pydatamax 0.1.15__py3-none-any.whl → 0.1.16__py3-none-any.whl

datamax/parser/core.py

@@ -87,7 +87,7 @@ class ParserFactory:
                 )
             elif parser_class_name == 'DocxParser' or parser_class_name == 'DocParser':
                 return parser_class(
-                    file_path=file_path, to_markdown=to_markdown
+                    file_path=file_path, to_markdown=to_markdown, use_uno=True
                 )
             elif parser_class_name == 'XlsxParser':
                 return parser_class(

datamax/parser/doc_parser.py

@@ -26,6 +26,21 @@ try:
     from datamax.utils.uno_handler import HAS_UNO, convert_with_uno
 except ImportError:
     HAS_UNO = False
+    logger.error(
+        "❌ UNO处理器导入失败！\n"
+        "🔧 解决方案：\n"
+        "1. 安装LibreOffice和python-uno：\n"
+        "   - Ubuntu/Debian: sudo apt-get install libreoffice python3-uno\n"
+        "   - CentOS/RHEL: sudo yum install libreoffice python3-uno\n"
+        "   - macOS: brew install libreoffice\n"
+        "   - Windows: 下载并安装LibreOffice\n"
+        "2. 确保Python可以访问uno模块：\n"
+        "   - Linux: export PYTHONPATH=/usr/lib/libreoffice/program:$PYTHONPATH\n"
+        "   - Windows: 添加LibreOffice\\program到系统PATH\n"
+        "3. 验证安装：python -c 'import uno'\n"
+        "4. 如果仍有问题，请查看完整文档：\n"
+        "   https://wiki.documentfoundation.org/Documentation/DevGuide/Installing_the_SDK"
+    )
 
 
 class DocParser(BaseLife):
@@ -46,7 +61,11 @@ class DocParser(BaseLife):
         else:
             self.use_uno = False
             if use_uno and not HAS_UNO:
-                logger.warning(f"⚠️ UNO不可用，回退到传统命令行方式")
+                logger.warning(
+                    f"⚠️ UNO不可用，回退到传统命令行方式\n"
+                    f"💡 提示：UNO转换更快更稳定，强烈建议安装和配置UNO\n"
+                    f"📖 请参考上述错误信息中的安装指南"
+                )
             else:
                 logger.info(f"🚀 DocParser初始化完成 - 使用传统命令行方式")
 
@@ -265,52 +284,78 @@ class DocParser(BaseLife):
             return ""
 
     def _clean_extracted_text(self, text: str) -> str:
-        """清理提取的文本，移除控制字符和格式化，但保留中文"""
+        """清理提取的文本，彻底移除所有XML标签和控制字符，只保留纯文本"""
         try:
-            # 移除NULL字符和其他控制字符（但不移除换行等）
+            # 1. 解码HTML/XML实体
+            text = html.unescape(text)
+            
+            # 2. 移除所有XML/HTML标签
+            text = re.sub(r'<[^>]+>', '', text)
+            
+            # 3. 移除XML命名空间前缀
+            text = re.sub(r'\b\w+:', '', text)
+            
+            # 4. 移除NULL字符和其他控制字符
             text = re.sub(r'[\x00-\x08\x0b-\x0c\x0e-\x1f\x7f]', '', text)
             
-            # 移除连续的特殊字符（但保留中文和常用标点）
-            # 修改正则表达式，确保不会误删中文
-            text = re.sub(r'[^\w\s\u4e00-\u9fff\u3040-\u30ff\uac00-\ud7af，。！？；：""''（）【】《》、·…—.,!?;:()-]+', ' ', text)
+            # 5. 移除特殊的XML字符序列
+            text = re.sub(r'&[a-zA-Z]+;', '', text)
+            text = re.sub(r'&#\d+;', '', text)
+            text = re.sub(r'&#x[0-9a-fA-F]+;', '', text)
+            
+            # 6. 保留有意义的字符，移除其他特殊字符
+            # 保留：中文、日文、韩文、英文、数字、常用标点和空白
+            allowed_chars = (
+                r'\w\s'  # 字母数字和空白
+                r'\u4e00-\u9fff'  # 中文
+                r'\u3040-\u30ff'  # 日文
+                r'\uac00-\ud7af'  # 韩文
+                r'，。！？；：""''（）【】《》、·…—'  # 中文标点
+                r'.,!?;:()[\]{}"\'`~@#$%^&*+=\-_/\\'  # 英文标点和常用符号
+            )
+            
+            # 使用更严格的过滤，但保留所有有意义的字符
+            cleaned_text = ''.join(char for char in text if re.match(f'[{allowed_chars}]', char))
             
