File manager - Edit - /opt/saltstack/salt/lib/python3.10/site-packages/charset_normalizer/__pycache__/models.cpython-310.pyc
Back
o ;j�, � @ s� d dl mZ d dlmZ d dlmZ d dlmZmZm Z m Z m Z m Z m Z ddlmZ ddlmZmZmZ G dd � d �ZG d d � d �Ze eef Ze e ZG d d � d �ZdS )� )�aliases)�sha256)�dumps)�Any�Dict�Iterator�List�Optional�Tuple�Union� )�TOO_BIG_SEQUENCE)� iana_name�is_multi_byte_encoding� unicode_rangec @ s� e Zd Z d=dededededddee f d d �Zd e d efd d�Z d e d efdd�Z e d efdd��Z d efdd�Zd efdd�Zd>dd�Ze d efdd��Ze d ee fdd ��Ze d efd d ��Ze d efd d ��Ze d ee fd!d"��Ze d efd#d$��Ze d efd%d&��Ze d efd'd(��Ze d efd)d*��Ze d efd+d,��Ze d efd-d.��Z e d ed fd/d0��Z e d efd1d2��Z e d ee fd3d4��Z e d ee fd5d6��Z d?d8ed efd9d:�Z!e d efd;d<��Z"dS )@� CharsetMatchN�payload�guessed_encoding�mean_mess_ratio�has_sig_or_bom� languages�CoherenceMatches�decoded_payloadc C sF || _ || _|| _|| _|| _d | _g | _d| _d | _d | _ || _ d S )N� ) �_payload� _encoding�_mean_mess_ratio� _languages�_has_sig_or_bom�_unicode_ranges�_leavesZ_mean_coherence_ratio�_output_payload�_output_encoding�_string)�selfr r r r r r � r% �M/opt/saltstack/salt/lib/python3.10/site-packages/charset_normalizer/models.py�__init__ s zCharsetMatch.__init__�other�returnc C s> t |t�std�t|j�t| j����| j|jko | j|jkS )Nz&__eq__ cannot be invoked on {} and {}.)� isinstancer � TypeError�format�str� __class__�encoding� fingerprint�r$ r( r% r% r& �__eq__$ s ��zCharsetMatch.__eq__c C sv t |t�st�t| j|j �}t| j|j �}|dk r5|dkr5|dkr/| j|jkr/| j|jkS | j|jkS | j|jk S )zQ Implemented to make sorted available upon CharsetMatches items. g{�G�z�?g{�G�z�?r )r* r � ValueError�abs�chaos� coherence�multi_byte_usage)r$ r( Zchaos_differenceZcoherence_differencer% r% r& �__lt__- s zCharsetMatch.__lt__c C s dt t| ��t | j� S )Ng �?)�lenr- �raw�r$ r% r% r& r7 @ s z CharsetMatch.multi_byte_usagec C s" | j d u rt| j| jd�| _ | j S )N�strict)r# r- r r r; r% r% r& �__str__D s zCharsetMatch.__str__c C s d� | j| j�S )Nz <CharsetMatch '{}' bytes({})>)r, r/ r0 r; r% r% r& �__repr__J � zCharsetMatch.__repr__c C s8 t |t�r || krtd�|j���d |_| j�|� d S )Nz;Unable to add instance <{}> as a submatch of a CharsetMatch)r* r r3 r, r. r# r �appendr1 r% r% r& � add_submatchM s ��zCharsetMatch.add_submatchc C � | j S �N)r r; r% r% r& r/ X � zCharsetMatch.encodingc C sD g }t �� D ]\}}| 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. )r �itemsr/ r@ )r$ Z also_known_as�u�pr% r% r& �encoding_aliases\ s �z CharsetMatch.encoding_aliasesc C rB rC �r r; r% r% r& �bomi rD zCharsetMatch.bomc C rB rC rI r; r% r% r& �byte_order_markm rD z CharsetMatch.byte_order_markc C s dd� | j D �S )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'. c S s g | ]}|d �qS )r r% )�.0�er% r% r& � <listcomp>w � z*CharsetMatch.languages.<locals>.<listcomp>�r r; r% r% r& r q s zCharsetMatch.languagesc C sp | j s1d| jv r dS ddlm}m} t| j�r || j�n|| j�}t|�dks+d|v r-dS |d S | j d d S )z� Most probable language found in decoded sequence. If none were detected or inferred, the property will return "Unknown". �asciiZEnglishr )�encoding_languages�mb_encoding_languagesz Latin BasedZUnknown)r �could_be_from_charsetZcharset_normalizer.cdrR rS r r/ r9 )r$ rR rS r r% r% r& �languagey s ��zCharsetMatch.languagec C rB rC )r r; r% r% r& r5 � rD zCharsetMatch.chaosc C s | j sdS | j d d S )Nr r r rP r; r% r% r& r6 � s zCharsetMatch.coherencec C � t | jd dd�S �N�d � )�ndigits)�roundr5 r; r% r% r& � percent_chaos� � zCharsetMatch.percent_chaosc C rV rW )r[ r6 r; r% r% r& �percent_coherence� r] z CharsetMatch.percent_coherencec C rB )z+ Original untouched bytes. )r r; r% r% r& r: � s zCharsetMatch.rawc C rB rC )r r; r% r% r& �submatch� rD zCharsetMatch.submatchc C � t | j�dkS �Nr )r9 r r; r% r% r& � has_submatch� s zCharsetMatch.has_submatchc C s@ | j d ur| j S dd� t| �D �}ttdd� |D ���| _ | j S )Nc S s g | ]}t |��qS r% )r )rL �charr% r% r& rN � s �z*CharsetMatch.alphabets.