*pj4SSKJr SSKJr SSKJr SSKJrJrJ r J r SSK J r J r SSKJrJrJr "SS 5r"S S 5r\ \\4r\ \r"S S 5rg)) annotations)aliases)sub)AnyIteratorListTuple)RE_POSSIBLE_ENCODING_INDICATIONTOO_BIG_SEQUENCE) iana_nameis_multi_byte_encoding unicode_rangec\rSrSrSSSjjrSSjrSSjr\SSj5rS Sjr S Sjr S!S jr \S S j5r \S"S j5r \S#S j5r\S#S j5r\S"Sj5r\S Sj5r\SSj5r\SSj5r\SSj5r\SSj5r\S$Sj5r\S%Sj5r\S#Sj5r\S"Sj5r\S"Sj5rS&S'Sjjr\S(Sj5rSrg)) CharsetMatch NcXlX lX0lXPlX@lSUl/UlSUlSUlSUl X`l Xpl g)N) _payload _encoding_mean_mess_ratio _languages_has_sig_or_bom_unicode_ranges_leaves_mean_coherence_ratio_output_payload_output_encoding_string_preemptive_declaration)selfpayloadguessed_encodingmean_mess_ratiohas_sig_or_bom languagesdecoded_payloadpreemptive_declarations U/var/www/html/pdf-tiff/venv/lib/python3.13/site-packages/charset_normalizer/models.py__init__CharsetMatch.__init__ sQ,3 .'6,5%315+- ,/"-1,0#2 3I$c[U[5(d/[U[5(a[US5UR:H$gURUR:H=(a UR UR :H$)NF) isinstancerstrr encoding fingerprintr!others r)__eq__CharsetMatch.__eq__(s`%..%%% !.$--??}}.X43C3CuGXGX3XXr,c[U[5(d[e[URUR- 5n[UR UR - 5nUS:aUS:aUR UR :$US:aUUS::aO[ UR5[:aURUR:$URUR:$URUR:$)zA Implemented to make sorted available upon CharsetMatches items. g{Gzt?g{Gz?) r.r ValueErrorabschaos coherencelenrr multi_byte_usage)r!r3chaos_differencecoherence_differences r)__lt__CharsetMatch.__lt__3s%.. "%djj5;;&>"?&)$..5??*J&K e #(>EOO3 3  %*>$*F4==!%55zzEKK//((5+A+AA AzzEKK''r,cn[UR5nUS:XagS[[U55U- - $)Nrrg?)r;rawr/)r!raw_lens r)r<CharsetMatch.multi_byte_usageIs3dhh- a<c#d)nw.//r,c2URc[URURS5UlUR(aHURS:Xa8UR(a'URSS:XaURSSUlUR$)Nstrictutf_7rur )rr/rrrr!s r)__str__CharsetMatch.__str__Qsn << t}}dnnhGDL $$NNg-LLLLOx/#||AB/ ||r,c<SURSURS3$)Nz)r0r1rHs r)__repr__CharsetMatch.__repr__as" uT5E5E4FbIIr,c[U[5(aX:Xa$[SRUR55eSUlUR RU5 g)Nz;Unable to add instance <{}> as a submatch of a CharsetMatch)r.rr7format __class__rrappendr2s r) add_submatchCharsetMatch.add_submatchdsP%..%-MTTOO    E"r,cUR$N)rrHs r)r0CharsetMatch.encodingos ~~r,c/n[R"5HKup#URU:XaURU5 M(URU:XdM:URU5 MM U$)zr Encoding name are known by many name, using this could help when searching for IBM855 when it's listed as CP855. )ritemsr0rQ)r! also_known_asups r)encoding_aliasesCharsetMatch.encoding_aliasesssW $& MMODA}}!$$Q'!#$$Q' $ r,cUR$rUrrHs r)bomCharsetMatch.bom###r,cUR$rUr_rHs r)byte_order_markCharsetMatch.byte_order_markrbr,cJURVs/sHoSPM sn$s snf)z Return the complete list of possible languages found in decoded sequence. Usually not really useful. Returned list may be empty even if 'language' property return something != 'Unknown'. rr)r!es r)r&CharsetMatch.languagess" #oo.o!o...s c,UR(drSUR;agSSKJnJn [ UR 5(aU"UR 5OU"UR 5n[U5S:XdSU;agUS$URSS$)zz Most probable language found in decoded sequence. If none were detected or inferred, the property will return "Unknown". asciiEnglishr)encoding_languagesmb_encoding_languagesz Latin BasedUnknown)rcould_be_from_charsetcharset_normalizer.cdrmrnrr0r;)r!rmrnr&s r)languageCharsetMatch.languages $444  X*$--88&dmm4' 6  9~"my&@ Q< q!!$$r,cUR$rU)rrHs r)r9CharsetMatch.chaoss$$$r,cJUR(dgURSS$)Nrrr rgrHs r)r:CharsetMatch.coherences q!!$$r,c0[URS-SS9$Nd)ndigits)roundr9rHs r) percent_chaosCharsetMatch.percent_chaossTZZ#%q11r,c0[URS-SS9$ry)r}r:rHs r)percent_coherenceCharsetMatch.percent_coherencesT^^c)155r,cUR$)z Original untouched bytes. )rrHs r)rBCharsetMatch.raws }}r,cUR$rU)rrHs r)submatchCharsetMatch.submatchs ||r,c2[UR5S:$Nr)r;rrHs r) has_submatchCharsetMatch.has_submatchs4<< 1$$r,c URb UR$[U5Vs/sHn[U5PM nn[[ UVs1sH o3(dM UiM sn55UlUR$s snfs snfrU)rr/rsortedlist)r!chardetected_rangesrs r) alphabetsCharsetMatch.alphabetssp    +'' 'MPQUY,WYT]4-@Y,W%d+L!!A+L&MN###-X+LsA< B B ctUR/URVs/sHoRPM