Wwg @dZddlZddlZddlZddlZddlZddlZddlZddlZddl m Z m Z ddl m Z m Z mZddlmZddlmZmZmZmZmZmZddlmZmZmZmZmZmZmZmZm Z m!Z!m"Z"ejFrdd l$m%Z%gd Z&d Z'd Z(ejRd jUdZ+ejXe-Z.Gddej^Z0de0fdZ1GddZ2Gdde2Z3e dGdde2Z4 dWddde5ddfdZ6Gd dZ7Gd!d"e7Z8Gd#d$e7Z9Gd%d&e:e7Z;Gd'd(e7e2Z<Gd)d*ejze7Z>Gd+d,e?e7Z@d-eeAeBeCfded.fd/ZDd0ZEded.fd1ZFdeBfd2ZGded.fd3ZHGd4d5eBe7ZIGd6d7ej^ZJd8eeBeCfdeJfd9ZKGd:d;eAe7ZLd<ZMd=eBdd>fd?ZNGd@d>eAe7ZOdAZPGdBdCeQe7ZRGdDdEeRZSdFZTdGZUdHZVeWeVdIk(sJeXeVDcic]\}}|| c}}ZYeOZZeLeIeRe;fZ[dJZ\GdKdLe7Z]dMZ^dNe deLfdOZ_ejRdPZ`ejRdQZaejRdRZbejRdSZcdXdTeAdUe5de fdVZdycc}}w)Yz Implementation of PDF object types and other generic functionality. The internals were imported from PyPDF2, with modifications. See :ref:`here ` for the original license of the PyPDF2 project. N) dataclassfield)datetime timedeltatimezone)BytesIO)AnyCallableIteratorOptionalTupleUnion) IndirectObjectExpectedPdfError PdfReadErrorPdfStreamErrorPdfStrictReadError PdfWriteErroris_regular_characterread_non_whitespaceread_until_delimiterread_until_regexskip_over_whitespaceSecurityHandler)Dereferenceable ReferenceTrailerReference PdfObjectIndirectObject NullObject BooleanObject FloatObject NumberObjectByteStringObjectTextStringObject NameObject ArrayObjectDictionaryObject StreamObject read_objectpdf_name pdf_stringpdf_dateTextStringEncodingEncryptedObjAccessDecryptedObjectProxys/<[tf(n%s+-z(\d+)\s+(\d+)\s+R[^a-zA-Z]asciic eZdZdZdZ dZ dZy)r1zt Defines what to do when an encrypted object is encountered when retrieving an object from a container. rrN)__name__ __module__ __qualname____doc__PROXY TRANSPARENTRAWP/var/www/horilla/myenv/lib/python3.12/site-packages/pyhanko/pdf_utils/generic.pyr1r1Gs+ E K  Cr>r1eoact|tr>|tjk(r |jS|tj k(r |j S|SN) isinstancer2r1r; decryptedr< raw_object)objr@s r?_deproxy_decryptrGfsD#+, $00 0== &** *>> ! Jr>ceZdZdZddZdZy)raD Represents an opaque reference to a PDF object associated with a PDF Handler (see :class:`PdfHandler <.rw_common.PdfHandler>`). This can either be a reference to an object with an object ID (see :class:`.Reference`) or a reference to the trailer of a PDF document (see :class:`.TrailerReference`). ct)zgRetrieve the PDF object backing this dereferenceable. :return: A :class:`.PdfObject`. NotImplementedErrorselfs r? get_objectzDereferenceable.get_objecty "!r>ct)zzReturn the PDF handler associated with this dereferenceable. :return: a :class:`~.rw_common.PdfHandler`. rJrLs r?get_pdf_handlerzDereferenceable.get_pdf_handlerrOr>Nreturnr )r6r7r8r9rNrQr=r>r?rros""r>rc$eZdZdZdZddZdZy)rao A reference to the trailer of a PDF document. .. warning:: Since the trailer does not have a well-defined object ID in files with "classical" cross-reference tables (as opposed to cross-reference streams), this is not a subclass of :class:`.Reference`. :param reader: a :class:`~pyhanko.pdf_utils.reader.PdfFileReader` c||_yrBreader)rMrWs r?__init__zTrailerReference.__init__s  r>c.|jjSrB)rWtrailerrLs r?rNzTrailerReference.get_objects{{"""r>c|jSrBrVrLs r?rQz TrailerReference.get_pdf_handlers {{r>NrR)r6r7r8r9rXrNrQr=r>r?rrs #r>rT)frozenc`eZdZUdZeed< dZeed< eddddZe ed< d d Z d Z y) ra A reference to an object with a certain ID and generation number, with a PDF handler attached to it. .. warning:: Contrary to what one might expect, the generation number does *not* indicate the document revision in which the object was modified. In fact, nonzero generation numbers are exceedingly rare these days; in most real-world PDF files, objects are simply overridden without ever increasing the generation number. Except in very specific circumstances, dereferencing a :class:`.Reference` will return the most recent version of the object with the stated object ID and generation number. idnumr generationFN)reprhashcomparedefaultpdfc|j tSddlm}t |j|sJ|jj |j S)Nr PdfHandler)rdr"pyhanko.pdf_utils.rw_commonrgrCrN)rMrgs r?rNzReference.get_objectsI 88 < :$((J///xx""4(3355r>c|jSrB)rdrLs r?rQzReference.get_pdf_handlers xxr>rR) r6r7r8r9int__annotations__r_rrdobjectrNrQr=r>r?rrsJ JJUtLCL6r>r container_refas_metadata_streamrSr cT|jd}|jdtjtj |}|dk(rt j|}n|dk(r]|jd}|jdtj|dk(rtj|||}nit|}n\|dk(rtj||}n?