I am trying to do something like this -
const getData = (req, ctx) => (accounts) => R.compose
( R.andThen((result) => {
result.map( item => {
const xx = [...new Set(accounts.filter(x => x.id=== item.id).map(item => item.prop2))]; // need to use accounts variable i.e. the data parameter.
... do some code here based on xx
return result;
})
, callSomeApi(req, ctx)
, R.map(R.prop('id'))
);
let accounts = [ {id: '123", ....} , .... ]; // aaray of some data with property id
const testData= await getData(req,ctx) (accounts);
As per the above program, I will be passing accounts array to func getData and this param would act as data/tenet to Ramda func. I need to access this variable later also in pipe func when api returns the result. But the above syntax doesn't work and fails on first step itself i.e. - R.map(R.prop('id'))
What is the correct way to achieve this? I know one possible way is to create a separate func for filtering after receiving response but I want to understand if this is achievable in the single pipe func ?
I feel as though you're trying to squeeze something where it doesn't belong. Ramda certainly has some techniques to help make point-free code. But if the work doesn't fall into the sweet spot of a composition pipeline, then it's probably better not to try to force it.
I'm sure this isn't an exact match for what you want to do, but the above code looks to me as though it can be written with something like
const getData = (req, ctx) => (accounts) =>
callSomeApi (req, ctx) (pluck ('id') (accounts))
.then (map (item => pluck ('prop2') (uniq (filter (propEq ('id') (item)) (accounts)))))
Here we use a number of Ramda helpers, but we don't try to organize the whole thing into a pipeline, and I believe that simplifies things. (Even if I'm missing something (what else is supposed to happen with xx?), I think it's a simple enough structure to build on.
Related
I have been trying to use the value from the JSON that I have got added successfully using fs.write() function,
There are two test cases in the same fixture, one to create an ID and 2nd to use that id. I can wrote the id successfully in the json file using fs.write() function and trying to use that id using importing json file like var myid=require('../../resources/id.json')
The json file storing correct id of the current execution but I get the id of first test execution in 2nd execution.
For example, id:1234 is stored during first test execution and id:4567 is stored in 2nd test execution. During 2nd test execution I need the id:4567 but I get 1234 this is weird, isn't it?
I use it like
t.typeText(ele, myid.orid)
my json file contains only id like {"orid":"4567"}
I am new to Javascript and Testcafe any help would really be appreciated
Write File class
const fs = require('fs')
const baseClass =require('../component/base')
class WriteIntoFile{
constructor(orderID){
const OID = {
orderid: orderID
}
const jsonString = JSON.stringify(OID)
fs.writeFile(`resources\id.json`, jsonString, err => {
if (err) {
console.log('Error writing file', err)
} else {
console.log('Successfully wrote file')
}
})
}
}
export default WriteIntoFile
I created 2 different classes in order to separate create & update operations and call the functions of create & update order in single fixture in test file
Create Order class
class CreateOrder{
----
----
----
async createNewOrder(){
//get text of created ordder and saved order id in to the json file
-----
-----
-----
const orId= await baseclass.getOrderId();
new WriteIntoFile(orId)
console.log(orId)
-----
-----
-----
}
}export default CreateOrder
Update Order class
var id=require('../../resources/id.json')
class UpdateOrder{
async searchOrderToUpdate(){
await t
***//Here, I get old order id that was saved during previous execution***
.typeText(baseClass.searchBox, id.orderid)
.wait(2500)
.click(baseClass.searchIcon)
.doubleClick(baseClass.orderAGgrid)
console.log(id.ordderid)
----
----
async updateOrder(){
this.searchOrderToUpdate()
.typeText(baseClass.phNo, '1234567890')
.click(baseClass.saveBtn)
}
}export default UpdateOrder
Test file
const newOrder = new CreateOrder();
const update = new UpdateOrder();
const role = Role(`siteurl`, async t => {
await t
login('id')
await t
.wait(1500)
},{preserveUrl:true})
test('Should be able to create an Order', async t=>{
await newOrder.createNewOrder();
});
test('Should be able to update an order', async t=>{
await update.updateOrder();
});
I'll reply to this, but you probably won't be happy with my answer, because I wouldn't go down this same path as you proposed in your code.
