AiDOPWarehouse/doc/_gen_bid_final.py

# -*- coding: utf-8 -*-
"""把承诺书、软著一览表、团队表与技术／实施方案各章并入响应文件，成一册交付。

响应文件由 _gen_bid_response.py 生成，其中留有 @@MERGE:key@@ 占位段；本脚本用
Word 定位每个占位段，就地插入对应文件内容，然后统一页码、刷新目录并导出 PDF。

用法：
    python doc/_gen_bid_final.py            合并全部
    python doc/_gen_bid_final.py --list      只列出占位标记与对应文件，不改文档
"""
import glob
import io
import os
import shutil
import sys
import time

sys.path.insert(0, os.path.dirname(os.path.abspath(__file__)))

sys.stdout = io.TextIOWrapper(sys.stdout.buffer, encoding="utf-8")

HERE = os.path.dirname(os.path.abspath(__file__))
SYNC_DIR = r"C:\Users\skygu\OneDrive\Projects\AIDOP\项目\项目招投标"
MERGE_DIR = os.path.join(HERE, "_bid_merge_parts")

# Word 常量
WD_STORY = 6
WD_SECTION_BREAK_NEXT_PAGE = 2
WD_SECTION_BREAK_CONTINUOUS = 3
WD_FIND_STOP = 0
WD_COLLAPSE_START = 1
WD_FORMAT_DOCX = 16
WD_EXPORT_PDF = 17
WD_HEADER_FOOTER_PRIMARY = 1
WD_STAT_PAGES = 2

# 占位标记 → (待插入文件, 分节方式)
# 承诺书、一览表、团队表前面已有带段前分页的组织标题，用「连续」分节符可与标题
# 同页，不再多出一页只有标题的空页；各章无组织标题，用「下一页」分节符另起页。
PARTS = [
    ("invoice", "承诺书-能够提供增值税专用发票.docx", "cont"),
    ("consortium", "承诺书-不组成联合体及中选后不转包.docx", "cont"),
    ("credit", "承诺书-资格条件及信用状况.docx", "cont"),
    ("swtable", "软件著作权登记证书一览表.docx", "cont"),
    ("team", "项目团队岗位职责与人员配备表.docx", "cont"),
]
for _no in range(4, 13):
    _hit = glob.glob(os.path.join(MERGE_DIR, f"六3-{_no:02d} *.docx"))
    PARTS.append((f"ch{_no:02d}",
                  os.path.join("_bid_merge_parts", os.path.basename(_hit[0]))
                  if _hit else None, "next"))


def resolve():
    out = []
    for key, rel, brk in PARTS:
        if rel is None:
            raise SystemExit(f"缺少 {key} 对应文件（_bid_merge_parts 未生成？）")
        path = os.path.join(HERE, rel)
        if not os.path.exists(path):
            raise SystemExit(f"缺少待并入文件：{rel}")
        out.append((key, path, brk))
    return out


def response_doc():
    hits = glob.glob(os.path.join(HERE, "响应文件-*研发项目*.docx"))
    if not hits:
        raise SystemExit("未找到响应文件，请先运行 _gen_bid_response.py")
    return hits[0]


def insert_at_marker(word, doc, key, path, brk):
    """定位 @@MERGE:key@@ 所在段落，整段删除后在原位插入外部文件。

    必须先插分节符：被插入文件自带的节属性（页边距、横向页）会作用到它之前的
    内容，不隔开会把响应文件原有版式一起改掉。标记段整段删除，否则残留的空段
    落会各自占掉一页。
    """
    sel = word.Selection
    sel.HomeKey(WD_STORY)
    f = sel.Find
    f.ClearFormatting()
    f.Text = f"@@MERGE:{key}@@"
    f.Forward = True
    f.Wrap = WD_FIND_STOP
    f.MatchCase = True
    if not f.Execute():
        raise SystemExit(f"文档中找不到占位标记 {key}")
    sel.Paragraphs(1).Range.Select()
    sel.Delete()
    sel.InsertBreak(WD_SECTION_BREAK_CONTINUOUS if brk == "cont"
                    else WD_SECTION_BREAK_NEXT_PAGE)
    sel.InsertFile(path)
    return True


