AWS Amplify AppSync Subscription not working correctly - ionic-framework

I wrote a small application that subscribes to DB changes using AWS Amplify CLI / AppSync. All amplify api calls work perfectly (mutations, queries) but unfortunately the observer does not receive events. I can see that the MQTT socket receives periodically binaries but I can't obtain changed objects.
I configured Amplify for amplify use. I can see in the debugger that the AppSyncProvider has been initisalised. Also tried API and PubSub but makes no difference.
const awsmobile = {
"aws_appsync_graphqlEndpoint": "https://[...].appsync-api.[...]/graphql",
"aws_appsync_region": "[...]",
"aws_appsync_authenticationType": "AMAZON_COGNITO_USER_POOLS",
};
Amplify.configure(awsmobile);
ngOnInit()
{
try {
this.apiService.OnUpdateA.subscribe(
{
next: (x) => {[...]},
error: (e) => {[...]},
complete: () => {[...]}
});
}
catch (error) {[...] }
}
***Schema***
type A
#model
#auth(rules: [
{allow: owner},
{allow: groups, groups: ["A"], operations: [create, update, read]},
{allow: groups, groups: ["B"], operations: [read]},
])
{
id: ID!
entry: AnotherType! #connection(name: "AnotherConnection")
[...]
}
OnUpdateAListener: Observable<
OnUpdateASubscription
> = API.graphql(
graphqlOperation(
`subscription OnUpdateA($owner: String) {
onUpdateA(owner: $owner) {
__typename
id
owner
[...]
}
}`
)
) as Observable<OnUpdateASubscription>;
Anyone for any ideas?
**Logs:**
{mqttConnections: Array(1), newSubscriptions: {…}, provider: Symbol(INTERNAL_AWS_APPSYNC_PUBSUB_PROVIDER)}
mqttConnections: Array(1)
0: {url: "wss://[...]-ats.iot.[...].amazonaws…[...]%3D%3D", topics: Array(2), client: "[...]"}
length: 1
__proto__: Array(0)
newSubscriptions:
onUpdate:
expireTime: 1573313050000
topic: "[....]/22tmaezjv5555h4o7yreu24f7u/onUpdate/1cd033bad555ba55555a20690d3e04e901145776d3b8d8ac95a0aea447b273c3"
__proto__: Object
__proto__: Object
provider: Symbol(INTERNAL_AWS_APPSYNC_PUBSUB_PROVIDER)
__proto__: Object
However, not sure whether it is suspicious that the subscription Object has no queue?
Subscription {_observer: {…}, _queue: undefined, _state: "ready", _cleanup: ƒ}
_cleanup: ƒ ()
_observer:
next: (x) => {…}
__proto__: Object
_queue: ***undefined***
_state: "ready"
closed: (...)
__proto__: Object
Many thanks in advance.

For those who are experiencing the same behaviour. It was due to the fact that I had the owner in the auth section of the schema. Deleted the {allow: owner}, part and the subscriptions started to work immediately.

here is a working example of AWS Amplify Subscriptions:
import Amplify from 'aws-amplify';
import API from '#aws-amplify/api';
import PubSub from '#aws-amplify/pubsub';
import awsconfig from './aws-exports';
Amplify.configure(awsconfig);
API.configure(awsconfig);
PubSub.configure(awsconfig);
// put above in root
// below is example
import { API, graphqlOperation } from 'aws-amplify';
var onAddNote = `subscription OnCreateNote {
onCreateNote {
id
patient {
id
organization {
id
}
}
}
}
`;
listenForNoteAdd() {
return API.graphql(graphqlOperation(onAddNote) ).subscribe({next: (noteData) => {
console.log("new note so reload consider reload")
let note = noteData.value.data.onCreateNote
console.log(JSON.stringify(note))
// now that you have indication of something happening
// do what you must next
}})
}

I had the same problem with GraphQL. Thing is: we have to return Owner in mutation response in order to let Subscription know to whom to send this event.
removing {allow: owner} did worked for me but thats not the right way since we require it in order to have owner based access to data.
so the correct way i found is:
if subscription is:
subscription MySubscription {
onCreateuser(owner: "$userName") {
id
name
number
}
}
mutation should be:
mutation MyMutation {
createUser(input: {name: "xyz", id: "user123", number: "1234567890"}) {
id
name
number
owner
}
}
we must return owner in mutation's response in order to get the subscription to that event and all other properties as same as mutation response.

