o ZhNK @sUdZdZdZddlZddlZddlZddlZddlZddlm Z ddl m Z m Z ddl m Z ddlmZdd lmZmZmZmZmZmZmZmZmZejdd d kr[dd lmZndd lmZejd krmddlmZnddlmZddlmZm 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(dee'dee fddZ)de'de'de'fddZ*dZ+d,e+Z-de-dZ.dgde&d ee/de0fd!d"Z1de&de0fd#d$Z2de&de3fd%d&Z4d'e0de3fd(d)Z5de&ddfd*d+Z6de&d,ee0de0fd-d.Z7de&d/e/de0fd0d1Z8de&de0fd2d3Z9d4e#d5e#de#fd6d7Z:de&ddfd8d9Z;ed5e'de/fd:d;Zd;ZdFe'dGe'dHe'ddfdIdJZ?dFe'dGe'dHe'ddfdKdLZ@dMe'dHe'ddfdNdOZAdMe'dHe'ddfdPdQZBd@e'dRe'ddfdSdTZCd@e'dRe'ddfdUdVZD WdidXe'dYee'efdZee'e'fd[e3ddf d\d]ZEd^e/de'fd_d`ZFGdadbdbZGe GdcddddZHejIGdedfdfZJdS)jz"Utility functions for PDF library.zMathieu Fenniakzbiziqe@mathieu.fenniak.netN) dataclass)datetimetimezone)DEFAULT_BUFFER_SIZE)SEEK_CUR) IOAnyDictListOptionalPatternTupleUnionoverload) ) TypeAlias)r )Self)STREAM_TRUNCATED_PREMATURELYDeprecationErrorPdfStreamErrorTransformationMatrixTypeCompressedTransformationMatrixtextreturnc Cs|}|durdS|drd|}|dr|d7}|dddddd }t|d|d }|dkrD|t|d krD|d 7}d D]#}zt||}Wn tyYYqFw|drf|jt j d}|Std|)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)rZorgtextifdr2C/var/www/html/lang_env/lib/python3.10/site-packages/pypdf/_utils.pyparse_iso8824_dateQs*     r4header1header2cCsfd}g}||vr|||||vr|||t|dkr-td|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:)     []streammaxcharscCs>d} |d}|s|s |S||7}t||kr |Sq)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. rATr)readisspacer+)rDrEtxttokr2r2r3read_until_whitespaces   rJcCs(|d}|tvr|d}|tvs |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  rNcCs8|d}d}|tvr|d7}|d}|tvs |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  rOvaluecCstdd|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|]}|tvVqdSN)WHITESPACES_AS_BYTES).0br2r2r3 sz+check_if_whitespace_only..)all)rPr2r2r3check_if_whitespace_onlys rWcCsT|d}|dd|dkr&|dvr(|d}|dkr td|dvsdSdSdS)Nr%)r=r?rAzFile ended unexpectedly.)rFseekrrMr2r2r3skip_over_comments    r[regexcCsrd} |d}|s |S|||}|dur4||t|t|d||d|} |S||7}q)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. rATNr)rFsearchrZstartr+)rDr\namerImr2r2r3read_until_regexs   rbto_readcCs>||kr td|| 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}|dkrtt tt|}|dkrndt||}t|d}|sF|dkr@||dvr@|d8}|dkr@||dvs2|dkrFd}|rl|||dd|dkrk||dvrk|d8}|dkrk||dvs]n|||dkr~||dt nqd |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. FrTrs NrArX) rdrrminrrer+r7rZrjoin)rDZ line_contentZ found_crlfrcblockidxr2r2r3read_previous_lines8    "rjarTcstfdd|DS)Nc3s*|]tfddtDVqdS)c3s(|]}tddt|DVqdS)css$|] \}}t|t|VqdSrQ)float)rSr/jr2r2r3rUPs"z6matrix_multiply....N)sumzip)rScolrowr2r3rUPs&z,matrix_multiply...N)tuplero)rSrTrqr3rUOs  z"matrix_multiply..)rs)rkrTr2rtr3matrix_multiplyLs