@birthday8/doc-mcp 1.0.0

package/python/docx_converter.py

@@ -0,0 +1,807 @@
+from docx import Document
+from docx.shared import Pt, RGBColor, Inches, Cm
+from docx.enum.text import WD_PARAGRAPH_ALIGNMENT, WD_LINE_SPACING
+from docx.enum.section import WD_SECTION
+from docx.oxml.ns import qn
+from docx.oxml import OxmlElement
+from bs4 import BeautifulSoup
+import os
+import re
+
+def parse_color(color_str):
+    """解析颜色字符串为RGBColor"""
+    if not color_str or not color_str.startswith('#'):
+        return None
+    try:
+        r = int(color_str[1:3], 16)
+        g = int(color_str[3:5], 16)
+        b = int(color_str[5:7], 16)
+        return RGBColor(r, g, b)
+    except:
+        return None
+
+def set_font(run, font_name='微软雅黑', size=12, color=None, bold=False, italic=False, 
+            underline=False, strike=False, highlight_color=None):
+    """设置字体样式"""
+    run.font.name = font_name
+    run._element.rPr.rFonts.set(qn('w:eastAsia'), font_name)
+    run.font.size = Pt(size)
+    run.font.bold = bold
+    run.font.italic = italic
+    
+    if color:
+        run.font.color.rgb = color
+    
+    if underline:
+        run.font.underline = True
+    
+    if strike:
+        run.font.strike = True
+    
+    if highlight_color:
+        from docx.enum.text import WD_COLOR_INDEX
+        color_map = {
+            'yellow': WD_COLOR_INDEX.YELLOW,
+            'green': WD_COLOR_INDEX.BRIGHT_GREEN,
+            'cyan': WD_COLOR_INDEX.CYAN,
+            'magenta': WD_COLOR_INDEX.MAGENTA,
+            'blue': WD_COLOR_INDEX.TURQUOISE,
+            'red': WD_COLOR_INDEX.RED,
+            'darkblue': WD_COLOR_INDEX.BLUE,
+            'orange': WD_COLOR_INDEX.ORANGE,
+            'gray': WD_COLOR_INDEX.GRAY_25,
+        }
+        if highlight_color in color_map:
+            run.font.highlight_color = color_map[highlight_color]
+
+def process_inline_elements(element, parent_run=None):
+    """处理内联元素"""
+    from docx.text.paragraph import Paragraph
+    
+    runs = []
+    
+    for child in element.children:
+        if child.name is None:  # 文本节点
+            text = str(child).strip()
+            if text:
+                if parent_run:
+                    parent_run.add_text(text)
+                else:
+                    runs.append({'text': text})
+        elif child.name == 'strong' or child.name == 'b':
+            if parent_run:
+                parent_run.bold = True
+                process_inline_elements(child, parent_run)
+            else:
+                runs.append({'text': child.get_text(), 'bold': True})
+        elif child.name == 'em' or child.name == 'i':
+            if parent_run:
+                parent_run.italic = True
+                process_inline_elements(child, parent_run)
+            else:
+                runs.append({'text': child.get_text(), 'italic': True})
+        elif child.name == 'u':
+            if parent_run:
+                parent_run.underline = True
+                process_inline_elements(child, parent_run)
+            else:
+                runs.append({'text': child.get_text(), 'underline': True})
+        elif child.name == 's' or child.name == 'del':
+            if parent_run:
+                parent_run.strike = True
+                process_inline_elements(child, parent_run)
+            else:
+                runs.append({'text': child.get_text(), 'strike': True})
+        elif child.name == 'sup':
+            if parent_run:
+                parent_run.font.superscript = True
+                process_inline_elements(child, parent_run)
+            else:
+                runs.append({'text': child.get_text(), 'superscript': True})
+        elif child.name == 'sub':
+            if parent_run:
+                parent_run.font.subscript = True
+                process_inline_elements(child, parent_run)
+            else:
+                runs.append({'text': child.get_text(), 'subscript': True})
+        elif child.name == 'code':
+            code_text = child.get_text()
+            if parent_run:
+                parent_run.font.name = 'Consolas'
+                parent_run.font.size = Pt(10)
+                parent_run.add_text(code_text)
