proposa-ai/bid_proposal/proposal_writer.py

"""
proposal_writer.py - 投标文件DOCX输出模块

将 ProposalDocument 格式化为结构化的 Word 文档（.docx），
支持封面、目录、标题层级、表格、列表、页脚页码等排版要素。
"""

import logging
import re
from typing import Optional

from docx import Document
from docx.enum.table import WD_TABLE_ALIGNMENT
from docx.enum.text import WD_ALIGN_PARAGRAPH, WD_LINE_SPACING
from docx.oxml.ns import qn, nsdecls
from docx.oxml import parse_xml
from docx.shared import Cm, Pt, RGBColor
from docx.table import _Cell

from .models import ProposalDocument, ProposalSection

logger = logging.getLogger(__name__)

# ---- 常量 ----
PAGE_WIDTH_CM = 21.0
PAGE_HEIGHT_CM = 29.7
MARGIN_CM = 2.0
BODY_WIDTH_CM = PAGE_WIDTH_CM - 2 * MARGIN_CM  # 17.0 cm

FONT_NAME = "宋体"
FONT_NAME_ASCII = "Times New Roman"

HEADING_FONTS = {1: 16, 2: 14, 3: 12}
BODY_FONT_SIZE = 12  # pt


def _set_cell_border(cell: _Cell, **kwargs):
    """设置单元格边框。"""
    tc = cell._tc
    tcPr = tc.get_or_add_tcPr()
    tcBorders = parse_xml(
        f'<w:tcBorders {nsdecls("w")}>'
        '  <w:top w:val="single" w:sz="4" w:space="0" w:color="000000"/>'
        '  <w:left w:val="single" w:sz="4" w:space="0" w:color="000000"/>'
        '  <w:bottom w:val="single" w:sz="4" w:space="0" w:color="000000"/>'
        '  <w:right w:val="single" w:sz="4" w:space="0" w:color="000000"/>'
        '</w:tcBorders>'
    )
    tcPr.append(tcBorders)


def _set_paragraph_spacing(paragraph, line_spacing: float = 1.5, space_before: float = 0.5, space_after: float = 0.5):
    """设置段落间距。"""
    pf = paragraph.paragraph_format
    pf.line_spacing_rule = WD_LINE_SPACING.MULTIPLE
    pf.line_spacing = line_spacing
    pf.space_before = Pt(space_before * BODY_FONT_SIZE)
    pf.space_after = Pt(space_after * BODY_FONT_SIZE)


def _run_format(run, font_name: str = FONT_NAME, font_size: int = BODY_FONT_SIZE,
                bold: bool = False, color: Optional[str] = None):
    """设置 run 的字体格式。"""
    run.font.name = font_name
    run.font.size = Pt(font_size)
    run.font.bold = bold
    run._element.rPr.rFonts.set(qn("w:eastAsia"), font_name)
    if color:
        run.font.color.rgb = RGBColor(*bytes.fromhex(color.lstrip("#")))


def _setup_page(doc: Document):
    """设置页面格式：A4、2cm页边距。"""
    section = doc.sections[0]
    section.page_width = Cm(PAGE_WIDTH_CM)
    section.page_height = Cm(PAGE_HEIGHT_CM)
    section.top_margin = Cm(MARGIN_CM)
    section.bottom_margin = Cm(MARGIN_CM)
    section.left_margin = Cm(MARGIN_CM)
    section.right_margin = Cm(MARGIN_CM)


def _add_footer_page_number(doc: Document):
    """添加页脚页码。"""
    section = doc.sections[0]
    footer = section.footer
    footer.is_linked_to_previous = False
    p = footer.paragraphs[0]
    p.alignment = WD_ALIGN_PARAGRAPH.CENTER
    # "第 X 页 / 共 Y 页"
    run1 = p.add_run("第 ")
    _run_format(run1, font_size=10)

    fld_char1 = parse_xml(f'<w:fldChar {nsdecls("w")} w:fldCharType="begin"/>')
    run2 = p.add_run()
    run2._element.append(fld_char1)

    instr = parse_xml(f'<w:instrText {nsdecls("w")} xml:space="preserve"> PAGE </w:instrText>')
    run3 = p.add_run()
    run3._element.append(instr)

    fld_char2 = parse_xml(f'<w:fldChar {nsdecls("w")} w:fldCharType="end"/>')
    run4 = p.add_run()
    run4._element.append(fld_char2)

    run5 = p.add_run(" 页 / 共 ")
    _run_format(run5, font_size=10)

    fld_char3 = parse_xml(f'<w:fldChar {nsdecls("w")} w:fldCharType="begin"/>')
    run6 = p.add_run()
    run6._element.append(fld_char3)

    instr2 = parse_xml(f'<w:instrText {nsdecls("w")} xml:space="preserve"> NUMPAGES </w:instrText>')
    run7 = p.add_run()
    run7._element.append(instr2)

    fld_char4 = parse_xml(f'<w:fldChar {nsdecls("w")} w:fldCharType="end"/>')
    run8 = p.add_run()
    run8._element.append(fld_char4)

    run9 = p.add_run(" 页")
    _run_format(run9, font_size=10)


def _create_cover_page(doc: Document, title: str, date_str: str = ""):
    """创建封面页。

