Skip to content

README.md

Latest commit

 

History

History
178 lines (139 loc) · 5.94 KB

README.md

File metadata and controls

178 lines (139 loc) · 5.94 KB

Vallm Examples

This directory contains comprehensive examples demonstrating vallm's code validation and analysis capabilities.

📁 Example Structure

Each example is organized in its own folder with:

  • main.py - The main example script
  • README.md - Detailed explanation of what the example demonstrates
  • .vallm/ - Generated analysis data (created after running)

🚀 Quick Start

Run All Examples

# Run all examples sequentially
./run.sh

Run Individual Examples

cd 01_basic_validation
python main.py

📋 Examples Overview

1. Basic Validation (01_basic_validation/)

Demonstrates: Core validation pipeline (syntax, imports, complexity)

  • Good code: Clean Fibonacci implementation
  • Bad code: Syntax error example
  • Complex code: Deeply nested conditionals

2. AST Comparison (02_ast_comparison/)

Demonstrates: Abstract Syntax Tree analysis and similarity scoring

  • Python AST similarity: Comparing function implementations
  • Multi-language parsing: JavaScript and C code analysis
  • Structural diffing: Code change detection

3. Security Check (03_security_check/)

Demonstrates: Security vulnerability detection

  • Insecure patterns: eval(), exec(), os.system, pickle.loads
  • Safe alternatives: Secure coding practices
  • Risk assessment: Security scoring and issue reporting

4. Graph Analysis (04_graph_analysis/)

Demonstrates: Code graph building and structural analysis

  • Import graphs: Module dependency tracking
  • Call graphs: Function relationship mapping
  • Change detection: Breaking change identification

5. LLM Semantic Review (05_llm_semantic_review/)

Demonstrates: AI-powered semantic code analysis

  • Prerequisites: Ollama with qwen2.5-coder:7b model
  • Bug detection: Subtle logic error identification
  • Code quality: Style and practice evaluation

6. Multi-language Validation (06_multilang_validation/)

Demonstrates: Cross-language validation capabilities

  • JavaScript: ES6+ syntax and complexity
  • C: C99/C11 standard compliance
  • Universal interface: Consistent API across languages

7. Multi-Language (07_multi_language/)

Demonstrates: Comprehensive multi-language support with 30+ languages

  • Auto-detection: Language detection from file extensions
  • 30+ Languages: Python, Go, Rust, Java, C/C++, Zig, Julia, Nim, V, and more
  • Batch validation: Validate mixed-language projects

8. code2llm Integration (08_code2llm_integration/)

Demonstrates: Integration with code2llm for project analysis

  • Structure analysis: code2llm analyzes project layout
  • Quality validation: vallm validates each file
  • Combined reports: Unified structural + quality metrics

9. code2logic Integration (09_code2logic_integration/)

Demonstrates: Integration with code2logic for logical analysis

  • Control flow: Extract branches, loops, conditions
  • Call graphs: Visualize function dependencies
  • Complexity analysis: Identify high-complexity functions

10-12. Additional Examples

  • MCP Ollama Demo (10_mcp_ollama_demo/) - MCP server integration
  • Claude Code Autonomous (11_claude_code_autonomous/) - Autonomous workflows
  • Ollama Simple Demo (12_ollama_simple_demo/) - Basic Ollama integration

13. Batch Processing (13_batch_processing/)

Demonstrates: Batch validation of multiple files

  • Directory scanning: Automatic file discovery
  • Batch reports: TOON-formatted output
  • Result aggregation: Summary statistics

14. Advanced API Usage (14_api_advanced/)

Demonstrates: Programmatic API patterns

  • Proposal creation: Multiple creation methods
  • Settings customization: Validator configuration
  • Workflow integration: CI/CD quality gates

15. CLI Usage (15_cli_usage/)

Demonstrates: Command-line interface patterns

  • CLI invocation: Programmatic CLI usage
  • Output formats: Text and JSON handling
  • Exit codes: CI/CD automation

16. Configuration (16_configuration/)

Demonstrates: Configuration management

  • Config files: JSON-based configuration
  • Environment variables: Runtime configuration
  • Profiles: Predefined validator sets

📊 Analysis Data

Each example generates analysis data in its .vallm/ folder:

  • JSON reports: Structured validation results
  • Score breakdowns: Detailed metric analysis
  • Issue details: Specific problems found
  • Comparison data: Before/after analysis

🛠️ Configuration

Basic Examples

Examples 1-4 work out-of-the-box with vallm's default configuration.

LLM Example (Example 5)

Requires additional setup:

# Install Ollama
curl -fsSL https://ollama.ai/install.sh | sh

# Pull the model
ollama pull qwen2.5-coder:7b

# Start Ollama service
ollama serve

# Install Python package
pip install ollama

Multi-language Example (Example 6)

Supports JavaScript and C out-of-the-box using tree-sitter parsers.

📈 Expected Output

Each example provides:

  • Real-time validation: Live output during execution
  • Summary reports: End-of-run analysis summary
  • Data persistence: Results saved to .vallm/ folder
  • Error handling: Graceful failure reporting

🔧 Troubleshooting

Common Issues

  1. Import errors: Ensure vallm is properly installed
  2. LLM connection: Check Ollama service is running for example 5
  3. Permission errors: Ensure write access for .vallm/ folder creation

Debug Mode

Run individual examples to isolate issues:

cd 01_basic_validation
python main.py

📚 Further Reading

🤝 Contributing

To add new examples:

  1. Create a new folder with numbered prefix (17_*)
  2. Add main.py with example code
  3. Create README.md with documentation
  4. Include analysis data generation
  5. Update this overview with the new example description