+            else:
+                runs.append({'text': code_text, 'font': 'Consolas', 'size': 10})
+        elif child.name == 'a':
+            link_text = child.get_text()
+            href = child.get('href', '')
+            if parent_run:
+                parent_run.add_text(link_text)
+            else:
+                runs.append({'text': link_text, 'link': href})
+        elif child.name == 'span':
+            style = child.get('style', '')
+            color_match = re.search(r'color:\s*([^;]+)', style)
+            bg_match = re.search(r'background(?:-color)?:\s*([^;]+)', style)
+            
+            props = {'text': child.get_text()}
+            if color_match:
+                color = parse_color(color_match.group(1).strip())
+                if color:
+                    props['color'] = color
+            if bg_match:
+                bg_color = bg_match.group(1).strip()
+                if bg_color.startswith('#'):
+                    bg_rgb = parse_color(bg_color)
+                    if bg_rgb:
+                        props['highlight'] = str(bg_rgb)
+            
+            if parent_run:
+                if 'color' in props:
+                    parent_run.font.color.rgb = props['color']
+                process_inline_elements(child, parent_run)
+            else:
+                runs.append(props)
+        else:
+            process_inline_elements(child, parent_run)
+    
+    return runs
+
+def _apply_highlight(run, bg_color):
+    """为run应用背景色/高亮"""
+    from docx.enum.text import WD_COLOR_INDEX
+    
+    # 颜色名称到WD_COLOR_INDEX的映射（只使用可用的枚举值）
+    color_map = {
+        'yellow': WD_COLOR_INDEX.YELLOW,
+        'green': WD_COLOR_INDEX.GREEN,
+        'brightgreen': WD_COLOR_INDEX.BRIGHT_GREEN,
+        'blue': WD_COLOR_INDEX.BLUE,
+        'darkblue': WD_COLOR_INDEX.DARK_BLUE,
+        'red': WD_COLOR_INDEX.RED,
+        'darkred': WD_COLOR_INDEX.DARK_RED,
+        'darkyellow': WD_COLOR_INDEX.DARK_YELLOW,
+        'lightgray': WD_COLOR_INDEX.GRAY_25,
+        'gray': WD_COLOR_INDEX.GRAY_50,
+        'black': WD_COLOR_INDEX.BLACK,
+        'white': WD_COLOR_INDEX.WHITE,
+        'pink': WD_COLOR_INDEX.PINK,
+        'teal': WD_COLOR_INDEX.TEAL,
+        'turquoise': WD_COLOR_INDEX.TURQUOISE,
+        'violet': WD_COLOR_INDEX.VIOLET,
+        'cyan': WD_COLOR_INDEX.TURQUOISE,
+        'magenta': WD_COLOR_INDEX.VIOLET,
+    }
+    
+    # 标准化颜色名称
+    bg_lower = bg_color.lower().strip()
+    
+    if bg_lower in color_map:
+        # 使用预定义的高亮色
+        run.font.highlight_color = color_map[bg_lower]
+    elif bg_lower.startswith('#'):
+        # 十六进制颜色，直接使用字符串
+        shading_elm = OxmlElement('w:shd')
+        shading_elm.set(qn('w:fill'), bg_lower[1:].upper())
+        run._element.get_or_add_rPr().append(shading_elm)
+    else:
+        # 尝试其他常见颜色名称映射到相近的预定义颜色
+        similar_colors = {
+            'lightblue': WD_COLOR_INDEX.TURQUOISE,
+            'lightyellow': WD_COLOR_INDEX.YELLOW,
+            'lightgreen': WD_COLOR_INDEX.BRIGHT_GREEN,
+            'orange': WD_COLOR_INDEX.YELLOW,  # 橙色映射到黄色
+            'purple': WD_COLOR_INDEX.VIOLET,
+            'brown': WD_COLOR_INDEX.DARK_YELLOW,
+        }
+        if bg_lower in similar_colors:
+            run.font.highlight_color = similar_colors[bg_lower]
+
+def process_paragraph(paragraph, doc, default_font='微软雅黑', default_size=12, 
+                     indent=None, align=None, line_spacing=None):
+    """处理段落及其内联元素"""
+    para = doc.add_paragraph()
+
+    # 设置段落格式
+    if align:
+        para.alignment = align
+
+    # 优先使用传入的indent参数，否则尝试从data-indent属性读取
+    if indent is not None:
+        para.paragraph_format.first_line_indent = Inches(indent)
+    else:
+        # 从data-indent属性读取缩进（单位：em）
+        data_indent = paragraph.get('data-indent', '')
+        if data_indent:
+            try:
+                em_count = float(data_indent)
+                # 缩进 = 字体大小(pt) × em数
+                indent_pt = default_size * em_count
+                para.paragraph_format.first_line_indent = Pt(indent_pt)
+            except ValueError:
+                pass
+
+    if line_spacing:
