runtime-visualizer

A visualizer for analyzing Python code.

Preview

Enter a Python project repository and the command to execute tests, then voila!

Design Video

Visualization Description

1. Our visualization was implemented using react-force-graph-2d
   • On top of this, we created customized rendering for node, and link metadata to display as an overlay
   • To prevent clutter, we only display data on hover
     • The hovered node and its children are highlighted and their metadata is displayed

Analysis Description

1. Function Call Graph — base graph shown in visualization

   • Using static syntactic analysis we grabbed all the name of functions
   • Using static lexical analysis, we look for function uses in each function of each file

2. Runtime Analysis — adds hover details and representitive edge widths to graph

   • Using static syntactical analysis, we determined where functions started and saved the name and make a call to our injected code to call our analyzeCode package
   • Using dynamic semantic analysis we ran tests that would trigger the injected lines of code naturally with reasonable inputs and save the data in a .json file to be processed later in JavaScript
   • Made a Python class to store our call information
   • Separate injections were made for test classes to instantiate our code analyzer per test

3. Memory Analysis — adds node size to graph

   • Using static syntatical analysis, we find function definitions and inject code before each
   • For our dynamic analysis we run tests and lexically analyze the results from memory_profiler, tracking peak memory usage for each function call

Group members & Work Contributed

1. Johnson Vu

   • Visualization
     • d3 setup
   • Front & Back-end configuration
   • General application data workflow/pipelining
     • integration of everyones' parts
   • Memory analysis
     • Injection into Python files
   • User tests

2. Justin Kwan

   • Call graph generation
   • Run-time injection and analysis
     • Brainstorming dynamic analysis process
     • Analysis of results
     • Injection into Python files
   • User tests

3. Leon Lui (our Python expert 🙂)

   • Run-time injection and analysis
     • Code analysis Python class
     • Injection into Python files
   • Example code mock-up

4. MengXin Zhao

5. Varun Belani

   • Memory analysis
     • Analysis of memory_profiler results
   • Project video
     • Video editing
     • Audio recording

Running the application

First install docker and docker-compose

Now run the following command in the main project directory:

$ sudo docker-compose up