import re import pandas as pd def normalize(col_name): """Normalize nama kolom - lowercase, collapse spaces, and strip whitespace""" if pd.isna(col_name): return "" # Collapse multiple whitespace characters into single space cleaned = re.sub(r'\s+', ' ', str(col_name)) return cleaned.strip().lower() def find_column(source_cols, possible_names): """ Cari kolom dari beberapa kemungkinan nama. Args: source_cols: list kolom dari DataFrame possible_names: list kemungkinan nama kolom Returns: Nama kolom yang ditemukan, atau None """ source_cols_map = {normalize(c): c for c in source_cols} # Step 1: Direct normalized match for name in possible_names: norm_name = normalize(name) if norm_name in source_cols_map: return source_cols_map[norm_name] # Step 2: Code token match (e.g. 'f1761', 'f5a1', 'f1101', 'f502', etc.) for name in possible_names: norm_name = normalize(name) # Extract question code if present (e.g., 'f1761', 'f5a1', 'f8', etc.) match = re.search(r'\b(f\d+[a-z]?\d*)\b', norm_name) if match: code = match.group(1) for orig_norm, orig_col in source_cols_map.items(): if re.search(r'\b' + re.escape(code) + r'\b', orig_norm): return orig_col return None def get_value(row, source_cols, possible_names, default=None): """ Ambil nilai dari row dengan cari kolom dulu. Args: row: pandas Series (satu row) source_cols: list kolom dari DataFrame possible_names: list kemungkinan nama kolom default: nilai default jika tidak ditemukan Returns: Nilai dari row, atau default """ col = find_column(source_cols, possible_names) if col and pd.notna(row.get(col)): val = row[col] # Convert int/float if isinstance(val, (int, float)): return int(val) if val == int(val) else val return str(val).strip() return default def check_available_columns(df_cols: list) -> dict: """ Check kolom apa saja yang tersedia di dataframe. Berguna untuk debugging. """ result = { 'total': len(df_cols), 'has_email': any('email' in str(c).lower() for c in df_cols), 'has_nim': any('nim' in str(c).lower() for c in df_cols), 'has_f1102': 'F1102' in df_cols, 'has_f5b': 'F5b' in df_cols, 'has_f1762': 'F1762' in df_cols, } return result