+        para.paragraph_format.line_spacing = line_spacing
+
+    # 处理段落内容 - 递归处理所有子元素
+    _process_element_to_runs(paragraph, para, default_font, default_size)
+
+    return para
+
+def _process_element_to_runs(element, para, default_font='微软雅黑', default_size=12, 
+                              bold=False, italic=False, underline=False, strike=False,
+                              color=None, bg_color=None, font_name=None, font_size=None):
+    """递归处理元素，为不同格式的文本创建独立的runs"""
+    current_font = font_name or default_font
+    current_size = font_size or default_size
+    
+    for child in element.children:
+        if child.name is None:  # 文本节点
+            text = str(child)
+            # 去除多余空白但保留单个空格
+            if text:
+                # 替换换行和制表符为空格，然后合并多个空格
+                text = ' '.join(text.replace('\n', ' ').replace('\t', ' ').split())
+                if text:  # 再次检查，因为去除空白后可能为空
+                    run = para.add_run(text)
+                    set_font(run, font_name=current_font, size=current_size, 
+                            bold=bold, italic=italic, underline=underline, strike=strike)
+                    if color:
+                        run.font.color.rgb = color
+                    # 应用背景色
+                    if bg_color:
+                        _apply_highlight(run, bg_color)
+        elif child.name == 'strong' or child.name == 'b':
+            _process_element_to_runs(child, para, default_font, default_size,
+                                    bold=True, italic=italic, underline=underline, strike=strike,
+                                    color=color, bg_color=bg_color)
+        elif child.name == 'em' or child.name == 'i':
+            _process_element_to_runs(child, para, default_font, default_size,
+                                    bold=bold, italic=True, underline=underline, strike=strike,
+                                    color=color, bg_color=bg_color)
+        elif child.name == 'u':
+            _process_element_to_runs(child, para, default_font, default_size,
+                                    bold=bold, italic=italic, underline=True, strike=strike,
+                                    color=color, bg_color=bg_color)
+        elif child.name == 's' or child.name == 'del':
+            _process_element_to_runs(child, para, default_font, default_size,
+                                    bold=bold, italic=italic, underline=underline, strike=True,
+                                    color=color, bg_color=bg_color)
+        elif child.name == 'sup':
+            for sub_child in child.children:
+                if sub_child.name is None:
+                    run = para.add_run(str(sub_child))
+                    set_font(run, font_name=current_font, size=current_size,
+                            bold=bold, italic=italic, underline=underline, strike=strike)
+                    run.font.superscript = True
+                    if color:
+                        run.font.color.rgb = color
+                else:
+                    _process_element_to_runs(child, para, default_font, default_size,
+                                            bold=bold, italic=italic, underline=underline, strike=strike,
+                                            color=color, bg_color=bg_color)
+        elif child.name == 'sub':
+            for sub_child in child.children:
+                if sub_child.name is None:
+                    run = para.add_run(str(sub_child))
+                    set_font(run, font_name=current_font, size=current_size,
+                            bold=bold, italic=italic, underline=underline, strike=strike)
+                    run.font.subscript = True
+                    if color:
+                        run.font.color.rgb = color
+                else:
+                    _process_element_to_runs(child, para, default_font, default_size,
+                                            bold=bold, italic=italic, underline=underline, strike=strike,
+                                            color=color, bg_color=bg_color)
+        elif child.name == 'code':
+            code_text = child.get_text()