For those coming here experiencing this error, do not delete {allow: owner}. Allow owner ensures only a user authenticated with Cognito User Pool can run queries, mutations, etc.
It looks like the OP is using amplify codegen to generate his API service, and if you look the listener has param for owner. It's optional, but if your #auth is {allow: owner} it is required.
Additional note: Not sure if he is using the correct owner field stored in his datastore or not. If there is not already an owner field created (or a different field specified), it will create one with a unique uuid. So he could be passing the incorrect owner or none at all.
Run a simple call to get the owner...
export const GetOwner = async (id: string): Promise<GetOwnerQuery> => {
const statement = `query GetAppData($id: ID!) {
getAppData(id: $id) {
owner
}
}`;
const gqlAPIServiceArguments: any = {
id,
};
const response = (await API.graphql(graphqlOperation(statement, gqlAPIServiceArguments))) as any;
return <GetOwnerQuery>response.data.getAppData;
};
...and pass that in your subscription.
const { owner } = await GetOwner(id);
if (owner) {
this.apiUpdateListener$ = this.api.OnUpdateAppDataListener(owner).subscribe((data) => {
console.log('api update listener ===> ', data);
});
}

Related

How to pass data for sync function using watermelondb

Good day everyone, I am working with watermelondb and I have the code below, but I don't know how to actually use it. I am new in watermelondb and I don't know how to pass data as props to the pullChanges and pushChanges objects. How do I pass necessary data like changes and lastPulledAt from the database into the sync function when I call it. And I need more explanation on the migrationsEnabledAtVersion: 1 too. Thanks in advance for your gracious answers.
import { synchronize } from '#nozbe/watermelondb/sync'
async function mySync() {
await synchronize({
database,
pullChanges: async ({ lastPulledAt, schemaVersion, migration }) => {
const urlParams = `last_pulled_at=${lastPulledAt}&schema_version=${schemaVersion}&migration=${encodeURIComponent(JSON.stringify(migration))}`
const response = await fetch(`https://my.backend/sync?${urlParams}`)
if (!response.ok) {
throw new Error(await response.text())
}
const { changes, timestamp } = await response.json()
return { changes, timestamp }
},
pushChanges: async ({ changes, lastPulledAt }) => {
const response = await fetch(`https://my.backend/sync?last_pulled_at=${lastPulledAt}`, {
method: 'POST',
body: JSON.stringify(changes)
})
if (!response.ok) {
throw new Error(await response.text())
}
},
migrationsEnabledAtVersion: 1,
})
}
Watermelondb's documentation is terrible and its link to typescript even worse.
I spent almost a week to get 100% synchronization with a simple table, now I'm having the same problems to solve the synchronization with associations.
Well, the object you need to return in pullChanges is of the following form:
return {
changes: {
//person is the name of the table in the models
person: {
created: [
{
// in created you need to send null in the id, if you don't send the id it doesn't work
id: null,
// other fields of your schema, not model
}
],
updated: [
{
// the fields of your schema, not model
}
],
deleted: [
// is a string[] consisting of the watermelondb id of the records that were deleted in the remote database
],
}
},
timestamp: new Date().getTime() / 1000
}
In my case, the remote database is not a watermelondb, it's a mySQL, and I don't have an endpoint in my API that returns everything in the watermelon format. For each table I do a search with deletedAt, updatedAt or createdAt > lastPulledAt and do the necessary filtering and preparations so that the data from the remote database is in the schema format of the local database.
In pushChanges I do the reverse data preparation process by calling the appropriate creation, update or deletion endpoints for each of the tables.
It's costly and annoying to do, but in the end it works fine, the biggest problem is watermelon's documentation which is terrible.

