CodeGraph

CodeGraph is designed with several extension points for developers looking to build on top of it:

1. New Visualizers: Add new rendering logic or export formats (like JSON or DOT) by adding functions to codegraph/vizualyzer.py.
2. New Parsers: Extend codegraph/parser.py to support languages other than Python.
3. New Link Types: Update the convert_to_d3_format() function in vizualyzer.py to handle new relationship definitions.

CodeGraph follows a linear pipeline to transform Python source code into an interactive dependency graph:

1. Path Discovery: The CLI receives paths, and utils.get_python_paths_list() recursively finds all .py files.
2. Parsing: parser.create_objects_array() parses each file to extract functions, classes, methods, and import statements.
3. Graph Building: core.CodeGraph.usage_graph() maps entities to line ranges, identifies where they are used, and creates dependency edges.
4. Visualization: vizualyzer.draw_graph() converts the graph into a D3.js format and generates an HTML file with embedded JavaScript, which is then opened in a browser.

Core Data Flow Summary: Python Files $\rightarrow$ Parser $\rightarrow$ Code Objects $\rightarrow$ Import Analysis $\rightarrow$ Entity Usage $\rightarrow$ Dependency Graph

When viewing a CodeGraph visualization, nodes and links are categorized by type to represent the structure of your code:

Node Types

Link Types

Install the CodeGraph static code analyzer using the Python package manager:

pip install codegraph

To analyze a Python project and generate an interactive HTML diagram of its structure (showing dependencies between modules, classes, and functions), run the codegraph command followed by the path to your code. This will automatically open the visualization in your default browser.

codegraph /path/to/your_python_code

Install the core CodeGraph package using pip:

pip install codegraph

If you require legacy matplotlib visualization support, install the optional dependency:

pip install codegraph[matplotlib]

CodeGraph is a static code analyzer that creates a graph of code entities (methods, classes, etc.) based on lex and syntax parsing without executing the code. You can use the CLI to generate interactive D3.js HTML visualizations, Matplotlib plots, or export graph data to CSV.

Basic usage requires providing at least one path to a codebase:

codegraph /path/to/your/code

You can export the analyzed code structure to a CSV file for use in spreadsheets or other data analysis tools using the --csv flag.

CSV Columns:

• name: Entity name
• type: Entity type (module, function, class, or external)
• parent_module: The parent module (applicable to functions and classes)
• full_path: The file path
• links_out: Count of outgoing dependencies
• links_in: Count of incoming dependencies
• lines: Lines of code in the entity

codegraph /path/to/code --csv output.csv

The codegraph CLI tool supports the following options:

| Option | Description |
|-------|-------------|
| `--output PATH` | Custom output path for the HTML file (default: `./codegraph.html`) |
| `--matplotlib` | Use legacy matplotlib visualization instead of D3.js (requires `codegraph[matplotlib]`)
| `-o, --object-only` | Print dependencies to console only, no visualization |

Run the codegraph command followed by the path to the Python code you wish to analyze. This generates an interactive HTML visualization and opens it in your default web browser.

CodeGraph performs static analysis using lexical and syntax parsing, meaning it does not execute your code and does not require you to activate environments or install the target code's dependencies.

codegraph /path/to/your_python_code