def unify_page_numbers(doc):
    """全册连续编号：各节页脚继承首节，且除首节外都不重启编号。"""
    for i in range(2, doc.Sections.Count + 1):
        sec = doc.Sections(i)
        ft = sec.Footers(WD_HEADER_FOOTER_PRIMARY)
        ft.LinkToPrevious = True
        ft.PageNumbers.RestartNumberingAtSection = False
    pn = doc.Sections(1).Footers(WD_HEADER_FOOTER_PRIMARY).PageNumbers
    pn.RestartNumberingAtSection = True
    pn.StartingNumber = 1


def main():
    parts = resolve()
    if "--list" in sys.argv:
        for key, path, brk in parts:
            print(f"  {key:<10} [{brk}] → {os.path.relpath(path, HERE)}")
        return

    src = response_doc()
    out = os.path.join(HERE, "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")
    pdf = os.path.join(HERE, "_bid_preview", os.path.basename(out)[:-5] + ".pdf")
    os.makedirs(os.path.dirname(pdf), exist_ok=True)
    for f in (out, pdf):
        if os.path.exists(f):
            os.remove(f)
    shutil.copy2(src, out)

    import win32com.client as win32
    word = None
    for attempt in range(4):
        try:
            word = win32.gencache.EnsureDispatch("Word.Application")
            word.Visible = False
            word.DisplayAlerts = 0
            break
        except Exception as e:                       # Word COM 偶发拒绝调用
            print(f"  启动 Word 重试 {attempt + 1}/4：{e}")
            time.sleep(4)
    if word is None:
        raise SystemExit("无法启动 Word")

    try:
        doc = word.Documents.Open(out)
        doc.Activate()
        for key, path, brk in parts:
            insert_at_marker(word, doc, key, path, brk)
            print(f"  已并入 {key:<10} {os.path.basename(path)}")

        unify_page_numbers(doc)
        for _ in range(3):        # 目录页数变化会反过来影响页码，多刷几遍
            if doc.TablesOfContents.Count:
                doc.TablesOfContents(1).Update()
            doc.Fields.Update()
        doc.Repaginate()

        pages = doc.ComputeStatistics(WD_STAT_PAGES)
        toc_items = (doc.TablesOfContents(1).Range.Paragraphs.Count
                     if doc.TablesOfContents.Count else 0)
        secs = doc.Sections.Count
        doc.SaveAs2(out, FileFormat=WD_FORMAT_DOCX)
        doc.ExportAsFixedFormat(pdf, WD_EXPORT_PDF)
        doc.Close(False)
    finally:
        word.Quit()

    size = os.path.getsize(out) / 1024 / 1024
    print(f"\n[完成] {os.path.basename(out)}")
    print(f"       {pages} 页　{secs} 节　目录 {toc_items} 条　{size:.1f} MB")
    if os.path.isdir(SYNC_DIR):
        dst = os.path.join(SYNC_DIR, os.path.basename(out))
        # 交付目录里的总册会被手工补材料（业绩扫描件、查询截图等）。若那份比本次
        # 生成的更新，直接覆盖会丢掉手工内容，故先备份再同步。
        if os.path.exists(dst) and os.path.getmtime(dst) > os.path.getmtime(src):
            bak = dst[:-5] + time.strftime("_手工修改备份_%m%d%H%M") + ".docx"
            shutil.copy2(dst, bak)
            print(f"[注意] 交付目录那份比生成源更新，疑似手工改过，已备份为\n"
                  f"       {os.path.basename(bak)}")
        shutil.copy2(out, dst)
        print(f"[已同步] {SYNC_DIR}")


if __name__ == "__main__":
    main()