Sign Up not working and throwing params errors

I am currently able to sign in just fine with previously created user credentials and use the app as normal, but am unable to create a new user. I am using React.js on the client side and an Express api on the backend. I am getting a mongoose validation error. All of the authentication came with the template the course has us use and I haven't touched any of those files. I even went back and compared commit history trees to ensure that nothing was changed.
Here is my user schema and sign-up route. I tried eliminating uniqueness from the model and that didn't impact it. I know there is a lot of potential places something could be going wrong, but if anyone has any suggestions on potential issues I would be forever grateful! I console logged the req.body.credentials within sign up and the data being sent over looks good.
Error code: 422 Unprocessable Entity
Server side error: 'The received params failed a Mongoose validation'
const mongoose = require('mongoose')
const { petSchema } = require('./pet.js')
const { pictureSchema } = require('./picture.js')
const userSchema = new mongoose.Schema({
email: {
type: String,
required: true,
unique: true
},
hashedPassword: {
type: String,
required: true
},
token: String,
pets: [petSchema],
pictures: [pictureSchema]
}, {
timestamps: true,
toObject: {
// remove `hashedPassword` field when we call `.toObject`
transform: (_doc, user) => {
delete user.hashedPassword
return user
}
}
})
module.exports = mongoose.model('User', userSchema)
// SIGN UP
// POST /sign-up
router.post('/sign-up', (req, res) => {
// start a promise chain, so that any errors will pass to `handle`
console.log(req.body.credentials)
Promise.resolve(req.body.credentials)
// reject any requests where `credentials.password` is not present, or where
// the password is an empty string
.then(credentials => {
if (!credentials ||
!credentials.password ||
credentials.password !== credentials.password_confirmation) {
throw new BadParamsError()
}
})
// generate a hash from the provided password, returning a promise
.then(() => bcrypt.hash(req.body.credentials.password, bcryptSaltRounds))
.then(hash => {
// return necessary params to create a user
return {
email: req.body.credentials.email,
hashedPassword: hash
}
})
// create user with provided email and hashed password
.then(user => User.create(user))
// send the new user object back with status 201, but `hashedPassword`
// won't be sent because of the `transform` in the User model
.then(user => res.status(201).json({ user: user.toObject() }))
// pass any errors along to the error handler
.catch(err => handle(err, res))
})
Solved. One of the subdocuments I had within user had a key with a value set to unique. I needed to eliminate that because my database was indexing users with a null value and throwing a duplicate error. I then needed to reset my database (I just renamed it to test it out) so that it didn't have any saved indexes with that configuration. I just deleted my collections within Heroku as well (luckily I didn't have significant amounts of data in there and this solution was perfectly fine for my situation). I am now able to sign up users again without any duplicate key errors.

