# -*- coding: utf-8 -*- import openpyxl, sys, io, glob, os sys.stdout=io.TextIOWrapper(sys.stdout.buffer,encoding='utf-8') def urls(F): wb=openpyxl.load_workbook(F); ws=wb.active s=set() for r in range(3,ws.max_row+1): u=ws.cell(r,11).value if u and str(u).startswith('http'): uu=str(u).split('#')[0].rstrip('/') s.add(uu) return s PAIRS={ '옥천':'5.옥천군\\충청북도_옥천군', '음성':'6.음성군\\충청북도_음성군', '제천':'7.제천시\\충청북도_제천시', '청주':'10.청주시\\충청북도_청주시', } for nm,base in PAIRS.items(): new=base+'.xlsx'; bak=base+'_backup_재구축전.xlsx' if not os.path.exists(bak): print(nm,'백업없음'); continue un=urls(new); ub=urls(bak) lost=ub-un; gain=un-ub print('%s | 신규행url %d · 원본url %d | 원본에만(손실) %d · 신규에만(추가) %d'%(nm,len(un),len(ub),len(lost),len(gain))) for u in list(lost)[:6]: print(' 손실:',u[:75])