j5jOFdZddlmZddlZddlZddlZddlmZddlm Z ddl m Z m Z m Z mZddlmZddlmZdd lmZdd lmZdd lmZdd lmZmZdd lmZmZddlm Z m!Z!ddl"m#Z#ddl$m%Z%ejLe'Z(ddZ)Gdde Z*edZ+edZ,edZ-edZ.eddZ/Gdde#Z0edZ1y)z1Built-in plugin to implement OCR using Tesseract.) annotationsN) Annotated)Image) BaseModelFieldfield_validatormodel_validator)hookimpl tesseract)ThresholdingMethod) PageContext)numeric) BadArgsErrorMissingDependencyError)available_cpu_countclamp)calculate_downsampledownsample_image) OcrEngine)check_external_programc@tjtjtjtjd}|j |vr;ddl}dj|j}|jd|d|||j S)a+Convert string argument to ThresholdingMethod enum. Args: value: String name of thresholding method (auto, otsu, adaptive-otsu, sauvola) Returns: ThresholdingMethod enum value Raises: argparse.ArgumentTypeError: If value is not a valid thresholding method )autootsuz adaptive-otsusauvolarN, zInvalid thresholding method 'z'. Must be one of: ) r AUTOOTSU ADAPTIVE_OTSUSAUVOLAlowerargparsejoinkeysArgumentTypeError)value method_mapr"valids \/var/www/html/qr/venv/lib/python3.12/site-packages/ocrmypdf/builtin_plugins/tesseract_ocr.py_thresholding_method_converterr*s#''"''+99%-- J  {{}J& *//+,(h((+E72EeW M   ekkm $$cHeZdZUdZgZded<dZded<dZded<ejZ d ed <d Z d ed <d Z ded<dZ ded<dZded<dZded<dZded<dZded<ed#d$dZed dedZededZed d!Zd%d"Zy)&TesseractOptionsz)Options specific to Tesseract OCR engine.zSAnnotated[list[str], Field(description='Additional Tesseract configuration files')]configNz]Annotated[int | None, Field(ge=0, le=13, description='Set Tesseract page segmentation mode')] pagesegmodezUAnnotated[int | None, Field(ge=0, le=3, description='Set Tesseract OCR engine mode')]oemz_Annotated[ThresholdingMethod, Field(description='Set Tesseract input image thresholding mode')] thresholdingf@zRAnnotated[float, Field(ge=0, description='Timeout for OCR operations in seconds')]timeoutzVAnnotated[float, Field(ge=0, description='Timeout for non-OCR operations in seconds')]non_ocr_timeoutTzHAnnotated[bool, Field(description='Downsample large images before OCR')]downsample_large_imageszdAnnotated[int, Field(ge=100, le=32767, description='Downsample images larger than this pixel size')]downsample_abovezMAnnotated[str | None, Field(description='Path to Tesseract user words file')] user_wordszPAnnotated[str | None, Field(description='Path to Tesseract user patterns file')] user_patternszpAnnotated[int | None, Field(description='Calculated OMP_THREAD_LIMIT for Tesseract subprocesses', exclude=True)]omp_thread_limitc 0|jdd}|jd|dddg|dd |jd|d d td td d|dd|jd|dd tdtd d|dd|jd|dd td|dd|jd|ddt t d d|dd |d!"|jd|d#dt t d d|d$d%|d&"|jd|d'tjd(|d)d*|d+,|jd|d-d t td.d/d/|d0d1|d2|jd3d4d5d67|jd8d4d9d:7y;)Additional Tesseract configuration files -- see documentation.)actionmetavardefaultdesthelpz -pagesegmodestorePSMr _pagesegmodez-timeout because these operations are not as expensive as OCR.z-downsample-large-imagesT_downsample_large_imagesaDownsample large images before OCR. Tesseract has an upper limit on the size images it will support. If this argument is given, OCRmyPDF will downsample large