rucCsxd}|| dtdd}||||d|||Wdn1s.wY|| ddS)z5Create text file showing current location in context.irzpypdf_pdfLocation.txtwbsHEREN)rZopenwriterF)rDZradiusZ output_fhr2r2r3 mark_locationUs  rycCdSrQr2rtr2r2r3ord_ar{cCrzrQr2rtr2r2r3r{fr|cCrzrQr2rtr2r2r3r{kr|cCst|tr t|S|SrQ) isinstancestrordrtr2r2r3r{ps rmsg stacklevelcCstj|t|ddS)N)r)warningswarnDeprecationWarning)rrr2r2r3 deprecatevsrcCst|rQ)r)rr2r2r3 deprecationzsrold_namenew_name removed_incCs 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_replacementsrr`cCst|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_replacementsrsrccCt||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 getLoggererrorrrr2r2r3 logger_errors rcCr)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_namekwargsaliasesfailc Cs|D]?\}}||vrC|rt|d|d||vr/t|d|d|d|d|d ||||<tj|d|dtdqd 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 r)messagecategoryN)itemsr TypeErrorpoprrr)rrrrZold_termZnew_termr2r2r3 rename_kwargss* rbytescCsR|dkr |dS|dkr|dddS|dkr!|dddS|dddS) Niz Bytei@Bz.1fz kBiʚ;z MBz GBr2)rr2r2r3_human_readable_bytess rc@s:eZdZdZd d ddZd defddZdefdd 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__  zclassproperty.__init__cCs ||SrQr)rinstanceclsr2r2r3__get__rzclassproperty.__get__cCs ||_|SrQrrr2r2r3getterszclassproperty.getterrQ)rN) __name__ __module__ __qualname____doc__rrrrrr2r2r2r3rs  rc@sdeZdZUddlmZdZeed< dZe ed< dZ e eed< d efd d Z d efd d Z dS)Filer)IndirectObjectr"r`rAdataNindirect_referencercCs&|jjd|jdtt|jdS)Nz(name=z, data: )) __class__rr`rr+rrr2r2r3__str__'s&z File.__str__cCs"|dddt|jdS)NrXz, hash: r)rhashrrr2r2r3__repr__*s"z File.__repr__)rrrZgenericrr`r~__annotations__rrrr rrr2r2r2r3rs   rc@sjeZdZedZdeddfddZdedee e effddZ d e de fd d Zd ede fd d ZdS)Versionz ^(\d+)(.*)$ version_strrNcCs||_|||_dSrQ)r_parse_version components)rrr2r2r3r2szVersion.__init__cCsn|d}g}|D]+}tj|}|s|d|fq |d}|d}|dur+d}|t||fq |S)Nrrrr)splitrCOMPONENT_PATTERNmatchr7groupint)rrrZparsed_components componentrZinteger_prefixsuffixr2r2r3r6s    zVersion._parse_versionothercCst|tsdS|j|jkS)NF)r}rr)rrr2r2r3__eq__Es  zVersion.__eq__cCst|tstdt|t|j|jD](\}}|\}}|\}}||kr(dS||kr/dS||kr6dS||kr=dSqt|jt|jkS)Nz#Version cannot be compared against TF)r}rr-typerorr+)rrZself_componentZother_componentZ self_valueZ self_suffixZ other_valueZ other_suffixr2r2r3__lt__Js zVersion.__lt__)rrrrecompilerr~rr r rrobjectboolrrrr2r2r2r3r.s  rrQ)r)F)Kr __author____author_email__ functoolsrrsysr dataclassesrrriorosrtypingrrr r r r r rr version_infortyping_extensionsrerrorsrrrrlrrrZ StreamTyper~Z StrByteTyper4r:rLrgrRZWHITESPACES_AS_REGEXPrrrJrNrrOrWr[rbrerjruryr{rrrrrrrrrrrrtotal_orderingrr2r2r2r3s   ,    &       :   $    "-