Ticket Selling System

A simple ticket selling system project using flask, sqlalchemy, mysql, deployed with docker

Built and tested on

• docker 17.05.0-ce, build 89658be
• docker-compose version 1.8.0, build f3628c7

Design

• Database Design: Design diagram
• API Documentation: API

Quick Start

Deploy to host

Prerequisite

• python3.5
• mysql5.7
• python modules
  • flask==0.12
  • sqlalchemy==1.1.10
  • flask_sqlalchemy==2.2
  • flask_hashing==1.1
  • configparser==3.5.0
  • pymysql==0.7.9
  • mixer==5.6.6

Step

1. Goto project root directory, execute sudo pip3 install -r requirements.txt to install python modules
2. Modify configuration file in src/config.d/config.conf with your special needs
3. Ensure mysql database is set as in config.conf file
4. Goto src/ and execute python3 server.py start and you are all set

Deploy to docker

Prerequisite

• docker
• docker-compose

Step

1. Goto docker/
2. Put .env file in it. The contents are documented later
3. Execute docker-compose up -d and you are all set

Note

Do not modify src/config.d/config.conf when you deploy to docker

.env File

• TICKET_SERVER_CMD
  • start Run server
  • test Run test and exit
• TICKET_SERVER_PORT
  • 9999 Deloying port

Look into deeper

Flask Session

Flask session is based on cookies, and you must set a secret key to it so that the cookies are encrypted and cannot be read by others The secret key is set in src/app/__init__.py and is not configurable in config.conffile. It is recommanded to use a random string as secret key, just as documented in Flask quick start

Database fake data

Python module mixer is a fake data generator. To auto-generate data and store them into database, you can start the server with test parameter (pass test as parameter when run in host, or set TICKET_SERVER_PORT=testin .env file when run in docker) Problem is that it cannot deal with FK in a good manner, so to create tickets, you must insert to database on your own. More usage is documented in here

Password hashing

Passwords are hashed and not salted before they are stored into database. To have customed pasword hashing, reference flask_hashing

Things can be improved

1. Seperate routers and handlers
2. Some basic operation can be encapsulated into models
3. Log system
4. Distributed system, with nginx as proxy server
5. Database cluster and data backup
6. Cache movie and cinema infomation in Redis

...