Flask bootstrap project

You could use this template to easy and fast start new project for API, Web, tasks, etc without configuration some basic things like logging, celery tasks, endpoints routing, web templates configuration etc.

Features:

• Template for web pages with static
• Template for API endpoints
• Scheduled tasks with Celery
• Logging
• Configuration with ENV variables support

Running:

Docker

First you need to build the image:

$ docker build -t my_project .

After that you can run the container:

$ docker run -d \
  --name=clicker \
  --rm \
  -ti \
  -p 80:80 \
  -v /local/path:/app/logs \
  my_project

Change the port if you need and /local/path to your own path in host os where you want to store logs.

Docker-compose

Run docker-compose configuration:

$ docker-compose up --build -d

It will use docker-compose.yml file from the current directory. If you don't need periodic tasks you can remove tasks and redis services from docker-compose.yml file. Redis is mandatory for periodic tasks. It used as celery broker and as cache for celery tasks.

Configuration

Web pages

To add new web page you need to add new template to templates directory. Then you need to add python file to web directory and add necessary imports to __init__.py file. Or you can extend existing configuration.

API endpoints

To add new API endpoint you need to add new python file to api/v1 directory and add necessary imports to __init__.py file. Or you can extend existing configuration. The main difference between API endpoints and web pages is that API endpoints are not able to serve static files.

Scheduled tasks

To add new scheduled task you need to add new task file with to tasks directory. Then extend imports section in celery app configuration in scheduler.py file.

app.conf.update(
    ...
    imports=('tasks.hello_world', 'tasks.new_task')
)

And add new task to scheduler section in same file.

app.conf.beat_schedule = {
    ...
    'new_task': {
        'task': 'tasks.new_task.new_task',
        'schedule': int(APP_CONFIG.get('TASK_INTERVAL_SECONDS', 10)),
    },
}

To start scheduler you need to run:

$ celery -A scheduler worker -B

or if you run it in production split it in two processes:

$ celery -A scheduler worker

and

$ celery -A scheduler beat

Logging

Logging configuration located in config/logging.py file. You could add more loggers, formatters and handlers to it. Just use existing configuration as example.

Also, one filter already added to logging configuration, this is RequestIdFilter. It adds request id to all log records so you can track logs by request id from beggining of request to end.

To start logging anywhere in your code you need to initialize necessary logger. As parameter, you need to pass logger name as it defined in loggers section of your logging configuration.

import logging
...
debug_logger = logging.getLogger('debug')

then you can use logger to log messages:

debug_logger.debug('message')

Overriding variables

If you need to override some config variable without changing it in common.py you could set necessary variable as environment variable on your system.

Or if you're using docker or docker-compose you could set them in .env file or pass as environment variables to docker container.