|dk(s|d k(rtj|}n|d k(r t|}n |d k(rtj|}n|d k(rR|d vr|jd}|d vrt||jdtjt||}n|t vrt"j|}n||jd}|jt%| tjt&j)|t*j||}nt"j|}||_|S)a Read a PDF object from an input stream. .. note:: The `container_ref` parameter tells the API which reference to register when the returned object is modified in an incremental update. See also here :ref:`here ` for further information. :param stream: An input stream. :param container_ref: A reference to an object containing this one. *Note:* It is perfectly possible (and common) for `container_ref` to resolve to the return value of this function. :param as_metadata_stream: Whether to dereference the object as an XMP metadata stream. :return: A :class:`.PdfObject`. rrr5<<rn)  )readseekosSEEK_CUROBJECT_PREFIXESfindr(read_from_streamr*read_hex_string_from_streamr)r#read_string_from_streamr"rr, NUMBER_SIGNSr%lenINDIRECT_PATTERNmatchr!rm)streamrmrntokidxresultpeeks r?r,r,s2 ++a.C KKBKK   s #C ax,,V4 {{1~ B $ 5=%66 :L7F18F --fmD SAX//7 (0 ,,V4 '++a.C'F# B $V]3 , !226:F;;r?D KKT BKK 0%%d+7'88O%66v>(F Mr>c\eZdZUdZdZeeed< defdZdZ d deddee fd Z y) r zSuperclass for all PDF objects.NrmrSc8|j}| td|S)z Return a reference to the closest parent object containing this object. Raises an error if no such reference can be found. zONo container reference available. This object probably wasn't read from a file.)rmr)rMrefs r?get_container_refzPdfObject.get_container_ref=s-    ;,  r>c|S)zpResolves indirect references. :return: `self`, unless an instance of :class:`.IndirectObject`. r=rLs r?rNzPdfObject.get_objectJs  r>handlerrct)z Abstract method to render this object to an output stream. :param stream: An output stream. :param container_ref: Local encryption key. :param handler: Security handler rJrMrrrms r?write_to_streamzPdfObject.write_to_streamQs "!r>NN) r6r7r8r9rmr rrkrrNrrr=r>r?r r sV)/3M8O,34 ? 04-1 "+," * "r>cJeZdZdZ d dedfdZedZdZdZ d Z y) r"zO PDF `null` object. All instances are treated as equal and falsy. Nrrc&|jdy)Nnull)writers r?rzNullObject.write_to_streamks  Wr>cX|jd}|dk7r tdtS)NrurzCould not read Null object)r|rr")rnulltxts r?rzNullObject.read_from_streamss+++a. g ;< <|r>c.||uxst|tSrB)rCr"rMothers r?__eq__zNullObject.__eq__zsu}= 5* ==r>ctdSrB)rarLs r?__hash__zNullObject.__hash__}s Dzr>cy)NFr=rLs r?__bool__zNullObject.__bool__sr>r) r6r7r8r9r r staticmethodrrrrr=r>r?r"r"dsG04 +, >r>r"cVeZdZdZdZ d dedfdZedZdZ d Z d Z d Z y) r#zPDF boolean value.c||_yrB)value)rMrs r?rXzBooleanObject.__init__s  r>Nrrcb|jr|jdy|jdy)Ntruesfalse)rrrs r?rzBooleanObject.write_to_streams# :: LL ! LL "r>c|jd}|dk(r tdS|dk(r|jddk(r tdStd) NrurTsfalsreFzCould not read Boolean object)r|r#r)rwords r?rzBooleanObject.read_from_streamsO{{1~ 7? & & W_{{1~%$U++:;;r>c,t|jSrB)boolrrLs r?rzBooleanObject.__bool__sDJJr>c`t|ttfxrt|t|k(SrB)rCr#rrs r?rzBooleanObject.__eq__s1%-!67 DJ$ K =  r>c*tt|SrB)strrrLs r?__str__zBooleanObject.__str__s4:r>ct|SrB)rrLs r?__repr__zBooleanObject.__repr__s 4yr>r) r6r7r8r9rXr rrrrrrrr=r>r?r#r#sO 04 #+, #<<  r>r#cbeZdZdZdZej fdefdZ d dedfdZ e d Z y) r)a PDF array object. This class extends from Python's list class, and supports its interface. .. warning:: Contrary to the case of dictionary objects, PyPDF2 does not transparently dereference array entries when accessed using :meth:`__getitem__`. For usability & consistency reasons, I decided to depart from that and dereference automatically. This makes the behaviour of :class:`.ArrayObject` consistent with :class:`.DictionaryObject`. That said, some vestiges of the old PyPDF2 behaviour may linger in the codebase. I'll fix those as I get to them. c@|j|jSrB)raw_getrN)rMindexs r? __getitem__zArrayObject.