images to fit Tesseract. This may reduce OCR quality, on large images the most desirable text is usually larger. If this parameter is not supplied, Tesseract will error out and produce no OCR on the page in question. This argument should be used with a high value of --z0-timeout to ensure Tesseract has enough to time.)r@rBrCrDz-downsample-abovedr6_downsample_abovezTDownsample images larger than this size pixel size (either dimension) before OCR. --a-downsample-large-images downsamples when an image exceeds Tesseract's internal limits. This argument causes downsampling to occur when an image exceeds the given size. This may reduce OCR quality, but on large images the most desirable text is usually larger.z --user-wordsFILEr8aSpecify the location of the Tesseract user words file. This is a list of words Tesseract should consider while performing OCR in addition to its standard language dictionaries. This can improve OCR quality especially for specialized and technical documents.)rArCrDz--user-patternsr9z9Specify the location of the Tesseract user patterns file.N) add_argument_group add_argumentintranger*rfloatr"BooleanOptionalAction)clsparser namespacetesss r)add_arguments_to_parserz(TesseractOptions.add_arguments_to_parseris(( <   7 #;g&Q    < (!RL;l+O    4 !QK;d#   = )/;m,O   8 $";h'K0 2    + ,";./DDM;OHH    3 411;67 "{#& &  (  , -c5);/0!!* ,""   N     L  r+cB|dkDrtjd|d|S)z'Validate timeout values are reasonable.iz Timeout of z* seconds is very long and may cause issueslogwarningr\vs r)validate_timeout_reasonablez,TesseractOptions.validate_timeout_reasonables& t8 KK+aS(RS Tr+c8|dvrtjd|S)zBValidate page segmentation mode and warn about problematic values.)rz[The tesseract-pagesegmode you selected will disable OCR. This may cause processing to fail.rbres r)validate_pagesegmode_warningz-TesseractOptions.validate_pagesegmode_warnings# ; KK5 r+after)modecf|jdk7r!|jstjd|S)z+Validate downsample options are consistent.r6wThe --tesseract-downsample-above argument will have no effect unless --tesseract-downsample-large-images is also given.)r7r5rcrdselfs r)validate_downsample_consistencyz0TesseractOptions.validate_downsample_consistency s3  E )$2N2N KKE  r+c |ddh}|t|zr*tddj|t|zdy)zValidate options that require external context. Args: languages: List of languages being used for OCR equosdz^The following languages are for Tesseract's internal use and should not be issued explicitly: rz- Remove them from the -l/--language argument.N)setrr#)rp languagesDENIED_LANGUAGESs r)validate_with_contextz&TesseractOptions.validate_with_contextsU"5> c)n ,899-I>?@A??  -r+r )r^str)rvz list[str]returnNone)__name__ __module__ __qualname____doc__r.