    封面排版：居中，从上到下依次为：
    - "投标文件"（大号标题）
    - 项目名称
    - 编制日期
    """
    import datetime
    if not date_str:
        date_str = datetime.date.today().strftime("%Y年%m月%d日")

    # 上部留白
    for _ in range(6):
        doc.add_paragraph()

    # 大标题：投标文件
    p_title = doc.add_paragraph()
    p_title.alignment = WD_ALIGN_PARAGRAPH.CENTER
    run = p_title.add_run("投 标 文 件")
    _run_format(run, font_name=FONT_NAME, font_size=28, bold=True)

    # 装饰线
    p_line = doc.add_paragraph()
    p_line.alignment = WD_ALIGN_PARAGRAPH.CENTER
    run_line = p_line.add_run("━" * 20)
    _run_format(run_line, font_size=14, color="333333")

    doc.add_paragraph()

    # 项目名称
    p_proj = doc.add_paragraph()
    p_proj.alignment = WD_ALIGN_PARAGRAPH.CENTER
    run_proj = p_proj.add_run(f"项目名称：{title}")
    _run_format(run_proj, font_size=16)

    doc.add_paragraph()
    doc.add_paragraph()

    # 编制日期
    p_date = doc.add_paragraph()
    p_date.alignment = WD_ALIGN_PARAGRAPH.CENTER
    run_date = p_date.add_run(f"编制日期：{date_str}")
    _run_format(run_date, font_size=14)

    # 分页
    doc.add_page_break()


def _add_heading(doc: Document, text: str, level: int = 1):
    """添加格式化标题。

    使用预定义的字体大小：Heading1=16pt, Heading2=14pt, Heading3=12pt。
    """
    heading = doc.add_heading(text, level=min(level, 3))
    font_size = HEADING_FONTS.get(min(level, 3), 12)
    for run in heading.runs:
        run.font.name = FONT_NAME
        run.font.size = Pt(font_size)
        run._element.rPr.rFonts.set(qn("w:eastAsia"), FONT_NAME)
    return heading


def _add_body_paragraph(doc: Document, text: str):
    """添加格式化正文段落。"""
    p = doc.add_paragraph()
    p.alignment = WD_ALIGN_PARAGRAPH.JUSTIFY
    _set_paragraph_spacing(p)

    # 处理内联格式：**粗体**
    parts = re.split(r'(\*\*.+?\*\*)', text)
    for part in parts:
        if part.startswith("**") and part.endswith("**"):
            run = p.add_run(part[2:-2])
            _run_format(run, bold=True)
        else:
            run = p.add_run(part)
            _run_format(run)

    p.style = doc.styles["Normal"]
    return p


def _add_bullet_list(doc: Document, items: list[str]):
    """添加无序列表。"""
    for item in items:
        p = doc.add_paragraph(style="List Bullet")
        _set_paragraph_spacing(p, line_spacing=1.3, space_before=0, space_after=0)
        p.alignment = WD_ALIGN_PARAGRAPH.JUSTIFY
        run = p.add_run(item.strip())
        _run_format(run)


def _add_numbered_list(doc: Document, items: list[str]):
    """添加有序列表。"""
    for item in items:
        p = doc.add_paragraph(style="List Number")
        _set_paragraph_spacing(p, line_spacing=1.3, space_before=0, space_after=0)
        p.alignment = WD_ALIGN_PARAGRAPH.JUSTIFY
        run = p.add_run(item.strip())
        _run_format(run)


def _parse_markdown_table(md_table: str) -> list[list[str]]:
    """解析Markdown表格为二维列表。

    Args:
        md_table: Markdown表格文本

    Returns:
        list[list[str]]: 二维字符串列表，第一行为表头
    """
    lines = [line.strip() for line in md_table.strip().split("\n")]
    lines = [line for line in lines if line and not line.startswith("| ---") and not line.startswith("|---")]

    if not lines:
        return []

    table_data = []
    for line in lines:
        line = line.strip()
        if line.startswith("|"):
            line = line[1:]
        if line.endswith("|"):
            line = line[:-1]
        cells = [cell.strip() for cell in line.split("|")]
        table_data.append(cells)

    return table_data


def _write_table(doc: Document, table_lines: list[str]):
    """将Markdown表格写入DOCX表格。"""
    table_text = "\n".join(table_lines)
    data = _parse_markdown_table(table_text)

    if not data or len(data) < 2:
        logger.warning("表格数据不足，跳过表格写入")
        return

    rows = len(data)
    cols = max(len(row) for row in data)

    doc_table = doc.add_table(rows=rows, cols=cols)
    doc_table.autofit = True
    doc_table.alignment = WD_TABLE_ALIGNMENT.CENTER

