:'a0ddlZddlmZddlmZddlmZddlmZm Z m Z m Z ddl m Z ddlmZmZddlmZdd lmZdd lmZmZmZGd d ZGd dZe eefZe eZGddZ eZ!dS)N)aliases)sha256)dumps)OptionalListTupleSet)Counter)subcompile)TOO_BIG_SEQUENCE) mess_ratio) iana_nameis_multi_byte_encoding unicode_rangeceZdZ 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 efd Z e d eefd!Z!e d eefd"Z"d+d#Z#d+d$Z$d,d&ed efd'Z%e d efd(Z&dS)- CharsetMatchNpayloadguessed_encodingmean_mess_ratiohas_sig_or_bom languagesCoherenceMatchesdecoded_payloadc||_||_||_||_||_d|_g|_d|_d|_d|_ ||_ dS)N) _payload _encoding_mean_mess_ratio _languages_has_sig_or_bom_unicode_ranges_leaves_mean_coherence_ratio_output_payload_output_encoding_string)selfrrrrrrs ]/mounts/lovelace/software/anaconda3/lib/python3.11/site-packages/charset_normalizer/models.py__init__zCharsetMatch.__init__s[  ) /#-# %'"# $& returnct|tsGtdt |jt |j|j|jko|j|jkS)Nz&__eq__ cannot be invoked on {} and {}.) isinstancer TypeErrorformatstr __class__encoding fingerprintr(others r)__eq__zCharsetMatch.__eq__(sl%.. xDKKCPUP_L`L`befjftbubuvvww w}.X43CuGX3XXr+ct|tstt|j|jz }|dkr|j|jkS|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^%..  tzEK788 d " ">EO3 3zEK''r+cptjdttt |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<s6  UWijjj II    r+c:tjdtdS)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.55dmTEUVVVr+r6ct|tr||kr'td|jd|_|j|dS)Nz;Unable to add instance <{}> as a submatch of a CharsetMatch)r.rr9r0r2r'r#appendr5s r) add_submatchzCharsetMatch.add_submatchgsa%.. t%4--Zaabgbqrrss s  E"""""r+c|jSN)rrCs r)r3zCharsetMatch.encodingns ~r+cg}tjD]F\}}|j|kr||&|j|kr||G|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_aliasesrsn  MOO ( (DAq}!!$$Q''''!##$$Q'''r+c|jSrXr!rCs r)bomzCharsetMatch.bom ##r+c|jSrXr`rCs r)byte_order_markzCharsetMatch.byte_order_markrbr+c$d|jDS)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'. cg|] }|d S)r).0es r) z*CharsetMatch.languages..s...!...r+r rCs r)rzCharsetMatch.languagess /.do....r+c|jshd|jvrdSddlm}m}t |jr||jn||j}t|dksd|vrdS|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.cdrorprr3len)r(rorprs r)languagezCharsetMatch.languages  $444 y X W W W W W W W@VW[Wd@e@eM--dm<<.s"D"D"D!1:"D"D"Dr+)rr#rCs r)rrz"CharsetMatch.could_be_from_charsets%"D"Dt|"D"D"DDDr+c|Sz> Kept for BC reasons. Will be removed in 3.0. rgrCs r)firstzCharsetMatch.first  r+c|SrrgrCs r)bestzCharsetMatch.bestrr+utf_8r3c|j |j|kr/||_t||d|_|jS)z Method to get re-encoded bytes payload using given target encoding. Default to UTF-8. Any errors will be simply ignored by the encoder NOT replaced. Nreplace)r&r1encoder%)r(r3s r)outputzCharsetMatch.outputsI  (D,AX,M,M$,D !#&t99#3#3Hi#H#HD ##r+cht|S)zw Retrieve the unique SHA256 computed using the transformed (re-encoded) payload. Not the original one. )rr hexdigestrCs r)r4zCharsetMatch.fingerprints& dkkmm$$..000r+rX)r6rr,N)r,r)r)'__name__ __module__ __qualname__bytesr1floatboolrr*r7r>propertyrDrFr rNrQrSrVr3rr^rardrrur;r<rrrrrrrrrrrr4rgr+r)rrsH.2'''"'# ' ! ' * '&c]''''2YtYYYY (t ( ( ( (  e    X  UX 67 6 6 6X 6 W#WWWW#####X $s)   X $T$$$X$$$$$X$/49///X/%#%%%X%.%u%%%X%%5%%%X% 2u222X265666X6UX $~.X%d%%%X% $49 $ $ $X $EtCyEEEXE  $ $s $ $ $ $ $1S111X111r+rceZdZdZddeefdZdZdefdZde fdZ d eddfd Z de d fd Z de d fd ZdS)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. Nresultsc6|rt|ng|_dSrX)r_results)r(rs r)r*zCharsetMatches.__init__s+2:w r+c#&K|jD]}|VdSrXr)r(results r)__iter__zCharsetMatches.__iter__s,m  FLLLL  r+r,ct|tr |j|St|tr't |d}|jD]}||jvr|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.intrr1rrrKeyError)r(itemrs r) __getitem__zCharsetMatches.__getitem__ sv dC  '=& & dC  "T5))D- " "6777!MMM8r+c*t|jSrX)rtrrCs r)__len__zCharsetMatches.__len__s4=!!!r+rct|ts4tdt |jt |jtkrB|j D]:}|j |j kr(|j |j kr| |dS;|j |t|j |_ dS)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.rr9r0r1r2rtrr rr4r;rVrUr)r(rmatchs r)rUzCharsetMatches.appends $ -- jLSSTWX\XfTgTghhii i tx==, , ,  $(888U[DJ=V=V&&t,,,FF T"""t}-- r+rc.|jsdS|jdS)zQ Simply return the first match. Strict equivalent to matches[0]. NrrrCs r)rzCharsetMatches.best+s } 4}Qr+c*|S)zP Redundant method, call the method best(). Kept for BC reasons. )rrCs r)rzCharsetMatches.first3syy{{r+rX)rrr__doc__rrr*rrrrrUrrrrgr+r)rrs;;\ 2;;;; <    """"".<.D....  h~.    x/r+rceZdZdededeedeededeededed ed eed efd Ze d Z defdZ dS)CliDetectionResultpathr3r^alternative_encodingsrurrr;r< unicode_path is_preferredc ||_| |_||_||_||_||_||_||_||_| |_ | |_ dSrX) rrr3r^rrurrr;r<r) r(rr3r^rrurrr;r<rrs r)r*zCliDetectionResult.__init__@sZ (  0%:"  ", "(r+c |j|j|j|j|j|j|j|j|j|j |j d S)N rr3r^rrurrr;r<rrrrCs r)__dict__zCliDetectionResult.__dict__MsOI $ 5%)%? "1Z - -   r+r,c0t|jddS)NT) ensure_asciiindent)rrrCs r)to_jsonzCliDetectionResult.to_json]s$ M    r+N) rrrr1rrrrr*rrrrgr+r)rr>s )S )C )49 )eijmen )z} )KOPSKT )fj )sx )EJ )ZbcfZg )w{ ) ) ) )   X         r+r)"r@encodings.aliasesrhashlibrjsonrtypingrrrr collectionsr rer r rIcharset_normalizer.constantr charset_normalizer.mdrcharset_normalizer.utilsrrrrrr1rCoherenceMatchrrCharsetNormalizerMatchrgr+r)rs%%%%%%------------))))))))888888,,,,,,UUUUUUUUUUm1m1m1m1m1m1m1m1`99999999xsEz"'$ $ $ $ $ $ $ $ N&r+