<locals>.<listcomp>c S s h | ]}|r|�qS r% r% )rL �rr% r% r& � <setcomp>� rO z)CharsetMatch.alphabets.<locals>.<setcomp>)r r- �sorted�list)r$ Zdetected_rangesr% r% r& � alphabets� s �zCharsetMatch.alphabetsc C s | j gdd� | jD � S )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'. c S s g | ]}|j �qS r% )r/ )rL �mr% r% r& rN � s z6CharsetMatch.could_be_from_charset.<locals>.<listcomp>)r r r; r% r% r& rT � s z"CharsetMatch.could_be_from_charset�utf_8r/ c C s2 | j du s | 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. N�replace)r" r- �encoder! )r$ r/ r% r% r& �output� s zCharsetMatch.outputc C s t | �� ��� S )zw Retrieve the unique SHA256 computed using the transformed (re-encoded) payload. Not the original one. )r rm � hexdigestr; r% r% r& r0 � s zCharsetMatch.fingerprintrC )r( r r) N)rj )#�__name__� __module__� __qualname__�bytesr- �float�boolr r' �objectr2 r8 �propertyr7 r= r> rA r/ r rH rJ rK r rU r5 r6 r\ r^ r: r_ rb rh rT rm r0 r% r% r% r& r sn ������ � r c @ s� e Zd ZdZddeee fdd�Zdee fdd�Z d e e e f defd d �Z de fd d �Zdefdd�Zd eddfdd�Zded fdd�Zded fdd�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. N�resultsc C s |r t |�| _d S g | _d S rC )rf �_results)r$ rx r% r% r& r' � s zCharsetMatches.__init__r) c c s � | j E d H d S rC �ry r; r% r% r& �__iter__� s �zCharsetMatches.__iter__�itemc C sJ t |t�r | j| S t |t�r#t|d�}| jD ] }||jv r"| S qt�)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* �intry r- r rT �KeyError)r$ r| �resultr% r% r& � __getitem__� s �zCharsetMatches.__getitem__c C s t | j�S rC �r9 ry r; r% r% r& �__len__� s zCharsetMatches.__len__c C r` ra r� r; r% r% r& �__bool__� s zCharsetMatches.__bool__c C s| t |t�std�t|j����t|j�tkr0| j D ]}|j |j kr/|j |j kr/|� |� dS q| 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* r r3 r, r- r. r9 r: r ry r0 r5 rA r@ rf )r$ r| �matchr% r% r& r@ � s �� � zCharsetMatches.appendr c C s | j sdS | j d S )zQ Simply return the first match. Strict equivalent to matches[0]. Nr rz r; r% r% r& �best s zCharsetMatches.bestc C s | � � S )zP Redundant method, call the method best(). Kept for BC reasons. )r� r; r% r% r& �first s zCharsetMatches.firstrC )ro rp rq �__doc__r r r r' r r{ r r} r- r� r� rt r� r@ r� r� r% r% r% r& rw � s rw c @ s~ e Zd Zdedee dee dee dedee deded ed ee d efd d �Ze de ee f fdd��Z defdd�Z dS )�CliDetectionResult�pathr/ rH �alternative_encodingsrU rh r r5 r6 � unicode_path� is_preferredc C sF || _ | | _|| _|| _|| _|| _|| _|| _|| _| | _ | | _ d S rC ) r� r� r/ rH r� rU rh r r5 r6 r� ) r$ r� r/ rH r� rU rh r r5 r6 r� r� r% r% r& r' & s zCliDetectionResult.__init__r) c C s2 | j | j| j| j| j| j| j| j| j| j | j d� S )N� r� r/ rH r� rU rh r r5 r6 r� r� r� r; r% r% r& �__dict__@ s �zCliDetectionResult.__dict__c C s t | jddd�S )NT� )� ensure_ascii�indent)r r� r; r% r% r& �to_jsonP r? zCliDetectionResult.to_jsonN)ro rp rq r- r r rt rs r' rv r r r� r� r% r% r% r& r� % s6 ������� � � � �r� N)Zencodings.aliasesr �hashlibr �jsonr �typingr r r r r r r Zconstantr �utilsr r r r rw r- rs ZCoherenceMatchr r� r% r% r% r& �<module> s $ U C
| ver. 1.4 |
Github
|
.
| PHP 8.2.30 | Generation time: 0 |
proxy
|
phpinfo
|
Settings