pre-commit

#!/bin/sh

# REGEX PATTERNS 
FORBIDDEN_ADDRESS="^[A-PR-UWYZ]([0-9]{1,2}|([A-HK-Y][0-9]([0-9ABEHMNPRV-Y])?)|[0-9][A-HJKPS-UW])?[[:space:]]?[0-9][ABD-HJLNP-UW-Z]{2}$"
FORBIDDEN_PATIENT_ID="^[a-zA-Z][0-9]{6}$|^[0-9]{7}[a-zA-Z]$"
FORBIDDEN_PHONE="(\+44|07)[0-9]{9}"
FORBIDDEN_ACCESSION_NUMBER="sp-[0-9]{2}-[0-9]{7}"
FORBIDDEN_DOB="[0-1][0-9]/[0-3][0-9]/[1-2][0-9]{3}"
FORBIDDEN_DATE_TIME="[0-1][0-9]/[0-3][0-9]/[1-2][0-9]{3}\s[0-2][0-9]:[0-5][0-9]:[0-5][0-9]"
FORBIDDEN_NHS="[0-9]{10}"

git_verification_patterns=( $FORBIDDEN_ADDRESS $FORBIDDEN_PATIENT_ID $FORBIDDEN_PHONE $FORBIDDEN_ACCESSION_NUMBER $FORBIDDEN_DATE_TIME $FORBIDDEN_DOB $FORBIDDEN_NHS)
git_verification_patterns_desc=("Address" "Patient ID" "Phone number" "Accession Number" "Date and Time" "Date of Birth" "NHS Number")

# Get modified files
FILES_MODIFIED=$(git diff --cached --name-only)

NUM_FILES_CHECKED=0
NUM_FILES_OFFENCES=0

# Exceptions
exception_file=".sensitive_exceptions"
exclusion_file=".files_exceptions"

echo "-- RUNNING SENSITIVE DATA CHECKS ----------------------------------------"

for F in $FILES_MODIFIED
do
  F_basename=$(basename $F)
  if grep -Fiq -- "$F_basename" $exclusion_file; then
    continue
  fi

  for i in "${!git_verification_patterns[@]}"; do
    MATCHES=$(egrep -i --line-number "${git_verification_patterns[$i]}" "$F" || true)

    for MATCH in $MATCHES; do
      IFS=':' read -ra PARTS <<< "$MATCH"
      LINE_NUMBER=${PARTS[0]}
      CONTENT=${PARTS[1]}

      # Skip exceptions
      if echo "$CONTENT" | grep -Fiq -f $exception_file; then
        continue
      fi

      echo "FILE: $F"
      echo "   DESC: ${git_verification_patterns_desc[$i]}"
      echo "   MATCH: $MATCH"      
      echo "  "

      NUM_FILES_OFFENCES=$((NUM_FILES_OFFENCES+1))
    done
  done

  NUM_FILES_CHECKED=$((NUM_FILES_CHECKED+1))
done

echo "-- SUMMARY --------------------------------------------------------------"
echo ""
echo "   Files Checked: $NUM_FILES_CHECKED"
echo "   Num File Offences: $NUM_FILES_OFFENCES"
if [ $NUM_FILES_OFFENCES -gt 0 ]; then
  echo "   Status: FAIL"  
  echo "   "
else
  echo "   Status: OK"
  echo "   "
fi

if [ $NUM_FILES_OFFENCES -gt 0 ]; then
  exit 1
else
  exit 0
fi