Dj4ddlmZddlmZddlmZddlmZmZm Z m Z ddl m Z m Z ddlmZmZmZGdd ZGd d Ze eefZe eZGd d Zy)) annotations)aliases)sub)AnyIteratorListTuple)RE_POSSIBLE_ENCODING_INDICATIONTOO_BIG_SEQUENCE) iana_nameis_multi_byte_encoding unicode_rangeceZdZ d ddZddZddZeddZddZddZ ddZ edd Z ed d Z ed!d Z ed!d Zed d ZeddZeddZeddZeddZeddZed"dZed#dZed!dZed dZed dZd$d%dZed&dZy)' CharsetMatchNc||_||_||_||_||_d|_g|_d|_d|_d|_ ||_ ||_ y)N) _payload _encoding_mean_mess_ratio _languages_has_sig_or_bom_unicode_ranges_leaves_mean_coherence_ratio_output_payload_output_encoding_string_preemptive_declaration)selfpayloadguessed_encodingmean_mess_ratiohas_sig_or_bom languagesdecoded_payloadpreemptive_declarations W/root/workspace/ytshorts/venv/lib/python3.12/site-packages/charset_normalizer/models.py__init__zCharsetMatch.__init__ s_,3 .'6,5%315+- ,/"-1,0#2 3I$ct|ts*t|trt|d|jk(Sy|j|jk(xr|j |j k(S)NF) isinstancerstrr encoding fingerprintr others r(__eq__zCharsetMatch.__eq__(sZ%.%% !.$--??}}.X43C3CuGXGX3XXr*ct|tstt|j|jz }t|j |j z }|dkr|dkDr|j |j kDS|dkrS|dkrNt |jtk\r|j|jkS|j|jkDS|j|jkS)zQ Implemented to make sorted available upon CharsetMatches items. g{Gzt?g{Gz?) r,r ValueErrorabschaos coherencelenrr multi_byte_usage)r r1chaos_differencecoherence_differences r(__lt__zCharsetMatch.__lt__3s%. "%djj5;;&>"?&)$..5??*J&K e #(>EOO3 3  %*>$*F4==!%55zzEKK//((5+A+AA AzzEKK''r*clt|j}|dk(rydtt||z z S)Nrrg?)r8rawr-)r raw_lens r(r9zCharsetMatch.multi_byte_usageIs3dhh- a<c#d)nw.//r*c|jst|j|jd|_|jrA|jdk(r2|jr&|jddk(r|jdd|_|jS)Nstrictutf_7rur )rr-rrrr s r(__str__zCharsetMatch.__str__Qsn << t}}dnnhGDL $$NNg-LLLLOx/#||AB/ ||r*c<d|jd|jdS)Nz)r.r/rCs r(__repr__zCharsetMatch.__repr__as" uT5E5E4FbIIr*ct|tr||k(r$tdj|jd|_|j j|y)Nz;Unable to add instance <{}> as a submatch of a CharsetMatch)r,rr4format __class__rrappendr0s r( add_submatchzCharsetMatch.add_submatchdsO%.%4-MTTOO    E"r*c|jSN)rrCs r(r.zCharsetMatch.encodingos ~~r*cg}tjD]G\}}|j|k(r|j|'|j|k(s7|j|I|S)z Encoding name are known by many name, using this could help when searching for IBM855 when it's listed as CP855. )ritemsr.rJ)r also_known_asups r(encoding_aliaseszCharsetMatch.encoding_aliasesss^ $& MMO (DAq}}!$$Q'!#$$Q'  ( r*c|jSrMrrCs r(bomzCharsetMatch.bom###r*c|jSrMrUrCs r(byte_order_markzCharsetMatch.byte_order_markrWr*cF|jDcgc]}|d c}Scc}w)z Return the complete list of possible languages found in decoded sequence. Usually not really useful. Returned list may be empty even if 'language' property return something != 'Unknown'. rr)r es r(r%zCharsetMatch.languagess #oo.!...s c|jshd|jvryddlm}m}t |j r||j n||j }t|dk(sd|vry|dS|jddS)z Most probable language found in decoded sequence. If none were detected or inferred, the property will return "Unknown". asciiEnglishr)encoding_languagesmb_encoding_languagesz Latin BasedUnknown)rcould_be_from_charsetcharset_normalizer.cdr`rarr.r8)r r`rar%s r(languagezCharsetMatch.languages $444  X*$--8&dmm4' 6  9~"my&@ Q< q!!$$r*c|jSrM)rrCs r(r6zCharsetMatch.chaoss$$$r*c@|jsy|jddS)Nrrr r[rCs r(r7zCharsetMatch.coherences q!!