Release v1.1.0

News feed server, reusable parsing specs, caching, and multiple output formats.

Added

• serve — local HTTP feed server (standard-library only) that turns any page URL into a feed on demand over GET (GET /feed?url=<page>&format=atom), plus /health and / endpoints. Feed URLs can be pasted straight into any RSS reader.
• analyze — runs the dynamic heuristics once and distills them into a reusable YAML parsing spec; new --spec / -s option for extract runs fast, deterministic extraction from a pre-built spec.
• Reusable parsing specs via newsworker.spec (FeedSpec, SpecAnalyzer, SpecExtractor) using deterministic CSS/XPath selectors.
• Caching layer (newsworker.cache) with a spec cache and a content cache (configurable TTL). New --no-cache, --refresh, --config / -c options for extract, and --cache-dir / --content-ttl for serve.
• Settings support (newsworker.settings) backed by a YAML config file at ~/.newsworker/config.yaml.
• High-level newsworker.service.FeedService shared by extract and the server.
• Multiple output formats: extract supports json/rss/atom/csv; scan supports json/rss/atom/csv/opml. New --format / -f and --output / -o options, plus a new newsworker.formats module.

Changed

• Rewrote README.md with a modern structure and full CLI/caching documentation.
• extract now builds and caches a parsing spec (plus page content) on first use for faster subsequent runs.
• scan now emits structured, format-aware output.
• Added cssselect, pyyaml, requests, urllib3 dependencies (and declared feedparser explicitly).

Fixed

• Naive datetimes are normalized to UTC when rendering RSS/Atom feeds.

Full changelog: https://github.com/ivbeg/newsworker/blob/master/CHANGELOG.md