hj0SSKJr SSKJr SSKJr SSKJr SSKJ r SSK J r J r J r Jr SSKJrJr SS KJrJrJr "S S 5r"S S 5r\\\4r\ \r"SS5rg)) annotations)aliases)sha256)dumps)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$Sj5r\S%Sj5r\S#Sj5r\S"Sj5r\S"Sj5rS&S'Sjjr\S Sj5rSrg)( CharsetMatch NcXlX lX0lXPlX@lSUl/UlSUlSUlSUl X`l Xpl g)N) _payload _encoding_mean_mess_ratio _languages_has_sig_or_bom_unicode_ranges_leaves_mean_coherence_ratio_output_payload_output_encoding_string_preemptive_declaration)selfpayloadguessed_encodingmean_mess_ratiohas_sig_or_bom languagesdecoded_payloadpreemptive_declarations ;/usr/lib/python3/dist-packages/charset_normalizer/models.py__init__CharsetMatch.__init__sQ ' .'6,5%315+- ,/"-1,0#2 3I$c[U[5(d.[U[5(a[U5UR:H$gURUR:H=(a UR UR :H$)NF) isinstancerstrrencoding fingerprintr#others r+__eq__CharsetMatch.__eq__*s\%..%%% '4==88}}.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{Gz?g{Gz?) r0r ValueErrorabschaos coherencelenrrmulti_byte_usage)r#r5chaos_differencecoherence_differences r+__lt__CharsetMatch.__lt__1s%.. "%djj5;;&>"?&)$..5??*J&K d "';d'B>>EOO3 3  $)=)E4==!%55zzEKK//((5+A+AA AzzEKK''r.c\S[[U55[UR5- - $)Ng?)r=r1rawr#s r+r>CharsetMatch.multi_byte_usageGs"c#d)ns488}455r.cURc&[URURS5UlUR$)Nstrict)r!r1rrrEs r+__str__CharsetMatch.__str__Ks. << t}}dnnhGDL||r.c<SURSURS3$)Nz)r2r3rEs r+__repr__CharsetMatch.__repr__Qs" x8H8H7ILLr.c[U[5(aX:Xa$[SRUR55eSUlUR RU5 g)Nz;Unable to add instance <{}> as a submatch of a CharsetMatch)r0rr9format __class__r!rappendr4s r+ add_submatchCharsetMatch.add_submatchTsP%..%-MTTOO    E"r.cUR$N)rrEs r+r2CharsetMatch.encoding_s ~~r.c/n[R"5HKup#URU:XaURU5 M(URU:XdM:URU5 MM U$)zr Encoding name are known by many name, using this could help when searching for IBM855 when it's listed as CP855. )ritemsr2rQ)r# also_known_asups r+encoding_aliasesCharsetMatch.encoding_aliasescsW $& MMODA}}!$$Q'!#$$Q' $ r.cUR$rUrrEs r+bomCharsetMatch.bomp###r.cUR$rUr_rEs r+byte_order_markCharsetMatch.byte_order_marktrbr.cJURVs/sHoSPM sn$s snf)z Return the complete list of possible languages found in decoded sequence. Usually not really useful. Returned list may be empty even if 'language' property return something != 'Unknown'. rr)r#es r+r(CharsetMatch.languagesxs" #oo.o!o...s c,UR(drSUR;agSSKJnJn [ UR 5(aU"UR 5OU"UR 5n[U5S:XdSU;agUS$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.cdrmrnrr2r=)r#rmrnr(s r+languageCharsetMatch.languages $444  X*$--88&dmm4' 6  9~"my&@ Q< q!!$$r.cUR$rU)rrEs r+r;CharsetMatch.chaoss$$$r.cJUR(dgURSS$)Nrrr rgrEs r+r<CharsetMatch.coherences q!!$$r.c0[URS-SS9$Nd)ndigits)roundr;rEs r+ percent_chaosCharsetMatch.percent_chaossTZZ#%q11r.c0[URS-SS9$ry)r}r<rEs r+percent_coherenceCharsetMatch.percent_coherencesT^^c)155r.cUR$)z Original untouched bytes. )rrEs r+rDCharsetMatch.raws }}r.cUR$rU)rrEs r+submatchCharsetMatch.submatchs ||r.c2[UR5S:$Nr)r=rrEs r+ has_submatchCharsetMatch.has_submatchs4<< 1$$r.c URb UR$[U5Vs/sHn[U5PM nn[[ UVs1sH o3(dM UiM sn55UlUR$s snfs snfrU)rr1rsortedlist)r#chardetected_rangesrs r+ alphabetsCharsetMatch.alphabetssp    +'' 'MPQUY,WYT]4-@Y,W%d+L!!A+L&MN###-X+LsA< B B ctUR/URVs/sHoRPM sn-$s snf)z The complete list of encoding that output the exact SAME str result and therefore could be the originating encoding. This list does include the encoding available in property 'encoding'. )rrr2)r#ms r+rp"CharsetMatch.could_be_from_charsets.t||"D|!::|"DDD"Ds5c:^TRbTRU:warUTl[T5nTRb<TRR5S;a[ [ U4SjUSSSS9nX2SS-nUR US5TlTR$)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$)Nrr _-)stringspanreplacegroupsrr )rr#s r+%CharsetMatch.output..sWahhqvvx{QVVXa[AII 1 !