$$r*c6t|jdzdSNd)ndigits)roundr6rCs r( percent_chaoszCharsetMatch.percent_chaossTZZ#%q11r*c6t|jdzdSri)rmr7rCs r(percent_coherencezCharsetMatch.percent_coherencesT^^c)155r*c|jS)z+ Original untouched bytes. )rrCs r(r>zCharsetMatch.raws }}r*c|jSrM)rrCs r(submatchzCharsetMatch.submatchs ||r*c2t|jdkDSNr)r8rrCs r( has_submatchzCharsetMatch.has_submatchs4<< 1$$r*c|j |jSt|Dcgc] }t|}}tt |Dchc]}|s| c}|_|jScc}wcc}wrM)rr-rsortedlist)r chardetected_rangesrs r( alphabetszCharsetMatch.alphabetssk    +'' 'MPQUY,WT]4-@,W,W%d+L!!A+L&MN###-X+LsA0A5 A5 cp|jg|jDcgc]}|jc}zScc}w)z The complete list of encoding that output the exact SAME str result and therefore could be the originating encoding. This list does include the encoding available in property 'encoding'. )rrr.)r ms r(rcz"CharsetMatch.could_be_from_charsets,t||"D!1::"DDD"Ds3c6jj|k7rr|_t}j=jjdvr!t t fd|ddd}||ddz}|j |d_jS)z Method to get re-encoded bytes payload using given target encoding. Default to UTF-8. Any errors will be simply ignored by the encoder NOT replaced. N)zutf-8utf8utf_8c|j|jd|jdj|jdt j jddS)Nrr _-)stringspanreplacegroupsr r)rr s r(z%CharsetMatch.output..sYahhqvvx{QVVXa[AII 1 !$"7"78@@cJr*i r )countr)rr-rlowerrr encoder)r r.decoded_stringpatched_headers` r(outputzCharsetMatch.outputs  (D,A,AX,M$,D ! YN,,80066812"%3#5D)""0.2G!G#1#8#89#MD ###r*c*tt|S)z] Retrieve a hash fingerprint of the decoded payload, used for deduplication. )hashr-rCs r(r/zCharsetMatch.fingerprints CIr*)NN)r!bytes | bytearrayr"r-r#floatr$boolr%CoherenceMatchesr& str | Noner'r)r1objectreturnr)rrrr-)r1rrNone)r list[str]rr)rr)rzlist[CharsetMatch])r)r.r-rbytesrint)__name__ __module__ __qualname__r)r2r<propertyr9rDrFrKr.rSrVrYr%rer6r7rnrpr>rsrvr}rcrr/r*r(rr s'+-1J"JJ J  J $ J$J!+J8 Y(,00 J #  $$$$//%%6%%%% 2266 %%$$EE$:r*rcZeZdZdZd d dZddZddZddZddZddZ dd Z dd Z dd Z y)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. Nc:|r t|ng|_d|_yNT)rx_results _is_sorted)r resultss r(r)zCharsetMatches.__init__ s?FF7OB $r*c`|js"|jjd|_yyr)rrsortrCs r(_ensure_sortedzCharsetMatches._ensure_sorteds% MM   "DOr*c#XK|j|jEd{y7wrM)rrrCs r(__iter__zCharsetMatches.__iter__s ==  s *(*ct|tr|j|j|St|tr/t |d}|jD]}||j vs|cSt)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 rcKeyError)r itemresults r( __getitem__zCharsetMatches.__getitem__sl dC    !==& & dC T5)D-- "6777!M "r*c,t|jSrMr8rrCs r(__len__zCharsetMatches.__len__'s4==!!r*c2t|jdkDSrurrCs r(__bool__zCharsetMatches.__bool__*s4==!A%%r*ct|ts-tdjt |j t |jtkrW|jD]H}|j|jk(s|j|jk(s7|j|y|jj|d|_y)z~ Insert a single match. Will be inserted accordingly to preserve sort. Can be inserted as a submatch. z-Cannot append instance '{}' to CharsetMatchesNF)r,rr4rHr-rIr8r>r rr/r6rKrJr)r rmatchs r(rJzCharsetMatches.append-s $ -?FF'  txx=+ + $$(8(88U[[DJJ=V&&t,  T"r*cZ|jsy|j|jdS)zQ Simply return the first match. Strict equivalent to matches[0]. Nr)rrrCs r(bestzCharsetMatches.bestAs)}} }}Qr*c"|jS)zP Redundant method, call the method best(). Kept for BC reasons. )rrCs r(firstzCharsetMatches.firstJsyy{r*rM)rzlist[CharsetMatch] | None)rr)rzIterator[CharsetMatch])rz int | strrrrr)rrrr)rzCharsetMatch | None) rrr__doc__r)rrrrrrJrrrr*r(rrs4 %# ! "& ( r*rcZeZdZ ddZeddZddZy)CliDetectionResultc ||_| |_||_||_||_||_||_||_||_| |_ | |_ yrM) path unicode_pathr.rSalternative_encodingsrer}r$r6r7 is_preferred) r rr.rSrrer}r$r6r7rrs r(r)zCliDetectionResult.__init__VsV (4$, +;0E"% $-$2! )".r*c |j|j|j|j|j|j |j |j|j|j|jd S)N rr.rSrrer}r$r6r7rrrrCs r(__dict__zCliDetectionResult.__dict__pseII $ 5 5%)%?%? "11ZZ -- --  r*c8ddlm}||jddS)Nr)dumpsT) ensure_asciiindent)jsonrr)r rs r(to_jsonzCliDetectionResult.to_jsonsT]]a@@r*N)rr-r.rrSrrrrer-r}rr$rr6rr7rrrrr)rzdict[str, Any]r)rrrr)rrrrr*r(rrUs///$ / ) /  /////!//4    Ar*rN) __future__rencodings.aliasesrrertypingrrrr constantr r utilsr rrrrr-rCoherenceMatchrrrr*r(rs^"%--GCCwwtIIXsEz"'.A.Ar*