AiDOPWarehouse/doc/_gen_seal_list.py

# -*- coding: utf-8 -*-
"""校验生成文件未改动原文，并通过 Word COM 取签章位真实页码，生成签章位置清单。"""

import json
import os
import re
import sys
import time

sys.stdout.reconfigure(encoding="utf-8", errors="replace")

from docx import Document
from docx.oxml.ns import qn

DOC_DIR = os.path.dirname(os.path.abspath(__file__))
SRC = os.path.join(DOC_DIR, "_tmp_xunbi.docx")
OUT_NAME = "响应文件-2026年河南产互联制造业数据智能运营平台研发项目-北京智造易科技有限公司.docx"
OUT = os.path.join(DOC_DIR, OUT_NAME)
# 页码要按实际装订的总册算：承诺书、软著一览表、团队表与技术方案各章已并入其中。
# 总册在 OneDrive 里会被手工补材料（业绩扫描件、查询截图等），页码随之变化，
# 因此优先取两处中较新的那一份，并在输出里注明用的是哪份。
FINAL_NAME = "响应文件（完整版）-2026年河南产互联制造业数据智能运营平台研发项目.docx"
LIST_NAME = "响应文件签章位置清单.docx"
LIST_PATH = os.path.join(DOC_DIR, LIST_NAME)
ONEDRIVE_DIR = r"C:\Users\skygu\OneDrive\Projects\AIDOP\项目\项目招投标"


def pick_final():
    """在本地生成版与 OneDrive 手工修改版之间取较新的一份；可用 argv 直接指定。"""
    if len(sys.argv) > 1 and os.path.exists(sys.argv[1]):
        return sys.argv[1], "命令行指定"
    cands = [(os.path.join(DOC_DIR, FINAL_NAME), "本地生成版"),
             (os.path.join(ONEDRIVE_DIR, FINAL_NAME), "OneDrive 修改版")]
    cands = [(p, w) for p, w in cands if os.path.exists(p)]
    if not cands:
        return None, None
    return max(cands, key=lambda x: os.path.getmtime(x[0]))


FINAL, FINAL_FROM = pick_final()

# ---------------------------------------------------------------- 1) 原文校验
src = Document(SRC)


def ptext(el):
    return "".join(n.text or "" for n in el.iter(qn("w:t")))


ch = list(src.element.body.iterchildren())
c6 = [i for i, e in enumerate(ch)
      if e.tag == qn("w:p") and ptext(e).strip().startswith("第六章")
      and "响应文件格式" in ptext(e)]
c7 = [i for i, e in enumerate(ch)
      if e.tag == qn("w:p") and ptext(e).strip().startswith("第七章")]
s = c6[-1]
t = next(i for i in c7 if i > s)

orig_paras = [ptext(e) for e in ch[s:t] if e.tag == qn("w:p")]
gen = Document(OUT)
gen_paras = [p.text for p in gen.paragraphs]

# 原文中作为「填空提示」的固定文字，填写后必然改变，属预期
FILLED_KEYS = [
    "项目编号：", "供应商：", "日  期：", "致：", "1、根据已收到的项目编号为",
    "4、一旦我公司成交", "单位名称：", "单位性质：", "地      址：", "成立时间：",
    "经营期限：", "姓    名：", "系", "法定代表人：", "本人", "供  应  商：",
    "法定代表人身份证号码：", "委托代理人：", "委托代理人身份证号码：",
    "供应商名称（盖章）：", "日期：", "承诺单位（签章）：", "年  月  日",
    "对第四章", "对第五章", "供应商名称：", "（采购单位）", "为营造公开",
    "八、本人作为", "项目", "        年     月     日",
]