I can see a couple of problems. Some of them might not be problems right now, but in a month, you could struggle with this.
1/ You are creating separate test cases that are dependent on each other.
This is a problem because of these reasons:
what if Should be able to create an Order doesn't run? or what if it fails? then Should be able to update an order fails as well, and this information is useless, because it wasn't the update operation that failed, but the fact that you didn't meet all preconditions for the test case
how do you make sure Should be able to create an Order always runs before hould be able to update an order? There's no way! You can do it like this when one comes before the other and I think it will work, but in some time you decide to move one test somewhere else and you are in trouble and you'll spend hours debugging it. You have prepared a trap for yourself. I wrote this answer on this very topic, you can read it.
you can't run the tests in parallel
when I read your test file, there's no visible hint that the tests are dependent on each other. Therefore as a stranger to your code, I could easily mess things up because I have no way of knowing about it without going deeper in the code. This is a big trap for anyone who might come to your code after you. Don't do this to your colleagues.
2/ Working with files when all you need to do is pass a value around is too cumbersome.
I really don't see a reason why you need to same the id into a file. A slightly better approach (still violating 1/) could be:
const newOrder = new CreateOrder();
const update = new UpdateOrder();
// use a variable to pass the orderId around
// it's also visible that the tests are dependent on each other
let orderId = undefined;
const role = Role(`siteurl`, async t => {
// some steps, I omit this for better readability
}, {preserveUrl: true})
test('Should be able to create an Order', async t=>{
orderId = await newOrder.createNewOrder();
});
test('Should be able to update an order', async t=>{
await update.updateOrder(orderId);
});
Doing it like this also slightly remedies what I wrote in 1/, that is that it's not visible at first sight that the tests are dependent on each other. Now, this is a bit improved.
Some other approaches how you can pass data around are mentioned here and here.
Perhaps even a better approach is to use t.fixtureCtx object:
const newOrder = new CreateOrder();
const update = new UpdateOrder();
const role = Role(`siteurl`, async t => {
// some steps, I omit this for better readability
}, {preserveUrl:true})
test('Should be able to create an Order', async t=>{
t.fixtureCtx.orderId = await newOrder.createNewOrder();
});
test('Should be able to update an order', async t=>{
await update.updateOrder(t.fixtureCtx.orderId);
});
Again, I can at least see the tests are dependent on each other. That's already a big victory.
Now back to your question:
During 2nd test execution I need the id:4567 but I get 1234 this is weird, isn't it?
No, it's not weird. You required the file:
var id = require('../../resources/id.json')
and so it's loaded once and if you write into the file later, you won't read the new content unless you read the file again. require() is a function in Node to load modules, and it makes sense to load them once.
This demonstrates the problem:
const idFile = require('./id.json');
const fs = require('fs');
console.log(idFile); // { id: 5 }
const newId = {
'id': 7
};
fs.writeFileSync('id.json', JSON.stringify(newId));
// it's been loaded once, you won't get any other value here
console.log(idFile); // { id: 5 }
What you can do to solve the problem?
You can use fs.readFileSync():
const idFile = require('./id.json');
const fs = require('fs');
console.log(idFile); // { id: 5 }
const newId = {
'id': 7
};
fs.writeFileSync('id.json', JSON.stringify(newId));
// you need to read the file again and parse its content
const newContent = JSON.parse(fs.readFileSync('id.json'));
console.log(newContent); // { id: 7 }
And this is what I warned you against in the comment section. That this is too cumbersome, inefficient, because you write to a file and then read from the file just to get one value.