-            # 移除过长的无意义字符序列（通常是乱码）
-            text = re.sub(r'[\x80-\xff]{10,}', ' ', text)
+            # 7. 移除过长的无意义字符序列（通常是二进制垃圾）
+            cleaned_text = re.sub(r'([^\s\u4e00-\u9fff])\1{5,}', r'\1', cleaned_text)
             
-            # 移除重复的空白
-            text = re.sub(r'\s+', ' ', text)
-            text = re.sub(r'\n\s*\n\s*\n', '\n\n', text)
+            # 8. 清理多余的空白，但保留段落结构
+            cleaned_text = re.sub(r'[ \t]+', ' ', cleaned_text)  # 多个空格/制表符变为单个空格
+            cleaned_text = re.sub(r'\n\s*\n\s*\n+', '\n\n', cleaned_text)  # 多个空行变为双空行
+            cleaned_text = re.sub(r'^\s+|\s+$', '', cleaned_text, flags=re.MULTILINE)  # 移除行首行尾空白
             
-            # 确保段落分隔
-            lines = text.split('\n')
+            # 9. 进一步清理：移除独立的标点符号行
+            lines = cleaned_text.split('\n')
             cleaned_lines = []
             
             for line in lines:
                 line = line.strip()
                 if line:
-                    # 检查行是否主要是乱码
-                    printable_chars = sum(1 for c in line if c.isprintable() or '\u4e00' <= c <= '\u9fff')
-                    total_chars = len(line)
+                    # 检查行是否主要是有意义的内容
+                    # 计算中文、英文字母和数字的比例
+                    meaningful_chars = sum(1 for c in line if (
+                        c.isalnum() or '\u4e00' <= c <= '\u9fff'
+                    ))
                     
-                    # 如果可打印字符（包括中文）占比超过60%，则保留该行
-                    if total_chars > 0 and printable_chars / total_chars > 0.6:
+                    # 如果有意义字符占比超过30%，或者行长度小于5（可能是标题），则保留
+                    if (len(line) < 5 or 
+                        (meaningful_chars > 0 and meaningful_chars / len(line) > 0.3)):
                         cleaned_lines.append(line)
-                elif cleaned_lines and cleaned_lines[-1]:
-                    cleaned_lines.append('')  # 保留段落分隔
+                elif cleaned_lines and cleaned_lines[-1]:  # 保留段落分隔
+                    cleaned_lines.append('')
             
             result = '\n'.join(cleaned_lines).strip()
             
-            # 最后检查：如果结果太短或包含太多乱码，返回空
-            if len(result) < 20:
+            # 10. 最终检查
+            if len(result) < 10:
+                logger.warning("⚠️ 清理后的文本过短，可能存在问题")
                 return ""
             
-            # 检查乱码比例
-            weird_chars = sum(1 for c in result if ord(c) > 127 and not ('\u4e00' <= c <= '\u9fff' or c in '，。！？；：""''（）【】《》、·…—'))
-            if len(result) > 0 and weird_chars / len(result) > 0.3:
-                logger.warning(f"⚠️ 文本包含过多乱码字符 ({weird_chars}/{len(result)})")
-                # 尝试只保留ASCII和中文部分
-                result = re.sub(r'[^\x00-\x7f\u4e00-\u9fff\u3040-\u30ff\uac00-\ud7af，。！？；：""''（）【】《》、·…—\s]+', ' ', result)
-                result = re.sub(r'\s+', ' ', result).strip()
+            # 检查是否还包含XML标签
+            if re.search(r'<[^>]+>', result):
+                logger.warning("⚠️ 清理后仍包含XML标签，进行二次清理")
+                result = re.sub(r'<[^>]+>', '', result)
             
             return result
             
@@ -355,7 +400,20 @@ class DocParser(BaseLife):
                     return txt_path
 