# 原文第 768 段的手工目录条目（无页码），已整体换成自动目录域，属预期删除
MANUAL_TOC = {"响应函及响应函附录", "法定代表人身份证明书", "法定代表人授权委托书",
              "资格审查资料", "商务响应文件", "技术响应文件", "其他"}
# 原文漏排编号、生成时补齐前缀的段落，比对时按「原文是生成文的后缀」放行
PREFIXED = {"包括但不限于以下资料（根据评分办法要求格式自拟）："}
# 原文编号重复、生成时改号的段落（原文 → 生成文）
RENUMBERED = {"（一）技术要求的汇总应答": "（二）技术要求的汇总应答"}


def is_fill_para(txt):
    return any(k and txt.strip().startswith(k) for k in FILLED_KEYS)


def norm(txt):
    """比对时忽略空白差异：制表符与连续空格在原文/生成文件中写法不一，
    仅因此报「原文缺失」属误报（如封面署名行原文用空格、模板里是制表符）。"""
    return re.sub(r"\s+", "", txt)


# 已按填写要求填入日期的段落，原文的空日期行必然消失，属预期
DATE_LINE = re.compile(r"^年+\s*月+\s*日$")

missing = []
gen_set = {norm(x) for x in gen_paras}
for op in orig_paras:
    o = op.strip()
    if not o or is_fill_para(o) or DATE_LINE.match(norm(o)) or o in MANUAL_TOC:
        continue
    if o in PREFIXED and any(x.endswith(norm(o)) for x in gen_set):
        continue
    if o in RENUMBERED and norm(RENUMBERED[o]) in gen_set:
        continue
    if norm(o) not in gen_set:
        missing.append(o)

print("=" * 66)
print("【原文完整性校验】")
print(f"  第六章原文非空段落：{len([x for x in orig_paras if x.strip()])}")
print(f"  生成文件段落总数：{len(gen_paras)}")
if missing:
    print(f"  ！以下 {len(missing)} 条原文未在生成文件中找到：")
    for m in missing[:20]:
        print(f"     - {m[:70]}")
else:
    print("  OK：所有未涉及填写的原文段落均原样保留")

# 检查是否混入了非原文的新增条款（响应函应仍为 6 条）
# 按「（一）响应函」到「（二）响应函附录」之间的编号段落计数；
# 早先用「含我公司」筛选会漏掉以「我们」「贵方」开头的第 5、6 条，造成误报。
def _section(paras, start_key, end_key):
    try:
        a = next(i for i, x in enumerate(paras) if start_key in x)
        b = next(i for i, x in enumerate(paras) if i > a and end_key in x)
    except StopIteration:
        return paras
    return paras[a:b]


resp_items = [x for x in _section(gen_paras, "（一）响应函", "（二）响应函附录")
              if re.match(r"^[1-9]、", x.strip())]
print(f"  响应函条目数：{len(resp_items)}（原文为 6 条）")
for x in resp_items:
    print(f"     {x.strip()[:40]}")

# ---------------------------------------------------------------- 2) 取页码
with open(os.path.join(DOC_DIR, "_tmp_seals.json"), encoding="utf-8") as f:
    seals = json.load(f)

print(f"\n【页码依据】{FINAL_FROM}：{os.path.basename(FINAL or OUT)}")
if FINAL:
    print("           修改时间 " + time.strftime(
        "%Y-%m-%d %H:%M", time.localtime(os.path.getmtime(FINAL))))

pages = {}
try:
    import win32com.client as win32

    word = win32.DispatchEx("Word.Application")
    word.Visible = False
    word.DisplayAlerts = 0
    d = word.Documents.Open(FINAL or OUT, ReadOnly=True)
    d.Repaginate()
    total = d.ComputeStatistics(2)  # wdStatisticPages
    for key, *_ in seals:
        try:
            pages[key] = d.Bookmarks(key).Range.Information(3)
        except Exception as e:
            pages[key] = f"?({e.__class__.__name__})"
    d.Close(False)
    word.Quit()
    print(f"\n【页码提取】文档总页数：{total}")
except Exception as e:
    total = "?"
    print(f"\n[Word COM 不可用] {type(e).__name__}: {e}")

seals.sort(key=lambda s: (pages.get(s[0], 999) if isinstance(pages.get(s[0]), int) else 999,
                          s[0]))
for key, section, *_ in seals:
    print(f"  {key}  第 {pages.get(key, '?')} 页  {section}")