__getitem__s||E"--//r>decryptcFtj||}t||S)ay .. versionchanged:: 0.14.0 ``decrypt`` parameter is no longer boolean Get a value from an array without dereferencing. In other words, if the value corresponding to the given key is of type :class:`.IndirectObject`, the indirect reference will not be resolved. :param index: Key to look up in the dictionary. :param decrypt: What to do when retrieving encrypted objects; see :class:`.EncryptedObjAccess`. The default is :attr:`.EncryptedObjAccess.TRANSPARENT`. :return: A :class:`.PdfObject`. )listrrG)rMrrvals r?rzArrayObject.raw_gets#0tU+W--r>Nrrc|jd|D]'}|jd|j|||)|jdy)N[ )rrms ])rr)rMrrrmdatas r?rzArrayObject.write_to_streamsS  T D LL   } !    Ur>ct}|jd}|dk7r td t|}|dk(r |S|j dt j |jt||O)NrrzCould not read array]rp) r)r|rrr}r~rappendr,)rrmarrtmp peekaheads r?rzArrayObject.read_from_streamstmkk!n $;56 6+F3ID   KKBKK ( JJ{6=9 :r>r) r6r7r8r9rr1r;rr rrrr=r>r?r)r)sW"0 '9&D&D.$.<04 +,   r>r)ceZdZdZdZdZdZedefdZ edZ dZ d Z d Z d Z dd edfdZeddZy )r!a Thin wrapper around a :class:`.Reference`, implementing both the :class:`.Dereferenceable` and :class:`.PdfObject` interfaces. .. warning:: For many purposes, this class is functionally interchangeable with :class:`.Reference`, with one important exception: :class:`.IndirectObject` instances pointing to the same reference but occurring at different locations in the file may have distinct `container_ref` values. c(t||||_yrB)r reference)rMr^r_rds r?rXzIndirectObject.__init__ s"5*c:r>cz|jj}t|tr|jS|S)zC :return: The PDF object this reference points to. )rrNrCr!)rMrFs r?rNzIndirectObject.get_object s3nn'') $.c>#Bs~~KKr>c6|jjSrB)rrQrLs r?rQzIndirectObject.get_pdf_handlers~~--//r>rSc.|jjS)z; :return: the object ID of this reference. )rr^rLs r?r^zIndirectObject.idnums ~~###r>c.|jjS)zC :return: the generation number of this reference. )rr_rLs r?r_zIndirectObject.generation"s ~~(((r>c<d|jd|jdS)NzIndirectObject(z, ))r^r_rLs r?rzIndirectObject.__repr__)s+/::tGGr>cDt|j|jfSrB)rar^r_rLs r?rzIndirectObject.__hash__/sTZZ122r>cd|duxr+t|txr|j|jk(SrB)rCr!rrs r?rzIndirectObject.__eq__2s3   25.1 2%//1 r>c&|j| SrB)rrs r?__ne__zIndirectObject.__ne__9s;;u%%%r>NrrcX|jd|j|jfzy)Ns%d %d R)rr^r_rs r?rzIndirectObject.write_to_stream<s"  Z4::t"??@r>cnd} |jd}|s td|jr|s2n||z }8d} |jd}|s td|jr|s2n||z }8t|}|dk7r't |j }t d|z t|t|}}|dkDr|dk\st tt|t||jS#t$r(t |j }t d|wxYw)Nr>rStream has ended unexpectedlyRz2Error reading indirect object reference at byte %srz0Parse error on indirect object reference around ) r|risspacerhextellrrj ValueErrorr!rQ) rrm idnum_strrgeneration_strrposr^r_s r?rzIndirectObject.read_from_streamDsU ++a.C$%DEE{{}   I++a.C$%DEE{{}% c !N  ' 9fkkm$CDsJ   #IN0C:EAI*/  #2  NC/1N1N1P   fkkm$CB3%H  s /&D1D4r)rmr)r6r7r8r9rXrNrQpropertyrjr^r_rrrrr rrrr=r>r?r!r!s ; L0$s$$ )) H 3 & 04 A+,A( ( r>r!c<eZdZdZd dZdZdZ d dedfdZy) r$z PDF Float object. Internally, these are treated as decimals (and therefore actually fixed-point objects, to be precise). cTtjj|t|SrB)decimalDecimal__new__r)clsrs r?rzFloatObject.__new__xs&&sCJ77r>c||jk(r-t|jtjdSt|S)Nr) to_integralrquantizerrrLs r?rzFloatObject.__repr__{s: 4##% %t}}W__Q%789 9t9 r>ct|S)zD :return: a Python ``float`` value for this object. )floatrLs r? as_numericzFloatObject.as_numericsT{r>NrrcV|jt|jdyNr3rr`encoders r?rzFloatObject.write_to_stream  T$Z&&w/0r>)0r) r6r7r8r9rrrr rr=r>r?r$r$ps38 04 1+,1r>r$cleZdZdZej dZdZdZdZ d de dfd Z e d Z y) r%zE PDF number object. This is the PDF type for integer values. s [^+-.0-9].cDt|}tj||SrB)rjr)rrrs r?rzNumberObject.