df16c98366
공공기관2 작업 중. _temp 몽타주(재생성가능)는 제외. Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
72 lines
2.9 KiB
Python
72 lines
2.9 KiB
Python
# -*- coding: utf-8 -*-
|
|
"""보은 재수집: www.boeun.go.kr(차단)→boeun.go.kr(apex) URL교정 + L빈칸 Phase2~4 채움.
|
|
사용: python -X utf8 _보은_refill.py [--write]
|
|
"""
|
|
import sys, os, shutil, importlib.util
|
|
from concurrent.futures import ThreadPoolExecutor, as_completed
|
|
import openpyxl
|
|
|
|
HERE=os.path.dirname(os.path.abspath(__file__))
|
|
sp=importlib.util.spec_from_file_location('cb',os.path.join(HERE,'_chungbuk_phase234_all.py'))
|
|
M=importlib.util.module_from_spec(sp); sp.loader.exec_module(M)
|
|
XLSX=r'D:\01.프로젝트\DB수집\작업파일\광역_사이트맵\충청북도\3.보은군\충청북도_보은군.xlsx'
|
|
BODY=['#contents','#txt','main']
|
|
sess=M.make_session()
|
|
|
|
def apexfix(u):
|
|
return u.replace('://www.boeun.go.kr','://boeun.go.kr')
|
|
|
|
def working(u):
|
|
s=M.fetch(sess,u)
|
|
if s is not None: return u,s
|
|
v=apexfix(u)
|
|
if v!=u:
|
|
s=M.fetch(sess,v)
|
|
if s is not None: return v,s
|
|
return None,None
|
|
|
|
def do_row(r,K):
|
|
wu,soup=working(K)
|
|
if soup is None: return r,None,K
|
|
return r,M.process_row(sess,wu,BODY),wu
|
|
|
|
def main():
|
|
write='--write' in sys.argv
|
|
wb=openpyxl.load_workbook(XLSX); ws=wb.active
|
|
# 1) www→apex URL 교정 (모든 행, 죽은링크 복구)
|
|
urlfix=0
|
|
for r in range(3,ws.max_row+1):
|
|
K=ws.cell(r,11).value
|
|
if isinstance(K,str) and '://www.boeun.go.kr' in K:
|
|
nk=apexfix(K); urlfix+=1
|
|
if write:
|
|
ws.cell(r,11).value=nk
|
|
if ws.cell(r,11).hyperlink: ws.cell(r,11).hyperlink.target=nk
|
|
# 2) L빈칸 http행 Phase2~4
|
|
jobs=[(r,ws.cell(r,11).value) for r in range(3,ws.max_row+1)
|
|
if ws.cell(r,12).value in (None,'') and isinstance(ws.cell(r,11).value,str) and ws.cell(r,11).value.startswith('http')]
|
|
res={}
|
|
with ThreadPoolExecutor(max_workers=6) as ex:
|
|
futs={ex.submit(do_row,r,K):r for r,K in jobs}
|
|
for f in as_completed(futs):
|
|
r,out,wu=f.result(); res[r]=(out,wu)
|
|
filled=fail=0
|
|
for r,(out,wu) in sorted(res.items()):
|
|
if out is None: fail+=1; continue
|
|
if write:
|
|
if wu!=ws.cell(r,11).value:
|
|
ws.cell(r,11).value=wu
|
|
if ws.cell(r,11).hyperlink: ws.cell(r,11).hyperlink.target=wu
|
|
ws.cell(r,12).value=out['L']; ws.cell(r,13).value=out['M']; ws.cell(r,14).value=out['N']
|
|
ws.cell(r,15).value=out['O']; ws.cell(r,16).value=out['P']; ws.cell(r,17).value=out['Q']
|
|
filled+=1
|
|
print(f'www→apex 교정 {urlfix}행 · L빈칸채움 {filled} · 실패 {fail} · {"적용" if write else "DRY"}')
|
|
from collections import Counter
|
|
print('채운 L분포:',dict(Counter(v[0]['L'] for v in res.values() if v[0])))
|
|
if write and (urlfix or filled):
|
|
bak=XLSX.replace('.xlsx','_backup_보은재수집전.xlsx')
|
|
if not os.path.exists(bak): shutil.copy(XLSX,bak)
|
|
wb.save(XLSX); print('저장+백업')
|
|
|
|
if __name__=='__main__': main()
|