+            run = para.add_run(code_text)
+            set_font(run, font_name='Consolas', size=10)
+        elif child.name == 'a':
+            link_text = child.get_text()
+            run = para.add_run(link_text)
+            set_font(run, font_name=current_font, size=current_size)
+            run.font.underline = True
+            run.font.color.rgb = RGBColor(0, 0, 255)
+        elif child.name == 'span':
+            # 处理span的样式
+            style = child.get('style', '')
+            classes = child.get('class', [])
+            
+            span_color = color
+            span_bg = bg_color
+            
+            # 解析style中的颜色
+            color_match = re.search(r'color:\s*([^;]+)', style)
+            if color_match:
+                parsed = parse_color(color_match.group(1).strip())
+                if parsed:
+                    span_color = parsed
+            
+            # 解析class中的颜色
+            if 'red' in classes:
+                span_color = RGBColor(255, 0, 0)
+            elif 'blue' in classes:
+                span_color = RGBColor(0, 0, 255)
+            elif 'green' in classes:
+                span_color = RGBColor(0, 128, 0)
+            elif 'purple' in classes:
+                span_color = RGBColor(128, 0, 128)
+            
+            # 背景色
+            bg_match = re.search(r'background(?:-color)?:\s*([^;]+)', style)
+            if bg_match:
+                span_bg = bg_match.group(1).strip()
+            if 'highlight' in classes:
+                span_bg = 'yellow'
+            
+            _process_element_to_runs(child, para, default_font, default_size,
+                                    bold=bold, italic=italic, underline=underline, strike=strike,
+                                    color=span_color, bg_color=span_bg, 
+                                    font_name=current_font, font_size=current_size)
+        else:
+            # 其他标签递归处理
+            _process_element_to_runs(child, para, default_font, default_size,
+                                    bold=bold, italic=italic, underline=underline, strike=strike,
+                                    color=color, bg_color=bg_color)
+
+def process_list_items(items, doc, ordered=False, default_font='微软雅黑', default_size=12, level=0):
+    """处理列表项，支持嵌套"""
+    for item in items:
+        # 创建列表项段落
+        if ordered:
+            para = doc.add_paragraph(style='List Number')
+        else:
+            para = doc.add_paragraph(style='List Bullet')
+        
+        # 设置缩进：每级增加 0.25 英寸
+        para.paragraph_format.left_indent = Inches(0.25 * (level + 1))
+        
+        # 查找嵌套列表
+        nested_ul = item.find('ul', recursive=False)
+        nested_ol = item.find('ol', recursive=False)
+        
+        # 处理列表项的文本内容（排除嵌套列表）
+        # 创建一个临时副本用于提取文本
+        item_copy = BeautifulSoup(str(item), 'html.parser').find('li')
+        if item_copy:
+            # 移除嵌套列表
+            for nested in item_copy.find_all(['ul', 'ol'], recursive=False):
+                nested.decompose()
+            
+            # 处理剩余内容
+            if item_copy.get_text().strip():
+                _process_element_to_runs(item_copy, para, default_font, default_size)
+        
+        # 递归处理嵌套列表
+        if nested_ul:
+            nested_items = nested_ul.find_all('li', recursive=False)
+            process_list_items(nested_items, doc, ordered=False, 
+                             default_font=default_font, default_size=default_size, level=level+1)
+        if nested_ol:
+            nested_items = nested_ol.find_all('li', recursive=False)
+            process_list_items(nested_items, doc, ordered=True, 
+                             default_font=default_font, default_size=default_size, level=level+1)
+
+def _parse_style(style_str):
+    """解析style字符串为字典"""
+    styles = {}
+    if not style_str:
+        return styles
+    for item in style_str.split(';'):
+        if ':' in item:
+            key, value = item.split(':', 1)
+            styles[key.strip()] = value.strip()
+    return styles
+
+def _apply_cell_style(cell_elem, style_dict):
+    """应用单元格样式"""
+    # 背景色
+    bg_color = style_dict.get('background-color', '')
