:'a0ddlZddlmZddlmZddlmZddlmZm Z m Z m Z ddl m Z ddlmZmZddlmZdd lmZdd lmZmZmZGd d ZGd dZe eefZe eZGddZ eZ!y)N)aliases)sha256)dumps)OptionalListTupleSet)Counter)subcompile)TOO_BIG_SEQUENCE) mess_ratio) iana_nameis_multi_byte_encoding unicode_rangecPeZdZ d'dededededddeef d Zd efd Z d efd Z e d efd Z e d efdZ e d efdZd efdZd efdZd(dZe d efdZe d eefdZe d efdZe d efdZe d eefdZe d efdZe d efdZe d efdZe d efdZe d efdZe d efdZe d edfdZe d efdZ e d eefd Z!e d eefd!Z"d)d"Z#d)d#Z$d*d$ed efd%Z%e d efd&Z&y)+ CharsetMatchNpayloadguessed_encodingmean_mess_ratiohas_sig_or_bom languagesCoherenceMatchesdecoded_payloadc||_||_||_||_||_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)selfrrrrrrs 9lib/python3.12/site-packages/charset_normalizer/models.py__init__zCharsetMatch.__init__sW  ) /#-# %'"# $& returnc t|tsAtdjt |j t |j |j |j k(xr|j|jk(S)Nz&__eq__ cannot be invoked on {} and {}.) isinstancer TypeErrorformatstr __class__encoding fingerprintr(others r)__eq__zCharsetMatch.__eq__(sg%.DKKCPUP_P_L`befjftftbuvw w}}.X43C3CuGXGX3XXr+ct|tstt|j|jz }|dkr|j |j kDS|j|jkS)zQ Implemented to make sorted available upon CharsetMatches items. g{Gz?)r.r ValueErrorabschaos coherence)r(r6chaos_differences r)__lt__zCharsetMatch.__lt__-s\%. tzzEKK78 d ">>EOO3 3zzEKK''r+c`tjdttt |dS)z Check once again chaos in decoded text, except this time, with full content. Use with caution, this can be very slow. Notice: Will be removed in 3.0 z=chaos_secondary_pass is deprecated and will be removed in 3.0g?)warningswarnDeprecationWarningrr1r(s r)chaos_secondary_passz!CharsetMatch.chaos_secondary_pass<s+  UWij I   r+c8tjdty)zy Coherence ratio on the first non-latin language detected if ANY. Notice: Will be removed in 3.0 z)r0r3r4rCs r)__repr__zCharsetMatch.__repr__ds.55dmmTEUEUVVr+ct|tr||k(r$tdj|jd|_|j j|y)Nz;Unable to add instance <{}> as a submatch of a CharsetMatch)r.rr9r0r2r'r#appendr5s r) add_submatchzCharsetMatch.add_submatchgsI%.%4-Zaabgbqbqrs s  E"r+c|jSN)rrCs r)r3zCharsetMatch.encodingns ~~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. )ritemsr3rU)r( also_known_asups r)encoding_aliaseszCharsetMatch.encoding_aliasesrs^  MMO (DAq}}!$$Q'!#$$Q'  ( r+c|jSrXr!rCs r)bomzCharsetMatch.bom###r+c|jSrXr`rCs r)byte_order_markzCharsetMatch.byte_order_markrbr+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)rzCharsetMatch.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)mb_encoding_languagesencoding_languagesz Latin BasedUnknown)r could_be_from_charsetcharset_normalizer.cdrkrlrr3len)r(rkrlrs r)languagezCharsetMatch.languages $444  X@VW[WdWd@e-dmm Kept for BC reasons. Will be removed in 3.0. rCs r)firstzCharsetMatch.first  r+c|SrrrCs r)bestzCharsetMatch.bestrr+r3c|j|j|k7r'||_t|j|d|_|jS)z Method to get re-encoded bytes payload using given target encoding. Default to UTF-8. Any errors will be simply ignored by the encoder NOT replaced. replace)r&r1encoder%)r(r3s r)outputzCharsetMatch.outputsJ  (D,A,AX,M$,D !