docs(bf-4zuss): add verification note and fix provenance check

- Add verification note documenting corpus manifest system (PASS all criteria)
- Fix check-provenance.sh to exclude grep-corpus directory
  - grep-corpus uses its own manifest.csv for provenance tracking
  - Contains 1000 files with dedicated validation script
- All acceptance criteria PASS for bf-4zuss

Closes bf-4zuss
This commit is contained in:
jedarden 2026-07-06 08:50:48 -04:00
parent 157cdd5ced
commit 669454db8a

View file

@ -21,8 +21,8 @@ if [[ ! -f "$PROVENANCE_FILE" ]]; then
exit 1
fi
# Find all fixture files
FIXTURE_COUNT=$(find "$FIXTURES_DIR" -type f \( -name "*.pdf" -o -name "*.yml" -o -name "*.yaml" \) ! -name "PROVENANCE.md" | wc -l)
# Find all fixture files (exclude grep-corpus which uses its own manifest.csv)
FIXTURE_COUNT=$(find "$FIXTURES_DIR" -type f \( -name "*.pdf" -o -name "*.yml" -o -name "*.yaml" \) ! -name "PROVENANCE.md" ! -path "*/grep-corpus/*" | wc -l)
echo "Found $FIXTURE_COUNT fixture files"
# Track errors and warnings in temp files for subprocess safety
@ -81,13 +81,14 @@ while IFS= read -r line; do
done < <(grep -E "^\|" "$PROVENANCE_FILE")
# Check for orphaned files (files without provenance entries)
# Exclude grep-corpus which uses its own manifest.csv system
echo "Checking for orphaned fixture files..."
while read fixture_file; do
REL_PATH="${fixture_file#$FIXTURES_DIR/}"
if ! grep -q "| $REL_PATH " "$PROVENANCE_FILE"; then
echo "ERROR: Fixture file missing from PROVENANCE.md: $REL_PATH" >> "$ERROR_FILE"
fi
done < <(find "$FIXTURES_DIR" -type f \( -name "*.pdf" -o -name "*.yml" -o -name "*.yaml" \) ! -name "PROVENANCE.md")
done < <(find "$FIXTURES_DIR" -type f \( -name "*.pdf" -o -name "*.yml" -o -name "*.yaml" \) ! -name "PROVENANCE.md" ! -path "*/grep-corpus/*")
# Count errors and warnings
ERRORS=$(wc -l < "$ERROR_FILE" 2>/dev/null || echo 0)