+    if bg_color:
+        # 处理颜色值
+        if bg_color.startswith('#'):
+            shading_elm = OxmlElement('w:shd')
+            shading_elm.set(qn('w:fill'), bg_color[1:].upper())
+            cell_elem._element.get_or_add_tcPr().append(shading_elm)
+    
+    # 文字颜色
+    color = style_dict.get('color', '')
+    if color:
+        rgb = parse_color(color) if color.startswith('#') else None
+        if rgb:
+            for run in cell_elem.paragraphs[0].runs:
+                run.font.color.rgb = rgb
+
+def process_table(table, doc, default_font='微软雅黑', default_size=11):
+    """处理表格，支持内联样式"""
+    rows = table.find_all('tr')
+    if not rows:
+        return
+    
+    # 获取列数
+    cols = max(len(row.find_all(['td', 'th'])) for row in rows)
+    
+    # 创建表格
+    word_table = doc.add_table(rows=len(rows), cols=cols)
+    word_table.style = 'Table Grid'
+    
+    for row_idx, row in enumerate(rows):
+        # 处理行样式（如背景色）
+        row_style = _parse_style(row.get('style', ''))
+        row_bg = row_style.get('background-color', '')
+        
+        cells = row.find_all(['td', 'th'])
+        for col_idx, cell in enumerate(cells):
+            if col_idx < cols:
+                cell_elem = word_table.rows[row_idx].cells[col_idx]
+                cell_elem.paragraphs[0].text = cell.get_text().strip()
+                
+                # 解析单元格样式
+                cell_style = _parse_style(cell.get('style', ''))
+                
+                # 表头加粗
+                if cell.name == 'th':
+                    for run in cell_elem.paragraphs[0].runs:
+                        run.font.bold = True
+                
+                # 设置单元格对齐
+                align = cell_style.get('text-align', 'center')
+                if align == 'center':
+                    cell_elem.paragraphs[0].alignment = WD_PARAGRAPH_ALIGNMENT.CENTER
+                elif align == 'left':
+                    cell_elem.paragraphs[0].alignment = WD_PARAGRAPH_ALIGNMENT.LEFT
+                elif align == 'right':
+                    cell_elem.paragraphs[0].alignment = WD_PARAGRAPH_ALIGNMENT.RIGHT
+                
+                # 应用单元格样式（背景色、文字颜色）
+                _apply_cell_style(cell_elem, cell_style)
+                
+                # 如果行有背景色且单元格没有单独设置，应用行背景色
+                if row_bg and not cell_style.get('background-color'):
+                    if row_bg.startswith('#'):
+                        shading_elm = OxmlElement('w:shd')
+                        shading_elm.set(qn('w:fill'), row_bg[1:].upper())
+                        cell_elem._element.get_or_add_tcPr().append(shading_elm)
+
+def set_section_columns(section, cols_num=2, space=720):
+    """设置节的多栏布局
+
+    Args:
+        section: docx.section.Section对象
+        cols_num: 栏数，默认为2
+        space: 栏间距（单位：twips，1英寸=1440twips），默认720（0.5英寸）
+    """
+    sectPr = section._sectPr
+    cols = OxmlElement('w:cols')
+    cols.set(qn('w:num'), str(cols_num))
+    cols.set(qn('w:space'), str(space))
+    sectPr.append(cols)
+
+def add_columns_section(doc, cols_num=2, space=720):
+    """添加连续分节符并设置多栏布局（不换页）
+
+    Args:
+        doc: Document对象
+        cols_num: 栏数，默认为2
+        space: 栏间距（单位：twips），默认720（0.5英寸）
+
+    Returns:
+        section: 新创建的节对象
+    """
+    # 添加连续分节符（不换页）
+    section = doc.add_section(start_type=WD_SECTION.CONTINUOUS)
+    # 设置栏数
+    set_section_columns(section, cols_num, space)
+    return section
+
+def _process_blockquote(blockquote_elem, doc, level=0):
+    """递归处理嵌套引用"""
+    # 获取当前引用的直接文本内容（不包括嵌套引用）
+    direct_text = ''
+    for child in blockquote_elem.children:
+        if child.name is None:  # 文本节点
+            direct_text += str(child)
+        elif child.name != 'blockquote':  # 其他非引用标签
+            direct_text += child.get_text()
+
+    direct_text = ' '.join(direct_text.split())
+
+    # 如果有直接文本，创建段落
+    if direct_text:
+        para = doc.add_paragraph()
+        run = para.add_run(direct_text)
+        set_font(run, italic=True, color=RGBColor(100, 100, 100))
+        # 根据层级设置缩进
+        para.paragraph_format.left_indent = Inches(0.3 * level)
+        para.paragraph_format.right_indent = Inches(0.5)
