Update app.py

app.py

@@ -4,6 +4,7 @@ import pandas as pd
 import os
 import re
 import base64
+import difflib
 
 # 1️⃣ 저장소 설정
 UPLOAD_DIR = "uploaded_dbs"
@@ -110,6 +111,37 @@ def reset_base():
 def reset_comp():
     return None, None, None
 
+# --------------------------
+# 🔥 diff 함수 (핵심)
+# --------------------------
+def highlight_comp_only(base_text, comp_text):
+    if pd.isna(base_text): base_text = ""
+    if pd.isna(comp_text): comp_text = ""
+
+    base_words = str(base_text).split()
+    comp_words = str(comp_text).split()
+
+    d = difflib.ndiff(base_words, comp_words)
+
+    result = []
+
+    for diff in d:
+        code = diff[0]
+        word = diff[2:]
+
+        if code == ' ':
+            result.append(word)
+
+        elif code == '+':  # 추가
+            result.append(
+                f"<span style='color:#2ecc71; font-weight:600'>{word}</span>"
+            )
+
+        elif code == '-':
+            continue
+
+    return " ".join(result)
+
 # --------------------------
 # 데이터 조회
 # --------------------------
@@ -202,18 +234,22 @@ def unified_search(bs, bv, bc, cs, cv, cc):
         if 'section' not in df_base.columns or 'section' not in df_comp.columns:
             return pd.DataFrame({"Error": ["section 컬럼 없음"]})
 
-        merged = pd.merge(
-            df_base,
-            df_comp,
-            on="section",
-            how="outer"
-        )
+        merged = pd.merge(df_base, df_comp, on="section", how="outer")
 
         merged = merged.rename(columns={
             "description_x": f"Description_{bv}",
             "description_y": f"Description_{cv}"
         })
 
+        # 🔥 diff 적용 (comp에만)
+        base_col = f"Description_{bv}"
+        comp_col = f"Description_{cv}"
+
+        for idx, row in merged.iterrows():
+            base_text = row.get(base_col, "")
+            comp_text = row.get(comp_col, "")
+            merged.at[idx, comp_col] = highlight_comp_only(base_text, comp_text)
+
         merged = merged.sort_values(
             by="section",
             key=lambda col: col.map(natural_sort_key)
@@ -230,7 +266,6 @@ with gr.Blocks() as demo:
 
     gr.Markdown("# 📜 Regulation Viewer")
 
-    # 선택 영역
     with gr.Row():
         with gr.Column():
             gr.Markdown("### 📌 기준 법규")
@@ -246,20 +281,13 @@ with gr.Blocks() as demo:
             comp_category = gr.Dropdown(label="Category")
             comp_reset_btn = gr.Button("↺ 초기화")
 
-    # 버튼
     with gr.Row():
         search_btn = gr.Button("🔍 조회")
 
-    # 결과 영역
     gr.Markdown("### 📊 비교 결과")
 
-    output_df = gr.Dataframe(
-        wrap=True,
-        interactive=False,
-        datatype="html"
-    )
+    output_df = gr.Dataframe(wrap=True, interactive=False, datatype="html")
 
-    # 이벤트
     demo.load(on_load, None, base_standard)
     demo.load(on_load, None, comp_standard)
 
@@ -276,7 +304,6 @@ with gr.Blocks() as demo:
         output_df
     )
 
-    # 초기화 버튼 연결
     base_reset_btn.click(reset_base, None, [base_standard, base_version, base_category])
     comp_reset_btn.click(reset_comp, None, [comp_standard, comp_version, comp_category])
 
@@ -293,7 +320,6 @@ if __name__ == "__main__":
     font-family: 'Pretendard', sans-serif;
 }
 
-/* 헤더 고정 */
 thead th {
     position: sticky !important;
     top: 0;
@@ -302,13 +328,11 @@ thead th {
     border-bottom: 2px solid #ddd;
 }
 
-/* 스크롤 */
 .dataframe {
     max-height: 700px;
     overflow-y: auto;
 }
 
-/* 텍스트 */
 td {
     white-space: normal !important;
     word-break: break-word !important;