Methodology

Official source files

Names, general categories, scripts, and block assignments are generated from Unicode 15.1 UnicodeData.txt, Blocks.txt, and Scripts.txt. The same official files are stored under scripts/ucd/ for reproducible builds.

Curated editorial layer

Editorial summaries, aliases, descriptions, and related-character links are stored separately from imported facts. A source refresh can update facts without overwriting editorial copy, and every published record passes schema and referential-integrity validation.

Review and exclusion rules

Records without a complete name, category, block, script, aliases, evidence, and reviewed date are excluded from page generation. The dataset version and last reviewed date appear on character pages and in this site's footer.