o Fhh-@sddlmZddlmZddlmZddlmZmZm Z m Z m Z m Z m Z ddlmZddlmZmZmZGdd d ZGd d d Ze eefZe eZGd d d ZdS))aliases)sha256)dumps)AnyDictIteratorListOptionalTupleUnion)TOO_BIG_SEQUENCE) iana_nameis_multi_byte_encoding unicode_rangec@seZdZ d=dededededddeef d d Zd e d efd dZ d e d efddZ e d efddZ d efddZd efddZd>ddZe d efddZe d eefddZe d efddZe d efdd Ze d eefd!d"Ze d efd#d$Ze d efd%d&Ze d efd'd(Ze d efd)d*Ze d efd+d,Ze d efd-d.Ze d edfd/d0Ze d efd1d2Ze d eefd3d4Ze d eefd5d6Z d?d8ed efd9d:Z!e d efd;d<Z"dS)@ CharsetMatchNpayloadguessed_encodingmean_mess_ratiohas_sig_or_bom languagesCoherenceMatchesdecoded_payloadcCsF||_||_||_||_||_d|_g|_d|_d|_d|_ ||_ dS)N) _payload _encoding_mean_mess_ratio _languages_has_sig_or_bom_unicode_ranges_leaves_mean_coherence_ratio_output_payload_output_encoding_string)selfrrrrrrr&R/var/www/html/loop/nvenv/lib/python3.10/site-packages/charset_normalizer/models.py__init__ s  zCharsetMatch.__init__otherreturncCs>t|tstdt|jt|j|j|jko|j|jkS)Nz&__eq__ cannot be invoked on {} and {}.) isinstancer TypeErrorformatstr __class__encoding fingerprintr%r)r&r&r'__eq__$s zCharsetMatch.__eq__cCst|tstt|j|j}t|j|j}|dkr%|dkr%|j|jkS|dkr@|dkr@t|jtkr:|j|jkS|j |j kS|j|jkS)zQ Implemented to make sorted available upon CharsetMatches items. g{Gz?g{Gz?) r+r ValueErrorabschaos coherencelenrr multi_byte_usage)r%r)chaos_differencecoherence_differencer&r&r'__lt__-s     zCharsetMatch.__lt__cCsdtt|t|jS)Ng?)r8r.rawr%r&r&r'r9CszCharsetMatch.multi_byte_usagecCs"|jdurt|j|jd|_|jS)Nstrict)r$r.rrr>r&r&r'__str__Gs zCharsetMatch.__str__cCsd|j|jS)Nz)r-r0r1r>r&r&r'__repr__MzCharsetMatch.__repr__cCs8t|tr ||krtd|jd|_|j|dS)Nz;Unable to add instance <{}> as a submatch of a CharsetMatch)r+rr4r-r/r$r appendr2r&r&r' add_submatchPszCharsetMatch.add_submatchcC|jSN)rr>r&r&r'r0[zCharsetMatch.encodingcCsDg}tD]\}}|j|kr||q|j|kr||q|S)z Encoding name are known by many name, using this could help when searching for IBM855 when it's listed as CP855. )ritemsr0rC)r% also_known_asupr&r&r'encoding_aliases_s    zCharsetMatch.encoding_aliasescCrErFrr>r&r&r'bomlrGzCharsetMatch.bomcCrErFrMr>r&r&r'byte_order_markprGzCharsetMatch.byte_order_markcCsdd|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'. cSsg|]}|dqS)rr&).0er&r&r' zz*CharsetMatch.languages..rr>r&r&r'rtszCharsetMatch.languagescCsp|js1d|jvr dSddlm}m}t|jr||jn||j}t|dks+d|vr-dS|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.cdrWrXrr0r8)r%rWrXrr&r&r'language|s  zCharsetMatch.languagecCrErF)rr>r&r&r'r6rGzCharsetMatch.chaoscCs|jsdS|jddS)Nrrr rTr>r&r&r'r7szCharsetMatch.coherencecCt|jdddSNd)ndigits)roundr6r>r&r&r' percent_chaoszCharsetMatch.percent_chaoscCr]r^)rbr7r>r&r&r'percent_coherencerdzCharsetMatch.percent_coherencecCrE)z+ Original untouched bytes. )rr>r&r&r'r=szCharsetMatch.rawcCrErF)r r>r&r&r'submatchrGzCharsetMatch.submatchcCt|jdkSNr)r8r r>r&r&r' has_submatchszCharsetMatch.has_submatchcCs@|jdur|jSddt|D}ttdd|D|_|jS)NcSsg|]}t|qSr&)r)rPcharr&r&r'rRsz*CharsetMatch.alphabets..cSsh|]}|r|qSr&r&)rPrr&r&r' rSz)CharsetMatch.alphabets..)rr.sortedlist)r%detected_rangesr&r&r' alphabetss zCharsetMatch.alphabetscCs|jgdd|jDS)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'. cSsg|]}|jqSr&)r0)rPmr&r&r'rRsz6CharsetMatch.could_be_from_charset..)rr r>r&r&r'rZsz"CharsetMatch.could_be_from_charsetutf_8r0cCs2|jdus |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#r.encoder")r%r0r&r&r'outputszCharsetMatch.outputcCst|S)zw Retrieve the unique SHA256 computed using the transformed (re-encoded) payload. Not the original one. )rru hexdigestr>r&r&r'r1szCharsetMatch.fingerprintrF)r)rr*N)rr)#__name__ __module__ __qualname__bytesr.floatboolr r(objectr3r<propertyr9r@rArDr0rrLrNrOrr\r6r7rcrer=rfrirprZrur1r&r&r&r'r sn       rc@seZdZdZddeeefddZdeefddZ d e e e fdefd d Z de fd d ZdefddZd eddfddZdedfddZdedfddZdS)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. NresultscCs|r t||_dSg|_dSrF)rm_results)r%rr&r&r'r(szCharsetMatches.__init__r*ccs|jEdHdSrFrr>r&r&r'__iter__szCharsetMatches.__iter__itemcCsJt|tr |j|St|tr#t|d}|jD] }||jvr"|Sqt)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+intrr.rrZKeyError)r%rresultr&r&r' __getitem__s      zCharsetMatches.__getitem__cCs t|jSrFr8rr>r&r&r'__len__s zCharsetMatches.__len__cCrgrhrr>r&r&r'__bool__szCharsetMatches.__bool__cCs|t|tstdt|jt|jtkr0|j D]}|j |j kr/|j |j kr/| |dSq|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+rr4r-r.r/r8r=r rr1r6rDrCrm)r%rmatchr&r&r'rCs    zCharsetMatches.appendrcCs|jsdS|jdS)zQ Simply return the first match. Strict equivalent to matches[0]. Nrrr>r&r&r'bests zCharsetMatches.bestcCs|S)zP Redundant method, call the method best(). Kept for BC reasons. )rr>r&r&r'firstszCharsetMatches.firstrF)rwrxry__doc__r rrr(rrr rr.rrr|rrCrrr&r&r&r'rsrc@s~eZdZdedeedeedeededeededed ed eed efd d Ze de ee ffddZ defddZ dS)CliDetectionResultpathr0rLalternative_encodingsr\rprr6r7 unicode_path is_preferredc CsF||_| |_||_||_||_||_||_||_||_| |_ | |_ dSrF) rrr0rLrr\rprr6r7r) r%rr0rLrr\rprr6r7rrr&r&r'r()s zCliDetectionResult.__init__r*c Cs2|j|j|j|j|j|j|j|j|j|j |j d S)N rr0rLrr\rprr6r7rrrr>r&r&r'__dict__CszCliDetectionResult.__dict__cCst|jdddS)NT) ensure_asciiindent)rrr>r&r&r'to_jsonSrBzCliDetectionResult.to_jsonN)rwrxryr.r rr|r{r(r~rrrrr&r&r&r'r(s6    rN)encodings.aliasesrhashlibrjsonrtypingrrrrr r r constantr utilsrrrrrr.r{CoherenceMatchrrr&r&r&r's  $ X C