#&t9#3#3Hi#HD ###r+cPt|jjS)zw Retrieve the unique SHA256 computed using the transformed (re-encoded) payload. Not the original one. )rr hexdigestrCs r)r4zCharsetMatch.fingerprints dkkm$..00r+rX)r6rr,N)r,r)utf_8)'__name__ __module__ __qualname__bytesr1floatboolrr*r7r>propertyrDrFr rNrQrSrVr3rr^rardrrqr;r<r{r}rrrrrnrrrr4rr+r)rrs.2''"'# ' ! ' * '&c]'2YtY (t (  e    U 67 6 6 W#W## $s)  $T$$$$$/49//%#%%.%u%%%5%% 2u226566U $~.%d%% $49 $ $EtCyEE  $s $ $1S11r+rcveZdZdZddeefdZdZdefdZde fdZ d eddfd Z de d fd Z de d fd Zy)CharsetMatchesz Container with every CharsetMatch items ordered by default from most probable to the less one. Act like a list(iterable) but does not implements all related methods. Nresultsc8|rt||_yg|_yrX)r_results)r(rs r)r*zCharsetMatches.__init__s+2w  r+c#6K|jD]}|ywrXr)r(results r)__iter__zCharsetMatches.__iter__smm FL sr,ct|tr|j|St|tr/t |d}|jD]}||j vs|cSt )z Retrieve a single item either by its position or encoding name (alias may be used here). Raise KeyError upon invalid index or encoding not present in results. F)r.intrr1rrnKeyError)r(itemrs r) __getitem__zCharsetMatches.__getitem__ s` dC ==& & dC T5)D-- "6777!M "r+c,t|jSrX)rprrCs r)__len__zCharsetMatches.__len__s4==!!r+rct|ts-tdjt |j t |jtkrW|jD]H}|j|jk(s|j|jk(s7|j|y|jj|t|j|_ y)z~ Insert a single match. Will be inserted accordingly to preserve sort. Can be inserted as a submatch. z-Cannot append instance '{}' to CharsetMatchesN)r.rr9r0r1r2rprr rr4r;rVrUr)r(rmatchs r)rUzCharsetMatches.appends $ -LSSTWX\XfXfTghi i txx=, , $$(8(88U[[DJJ=V&&t,  T"t}}- r+rc:|jsy|jdS)zQ Simply return the first match. Strict equivalent to matches[0]. NrrrCs r)rzCharsetMatches.best+s}}}}Qr+c"|jS)zP Redundant method, call the method best(). Kept for BC reasons. )rrCs r)rzCharsetMatches.first3syy{r+rX)rrr__doc__rrr*rrrrrUrrrrr+r)rrsj;\ 2; < "".<.D.  h~. x/r+rcteZdZdededeedeededeededed ed eed efd Ze d Z defdZ y)CliDetectionResultpathr3r^alternative_encodingsrqrrr;r< unicode_path is_preferredc ||_| |_||_||_||_||_||_||_||_| |_ | |_ yrX) rrr3r^rrqrrr;r<r) r(rr3r^rrqrrr;r<rrs r)r*zCliDetectionResult.__init__@sT (  0%:"  ", "(r+c |j|j|j|j|j|j |j |j|j|j|jd S)N rr3r^rrqrrr;r<rrrrCs r)__dict__zCliDetectionResult.__dict__MseII $ 5 5%)%?%? "11ZZ -- --  r+r,c2t|jddS)NT) ensure_asciiindent)rrrCs r)to_jsonzCliDetectionResult.to_json]s MM  r+N) rrrr1rrrrr*rrrrr+r)rr>s )S )C )49 )eijmen )z} )KOPSKT )fj )sx )EJ )ZbcfZg )w{ )      r+r)"r@encodings.aliasesrhashlibrjsonrtypingrrrr collectionsr rer r rIcharset_normalizer.constantr charset_normalizer.mdrcharset_normalizer.utilsrrrrrr1rCoherenceMatchrrCharsetNormalizerMatchrr+r)rso%--)8,UUm1m1`99xsEz"'$ $ N&r+