+        # 添加灰色左边框
+        pBdr = OxmlElement('w:pBdr')
+        left_border = OxmlElement('w:left')
+        left_border.set(qn('w:val'), 'single')
+        left_border.set(qn('w:sz'), '18')
+        left_border.set(qn('w:color'), 'CCCCCC')
+        pBdr.append(left_border)
+        para.paragraph_format._element.get_or_add_pPr().append(pBdr)
+
+    # 递归处理嵌套引用
+    nested_quotes = blockquote_elem.find_all('blockquote', recursive=False)
+    for nested in nested_quotes:
+        _process_blockquote(nested, doc, level + 1)
+
+def add_page_break(doc):
+    """添加分页符"""
+    doc.add_page_break()
+
+def add_horizontal_rule(doc):
+    """添加水平线"""
+    para = doc.add_paragraph()
+    para.alignment = WD_PARAGRAPH_ALIGNMENT.CENTER
+    run = para.add_run('_' * 50)
+    run.font.color.rgb = RGBColor(200, 200, 200)
+
+def convert_html_to_docx(html_file, output_file, default_font='微软雅黑', default_size=12):
+    """将HTML文件转换为DOCX文件"""
+    # 读取HTML文件
+    with open(html_file, 'r', encoding='utf-8') as f:
+        html_content = f.read()
+    
+    # 解析HTML
+    soup = BeautifulSoup(html_content, 'html.parser')
+    
+    # 创建Word文档
+    doc = Document()
+    
+    # 设置默认字体
+    doc.styles['Normal'].font.name = default_font
+    doc.styles['Normal']._element.rPr.rFonts.set(qn('w:eastAsia'), default_font)
+    doc.styles['Normal'].font.size = Pt(default_size)
+    
+    # 处理页面设置
+    section = doc.sections[0]
+    section.page_height = Cm(29.7)
+    section.page_width = Cm(21)
+    section.left_margin = Cm(2.54)
+    section.right_margin = Cm(2.54)
+    section.top_margin = Cm(2.54)
+    section.bottom_margin = Cm(2.54)
+    
+    # 遍历所有顶级元素
+    for element in soup.body.find_all(recursive=False):
+        if element.name in ['h1', 'h2', 'h3', 'h4', 'h5', 'h6']:
+            level = int(element.name[1])
+            heading = doc.add_heading(element.get_text().strip(), level=level)
+            heading.alignment = WD_PARAGRAPH_ALIGNMENT.LEFT
+            
+            # 标题样式
+            for run in heading.runs:
+                run.font.name = default_font
+                run._element.rPr.rFonts.set(qn('w:eastAsia'), default_font)
+                if level == 1:
+                    run.font.size = Pt(18)
+                    run.font.bold = True
+                    run.font.color.rgb = RGBColor(74, 63, 107)
+                    heading.alignment = WD_PARAGRAPH_ALIGNMENT.CENTER
+                elif level == 2:
+                    run.font.size = Pt(16)
+                    run.font.bold = True
+                    run.font.color.rgb = RGBColor(91, 78, 140)
+                else:
+                    run.font.size = Pt(14)
+                    run.font.bold = True
+        
+        elif element.name == 'p':
+            # 检查特殊段落样式
+            classes = element.get('class', [])
+            
+            if 'center' in classes:
+                para = process_paragraph(element, doc, align=WD_PARAGRAPH_ALIGNMENT.CENTER)
+            elif 'right' in classes:
+                para = process_paragraph(element, doc, align=WD_PARAGRAPH_ALIGNMENT.RIGHT)
+            elif 'dialogue' in classes:
+                para = process_paragraph(element, doc, indent=0.5, line_spacing=1.5)
+                for run in para.runs:
+                    set_font(run, italic=True, color=RGBColor(107, 91, 122))
+            elif 'quote' in classes or element.get('style', '').find('background') != -1:
+                para = process_paragraph(element, doc)
+                para.paragraph_format.left_indent = Inches(1)
+                para.paragraph_format.right_indent = Inches(1)
+                from docx.enum.text import WD_BORDER
+                for border in para.paragraph_format._element.xpath('./w:pBdr'):
+                    border.getparent().remove(border)
+                # 添加边框效果（使用浅灰色背景模拟）
+                shading_elm = OxmlElement('w:shd')
+                shading_elm.set(qn('w:fill'), 'F5F5F5')
+                para.paragraph_format._element.get_or_add_pPr().append(shading_elm)
+            else:
+                process_paragraph(element, doc, indent=0.5, line_spacing=1.5)
+        
