AiDOPWarehouse/doc/_tmp_scan_int.py

# -*- coding: utf-8 -*-
from pathlib import Path
from docx import Document

d = Path(r"C:\Users\skygu\OneDrive\Projects\AIDOP\项目\项目管理\产互联项目管理\交付文档\系统集成")
OPENING = "本模块对接的第三方系统涉及 ERP、MES、WMS"
for p in sorted(d.glob("*V0.4.docx")):
    if p.name.startswith("~$"):
        continue
    doc = Document(str(p))
    hits = []
    for para in doc.paragraphs:
        t = para.text.strip()
        if any(x in t for x in ("ERP", "MES", "WMS", "CRM", "SRM", "TMS")):
            hits.append(t)
    for table in doc.tables:
        for row in table.rows:
            for cell in row.cells:
                t = cell.text.strip()
                if any(x in t for x in ("ERP", "MES", "WMS", "CRM", "SRM", "TMS")):
                    hits.append(t)
    leftover = [h for h in hits if OPENING not in h]
    print("====", p.name, "named", len(hits), "leftover", len(leftover))
    for h in leftover:
        print(" ", h[:160])