uPjJUdZddlmZmZddlZddlmZddlmZm Z m Z m Z ddl m Z erddlmZdZeed <Gd d eZy) z, Mistral OCR transformation implementation. ) TYPE_CHECKINGFinalN)verbose_logger) BaseOCRConfig DocumentTypeOCRRequestData OCRResponse)get_secret_str)LoggingMISTRAL_API_KEYMISTRAL_OCR_API_KEY_ENV_VARc eZdZdZdfd ZdedefdZdedzfdZde d e dede fd Z dd e ded edzd edzde dzde f dZ dd edzded e de dzdef dZ dede d e d e def dZdedej"dddefdZxZS)MistralOCRConfigzh Mistral OCR transformation configuration. Reference: https://docs.mistral.ai/api/#tag/ocr returnNc"t|yN)super__init__)self __class__s l/home/oem/CyberBrain26-unified/.venv/lib/python3.12/site-packages/litellm/llms/mistral/ocr/transformation.pyrzMistralOCRConfig.__init__s modelc gdS)a Get supported OCR parameters for Mistral OCR. Mistral OCR supports: - pages: List of page numbers to process - include_image_base64: Whether to include base64 encoded images - image_limit: Maximum number of images to return - image_min_size: Minimum size of images to include - bbox_annotation_format: Format for bounding box annotations - document_annotation_format: Format for document annotations - document_annotation_prompt: Prompt for document annotation extraction - extract_header: Whether to extract document header - extract_footer: Whether to extract document footer - table_format: Table output format ("markdown" or "html") - confidence_scores_granularity: Confidence score level ("word" or "page") - include_blocks: Whether to return paragraph-level bounding boxes and typed content blocks (OCR 4) - id: Request identifier ) pagesinclude_image_base64 image_limitimage_min_sizebbox_annotation_formatdocument_annotation_formatdocument_annotation_promptextract_headerextract_footer table_formatconfidence_scores_granularityinclude_blocksid)rrs rget_supported_ocr_paramsz)MistralOCRConfig.get_supported_ocr_params"s &  rctSr)r )rs rget_api_key_env_varz$MistralOCRConfig.get_api_key_env_varEs**rnon_default_paramsoptional_paramscr|j|}i}|jD]\}}||vs |||<|S)z Map OCR parameters to Mistral-specific format. Mistral accepts these parameters directly, so no transformation needed. Just filter out unsupported params. )r)r)items)rr,r-rsupported_params mapped_paramsparamvalues rmap_ocr_paramszMistralOCRConfig.map_ocr_paramsHsV#'"?"?e"?"L " .446 -LE5((', e$ -rheadersapi_keyapi_baselitellm_paramsc T|tt}| tddd|i|}|S)zJ Validate environment and return headers for Mistral OCR. z}Missing Mistral API Key - A call is being made to Mistral but no key is set either in the environment variables or via params AuthorizationzBearer )r r ValueError)rr5rr6r7r8kwargss rvalidate_environmentz%MistralOCRConfig.validate_environment^sP ?$%@AG ?P  wwi0  rc b|d}|jd}|jdr|dS|dS)zl Get complete URL for Mistral OCR endpoint. Returns: https://api.mistral.ai/v1/ocr zhttps://api.mistral.ai/v1/z/v1z/ocrz/v1/ocr)rstripendswith)rr7rr-r8r<s rget_complete_urlz!MistralOCRConfig.get_complete_url|sI  2H??3'   U #Zt$ $7##rdocumentc tjd|t|tst dt |||d}|j |t|dS)a Transform OCR request to Mistral-specific format. Mistral OCR API accepts: { "model": "mistral-ocr-latest", "document": { "type": "document_url", "document_url": "" }, "pages": [0], # optional "include_image_base64": false, # optional ... } Args: model: Model name (e.g., "mistral-ocr-latest") document: Document dict from user (Mistral format) - already validated in main.py optional_params: Already mapped optional parameters headers: Request headers Returns: OCRRequestData with JSON data z-Mistral OCR transform_ocr_request - model: %szExpected document dict, got )rrCN)datafiles)rdebug isinstancedictr;typeupdater)rrrCr-r5r<rEs rtransform_ocr_requestz&MistralOCRConfig.transform_ocr_requestsf@ LeT(D);DN;KLM M   O$4t44r raw_response logging_objLiteLLMLoggingObjc Z |j}tjd|jt |j dg|j d||j d|j ddS#t $r}tjd||d }~wwxYw) av Return Mistral OCR response in native format. Mistral OCR is the standard format for LiteLLM OCR responses. No transformation needed - return native response. Mistral OCR returns: { "pages": [ { "index": 0, "markdown": "extracted text content", "images": [...], "dimensions": {...} }, ... ], "model": "mistral-ocr-2505-completion", "document_annotation": null, "usage_info": {...} } zMistral OCR response keys: %srrdocument_annotation usage_infoocr)rrrQrRobjectz&Error parsing Mistral OCR response: %sN)jsonrrGkeysr get Exceptionerror)rrrMrNr< response_jsones rtransform_ocr_responsez'MistralOCRConfig.transform_ocr_responses: #/#4#4#6M  !@-BTBTBV W#''4#''7$1$5$56K$L(,,\:     !I1 MG sBB B* B%%B*)rN)NNNr)__name__ __module__ __qualname____doc__rstrlistr)r+rIr4r=rBrrrLhttpxResponser r\ __classcell__)rs@rrrsZ ! c! d! F+S4Z+    4##&* t  *  t  F'+ $*$$ $ t $ $2151515 15  15 15f,,nn,) ,  ,rr)r`typingrrrclitellm._loggingr(litellm.llms.base_llm.ocr.transformationrrrr litellm.secret_managers.mainr *litellm.litellm_core_utils.litellm_loggingr rOr __annotations__rr(rrrlsC( + 8W%6U6\}\r