sn-$s snf)z The complete list of encoding that output the exact SAME str result and therefore could be the originating encoding. This list does include the encoding available in property 'encoding'. )rrr0)r!ms r)rp"CharsetMatch.could_be_from_charsets.t||"D|!::|"DDD"Ds5c:^TRbTRU:warUTl[T5nTRb<TRR5S;a[ [ U4SjUSSSS9nX2SS-nUR US5TlTR$)z Method to get re-encoded bytes payload using given target encoding. Default to UTF-8. Any errors will be simply ignored by the encoder NOT replaced. N)zutf-8utf8utf_8c>URUR5SUR5SRUR5S[ TR 5RSS55$)Nrr _-)stringspanreplacegroupsr r)rr!s r)%CharsetMatch.output..sWahhqvvx{QVVXa[AII 1 !$"7"78@@cJr,i r )countr)rr/r lowerrr encoder)r!r0decoded_stringpatched_headers` r)outputCharsetMatch.outputs  (D,A,AX,M$,D ! YN,,80066812"%3#5D)""02G!G#1#8#89#MD ###r,c*[[U55$)zM Retrieve a hash fingerprint of the decoded payload, used for deduplication. )hashr/rHs r)r1CharsetMatch.fingerprints CIr,) rrrrrrrrrr rr)NN)r"bytes | bytearrayr#r/r$floatr%boolr&CoherenceMatchesr' str | Noner(r)r3objectreturnr)rrrr/)r3rrNone)r list[str]rr)rr)rzlist[CharsetMatch])r)r0r/rbytesrint)__name__ __module__ __qualname____firstlineno__r*r4r?propertyr<rIrLrRr0r\r`rdr&rrr9r:r~rrBrrrrprr1__static_attributes__r,r)rr s'+-1J"JJ J  J $ J$J!+J8 Y(,00 J #  $$$$//%%6%%%% 2266 %%$$EE$:r,rcv\rSrSrSrSSSjjrSSjrSSjrSSjrSSjr SS jr SS jr SS jr SS jr S rg)CharsetMatchesiz Container with every CharsetMatch items ordered by default from most probable to the less one. Act like a list(iterable) but does not implements all related methods. NcDU(a [U5O/UlSUlgNT)r_results _is_sorted)r!resultss r)r*CharsetMatches.__init__ s?FF7OB $r,cjUR(d"URR5 SUlggr)rrsortrHs r)_ensure_sortedCharsetMatches._ensure_sorteds% MM   "DOr,c#X# UR5 URShvN gN7frU)rrrHs r)__iter__CharsetMatches.__iter__s ==  s *(*c[U[5(aUR5 URU$[U[5(a2[ US5nURHnXR ;dMUs $ [e)z Retrieve a single item either by its position or encoding name (alias may be used here). Raise KeyError upon invalid index or encoding not present in results. F)r.rrrr/r rpKeyError)r!itemresults r) __getitem__CharsetMatches.__getitem__sk dC    !==& & dC T5)D--777!M(r,c,[UR5$rUr;rrHs r)__len__CharsetMatches.__len__'s4==!!r,c2[UR5S:$rrrHs r)__bool__CharsetMatches.__bool__*s4==!A%%r,c[U[5(d-[SR[ UR 555e[ UR5[:a\URHLnURUR:XdMURUR:XdM;URU5 g URRU5 SUlg)zf Insert a single match. Will be inserted accordingly to preserve sort. Can be inserted as a submatch. z-Cannot append instance '{}' to CharsetMatchesNF)r.rr7rOr/rPr;rBr rr1r9rRrQr)r!rmatchs r)rQCharsetMatches.append-s $ --?FF'  txx=+ +$$(8(88U[[DJJ=V&&t,' T"r,cdUR(dgUR5 URS$)zA Simply return the first match. Strict equivalent to matches[0]. Nr)rrrHs r)bestCharsetMatches.bestAs)}} }}Qr,c"UR5$)z@ Redundant method, call the method best(). Kept for BC reasons. )rrHs r)firstCharsetMatches.firstJsyy{r,)rrrU)rzlist[CharsetMatch] | None)rr)rzIterator[CharsetMatch])rz int | strrrrr)rrrr)rzCharsetMatch | None)rrrr__doc__r*rrrrrrQrrrrr,r)rrs4 %# ! "& ( r,rch\rSrSrSSjr\SSj5rS SjrSrg) CliDetectionResultiUc XlXlX lX0lX@lXPlX`lXplXlXl Xl grU) path unicode_pathr0r\alternative_encodingsrrrr%r9r: is_preferred) r!rr0r\rrrrr%r9r:rrs r)r*CliDetectionResult.__init__Vs@ (4$, +;0E"% $-$2! )".r,c URURURURURUR UR URURURURS. $)N rr0r\rrrrr%r9r:rrrrHs r)__dict__CliDetectionResult.__dict__pseII $ 5 5%)%?%? "11ZZ -- --  r,c2SSKJn U"URSSS9$)Nr)dumpsT) ensure_asciiindent)jsonrr)r!rs r)to_jsonCliDetectionResult.to_jsonsT]]a@@r,) rrr9r:r0r\r%rrrrrN)rr/r0rr\rrrrrr/rrr%rr9rr:rrrrr)rzdict[str, Any]r) rrrrr*rrrrrr,r)rrUs///$ / ) /  /////!//4    Ar,rN) __future__rencodings.aliasesrrertypingrrrr constantr r utilsr rrrrr/rCoherenceMatchrrrr,r)rs^"%--GCCwwtIIXsEz"'.A.Ar,