# ---------------------------------------------------------------- 3) 生成清单
from docx.enum.table import WD_TABLE_ALIGNMENT
from docx.enum.text import WD_ALIGN_PARAGRAPH
from docx.shared import Cm, Pt

lst = Document()
sec = lst.sections[0]
sec.page_width, sec.page_height = Cm(29.7), Cm(21.0)  # A4 横向
sec.orientation = 1
sec.left_margin = sec.right_margin = Cm(1.8)
sec.top_margin = sec.bottom_margin = Cm(1.8)

st = lst.styles["Normal"]
st.font.name = "Times New Roman"
st.font.size = Pt(10.5)
st.element.rPr.rFonts.set(qn("w:eastAsia"), "宋体")

h = lst.add_paragraph()
h.alignment = WD_ALIGN_PARAGRAPH.CENTER
r = h.add_run("响应文件 签章位置清单")
r.font.size = Pt(18)
r.font.bold = True
r._element.get_or_add_rPr().find(qn("w:rFonts")).set(qn("w:eastAsia"), "黑体") \
    if r._element.get_or_add_rPr().find(qn("w:rFonts")) is not None else None

sub = lst.add_paragraph()
sub.alignment = WD_ALIGN_PARAGRAPH.CENTER
rs = sub.add_run(f"{FINAL_NAME if FINAL else OUT_NAME}"
                 f"　｜　共 {total} 页　｜　"
                 f"下列 {len(seals)} 处均为询比文件明文要求，非明文要求者未列入")
rs.font.size = Pt(9)
lst.add_paragraph()

tb = lst.add_table(rows=1, cols=6)
tb.style = "Table Grid"
tb.alignment = WD_TABLE_ALIGNMENT.CENTER
hdrs = ["序号", "页码", "所在章节 / 位置", "盖章要求", "签字要求", "依据出处（询比文件原文）"]
widths = [1.1, 1.8, 5.4, 3.9, 5.4, 8.5]
from docx.oxml import OxmlElement

tb.autofit = False
_lay = OxmlElement("w:tblLayout")
_lay.set(qn("w:type"), "fixed")
tb._tbl.tblPr.append(_lay)
for i, htxt in enumerate(hdrs):
    c = tb.rows[0].cells[i]
    c.text = ""
    rr = c.paragraphs[0].add_run(htxt)
    rr.font.bold = True
    rr.font.size = Pt(10)

for n, (key, section, stamp, sign, basis) in enumerate(seals, 1):
    cells = tb.add_row().cells
    for i, v in enumerate([str(n), f"第 {pages.get(key, '?')} 页", section,
                           stamp, sign, basis]):
        cells[i].text = ""
        rr = cells[i].paragraphs[0].add_run(v)
        rr.font.size = Pt(9.5)
    cells[0].paragraphs[0].alignment = WD_ALIGN_PARAGRAPH.CENTER
    cells[1].paragraphs[0].alignment = WD_ALIGN_PARAGRAPH.CENTER

for row in tb.rows:
    for i, w in enumerate(widths):
        row.cells[i].width = Cm(w)

lst.add_paragraph()
p2 = lst.add_paragraph()
r2 = p2.add_run("另：明文要求「加盖单位公章」的复印件（共 2 处）")
r2.font.bold = True
r2.font.size = Pt(11)
for txt in [
    "1. 法定代表人身份证复印件　——　第六章「二、法定代表人身份证明书」末尾原文："
    "附：法定代表人身份证复印件，加盖单位公章。",
    "2. 委托代理人身份证复印件　——　第六章「三、法定代表人授权委托书」末尾原文："
    "附：委托代理人身份证复印件，加盖单位公章。",
]:
    pp = lst.add_paragraph()
    pp.paragraph_format.left_indent = Pt(18)
    pp.add_run(txt).font.size = Pt(10)

lst.save(LIST_PATH)
print(f"\n[完成] {LIST_PATH}")

import shutil

if os.path.isdir(ONEDRIVE_DIR):
    shutil.copy2(LIST_PATH, os.path.join(ONEDRIVE_DIR, LIST_NAME))
    print(f"[已同步] {os.path.join(ONEDRIVE_DIR, LIST_NAME)}")