Skip to content

CHANGELOG.md

Latest commit

 

History

History
187 lines (147 loc) · 5.68 KB

CHANGELOG.md

File metadata and controls

187 lines (147 loc) · 5.68 KB

[0.2.1] - 2026-02-28

Summary

refactor(docs): code analysis engine

Docs

  • docs: update TODO.md
  • docs: update context.md
  • docs: update context_final.md
  • docs: update context_fixed.md
  • docs: update fast_analysis_report.md

Config

  • config: update goal.yaml

Other

  • update debug/.code2flow_cache/__init___067a3ea9a806bdcd.pkl
  • update debug/.code2flow_cache/__init___06ee3b304cbac344.pkl
  • update debug/.code2flow_cache/__init___07004ae5fc0b63a4.pkl
  • update debug/.code2flow_cache/__init___092c164e1ea3ed2a.pkl
  • update debug/.code2flow_cache/__init___1306939d2650ad0a.pkl
  • update debug/.code2flow_cache/__init___1435b739d4a93c01.pkl
  • update debug/.code2flow_cache/__init___1a3f34073e505d94.pkl
  • update debug/.code2flow_cache/__init___20b71d7ad5e01760.pkl
  • update debug/.code2flow_cache/__init___385814d063e205eb.pkl
  • update debug/.code2flow_cache/__init___563e1960e3f8fe02.pkl
  • ... and 222 more

Changelog

All notable changes to this project will be documented in this file.

The format is based on Keep a Changelog, and this project adheres to Semantic Versioning.

[0.2.0] - 2025-02-28

Added

Core Analysis Engine

  • Optimized ProjectAnalyzer with caching and parallel processing

    • FileCache with TTL support for AST parsing results
    • Parallel file analysis using ProcessPoolExecutor
    • Configurable performance settings (FAST_CONFIG, DETAILED_CONFIG)
    • Memory-efficient depth limiting for CFG generation

  • Enhanced Filtering

    • FastFileFilter with glob pattern matching
    • Exclude test files, private methods, properties
    • Min function lines threshold
    • Configurable include/exclude patterns

  • Comprehensive Code Model

    • FlowNode, FlowEdge for CFG representation
    • FunctionInfo, ClassInfo, ModuleInfo with metadata
    • Pattern detection (recursion, state machines)
    • Compact JSON/YAML output with optional full details

NLP Processing Pipeline

  • Query Normalization (1a-1e)

    • Unicode NFKC normalization
    • Lowercase conversion
    • Punctuation removal
    • Whitespace normalization
    • Stopword removal (multilingual)

  • Intent Matching (2a-2e)

    • Fuzzy matching with configurable algorithms
    • Keyword matching with weighted scoring
    • Context window scoring for disambiguation
    • Multi-intent resolution strategies (best_match, combine, sequential)

  • Entity Resolution (3a-3e)

    • Type-based entity extraction (function, class, module, variable, file)
    • Name matching with similarity threshold
    • Context-aware disambiguation
    • Hierarchical resolution (Class.method -> method)
    • Alias resolution (short -> qualified names)

  • Pipeline Integration (4a-4e)

    • Orchestration with stage tracking
    • Result aggregation and confidence scoring
    • Fallback handling for low-confidence queries
    • Formatted output with action recommendations

  • Multilingual Support

    • English and Polish query support
    • Cross-language fuzzy matching
    • Language-specific stopwords
    • YAML-driven configuration

Export Formats

  • JSON Exporter - Machine-readable analysis output
  • YAML Exporter - Human-readable with compact/full modes
  • Mermaid Exporter - Flowchart and call graph visualization
  • LLMPromptExporter - LLM-ready analysis summaries
  • GraphVisualizer - NetworkX/matplotlib PNG generation

CLI Enhancements

  • Improved argument parsing with subcommands
  • Automatic PNG generation from Mermaid files
  • LLM flow generation command
  • Verbose output with progress reporting
  • Multiple output format support

Testing & Quality

  • Comprehensive Test Suite

    • Unit tests for all core components
    • Edge case tests (empty projects, syntax errors, unicode)
    • Performance benchmarks
    • Integration tests (NLP + Analysis workflow)
    • NLP pipeline tests (steps 1a-4e validation)

  • Benchmarking

    • Performance tests for large projects
    • Cache effectiveness measurement
    • Parallel vs sequential comparison
    • Memory usage validation

Documentation

  • Complete API documentation
  • Usage examples and tutorials
  • Performance optimization guide
  • Multilingual query examples
  • Configuration reference

Changed

  • Refactored monolithic flow.py into modular package structure
  • Improved error handling throughout codebase
  • Enhanced type hints for better IDE support
  • Updated setup.py for PyPI publication readiness

Fixed

  • Import errors in CLI module
  • Attribute mismatches between models and exporters
  • Parallel processing pickle compatibility issues
  • FlowEdge attribute access (condition -> conditions)

[0.1.0] - 2025-02-20

Added

  • Initial project structure
  • Basic AST-based code analysis
  • Control flow graph generation
  • Call graph extraction
  • Pattern detection (recursion, loops)
  • Mermaid diagram export
  • Command-line interface
  • Initial test suite

Future Roadmap

Planned for 0.3.0

  • Semantic code search using embeddings
  • Advanced pattern detection (factory, singleton, observer)
  • Interactive web UI (Streamlit/Gradio)
  • VS Code extension
  • Support for additional languages (JavaScript, TypeScript)

Planned for 0.4.0

  • Real-time code analysis via file watching
  • Integration with Git for diff analysis
  • Custom pattern definition via YAML
  • Plugin system for third-party extensions
  • Docker container for easy deployment

Planned for 1.0.0

  • Complete API stability
  • Comprehensive security audit
  • Enterprise features (SSO, audit logs)
  • Performance optimizations for 100k+ LOC projects
  • Full documentation with video tutorials

Contributing

Please report issues and suggest features via GitHub Issues.

Credits

Developed by the STTS Project team.