Update app.py

app.py

@@ -3,6 +3,20 @@ import re
 import gradio as gr
 from transformers import AutoTokenizer, AutoModelForTokenClassification, pipeline
 
+def sanitize_sensitive_info(text, keyword_string, replace_word):
+    # 📍 기관 키워드 치환
+    keywords = [k.strip() for k in keyword_string.split(",") if k.strip()]
+    for kw in keywords:
+        pattern = rf"{re.escape(kw)}(?=\W|$)"
+        text = re.sub(pattern, replace_word, text, flags=re.IGNORECASE)
+
+    # 📍 기본 민감정보 마스킹 예시 (이메일 앞부분 마스킹)
+    text = re.sub(r"\b[\w\.-]+@", "******@", text)
+
+    return text
+
+
+
 # =============================================
 # Configurable Constants
 # =============================================