o j4@sddlmZddlmZddlmZddlmZmZm Z m Z ddl m Z m Z ddlmZmZmZGdd d ZGd d d Ze eefZe eZGd d d ZdS)) annotations)aliases)sub)AnyIteratorListTuple)RE_POSSIBLE_ENCODING_INDICATIONTOO_BIG_SEQUENCE) iana_nameis_multi_byte_encoding unicode_rangec@sHeZdZ  dIdJddZdKddZdKddZedLddZdMddZdMddZ dNdd Z edMd!d"Z edOd$d%Z edPd&d'Z edPd(d)ZedOd*d+ZedMd,d-ZedLd.d/ZedLd0d1ZedLd2d3ZedLd4d5ZedQd6d7ZedRd9d:ZedPd;d<ZedOd=d>ZedOd?d@ZdSdTdDdEZedUdGdHZdS)V CharsetMatchNpayloadbytes | bytearrayguessed_encodingstrmean_mess_ratiofloathas_sig_or_bombool languagesCoherenceMatchesdecoded_payload str | Nonepreemptive_declarationcCsL||_||_||_||_||_d|_g|_d|_d|_d|_ ||_ ||_ dS)N) _payload _encoding_mean_mess_ratio _languages_has_sig_or_bom_unicode_ranges_leaves_mean_coherence_ratio_output_payload_output_encoding_string_preemptive_declaration)selfrrrrrrrr+/root/aizidognhua/tmp/workspace/projects/ec89d86c-575f-41c9-af57-ac45cbdbf775/venv/lib/python3.10/site-packages/charset_normalizer/models.py__init__ s  zCharsetMatch.__init__otherobjectreturncCs@t|tst|trt|d|jkSdS|j|jko|j|jkS)NF) isinstancerrr encoding fingerprintr*r.r+r+r,__eq__(s  zCharsetMatch.__eq__cCst|tstt|j|j}t|j|j}|dkr%|dkr%|j|jkS|dkr@|dkr@t|jtkr:|j|jkS|j |j kS|j|jkS)zQ Implemented to make sorted available upon CharsetMatches items. g{Gzt?g{Gz?) r1r ValueErrorabschaos coherencelenrr multi_byte_usage)r*r.chaos_differencecoherence_differencer+r+r,__lt__3s     zCharsetMatch.__lt__cCs*t|j}|dkr dSdtt||S)Nrrg?)r:rawr)r*raw_lenr+r+r,r;Is zCharsetMatch.multi_byte_usagecCsV|jdur(t|j|jd|_|jr(|jdkr(|jr(|jddkr(|jdd|_|jS)Nstrictutf_7rur )r(rrrr"r*r+r+r,__str__Qs  zCharsetMatch.__str__cCsd|jd|jdS)Nz)r2r3rCr+r+r,__repr__aszCharsetMatch.__repr__NonecCs8t|tr ||krtd|jd|_|j|dS)Nz;Unable to add instance <{}> as a submatch of a CharsetMatch)r1rr6format __class__r(r$appendr4r+r+r, add_submatchdszCharsetMatch.add_submatchcC|jSN)rrCr+r+r,r2ozCharsetMatch.encoding list[str]cCsDg}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. )ritemsr2rI)r* also_known_asupr+r+r,encoding_aliasesss    zCharsetMatch.encoding_aliasescCrKrLr"rCr+r+r,bomrMzCharsetMatch.bomcCrKrLrTrCr+r+r,byte_order_markrMzCharsetMatch.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, z*CharsetMatch.languages..r!rCr+r+r,rszCharsetMatch.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)r!could_be_from_charsetcharset_normalizer.cdr^r_r r2r:)r*r^r_rr+r+r,languages  zCharsetMatch.languagecCrKrL)r rCr+r+r,r8rMzCharsetMatch.chaoscCs|jsdS|jddS)Nrrr r[rCr+r+r,r9szCharsetMatch.coherencecCt|jdddSNd)ndigits)roundr8rCr+r+r, percent_chaoszCharsetMatch.percent_chaoscCrdre)rir9rCr+r+r,percent_coherencerkzCharsetMatch.percent_coherencecCrK)z+ Original untouched