Files
Understand-Anything/understand-anything-plugin/skills/understand/languages/python.md
T
Sreeram b5eb2e6041 feat: language-agnostic analysis with config-driven registry and framework detection
Replace the TypeScript/JavaScript-biased analysis pipeline with a truly
language-agnostic system. The core architecture was already language-neutral
(graph schema, dashboard, search) — the bias lived in agent prompts,
tree-sitter plugin, and language-lesson system.

Core changes:
- LanguageConfig + FrameworkConfig types with Zod validation
- LanguageRegistry (12 languages) and FrameworkRegistry (10 frameworks)
- Config-driven TreeSitterPlugin replacing hardcoded TS/JS grammars
- PluginRegistry now delegates to LanguageRegistry for extension mapping
- Language-lesson system uses config for display names and concepts

Prompt system:
- SKILL.md generalized: dynamic injection of language snippets and
  framework addendums instead of hardcoded if/else conditionals
- 12 language prompt snippets (languages/*.md) with concepts, patterns,
  frameworks per language
- 10 framework addendums (frameworks/*.md) with canonical file roles,
  edge patterns, architectural layers — Django/FastAPI/Flask preserved
  and split, plus React/Next.js/Express/Vue/Spring/Rails/Gin added
- Extended entry points, directory patterns, and test patterns across
  all 12 language ecosystems in base prompts
2026-03-23 12:35:09 +05:30

49 lines
2.6 KiB
Markdown

# Python Language Prompt Snippet
## Key Concepts
- **Decorators**: Functions that wrap other functions or classes using `@decorator` syntax
- **List/Dict Comprehensions**: Concise syntax for creating collections from iterables
- **Generators and Yield**: Lazy iterators using `yield` for memory-efficient data processing
- **Context Managers**: `with` statement for resource management via `__enter__`/`__exit__`
- **Type Hints and Typing Module**: Optional static type annotations for tooling and documentation
- **Dunder Methods**: Special methods like `__init__`, `__repr__`, `__eq__` defining object behavior
- **Metaclasses**: Classes that define how other classes are created (type as default metaclass)
- **Dataclasses**: `@dataclass` decorator auto-generating boilerplate from field annotations
- **Protocols**: Structural subtyping via `typing.Protocol` for duck-type-safe interfaces
- **Descriptors**: Objects defining `__get__`, `__set__`, `__delete__` to customize attribute access
- **Async/Await with Asyncio**: Cooperative concurrency using coroutines and an event loop
## Import Patterns
- `from module import name` — import specific name from module
- `import module` — import entire module, access via `module.name`
- `from package.module import name` — absolute import from nested package
- `from . import relative` — relative import within a package
## File Patterns
- `__init__.py` — package initializer (barrel equivalent), can re-export public API
- `__main__.py` — package entry point when run with `python -m package`
- `conftest.py` — pytest shared fixtures and hooks (auto-discovered)
- `setup.py` / `pyproject.toml` — project configuration and build metadata
- `requirements.txt` — pinned dependency list
## Common Frameworks
- **Django** — Full-stack web framework with ORM, admin, and batteries included
- **FastAPI** — Modern async API framework with automatic OpenAPI docs
- **Flask** — Lightweight WSGI micro-framework for web applications
- **SQLAlchemy** — SQL toolkit and ORM with unit-of-work pattern
- **Celery** — Distributed task queue for background job processing
- **Pydantic** — Data validation and settings management using type annotations
## Example Language Notes
> Uses `@dataclass` decorator to auto-generate `__init__`, `__repr__`, and `__eq__` from
> field annotations. This eliminates boilerplate while keeping the class definition
> readable and the generated methods consistent.
>
> When `__init__.py` re-exports symbols, it acts as the package's public API surface —
> consumers import from the package rather than reaching into internal modules.