I'm working on a simple registration system using Firebase as a backend. I am successfully authenticating users and writing to the database. I have an index of courses and users with the following structure:
{
courses: { // index all the courses available
key1: {
title: "Course 1",
desc: "This is a description string.",
date: { 2018-01-01 12:00:00Z }
members: {
user1: true
...
}
},
key2 { ... },
},
users: { // track individual user registrations
user1: {
key1: true,
...
},
user2: { ... }
}
}
I have a cloud function that watches for the user to add a course and it builds an array with the corresponding courseId that will look at the courses node to return the appropriate items.
exports.listenForUserClasses = functions.database.ref('users/{userId}')
.onWrite(event => {
var userCourses = [];
var ref = functions.database.ref('users/{userId}');
for(var i=0; i<ref.length; i++) {
userCourses.push(ref[i])
}
console.log(userCourses); // an array of ids under the user's node
});
So, my question has two parts:
How can I build the updated object when the page is loaded?
How do I return the function to the client script?
Question 1: From the client side you want to get the reference to the database path. Then you want to call the child_added event. Keep it in-memory, this will be called whenever one is add then you can update your UI.
var ref = db.ref("path/to/courses");
ref.on("child_added", function(snapshot, prevChildKey) {
var newClass = snapshot.val();
});
If you are completely refreshing the page then you can always grab the data again from the database path by using the value option and calling once
Questions 2: You don't. This is considered an asynchronous function. If you wanted a response from the function then you would setup an HTTP trigger and wait for the response from that function.
Related
what is the best way to update many records with different data ?
I'm doing it like this
const updateBody = JSON.parse(req.body);
try {
for (let object of updateBody) {
await prisma.comissions.upsert({
where: {
producer: object.producer,
},
update: {
rate: object.rate,
},
create: object,
});
}
I'm being able to update it, but it's taking a really long time to do so. I'm aware of transaction, but i'm not sure how to use it.
In Prisma transaction query is used in two ways.
Sequential operations: Pass an array of Prisma Client queries to be executed sequentially inside of a transaction.
Interactive transactions: Pass a function that can contain user code including Prisma Client queries, non-Prisma code, and other control flow to be executed in a transaction.
In our case we should use the interactive transaction, Because it contain user code, To use the callback function in the Prisma transaction, we need to add a preview feature to the Prisma.schema file
generator client {
provider = "prisma-client-js"
previewFeatures = ["interactiveTransactions"]
}
prisma.$transaction(async(prisma) => {
try {
for (let object of updateBody) {
await prisma.comissions.upsert({
where: {
producer: object.producer,
},
update: {
rate: object.rate,
},
create: object,
});
}
});
What I am trying to do
I am creating a social media app with react native and firebase. I am trying to call a function, and have that function return a list of posts from off of my server.
Problem
Using the return method on a firebase query gives me a hard to use object array:
Array [
Object {
"-L2mDBZ6gqY6ANJD6rg1": Object {
//...
},
},
]
I don't like how there is an object inside of an object, and the whole thing is very hard to work with. I created a list inside my app and named it items, and when pushing all of the values to that, I got a much easier to work with object:
Array [
Object {
//...
"key": "-L2mDBZ6gqY6ANJD6rg1",
},
]
This object is also a lot nicer to use because the key is not the name of the object, but inside of it.
I would just return the array I made, but that returns as undefined.
My question
In a function, how can I return an array I created using a firebase query? (to get the objects of an array)
My Code
runQ(group){
var items = [];
//I am returning the entire firebase query...
return firebase.database().ref('posts/'+group).orderByKey().once ('value', (snap) => {
snap.forEach ( (child) => {
items.push({
//post contents
});
});
console.log(items)
//... but all I want to return is the items array. This returns undefined though.
})
}
Please let me know if I'm getting your question correctly. So, the posts table in database looks like this right now:
And you want to return these posts in this manner:
[
{
"key": "-L1ELDwqJqm17iBI4UZu",
"message": "post 1"
},
{
"key": "-L1ELOuuf9hOdydnI3HU",
"message": "post 2"
},
{
"key": "-L1ELqFi7X9lm6ssOd5d",
"message": "post 3"
},
{
"key": "-L1EMH-Co64-RAQ1-AvU",
"message": "post 4"
}
...
