j5jJddlmZddlZddlZddlZddlZddlZddlZddlZddl m Z m Z ddl m Z ddlmZddlmZddlmZddlmZdd lmZmZmZerdd lmZddlZddlZdd lmZdd l m!Z!dd l"m#Z#m$Z$ddl%m&Z&m'Z'ddl(m)Z)ddl*m+Z+ddl,m-Z-ddl.m/Z/m0Z0m1Z1m2Z2m3Z3m4Z4m5Z5m6Z6m7Z7m8Z8m9Z9m:Z:m;Z;mZ>m?Z?m@Z@ddlAmBZBddlCmDZDddlEmFZFmGZGddlHmIZImJZJmKZKmLZLmMZMddlNmOZOddlPmQZQejeSZTejZVdeV_WdZXeXeVd0dZYGddeZZGddejZ\Gdd ejZ^eGd!d"Z_ d1 d2d#Z`d3d$Zae d4d%Zbd5d&Zced6d'Zd d7d(Ze d8d)Zfd9d*Zg d:d+Zh d;d,Zidd/Zly)?) annotationsN)CallableSequence)BrokenProcessPool)BrokenThreadPool)contextmanager) dataclassPath) TYPE_CHECKING NamedTuplecast) OcrElement)Pdf)remove_broken_goto_annotations)Executorsetup_executor) PageContext PdfContext)PageNumberFilter)metadata_fixup) OcrOptions)convert_to_pdfacreate_ocr_imagecreate_pdf_page_from_imagecreate_visible_page_jpggenerate_postscript_stubget_orientation_correctionget_pdf_save_settings get_pdfinfo optimize_pdfpreprocess_cleanpreprocess_deskewpreprocess_remove_background rasterizerasterize_previewshould_linearize!should_visible_page_image_use_jpg try_auto_pdfatry_speculative_pdfa)OcrmypdfPluginManager)report_output_file_size)ExitCodeExitCodeException) check_pdfpikepdf_enable_mmaprunning_in_dockerrunning_in_snapsamefile)file_claims_pdfa)PdfInfocftjfd}tj|y)z=Inject current page number (when available) into log records.cR|i|}tdrj|_|S)Npageno)hasattrr8)argskwargsrecord old_factorytlss Q/var/www/html/qr/venv/lib/python3.12/site-packages/ocrmypdf/_pipelines/_common.pywrapperz!_set_logging_tls..wrapperNs,d-f- 3 !JJFM N)logginggetLogRecordFactorysetLogRecordFactory)r>r@r=s` @r?_set_logging_tlsrEJs&--/K  (rAc|t_y)z@Set page number (1-based) that the current thread is processing.N)r>r8)r8s r?set_thread_pagenorGZs CJrAcleZdZUdZded< dZded< dZded< dZded< d Zded < dZ d ed <y) PageResultz*Result when a page is finished processing.intr8N Path | Nonepdf_page_from_imageocrtextrorientation_correctionOcrElement | Noneocr_tree) __name__ __module__ __qualname____doc____annotations__rLrMrNrOrQrAr?rIrI_sQ4 K'++%CD+ "#C#,"&H&ArArIceZdZfdZxZS)HOCRResultEncoderc\t|tr dt|iSt||SNr ) isinstancer strsuperdefault)selfobj __class__s r?r_zHOCRResultEncoder.defaultvs+ c4 CH% %ws##rA)rRrSrTr_ __classcell__rbs@r?rYrYus $$rArYc$eZdZfdZdZxZS)HOCRResultDecodercB|j|d<t||i|y)N object_hook)dict_to_objectr^__init__)r`r:r;rbs r?rjzHOCRResultDecoder.__init__}s% $ 3 3} $)&)rAc*d|vrt|dS|Sr[r )r`ds r?riz HOCRResultDecoder.dict_to_objects Q;& ? "rA)rRrSrTrjrircrds@r?rfrf|s *rArfceZdZUdZded< dZded< dZded< dZded< d Zded < dZ d ed < e dd Z ddZ y) HOCRResultz(Result when hOCR is finished processing.rJr8NrKrLhocrtextpdfrrOrPrQcD|ditj|tS)zCreate an instance from a dict.clsrW)jsonloadsrf)rsjson_strs r? from_jsonzHOCRResult.from_jsonsATZZ.?