Redactorator

Human-auditable regex helpers for identifying labeled sensitive data like dates of birth (DOB) and Social Security Numbers (SSNs).

Features

• Labeled DOB detection (e.g., DOB: 01/02/1980, Born on Jan 1, 1980)
• Labeled SSN detection (e.g., SSN: 123-45-6789, Social Security # 123456789)
• Labeled phone detection (e.g., Phone: 123-456-7890, Tel: (123) 456-7890)
• Labeled phone detection (e.g., Phone: 123-456-7890, Tel: (123) 456-7890)
• Generic labeled ID detection (e.g., ID: ABC-12345, Passport: A1234567, Member ID: 123456789) with an optional aggressive unlabeled mode for long alphanumeric tokens
• Small helper API for matching and detection
• Redaction helpers with fixed or length-based masking

Usage

import redactorator as redact

text = "DOB: 01/02/1980, Phone: 123-456-7890, SSN: 123-45-6789"

# Find all matches grouped by type
print(redact.find_all(text))

# Redaction (fixed mask)
print(redact.redact_all(text))

# Redaction (length-based mask)
print(redact.redact_all(text, mask_mode="length"))

# Redaction (length-based with custom char)
print(redact.redact_all(text, mask_mode="length", mask_char="#"))

# Pattern groups
dob_group = redact.PATTERNS["dobs"]
print(dob_group.find("DOB: 01/02/1980"))
print(dob_group.contains("DOB: 01/02/1980"))

ssn_group = redact.PATTERNS["ssns"]
print(ssn_group.redact("SSN: 123-45-6789"))
print(ssn_group.find("123-45-6789", mode="aggressive"))

phone_group = redact.PATTERNS["phones"]
print(phone_group.redact("PHONE: 123-456-7890")
print(phone_group.find("123-456-7890", mode="aggressive"))

ids_group = redact.PATTERNS["ids"]
print(ids_group.redact("ID: ABC-12345"))
print(ids_group.find("ABC12345678", mode="aggressive"))

Helper API

• PATTERNS: dict of pattern groups (e.g., PATTERNS["dobs"], PATTERNS["ssns"])
• find_all(text: str, mode: str = "strict") -> Dict[str, List[str]]
• redact_all(text: str, mask: str = "***", mask_mode: str = "fixed", mask_char: str = "*", mode: str = "strict") -> str

PatternGroup methods

• find(text: str, mode: str = "strict") -> List[str]
• contains(text: str, mode: str = "strict") -> bool
• redact(text: str, mask: str = "***", mask_mode: str = "fixed", mask_char: str = "*", mode: str = "strict") -> str

Masking options

• mask_mode="fixed" uses the provided mask string (default: ***).
• mask_mode="length" replaces each character in the matched value with mask_char.

Pattern modes

• mode="strict" (default) will only match when a clear label is present. (e.g., DOB:, Passport:, Phone:).
• mode="aggressive" also matches unlabeled values (DOBs, phones, SSNs).

Redaction order

redact_all applies pattern groups in the order defined by PATTERN_ORDER in redact.py.

Adding new pattern modules

1. Create a new module under patterns/ with a PatternGroup instance (e.g., PHONE_GROUP).
2. Import and register it in redact.PATTERNS and PATTERN_ORDER.

Running tests

C:/Projects/redactorator/.venv/Scripts/python.exe -m unittest discover -s tests -p "test_*.py" -v

python -m unittest discover -s tests -p "test_*.py" -v

Notes

• Strict mode is the default for all APIs and only matches labeled values.
• To use aggressive mode - call it explicitly:
  • redact.find_all(text, mode="aggressive")
  • redact.redact_all(text, mode="aggressive")
  • PATTERNS["dobs"].find(text, mode="aggressive")
  • PATTERNS["phones"].contains(text, mode="aggressive")
  • PATTERNS["ssns"].redact(text, mode="aggressive")
  • PATTERNS["ids"].find(text, mode="aggressive")