__new__s%j{{3$$r>ct|S)zB :return: a Python ``int`` value for this object. )rjrLs r?rzNumberObject.as_numerics4yr>NrrcV|jt|jdyrrrs r?rzNumberObject.write_to_streamrr>ct|tjd}|jtjdk7rt |j dSt|j dS)NT)regex ignore_eofrpr3)rr% NumberPatternrByteDotr$decode)rnums r?rzNumberObject.read_from_streams\ ,,   88L(( )R /szz'23 3 7 34 4r>r)r6r7r8r9recompilerrrrr rrrr=r>r?r%r%sYBJJ|,MG%04 1+,1 5 5r>r%string)r&r'c t|tr t|St|ttfr/t |} t|j |}||_|Std#t$rt|cYSwxYw)z Encode a string as a :class:`.TextStringObject` if possible, or a :class:`.ByteStringObject` otherwise. :param string: A Python string. z'pdf_string should have str or bytes arg) rCrr'bytes bytearray_guess_enc_by_bomrautodetected_encodingUnicodeDecodeErrorr& TypeError)rguessedretvals r?r.r.s&#'' FUI. /#F+ ,%gnnV&<=F+2F (MABB" ,#F+ + ,s"A,,BBs0123456789abcdefABCDEFcjddfd}tjdj|rdndz}t |S)zp Read a hex string from a stream into a PDF string object. :param stream: An input stream. rFc3~K t}|dk(ry|tvrtdt|z| 8w)N>z Unexpected token in hex string: )r HEX_DIGITSrr`)roddrs r? read_tokensz0read_hex_string_from_stream..read_tokenssR%f-Cd{J&$6cBI'Cs:=r>0)r|binascii unhexlifyjoinr.)rrrrs` @r?rrsP KKN C    34C8F f r>c,|jdd}t} |jd}|s td|dk(r|dz }n:|dk(r|dz}|dk(r* |jS|dk(r|jd}|dvrn|dk(rd }n|d k(rd }n|d k(rd }n|dk(rd}n|dk(rd}n|jrrt dD]J}|jd}|jr||z }*|j dt jnt|d}t|f}nS|dvr8|jd}|dvr |j dt jd}ntdt|z|j|w)Nrr()r\s() /%<>[]#_&$\nrzrryt bf r5rp)bases r>zUnexpected escaped string: )r|rrisdigitranger}r~rrjrrr`rgetvalue)rparenstxtrintokoctals r?_read_string_literal_bytesr.s KKN F )C kk!n !@A A $; aKF D[ aKF{Z <<>YE\++a.C(( qA!;;q>D||~t  B 4Ca(UHokk!ng%KKBKK0"#@49#LMM #m r>c*tt|S)z Read a PDF string literal from a stream. Attempt to decode it into a text string by autodetecting the encoding, or failing that, return it as a byte string instead. :param stream: An input stream. )r.r.)rs r?rr5s 08 99r>c<eZdZdZedZ ddedfdZy)r&zPDF bytestring class.c|SrBr=rLs r?zByteStringObject.Gs4r>Nrrc,|}|J|H|j}|j|j|j}|j ||}|j d|j t j||j dy)Nr)r6r7r8r9roriginal_bytesr rr=r>r?r&r&Ds2/0N04 +,r>r&ceZdZdZdZ ej dfZ ejdfZ ejdfZ de de fdZdee efde fd Zy) r0z) Encodings for PDF text strings. Nutf-16bezutf-8zutf-16lerrSc|tjk(r t|S|j\}}||j |zS)z Encode a string with BOM. :param string: The string to encode. :return: The encoded string. )r0PDF_DOCencode_pdfdocencodingrr)rMrbomencs r?rzTextStringEncoding.encode|s? %-- -(0 0zzHCs++ +r>c|tjk(r t|S|tjk(r|j dS|j dS)z Decode a string with BOM. :param string: The string to encode. :return: The encoded string. :raise UnicodeDecodeError: Raised if decoding fails. z utf-8-sigzutf-16)r0r@decode_pdfdocencodingUTF8r)rMrs r?rzTextStringEncoding.decodesK %-- -(0 0 ',, ,==- -==* *r>)r6r7r8r9r@codecs BOM_UTF16_BEUTF16BEBOM_UTF8rF BOM_UTF16_LEUTF16LErrrrrrr=r>r?r0r0^sG""J/G OOW %D""J/G ,S ,U ,+U5)#34++r>r0encodedc<|jtjrtjS|jtj rtj S|jtjrtjStjSrB) startswithrGrHr0rIrKrLrJrFr@)rMs r?rrsn&--.!)))   F// 0!)))   FOO ,!&&& ")))r>cfeZdZUdZdZeeed< dZeeed< e dZ d dedfdZ y) r'z! PDF text string object. Nr force_output_encodingcf|jr|jj|Std)z Retrieve the original bytes of the string as specified in the source file. This may be necessary if this string was misidentified as a text string. z#No information about original bytes)r rrrLs r?r<zTextStringObject.original_bytess0  % %--44T: :@A Ar>rrc|j|jj|}n t|}d}|/|-|j j}|dk(rdn|j}|U|j|j|j}|j||}t|}|j|y|jd|D]I} t!| g} | j#s| dk7r|jd| z9|j| K|jdy#t$r&tj |jdz}Y)wxYw)Nr> /Identityr s\%03or)rQrrAUnicodeEncodeErrorrGrHcrypt_filter_configstring_filter_namer5r6r^r_r7r&rrrisalnum) rMrrrmrMr:cf_namer;rFcc_s r?rz TextStringObject.write_to_streamsM  % % 10077=G  H/5   =#<11DDG K/W5N5N5PB >,,##]%=%=IjjG4G"7+C    ' LL  %A3Zzz|T LLQ/LL$  % LL 5& H!