o "iy-@sjdZddlZddlmZddlZddlZddlZddlmZe e Z GdddZ da de fdd ZdS) u_ 文件解析服务 使用 LangChain 的文档加载器支持多种文件格式的文本提取 N)Optional)FileProcessingErrorc @seZdZdZeddededefddZeddededefd d Zeddededefd d Z eddededefddZ ed dededefddZ ed!dededefddZ e d"dedededefddZdS)#FileParserServiceu8文件解析服务类 - 使用 LangChain 文档加载器file.txtcontentfilenamereturnc Cz!dD]}z ||WWStyYqwtdt|dty)tyC}ztd|t|t|d}~ww)uc解析TXT文件 Raises: FileProcessingError: 如果无法解码文件 )utf-8gbkgb2312latin-1z5Unable to decode TXT file with any supported encodingu-无法使用任何支持的编码解码文件zFailed to parse TXT file: NdecodeUnicodeDecodeErrorloggererrorr ExceptionstrrrencodingerJ/lsinfo/ai/hellotax_ai/base_platform/app/services/knowledge/file_parser.py parse_txt     zFileParserService.parse_txtfile.pdfc Cszhddlm}tjddd}|||j}Wdn1s!wYz5||}|}dd|D}d |}t d t |d t |d |Wt j |rZt |WSWSt j |rht |wwtypty} ztd | t|t| d} ~ ww)u解析PDF文件 - 使用 LangChain PyPDFLoader Raises: FileProcessingError: 如果无法解析PDF文件 r) PyPDFLoaderFz.pdfdeletesuffixNcSg|] }|jr|jqSr page_contentstrip.0docrrr =z/FileParserService.parse_pdf.. zSuccessfully parsed PDF, z pages, extracted characterszFailed to parse PDF file: )$langchain_community.document_loadersrtempfileNamedTemporaryFilewritenameloadjoinrinfolenospathexistsunlinkrrrr) rrrtmp_file tmp_file_pathloader documents text_partsresultrrrr parse_pdf(s2       zFileParserService.parse_pdf file.docxc Cs<z{ddl}ddlm}|||}dd|jD}g}|jD]}|jD]}ddd|jD} | r8| | q$q||} d | } | sTt d |t |d | d sad | ddvrnt d|t |dt dt| d| WSt yty} zt d| t |t| d} ~ ww)u解析Word文档 (.docx) - 使用 python-docx 库 Raises: FileProcessingError: 如果无法解析DOCX文件 rN)BytesIOcSr!rtextr$)r&pararrrr(\r)z0FileParserService.parse_docx..z | css$|] }|jr|jVqdS)NrB)r&cellrrr bs"z/FileParserService.parse_docx..r*zExtracted empty content from u文档内容为空u ÐÏࡱádzExtracted binary content from u<提取的内容包含二进制数据，文件可能已损坏z.Successfully parsed .docx document, extracted r+z Failed to parse .docx document: )docxiorADocument paragraphstablesrowsr2cellsappendr$rwarningr startswithrr3r4rr) rrrIrAr'rL tables_texttablerowrow_textall_textr>rrrr parse_docxMs<        zFileParserService.parse_docxfile.docc Csddl}tjddd}|||j}Wdn1swYz'z@|jd|gdddd }|jdkr[|jr[|j}t d t |d |WWt j |rZt |SSt d |jWnty}zt d |WYd}~nd}~wwz>ddlm}t d||} | } dd| D} d| }|rt dt |d |WWt j |rt |SSWnty} zt d| WYd} ~ nd} ~ wwzMt ddD]C} z5|| }|rt |dkrt d| dt |d |WWWt j |rt |SSWqttfy$YqwWntyA}zt d|WYd}~nd}~wwt dt|dt j |rYt |ww)u解析旧版Word文档 (.doc) - 优先使用 antiword Raises: FileProcessingError: 如果无法解析DOC文件 rNFz.docrantiwordT)capture_outputrCtimeoutz2antiword successfully parsed .doc file, extracted r+zantiword parsing failed: zantiword execution failed: )Docx2txtLoaderz#Attempting to use Docx2txtLoader...cSr!rr"r%rrrr(r)z/FileParserService.parse_doc..r*z3Docx2txtLoader successfully parsed file, extracted zDocx2txtLoader also failed: z#Attempting to read as plain text...)r r r r cp1252 zSuccessfully read file with z encoding, extracted z Plain text reading also failed: zAll .doc parsing methods failed. Suggestions: 1) Verify file is standard Word format; 2) Try saving file as .docx format; 3) Open and re-save with another text editoruJ所有解析方法均失败，请验证文件格式或转换为.docx格式) subprocessr-r.r/r0run returncodestdoutr$rr3r4r5r6r7r8rQstderrrr,r^r1r2rrAttributeErrorrr)rrrar9r:r>rCrr^r;r<r= docx_errorr text_errorrrr parse_doc}s   )             zFileParserService.parse_doc file.pptxc Cs zcddlm}tjddd}|||j}Wdn1s!wYz0||}|}dd|D}d |}t d t |d |Wt j |rUt |WSWSt j |rct |wwtykty} ztd | t|t| d} ~ ww) u解析PowerPoint文档 - 使用 LangChain UnstructuredPowerPointLoader Raises: FileProcessingError: 如果无法解析PPTX文件 r)UnstructuredPowerPointLoaderFz.pptxrNcSr!rr"r%rrrr(r)z0FileParserService.parse_pptx..r*z#Successfully parsed PPT, extracted r+zFailed to parse PPT document: )r,rkr-r.r/r0r1r2rr3r4r5r6r7r8rrrr) rrrkr9r:r;r<r=r>rrrr parse_pptxs2      zFileParserService.parse_pptxfile.mdc Cr )up解析Markdown文件 Raises: FileProcessingError: 如果无法解析Markdown文件 )r r r zUnable to decode Markdown fileu5无法使用任何支持的编码解码Markdown文件zFailed to parse Markdown file: Nrrrrrparse_mdrzFileParserService.parse_mdfile file_typec Cs||d}|j|j|j|j|j|j|j|j|jd }||}|r.|||}| ddSt d|t |d|)u 根据文件类型解析文件内容 Args: content: 文件二进制内容 file_type: 文件类型（扩展名，如 'pdf', 'docx'） filename: 文件名（用于错误消息） Returns: 提取的文本内容 Raises: FileProcessingError: 如果文件类型不支持或解析失败 .) txtrCpdfr'rIpptpptxmdmarkdownrGzUnsupported file type: u不支持的文件类型: ) lowerr$rr?rirXrlrngetreplacerrQr)clsrrprparsersparserrCrrr parse_files"   zFileParserService.parse_fileN)r)r)r@)rY)rj)rm)ro)__name__ __module__ __qualname____doc__ staticmethodbytesrrr?rXrirlrn classmethodrrrrrrs $/F$"rrcCstdurtatS)u0获取文件解析服务实例（单例模式）N)_file_parser_servicerrrrrget_file_parser_service+sr)rloggingtypingrrJr-r5app.core.exceptionsr getLoggerrrrrrrrrrs