+        elif element.name == 'ul':
+            items = element.find_all('li', recursive=False)
+            process_list_items(items, doc, ordered=False)
+        
+        elif element.name == 'ol':
+            items = element.find_all('li', recursive=False)
+            process_list_items(items, doc, ordered=True)
+        
+        elif element.name == 'blockquote':
+            # 递归处理嵌套引用
+            _process_blockquote(element, doc, level=0)
+        
+        elif element.name == 'pre':
+            code_text = element.get_text()
+            para = doc.add_paragraph()
+            para.paragraph_format.left_indent = Inches(0.5)
+            run = para.add_run(code_text)
+            set_font(run, font_name='Consolas', size=10, color=RGBColor(0, 0, 128))
+            # 添加灰色背景
+            shading_elm = OxmlElement('w:shd')
+            shading_elm.set(qn('w:fill'), 'F0F0F0')
+            para.paragraph_format._element.get_or_add_pPr().append(shading_elm)
+        
+        elif element.name == 'hr':
+            # 检查是否有分页符class或style
+            classes = element.get('class', [])
+            style = element.get('style', '')
+            if 'page-break' in classes or 'page-break-after' in style:
+                add_page_break(doc)
+            else:
+                add_horizontal_rule(doc)
+        
+        elif element.name == 'table':
+            process_table(element, doc)
+        
+        elif element.name == 'div':
+            # 检查是否是特殊div
+            classes = element.get('class', [])
+            if 'chapter' in classes:
+                # 处理章节
+                h2 = element.find('h2')
+                if h2:
+                    heading = doc.add_heading(h2.get_text().strip(), level=2)
+                    for run in heading.runs:
+                        run.font.color.rgb = RGBColor(91, 78, 140)
+                        run.font.size = Pt(16)
+                        run.font.name = default_font
+                        run._element.rPr.rFonts.set(qn('w:eastAsia'), default_font)
+                
+                paragraphs = element.find_all('p')
+                for p in paragraphs:
+                    first_span = p.find('span', class_='first-line')
+                    if first_span:
+                        para = doc.add_paragraph()
+                        first_char_run = para.add_run(first_span.text)
+                        set_font(first_char_run, size=20, bold=True, color=RGBColor(102, 126, 234))
+                        remaining_text = p.get_text().replace(first_span.text, '', 1)
+                        run = para.add_run(remaining_text)
+                        set_font(run)
+                    else:
+                        process_paragraph(p, doc, indent=0.5, line_spacing=1.5)
+            
+            elif 'ending' in classes:
+                para = process_paragraph(element, doc, align=WD_PARAGRAPH_ALIGNMENT.CENTER)
+                for run in para.runs:
+                    set_font(run, italic=True, size=14, color=RGBColor(91, 78, 140))
+            
+            elif 'page-break' in classes:
+                add_page_break(doc)
+
+            elif 'columns' in classes:
+                # 处理多栏布局（使用连续分节符，不换页）
+                cols_num = int(element.get('data-cols', '2'))
+                # 添加连续分节符并设置栏数
+                add_columns_section(doc, cols_num)
+                # 处理其中的段落
+                for p in element.find_all('p', recursive=False):
+                    process_paragraph(p, doc, default_font=default_font, default_size=default_size)
+
+            elif 'info' in classes or 'warning' in classes or 'success' in classes:
+                # 处理提示框
+                para = doc.add_paragraph()
+                para.paragraph_format.right_indent = Inches(0.3)
+                
+                # 设置背景色和左边框颜色
+                if 'info' in classes:
+                    bg_color = 'E3F2FD'  # 浅蓝
+                    border_color = '2196F3'  # 蓝色
+                elif 'warning' in classes:
+                    bg_color = 'FFF3CD'  # 浅黄
+                    border_color = 'FFC107'  # 黄色
+                else:  # success
+                    bg_color = 'D4EDDA'  # 浅绿