-- J0GG Hs D00+EEr) r6r7r8r9r r r0rkrQrr<rr=r>r?r'r'sj;?8$67>;?8$67> BB*04 )+,)r>r'cd|cxkr dkr|dz Sd|cxkr dkr|dz Sd|cxkr dkr|dz Std td ) N09AF7afWz6Numeric escape in PDF name must use hexadecimal digits)r) ascii_chars r? _as_hex_digitrgsz z!T!D  "  #t #D   $  #t #D   $ D  l D  r> name_bytesr(c^t}|jdt|} t|}|dk(r2 t|}t|}t|dzt|z}n)d|cxkrdkrnn t|st d|dd |jt|f#t$rt dt |wxYw#t$rYnwxYw|j}d }d }|D]$} |j|}n#t$rY"wxYw|Jt|S) zu Decode the bytes that make up a name object (minus the initial /), expanding all escapes along the way. /#z!Unterminated escape in PDF name /!~zByte (0x02xz) must be escaped in a PDF name)utf8latin1N)rriternext StopIterationrr`rgrrr(rrr() rhr name_itercur_bytedigit1digit2encodings_to_tryname_strrCs r? _decode_namer{ sj YF LLZ I IH4!)_F!)_F )025 f8MMh.$.7K8#xn,KL LL{+ ,'%&;Dcp|jd}|dk7r tdt|}t|S)NrrjzName object should start with /)r|rrr{)r name_startrhs r?rzNameObject.read_from_stream_s8[[^  @A A)&1 J''r>r)r6r7r8r9r rrrr=r>r?r(r(?s804 '+,'2((r>crt|ts&t|tr t|Std|S)Nzkey must be a name object)rCr(rr)keys r?_normalise_keyrhs1 c: & c3 c? "89 9 Jr>ceZdZdZdfd Zej fdeee fdefdZ dZ ddZ d Z d dd d eegeffd ZddefdZ ddedfdZe ddddefdZxZS)r*z A PDF dictionary object. Keys in a PDF dictionary are PDF names, and values are PDF objects. When accessing a key using the standard :meth:`__getitem__` syntax, :class:`.IndirectObject` references will be resolved. Nc |;t||jDcic]\}}t||c}}yt|ycc}}wrB)superrXitemsr)rM dict_datakv __class__s r?rXzDictionaryObject.__init__{sL  G 2;//2CD$!Q"A%D  G  EsA rrcFtj||}t||S)a{ .. versionchanged:: 0.14.0 ``decrypt`` parameter is no longer boolean Get a value from a dictionary without dereferencing. In other words, if the value corresponding to the given key is of type :class:`.IndirectObject`, the indirect reference will not be resolved. :param key: Key to look up in the dictionary. :param decrypt: What to do when retrieving encrypted objects; see :class:`.EncryptedObjAccess`. The default is :attr:`.EncryptedObjAccess.TRANSPARENT`. :return: A :class:`.PdfObject`. )dictrrG)rMrrrs r?rzDictionaryObject.raw_gets#.tS)W--r>ct|}t|ts td|j|j|_t j |||SNzvalue must be PdfObject)rrCr rrmr __setitem__rMrrs r?rzDictionaryObject.__setitem__sQS!%+67 7    )"&"4"4E c511r>ct|}t|ts td|j|j|_t j |||Sr)rrCr rrmr setdefaultrs r?rzDictionaryObject.setdefaultsOS!%+67 7    )"&"4"4E tS%00r>c4tj||}|dk(rQt|trAddlm}|j }t||sJ|j|jdS|j}t|tr t||S)Nz /MetadatarrfTrs) rrrCr!rhrgrQrNrr"KeyError)rMrraw_objrgr deref_objs r?rzDictionaryObject.__getitem__s""4- + *Wn"E >--/Ggz2 22%%!!d&  **,I)Z0sm#  r>F)rawrcfunctioncf |r|j|n||}||S#t$r|cYSwxYwrBrr)rMrrrrcrs r? get_and_applyzDictionaryObject.get_and_applys? ),DLL%$s)E N s " 00rScHd}|j||d}||st|S)NcFt|tr |jStrB)rCr!rr)rFs r?as_refz7DictionaryObject.get_value_as_reference..as_refs#~.}}$( (r>T)r)rr)rMroptionalrrs r?get_value_as_referencez'DictionaryObject.get_value_as_references1 ) ""3D"9 =N r>rrc|jdt|jD]M\}}|j||||jd|j||||jdO|jdy)Ns<< rrzs>>)rrrr)rMrrrmrrs r?rz DictionaryObject.write_to_streamsx  Wtzz|, JC    ? LL   ! !&'= A LL    Ur>rmrrnc|jd}|dk7r%tdt|jzi}|j } t |}|dk(r|jdn|j dtj tj|}t ||j dtjt||} ||vr| ||<nKd t|jd |} |jr t| tj!| |j} t |d } d} | d k(r |jd dk(rt#|d|t%d}t'|t(r2|j}|j+|}|j ||j|} t |}|jd}||zdk7rt|j}|j dtj|jd}|dk(r| dd} n(tdt|z|j | | t,}|r ddlm}|}||| St5|S#t$r+}tdt|jz|d}~wwxYw#t2$rYWwxYw)Nr5rrz?Dictionary read error at byte 0x%s: stream must begin with '<<'Trrrpz=Failed to read dictionary key at byte 0x%s; expected PDF namez+Multiple definitions in dictionary at byte z for key ) allow_eofsrvstream)stop_after_eol/Lengthr$s endstreami z:Unable to find 'endstream' marker after stream at byte %s.r)MetadataStream) encoded_data)r|rrrrQrr}r~rr(r Exceptionr,strictrloggerwarningrr-rCr!rNr+"pyhanko.pdf_utils.metadata.xmp_xmlr ImportErrorr*)rrmrnrrrrrexrerrrs stream_datalengthtendstreamorig_endstream_posendstm_clsrs r?rz!DictionaryObject.read_from_streams kk!n %<.03FKKM0BC //1%f-Cd{ A KKBKK (  11&9  ' KKBKK ( 6E$!S  (+6;;='93@>>,S11NN3'58kkm $ 7 9Q83 ! =(9-.F&.1KKM ++F3 A ++f-K#F+A{{1~HH -&,[[]" C-kk!n,&"-cr"2K&-/23E/FG KK   "#G!