Changelog

All notable changes to this project are documented in this file.

[1.0.0]

Production-stable release of the independent extraction engine.
CLI: news-fetch, news-fetch batch (JSONL), news-fetch discover.
fetch_iter for memory-conscious streaming bulk results.
Strategy plugin API: register_strategy / ExtractionStrategy / CallableStrategy.
Optional browser rendering: pip install news-fetch[browser] + render=True / browser_fallback=True.
Disk cache (cache=True) and respect_robots=True.
HTTP 429 Retry-After handling in the sync client.
Extra HTML regression fixtures.

Version / user-agent bumped to 1.0; classifier set to Production/Stable.
README optimized for PyPI discoverability (news scraper / article extractor queries).
In-repo docs/ removed (documentation lives in a separate repository).

Evidence → rank → confidence architecture (article.confidence, article.extraction).
Page-type detection and strict / threshold gates.
LowConfidenceExtractionError, article.to_json().

Proxy / proxy pools, rotation, request_delay, progress callbacks, iter_fetch.