""" Markdown 正文 → Word 段落转换。 将正文 Markdown 按标题层级拆分为带样式的段落序列, 再注入到渲染后 docx 文档的占位符位置。 """ import re from docx import Document _PAT_HEADING = re.compile(r"^(#{1,6})\s+(.+)$", re.MULTILINE) def body_to_paragraphs(md_text: str) -> list[dict]: """将 Markdown 正文按标题和段落拆分为结构化列表。 返回的每个元素:: {"text": str, "level": int, "style": str} 其中 ``style`` 为 ``Heading N`` 或 ``Normal``。 """ paragraphs: list[dict] = [] last_end = 0 for m in _PAT_HEADING.finditer(md_text): # 标题前的普通文本 if m.start() > last_end: pre = md_text[last_end : m.start()].strip() if pre: for block in re.split(r"\n\s*\n", pre): block = block.strip() if block: paragraphs.append( {"text": block, "level": 0, "style": "Normal"} ) level = len(m.group(1)) heading_text = m.group(2).strip() paragraphs.append( {"text": heading_text, "level": level, "style": f"Heading {level}"} ) last_end = m.end() # 最后一段 / 尾部文本 tail = md_text[last_end:].strip() if tail: for block in re.split(r"\n\s*\n", tail): block = block.strip() if block: paragraphs.append( {"text": block, "level": 0, "style": "Normal"} ) return paragraphs def replace_placeholder(doc: Document, placeholder: str, paragraphs: list[dict]): """在 *doc* 中找到包含 *placeholder* 的段落,替换为 *paragraphs* 列表。 每个段落的 ``style`` 字段会从文档样式中查找并应用。 """ placeholder_found = False for para in doc.paragraphs: if placeholder in para.text: placeholder_found = True parent = para._element.getparent() idx = list(parent).index(para._element) parent.remove(para._element) for pd_data in reversed(paragraphs): new_p = doc.add_paragraph(pd_data["text"]) style_name = pd_data["style"] try: new_p.style = doc.styles[style_name] except KeyError: matched = False for s in doc.styles: if s.name.lower() == style_name.lower(): new_p.style = s matched = True break if not matched: new_p.style = doc.styles["Normal"] parent.insert(idx, new_p._element) break if not placeholder_found: print(f"警告:未找到占位符 '{placeholder}',正文段落未注入。")