-G4k: :#D) )S "S&++-() J#s*J>K K$&K  K KKrBFr)r6r7r8r9rXr1r;rr(rrrrrr r r rrrr rrrr __classcell__rs@r?r*r*qs'9&D&D. :s? #.$.421!,  I;+,  Y 04 +, $)^*(^*!^*^*r>r*c DeZdZdZ ddeedeedeededffd Zd ed efd Z e d e fd Z e dddfdedfdZd eeeeeffdZdZdZe d efdZe d efdZ ddee fdZdZe dZ ddedffd ZxZS)r+a PDF stream object. Essentially, a PDF stream is a dictionary object with a binary blob of data attached. This data can be encoded by various filters (not all of which are currently supported, see :mod:`.filters`). A stream object can be initialised with encoded or decoded data. The former is used by :class:`.reader.PdfFileReader` to provide on-demand decoding, with :class:`.writer.BasePdfFileWriter` and its subclasses working the other way around. .. note:: The :class:`.StreamObject` class manages some of its dictionary keys by itself. This is partly the case for the various ``/Filter`` and ``/DecodeParms`` entries, but also for the ``/Length`` entry. The latter will be overwritten as necessary. :param dict_data: The dictionary data for this stream object. :param stream_data: The (unencoded) stream data. :param encoded_data: The encoded stream data. .. warning:: Ordinarily, a stream can be initialised either from decoded and from encoded data. If both `stream_data` and `encoded_data` are provided, the caller is responsible for making sure that both are compatible given the currently relevant filter configuration. :param handler: A reference to the currently active :class:`.pyhanko.pdf_utils.crypt.SecurityHandler`. This is only necessary if the stream requires crypt filters. NrrrrrcNt||||_||_||_yrB)rrX_data _encoded_data_handler)rMrrrrrs r?rXzStreamObject.__init__ns( #  ) r>rdecrypted_entriescL|||_|jr |j}no|jr|j }n|j }|j |j|j}|j||j}|j|||S)aJ Internal method to handle decrypting streams that are encrypted with the document's default encryption handler for streams and/or embedded files (i.e. not with any custom crypt filters). This routine is called deep in the object fetching stack, and you should never invoke it yourself. It's defined as a method in :class:`.StreamObject` because it needs to be able to preserve the type (subclass) of the stream object on which it is called, in order to properly feed into the logic surrounding metadata streams. )rr) r_has_crypt_filterris_embedded_file_streamget_embedded_file_filterget_stream_filterr6r^r_rr)rMrrrdecrypted_datar:r;s r? _implicit_decrypt_stream_contentz-StreamObject._implicit_decrypt_stream_contentzs  #DM  ! !"..N++557..0,,SYYGIZZ 43D3DEN~~ NG  r>rSc4dd|jDvS)N/Cryptc3&K|] \}}| ywrBr=).0name_s r? z1StreamObject._has_crypt_filter..s@WT1D@s)_filtersrLs r?rzStreamObject._has_crypt_filters@ @@@r>rTc|||_|j td||jjvrtd|d|xs t}t d|d<||d<|j t d|d y) Nz#There is no security handler aroundzThe crypt filter z& is not known to the security handler.z/CryptFilterDecodeParms/Type/NamerT)paramsallow_duplicates)rrrWr*r- apply_filter)rMrrrs r?add_crypt_filterzStreamObject.add_crypt_filters  #DM ==  !FG G t}}88 8 #D6)OP -+-"#<=ww  X v  r>c#K |td}t|tr|f}nt|ts t d |td}t|t r|g}t|tr-t|t|z }|dkDr|tg|zz }t|d|DEd{y#t$rYywxYw#t$rtgt|z}YMwxYw7:w)N/Filterz5/Filter should be a name object or an array of names. /DecodeParmsrc3<K|]}|jywrB)rN)r param_sets r?rz(StreamObject._filters..sOI--/Os) r-rrCr(r)rr*rrr"zip)rM filter_arr decode_paramslendiffs r?rzStreamObject._filterss hy12J j* -$JJ 4 G  = .!9:M-)9:!. -.j/C ,>>Q;!jl^g%==M  OO   5   * ='\NS_Z %CF(*zz';7;F Bz,7&88c"|jj|j}| td|jD]\}}|j ||}t |t r|j}||_|jJ|jS)z Return the decoded stream data as bytes. If the stream hasn't been decoded yet, it will be decoded on-the-fly. :raises .misc.PdfStreamError: If the stream could not be decoded. No data available.)rrrrrrC memoryviewtobytes)rMr filter_clsrs r?rzStreamObject.datas :: %%D|$%9::-1-B-B-D >) M!