From C++ to UML Class Diagram

Overview

This project offers a pipeline for reverse-engineering C++ codebases. By leveraging Clang-based parsing, it extracts data structures and their relationships across multiple translation units to generate a comprehensive PlantUML Class Diagram.

Workflow

graph LR
    %% Step 1
    subgraph S1 [Step 1: Generate Compilation Database]
        A[C++ Codebase] --> B(compile_commands.json)
    end

    %% Step 2
    subgraph S2 [Step 2: Extraction]
        B --> C[extract.py]
        C --> D[Class JSONs]
    end

    %% Step 3
    subgraph S3 [Step 3: Visualization]
        D --> E[plantuml.py]
        E --> F[PlantUML File]
        F --> G((Final Diagram))
    end

    %% Styling
    style S1 fill:#f5f5f5,stroke:#333,stroke-dasharray: 5 5
    style S2 fill:#e1f5fe,stroke:#01579b
    style S3 fill:#e8f5e9,stroke:#2e7d32
    style G fill:#fff,stroke:#2e7d32,stroke-width:2px

Loading

Step 1: Generate Compilation Database

Configure your CMake project to export compilation commands:

cmake -DCMAKE_EXPORT_COMPILE_COMMANDS=ON /path/to/source

Note: Specific CMake commands are based on standard practices as the context focuses on the Python scripts.

Step 2: Extract Data Structures

Execute extract.py to parse source files and generate JSON representations of the classes.

python extract.py --input /path/to/compile_commands.json --output /path/to/output_dir [--namespace <prefix>]

Arguments:

• --input: Path to compile_commands.json.
• --output, -o: Path to the output directory (default: output).
• --namespace: Optional filter to to limit the scope of analysis.

Step 3: Generate PlantUML Diagram

Execute plantuml.py to convert the extracted JSON files into a PlantUML Class Diagram.

python plantuml.py --input /path/to/output_dir --output /path/to/diagram.puml [--private]

Arguments:

• --input, -i: Path to the directory containing JSON files produced by extract.py (required) [3].
• --output, -o: Output .puml file path (default: output/classes.puml) [3].
• --private: Include private members in the diagram (default: public + protected only) [3].

Prerequisites

• Python 3 and Clang Python Bindings
  • Ensure libclang library and the corresponding Python bindings are compatible. The project was tested with python clang 18.1.8 and libclang-18.1.3.
• CMake: Required to generate the compile_commands.json file.
• Graphviz: Used for rendering the generated .puml files (optional, for viewing).
• Recommendation

    sudo apt update
    sudo apt install clang libclang-dev graphviz
  
    cd <git repo directory>
    python -m venv .venv
    source .venv/bin/activate
    pip install clang
  

Project Structure

• ast_visitor.py: Contains logic for traversing the Clang AST to extract data structures and their relationships.
• extract.py: Main entry point for parsing the compilation database and orchestrating the extraction process.
• plantuml.py: Generates the PlantUML script from the extracted JSON data.