@AArAcLtj|jtS)zSerialize to a JSON string.rr)rtdumps__dict__rY)r`s r?to_jsonzHOCRResult.to_jsonszz$--->??rA)rvr]returnrn)r|r]) rRrSrTrUrVrLrorprOrQ classmethodrwr{rWrAr?rnrnsk2 K'++%D+ G[,"#C#,"&H&ABB@rArncLtj|djtjtjd}j |j ttjjfd}|fS)aCreate a debug log file at a specified location. Returns the log handler, and a function to remove the handler. Args: log_filename: Where to the put the log file. prefix: The logging domain prefix that should be sent to the log. T)delayzA[%(asctime)s] - %(name)s - %(levelname)7s -%(pageno)s %(message)sc tjjjy#t$r%}t |t jYd}~yd}~wwxYw)Nfile)rB getLogger removeHandlercloseOSErrorprintsysstderr)elog_file_handlerprefixs r?removerz(configure_debug_logging..removersO &   f % 3 34D E  " " $ & !#** % % &s48 A&A!!A&) rB FileHandlersetLevelDEBUG Formatter setFormatter addFilterrr addHandler) log_filenamer formatterrrs ` @r?configure_debug_loggingrs** + %b' #G  I 7 I sAA2A# A2# A//A2cd|g}tr|jdntr|jdtdj |t j y)z0Print the location of the temporary work folder.z%Temporary working files retained at: zUOCRmyPDF is running in a Docker container, so the files will be inside the container.zSOCRmyPDF is running in a Snap container, so the files will be inside the container. rN)r1appendr2rjoinrr)rmsgss r?_print_temp_folder_locationrsY4[M B CD  9     9  $))D/ +rAc#K ||r|r$t|ytj|dyy#|r|r t|wwtj|dwxYww)NT) ignore_errors)rshutilrmtree)rretainprint_locations r?manage_work_folderrsX; +K8 MM+T : +K8 MM+T :sA1)A*AAc |||S#t$rM|jdk\rtjdntj dt j cYSt$r}tt|}|jdk\rtjdngt|r4tj dt|jt|n(tj t|j|jcYd}~Sd}~wt$r}dt|jvsdt|jvrY|jdk\rtjdntj d t|t jcYd}~Sd}~wt j"j$$r(tjd t j&cYSt(t*f$r(tjd t j,cYSt.$r(tjd t j&cYSwxYw) aYConvert exceptions into command line error messages and exit codes. When known exceptions are raised, the exception message is printed to stderr and the program exits with a non-zero exit code. When unknown exceptions are raised, the exception traceback is printed to stderr and the program exits with a non-zero exit code. rKeyboardInterruptr.z%s: %sNzvalidation errorz value errorzValidation errorzInvalid argument: %szA decompression bomb error was encountered while executing the pipeline. Use the argument --max-image-mpixels to raise the maximum image pixel limit.zA worker process was terminated unexpectedly. This is known to occur if processing your file takes all available swap space and RAM. It may help to try again with a smaller number of jobs, using the --jobs argument.z2An exception occurred while executing the pipeline)rrlog exceptionerrorr-ctrl_cr.rr]typerR exit_code ValueErrorlowerbad_argsrrDecompressionBombError other_errorrrchild_process_error Exception)fnrplugin_managerrs r?cli_exception_handlerrs2$'>**  ??a  MM- . II) *  "A & ??a  MM- . V IIhQ 0 0#a& 9 IId1g&& '{{   Q /=CFLLN3R!