U ߦwhL@sUdZdZdZddlZddlZddlZddlZddlZddlm Z ddl m Z m Z ddl m Z ddlmZdd lmZmZmZmZmZmZmZmZmZejdd d krdd lmZn dd lmZejd krddlmZn ddlmZddlmZm Z m!Z!eee"e"e"fee"e"e"fee"e"e"ffZ#ee$d<ee"e"e"e"e"e"fZ%ee$d<eeZ&ee'e&fZ(ee'ee dddZ)e'e'e'dddZ*dZ+d,e+Z-de-dZ.d`e&ee/e0dddZ1e&e0d d!d"Z2e&e3d d#d$Z4e0e3d%d&d'Z5e&dd d(d)Z6e&ee0e0d*d+d,Z7e&e/e0d-d.d/Z8e&e0d d0d1Z9e#e#e#d2d3d4Z:e&dd d5d6Z;ee'e/d7d8d9Zd?d@Z=e'ddAdBdCZ>e'e'e'ddDdEdFZ?e'e'e'ddDdGdHZ@e'e'ddIdJdKZAe'e'ddIdLdMZBe'e'ddNdOdPZCe'e'ddNdQdRZDdbe'ee'efee'e'fe3ddTdUdVZEe/e'dWdXdYZFGdZd[d[ZGe Gd\d]d]ZHejIGd^d_d_ZJdS)cz"Utility functions for PDF library.zMathieu Fenniakzbiziqe@mathieu.fenniak.netN) dataclass)datetimetimezone)DEFAULT_BUFFER_SIZE)SEEK_CUR) IOAnyDictListOptionalPatternTupleUnionoverload) ) TypeAlias)r )Self)STREAM_TRUNCATED_PREMATURELYDeprecationErrorPdfStreamErrorTransformationMatrixTypeCompressedTransformationMatrix)textreturnc Cs|}|dkrdS|dr$d|}|dr6|d7}|dddddd }t|d|d }|dkr|t|d kr|d 7}d D]N}zt||}Wntk rYqYqX|dr|jt j d}|Sqtd|dS)NrzD:)ZzZ0000r+r'-Z00)zD:%YzD:%Y%mzD:%Y%m%dz D:%Y%m%d%Hz D:%Y%m%d%H%MzD:%Y%m%d%H%M%SzD:%Y%m%d%H%M%S%zz+0000)tzinfozCan not convert date: ) isdigitendswithreplacemaxfindlenrstrptime ValueErrorrutc)rZorgtextifdr20/tmp/pip-unpacked-wheel-r8zeli8p/pypdf/_utils.pyparse_iso8824_dateQs(      r4)header1header2rcCsfd}g}||kr |||||kr8|||t|dkrZtd|d|d|t|S)N)z%PDF-1.3z%PDF-1.4z%PDF-1.5z%PDF-1.6z%PDF-1.7z%PDF-2.0rzNeither z nor z are proper headers)appendindexr+r-r))r5r6versionsZpdf_header_indicesr2r2r3_get_max_pdf_version_headerqs r:)     [])streammaxcharsrcCs8d}|d}|s4|sq4||7}t||krq4q|S)a< Read non-whitespace characters and return them. Stops upon encountering whitespace or when maxchars is reached. Args: stream: The data stream from which was read. maxchars: The maximum number of bytes returned; by default unlimited. Returns: The data which was read. rAr)readisspacer+)rDrEtxttokr2r2r3read_until_whitespaces   rJ)rDrcCs"|d}|tkr|d}q |S)z Find and read the next non-whitespace character (ignores whitespace). Args: stream: The data stream from which was read. Returns: The data which was read. rrF WHITESPACESrDrIr2r2r3read_non_whitespaces  rNcCs2|d}d}|tkr*|d7}|d}q|dkS)a  Similar to read_non_whitespace, but return a boolean if at least one whitespace character was read. Args: stream: The data stream from which was read. Returns: True if one or more whitespace was skipped, otherwise return False. rrrK)rDrIZcntr2r2r3skip_over_whitespaces  rO)valuercCstdd|DS)z Check if the given value consists of whitespace characters only. Args: value: The bytes to check. Returns: True if the value only has whitespace characters, otherwise return False. css|]}|tkVqdSN)WHITESPACES_AS_BYTES).0br2r2r3 sz+check_if_whitespace_only..)all)rPr2r2r3check_if_whitespace_onlys rWcCsF|d}|dd|dkrB|dkrB|d}|dkrtdqdS)Nr%)r=r?rAzFile ended unexpectedly.)rFseekrrMr2r2r3skip_over_comments   r[)rDregexrcCspd}|d}|s|S|||}|dk