DB_JOB/작업파일/완료/광역_사이트맵/_검수체크.py

# -*- coding: utf-8 -*-
"""기관 구조(양식) 검수 — 읽기전용. 매뉴얼 1-8 체크리스트.
사용: python _검수체크.py <기관명>
출력: PASS (이상없음) 또는 FAIL + 이슈목록.
검수 항목(구조/양식만; 내용 정확도는 대상 아님):
 - 헤더병합 B1:R1·S1:W1·Y1:AA1
 - 시트명 {번호}_{기관명}
 - B순번 1..N 연속(gap/중복/비정수 없음)
 - D/E/F 병합이 그룹 경계 안 넘음(E⊆D, F⊆D+E)
 - K열 빈값/javascript/# 없음
 - 병합 범위 겹침 없음
 - 행높이 15(표본)
"""
import sys, glob, os, re
import openpyxl
from openpyxl.utils import range_boundaries

HEADER_MERGES = {'B1:R1', 'S1:W1', 'Y1:AA1'}


def find_file(name):
    base = os.path.dirname(os.path.abspath(__file__))
    cands = []
    for p in glob.glob(os.path.join(base, '*', '*', '*.xlsx')):
        if 'backup' in p or '_r2' in p or os.path.basename(p).startswith('_'):
            continue
        if os.path.basename(p).replace('.xlsx', '') == name:
            cands.append(p)
    if not cands:  # 부분일치 fallback
        for p in glob.glob(os.path.join(base, '*', '*', '*.xlsx')):
            if 'backup' in p or '_r2' in p:
                continue
            if name in os.path.basename(p):
                cands.append(p)
    return cands[0] if cands else None


