Update app.py

app.py

@@ -434,10 +434,9 @@ def execute_unified_search(base_std, base_ver, base_cat, comp_std, comp_ver, com
             df_comp['merge_key'] = df_comp[real_anchors_c].apply(lambda row: '-'.join([clean_key_val(x) for x in row]), axis=1)
 
             # =========================================================
-            # 💡 [핵심 업그레이드] 매핑 라우터: 어떤 장부를 읽을지 결정합니다!
+            # 💡 [방어막 1] Target_Table 이름 깔끔하게 가져오기
             # =========================================================
             conn_map = sqlite3.connect(os.path.join(UPLOAD_DIR, "mapping.db"))
-            
             registry_query = """
                 SELECT Target_Table FROM Mapping_registry 
                 WHERE TRIM(Base_std)=? AND TRIM(Base_ver)=? AND TRIM(Comp_std)=? AND TRIM(Comp_ver)=?
@@ -445,16 +444,25 @@ def execute_unified_search(base_std, base_ver, base_cat, comp_std, comp_ver, com
             """
             reg_df = pd.read_sql(registry_query, conn_map, params=[base_std.strip(), base_ver.strip(), comp_std.strip(), comp_ver.strip()])
             
-            # DB에 Target_Table이 적혀있으면 그걸 쓰고, 없거나 비어있으면 기본 장부(Mapping_table)를 씁니다.
-            target_table_name = reg_df.iloc[0]['Target_Table'] if not reg_df.empty and pd.notna(reg_df.iloc[0]['Target_Table']) else "Mapping_table"
+            target_table_name = "Mapping_table"
+            if not reg_df.empty and pd.notna(reg_df.iloc[0]['Target_Table']):
+                val = str(reg_df.iloc[0]['Target_Table']).strip()
+                if val and val.lower() not in ["none", "nan"]:
+                    target_table_name = val
 
-            q_fw = f"SELECT * FROM [{target_table_name}] WHERE TRIM(Base_std)=? AND TRIM(Base_ver)=? AND TRIM(Comp_std)=? AND TRIM(Comp_ver)=?"
-            df_fw = pd.read_sql(q_fw, conn_map, params=[base_std.strip(), base_ver.strip(), comp_std.strip(), comp_ver.strip()])
-            
-            q_rv = f"SELECT * FROM [{target_table_name}] WHERE TRIM(Comp_std)=? AND TRIM(Comp_ver)=? AND TRIM(Base_std)=? AND TRIM(Base_ver)=?"
-            df_rv = pd.read_sql(q_rv, conn_map, params=[base_std.strip(), base_ver.strip(), comp_std.strip(), comp_ver.strip()])
-            conn_map.close()
             # =========================================================
+            # 💡 [방어막 2] 매핑 장부 안전하게 읽어오기 (오타나면 에러 띄움)
+            # =========================================================
+            try:
+                q_fw = f"SELECT * FROM [{target_table_name}] WHERE TRIM(Base_std)=? AND TRIM(Base_ver)=? AND TRIM(Comp_std)=? AND TRIM(Comp_ver)=?"
+                df_fw = pd.read_sql(q_fw, conn_map, params=[base_std.strip(), base_ver.strip(), comp_std.strip(), comp_ver.strip()])
+                
+                q_rv = f"SELECT * FROM [{target_table_name}] WHERE TRIM(Comp_std)=? AND TRIM(Comp_ver)=? AND TRIM(Base_std)=? AND TRIM(Base_ver)=?"
+                df_rv = pd.read_sql(q_rv, conn_map, params=[base_std.strip(), base_ver.strip(), comp_std.strip(), comp_ver.strip()])
+            except Exception as sql_e:
+                conn_map.close()
+                return pd.DataFrame({"Error": [f"매핑 장부 '{target_table_name}'을 여는 데 실패했습니다. 테이블 이름을 확인하세요: {str(sql_e)}"]})
+            conn_map.close()
 
             cols_fw_lower = {c.lower(): c for c in df_fw.columns}
             
@@ -474,6 +482,9 @@ def execute_unified_search(base_std, base_ver, base_cat, comp_std, comp_ver, com
             b_sec = cols_fw_lower.get('base_section', 'Base_section')
             c_sec = cols_fw_lower.get('comp_section', 'Comp_section')
             
+            if b_sec not in df_fw.columns or c_sec not in df_fw.columns:
+                return pd.DataFrame({"Error": [f"'{target_table_name}' 장부에 '{b_sec}' 또는 '{c_sec}' 열이 없습니다. 대소문자를 확인하세요."]})
+            
             df_fw = df_fw[[b_sec, c_sec]].rename(columns={b_sec: 'Base_section', c_sec: 'Comp_section'})
             if not df_rv.empty:
                 df_rv = df_rv[[b_sec, c_sec]].rename(columns={b_sec: 'Comp_section', c_sec: 'Base_section'})
@@ -489,6 +500,11 @@ def execute_unified_search(base_std, base_ver, base_cat, comp_std, comp_ver, com
                 
                 df_mapping['Base_section'] = df_mapping['Base_section'].apply(clean_key_val)
                 df_mapping['Comp_section'] = df_mapping['Comp_section'].apply(clean_key_val)
+                
+                # =========================================================
+                # 💣 [방어막 3] 카테시안 폭발 방지: 빈칸끼리 연결된 가짜 매핑 삭제
+                # =========================================================
+                df_mapping = df_mapping[(df_mapping['Base_section'] != "") & (df_mapping['Comp_section'] != "")]
                 df_mapping = df_mapping.dropna().drop_duplicates()
             else:
                 df_mapping = pd.DataFrame(columns=['Base_section', 'Comp_section'])
@@ -523,16 +539,12 @@ def execute_unified_search(base_std, base_ver, base_cat, comp_std, comp_ver, com
                 for c in rename_b.values(): row_dict[c] = str(row[c]) if has_b and not pd.isna(row[c]) else ""
                 for c in rename_c.values(): row_dict[c] = str(row[c]) if has_c and not pd.isna(row[c]) else ""
                 
-                # =======================================================
-                # 💡 [깐깐한 빈칸 검사] 양쪽 다 글자가 채워져 있는지 확인!
-                # =======================================================
                 if mapped_only:
                     b_has_val = any(str(row_dict[c]).strip() not in ["", "nan", "None", " "] for c in rename_b.values())
                     c_has_val = any(str(row_dict[c]).strip() not in ["", "nan", "None", " "] for c in rename_c.values())
                     
                     if not (b_has_val and c_has_val):
                         continue
-                # =======================================================
                 
                 if has_b and has_c:
                     for orig_col in rename_b.keys():
@@ -565,8 +577,11 @@ def execute_unified_search(base_std, base_ver, base_cat, comp_std, comp_ver, com
     except Exception as e:
         import traceback
         traceback.print_exc()
-        return pd.DataFrame({"Error": [f"시스템 오류 발생: {str(e)}"]})
-
+        # 시스템 에러 대신 화면에 깔끔하게 원인을 띄워줍니다.
+        error_msg = str(e)
+        if "database is locked" in error_msg.lower():
+            return pd.DataFrame({"Error": ["🚨 DB가 잠겨있습니다! 켜놓으신 'DB Browser' 프로그램을 완전히 종료한 뒤 다시 조회해 주세요."]})
+        return pd.DataFrame({"Error": [f"시스템 오류 발생: {error_msg}"]})
 
 # ==========================================
 # 4. UI Layout & Event Binding