# Normalize line endings so release-source fingerprints and evidence
# SHA-256 digests are stable across platforms. Release-relevant files
# (Python sources, TOML metadata, release/measurement scripts, and all
# benchmark evidence JSON) must stay LF in every working tree; Windows
# checkouts otherwise rewrite them to CRLF and change read_bytes() digests.
*.py text eol=lf
*.toml text eol=lf
*.json text eol=lf
*.yml text eol=lf
*.yaml text eol=lf
*.cff text eol=lf
*.md text eol=lf
*.sh text eol=lf
*.csv text eol=lf