+                    border_color = '28A745'  # 绿色
+                
+                # 处理内容
+                _process_element_to_runs(element, para, default_font, default_size)
+                
+                # 添加背景色
+                shading_elm = OxmlElement('w:shd')
+                shading_elm.set(qn('w:fill'), bg_color)
+                para.paragraph_format._element.get_or_add_pPr().append(shading_elm)
+                
+                # 添加左边框
+                pBdr = OxmlElement('w:pBdr')
+                left_border = OxmlElement('w:left')
+                left_border.set(qn('w:val'), 'single')
+                left_border.set(qn('w:sz'), '24')  # 边框粗细
+                left_border.set(qn('w:color'), border_color)
+                pBdr.append(left_border)
+                para.paragraph_format._element.get_or_add_pPr().append(pBdr)
+            
+            else:
+                # 处理普通div，检查是否有内联样式（如提示框）
+                style = element.get('style', '')
+                style_dict = _parse_style(style)
+                
+                # 检查是否有背景色和左边框（提示框特征）
+                bg_color = style_dict.get('background-color', '')
+                border_left = style_dict.get('border-left', '')
+                
+                if bg_color and border_left:
+                    # 这是提示框
+                    para = doc.add_paragraph()
+                    para.paragraph_format.right_indent = Inches(0.3)
+                    
+                    # 处理内容
+                    _process_element_to_runs(element, para, default_font, default_size)
+                    
+                    # 添加背景色
+                    if bg_color.startswith('#'):
+                        shading_elm = OxmlElement('w:shd')
+                        shading_elm.set(qn('w:fill'), bg_color[1:].upper())
+                        para.paragraph_format._element.get_or_add_pPr().append(shading_elm)
+                    
+                    # 解析左边框颜色
+                    border_color = ''
+                    if 'solid' in border_left:
+                        parts = border_left.split()
+                        for i, part in enumerate(parts):
+                            if part.startswith('#'):
+                                border_color = part[1:]
+                                break
+                    
+                    # 添加左边框
+                    if border_color:
+                        pBdr = OxmlElement('w:pBdr')
+                        left_border = OxmlElement('w:left')
+                        left_border.set(qn('w:val'), 'single')
+                        left_border.set(qn('w:sz'), '24')
+                        left_border.set(qn('w:color'), border_color.upper())
+                        pBdr.append(left_border)
+                        para.paragraph_format._element.get_or_add_pPr().append(pBdr)
+                else:
+                    # 普通div，处理其中的段落
+                    for p in element.find_all('p', recursive=False):
+                        process_paragraph(p, doc)
+        
+        elif element.name == 'img':
+            src = element.get('src', '')
+            alt = element.get('alt', '图片')
+            if src and os.path.exists(src):
+                try:
+                    doc.add_picture(src, width=Inches(5))
+                    last_para = doc.paragraphs[-1]
+                    last_para.alignment = WD_PARAGRAPH_ALIGNMENT.CENTER
+                except:
+                    para = doc.add_paragraph(f'[图片: {alt}]')
+                    para.alignment = WD_PARAGRAPH_ALIGNMENT.CENTER
+            else:
+                para = doc.add_paragraph(f'[图片: {alt} - 路径: {src}]')
+                para.alignment = WD_PARAGRAPH_ALIGNMENT.CENTER
+    
+    # 保存文档
+    doc.save(output_file)
+    print(f"转换完成！文件已保存为 {output_file}")
+
+if __name__ == '__main__':
+    import sys
+    
+    if len(sys.argv) > 2:
+        html_file = sys.argv[1]
+        output_file = sys.argv[2]
+    else:
+        html_file = r'C:\Users\birth\Desktop\tmp\test\sample.html'
+        output_file = r'C:\Users\birth\Desktop\tmp\test\sample.docx'
+    
+    convert_html_to_docx(html_file, output_file)