rb||t|t|d||d|}ql||7}q|S)z Read until the regular expression pattern matched (ignore the match). Treats EOF on the underlying stream as the end of the token to be matched. Args: regex: re.Pattern Returns: The read bytes. rANr)rFsearchrZstartr+)rDr\namerImr2r2r3read_until_regexs    rb)rDto_readrcCs>||krtd|| t||}|| t|S)a Given a stream at position X, read a block of size to_read ending at position X. This changes the stream's position to the beginning of where the block was read. Args: stream: to_read: Returns: The data which was read. z!Could not read malformed PDF file)tellrrZrrF)rDrcrFr2r2r3read_block_backwardss   recCsg}d}|dkrtttt|}|dkr4qt||}t|d}|sx|dkrl||dkrl|d8}qN|dkrxd}|r|||dd|dkr||dkr|d8}qn |||dkr||dt qqd |dddS) a Given a byte stream with current position X, return the previous line. All characters between the first CR/LF byte found before X (or, the start of the file, if no such byte is found) and position X After this call, the stream will be positioned one byte after the first non-CRLF character found beyond the first CR/LF byte before X, or, if no such byte is found, at the beginning of the stream. Args: stream: StreamType: Returns: The data which was read. Frrs TNrArX) rdrrminrrer+r7rZrjoin)rDZ line_contentZ found_crlfrcblockidxr2r2r3read_previous_lines.      rj)arTrcstfdd|DS)Nc3s(|] tfddtDVqdS)c3s&|]}tddt|DVqdS)css"|]\}}t|t|VqdSrQ)float)rSr/jr2r2r3rUPsz6matrix_multiply....N)sumzip)rScolrowr2r3rUPsz,matrix_multiply...N)tuplero)rSrTrqr3rUOsz"matrix_multiply..)rs)rkrTr2rtr3matrix_multiplyLs ruc Csdd}|| dtdd0}||||d|||W5QRX|| ddS)z5Create text file showing current location in context.irzpypdf_pdfLocation.txtwbsHEREN)rZopenwriterF)rDZradiusZ output_fhr2r2r3 mark_locationUs  ry)rTrcCsdSrQr2rtr2r2r3ord_asrzcCsdSrQr2rtr2r2r3rzfscCsdSrQr2rtr2r2r3rzkscCst|trt|S|SrQ) isinstancestrordrtr2r2r3rzps r)msg stacklevelrcCstj|t|ddS)N)r)warningswarnDeprecationWarning)r~rr2r2r3 deprecatevsr)r~rcCs t|dSrQ)r)r~r2r2r3 deprecationzsr)old_namenew_name removed_inrcCs t|d|d|dddS)zFIssue a warning that a feature will be removed, but has a replacement., is deprecated and will be removed in pypdf . Use instead.Nrrrrr2r2r3deprecate_with_replacement~srcCst|d|d|ddS)zMRaise an exception that a feature was already removed, but has a replacement.( is deprecated and was removed in pypdf rrNrrr2r2r3deprecation_with_replacementsr)r`rrcCst|d|dddS)zCIssue a warning that a feature will be removed without replacement.r.rNrr`rr2r2r3deprecate_no_replacementsrcCst|d|ddS)zJRaise an exception that a feature was already removed without replacement.rrNrrr2r2r3deprecation_no_replacementsr)r~srcrcCst||dS)z Use this instead of logger.error directly. That allows people to overwrite it more easily. See the docs on when to use which: https://pypdf.readthedocs.io/en/latest/user/suppress-warnings.html N)logging getLoggererrorr~rr2r2r3 logger_errors rcCst||dS)a Use this instead