use once_cell::sync::Lazy; use regex::Regex; use serde::{Deserialize, Serialize}; #[derive(Debug, Clone, Serialize, Deserialize)] pub struct PiiMatch { pub entity_type: String, pub original: String, pub placeholder: String, pub start: usize, pub end: usize, } #[derive(Debug, Clone, Serialize, Deserialize)] pub struct RedactionResult { pub redacted_text: String, pub pii_map: Vec, } static PATTERNS: Lazy> = Lazy::new(compile_patterns); fn compile_patterns() -> Vec<(Regex, &'static str)> { let mut patterns = Vec::new(); for (pattern, label) in [ (r"\b(?:Dr\.|Dr|Professor|Prof\.)\s+[A-Z][a-z]+\b", "PROVIDER_NAME"), (r"\b(?:Patient|Pt\.|Mr\.|Mrs\.|Ms\.)\s+[A-Z][a-z]+(?:\s+[A-Z][a-z]+)+\b", "PERSON_NAME"), (r"\b[A-Z][a-z]+(?:\s+[A-Z][a-z]+)+\b", "PERSON_NAME"), (r"\b\d{1,2}/\d{1,2}/\d{2,4}\b", "DATE"), (r"\b\d{4}-\d{2}-\d{2}\b", "DATE"), (r"\b\d{3}-\d{2}-\d{4}\b", "SSN"), (r"\bMRN[:\s-]*[A-Z0-9-]{3,}\b", "MRN"), (r"\b\d{10}\b", "PHONE"), (r"\b\+?\d[\d\s().-]{7,}\b", "PHONE"), (r"\b[A-Z]{2}\d{2}\s?\d{2}\s?\d{2}\s?\d\b", "NHS_NUMBER"), (r"\b\d{2,3}\s?(?:years?|yo|yrs?)\b", "AGE"), (r"\b[A-Za-z0-9._%+-]+@[A-Za-z0-9.-]+\.[A-Za-z]{2,}\b", "EMAIL"), (r"\b(?:insurance|member|policy)\s*(?:id|number|#)?\s*[:=]?\s*[A-Z0-9-]{5,}\b", "INSURANCE_ID"), (r"\b(?:\d{1,3}\s)?[A-Za-z]+\s(?:Street|St\.|Road|Rd\.|Avenue|Ave\.|Boulevard|Blvd\.|Lane|Ln\.)\b", "ADDRESS"), ] { if let Ok(re) = Regex::new(pattern) { patterns.push((re, label)); } } patterns } pub fn redact_pii(raw_text: &str) -> RedactionResult { let mut spans: Vec<(usize, usize, &'static str, String)> = Vec::new(); for (re, label) in PATTERNS.iter() { for mat in re.find_iter(raw_text) { spans.push((mat.start(), mat.end(), *label, mat.as_str().to_string())); } } spans.sort_by(|a, b| a.0.cmp(&b.0).then_with(|| (b.1 - b.0).cmp(&(a.1 - a.0)))); let mut selected: Vec<(usize, usize, &'static str, String)> = Vec::new(); let mut cursor = 0usize; for span in spans { if span.0 >= cursor { cursor = span.1; selected.push(span); } } let mut counts = std::collections::HashMap::<&'static str, usize>::new(); let mut redacted = String::with_capacity(raw_text.len() + 64); let mut pii_map = Vec::with_capacity(selected.len()); let mut index = 0usize; for (start, end, label, original) in selected { if start > index { redacted.push_str(&raw_text[index..start]); } let count = counts.entry(label).or_insert(0); *count += 1; let placeholder = format!("[{}_{}]", label, count); redacted.push_str(&placeholder); pii_map.push(PiiMatch { entity_type: label.to_string(), original, placeholder, start, end, }); index = end; } if index < raw_text.len() { redacted.push_str(&raw_text[index..]); } RedactionResult { redacted_text: redacted, pii_map } } #[cfg(test)] mod tests { use super::*; #[test] fn redaction_replaces_sensitive_text() { let input = "Patient John Smith, 45 yo, MRN 847291A, phone 5551234567, SSN 123-45-6789."; let result = redact_pii(input); assert!(!result.redacted_text.contains("John Smith")); assert!(!result.redacted_text.contains("5551234567")); assert!(result.redacted_text.contains("[PERSON_NAME_1]") || result.redacted_text.contains("[PERSON_NAME_2]")); assert!(result.redacted_text.contains("[PHONE_1]")); assert!(result.redacted_text.contains("[SSN_1]")); } }