What you created is not very readable either:
const fs = require('fs')
const baseClass =require('../component/base')
class WriteIntoFile{
constructor(orderID){
const OID = {
orderid: orderID
}
const jsonString = JSON.stringify(OID)
fs.writeFile(`resources\id.json`, jsonString, err => {
if (err) {
console.log('Error writing file', err)
} else {
console.log('Successfully wrote file')
}
})
}
}
export default WriteIntoFile
All these operations for writing into a file are in a constructor, but a constructor is not the best place for all this. Ideally you have only variable assignments in it. I also don't see much reason for why you need to create a new class when you are doing only two operations that can easily fit on one line of code:
fs.writeFileSync('orderId.json', JSON.stringify({ orderid: orderId }));
Keep it as simple as possible. it's more readable like so than having to go to a separate file with the class and decypher what it does there.
I have a Serverless Lambda function that, in response to an S3 s3:ObjectCreated event, tries to check if a separate item exists in an S3 bucket using the following bit of code using the AWS JavaScript SDK:
exports.somethingSomeSomething = async (event) => {
event.Records.forEach(async (record) => {
let tst = await s3.headObject({
Bucket: "mybucket",
Key: "something.gz"
}).promise()
console.log(tst)
})
};
I'm quite rusty with promises in JS, so I'm not sure why this bit of code doesn't work. For reference, it just dies without outputting anything.
However, the following does work:
exports.somethingSomething = async (event) => {
let tst = await s3.headObject({
Bucket: "mybucket",
Key: "something.gz"
}).promise()
console.log(tst)
console.log("RED")
};
How can I get the initial bit of code working, and what am I doing wrong?
It's because your code is async, but the function passed to your forEach loop is also async, so you have an async function invoking another chunk of async code, therefore you lose control of the flow. Whatever is inside forEach will run (although anything after forEach will run before whatever is inside forEach), but it will execute asynchronously and you are unable to keep track of its execution.
But if the code, as I said, will run, why don't you see the results?
Well, that's because Lambda will terminate before that code has the chance to execute. If you run the same piece of code locally, you'll see it will run just fine, but since the original code runs on top of Lambda, you don't have control when it terminates.
You have two options here:
The easiest is to grab the first item in the Records array because s3 events send one and only one event per invocation. The reason it is an array is because the way AWS works (a common interface for all events). Anyways, your forEach is not using anything of the Record object, but still if you wanted to use any properties of it, simply reference the 0th position, like so:
exports.somethingSomeSomething = async (event) => {
const record = event.Records[0]
//do something with record
const tst = await s3.headObject({
Bucket: "mybucket",
Key: "something.gz"
}).promise()
console.log(tst)
};
If you still want to use a for loop to iterate through the records (although, again, unnecessary for s3 events), use a for of loop instead:
exports.somethingSomeSomething = async (event) => {
for (const record of event.Records) {
// do something with record
const tst = await s3.headObject({
Bucket: "mybucket",
Key: "something.gz"
}).promise()
console.log(tst)
}
};
Since for of is just a regular loop, it will use the async from the function it's being executed on, so await is perfectly valid inside it.
More on async/await and for..of
I'm currently developing an app with vue v2, and something got me thinking. It's kind of hard to find information for good practices so I decided I will ask here.
I have a laravel api and vue for front end. Until now my models were such that will hold only a few objects (< 100), so I just put the in the state, and get what ever needed with a getter. Now I have a model that will hold > 10000 objects, so putting them all into the state is pretty much out the topic. The way I solve this problem is a have with action that gets a single object from the api by id, and put it in the state, after a check if its not in there already. The check happens with a getter that I'm also using when I need this object. So I have similar structure:
// mutation
[mutation_types.FETCH_SOMETHING](state, {something}) {
if (!this.getters.getSomethingById(something.id)) {
state.somethings.push(something)
}
}
// action
const {data} = await axios.get('~route~')
commit(mutation_types.FETCH_SOMETHING, {something: data.something})
// getter
getSomethingById: state => id => {
return state.somethings.find(something => something.id === id)
}
So this is working, but there are 2 problems that I see. The first is every time when I need a something I should call the dispatch to get the item from the api into the state and then the getter to get the actual object and the second problem is the check is happening after the api call, so even if I have this specific something in the state, I'm doing a request for it.
