AiDOPWarehouse/doc/_unify_doc_format.py

# -*- coding: utf-8 -*-
"""Unify S0-S4 delivery docx front matter to S0 blueprint style.

WARNING: Do not run blindly — rebuilds front matter and strips original formatting.
Prefer surgical edits via _restore_cover_format.py after manual structure changes.
"""
from __future__ import annotations

from pathlib import Path

from docx import Document
from docx.oxml import OxmlElement
from docx.shared import Pt
from docx.table import Table
from docx.text.paragraph import Paragraph

BASE = 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"
)
SYSTEM_NAME = "Ai-DOP智慧运营管理系统"
V01, V02 = "V0.1", "V0.2"
NOTE_V01, NOTE_V02 = "初版建立", "增加版本号"

DOC_META = {
    ("S0", "req"): {
        "title1": "S0 基础主数据",
        "title2": "业务需求描述",
        "doc_no": "S0-BRD-001",
        "author": "智造易项目组",
        "created": "2026-06-10",
        "updated": "2026-06-26",
    },
    ("S0", "blue"): {
        "title1": "S0 基础主数据建模",
        "title2": "业务蓝图设计方案",
        "doc_no": "S0-BBP-001",
        "author": "智造易项目组",
        "created": "2026-06-10",
        "updated": "2026-06-26",
    },
    ("S1", "req"): {
        "title1": "S1 产销协同模块",
        "title2": "业务需求描述",
        "doc_no": "S1-BRD-001",
        "author": "彭熙玉",
        "created": "2026-06-09",
        "updated": "2026-06-09",
    },
    ("S1", "blue"): {
        "title1": "S1 产销协同模块",
        "title2": "业务蓝图设计方案",
        "doc_no": "S1-BBP-001",
        "author": "彭熙玉",
        "created": "2026-06-09",
        "updated": "2026-06-09",
    },
    ("S2", "req"): {
        "title1": "S2 制造协同模块",
        "title2": "业务需求描述",
        "doc_no": "S2-BRD-001",
        "author": "彭熙玉",
        "created": "2026-06-09",
        "updated": "2026-06-09",
    },
    ("S2", "blue"): {
        "title1": "S2 制造协同模块",
        "title2": "业务蓝图设计方案",
        "doc_no": "S2-BBP-001",
        "author": "彭熙玉",
        "created": "2026-06-09",
        "updated": "2026-06-09",
    },
    ("S3", "req"): {
        "title1": "S3 供应协同模块",
        "title2": "业务需求描述",
        "doc_no": "S3-BRD-001",
        "author": "彭熙玉",
        "created": "2026-06-10",
        "updated": "2026-06-11",
    },
    ("S3", "blue"): {
        "title1": "S3 供应协同模块",
        "title2": "业务蓝图设计方案",
        "doc_no": "S3-BBP-001",
        "author": "彭熙玉",
        "created": "2026-06-10",
        "updated": "2026-06-10",
    },
    ("S4", "req"): {
        "title1": "S4 采购执行模块",
        "title2": "业务需求描述",
        "doc_no": "S4-BRD-001",
        "author": "彭熙玉",
        "created": "2026-06-11",
        "updated": "2026-06-11",
    },
    ("S4", "blue"): {
        "title1": "S4 采购执行模块",
        "title2": "业务蓝图设计方案",
        "doc_no": "S4-BBP-001",
        "author": "彭熙玉",
        "created": "2026-06-11",
        "updated": "2026-06-11",
    },
}

REMOVE_HEADINGS = {"审核", "发布", "版本记录"}
REMOVE_LINE_PREFIXES = (
    "文档作者：",
    "创建日期：",
    "更新日期：",
    "当前版本：",
    "文档编号：",
    "版本：",
)


def body_blocks(doc: Document):
    body = doc.element.body
    for child in body:
        if child.tag.endswith("p"):
            yield "p", Paragraph(child, doc)
        elif child.tag.endswith("tbl"):
            yield "t", Table(child, doc)


def block_text(kind: str, obj) -> str:
    if kind == "p":
        return obj.text.strip()
    return ""


def is_change_log_table(table: Table) -> bool:
    if not table.rows:
        return False
    header = "".join(c.text.strip() for c in table.rows[0].cells)
    return "变更说明" in header and ("版本" in header or "日期" in header)


def is_cover_meta_table(table: Table) -> bool:
    if not table.rows:
        return False
    return table.rows[0].cells[0].text.strip() == "文档编号"


def is_audit_or_release_table(table: Table) -> bool:
    if not table.rows:
        return False
    header = [c.text.strip() for c in table.rows[0].cells]
    return header[:3] == ["姓名", "职位", "签字/日期"] or header[:3] == ["编号", "名称", "地点"]


