Access Firebase Cloud Firestore using Google Cloud Functions - Part 2

February 03, 2020

In this tutorial, we will write cloud functions and make API requests. Later we will deploy the cloud functions to Firebase.

Series

There are two parts to this series. Please read here for Part 1.

Prerequisites

You need to know how to create a Firebase project; You can refer to the starter guide here.
Familiar with JavaScript/TypeScript
Familiar with Python

Overview

In this tutorial, we will first write cloud functions to access Cloud Firestore. Then we will write python to make API requests to cloud functions that we had written. After that, we will verify the output result of each function call. Finally, we will deploy those cloud functions to Firebase.

Writing Cloud Functions

From the project that you just created, open index.ts file under YOUR_PROJECT/functions/src. Let’s start writing code.

We have two functions here.

export const addUser
export const getUser

We will be using json when we are adding the user name. Therefore, you have to set the content-type to application/json. Then we will be adding in users collection as shown in this line const docRef = db.collection("users").doc(name);. We also add server time by using this line FieldValue.serverTimestamp() so that we know when the user is added to the system.

When we need to user back, we just get it back from users collection also. You can also see that there are console.log to help you to debug if you have any problem getting it right.

import * as functions from "firebase-functions";
import * as admin from "firebase-admin";

admin.initializeApp();

export const addUser = functions.https.onRequest((request, response) => {
  console.log(request.get("content-type"));
  console.log('the request name' + request.body.name);

  const name = request.body.name;
  switch (request.get("content-type")) {
    case "application/json":
    const db = admin.firestore();
      const FieldValue = require("firebase-admin").firestore.FieldValue;
      const docRef = db.collection("users").doc(name);
      docRef
        .set({
          name: name,
          added_on: FieldValue.serverTimestamp()
        })
        .catch(error => {
          console.log(error);
          response.status(500).send(error);
        });
      break;
  }
  response.status(200).send(`${name} has been added`);
});

export const getUser = functions.https.onRequest((request, response) => {
  console.log('the request name' + request.body.name);
  const db = admin.firestore();
  const name = request.body.name;
  db.collection("users")
    .doc(name)
    .get()
    .then(snapshot => {
      response.send(snapshot.data());
    })
    .catch(error => {
      console.log(error);
      response.status(500).send(error);
    });
});

Complie Cloud Functions

Before, we call any cloud functions, we need to compile the code from the previous section. Make sure you are in <YOUR_PROJECT>/functions folder.

To check any error/subtle bug in your codeindex.ts file. Make sure there is no error after you run it.
```
$ npm run-script lint
```
To compile your code and generate a native JavaScript file.
```
$ npm run-script build
```
Now we are going to run those functions that we wrote. Take note of two links given at the console after you run it. Those functions will be running locally (You can see that those links started with localhost) but firebase admin will be accessing your Cloud Firestore database from Google.
```
$ firebase serve --only functions
```

Call Cloud Functions

Calling cloud function can be done by using curl but I think it is a little bit harder to see with json data. So we will be using python and requests to call those functions. Please make sure you have installed requests via pip.

1. To Add User

Make sure you change url with the generated link from the console you run it before. We are going to add Jane as an user to our database.

import requests
url = 'http://localhost:5000/<YOUR PROJECT LINK>/addUser'
payload = '{"name":"Jane"}'
headers = {'content-type': 'application/json', 'Accept-Charset': 'UTF-8'}
r = requests.post(url, data=payload, headers=headers)
print(r.content)

Result

You should see the following output when the user is added successfully.


b'Jane has been added'

2. To Get User

Again make sure you change url with the generated link from the console you run it before. This time, we should get the user back.

import requests
url = 'http://localhost:5000/<YOUR PROJECT LINK>/getUser'
payload = '{"name":"Jane"}'
headers = {'content-type': 'application/json', 'Accept-Charset': 'UTF-8'}
r = requests.post(url, data=payload, headers=headers)
print(r.content)

Result

You should see the following output when the request is successful. Of course, the timestamp will be different.


b'{"name":"Jane","added_on":{"_seconds":1580635800,"_nanoseconds":625000000}}'

Deploy Cloud Functions

You can deploy your cloud functions to Google server at your <YOUR_PROJECT>/functions by typing the following command at your console. You should see that links are generated. You can use those links to access remotely. Just replace your url in Python code from above and verify your output.

$ firebase deploy

NOTE: Take care of your Firebase security in general. Anything access from the outside world can incur charges if you exceed certain thresholds. If you don’t need it, just delete the functions; just to be safe. You can delete functions at Firebase console.

Conclusion

We first set up gcloud SDK and Firebase CLI at Part 1. Then we wrote cloud functions and access Cloud Firestore with Python and verify our results. After this part is done, we deploy cloud functions to Firebase to be able to access globally. With this, you should able to write and access cloud functions using Firebase.