# 01 0#a&9$$ $ 99 + +$  ! ###  ,   +++ $ JK###$sO AI I(B'DI I!B F1*I0F11AI83I--IIc|j]t|jdztj_tjjdk(rdtj_t t |}|S)Ni@Br)max_image_mpixelsrJrrrr0r)rrexecutors r?setup_pipelinerBs_  ,%()B)BY)N%O " 99 % % *)-CII &n-H OrAc Z|j}t|trJddlm}t j |5}t|j}ddd||}||_t|||j|j|j|j|S#1swYTxYw)Nr)_pages_from_ranges) total_pages)rdetailed_analysisprogbar max_workers use_threads check_pages) pagesr\r]ocrmypdf._optionsrropenlenr redo_ocr progress_barjobsr)pdf_pathrrrrpdfrs r?do_get_pdfinforYs--K+s#8 XXh  )3cii.K )(+N # !**$$LL''   ) )s B!!B*cZ|r t||}|r t||}|r t||}|S)zPreprocess an image.)r$r#r") page_contextimageremove_backgrounddeskewcleans r? preprocessrqs6,ULA !%6   5 LrAcv|j}dx}}t|j||d}t|j|j |j gs*t|||j|jdx}}||fS|js.t|||j|j|j }|j rt|j||dd}n|}|r$||k(r|j|j k(r|}||fSt|||j|j|j}||fS)z4Create intermediate and preprocessed images for OCR.NF) correctionremove_vectors)rT_ocr)rr output_tag) rr%originanyr clean_finalrrrrlossless_reconstruction)rrOr ocr_imagepreprocess_out rasterize_outrasterize_ocr_outs r?make_intermediate_imagesrs_""G!%%I) M  w22G4J4JK L%/    % % NN &  NT n $$E..')))) N  ! ! )##1#! ! !.  !]2 !4!44'I n $$#!))mm I n $$rAc|j}d}|jr"t|j|}t ||}t ||\}}t ||}d}|jsU|sJ|}t|jr t||}|jj||} | | }t|||}|||fS)zEProcess page to create OCR image, visible page image and orientation.rN)pageimage_filename)r rotate_pagesr&rrrrrr(pageinforrfilter_page_imager) rrrOrasterize_preview_outrr ocr_image_outpdf_page_from_image_outvisible_image_outfiltered_images r? process_pagers""G 1,2E2E| T!; !<" !9,!I~%Y =M"  * *~* ,\-B-B C 78I< X %44FF.?G   % . "< |-C#  13I IIrActj|5}|jd}t|r|j ||}n|}ddd|j j dk(r)t|\}}||j jd<nN|j j jdr)t|}||}nt|}t|||}|jj|} t|j j } | xr t!|| d<t#|| }t%|||S#1swYxYw) zPostprocess the PDF file.zfix_annots.pdfNauto_actual_output_typepdfa)context linearize)pdf_save_settings)rrget_pathrsaver output_typer) extra_attrs startswithr*rrris_optimization_enabledrr'rr!) pdf_filerrr fix_annotspdf_out actual_typespeculative_result ps_stub_out optimizing save_settingss r? postprocessrs= ( s%%&67 )# . HHZ  GG ""f,,Wg>=H##$9:  $ $ / / 71'7C  )(G37;K%g{GDG''???PJ)'//*E*EFM%/!V4DWg4VM+WgOG ( 337s 3E  Ecv|jdk(r%tjdtj(St|jdr?|jj r%tjdtj(St |jt tjr tj(S|jdk(r|jjdd}t|j}|dk(r|d rtjd |d n|d rtjd |d ntjd nx|jjdr]t|j}|d rtjd|d n)tjd|d tj St#|js%tjdtj$St'|||j|tj(S)N-zOutput sent to stdoutwritablezOutput written to streamrrrrpassz.Output file is a %s (auto mode achieved PDF/A) conformancezOutput file is a %sz Output file is a PDF (auto mode)z!Output file is a %s (as expected)zOOutput file is a valid PDF, but conversion to PDF/A did not succeed (issue: %s)z)Output file: The generated PDF is INVALID) output_filerinfor9rr3r rdevnullr r rr4r warningr-pdfa_conversion_failedr/invalid_output_pdfr,ok)rstart_input_fileoptimize_messagesr pdfa_infos r?report_output_pdfr&sc! ()L ;;K $$j 1g6I6I6R6R6T +,H ;;G '%%tBJJ'7 8 D ;;A   & (!--112GOK()<)<=If$6):Dm,6". -0HI;<  + +F 3()<)<=I VW *m,  666,,- KKC D.. . %w':':r8rErGrI JSONEncoderrY JSONDecoderrfrnrrrrrrrrrrrrr&rWrAr?rFs]#  .86%!221 ;98.-((;<+$g!ioo   ) BB,$(($(( @@ @@')%% #%3%< , , ;;>$?>$ >$*>$ >$B ).0       "9%9%7:9%9%xJB 4 4' 43; 4 4F(rA