def inspect(path):
    issues = []
    wb = openpyxl.load_workbook(path)
    ws = wb.active
    # 시트명
    if not re.match(r'^\d+_.+', ws.title or ''):
        issues.append(f'시트명 형식 이상: {ws.title!r} (기대 "번호_기관명")')
    # 헤더병합
    hm = {str(m) for m in ws.merged_cells.ranges}
    miss = [h for h in HEADER_MERGES if h not in hm]
    if miss:
        issues.append(f'헤더병합 누락: {miss}')
    # 데이터행 + flatten D/E/F
    data = [r for r in range(3, ws.max_row + 1) if ws.cell(r, 2).value not in (None, '')]
    fv = {r: {c: ws.cell(r, c).value for c in (4, 5, 6)} for r in data}
    for mr in ws.merged_cells.ranges:
        if mr.min_col in (4, 5, 6) and mr.min_row >= 3:
            top = ws.cell(mr.min_row, mr.min_col).value
            for rr in range(mr.min_row, mr.max_row + 1):
                if rr in fv and fv[rr].get(mr.min_col) in (None, ''):
                    fv[rr][mr.min_col] = top
    # B순번 연속
    bs = [ws.cell(r, 2).value for r in data]
    exp = list(range(1, len(bs) + 1))
    if bs != exp:
        nums = [b for b in bs if isinstance(b, (int, float))]
        s = set(nums)
        gaps = [i for i in range(1, int(max(nums)) + 1) if i not in s] if nums else []
        dups = sorted({x for x in nums if nums.count(x) > 1})
        nonint = [b for b in bs if not isinstance(b, (int, float))]
        issues.append(f'B순번 비연속: 행수={len(bs)} 최대={max(nums) if nums else None} 빠진={gaps[:15]} 중복={dups[:10]} 비정수={nonint[:5]}')
    # D/E/F 병합 경계
    for mr in ws.merged_cells.ranges:
        if mr.min_row < 3:
            continue
        rng = range(mr.min_row, mr.max_row + 1)
        if mr.min_col == 5:
            if len({fv.get(r, {}).get(4) for r in rng if r in fv}) > 1:
                issues.append(f'E병합 {mr}이 D경계 넘음')
        if mr.min_col == 6:
            if len({(fv.get(r, {}).get(4), fv.get(r, {}).get(5)) for r in rng if r in fv}) > 1:
                issues.append(f'F병합 {mr}이 D+E경계 넘음')
    # K열 빈/js/#
    badk = [r for r in data if (lambda k: k in (None, '') or (isinstance(k, str) and k.strip().startswith(('javascript:', '#'))))(ws.cell(r, 11).value)]
    if badk:
        issues.append(f'K열 빈값/js/#: {len(badk)}건 (행 {badk[:8]})')
    # 콘텐츠 행(순번 또는 URL 또는 메뉴 D~J 보유) — 연속행 포함
    content = [r for r in range(3, ws.max_row + 1)
               if ws.cell(r, 2).value not in (None, '') or ws.cell(r, 11).value not in (None, '')
               or any(ws.cell(r, c).value not in (None, '') for c in range(4, 11))]
    # B순번 빈칸(콘텐츠 행인데 B 비어있음) → 자동채움(모든 콘텐츠행 1..N 연속, 빈칸 없이)
    b_blank = [r for r in content if not isinstance(ws.cell(r, 2).value, int)]
    if b_blank:
        issues.append(f'[자동채움] B순번 빈칸(연속행 등): {len(b_blank)}건 (행 {b_blank[:12]})')
    # 사이트명(C) 빈칸 → 자동채움 대상(워크플로우에서 기관명 기입)
    c_empty = [r for r in content if ws.cell(r, 3).value in (None, '')]
    if c_empty:
        issues.append(f'[자동채움] 사이트명(C) 빈칸: {len(c_empty)}건 (행 {c_empty[:12]})')
    # 메뉴명/카테고리(D~J) 전무/중간구멍 → 보고(자동수정 X)
    # D/E/F=병합 상속, G~J=carry-forward(채워내림; G는 병합 안 되므로 첫행에만 값·형제행 None이 정상).
    mfill = {}
    for mr in ws.merged_cells.ranges:
        if mr.min_col in (4, 5, 6) and mr.min_row >= 3:
            top = ws.cell(mr.min_row, mr.min_col).value
            for rr in range(mr.min_row, mr.max_row + 1):
                mfill[(rr, mr.min_col)] = top
    def effdef(r, c):
        v = ws.cell(r, c).value
        return v if v not in (None, '') else mfill.get((r, c))
    nomenu, holes = [], []
    carry = {7: None, 8: None, 9: None, 10: None}
    prevdef = object()
    for r in content:
        eD, eE, eF = effdef(r, 4), effdef(r, 5), effdef(r, 6)
        if (eD, eE, eF) != prevdef:
            carry = {7: None, 8: None, 9: None, 10: None}; prevdef = (eD, eE, eF)
        res = [eD, eE, eF]
        for col in (7, 8, 9, 10):  # G~J carry-forward
            own = ws.cell(r, col).value
            if own not in (None, ''):
                carry[col] = own
                for dc in (8, 9, 10):
                    if dc > col:
                        carry[dc] = None
            res.append(carry[col])
        res = [v if v not in (None, '') else None for v in res]
        filled = [i for i, v in enumerate(res) if v is not None]
        if not filled:
            nomenu.append(r)
        elif any(res[i] is None for i in range(max(filled))):
            holes.append(r)
    if nomenu:
        issues.append(f'[보고] 메뉴명/카테고리 전무: {len(nomenu)}건 (행 {nomenu[:12]})')
    if holes:
        issues.append(f'[보고] 카테고리(대/중/소) 중간 빈칸: {len(holes)}건 (행 {holes[:12]})')
    # 병합 겹침
    mranges = [range_boundaries(str(m)) for m in ws.merged_cells.ranges]
    overlap = 0
    seen_cells = {}
    for (c1, r1, c2, r2) in mranges:
        for rr in range(r1, r2 + 1):
            for cc in range(c1, c2 + 1):
                if (rr, cc) in seen_cells:
                    overlap += 1
                seen_cells[(rr, cc)] = 1
    if overlap:
        issues.append(f'병합 겹침 셀 {overlap}개')
    # 행높이(표본 50)
    nonstd = [r for r in data[:50] if ws.row_dimensions[r].height not in (15, None)]
    if len(nonstd) > 5:
        issues.append(f'행높이≠15 다수(표본50중 {len(nonstd)})')
    return len(data), issues


def main():
    if len(sys.argv) < 2:
        print('사용: python _검수체크.py <기관명>'); return
    name = sys.argv[1]
    path = find_file(name)
    if not path:
        print(f'FAIL: 파일 못 찾음 ({name})'); return
    n, issues = inspect(path)
    rel = os.path.relpath(path, os.path.dirname(os.path.abspath(__file__)))
    print(f'기관: {name} | 파일: {rel} | 데이터 {n}행')
    if issues:
        print(f'FAIL — 구조 이상 {len(issues)}건:')
        for x in issues:
            print(f'   ✗ {x}')
    else:
        print('PASS — 구조/양식 이상 없음 (헤더병합·시트명·B순번연속·D/E/F병합경계·K열·병합겹침 정상)')


if __name__ == '__main__':
    main()