             except Exception as e:
-                logger.error(f"💥 UNO转换失败: {str(e)}")
+                logger.error(
+                    f"💥 UNO转换失败: {str(e)}\n"
+                    f"🔍 诊断信息：\n"
+                    f"   - 错误类型: {type(e).__name__}\n"
+                    f"   - LibreOffice是否已安装？尝试运行: soffice --version\n"
+                    f"   - Python UNO模块是否可用？尝试: python -c 'import uno'\n"
+                    f"   - 是否有其他LibreOffice实例在运行？\n"
+                    f"   - 文件权限是否正确？\n"
+                    f"🔧 可能的解决方案：\n"
+                    f"   1. 确保LibreOffice正确安装\n"
+                    f"   2. 关闭所有LibreOffice进程\n"
+                    f"   3. 检查文件权限和路径\n"
+                    f"   4. 尝试手动运行: soffice --headless --convert-to txt \"{doc_path}\""
+                )
                 logger.warning("⚠️ 自动回退到传统命令行方式...")
                 return self._doc_to_txt_subprocess(doc_path, dir_path)
         else:

datamax/parser/docx_parser.py

@@ -18,7 +18,21 @@ try:
     from datamax.utils.uno_handler import HAS_UNO, convert_with_uno
 except ImportError:
     HAS_UNO = False
-    logger.warning("⚠️ UNO不可用，回退到传统命令行方式")
+    logger.error(
+        "❌ UNO处理器导入失败！\n"
+        "🔧 解决方案：\n"
+        "1. 安装LibreOffice和python-uno：\n"
+        "   - Ubuntu/Debian: sudo apt-get install libreoffice python3-uno\n"
+        "   - CentOS/RHEL: sudo yum install libreoffice python3-uno\n"
+        "   - macOS: brew install libreoffice\n"
+        "   - Windows: 下载并安装LibreOffice\n"
+        "2. 确保Python可以访问uno模块：\n"
+        "   - Linux: export PYTHONPATH=/usr/lib/libreoffice/program:$PYTHONPATH\n"
+        "   - Windows: 添加LibreOffice\\program到系统PATH\n"
+        "3. 验证安装：python -c 'import uno'\n"
+        "4. 如果仍有问题，请查看完整文档：\n"
+        "   https://wiki.documentfoundation.org/Documentation/DevGuide/Installing_the_SDK"
+    )
 
 
 
@@ -40,7 +54,11 @@ class DocxParser(BaseLife):
         else:
             self.use_uno = False
             if use_uno and not HAS_UNO:
-                logger.warning(f"⚠️ UNO不可用，回退到传统命令行方式")
+                logger.warning(
+                    f"⚠️ UNO不可用，回退到传统命令行方式\n"
+                    f"💡 提示：UNO转换更快更稳定，强烈建议安装和配置UNO\n"
+                    f"📖 请参考上述错误信息中的安装指南"
+                )
             else:
                 logger.info(f"🚀 DocxParser初始化完成 - 使用传统命令行方式")
 
@@ -64,7 +82,20 @@ class DocxParser(BaseLife):
                     return txt_path
 
             except Exception as e:
-                logger.error(f"💥 UNO转换失败: {str(e)}")
+                logger.error(
+                    f"💥 UNO转换失败: {str(e)}\n"
+                    f"🔍 诊断信息：\n"
+                    f"   - 错误类型: {type(e).__name__}\n"
+                    f"   - LibreOffice是否已安装？尝试运行: soffice --version\n"
+                    f"   - Python UNO模块是否可用？尝试: python -c 'import uno'\n"
+                    f"   - 是否有其他LibreOffice实例在运行？\n"
+                    f"   - 文件权限是否正确？\n"
+                    f"🔧 可能的解决方案：\n"
+                    f"   1. 确保LibreOffice正确安装\n"
+                    f"   2. 关闭所有LibreOffice进程\n"
+                    f"   3. 检查文件权限和路径\n"
+                    f"   4. 尝试手动运行: soffice --headless --convert-to txt \"{docx_path}\""
+                )
                 logger.warning("⚠️ 自动回退到传统命令行方式...")
                 return self._docx_to_txt_subprocess(docx_path, dir_path)
         else:
@@ -228,26 +259,54 @@ class DocxParser(BaseLife):
             return ""
 
     def _extract_standard_document_content(self, docx_zip: zipfile.ZipFile) -> str:
-        """提取标准document.xml内容"""
+        """提取标准document.xml内容 - 只提取纯文本"""
         try:
             if 'word/document.xml' in docx_zip.namelist():
                 doc_xml = docx_zip.read('word/document.xml').decode('utf-8', errors='replace')
                 