bytes. )rrCr+r+r,r?szCharsetMatch.rawlist[CharsetMatch]cCrKrL)r$rCr+r+r,submatchrMzCharsetMatch.submatchcCt|jdkSNr)r:r$rCr+r+r, has_submatchszCharsetMatch.has_submatchcCs@|jdur|jSddt|D}ttdd|D|_|jS)NcSsg|]}t|qSr+)r)rWcharr+r+r,rYrZz*CharsetMatch.alphabets..cSsh|]}|r|qSr+r+)rWrr+r+r, rZz)CharsetMatch.alphabets..)r#rsortedlist)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+)r2)rWmr+r+r,rYsz6CharsetMatch.could_be_from_charset..)rr$rCr+r+r,rasz"CharsetMatch.could_be_from_charsetutf_8r2bytescs~jdus j|kr<|_t}jdur5jdvr5ttfdd|dddd}||dd}||d_jS) z Method to get re-encoded bytes payload using given target encoding. Default to UTF-8. Any errors will be simply ignored by the encoder NOT replaced. N)zutf-8utf8rzcs<|j|d|d|dtjddS)Nrr _-)stringspanreplacegroupsr r')ryrCr+r,s z%CharsetMatch.output..i r )countr)r'rr)lowerrr encoder&)r*r2decoded_stringpatched_headerr+rCr,outputs     zCharsetMatch.outputintcCs tt|S)z] Retrieve a hash fingerprint of the decoded payload, used for deduplication. )hashrrCr+r+r,r3s zCharsetMatch.fingerprint)NN)rrrrrrrrrrrrrr)r.r/r0r)r0rr0r)r.rr0rF)r0rNr0r)r0r)r0rm)rz)r2rr0r{r0r)__name__ __module__ __qualname__r-r5r>propertyr;rDrErJr2rSrUrVrrcr8r9rjrlr?rnrqrxrarr3r+r+r+r,r sX                      rc@sleZdZdZd d!ddZd"d d Zd#d d Zd$ddZd%ddZd&ddZ d'ddZ d(ddZ d(ddZ dS))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. Nresultslist[CharsetMatch] | NonecCs|rt|ng|_d|_dSNT)ru_results _is_sorted)r*rr+r+r,r- s zCharsetMatches.__init__r0rFcCs|js |jd|_dSdSr)rrsortrCr+r+r,_ensure_sorteds  zCharsetMatches._ensure_sortedIterator[CharsetMatch]ccs||jEdHdSrL)rrrCr+r+r,__iter__szCharsetMatches.__iter__item int | strrcCsRt|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)r1rrrrr raKeyError)r*rresultr+r+r, __getitem__s      zCharsetMatches.__getitem__rcCs t|jSrLr:rrCr+r+r,__len__'s zCharsetMatches.__len__rcCrorprrCr+r+r,__bool__*szCharsetMatches.__bool__cCsvt|tstdt|jt|jtkr0|j D]}|j |j kr/|j |j kr/| |dSq|j |d|_dS)z~ Insert a single match. Will be inserted accordingly to preserve sort. Can be inserted as a submatch. z-Cannot append instance '{}' to CharsetMatchesNF)r1rr6rGrrHr:r?r rr3r8rJrIr)r*rmatchr+r+r,rI-s     zCharsetMatches.appendCharsetMatch | NonecCs|jsdS||jdS)zQ Simply return the first match. Strict equivalent to matches[0]. Nr)rrrCr+r+r,bestAs zCharsetMatches.bestcCs|S)zP Redundant method, call the method best(). Kept for BC reasons. )rrCr+r+r,firstJszCharsetMatches.firstrL)rr)r0rF)r0r)rrr0rrr)rrr0rF)r0r) rrr__doc__r-rrrrrrIrrr+r+r+r,rs         rc@s.eZdZdddZedddZdddZdS)CliDetectionResultpathrr2rrSrNalternative_encodingsrcrxrrr8rr9 unicode_path is_preferredc CsF||_| |_||_||_||_||_||_||_||_| |_ | |_ dSrL) rrr2rSrrcrxrr8r9r) r*rr2rSrrcrxrr8r9rrr+r+r,r-Vs zCliDetectionResult.__init__r0dict[str, Any]c Cs2|j|j|j|j|j|j|j|j|j|j |j d S)N rr2rSrrcrxrr8r9rrrrCr+r+r,__dict__pszCliDetectionResult.__dict__cCsddlm}||jdddS)Nr)dumpsT) ensure_asciiindent)jsonrr)r*rr+r+r,to_jsons zCliDetectionResult.to_jsonN)rrr2rrSrNrrNrcrrxrNrrr8rr9rrrrr)r0rr)rrrr-rrrr+r+r+r,rUs   rN) __future__rencodings.aliasesrrertypingrrrrconstantr r utilsr r rrrrrCoherenceMatchrrr+r+r+r,s  { L