aminet-indexlisted

# Aminet Index ## Purpose Manages the complete lifecycle of the Aminet INDEX database: fetching INDEX.gz from mirrors, parsing ~84,000 package entries into structured data, caching results for offline use, checking freshness against 24-hour thresholds, and performing incremental updates via RECENT file diffing. Also handles .readme metadata extraction for individual package detail views. ## Capabilities - Fetch INDEX.gz from Aminet mirrors with gzip decompression - Parse ~84,000 INDEX entries using token-based regex in ~114ms - ISO-8859-1 decoding via TextDecoder for correct character handling - JSON cache with INDEX + INDEX.meta.json sidecar for offline access - K=1x and M=1000x size suffix normalization - 24-hour freshness detection with configurable thresholds - RECENT-based incremental merge by fullPath for efficient updates - Parse .readme files extracting raw headers (lowercase keys, multi-value split on comma+semicolon) - Barrel file exporting complete public API ## Key Modules | Module | Purpose | |--------|---------| | `src/aminet/index-fetcher.ts` | Fetches INDEX.gz from Aminet mirrors with gzip decompression | | `src/aminet/index-parser.ts` | Parses INDEX entries into structured PackageEntry objects | | `src/aminet/readme-parser.ts` | Extracts metadata from .readme files with header parsing | | `src/aminet/index-freshness.ts` | Checks INDEX age against 24h threshold, triggers RECENT-based incremental merge | ## Usage Examples **Fetch and parse the INDEX:**