NgddlZddlZddlZddlZddlmZmZddlmZddl m Z e rddl Z GddeZ Gd d eZ dS) N) BaseOCRAgentBaseOCRElementType)load_dataframe)is_pytesseract_availablecReZdZdZdZdZdZdZdZe dZ e dZ d S) TesseractFeatureTypez7 The element types for Tesseract Detection API rrrc tjdtjdtjdtjdtjdi}||S)Npage_num block_numpar_numline_numword_num)r PAGEBLOCKPARALINEWORD)selfname_cvts \/var/www/html/ai-engine/env/lib/python3.11/site-packages/layoutparser/ocr/tesseract_agent.py attr_namezTesseractFeatureType.attr_name(s? ! %z & %y %z %z  ~c$gd}|d|dzS)N)rrrrrr)rlevelss r group_levelsz!TesseractFeatureType.group_levels3s MMMjqj!!rN) __name__ __module__ __qualname____doc__rrrrrpropertyrr rrrr r sm D E D D D X""X"""rr ceZdZdZdgZddZedZdZ dd Z e d Z e d Z e d Z d S)TesseractAgentz A wrapper for `Tesseract `_ Text Detection APIs based on `PyTesseract `_. pytesseractengc vt|tr|nd||_||_dS)afCreate a Tesseract OCR Agent. Args: languages (:obj:`list` or :obj:`str`, optional): You can specify the language code(s) of the documents to detect to improve accuracy. The supported language and their code can be found on `its github repo `_. It supports two formats: 1) you can pass in the languages code as a string of format like `"eng+fra"`, or 2) you can pack them as a list of strings `["eng", "fra"]`. Defaults to 'eng'. +N) isinstancestrjoinlangconfigs)r languageskwargss r__init__zTesseractAgent.__init__As5",Is!;!;TII)ATAT  rc 4|tj_|di|S)Nr)r( tesseract_cmd)clstesseract_cmd_pathr2s rwith_tesseract_executablez(TesseractAgent.with_tesseract_executableQs"1C -s}}V}}rci}tj|fd|ji|j|d<tj|fd|ji|j}t jtj|tj dddti|d<|S)Nr/textzutf-8 )quotingencodingsep convertersdata) r(image_to_stringr/r0 image_to_datapdread_csvioStringIOcsv QUOTE_NONEr-)r img_contentres_datas r_detectzTesseractAgent._detectWs!1   "i +/<  F )+VVDIVVVk K  N}    F  rFTNc||}|r|S|r|dS||||S|dS)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 all output (string and boxes info) from Tesseract. 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:`~TesseractFeatureType`, optional): When set, aggregate the GCV output with respect to the specified aggregation level. Defaults to `None`. r:)rL gather_data)rimagereturn_responsereturn_only_textagg_output_levelrJs rdetectzTesseractAgent.detectfs[(ll5!!  J  v;   '##C)9:: :6{rc t|ts Jd||d}||j|jdddddd d d d d  ddd dd g}t|S)zo Gather the OCR'ed text, bounding boxes, and confidence in a given aggeragation level. zInvalid agg_level r@c dtj|d|d|d|d|d|djdgS) Nlefttopwidthheightconfr: )r>)rCSeriesminmaxmeanr-cat)gps rz,TesseractAgent.gather_data..s296 ((5 7 ))8 ((**6 ))6 **s*33   rT)dropx_1y_1whscorer:id)rrrr r index)columnsc |j|jzSN)rdrfxs rrbz,TesseractAgent.gather_data..aeackrc |j|jzSrn)rergros rrbz,TesseractAgent.gather_data..rqr rectangle)x_2y_2 block_type) r,r r:isnagroupbyr apply reset_indexrenameassignrcr)response agg_levelrJdfs rrNzTesseractAgent.gather_datas#  +   , , + + + , , ,v  ! WY+ , , U    [d[ # # []] V!  V))))& T3*T % %C Hb!!!rc~t|d5}tj|}dddn #1swxYwY|S)Nrb)openpickleload)filenamefprJs r load_responsezTesseractAgent.load_responsesv (D ! ! "R+b//C " " " " " " " " " " " " " " " s 266ct|d5}tj||tjddddS#1swxYwYdS)Nwb)protocol)rrdumpHIGHEST_PROTOCOL)rJ file_namers r save_responsezTesseractAgent.save_responses)T " " Cb KR&*A B B B B C C C C C C C C C C C C C C C C C Cs"AAA)r))FTN)r!r"r#r$ DEPENDENCIESr3 classmethodr8rLrS staticmethodrNrrrrrr'r'9s "?L [    UYB-"-"\-"^\ CC\CCCrr')rErGrpandasrCbaserrr file_utilsrr(r r'rrrrs 22222222111111"""""-"""8HCHCHCHCHC\HCHCHCHCHCr