    # 设置表格边框
    tbl = doc_table._tbl
    tblPr = tbl.tblPr if tbl.tblPr is not None else parse_xml(
        f'<w:tblPr {nsdecls("w")}/>'
    )
    borders = parse_xml(
        f'<w:tblBorders {nsdecls("w")}>'
        '  <w:top w:val="single" w:sz="4" w:space="0" w:color="000000"/>'
        '  <w:left w:val="single" w:sz="4" w:space="0" w:color="000000"/>'
        '  <w:bottom w:val="single" w:sz="4" w:space="0" w:color="000000"/>'
        '  <w:right w:val="single" w:sz="4" w:space="0" w:color="000000"/>'
        '  <w:insideH w:val="single" w:sz="4" w:space="0" w:color="000000"/>'
        '  <w:insideV w:val="single" w:sz="4" w:space="0" w:color="000000"/>'
        '</w:tblBorders>'
    )
    tblPr.append(borders)

    for r_idx, row_data in enumerate(data):
        for c_idx in range(cols):
            cell_text = row_data[c_idx] if c_idx < len(row_data) else ""
            cell = doc_table.cell(r_idx, c_idx)

            # 清空并重新设置内容
            cell.text = ""
            p = cell.paragraphs[0]
            p.alignment = WD_ALIGN_PARAGRAPH.CENTER
            run = p.add_run(cell_text)
            is_header = r_idx == 0
            _run_format(run, font_size=10, bold=is_header)

    doc.add_paragraph()  # 表后空行


def _write_markdown_content(doc, markdown_text: str, base_level: int = 2):
    """将Markdown格式的内容写入DOCX。

    支持的Markdown元素：
    - # 标题 → Heading样式
    - **粗体** → 粗体
    - 普通段落 → Normal样式
    - - 列表项 → 无序列表
    - 1. 列表项 → 有序列表
    - | 表格 | → DOCX表格
    - 空行 → 段落分隔

    Args:
        doc: Document对象
        markdown_text: Markdown格式的文本
        base_level: 基础标题层级偏移（默认2，即# -> Heading 2）
    """
    lines = markdown_text.split("\n")

    i = 0
    in_table = False
    table_lines = []
    in_code_block = False
    code_lines = []

    while i < len(lines):
        line = lines[i]
        stripped = line.strip()

        # 代码块处理
        if stripped.startswith("```"):
            if in_code_block:
                # 结束代码块，作为普通段落写入
                for code_line in code_lines:
                    _add_body_paragraph(doc, code_line)
                code_lines = []
                in_code_block = False
            else:
                in_code_block = True
            i += 1
            continue
        if in_code_block:
            code_lines.append(line)
            i += 1
            continue

        # 表格处理：连续以 | 开头的行
        # 判断是否为 Markdown 表格分隔行（如 |---|---| 或 |:---|:---:|）
        _is_separator = bool(re.match(r'^\|[\s\-:]+\|?$', stripped))

        if stripped.startswith("|") and not _is_separator:
            in_table = True
            table_lines.append(line)
            i += 1
            continue
        elif _is_separator:
            # 分隔行，跳过
            i += 1
            continue
        else:
            if in_table:
                # 表格结束，写入表格
                _write_table(doc, table_lines)
                table_lines = []
                in_table = False

        # 空行
        if not stripped:
            i += 1
            continue

        # 标题
        heading_match = re.match(r'^(#{1,6})\s+(.+)$', stripped)
        if heading_match:
            level = min(len(heading_match.group(1)) + base_level - 1, 3)
            # level最小为1
            level = max(level, 1)
            title_text = heading_match.group(2)

            # 去掉可能的内联格式标记（标题中保留文字）
            title_text = re.sub(r'\*\*(.+?)\*\*', r'\1', title_text)

            _add_heading(doc, title_text, level)
            i += 1
            continue

        # 无序列表
        if stripped.startswith("- ") or stripped.startswith("* "):
            items = []
            while i < len(lines):
                s = lines[i].strip()
                if s.startswith("- ") or s.startswith("* "):
                    items.append(s[2:])
                    i += 1
                else:
                    break
            _add_bullet_list(doc, items)
            continue

        # 有序列表
        ordered_match = re.match(r'^\d+\.\s+(.+)$', stripped)
        if ordered_match:
            items = []
            while i < len(lines):
                s = lines[i].strip()
                if re.match(r'^\d+\.\s+.+$', s):
                    items.append(re.match(r'^\d+\.\s+(.+)$', s).group(1))
                    i += 1
                else:
                    break
            _add_numbered_list(doc, items)
            continue