-                # 使用正则表达式提取文本内容
-                import xml.etree.ElementTree as ET
+                # 解码XML实体
+                doc_xml = html.unescape(doc_xml)
+                
+                # 提取所有<w:t>标签中的文本（包括各种命名空间前缀）
+                # 使用更宽松的正则表达式来匹配任何命名空间前缀
+                text_pattern = r'<[^:>]*:t[^>]*>([^<]*)</[^:>]*:t>'
+                text_matches = re.findall(text_pattern, doc_xml)
                 
-                # 移除命名空间前缀以简化处理
-                doc_xml_clean = re.sub(r'xmlns[^=]*="[^"]*"', '', doc_xml)
-                doc_xml_clean = re.sub(r'w:', '', doc_xml_clean)
-                doc_xml_clean = re.sub(r'[a-zA-Z0-9]+:', '', doc_xml_clean)
+                # 额外提取可能存在的无命名空间的<t>标签
+                text_matches.extend(re.findall(r'<t[^>]*>([^<]*)</t>', doc_xml))
                 
-                # 提取所有<t>标签中的文本
-                text_matches = re.findall(r'<t[^>]*>(.*?)</t>', doc_xml_clean, re.DOTALL)
                 if text_matches:
-                    content = ' '.join(text_matches)
-                    content = html.unescape(content)
-                    logger.info(f"📝 从document.xml提取文本: {len(content)} 字符")
-                    return content.strip()
+                    # 清理和组合文本
+                    cleaned_texts = []
+                    for text in text_matches:
+                        # 解码XML实体
+                        text = html.unescape(text)
+                        # 移除多余的空白字符，但保留单个空格
+                        text = re.sub(r'\s+', ' ', text.strip())
+                        if text:
+                            cleaned_texts.append(text)
+                    
+                    # 智能连接文本片段
+                    content = ''
+                    for i, text in enumerate(cleaned_texts):
+                        if i == 0:
+                            content = text
+                        else:
+                            # 如果前一个文本片段不是以标点结束，且当前文本不是以大写开头，则不加空格
+                            prev_char = content[-1] if content else ''
+                            curr_char = text[0] if text else ''
+                            
+                            if prev_char in '.!?。！？\n' or curr_char.isupper() or curr_char in '，。！？；：':
+                                content += ' ' + text
+                            else:
+                                content += text
+                    
+                    # 最终清理
+                    content = re.sub(r'\s+', ' ', content)
+                    content = content.strip()
+                    
+                    logger.info(f"📝 从document.xml提取纯文本: {len(content)} 字符")
+                    return content
             return ""
         except Exception as e:
             logger.error(f"💥 提取标准文档内容失败: {str(e)}")
@@ -271,7 +330,7 @@ class DocxParser(BaseLife):
             return ""
 
     def _extract_headers_footers(self, docx_zip: zipfile.ZipFile) -> str:
-        """提取页眉页脚内容"""
+        """提取页眉页脚内容 - 只提取纯文本"""
         try:
             header_footer_content = []
             
@@ -280,35 +339,66 @@ class DocxParser(BaseLife):
                     logger.debug(f"📄 处理页眉页脚: {filename}")
                     content = docx_zip.read(filename).decode('utf-8', errors='replace')
                     
-                    # 提取文本内容
-                    text_matches = re.findall(r'<w:t[^>]*>(.*?)</w:t>', content, re.DOTALL)
+                    # 解码XML实体
+                    content = html.unescape(content)
+                    
+                    # 提取文本内容 - 使用更宽松的模式
+                    text_pattern = r'<[^:>]*:t[^>]*>([^<]*)</[^:>]*:t>'
+                    text_matches = re.findall(text_pattern, content)
+                    text_matches.extend(re.findall(r'<t[^>]*>([^<]*)</t>', content))
+                    
                     if text_matches:
-                        header_footer_text = ' '.join(text_matches)
-                        header_footer_text = html.unescape(header_footer_text)
-                        if header_footer_text.strip():
-                            header_footer_content.append(header_footer_text.strip())
+                        # 清理和组合文本
+                        cleaned_texts = []
+                        for text in text_matches:
+                            text = html.unescape(text)
+                            text = re.sub(r'\s+', ' ', text.strip())
+                            if text:
+                                cleaned_texts.append(text)
+                        
+                        if cleaned_texts:
+                            # 合并文本片段
+                            header_footer_text = ' '.join(cleaned_texts)
+                            header_footer_text = re.sub(r'\s+', ' ', header_footer_text.strip())
+                            if header_footer_text:
+                                header_footer_content.append(header_footer_text)
             