A way to fix both of those problems that I though of calling the just working with the getter, which will check that state, if the searched ID is found in there it will be returned, if not, the dispatch will get it from the api and then return it.
I feel like I'm adding logic to the getter that doesn't belong there, but I can't think of another way to achieve this. Is there something that I'm missing / doing wrong somewhere?
Take advantage of the async nature of vuex actions and promises. Do the check for existence directly in the action along with the api call if needed:
state: {
items: []
}
mutations: {
addItem: (state, item) => {
state.items.push(item)
}
}
actions: {
fetchItem: async (context, id) => {
let item = context.state.items.find(i => i.id === id)
if (!item) {
item = await axios.get('/example/' + id).then(response => {
context.state.commit('addItem', response.data)
return response.data
})
}
return item
}
}
I am trying to write a Cloud Function in node where I can return a token from a parameter.
The URL I use is...
https://us-central1-nmnm03.cloudfunctions.net/GetAccount?taccount=Asd
my function is this... and its wrong. I suspect I am not assigning TT properly.
var functions = require('firebase-functions');
const admin = require('firebase-admin');
admin.initializeApp(functions.config().firebase);
exports.GetAccount = functions.https.onRequest((req, res) => {
const t = admin.database().ref('/newaccout/'+req.query.account)
const tt = t.child(token)
res.send( "res is " + tt );
});
req.query.account is the Key. One of the Items in the document is token
ideally, I would like to get something like...
{"token":"23453458885"}
Could I get a node hint please... thanks
Though, I am not a firebase geek. What it seems from the documentation is that you will have two events that you can use to listen for retrieving child data. You can read further more here. The given options are used for different cases. Please follow through the mentioned link to have clear view.
Inside your cloud function you can try doing following:
const t = admin.database().ref('/newaccout/'+req.query.account)
t.on('child_added', function(data) {
res.json({
token: data.token
})
})
Or maybe like this:
const t = admin.database().ref('/newaccout/'+req.query.account)
t.once('value', function(snapshot) {
//Process it like above
//But here you will get al child elements at once
});
It looks like you are expecting to query the value found at a database reference stored at t. Unfortunately, you haven't actually performed a query yet. tt is just yet another Reference object that points to a location in the database. You should use the once() method on Reference to query a database location. Also bear in mind that you are using a variable called token, but you haven't defined yet in your code. To me, that looks like it would generate an error to me.
You might be well served by looking at a bunch of the sample code.
I want to use Fetch API but i don' t really understand it's mecanism.
I have an in my HTML and i want to assign the result of my fetch with this code :
const weather = "http://api.apixu.com/v1/current.json?key=cba287f271e44f88a60143926172803&q=Paris";
const array = [];
fetch(weather)
.then(blob => blob.json())
.then(data => {
array.push(data.current.humidity)
console.log(array[0])
}
);
document.querySelector('h1').innerHTML = array[0];
i have the result with the console.log but the returns "undefined". can you explain why ?
thanks a lot
This is because the call to the API is asynchronous, meaning that the code is not executed just line by line as you write it. The callback only runs as soon as the call to the API has finished, basically meaning that
data => {
array.push(data.current.humidity)
console.log(array[0])
}
runs after
document.querySelector('h1').innerHTML = array[0];
So when you try to set your h1, array is still empty. If you want to set it as soon the data is available, you have to do it within the callback function:
data => {
array.push(data.current.humidity)
document.querySelector('h1').innerHTML = array[0];
}
This might seem weird at first, but keep in mind that you're only registering an anonymous function but not running it yet. You just define the function that you want to trigger as soon as something happens, in this case: when your API call has finished.