kh&ddlZddlZddlZddlZddlZddlmZddlm Z m Z ddl m Z m Z mZm Z ddlmZerddlmcmZddlmcmZdZGd d e ZGd d e Zy) N)imencode) BaseOCRAgentBaseOCRElementType)Layout TextBlock Quadrilateralr )is_gcv_availablec~tj|Dcgc]}|j|jgc}Scc}wN)nparrayxy)verticesvertexs V/var/www/teggl/fontify/venv/lib/python3.12/site-packages/layoutparser/ocr/gcv_agent.py_cvt_GCV_vertices_to_pointsr!s, 88Affhh)A BBAs:cDeZdZdZdZdZdZdZdZe dZ e dZ y ) GCVFeatureTypez8 The element types from Google Cloud Vision API rrrc tjdtjdtjdtjdtj di}||S)Npagesblocks paragraphswordssymbolsrPAGEBLOCKPARAWORDSYMBOL)selfname_cvts r attr_namezGCVFeatureType.attr_name0sN     (        ! !9  ~c  tjtjtjtjtjtjtjtj tj di}||Sr r )r& child_cvts r child_levelzGCVFeatureType.child_level;sf   !5!5  ."5"5   !4!4   !6!6  ! !4  r)N) __name__ __module__ __qualname____doc__r!r"r#r$r%propertyr(r,r)rrr%sH D E D D F r)rcneZdZdZdgZd dZedZdZ d dZ e dZ e d Z d Z d Zy)GCVAgentaJA wrapper for `Google Cloud Vision (GCV) `_ Text Detection APIs. Note: Google Cloud Vision API returns the output text in two types: * `text_annotations`: In this format, GCV automatically find the best aggregation level for the text, and return the results in a list. We use :obj:`~gather_text_annotations` to reterive this type of information. * `full_text_annotation`: To support better user control, GCV also provides the `full_text_annotation` output, where it returns the hierarchical structure of the output text. To process this output, we provide the :obj:`~gather_full_text_annotation` function to aggregate the texts of the given aggregation level. zgoogle-cloud-visionNc tj|_tj j ||_||_y#tjdYFxYw)aCreate a Google Cloud Vision OCR Agent. Args: languages (:obj:`list`, optional): You can specify the language code of the documents to detect to improve accuracy. The supported language and their code can be found on `this page `_. Defaults to None. ocr_image_decode_type (:obj:`str`, optional): The format to convert the input image to before sending for GCV OCR. Defaults to `".png"`. * `".png"` is suggested as it does not compress the image. * But `".jpg"` could also be a good choice if the input image is very large. zJThe GCV credential has not been set. You could not run the detect command.)language_hintsN) _visionImageAnnotatorClient_clientwarningswarntypes ImageContext_contextocr_image_decode_type)r& languagesr?s r__init__zGCVAgent.__init__`sP" "779DL  22)2L %:"   MM\ s AA!c 8|tjd<|di|S)zSpecifiy the credential to use for the GCV OCR API. Args: credential_path (:obj:`str`): The path to the credential file GOOGLE_APPLICATION_CREDENTIALSr2)osenviron)clscredential_pathkwargss rwith_credentialzGCVAgent.with_credentialzs 8G 34}V}r)ctjj|}|jj ||j }|S)N)content)image image_context)r7r<Imager9document_text_detectionr>)r& img_contentresponses r_detectzGCVAgent._detectsBmm))+)> <<77T]]8 r)ct|tjr(t|j|dj }n?t|t r/tj|d5}|j}ddd|j}|r|S|r|jjS||j||S|j|S#1swY[xYw)aSend the input image for OCR. Args: image (:obj:`np.ndarray` or :obj:`str`): The input image array or the name of the image file return_response (:obj:`bool`, optional): Whether directly return the google cloud response. Defaults to `False`. return_only_text (:obj:`bool`, optional): Whether return only the texts in the OCR results. Defaults to `False`. agg_output_level (:obj:`~GCVFeatureType`, optional): When set, aggregate the GCV output with respect to the specified aggregation level. Defaults to `None`. rrbN) isinstancerndarrayrr?tostringstrioopenreadrRfull_text_annotationtextgather_full_text_annotationgather_text_annotations)r&rLreturn_responsereturn_only_textagg_output_levelrP image_fileress rdetectzGCVAgent.detects, eRZZ ("4#=#=uEaHQQSK s #% 0(oo/  0ll;' J ++00 0  '33C9IJ J++C00 0 0s )CCc|jdd}t}t|D]T\}}t|jj }|j tt||j|V|S)a9Convert the text_annotations from GCV output to an :obj:`Layout` object. Args: response (:obj:`AnnotateImageResponse`): The returned Google Cloud Vision AnnotateImageResponse object. Returns: :obj:`Layout`: The reterived layout from the response. rN)blockr]id) text_annotationsr enumerater bounding_polyrappendr r description)rQdoc gathered_texti text_comppointss rr_z GCVAgent.gather_text_annotationssx''+ %cN LAy01H1H1Q1QRF   f 5I.iter_levels i&9&9: $ )E 5 55LL+!4!44!^%:%::EieY=R=RLL0ieY=R=R )89J9J9S9STF!*+F3 XXe_"oo% "J &&z2!OJA $r)r)rgr]) rr!rirrkrr r rmr\r)rQr}rnrrr~rrs @@rr^z$GCVAgent.gather_full_text_annotations"$)) ( $T ++ +++A.C01B1B1K1KLF$=+@sWXKk"" //CKJ sI{ 3k""r)ct|d5}|j}dddtjtj j dS#1swY=xYw)NrT)ignore_unknown_fields)rZr[ _json_formatParser7r<AnnotateImageResponse)r&filenamefdatas r load_responsezGCVAgent.load_responsesU (C  A668D !! '--557t    s AA#ctj|}t|d5}tj|}tj ||dddy#1swYyxYw)Nw)r MessageToJsonrZjsonloadsdump)r&rd file_namer json_files r save_responsezGCVAgent.save_responsesN((- )S ! $Q 3I IIi # $ $ $s ,AA )Nz.png)FFN)r-r.r/r0 DEPENDENCIESrA classmethodrIrRre staticmethodr_r^rrr2r)rr4r4Gsu,**L;4 (1T.D#D#L $r)r4)rYrDrr:numpyrcv2rbaserrelementsrr r file_utilsr google.protobuf.json_formatprotobuf json_formatrgoogle.cloud.visioncloudvisionr7rrr4r2r)rrsY 2BB)66))C'DY$|Y$r)