How to identify unique users with Diagflow

I am trying to make an assistant app and was using the cloud firestore service of firebase to send the response back to my app using webhook as fulfilment. I have used 'session' parameter in request JSON according to this documentation and sending fulfilmentText as response to the user. But whenever user launches the app, a new session is created which I don't want. I simply want, just a single entry for each user in my database so how to achieve that using dialogflow.
In Alexa Skill, we have deviceId as parameter by which we can uniquely identify the user irrespective of the session id but is there any parameter in the dialogflow request JSON. If not, then how to achieve this task without it.
The request JSON I am getting from Dialogflow has a userID in it, so can I use the userId or should I go with userStorage provided the userStorage parameter is not available in the request JSON.
request.body.originalDetectIntentRequest { source: 'google', version: '2', payload: { surface: { capabilities: [Object] },
inputs: [ [Object] ],
user:
{ locale: 'en-US',
userId: 'ABwppHG5OfRf2qquWWjI-Uy-MwfiE1DQlCCeoDrGhG8b0fHVg7GsPmaKehtxAcP-_ycf_9IQVtUISgfKhZzawL7spA' },
conversation:
{ conversationId: '1528790005269',
type: 'ACTIVE',
conversationToken: '["generate-number-followup"]' },
availableSurfaces: [ [Object] ] } }
EDIT : Thank You #Prisoner for the answer but I am unable to send the random ID generated in the response and set in in the payload. Below is the code where I am generating the uuid and storing it in firestore. What I am doing wrong in the below code due to which new uuid is generated for returning user and therefore response is shown as No document found in the database. I suppose I am not sending uuid appropriately. Please help.
exports.webhook = functions.https.onRequest((request, response) => {
console.log("request.body.queryResult.parameters", request.body.queryResult.parameters);
console.log("request.body.originalDetectIntentRequest.payload", request.body.originalDetectIntentRequest.payload);
let userStorage = request.body.originalDetectIntentRequest.payload.user.userStorage || {};
let userId;
console.log("userStorage", userStorage);
if (userId in userStorage) {
userId = userStorage.userId;
} else {
var uuid = require('uuid/v4');
userId = uuid();
userStorage.userId = userId
}
console.log("userID", userId);
switch (request.body.queryResult.action) {
case 'FeedbackAction': {
let params = request.body.queryResult.parameters;
firestore.collection('users').doc(userId).set(params)
.then(() => {
response.send({
'fulfillmentText' : `Thank You for visiting our ${params.resortLocation} hotel branch and giving us ${params.rating} and your comment as ${params.comments}.` ,
'payload': {
'google': {
'userStorage': userStorage
}
}
});
return console.log("resort location", params.resortLocation);
})
.catch((e => {
console.log('error: ', e);
response.send({
'fulfillmentText' : `something went wrong when writing to database`,
'payload': {
'google': {
'userStorage': userStorage
}
}
});
}))
break;
}
case 'countFeedbacks':{
var docRef = firestore.collection('users').doc(userId);
docRef.get().then(doc => {
if (doc.exists) {
// console.log("Document data:", doc.data());
var dat = doc.data();
response.send({
'fulfillmentText' : `You have given feedback for ${dat.resortLocation} and rating as ${dat.rating}`,
'payload': {
'google': {
'userStorage': userStorage
}
}
});
} else {
// doc.data() will be undefined in this case
console.log("No such document!");
response.send({
'fulfillmentText' : `No feedback found in our database`,
'payload': {
'google': {
'userStorage': userStorage
}
}
});
}
return console.log("userStorage_then_wala", userStorage);
}).catch((e => {
console.log("Error getting document:", error);
response.send({
'fulfillmentText' : `something went wrong while reading from the database`,
'payload': {
'google': {
'userStorage': userStorage
}
}
})
}));
break;
}
You have a couple of options, depending on your exact needs.
Simple: userStorage
Google provides a userStorage object which is persisted across conversations when it can identify a user. This lets you store your own identifier when you need to track when a user returns.
The easiest way to do this is to check the userStorage object for the identifier when your webhook is called. If it doesn't exist, create one using something like a v4 UUID and save it in the userStorage object.
If you are using the actions-on-google library, the code might look something like this:
let userId;
// if a value for userID exists un user storage, it's a returning user so we can
// just read the value and use it. If a value for userId does not exist in user storage,
// it's a new user, so we need to generate a new ID and save it in user storage.
if (userId in conv.user.storage) {
userId = conv.user.storage.userId;
} else {
// Uses the "uuid" package. You can get this with "npm install --save uuid"
var uuid = require('uuid/v4');
userId = uuid();
conv.user.storage.userId = userId
}
If you are using the dialogflow library, you can use the above, but you'll need this line first:
let conv = agent.conv();
If you're using the multivocal library, it does all of the above for you and will provide a UserID in the environment under the path User/Id.
If you're handling the JSON directly, and you are using the Dialogflow v2 protocol, you can get the userStorage object by examining originalDetectIntentRequest.payload.user.userStorage in the JSON request object. You'll set the payload.google.userStorage object in the JSON response. The code is similar to the above and might look something like this:
let userStorage = body.originalDetectIntentRequest.payload.user.userStorage || {};
let userId;
// if a value for userID exists un user storage, it's a returning user so we can
// just read the value and use it. If a value for userId does not exist in user storage,
// it's a new user, so we need to generate a new ID and save it in user storage.
if (userId in userStorage) {
userId = userStorage.userId;
} else {
// Uses the "uuid" package. You can get this with "npm install --save uuid"
var uuid = require('uuid/v4');
userId = uuid();
userStorage.userId = userId
}
// ... Do stuff with the userID
// Make sure you include the userStorage as part of the response
var responseBody = {
payload: {
google: {
userStorage: JSON.stringify(userStorage),
// ...
}
}
};
Note the first line of the code - if userStorage doesn't exist, use an empty object. It won't exist until you send a response that includes storing something in it for the first time, which will happen in the last few lines of this code.
Advanced: Account Linking
You can request users to sign in to your Action using Google Sign In. This can be done just using voice for the simplest of cases and would only interrupt the flow the first time.
After this, your Action is given a JWT which contains their Google ID which you can use as their identifier.
If you're using the actions-on-google library, you can get the ID from the decoded JWT with a line such as:
const userId = conv.user.profile.payload.sub;
In the multivocal library, the ID from the decoded JWT is available in the environment under the path User/Profile/sub
Deprecated: Anonymous User ID
You'll see some answers here on StackOverflow that reference an Anonymous User ID. Google has deprecated this identifier, which was not always a reliable way to verify returning users, and will be removing it 1 Jun 2019.
This code is currently still being sent, but this will be removed starting 1 Jun 2019.

