An async token-bucket rate limiter with configurable rate and burst, plus random jitter. It is fully self-contained, imported by downloaders and API clients to throttle requests. The implementation uses an asyncio lock and a simple linear token model, making it easy to extract into any async Python project requiring rate-limited third-party API calls.

An adjustable-size async worker pool that manages concurrent download tasks. It supports dynamic max_workers changes, safe cancellation (CancelledError handling), and a release method that is explicitly synchronous to avoid event loop issues. This pattern is reusable for any async batch processing where concurrency must be controlled and tasks may be added or aborted mid-flight.

A cross-platform filename sanitizer and template renderer for output file naming. It handles Windows reserved names, control characters, path length limits, and template variables (e.g., {nickname}, {aweme_id}) with auto-truncation. The core functions are pure, stateless, and depend only on Python stdlib, making them directly reusable in any project that needs safe, configurable file naming.

A YAML configuration loader with environment variable overrides (prefixed DOUYIN_), alias normalization, schema validation, and logging of masked credentials. It merges default config, file config, and env overrides in a well-structured pipeline. The pattern—loading from YAML, overriding with env vars, validating, and merging—is generic and could be extracted for any Python application needing flexible, layered configuration.