$"7"78@@cJr.i r )countr)r r1r"lowerrr encoder)r#r2decoded_stringpatched_headers` r+outputCharsetMatch.outputs  (D,A,AX,M$,D ! YN,,80066812"%3#5D)""02G!G#1#8#89#MD ###r.cP[UR55R5$)zg Retrieve the unique SHA256 computed using the transformed (re-encoded) payload. Not the original one. )rr hexdigestrEs r+r3CharsetMatch.fingerprints dkkm$..00r.) rrrrrrr rrr"r!r)NN)r$bytesr%r1r&floatr'boolr(CoherenceMatchesr) str | Noner*r)r5objectreturnr)rrrr1)r5rrNone)r list[str]rr)rr)rzlist[CharsetMatch])r)r2r1rr)__name__ __module__ __qualname____firstlineno__r,r6rApropertyr>rIrLrRr2r\r`rdr(rrr;r<r~rrDrrrrprr3__static_attributes__r.r+rr s'+-1JJJ J  J $ J$J!+J8Y(,66 M #  $$$$//%%6%%%% 2266 %%$$EE$:11r.rcl\rSrSrSrS SSjjrSSjrSSjrSSjrSSjr SS jr SS jr SS jr S r g)CharsetMatchesz Container with every CharsetMatch items ordered by default from most probable to the less one. Act like a list(iterable) but does not implements all related methods. NcBU(a[U5Ulg/UlgrU)r_results)r#resultss r+r,CharsetMatches.__init__s?FF7O B r.c#8# URShvN gN7frUrrEs r+__iter__CharsetMatches.__iter__s==  s c[U[5(aURU$[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)r0intrr1rrpKeyError)r#itemresults r+ __getitem__CharsetMatches.__getitem__s_ dC ==& & dC T5)D--777!M(r.c,[UR5$rUr=rrEs r+__len__CharsetMatches.__len__s4==!!r.c2[UR5S:$rrrEs r+__bool__CharsetMatches.__bool__s4==!A%%r.c[U[5(d-[SR[ UR 555e[ UR5[:a\URHLnURUR:XdMURUR:XdM;URU5 g URRU5 [UR5Ul g)zf Insert a single match. Will be inserted accordingly to preserve sort. Can be inserted as a submatch. z-Cannot append instance '{}' to CharsetMatchesN)r0rr9rOr1rPr=rDrrr3r;rRrQr)r#rmatchs r+rQCharsetMatches.appends $ --?FF'  txx=+ +$$(8(88U[[DJJ=V&&t,' T"t}}- r.cDUR(dgURS$)zA Simply return the first match. Strict equivalent to matches[0]. NrrrEs r+bestCharsetMatches.best)s}}}}Qr.c"UR5$)z@ Redundant method, call the method best(). Kept for BC reasons. )rrEs r+firstCharsetMatches.first1syy{r.rrU)rzlist[CharsetMatch] | None)rzIterator[CharsetMatch])rz int | strrr)rrr)rrrr)rzCharsetMatch | None)rrrr__doc__r,rrrrrQrrrrr.r+rrs0 O! "&.( r.rch\rSrSrSSjr\SSj5rS SjrSrg) CliDetectionResulti<c XlXlX lX0lX@lXPlX`lXplXlXl Xl grU) path unicode_pathr2r\alternative_encodingsrrrr'r;r< is_preferred) r#rr2r\rrrrr'r;r<rrs r+r,CliDetectionResult.__init__=s@ (4$, +;0E"% $-$2! )".r.c URURURURURUR UR URURURURS. $)N rr2r\rrrrr'r;r<rrrrEs r+__dict__CliDetectionResult.__dict__WseII $ 5 5%)%?%? "11ZZ -- --  r.c,[URSSS9$)NT) ensure_asciiindent)rrrEs r+to_jsonCliDetectionResult.to_jsongsT]]a@@r.) rrr;r<r2r\r'rrrrrN)rr1r2rr\rrrrrr1rrr'rr;rr<rrrrr)rzdict[str, Any]r) rrrrr,rrrrrr.r+rr<s///$ / ) /  /////!//4    Ar.rN) __future__rencodings.aliasesrhashlibrjsonrrertypingrr r r constantr rutilsrrrrrr1rCoherenceMatchrrrr.r+rsd"%--GCCe1e1P@@FsEz"',A,Ar.