CodeAnalysis

python v2_analyse_clusters.py --api-key APIkey --clusters 1 --output ast_cluster_1_report_enhanced.md

JavaScript code analysis toolkit with visualization tools.

Tools

• Analyzers

  • acorn-analyzer.js: AST analysis
  • tree-sitter-analyzer.js: Syntax analysis
  • codebase-analyzer.js: Metrics generation

• Visualizers

  • json-visualizer/: JSON data visualization
  • visualizer/: Code metrics visualization

Installation

npm init -y
npm install acorn acorn-walk acorn-jsx

Environment Setup

1. Clone repository:

git clone https://github.com/NikhilKalloli/CodeAnalysis
cd code-analysis-toolkit

2. Install dependencies:

npm install

Basic Usage

const Analyzer = require('./acorn-analyzer.js');

async function main() {
  const analyzer = new Analyzer('./src');
  await analyzer.analyzeDirectory();
  await analyzer.generateCSVReport('analysis-report.csv');
  analyzer.printSummary();
}

main();

Command Line Usage

Analyze a directory and generate report:

node analyze.js --input ./src --output report.csv

Options:

• --input: Path to analyze (required)
• --output: Report output path (default: analysis-report.csv)
• --verbose: Show detailed progress

Report Interpretation

CSV columns:

• Category: dataSinks/dataSources/authFlows/dataModels
• Type: Operation type (write_operation/read_operation/etc)
• Context: 3 lines of surrounding code

Key Features

• 🔍 Automatic detection of data flows (sources/sinks)
• 🔐 Authentication flow analysis
• 🗄️ Data model extraction
• 📊 CSV report generation

Configuration

Modify detection patterns in acorn-analyzer.js:

this.patterns = {
  sinks: ['write', 'save', 'create', 'update', 'delete'],
  sources: ['get', 'find', 'fetch', 'read'],
  auth: ['jwt', 'oauth', 'auth']
};

Supported File Types

• JavaScript (.js)
• TypeScript (.ts)
• JSX/TSX

Requirements

• Node.js 16+
• npm/yarn

Usage

npm install
node codebase-analyzer.js

Output: codebase-analysis.csv, output.json

1. Acorn Analyzer (acorn-analyzer.js) Approach: Uses Acorn, a JavaScript parser with JSX support through extensions. It uses acorn-walk for AST traversal.

Pros:

Fast and lightweight parser specifically designed for JavaScript Easy to extend with plugins (like JSX support) Well-established in the ecosystem with good documentation Simple visitor pattern via acorn-walk for traversing the AST Detailed context capture for reporting with line/column info and surrounding code Cons:

Limited to JavaScript/JSX (no TypeScript support by default) Manual pattern matching using arrays of keywords Less structured query syntax compared to Tree-sitter Requires more boilerplate code for traversing complex AST structures Not as powerful for complex language constructs 2. TypeScript ESTree Analyzer (codebase-analyzer.js) Approach: Uses @typescript-eslint/typescript-estree as the parser, which is designed for TypeScript compatibility.

Pros:

Native TypeScript support, handles both JS and TS formats Compatible with ESLint ecosystem Generally robust against syntax errors Custom tree traversal method that's more flexible than Acorn's walker Handles modern JavaScript features well Good for static analysis with type information Cons:

Requires manual recursion through the AST with the traverseTree method Simpler reporting format (doesn't include column info or context) Uses simple string pattern matching for detection Possibly slower than Acorn for large codebases Less mature than Acorn but more modern 3. Tree-sitter Analyzer (tree-sitter-analyzer.js) Approach: Uses Tree-sitter, a parser generator tool and incremental parsing library with a query language.

Pros:

Powerful query language specifically designed for code analysis Declarative pattern matching using specialized syntax Supports incremental parsing (efficient for real-time analysis) Can handle syntax errors gracefully Excellent for IDE-like features that need constant reparsing Potentially extensible to many languages beyond JavaScript Clean separation between query definitions and execution logic Cons:

More complex setup with language-specific modules Higher learning curve for the query syntax Currently only using JavaScript language pack (would need additional modules for TypeScript) Less widely adopted than ESLint/Acorn in the JavaScript ecosystem Potentially higher memory usage