Smart Pip Collector ^v0.1.0

• Create a local pip package repository
• Work in isolated environments without internet access
• Backup dependencies of all your projects
• Migrate projects between machines
• Protection against PyPI blocks

• Recursive search — finds all requirements.txt in nested folders
• Version awareness — downloads specific package versions, creating package/version/ structure
• No duplicates — groups same packages with different versions
• Multithreading — uses all CPU cores for maximum speed
• Detailed reports — JSON reports for each download + global report
• Graceful shutdown — proper Ctrl+C handling (1 press — stop, 2 — force)
• Project tracking — stores information about which projects use which dependency
• Resume mode — retry failed packages

git clone https://github.com/smartlegionlab/smart-pip-collector.git
cd smart-pip-collector

# Basic run
python app.py /path/to/your/projects

# With custom output directory
python app.py /path/to/your/projects --output ~/my_pip_backup

# With custom thread count
python app.py /path/to/your/projects --workers 16

# Resume failed downloads
python app.py /path/to/your/projects --resume

~/pip_packages/
├── requests/
│   └── 2.28.1/
│       ├── requests-2.28.1-py3-none-any.whl
│       └── .download_report.json
├── django/
│   └── 4.2.1/
│       ├── Django-4.2.1-py3-none-any.whl
│       └── .download_report.json
├── global_download_report.json
└── all_requirements.txt

Local report (in each package folder):

{
  "package": "requests",
  "version": "2.28.1",
  "download_time": "2024-01-15T10:30:00",
  "source_projects": ["project_a", "project_b"],
  "status": "success",
  "files": ["requests-2.28.1-py3-none-any.whl"]
}

Global report (global_download_report.json):

• Overall statistics
• Detailed information for each package
• Rate limiting and cancellation markers

• Language: Python 3.8+
• Platform: Windows / Linux / macOS
• License: BSD 3-Clause

• Packages requiring compilation (with C extensions) may fail without system dependencies
• Non-standard links (git+, http) are ignored
• PyPI rate limiting is possible with very large number of requests