Meteor: Publish using users profile properties rather than ID

I'm currently creating an app that will be used by multiple companies.
Each user has the following profile:
username: johnDoe
emails: [{address: "some#email.com", verified: true}],
profile: {
name: "John Doe",
companyId: "1234"
}
I then have a collection (called Companies) of company objects that contain configuration info, templates etc specific to that company.
{
id: "1234",
configuration: {},
templates: []
}
In order to isolate each companies data I want to only publish data that matches the users profile companyId to the companies id.
if (Meteor.isServer) {
// collection to store all customer accounts
Companies = new Mongo.Collection('Companies');
// publish collection
Meteor.publish("Company", function () {
return Companies.find({id: Meteor.user().profile.companyId});
})
}
This currently works if I hardcode the clinic Id
// publish collection
Meteor.publish("Company", function () {
return Companies.find({id: "1234");
})
But returns an empty cursor with the Meteor.user().profile.companyId.
This means that the issue is either that I'm using the wrong function or more probably, the publish is happening before the user().profile.companyId can run.
Anybody know what I'm doing wrong? and do you have any advice on what to read up about so that I have an understanding of this moving forwards?
Thanks
Try doing an explicit findOne() in your publish function:
// publish collection
Meteor.publish("Company", function () {
var user = Meteor.users.findOne({_id: this.userId});
if(user && user.profile && user.profile.companyId) {
return Companies.find({id: user.profile.companyId});
} else {
console.log(user);
return this.ready();
}
});

Accessing services field of Meteor.users

When I query Meteor.users I do not receive the services field or any other custom fields I have created outside of profile. Why is it that I only receive _id and profile on the client and how can I receive the entire Meteor.users object?
Thanks.
From the DOcs
By default, the current user's username, emails and profile are published to the client. You can publish additional fields for the current user with:
As said above If you want other fields you need to publish them
// server
Meteor.publish("userData", function () {
if (this.userId) {
return Meteor.users.find({_id: this.userId},
{fields: {'services': 1, 'others': 1}});
} else {
this.ready();
}
});
// client
Meteor.subscribe("userData");
The above answer does work, but it means you have to subscribe to said data, which you should do if you are getting data from users other than the currently logged in one.
But if all you care about is the logged in user's data, then you can instead use a null publication to get the data without subscribing.
On the server do,
Meteor.publish(null, function () {
if (! this.userId) {
return null;
}
return Meteor.users.find(this.userId, {
fields: {
services: 1,
profile: 1,
roles: 1,
username: 1,
},
});
});
And this is actually what the accounts package does under the hood