]
Is this correct? If so, here's what you're suppose to do:
var items = [];
firebase.database().ref('posts').orderByKey().once('value', (snapshot) => {
snapshot.forEach((child) => {
// 'key' might not be a part of the post, if you do want to
// include the key as well, then use this code instead
//
// const post = child.val();
// const key = child.key;
// items.push({ ...post, key });
//
// Otherwise, the following line is enough
items.push(child.val());
});
// Then, do something with the 'items' array here
})
.catch(() => { });
Off the topics here: I see that you're using firebase.database().... to fetch posts from the database, are you using cloud functions or you're fetching those posts in your App, using users' devices to do so? If it's the latter, you probably would rather use cloud functions and pagination to fetch posts, mainly because of 2 reasons:
There might be too many posts to fetch at one time
This causes security issues, because you're allowing every device to connect to your database (you'd have to come up with real good security rules to keep your database safe)
I aggregated some data and published it, but I'm not sure how/where to access the subscribed data. Would I be able to access WeeklyOrders client collection (which is defined as client-only collection i.e WeeklyOrders = new Mongo.Collection(null);)?
Also, I see "self = this;" being used in several examples online and I just used it here, but not sure why. Appreciate anyone explaining that as well.
Here is publish method:
Meteor.publish('customerOrdersByWeek', function(customerId) {
check(customerId, String);
var self = this;
var pipeline = [
{ $match: {customer_id: customerId} },
{ $group: {
_id : { week: { $week: "$_created_at" }, year: { $year: "$_created_at" } },
weekly_order_value: { $sum: "$order_value" }
}
},
{ $project: { week: "$_id.week", year: "$_id:year" } },
{ $limit: 2 }
];
var result = Orders.aggregate(pipeline);
result.forEach(function(wo) {
self.added('WeeklyOrders', objectToHash(wo._id), {year: wo.year, week: wo.week, order_value: wo.weekly_order_value});
});
self.ready();
});
Here is the route:
Router.route('/customers/:_id', {
name: 'customerOrdersByWeek',
waitOn: function() {
return [
Meteor.subscribe('customerOrdersByWeek', this.params._id)
];
},
data: function() { return Customers.findOne(this.params._id); }
});
Here is my template helper:
Template.customerOrdersByWeek.helpers({
ordersByWeek: function() {
return WeeklyOrders.find({});
}
});
You want var self = this (note the var!) so that call to self.added works. See this question for more details. Alternatively you can use the new es6 arrow functions (again see the linked question).
There may be more than one issue where, but in your call to added you are giving a random id. This presents two problems:
If you subscribe N times, you will get N of the same document sent to the client (each with a different id). See this question for more details.
You can't match the document by id on the client.
On the client, you are doing a Customers.findOne(this.params._id) where this.params._id is, I assume, a customer id... but your WeeklyOrders have random ids. Give this a try:
self.added('WeeklyOrders', customerId, {...});
updated answer
You'll need to add a client-only collection as a sort-of mailbox for your publisher to send WeeklyOrders to:
client/collections/weekly-orders.js
WeeklyOrders = new Meteor.Collection('WeeklyOrders');
Also, because you could have multiple docs for the same user, you'll probably need to:
Forget what I said earlier and just use a random id, but never subscribe more that once. This is an easy solution but somewhat brittle.
Use a compound index (combine the customer id + week, or whatever is necessary to make them unique).
Using (2) and adding a customerId field so you can find the docs on the client, results in something like this:
result.forEach(function (wo) {
var id = customerId + wo.year + wo.week;
self.added('WeeklyOrders', id, {
customerId: customerId,
year: wo.year,
week: wo.week,
order_value: wo.weekly_order_value,
});
});
Now on the client you can find all of the WeeklyOrders by customerId via WeeklyOrders.find({customerId: someCustomerId}).
Also note, that instead of using pub/sub you could also just do all of this in a method call. Both are no-reactive. The pub/sub gets you collection semantics (the ability to call find etc.), but it adds the additional complexity of having to deal with ids.
