v1.0.0
Release: Lazy execution–based crawling core
This release introduces lazy execution chains as the core model of antyr.
HTTP fetching, streaming, extraction, and persistence are now expressed through
result objects that defer execution until explicitly consumed. This establishes
predictable I/O boundaries, explicit lifetimes, and deterministic cleanup across
the toolkit.
Highlights
- Lazy, result-based execution model for HTTP operations
- Trio-based execution semantics with structured cancellation
- Simplified and finalized
HttpCrawlerinterface - Streaming-first processing for responses and extracted content
- ZIP handling built on the same streaming primitives
- Safe persistence of streamed content to disk
- IP rotation support for HTTP crawling
- Comprehensive test coverage, including property-based tests
- Complete module-level docstrings and aligned documentation
- Project renamed to
antyr, with updated packaging and dependencies
This release stabilizes the foundational execution model and prepares the project
for further extensions without introducing framework-level assumptions.
Full Changelog: v0.1.5...v1.0.0