of logger.warning directly. That allows people to overwrite it more easily. ## Exception, warnings.warn, logger_warning - Exceptions should be used if the user should write code that deals with an error case, e.g. the PDF being completely broken. - warnings.warn should be used if the user needs to fix their code, e.g. DeprecationWarnings - logger_warning should be used if the user needs to know that an issue was handled by pypdf, e.g. a non-compliant PDF being read in a way that pypdf could apply a robustness fix to still read it. This applies mainly to strict=False mode. N)rrwarningrr2r2r3logger_warningsrF) func_namekwargsaliasesfailrc Cs|D]\}}||kr|r0t|d|d||kr^t|d|d|d|d|d ||||<tj|d|dtdd qd S) z Helper function to deprecate arguments. Args: func_name: Name of the function to be deprecated kwargs: aliases: fail: z# is deprecated as an argument. Use z insteadz received both z and z as an argument. z is deprecated. Use rr)messagecategoryrN)itemsr TypeErrorpoprrr)rrrrZold_termZnew_termr2r2r3 rename_kwargss  r)bytesrcCsR|dkr|dS|dkr*|dddS|dkrB|dddS|dddS) Niz Bytei@Bz.1fz kBiʚ;z MBz GBr2)rr2r2r3_human_readable_bytess rc@s>eZdZdZd ddddZd edddZeddd ZdS) classpropertyz Decorator that converts a method with a single cls argument into a property that can be accessed directly from the class. NrcCs ||_dSrQfgetselfmethodr2r2r3__init__ szclassproperty.__init__cCs ||SrQr)rinstanceclsr2r2r3__get__szclassproperty.__get__cCs ||_|SrQrrr2r2r3getterszclassproperty.getter)N)N) __name__ __module__ __qualname____doc__rrrrrr2r2r2r3rsrc@s^eZdZUddlmZdZeed<dZe ed<dZ e eed<ed d d Z ed d d Z dS)Filer)IndirectObjectr"r`rAdataNindirect_referencercCs&|jjd|jdtt|jdS)Nz(name=z, data: )) __class__rr`rr+rrr2r2r3__str__(sz File.__str__cCs"|dddt|jdS)NrXz, hash: r)rhashrrr2r2r3__repr__+sz File.__repr__)rrrZgenericrr`r|__annotations__rrrr rrr2r2r2r3rs    rc@speZdZedZeddddZeee e efdddZ e e dd d Ze d d d Zee dddZdS)Versionz ^(\d+)(.*)$N) version_strrcCs||_|||_dSrQ)r_parse_version components)rrr2r2r3r3szVersion.__init__cCsn|d}g}|D]V}tj|}|s6|d|fq|d}|d}|dkrVd}|t||fq|S)Nrrrr)splitrCOMPONENT_PATTERNmatchr7groupint)rrrZparsed_components componentrZinteger_prefixsuffixr2r2r3r7s    zVersion._parse_version)otherrcCst|tsdS|j|jkS)NF)r{rr)rrr2r2r3__eq__Fs zVersion.__eq__rcCst|jt|jfSrQ)rrrsrrr2r2r3__hash__KszVersion.__hash__cCst|tstdt|t|j|jD]P\}}|\}}|\}}||krPdS||kr^dS||krldS||kr*dSq*t|jt|jkS)Nz#Version cannot be compared against TF)r{rr-typerorr+)rrZself_componentZother_componentZ self_valueZ self_suffixZ other_valueZ other_suffixr2r2r3__lt__Os zVersion.__lt__)rrrrecompilerr|rr r rrobjectboolrrrrr2r2r2r3r/s  r)N)r)F)Kr __author____author_email__ functoolsrrsysr dataclassesrrriorosrtypingrrr r r r r rr version_inforZtyping_extensionsrerrorsrrrrlrrrZ StreamTyper|Z StrByteTyper4r:rLrgrRZWHITESPACES_AS_REGEXPrrrJrNrrOrWr[rbrerjruryrzrrrrrrrrrrrrtotal_orderingrr2r2r2r3s   ,    &       ;  "   #-