             if header_footer_content:
-                logger.info(f"📑 提取页眉页脚内容: {len(header_footer_content)} 个")
+                logger.info(f"📑 提取页眉页脚纯文本: {len(header_footer_content)} 个")
             
-            return ' '.join(header_footer_content) if header_footer_content else ""
+            return '\n'.join(header_footer_content) if header_footer_content else ""
         except Exception as e:
             logger.error(f"💥 提取页眉页脚失败: {str(e)}")
             return ""
 
     def _extract_comments(self, docx_zip: zipfile.ZipFile) -> str:
-        """提取注释和批注内容"""
+        """提取注释和批注内容 - 只提取纯文本"""
         try:
             if 'word/comments.xml' in docx_zip.namelist():
                 comments_xml = docx_zip.read('word/comments.xml').decode('utf-8', errors='replace')
                 
-                # 提取注释文本
-                text_matches = re.findall(r'<w:t[^>]*>(.*?)</w:t>', comments_xml, re.DOTALL)
+                # 解码XML实体
+                comments_xml = html.unescape(comments_xml)
+                
+                # 提取注释文本 - 使用更宽松的模式
+                text_pattern = r'<[^:>]*:t[^>]*>([^<]*)</[^:>]*:t>'
+                text_matches = re.findall(text_pattern, comments_xml)
+                text_matches.extend(re.findall(r'<t[^>]*>([^<]*)</t>', comments_xml))
+                
                 if text_matches:
-                    comments_text = ' '.join(text_matches)
-                    comments_text = html.unescape(comments_text)
-                    logger.info(f"💬 提取注释内容: {len(comments_text)} 字符")
-                    return comments_text.strip()
+                    # 清理和组合文本
+                    cleaned_texts = []
+                    for text in text_matches:
+                        text = html.unescape(text)
+                        text = re.sub(r'\s+', ' ', text.strip())
+                        if text:
+                            cleaned_texts.append(text)
+                    
+                    if cleaned_texts:
+                        comments_text = ' '.join(cleaned_texts)
+                        comments_text = re.sub(r'\s+', ' ', comments_text.strip())
+                        logger.info(f"💬 提取注释纯文本: {len(comments_text)} 字符")
+                        return comments_text
             
             return ""
         except Exception as e:
@@ -316,7 +406,7 @@ class DocxParser(BaseLife):
             return ""
 
     def _extract_textbox_content(self, docx_zip: zipfile.ZipFile) -> str:
-        """提取文本框和图形对象中的文本"""
+        """提取文本框和图形对象中的文本 - 只提取纯文本"""
         try:
             textbox_content = []
             
@@ -325,26 +415,43 @@ class DocxParser(BaseLife):
                 if 'word/' in filename and filename.endswith('.xml'):
                     content = docx_zip.read(filename).decode('utf-8', errors='replace')
                     