Sorry if I'm not getting the terminology right. Here's what I have currently my MongoDB user docs db.users:
"liked" : [
"EBMKgrD4DjZxkxvfY",
"WJzAEF5EKB5aaHWC7",
"beNdpXhYLnKygD3yd",
"RHP3hngma9bhXJQ2g",
"vN7uZ2d6FSfzYJLmm",
"NaqAsFmMmnhqNbqbG",
"EqWEY3qkeJYQscuZJ",
"6wsrFW5pFdnQfoWMs",
"W4NmGXyha8kpnJ2bD",
"8x5NWZiwGq5NWDRZX",
"Qu8CSXveQxdYbyoTa",
"yLLccTvcnZ3D3phAs",
"Kk36iXMHwxXNmgufj",
"dRzdeFAK28aKg3gEX",
"27etCj4zbrKhFWzGS",
"Hk2YpqgwRM4QCgsLv",
"BJwYWumwkc8XhMMYn",
"5CeN95hYZNK5uzR9o"
],
And I am trying to migrate them to a new key that also captures the time that a user liked the post
"liked_time" : [
{
"postId" : "5CeN95hYZNK5uzR9o",
"likedAt" : ISODate("2015-09-23T08:05:51.957Z")
}
],
I am wondering if it might be possible to simply do this within the MongoDB Shell with a command that iterates over each user doc and then iterates over the liked array and then updates and $push the new postId and time.
Or would it be better to do this in JavaScript. I am using Meteor.
I almost got it working for individual users. But want to know if I could do all users at once.
var user = Meteor.users.findOne({username:"atestuser"});
var userLiked = user.liked;
userLiked.forEach(function(entry) {
Meteor.users.update({ username: "atestuser" },
{ $push: { liked_times: { postId: entry, likedAt: new Date() }}});
console.log(entry);
});
Still a bit of a newbie to MongoDB obviously......
Here is something i made real quick you should run this on the server side just put it into a file e.g. "migrate.js" in root meteor and run the meteor app
if (Meteor.isServer) {
Meteor.startup(function () {
var users = Meteor.users.find().fetch();
users.forEach(function (doc) {
liked.forEach(function (postId) {
Meteor.users.update(doc._id, { $push: { liked_times: { postId: postId, likedAt: new Date() } } });
});
});
console.log('finished migrating');
});
}
p.s I didn't test it
If this is a one time migration i would do something like this in a one time js script.
Get all users
Iterate over each user
Get all likes
Iterate over them, get likedAt
var liked_times = _.collect(likes, function (likeId) {
return {
'postId' : likeId,
'likedAt': // get post liked time from like id.
}
});
Insert the above in the collection of choice.
Note:
The above example makes use of lodash
I would rather just save likedAt as a timestamp.
In my social app(like as FB) I have a strange need to merge two cursors of the same collection users in one publish!
Meteor server print this error:
"Publish function returned multiple cursors for collection users".
Maybe this can not be done in Meteor 0.7.2, perhaps I'm mistaken approach.
But I have seen the structure of a cursor is pretty simple as I could make a simple array merge and return back a Cursor?
CLIENT
Meteor.subscribe('friendById', friend._id, function() {
//here show my friend data and his friends
});
SERVER
//shared functions in lib(NOT EDITABLE)
getUsersByIds = function(usersIds) {
return Meteor.users.find({_id: {$in: usersIds} },
{
fields: { // limited fields(FRIEND OF FRIEND)
username: 1,
avatar_url: 1
}
});
};
getFriendById = function(userId) {
return Meteor.users.find(userId,
{
fields: { // full fields(ONLY FOR FRIENDS)
username: 1,
avatar_url: 1,
online: 1,
favorites: 1,
follow: 1,
friends: 1
}
});
};
Meteor.publish('friendById', function(userId) { //publish user data and his friends
if(this.userId && userId)
{
var userCur = getFriendById(userId),
userFriends = userCur.fetch()[0].friends,
retCurs = [];
//every return friend data
retCurs.push( userCur );
//if user has friends! returns them but with limited fields:
if(userFriends.length > 0)
retCurs.push( getUsersByIds(userFriends) );
//FIXME ERROR "Publish function returned multiple cursors for collection users"
return retCurs; //return one or more cursor
}
else
this.ready();
});
See bold red text in the documentation:
If you return multiple cursors in an array, they currently must all be from different collections.
There is the smart-publish package which adds this ability to use on publish to manage multiple cursors on the same collection. It is relatively new.
Either that or manually manage the cursors by using 'this.added', 'this.removed', and 'this.changed' inside the publish.
SOLUTION
Meteor.publish('friendById', function(userId) {
if(this.userId && userId)
{
var userCur = getFriendById(userId), //user full fields
userData = userCur.fetch()[0],
isFriend = userData.friends.indexOf(this.userId) != -1,
retCurs = [];
//user and his friends with limited fields
retCurs.push( getUsersByIds( _.union(userId, userData.friends) ));
if(isFriend)
{
console.log('IS FRIEND');
this.added('users',userId, userData); //MERGE full fields if friend
//..add more fields and collections in reCurs..
}
return retCurs;
}
else
this.ready();
});