        # 普通段落
        paragraph_text = stripped
        i += 1
        # 合并后续非空、非特殊行
        while i < len(lines):
            next_line = lines[i]
            next_stripped = next_line.strip()
            if not next_stripped:
                break
            if next_stripped.startswith("#"):
                break
            if next_stripped.startswith("- ") or next_stripped.startswith("* "):
                break
            if re.match(r'^\d+\.\s+', next_stripped):
                break
            if next_stripped.startswith("|") and not next_stripped.startswith("| ---"):
                break
            paragraph_text += " " + next_stripped
            i += 1

        _add_body_paragraph(doc, paragraph_text)

    # 如果文档以表格结尾
    if in_table and table_lines:
        _write_table(doc, table_lines)


def write_proposal_docx(content: ProposalDocument, output_path: str) -> str:
    """将投标内容写入格式化的DOCX文件。

    文档结构：
    1. 封面（"投标文件"大标题 + 项目名称 + 日期）
    2. 目录
    3. 投标函（从 sections 中识别）
    4. 投标概要（content.summary）
    5. 正文章节（按层级组织）
    6. 页脚页码

    Args:
        content: 投标文档内容（ProposalDocument）
        output_path: 输出文件路径（应以 .docx 结尾）

    Returns:
        str: 输出文件路径

    Raises:
        ValueError: 如果 content 为空或 output_path 不合法
        IOError: 如果文件写入失败
    """
    if not content:
        raise ValueError("投标内容不能为空")
    if not output_path:
        raise ValueError("输出路径不能为空")

    logger.info("开始生成DOCX文档：%s", output_path)

    doc = Document()
    _setup_page(doc)

    # 从 sections 中分离出投标函和内容章节
    cover_letter_section = None
    content_sections: list[ProposalSection] = []
    for s in content.sections:
        if s.title == "投标函" and cover_letter_section is None:
            cover_letter_section = s
        else:
            content_sections.append(s)

    # 1. 封面
    _create_cover_page(doc, content.title or content.project_name or "投标文件")

    # 2. 目录
    _add_heading(doc, "目 录", 1)
    toc_num = 1
    if cover_letter_section:
        p = doc.add_paragraph(f"{_to_chinese_number(toc_num)}、{cover_letter_section.title}")
        _set_paragraph_spacing(p, line_spacing=1.3, space_before=0, space_after=0)
        toc_num += 1
    if content.summary:
        p = doc.add_paragraph(f"{_to_chinese_number(toc_num)}、编制说明")
        _set_paragraph_spacing(p, line_spacing=1.3, space_before=0, space_after=0)
        toc_num += 1
    for s in content_sections:
        indent = "  " * (s.level - 1) if s.level > 1 else ""
        p = doc.add_paragraph(f"{indent}{s.title}")
        _set_paragraph_spacing(p, line_spacing=1.3, space_before=0, space_after=0)
    doc.add_page_break()

    # 3. 投标函
    section_counter = 1
    if cover_letter_section:
        _add_heading(doc, f"{_to_chinese_number(section_counter)}、{cover_letter_section.title}", 1)
        section_counter += 1
        _write_markdown_content(doc, cover_letter_section.content, base_level=2)
        doc.add_page_break()

    # 4. 投标概要
    if content.summary:
        _add_heading(doc, f"{_to_chinese_number(section_counter)}、编制说明", 1)
        section_counter += 1
        _write_markdown_content(doc, content.summary, base_level=2)
        doc.add_page_break()

    # 5. 各章节
    for section in content_sections:
        numbered_title = f"{_to_chinese_number(section_counter)}、{section.title}"
        section_counter += 1
        _add_heading(doc, numbered_title, 1)
        _write_markdown_content(doc, section.content, base_level=2)

    # 6. 页脚页码
    _add_footer_page_number(doc)

    try:
        doc.save(output_path)
    except (OSError, IOError) as e:
        raise OSError(
            f"无法保存DOCX文件到路径: {output_path}\n"
            f"请检查目录是否存在、磁盘空间是否充足。\n"
            f"原始错误: {e}"
        ) from e
    logger.info("DOCX文档已保存至: %s", output_path)
    return output_path


def _to_chinese_number(n: int) -> str:
    """将阿拉伯数字转为中文数字（一、二、三...）。

    Args:
        n: 阿拉伯数字

    Returns:
        str: 中文数字
    """
    chinese_nums = ["〇", "一", "二", "三", "四", "五", "六", "七", "八", "九", "十",
                    "十一", "十二", "十三", "十四", "十五", "十六", "十七", "十八", "十九", "二十"]
    if 1 <= n <= len(chinese_nums):
        return chinese_nums[n]
    return str(n)