+                    # 解码XML实体
+                    content = html.unescape(content)
+                    
                     # 查找文本框内容 (w:txbxContent)
-                    textbox_matches = re.findall(r'<w:txbxContent[^>]*>(.*?)</w:txbxContent>', content, re.DOTALL)
+                    textbox_matches = re.findall(r'<[^:>]*:txbxContent[^>]*>(.*?)</[^:>]*:txbxContent>', content, re.DOTALL)
+                    
                     for match in textbox_matches:
-                        text_matches = re.findall(r'<w:t[^>]*>(.*?)</w:t>', match, re.DOTALL)
+                        # 从文本框内容中提取文本
+                        text_pattern = r'<[^:>]*:t[^>]*>([^<]*)</[^:>]*:t>'
+                        text_matches = re.findall(text_pattern, match)
+                        text_matches.extend(re.findall(r'<t[^>]*>([^<]*)</t>', match))
+                        
                         if text_matches:
-                            textbox_text = ' '.join(text_matches)
-                            textbox_text = html.unescape(textbox_text)
-                            if textbox_text.strip():
-                                textbox_content.append(textbox_text.strip())
+                            # 清理和组合文本
+                            cleaned_texts = []
+                            for text in text_matches:
+                                text = html.unescape(text)
+                                text = re.sub(r'\s+', ' ', text.strip())
+                                if text:
+                                    cleaned_texts.append(text)
+                            
+                            if cleaned_texts:
+                                textbox_text = ' '.join(cleaned_texts)
+                                textbox_text = re.sub(r'\s+', ' ', textbox_text.strip())
+                                if textbox_text:
+                                    textbox_content.append(textbox_text)
             
             if textbox_content:
-                logger.info(f"📦 提取文本框内容: {len(textbox_content)} 个")
+                logger.info(f"📦 提取文本框纯文本: {len(textbox_content)} 个")
             
-            return ' '.join(textbox_content) if textbox_content else ""
+            return '\n'.join(textbox_content) if textbox_content else ""
         except Exception as e:
             logger.error(f"💥 提取文本框内容失败: {str(e)}")
             return ""
 
     def _combine_extracted_content(self, content_list: list) -> str:
-        """合并提取到的各种内容"""
+        """合并提取到的各种内容 - 输出清晰的纯文本"""
         combined = []
         
         # 按重要性排序内容
@@ -353,14 +460,40 @@ class DocxParser(BaseLife):
         for content_type in priority_order:
             for item_type, content in content_list:
                 if item_type == content_type and content.strip():
-                    combined.append(content.strip())
+                    # 清理内容中的多余空白
+                    cleaned_content = re.sub(r'\s+', ' ', content.strip())
+                    cleaned_content = re.sub(r'\n\s*\n', '\n\n', cleaned_content)
+                    
+                    if cleaned_content:
+                        # 根据内容类型添加简单的标记（仅在有多种内容类型时）
+                        if len([1 for t, c in content_list if c.strip()]) > 1:
+                            if item_type == "header_footer":
+                                combined.append(f"[页眉页脚]\n{cleaned_content}")
+                            elif item_type == "comments":
+                                combined.append(f"[批注]\n{cleaned_content}")
+                            elif item_type == "textboxes":
+                                combined.append(f"[文本框]\n{cleaned_content}")
+                            else:
+                                combined.append(cleaned_content)
+                        else:
+                            combined.append(cleaned_content)
         
         # 添加其他未分类的内容
         for item_type, content in content_list:
             if item_type not in priority_order and content.strip():
-                combined.append(content.strip())
+                cleaned_content = re.sub(r'\s+', ' ', content.strip())
+                cleaned_content = re.sub(r'\n\s*\n', '\n\n', cleaned_content)
+                if cleaned_content:
+                    combined.append(cleaned_content)
+        
+        # 合并所有内容，使用双换行分隔不同部分
+        final_content = '\n\n'.join(combined) if combined else ""
+        
+        # 最终清理：确保没有过多的空行
+        final_content = re.sub(r'\n{3,}', '\n\n', final_content)
+        final_content = final_content.strip()
         
-        return '\n\n'.join(combined) if combined else ""
+        return final_content
 
     def _extract_html_from_mht(self, mht_content: str) -> str:
         """从MHT内容中提取HTML部分并转换为简洁文本"""

{pydatamax-0.1.15.dist-info → pydatamax-0.1.16.dist-info}/METADATA

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: pydatamax
-Version: 0.1.15
+Version: 0.1.16
 Summary: A library for parsing and converting various file formats.
 Home-page: https://github.com/Hi-Dolphin/datamax
 Author: ccy