__annotations__r/r0r rr1r3r4r5r7r8r9r: classmethodr`rrgrjr rqrxr+r)r-r-9sa3                  L L \Y 123 ]#$'"#r+r-cdtiS)z Register Tesseract option model.r )r-rr+r)register_optionsr&s ) **r+c.tj|yN)r-r`)r]s r) add_optionsr,s,,V4r+crtddditjdtjtj}|tjdk(r t dtj s>|jj tjk7rtjdy y y ) zDCheck external dependencies and version compatibility for Tesseract.r linuxz tesseract-ocrz4.1.1)programpackageversion_checker need_versionversion_parserz5.4.0zzTesseract 5.4.0 is not supported due to regressions in this version. Please upgrade to a newer or supported older version.zThe installed version of Tesseract does not support changes to its thresholding method. The --tesseract-threshold argument will be ignored.N) rr versionTesseractVersionrhas_thresholdingr1r rrcrd)options tess_versions r) check_optionsr2s/*!)) 11 $$&Ly11'::$ D    & & (    * *.@.E.E E    F )r+ctjjddjs2|jxs t }t |t|zdd}nttjd}||j_ tjd||jjdk7r-|jjstjdyyy)NOMP_THREAD_LIMITz&Using Tesseract OpenMP thread limit %dr6rn)osenvironget isnumericjobsrrlenrXr r:rcdebugr7r5rd)pdfinforr tess_threadss r)validaterOs ::>>,b 1 ; ; =||424TS\11a8 2::&89: )5G&II6 E **e3!!99  A : 4r+c|j}t|dd|St|jjd}|jj rt |||fd}t||}|S)zFilter the image before OCR. Tesseract cannot handle images with more than 32767 pixels in either axis, or more than 2**31 bytes. This function resizes the image to fit within those limits. r Nr6i)max_size max_bytes)rgetattrminr r7r5rr)pageimager thresholdsizes r)filter_ocr_imagerlsqllGw T*2 G%%66>I00# Y 2k !- Lr+ceZdZdZedZedZedZdZedZ edZ ed dZ ed Z ed Z y ) TesseractOcrEnginezImplements OCR with Tesseract.c<ttjSr)ryr rrr+r)rzTesseractOcrEngine.versions9$$&''r+c:|jdk(ry|jS)zADetermine the PDF renderer to use based on options and languages.rfpdf2) pdf_rendererrs r)_determine_rendererz&TesseractOcrEngine._determine_renderers!   6 )###r+ctj|}|xdk(rdtjSxdk(rdtjSdk(rdtjS dtjS)NhocrzOCRmyPDF hOCR + Tesseract OCR rzOCRmyPDF fpdf2 + Tesseract OCR sandwichzTesseract OCR + PDF Tesseract OCR )rrr)rrenderers r) creator_tagzTesseractOcrEngine.creator_tags%99'B78J8R8R8T7UVV89K9S9S9U8VWW-.@.H.H.J-KLL'(:(B(B(D'EFFr+c0dtjS)Nr)rrros r)__str__zTesseractOcrEngine.__str__s 2 : : <=>>r+c*tjSr)r get_languagesrs r)rvzTesseractOcrEngine.languagess&&((r+ctj||jj|jj|jjS)N) engine_moder3r:)r get_orientationr0r4r: input_filers r)rz"TesseractOcrEngine.get_orientationsE(( ))--%%55$..??   r+ctj||j|jj|jj|jj S)N)rvrr3r:)r get_deskewrvr0r4r:rs r)rzTesseractOcrEngine.get_deskewsN## ''))--%%55$..??   r+ctj||||j|jj|jj|jj |jj |jj|jj|jj|jj y)N) r output_hocr output_textrvr tessconfigr3r/r1r8r9r:) r generate_hocrrvr0r.r3r/r1r8r9r:)rrrrs r)rz TesseractOcrEngine.generate_hocrs!##''))--((//%%--))55 **77((33!++99$..?? r+ctj||||j|jj|jj|jj |jj |jj|jj|jj|jj y)N) r output_pdfrrvrrr3r/r1r8r9r:) r generate_pdfrvr0r.r3r/r1r8r9r:)rrrrs r)rzTesseractOcrEngine.generate_pdfs!!#''))--((//%%--))55 **77((33!++99$..?? r+N)rzrZ)r|r}r~r staticmethodrrrrrvrrrrrr+r)rrs((($$  G G?))        r+rc>|t|dd}|dvrytS)z6Return TesseractOcrEngine when selected or as default.N ocr_enginer)rr )rr)rrs r)get_ocr_enginers-WlF; 2 2  r+)r&ryrzr )rrr Image.Imagerzr)2r __future__rr"loggingrtypingrPILrpydanticrrrr ocrmypdfr ocrmypdf._execr ocrmypdf._exec.tesseractr ocrmypdf._jobcontextr ocrmypdf.clirocrmypdf.exceptionsrrocrmypdf.helpersrrocrmypdf.imageopsrrocrmypdf.pluginspecrocrmypdf.subprocessr getLoggerr|rcr*r-rrrrrrrrr+r)rs8" GG$7, D7D)6g!%8jyjZ + +  5 5     8    8  (U U p    r+