This repository holds examples of Python 3 web apps that run on the Google App Engine (GAE) 2nd Generation environment.

If you find this repository useful, please star it ;)

Recommended: If you're looking for a GAE/Flask/Py3 Starter Project template, check out this link: GAE/Py3 Boilerplate.

Install the following on your computer:

• Install both Python 2 and Python 3. Python 2 is needed because Cloud SDK depends on it.
• Install Java JDK (I recommend OpenJDK 11 from AdoptOpenJDK) - this is needed to run the Datastore or Firestore emulator (via Cloud SDK).
• Install Cloud SDK.

Make sure you have the following Cloud SDK components installed:

• Cloud SDK Core Libraries (core)
• gcloud app Python Extensions (app-engine-python)
• gcloud app Python Extensions - Extra Libraries (app-engine-python-extras)
• gcloud Beta Commands (beta)

If you'll use Datastore or Firestore, you'll need to install one of these (or both) components too:

• Cloud Datastore Emulator (cloud-datastore-emulator)
• Cloud Firestore Emulator (cloud-firestore-emulator)

Firstly, check which of these components you have already installed:

gcloud components list

At least one or two of them should already be installed. For the others, install them using this command:

gcloud components install <component-name>

For example, if you want to install the Datastore Emulator, use this command:

gcloud components install cloud-datastore-emulator

Also, make sure that your already installed components are up-to-date:

gcloud components update

If you need more instructions, check here: (Cloud SDK components documentation)

• simple-gae-app: A simple Flask app with all the basic GAE settings needed. No database included here.
• simple-app-datastore: A simple Flask app that allows you to store items in Datastore (Datastore Emulator is used).
• simple-app-datastore-tests: Flask app with Datastore and tests (pytest).
• simple-app-firestore: Flask app with the Firestore database.
• simple-app-firestore-tests: Flask app with Firestore + tests.
• simple-app-sql: A simple Flask app with an SQL database (localhost: SQLite; Google Cloud: PostgreSQL).

• structured-app-sql: A structured Flask app with SQL database and tests.
• structured-app-firestore: A structured Flask app with Firestore + tests. Gives you an idea how to structure your web app.
• structured-app-ndb: An example using the new ndb library for Python 3 GAE runtime.
• env-var-example: GAE does not support environment variables, so the alternative is to write them in the database instead. Here's an example how.
• auth-username-password: Username/password login example built with Flask, Datastore, ndb library and Python 3 GAE runtime.

• (recommended) smartninja/gae-py3-boilerplate: this is a starter project boilerplate that has the auth system, translations, emailing system and some other goodies already implemented.
• ramuta/ninjatodo: A task management system (TODO app) built with Python 3, Flask, Datastore, ndb and running on GAE second generation.

Click on each example and check the README.md there - it has all the instructions you need.

In order to deploy the web app on Google Cloud, you will probably need to connect your credit card to it. But this doesn't mean Google will start charging you right away - the free quota is pretty generous and you can also set a daily spending limit to 0 USD (see step 4 below in the "Deployment to GAE" section).

Open the Terminal in the root of the project and type in:

gcloud init

If this is a new project, select the choice no.2: Create a new configuration. Then enter a name for this configuration (it needs to be unique only on your computer, not globally).

If this is an already existing project with an existing configuration, select it and skip some of the next steps.

After this step you'll need to log in with your Google account.

If you already have a Google Cloud project for this repository, select it. If not, create a new one.

Now it's time to create a Google App Engine instance and select the region where it will run (See the list of possible regions here).

Important: The region cannot be changed later.

Once you've chosen the region, enter the following command:

gcloud app create --region=europe-west

In this case we chose the "europe-west" region, but replace it with some other if you want.

Go to Google Cloud Console, open your project and type this in the Search box: Cloud Build API. Then enable it.

You will probably need to enable billing for your project, but don't worry - this does not mean Google will start charging you. Google has a very generous free quota and you will very likely stay within that quote.

But it doesn't hurt to set up the daily limit for your GAE app on Google Cloud Console (type "App Engine settings" in the Search box and then enter the daily spending limit (it can be 0).

The next step is to deploy your code to Google Cloud:

gcloud app deploy --version production

You could do it without the version flag, but it's a good practice so that GAE does not create a new version for each of your deployments. You can also name versions after your Git branches (for example: master, develop).

If you'll use Datastore or Firestore as a database, you have to make sure the database is enabled on Google Cloud Platform.

Important: Only one of these two can be enabled in the same project! Either Datastore, or Firestore. Both cannot be enabled and you cannot switch from one to another once you have data in the database!

What's the difference between the two? Firestore is the newer version of Datastore. All Datastores will eventually be converted into Firestore, but they will continue working via the, so called, "Datastore Mode", which has less features than native Firestore. So if you're creating a new project with an empty database, choose Firestore.

How to enable Datastore/Firestore?

If it's not enabled automatically, just type either "Datastore" or "Firestore" in the Search box on Google Cloud Platform. Once you click on the selection, the Datastore/Firestore will be automatically enabled.

Note that google-cloud-datastore requires a google-cloud-core library version less than 0.30.0 (currently 0.28.1), while google-cloud-firestore requires a version bigger than 1.0.0. (currently 1.0.3). That's why you might need to use two separate virtual environments (one with the datastore lib and the other with the firestore lib).

Or to run pip install google-cloud-datastore or google-cloud-firestore each time you would want to switch the project.

But in my experience this wasn't needed. Try it out and you'll see.

Please create a new issue in case there's some bug or something should be improved.

Happy to receive pull requests, too! :)