{pydatamax-0.1.15.dist-info → pydatamax-0.1.16.dist-info}/RECORD

@@ -5,10 +5,10 @@ datamax/loader/minio_handler.py,sha256=e7ZUlwoStVe5iQfAVfNgEwRLxen4NbxwokpJZl6AR
 datamax/loader/oss_handler.py,sha256=ZO8ZbbA1oyuNN4Z7iVgSbMArYgJ1gvpqADkXDNDw4y0,7648
 datamax/parser/__init__.py,sha256=sIB1N4B_fVguSBN-Uw9tGxAO6s0oi4Tq4kRZ59YlUKo,52
 datamax/parser/base.py,sha256=FamDV6csc3aXVbobMR1lPNtVpvYMO19koRJW9poj_gE,2590
-datamax/parser/core.py,sha256=2Gkz08WrRPt9ga0DisXrV1Aa_Yu7jUMlECOXNaexpwQ,16831
+datamax/parser/core.py,sha256=pySissrF6kVVAzT5abIlQ-4cUliFu1HBWjcD6psNkYA,16845
 datamax/parser/csv_parser.py,sha256=lHQs1MHK9WM4Vl0p9nsE3fFhewF0EoXZUhtk8ixznRw,1028
-datamax/parser/doc_parser.py,sha256=g2vZEdwcA-5AM2y0IHBy0bx1cOflkvLBobQ6tljX1fo,27940
-datamax/parser/docx_parser.py,sha256=fehvMdOexWRRm1HTOCkVa_zDWI8A_LbGJdEFwW-MGss,29613
+datamax/parser/doc_parser.py,sha256=qPKpZy_p1veV2AodqEQU6LzqmT7y1PANlPtt0CYoHeg,30837
+datamax/parser/docx_parser.py,sha256=wdDGgeYIDg1Se493XZhlduxKjtYZ58Uqxltm2vt9Dy4,36691
 datamax/parser/epub_parser.py,sha256=K4eCS4wIXJzDicvtVAfQT8yt1gFHeibZN5-EdQZfJe8,1621
 datamax/parser/html_parser.py,sha256=5ACrVc03Q9pJqWI_b0EtRgOYy0eMYJq4podgHGD68Z8,1453
 datamax/parser/image_parser.py,sha256=UH3duPvB7Xu6CFlEeAukX5uJ8VlqnMR89hcLsW2O-aU,1281
@@ -31,8 +31,8 @@ datamax/utils/ppt_extract.py,sha256=jBVGYEsBGPjHqyq7drHTOM8MnFOwqKyHhbkKmEAryAk,
 datamax/utils/qa_generator.py,sha256=pXxdFm_EnWgMuilfmLKgy2c6NDexQZN8nWxT-bYBt74,12548
 datamax/utils/tokenizer.py,sha256=o78GPmeJ3vs3-SF0b2nMm35XtbrCKbrhDW0gI9gqGl4,880
 datamax/utils/uno_handler.py,sha256=gDm42OQQQoCiOP0SB7xZ9TRF6A_XBHNavwG5ycj6kEQ,14807
-pydatamax-0.1.15.dist-info/licenses/LICENSE,sha256=RltoeMa9c1956S08ThvZE2yJSPbnem68Y8cmiIfOgco,1088
-pydatamax-0.1.15.dist-info/METADATA,sha256=ySaiq1-bWbeW8W5ECuRTSPXzeNxTxaaJEAdqGnWvw0M,9795
-pydatamax-0.1.15.dist-info/WHEEL,sha256=_zCd3N1l69ArxyTb8rzEoP9TpbYXkqRFSNOD5OuxnTs,91
-pydatamax-0.1.15.dist-info/top_level.txt,sha256=N9TrwI3GKnWW07RRXHr0xX5Bm8dIM_sahfAnf9j8J9M,8
-pydatamax-0.1.15.dist-info/RECORD,,
+pydatamax-0.1.16.dist-info/licenses/LICENSE,sha256=RltoeMa9c1956S08ThvZE2yJSPbnem68Y8cmiIfOgco,1088
+pydatamax-0.1.16.dist-info/METADATA,sha256=nwHHLzsxwm_Za3aVbLMFfL-3JIZSZclp9KI0XL6rOHE,9795
+pydatamax-0.1.16.dist-info/WHEEL,sha256=_zCd3N1l69ArxyTb8rzEoP9TpbYXkqRFSNOD5OuxnTs,91
+pydatamax-0.1.16.dist-info/top_level.txt,sha256=N9TrwI3GKnWW07RRXHr0xX5Bm8dIM_sahfAnf9j8J9M,8
+pydatamax-0.1.16.dist-info/RECORD,,