Add comprehensive test suite (174 tests) covering encoding, schema,
image, database CRUD, and PNG round-trip. Fix critical bugs:
- PNG compression: replace non-functional zlibCompress with pako.deflate
- PNG import: add CRC validation, support all filter types (Sub/Up/Avg/Paeth)
- Input validation: validate records against schema before insert
- Index overflow: dynamic dataStartRing prevents index/data ring overlap
- Image expansion: expand before writes instead of after to prevent OOB
- update() read bug: search index from end to find latest entry, not deleted one
- String encoding: enforce 511-byte max length
- Index ring count: use 6 bits (2 pixels) instead of 3 bits for >7 ring support
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
Implement SpiralDB - a novel data storage format that encodes structured
data into PNG images using an 8-color palette. Data grows in a spiral
pattern from the center outward, enabling infinite expansion.
Features:
- 8-color palette (3-bit per pixel) for compression robustness
- Spiral coordinate system with ring-based organization
- Schema-based serialization (int, bool, string, timestamp, arrays)
- Record management with CRUD operations and status tracking
- PNG export/import with pako compression
- Browser support (Canvas, Blob, DataURL)
- Todo schema as reference implementation
Storage structure:
- Ring 0: Magic byte (validation)
- Ring 1: Database header (version, flags, counts)
- Ring 2: Schema definition
- Ring 3+: Master index (spans multiple rings if needed)
- Ring 4+: Record data
