shell bypass 403
o
_ b�C � @ s� d Z ddlZddlZddlmZ dgZe�d�Ze�d�Ze�d�Z e�d�Z
e�d �Z e�d
�Z e�d �Z
e�d �Ze�d
�Ze�dej�Ze�d
�Ze�d�ZG dd� dej�ZdS )z A parser for HTML and XHTML.� N)�unescape�
HTMLParserz[&<]z
&[a-zA-Z#]z%&([a-zA-Z][-.a-zA-Z0-9]*)[^a-zA-Z0-9]z)&#(?:[0-9]+|[xX][0-9a-fA-F]+)[^0-9a-fA-F]z <[a-zA-Z]�>z--\s*>z+([a-zA-Z][^\t\n\r\f />\x00]*)(?:\s|/(?!>))*z]((?<=[\'"\s/])[^\s/>][^\s/=>]*)(\s*=+\s*(\'[^\']*\'|"[^"]*"|(?![\'"])[^>\s]*))?(?:\s|/(?!>))*aF
<[a-zA-Z][^\t\n\r\f />\x00]* # tag name
(?:[\s/]* # optional whitespace before attribute name
(?:(?<=['"\s/])[^\s/>][^\s/=>]* # attribute name
(?:\s*=+\s* # value indicator
(?:'[^']*' # LITA-enclosed value
|"[^"]*" # LIT-enclosed value
|(?!['"])[^>\s]* # bare value
)
\s* # possibly followed by a space
)?(?:\s|/(?!>))*
)*
)?
\s* # trailing whitespace
z#</\s*([a-zA-Z][-.a-zA-Z0-9:_]*)\s*>c @ s� e Zd ZdZdZdd�dd�Zdd� Zd d
� Zd d � Zd
Z dd� Z
dd� Z dd� Z dd� Z
dd� Zd7dd�Zdd � Zd d � Zd d � Zd!d"� Zd#d$� Zd%d&� Zd'd(� Zd)d*� Zd+d,� Zd-d.� Zd/d0� Zd1d2� Zd3d4� Z d5d6� Z d
S )8r aE Find tags and other markup and call handler functions.
Usage:
p = HTMLParser()
p.feed(data)
...
p.close()
Start tags are handled by calling self.handle_starttag() or
self.handle_startendtag(); end tags by self.handle_endtag(). The
data between tags is passed from the parser to the derived class
by calling self.handle_data() with the data as argument (the data
may be split up in arbitrary chunks). If convert_charrefs is
True the character references are converted automatically to the
corresponding Unicode character (and self.handle_data() is no
longer split in chunks), otherwise they are passed by calling
self.handle_entityref() or self.handle_charref() with the string
containing respectively the named or numeric reference as the
argument.
)Zscript�styleT)�convert_charrefsc C s || _ | �� dS )z�Initialize and reset this instance.
If convert_charrefs is True (the default), all character references
are automatically converted to the corresponding Unicode characters.
N)r �reset)�selfr � r �4/usr/local/python-3.10/lib/python3.10/html/parser.py�__init__V s zHTMLParser.__init__c C s( d| _ d| _t| _d| _tj�| � dS )z1Reset this instance. Loses all unprocessed data.� z???N)�rawdata�lasttag�interesting_normal� interesting�
cdata_elem� _markupbase�
ParserBaser �r r r r
r _ s
zHTMLParser.resetc C s | j | | _ | �d� dS )z�Feed data to the parser.
Call this as often as you want, with as little or as much text
as you want (may include '\n').
r N)r
�goahead�r �datar r r
�feedg s zHTMLParser.feedc C s | � d� dS )zHandle any buffered data.� N)r r r r r
�closep s zHTMLParser.closeNc C s | j S )z)Return full source of start tag: '<...>'.)�_HTMLParser__starttag_textr r r r
�get_starttag_textv s z HTMLParser.get_starttag_textc C s$ |� � | _t�d| j tj�| _d S )Nz </\s*%s\s*>)�lowerr �re�compile�Ir )r �elemr r r
�set_cdata_modez s
zHTMLParser.set_cdata_modec C s t | _d | _d S �N)r r r r r r r
�clear_cdata_mode~ s
zHTMLParser.clear_cdata_modec C s: | j }d}t|�}||k �r�| jr;| js;|�d|�}|dk r:|�dt||d ��}|dkr8t�d�� ||�s8�q�|}n| j
� ||�}|rI|� � }n| jrN�q�|}||k ro| jrf| jsf| � t
|||� �� n | � |||� � | �||�}||kr{�q�|j}|d|��r t�||�r�| �|�} n>|d|�r�| �|�} n3|d|�r�| �|�} n(|d|�r�| �|�} n |d |�r�| �|�} n|d
|k r�| � d� |d
} n�q�| dk �r|sאq�|�d |d
�} | dk r�|�d|d
�} | dk r�|d
} n| d
7 } | j�r | j�s | � t
||| � �� n | � ||| � � | �|| �}n�|d |��rmt�||�}|�rO|�� d
d� }
| �|
