I don't know why the backend side should add something to the request? Maybe because of the internal transactions between the different middleware this happens? I mean middleware1 adds something to the request object and sends it to the middleware2 in backend side?
For example I can't understand why do we add something to req.session not res.session? Because as I understand this data should be passed to the frontend side to be added into their next request.
But if there is any other reason for adding properties to req object instead of the res object, please let me know?
Mutating the request object by adding additional properties is the primary method of passing additional data between middlewares within Express. Since the request and response are passed to every middleware, this is the only chain that gives you continuity between middlewares.
(You will also see some passed on the response object as well, but in general it's most common to see it on req).
Note that you need to avoid collisions with Express's properties and methods as well as the underlying connection objects from Node.
It might sound like a repeated question at first, but I've gone through all the blogs/ tutorials/ videos I found but none of them actually says how you run that request. Example: for a RESTful request, you code in NodeJs, hit the route(https://localhost/3000/api/getStudent) and get the response. In code you use router.post('/getStudent', async (req,res) => {
// Check response here or manipulate
})
But in soap after coding all the functions like in here-
https://medium.com/metrosystemsro/with-node-js-wrap-backend-soap-webservices-in-a-restful-api-a96887575046
https://dafabulousteach.wpcomstaging.com/2016/05/19/making-a-soap-call-with-node/
https://betterprogramming.pub/how-to-perform-soap-requests-with-node-js-4a9627070eb6
Where do I call the the function and how do I pass parameters and test it? And how do I check the response?
To place this in the context of what you are asking about, a SOAP service is just a server listening on an address for POST requests that have an XML payload. That's it.
So if your SOAP web service address is http://example.com/service_endpoint then you can call the web service by making a POST HTTP request at this address and sending it a SOAP XML message as payload.
Obviously, the XML message in the request must match something that the service expects and you know how to build that XML either by reading documentation or by using the WSDL of the SOAP web service.
So if you know how to make a POST HTTP request to a REST service address and send it a XML payload (although most commonly for REST you use JSON), you already know how to call a SOAP web service.
Now, for convenience, because SOAP is a protocol, the way you call the service and what it expects as XML payload is described by the WSDL, and you can use tools that read the WSDL and create a client API that you can invoke like any other method in your code. The tools handle the details of the HTTP POST request for you, and also the marshalling of any parameters to XML. This is probably what you found confusing.
So let's take an example.
If you have, say, a service that has an operation named savePerson which accepts firstName and lastName as parameters and is described in the WSDL, then your tooling might generate a client that has such a method and accepts a Person object and you can call it like:
var response = client.savePerson({ "firstName": "Kim", "lastName": "Seokjin" });
or some variation of this. You then get a response that you can read just like any other object. You might get a promise instead, or an event, or whatever way the client chooses for how to work. When you use this code, what happens is that the client performs a HTTP POST request behind the scenes for you and marshals the person object to XML, something like this:
POST /service_endpoint
Host: http://example.com
Content-Type: text/xml; charset="utf-8"
Content-Length: nnnn
<SOAP-ENV:Envelope xmlns:SOAP-ENV="http://schemas.xmlsoap.org/soap/envelope/">
<SOAP-ENV:Body>
<savePerson>
<person>
<firstName>Kim</firstName>
<lastName>Seokjin</lastName>
</person>
</savePerson>
</SOAP-ENV:Body>
</SOAP-ENV:Envelope>
You can obviously send this request yourself with whatever HTTP library you prefer if you don't want to generate a client from the WSDL. But people prefer to use a generated client instead of dealing with these details directly (i.e. making HTTP requests, parsing XML, etc).
Using graphene, when sending an outbound graphql request with a client, how can I intercept + introspect query params and post data?
This request is going from one backend service to another backend service
Both queries an mutations are sent via POST
My use case is that I need to cryptographically sign a data payload(dict) but to do that I need graphene to go through the steps of converting the query int a GET or the mutation into a POST before I can grab the data from the query params or post body.
The input data needed to sign the request is the dict:
data = {
"operationName": "blah",
"variables": {},
"query": "fancy graphql query here"
}
This request data:
is available when calling client.execute
is NOT fully available in the graphene middleware because we do not have access to the string value of the query in the resolve info (at least in the test client)
So the only place that one can access these inputs is right before client.execute is called or by making a custom version of the Client class, that implements an execute method, bake the signing in there, then call super().execute...
I'm trying to validate a webhook via facebook. So facebook hits my url my-url/facebook/receive within my route in nodejs i'd do res.send(req.query['hub.challenge']); to send an http response.
I'm using KoaJS. From what i understand, Koajs merges the request and response object into ctx but when reading through the docs I can't find anything along the lines of ctx.send or similar to send a http response.
Can anyone give me some direction or links.
Thanks.
To send the body of a response, you can simply do ctx.response.body = 'Hello'. There are many aliases attached to ctx, so you don't necessarily have to reference the response or request yourself. Doing ctx.body = 'Hello' would be the same as the code above.
If you wanted to set headers, you would use the ctx.set() method. For example: ctx.set('Content-Type', 'text/plain').
To access the query parameters, you would use ctx.request.query['some-key'] (or simply the alias ctx.query['some-key']).
All of the different request/response methods are documented pretty well at the Koa website along with a list of aliases attached to ctx. I highly recommend you give it a read.