Graphwalker-based Test Generation Using Communities

These are current capabilities of the project:

• Show networkx graph in a visual view
• Apply community detection
  • Show communities
• Load graphwalker json model and convert it to a networkx graph
• Generate test cases from graphwalker json model

Requirements

• Python 3 or higher
• Virtual environment
  1. Create it with: py -3 -m venv env
  2. Activate it with: env\scripts\activate
  3. Install pip packages when the virtual environment is activated
• Download Graphwalker Studio and Cli from official site and copy them to project root

Required Pip Packages

• networkx => For graph operations
  • numpy==1.19.3 => There is a bug in the latest numpy version (1.19.4)
• python-louvain => For community detection
• matplotlib

Follow this steps to install the required packages

1. pip install networkx
2. pip install python-louvain
3. pip install matplotlib
4. pip install numpy==1.19.3

Graphwalker

• Graphwalker cli is required to generate test cases. You can install graphwalker cli from the following page.
• The cli jar must be available in the root of the project. In the future, path option will be available.
• Java 8 or higher must be available in your system.

How to start Graphwalker studio in local

1. Download GraphWalker studio jar from the official link
2. Type java -jar graphwalker-studio-4.*.jar in the command line
3. Open a browser and use url: http://localhost:9090/studio.html

Workflow

1. Create a model in Graphwalker studio
2. Save it as a json file such as: LoginSignUpForm.json
3. Load this json and convert it to a networkx graph
   • Generate test cases and measure time
4. Apply community detection and create multiple graphs
   • Generate test cases and measure time

References

1. Networkx
2. Community Detection for Networkx
3. Graphwalker