� |�� } |d| d
��sH| d
} | �|| �}q d||d � v �rk| � |||d
� � | �||d
�}�q�|d|��r�t�||�}|�r�|�d
�}
| � |
� |�� } |d| d
��s�| d
} | �|| �}q t �||�}|�r�|�r�|�� ||d � k�r�|�� } | |k�r�|} | �||d
�}�q�|d
|k �r�| � d� | �||d
�}�q�q�||k s|�r||k �r| j�s| j�r| j�s| � t
|||� �� n | � |||� � | �||�}||d � | _ d S )Nr �<�&�" z[\s;]z</�<!--z<?z<!r r z&#� ����;) r
�lenr r �find�rfind�maxr r �searchr �start� handle_datar Z updatepos�
startswith� starttagopen�match�parse_starttag� parse_endtag�
parse_comment�parse_pi�parse_html_declaration�charref�group�handle_charref�end� entityref�handle_entityref�
incomplete) r r> r
�i�n�jZampposr5 r3 �k�namer r r
r � s�
��
�
�k zHTMLParser.goaheadc C s� | j }|||d � dkr| �|�S |||d � dkr!| �|�S |||d � �� dkrJ|�d|d �}|dkr;dS | �||d |� � |d
S | �|�S ) N� r( � z<![� z <!doctyper r* r) r )r
r8 Zparse_marked_sectionr r- � handle_decl�parse_bogus_comment)r rB r
�gtposr r r
r: � s
z!HTMLParser.parse_html_declarationr c C sD | j }|�d|d �}|dkrdS |r | �||d |� � |d S )Nr r) r* r )r
r- �handle_comment)r rB �reportr
�posr r r
rK s z HTMLParser.parse_bogus_commentc C sH | j }t�||d �}|sdS |�� }| �||d |� � |�� }|S )Nr) r* )r
�picloser0 r1 � handle_pir> )r rB r
r5 rD r r r
r9 s zHTMLParser.parse_pic C s� d | _ | �|�}|dk r|S | j}|||� | _ g }t�||d �}|�� }|�d��� | _}||k r�t �||�}|s=q�|�ddd�\} }
} |
sLd } n-| d d� d kr^| dd � kssn | d d� d krq| dd � kryn n| dd� } | rt
| �} |� | �� | f� |�� }||k s4|||� � � } | dvr�| �
� \}
}d | j v r�|
| j �d � }
t| j �| j �d � }n|t| j � }| �|||� � |S | �d
�r�| �||� |S | �||� || jv r�| �|� |S ) Nr r r) rH �'r* �")r �/>�
rT )r �check_for_whole_start_tagr
�tagfind_tolerantr5 r> r< r r �attrfind_tolerantr �append�stripZgetpos�countr, r. r2 �endswith�handle_startendtag�handle_starttag�CDATA_CONTENT_ELEMENTSr" )r rB �endposr
�attrsr5 rE �tag�m�attrname�restZ attrvaluer> �lineno�offsetr r r
r6 , s\
& �
� �
�
�
zHTMLParser.parse_starttagc C s� | j }t�||�}|rU|�� }|||d � }|dkr |d S |dkr?|�d|�r-|d S |�d|�r5dS ||kr;|S |d S |dkrEdS |dv rKdS ||krQ|S |d S td ��)
Nr r �/rT r) r* r z6abcdefghijklmnopqrstuvwxyz=/ABCDEFGHIJKLMNOPQRSTUVWXYZzwe should not get here!)r
�locatestarttagend_tolerantr5 r> r3 �AssertionError)r rB r
rc rD �nextr r r
rV _ s. z$HTMLParser.check_for_whole_start_tagc C s | j }t�||d �}|sdS |�� }t�||�}|s`| jd ur+| �|||� � |S t�||d �}|sH|||d � dkrC|d S | � |�S |�
d�� � }|� d|�� �}| �
|� |d S |�
d�� � }| jd ur||| jkr|| �|||� � |S | �
|� | �� |S )Nr r* r) rH z</>r )r
� endendtagr0 r> �
endtagfindr5 r r2 rW rK r< r r- �
handle_endtagr$ )r rB r
r5 rL Z namematchZtagnamer! r r r
r7 � s6
zHTMLParser.parse_endtagc C s | � ||� | �|� d S r# )r^ rn �r rb ra r r r
r] � s z HTMLParser.handle_startendtagc C � d S r# r ro r r r
r^ � � zHTMLParser.handle_starttagc C rp r# r )r rb r r r
rn � rq zHTMLParser.handle_endtagc C rp r# r �r rF r r r
r= � rq zHTMLParser.handle_charrefc C rp r# r rr r r r
r@ � rq zHTMLParser.handle_entityrefc C rp r# r r r r r
r2 � rq zHTMLParser.handle_datac C rp r# r r r r r
rM � rq zHTMLParser.handle_commentc C rp r# r )r Zdeclr r r
rJ � rq zHTMLParser.handle_declc C rp r# r r r r r
rQ � rq zHTMLParser.handle_pic C rp r# r r r r r
� unknown_decl� rq zHTMLParser.unknown_decl)r ) �__name__�
__module__� __qualname__�__doc__r_ r r r r r r r" r$ r r: rK r9 r6 rV r7 r] r^ rn r= r@ r2 rM rJ rQ rs r r r r
r > s8 z
3"( )rw r r Zhtmlr �__all__r r rA r? r; r4 rP Z commentcloserW rX �VERBOSEri rl rm r r r r r r
�<module> s,
��