def is_version_record_table(table: Table) -> bool:
    if not table.rows:
        return False
    header = [c.text.strip() for c in table.rows[0].cells]
    return header[:4] == ["版本", "日期", "修订人", "修订说明"]


def set_cell(cell, text: str, *, bold: bool = False) -> None:
    cell.text = ""
    run = cell.paragraphs[0].add_run(text)
    run.font.size = Pt(10.5)
    run.bold = bold


def build_cover_table(doc: Document, meta: dict) -> Table:
    rows = [
        ("文档编号", meta["doc_no"]),
        ("版本", V02),
        ("密级", "内部"),
        ("编制单位", meta["author"]),
        ("编制日期", meta["updated"]),
    ]
    table = doc.add_table(rows=len(rows), cols=2)
    for ri, (label, value) in enumerate(rows):
        set_cell(table.rows[ri].cells[0], label, bold=True)
        set_cell(table.rows[ri].cells[1], value)
    return table


def build_change_table(doc: Document, meta: dict) -> Table:
    table = doc.add_table(rows=3, cols=4)
    headers = ["日期", "版本", "修订人", "变更说明"]
    data = [
        (meta["created"], V01, meta["author"], NOTE_V01),
        (meta["updated"], V02, meta["author"], NOTE_V02),
    ]
    for ci, val in enumerate(headers):
        set_cell(table.rows[0].cells[ci], val, bold=True)
    for ri, row in enumerate(data, start=1):
        for ci, val in enumerate(row):
            set_cell(table.rows[ri].cells[ci], val)
    return table


def set_paragraph_text(p: Paragraph, text: str, style: str | None = None) -> None:
    p.text = text
    if style:
        p.style = style


def insert_paragraph_before(ref_el, doc: Document) -> Paragraph:
    el = OxmlElement("w:p")
    ref_el.addprevious(el)
    return Paragraph(el, doc)


def move_element_before(element, ref_el) -> None:
    parent = element.getparent()
    if parent is not None:
        parent.remove(element)
    ref_el.addprevious(element)


def remove_orphan_tables_after_toc(doc: Document, toc_el) -> None:
    body = doc.element.body
    for tbl_el in list(body):
        if not tbl_el.tag.endswith("tbl"):
            continue
        if list(body).index(tbl_el) <= list(body).index(toc_el):
            continue
        table = Table(tbl_el, doc)
        if is_version_record_table(table):
            body.remove(tbl_el)


def unify_document(path: Path, module: str, kind: str) -> None:
    meta = DOC_META[(module, kind)]
    doc = Document(path)
    blocks = list(body_blocks(doc))

    toc_idx = next(
        (i for i, (k, o) in enumerate(blocks) if k == "p" and block_text(k, o) == "目录"),
        None,
    )
    if toc_idx is None:
        raise RuntimeError(f"未找到目录: {path.name}")

    toc_el = blocks[toc_idx][1]._element

    for kind_b, obj in blocks[:toc_idx]:
        if kind_b == "p":
            obj._element.getparent().remove(obj._element)
        elif kind_b == "t":
            obj._tbl.getparent().remove(obj._tbl)

    p1 = insert_paragraph_before(toc_el, doc)
    p2 = insert_paragraph_before(toc_el, doc)
    p3 = insert_paragraph_before(toc_el, doc)
    move_element_before(p1._element, toc_el)
    move_element_before(p2._element, toc_el)
    move_element_before(p3._element, toc_el)

    set_paragraph_text(p1, meta["title1"])
    set_paragraph_text(p2, meta["title2"])
    set_paragraph_text(p3, SYSTEM_NAME)

    insert_paragraph_before(toc_el, doc)
    cover = build_cover_table(doc, meta)
    move_element_before(cover._tbl, toc_el)
    insert_paragraph_before(toc_el, doc)

    h1 = insert_paragraph_before(toc_el, doc)
    set_paragraph_text(h1, "文档控制", "Heading 1")
    h2 = insert_paragraph_before(toc_el, doc)
    set_paragraph_text(h2, "更改记录", "Heading 2")
    insert_paragraph_before(toc_el, doc)
    change = build_change_table(doc, meta)
    move_element_before(change._tbl, toc_el)
    insert_paragraph_before(toc_el, doc)

    remove_orphan_tables_after_toc(doc, toc_el)
    doc.save(path)


def main() -> None:
    for module in ["S0", "S1", "S2", "S3", "S4"]:
        mod_dir = BASE / module
        for path in sorted(mod_dir.glob("*.docx")):
            kind = "blue" if "蓝图" in path.name else "req"
            unify_document(path, module, kind)
            print(f"OK {module}/{path.name}")


if __name__ == "__main__":
    main()