Rajan Sharma
Create privacy.py
d0e3a07 verified
raw
history blame
667 Bytes
import re
# Simple, configurable redaction. Replace with your enterprise DLP when ready.
RE_PATTERNS = {
"MRN": re.compile(r"\b(?:MRN|Med(?:ical)?\s*Record)\s*[:#]?\s*\d{5,10}\b", re.I),
"DOB": re.compile(r"\b(?:DOB|DoB|Birth\s*Date)\s*[:#]?\s*(\d{4}[-/]\d{2}[-/]\d{2}|\d{2}[-/]\d{2}[-/]\d{4})\b", re.I),
"PHONE": re.compile(r"\b(?:\+?\d{1,2}\s*)?(?:\(?\d{3}\)?[\s.-]?\d{3}[\s.-]?\d{4})\b"),
"EMAIL": re.compile(r"\b[A-Za-z0-9._%+-]+@[A-Za-z0-9.-]+\.[A-Za-z]{2,}\b"),
}
def redact_text(text: str) -> str:
if not text:
return text
red = text
for pat in RE_PATTERNS.values():
red = pat.sub("[REDACTED]", red)
return red