((}= >$ +||~DJzz%%%zzr>c |j^|j}| tdt|j }t |D]\}}|j ||}||_|jJ|jS)z Return the encoded stream data as bytes. If the stream hasn't been encoded yet, it will be encoded on-the-fly. :raises .misc.PdfStreamError: If the stream could not be encoded. r)rrrtuplerreversedr)rMrdecodersrrs r?rzStreamObject.encoded_datas    %::D|$%9::T2245H-5h-? >) M!((}= >!%D !!---!!!r>rc2|j}||j |j}t|j }t |t s t|}t ts t|s||td<r~|td<not|\}|s||vr|durtd|dyt|f|z|td<s tr"fd}t||td<d|_||_y)a Apply a new filter to this stream. This filter will be prepended to any existing filters. This means that is is placed *last* in the encoding order, but *first* in the decoding order. *Note:* Calling this method on an encoded stream will first cause the stream to be decoded using the filters already present. The cached value for the encoded stream data will be cleared. :param filter_name: Name of the filter (see :const:`~pyhanko.pdf_utils.filters.DECODERS`) :param params: Parameters to the filter (will be written to ``/DecodeParms`` if not ``None``) :param allow_duplicates: If ``None``, silently ignore duplicate filters. If ``False``, raise ValueError when attempting to add a duplicate filter. If ``True`` (default), duplicate filters are allowed. NrrFzFilter z) has already been applied to this stream.c3\Kxs tD]}|xs tywrB)r")r param_setsrs r?_paramsz*StreamObject.apply_filter.._paramsjs0 0JL0%/8 '7:<78s),) rrrrrrCr(r-r*rrr)any) rM filter_namerrr cur_filters filter_namesrrs ` @r?rzStreamObject.apply_filter+s(4zz c<|jtddy)z Convenience method to add a ``/FlateDecode`` filter with default settings, if one is not already present. *Note:* compression is not actually applied until the stream is written. z /FlateDecodeN)r)rr-rLs r?compresszStreamObject.compressss (>2TJr>cJ |jddk(S#t$rYywxYw)Nrz /EmbeddedFileFrrLs r?rz$StreamObject.is_embedded_file_stream|s- <<(O; ;  s  ""c|j}|V|T|jsH|j}|j|j|j }|j ||}tt||td<t|-||||d=|jd|j||jdy)Nrs stream s endstream) rrrr6r^r_r7r%rr(rrr)rMrrrmrr:r;rs r?rzStreamObject.write_to_streams     )****,B,,##]%=%=I::i.D&23t9&=Z "# ? O ]# T ^$r>)NNNN)NTr)r6r7r8r9r rrrXrrrrrr(rr r rrrrrrrrrrrrs@r?r+r+Gsd$P%)'+(,/3  D>  e_  uo  +,  $ %$ :>$ LA4AA  $/3  +,  . (5htn)<#=> B:1e("e""(LPF:B4.FPK04 %+,%%r>r+c.fd}t|S)Nc 3lKD]} t|y#t$rtd|dddwxYww)Npdfdocencodingrp#does not exist in translation table)_pdfDocEncoding_revrrV)r[unicode_strings r?_buildz%encode_pdfdocencoding.._buildsR A )!,,  ($9  s4 414)r)rrs` r?rArAs  ?r>c:fd}dj|S)Nc3rKD]-}t|}|dk(rtdt|fddd|/yw)Nr rpr )_pdfDocEncodingr r)br[ byte_arrays r?rz%decode_pdfdocencoding.._buildsP A"AH}($1$K9 G s47r)r)rrs` r?rErEs  7768 r>(rrrrrrrrrrrrrrrrrrrrrrrru˘uˇuˆu˙u˝u˛u˚u˜ !"#$%&'(r*+,-./r123456789:;<=>?@ABCDEFGHIJKLMNOPQRSTUVWXYZ[\]^r`arr[drfghr+jrlmnopqrrrurwxyz{|}~ru•u†u‡u…u—u–uƒu⁄u‹u›u−u‰u„u“u”u‘u’u‚u™ufiufluŁuŒuŠuŸuŽuıułuœušužru€¡¢£¤¥¦§¨©ª«¬r®¯°±²³´µ¶·¸¹º»¼½¾¿ÀÁÂÃÄÅÆÇÈÉÊËÌÍÎÏÐÑÒÓÔÕÖ×ØÙÚÛÜÝÞßàáâãäåæçèéêëìíîïðñòóôõö÷øùúûüýþÿc>t|tr t||S|SrB)rC PROXYABLEr2) encrypted_objrs r?proxy_encrypted_objrs-+#M7;;r>cteZdZUdZeed< defdZedefdZ d de dfd Z d Z ed Z d Z y)r2a Internal proxy class that allows transparent on-demand encryption of objects. .. warning:: Most public-facing APIs won't leave you to deal with these *directly* (that's half the reason this class exists in the first place), and the API of this class is considered internal. However, for reasons related to the historical PyPDF2 codebase from which pyHanko's object handling code ultimately derives, there are some Python builtins that might cause these wrapper objects to inadvertently "leak". Please `tell us about such cases `_ so we can make those types of access more convenient and robust. .. danger:: The ``__eq__`` implementation on this class is not safe for general use, due to the fact that certain structures in PDF are exempt from encryption. Only compare proxy objects with ``==`` in areas of the document where these exemptions don't apply. :param raw_object: A raw object, typically as-parsed from a PDF file. :param handler: The security handler governing this object. rEc.||_d|_||_yrB)rE _decryptedr)rMrErs r?rXzDecryptedObjectProxy.__init__s$/3 r>rSc ^ |j |jSddlm}|j}|j |j }t |tstd|dt |tst |tr\ j}|j|j|j}t|j!||j"}nt |t$r]|j'Dcic]\}}|t)| } }}t |t*r|j- || }nOt%| }nCt |t.rt1 fd|} t/| }nt3dt5|d|j |_||_|Scc}}w)aG The decrypted PDF object exposed as a property. If this object is a container object, its constituent parts will be wrapped in :class:`.DecryptedObjectProxy` as well, in order to defer further decryption until the values are requested through a getter method on the container. rrzPProxyable objects must have a container ref pointing to a numbered object, not 'z'.ct|SrB)r)rrs r?r2z0DecryptedObjectProxy.decrypted..&s*=a*Ir>zObject of type z is not proxyable.)rcryptrrErrmrCrrr&r'r5r6r^r_r.rr<r*rrr+rr)mapr type) rMrrFrmr:r;rDdictkeyrr decrypted_maprs @r?rDzDecryptedObjectProxy.decrypteds ?? &?? "*oo#'<<)) -3))6r;  c+ , !1 **,B,,##]%=%=I#2::i9K9K#LMI - .'*iik!"GU,UG<<! !#|,@@],= -->? [ ) I3OM#M2Iod3i[8JKL L"%"3"3 ##!sF)Nrrc>|jj|||yrB)rDrrs r?rz$DecryptedObjectProxy.write_to_stream.s &&vw Fr>c6|jjSrB)rDrNrLs r?rNzDecryptedObjectProxy.get_object:s~~((**r>c.|jjSrB)rErmrLs r?rmz"DecryptedObjectProxy.container_ref=s,,,r>cXt|txr|j|jk(SrB)rCr2rDrs r?rzDecryptedObjectProxy.__eq__As) u2 3 24>>1 r>r)r6r7r8r9r rkrXrrDr rrNrmrr=r>r?r2r2sy89 3933p04 G+, G+-- r>r2zD:%Y%m%d%H%M%Sdtc|jt}d}|j}|G|j}|sd}n2d}|dkr d}t |}t |d\}}|dz}|d||fzz}t ||zS) a Convert a datetime object into a PDF string. This function supports both timezone-aware and naive datetime objects. :param dt: The datetime object to convert. :return: A :class:`TextStringObject` representing the datetime passed in. rrNr rr"i<z %02d'%02d')strftime ASN_DT_FORMAT utcoffset total_secondsabsdivmodr')rbase_dtutc_offset_string utc_offset tz_secondssignhrsminss r?r/r/Oskk-(GJ--/  # DA~ _ $Z6OC#D !% T{(B C  G&77 88r>z ^D:(\d{4})z^(?:D:)?(\d{4})z^(\d\d)z(\d\d)(?:'(\d\d))?'?date_strrc J|rtntj|}|st|dt |j d}||j dd}gd}tdD]=}tj|}|sn$t |j d||<|dd}?|\}}} } } d} |r|d} | dk(rt|dk(r td}n| dvr{|dd}tj|}|std |d |t |j d}t |j dxsd}t|| }| d k(r| }ntd |dt|} t|||| | | d| S#t$r}td|d}~wwxYw)Nz% does not appear to be a date string.rr)rrrrrrvr5rN)r r"z#Improper timezone specification in z: )hoursminutesr"z Improper trailing characters in r#)yearmonthdayhourminutesecond microsecondtzinfozImproper date value)MIN_DATE_REGEXMIN_DATE_REGEX_LENIENTrrrjgrouprr'TWO_DIGIT_STARTrr UTC_OFFSET fullmatchrrr)rrr[rdate_remaining lower_orderixrrrrrtz_infosgn tz_offsettz_spectz_matchtz_hours tz_minutesrs r?parse_pdf_daterys!'=DDXNA hZ'LMNN qwwqz?DaeeAhj)N!KAh,  ! !. 1 aggaj/ B'+ ,(3$E3ffGQ #:#n-2!! I J $QR(G!++G4H"9(2gYO8>>!,-HX^^A.3!4J!*EIcz&J !A(1MN N9% 5   50!445s5F F" FF"r)T)er9rrGrenumloggingr~rtyping dataclassesrrrrriorr r r r r rmiscrrrrrrrrrrr TYPE_CHECKING crypt.apir__all__rrrrr getLoggerr6rEnumr1rGrrrrr,r r"r#rr)r!rr$rjr%rrrr.rrr.rr&r0rr'rgr{r(rrr*r+rArErr enumerater r-rrr2rr/rrrrr)rchars00r?rst   (22BB     * 2 2::;BB7KL   8 $>1""2. $///fJOJ,JBFJJZB"B"J@'I'TL$ L^p Yp f1'//91@&53 &5XC #ui' (C 12C0'  12@;%;| : 12 :ui4=+=+@ *uUI%56 *;M *LsIL^  3 U3 |3 l&(i&(RS*tYS*lS%#S%l ""AF ?s"""09/0JKHBtRxK  /1A; O x 9x v!  9 9. 9HM*#$67"**Z( RZZ/ 0 65S65$65(65kLs J