B
    'Y'                 @   sT  d Z dddgZddlZddlmZmZ y0ddlmZmZm	Z	m
Z
mZmZmZ eefZW n8 ek
r   ddlmZmZm	Z	m
Z
mZmZ eZY nX dd	dZdd
dZdddZdd ZedejjZG dd dZdd Zdd ZyddlmZ W n" ek
r   ddlmZ Y nX edjZye  W n e!k
rF   e"Z Y nX dd Z#dS )z5External interface to the BeautifulSoup HTML parser.

fromstringparseconvert_tree    N)etreehtml)BeautifulSoupTagCommentProcessingInstructionNavigableStringDeclarationDoctype)r   r   r	   r
   r   r   c             K   s   t | ||f|S )a  Parse a string of HTML data into an Element tree using the
    BeautifulSoup parser.

    Returns the root ``<html>`` Element of the tree.

    You can pass a different BeautifulSoup parser through the
    `beautifulsoup` keyword, and a diffent Element factory function
    through the `makeelement` keyword.  By default, the standard
    ``BeautifulSoup`` class and the default factory of `lxml.html` are
    used.
    )_parse)databeautifulsoupmakeelementbsargs r   3lib/python3.7/site-packages/lxml/html/soupparser.pyr      s    c             K   s,   t | dst| } t| ||f|}t|S )aY  Parse a file into an ElemenTree using the BeautifulSoup parser.

    You can pass a different BeautifulSoup parser through the
    `beautifulsoup` keyword, and a diffent Element factory function
    through the `makeelement` keyword.  By default, the standard
    ``BeautifulSoup`` class and the default factory of `lxml.html` are
    used.
    read)hasattropenr   r   ZElementTree)filer   r   r   rootr   r   r   r   $   s    	
c             C   s.   t | |}| }x|D ]}|| qW |S )a  Convert a BeautifulSoup tree to a list of Element trees.

    Returns a list instead of a single root Element to support
    HTML-like soup with more than one root element.

    You can pass a different Element factory through the `makeelement`
    keyword.
    )_convert_treeZgetchildrenremove)beautiful_soup_treer   r   Zchildrenchildr   r   r   r   3   s
    	

c             K   s   |d krt }t|dr&d|kr&d|d< t|dr@d|kr@d|d< || f|}t||}t|dkrx|d jdkrx|d S d|_|S )	NZHTML_ENTITIESZconvertEntitiesr   ZDEFAULT_BUILDER_FEATURESZfeatureszhtml.parser   r   )r   r   r   lentag)sourcer   r   r   Ztreer   r   r   r   r   E   s    


r   z`(?:\s|[<!])*DOCTYPE\s*HTML(?:\s+PUBLIC)?(?:\s+(\'[^\']*\'|"[^"]*"))?(?:\s+(\'[^\']*\'|"[^"]*"))?c               @   s   e Zd Zdd Zdd ZdS )
_PseudoTagc             C   s   d| _ g | _|| _d S )Nr   )nameattrscontents)selfr%   r   r   r   __init__`   s    z_PseudoTag.__init__c             C   s
   | j  S )N)r%   __iter__)r&   r   r   r   r(   e   s    z_PseudoTag.__iter__N)__name__
__module____qualname__r'   r(   r   r   r   r   r"   ^   s   r"   c             C   s  |d krt jj}d  }}d  }}xft| D ]Z\}}t|trn|d krH|}|}|d kr|jr|j dkr|}q*|d kr*t|tr*|}q*W |d krg  }}	| j	}
n2| j	d | }| j	||d  }
| j	|d d  }	|d k	r|

|}|
d | |j	 |
|d d   |_	nt|
}t|}||}|}x2t|D ]&}||}|d k	r2|| |}q2W |}x.|	D ]&}||}|d k	rf|| |}qfW |d k	ry| }W n tk
r   |j}Y nX t|}|sn>| \}}| j}|o|dd |_|o|dd |_|S )Nr   r   )r   Zhtml_parserr   	enumerate
isinstancer   r#   lower_DECLARATION_OR_DOCTYPEr%   indexr"   _init_node_convertersreversedZaddpreviousZaddnextZoutput_readyAttributeErrorstring_parse_doctype_declarationgroupsZgetroottreedocinfoZ	public_idZ
system_url)r   r   Zfirst_element_idxZlast_element_idxZ	html_rootZdeclarationieZpre_rootZ	post_rootrootsconvert_nodeZres_rootprevZ	convertedZdoctype_stringmatchZexternal_idZsys_urir8   r   r   r   r   i   sd    


&








r   c                s   i g fdd}fdddfdd	dd d	d
  |t tfdd}|tdd }|tdd }|t fdd}S )Nc                 s    fdd}|S )Nc                s$   xD ]}|  |<  | qW | S )N)append)handlert)
convertersordered_node_typestypesr   r   add   s    
z5_init_node_converters.<locals>.converter.<locals>.addr   )rD   rE   )rB   rC   )rD   r   	converter   s    z(_init_node_converters.<locals>.converterc                s$   xD ]}t | |r | S qW d S )N)r.   )ZnoderA   )rB   rC   r   r   find_best_converter   s    

z2_init_node_converters.<locals>.find_best_converterc                sP   y t |  }W n( tk
r8   |  } t | < Y nX |d krFd S || |S )N)typeKeyError)bs_nodeparentr@   )rB   rG   r   r   r<      s    z+_init_node_converters.<locals>.convert_nodec             S   s\   t | trFi }xH|  D ](\}}t |tr4d|}t|||< qW ntdd | D }|S )N c             s   s   | ]\}}|t |fV  qd S )N)unescape).0kvr   r   r   	<genexpr>   s    z;_init_node_converters.<locals>.map_attrs.<locals>.<genexpr>)r.   dictitemslistjoinrM   )Zbs_attrsattribsrO   rP   r   r   r   	map_attrs   s    


z(_init_node_converters.<locals>.map_attrsc             S   s:   t | dkr| jpd| | _n| d jp*d| | d _d S )Nr    r,   )r   texttail)rK   rY   r   r   r   append_text   s    z*_init_node_converters.<locals>.append_textc          	      s   | j }|d k	r2|r|nd }tj|| j|d}n|r>|ni }| j|d}xP| D ]H}yt| }W n tk
r~   Y nX |d k	rV||| qV || qVW |S )N)Zattrib)r$   r   Z
SubElementr#   rH   rI   )rJ   rK   r$   rV   resr   r@   )r<   rB   r   rW   r   r   convert_tag   s     

z*_init_node_converters.<locals>.convert_tagc             S   s    t | }|d k	r|| |S )N)r   ZHtmlCommentr?   )rJ   rK   r\   r   r   r   convert_comment  s    

z._init_node_converters.<locals>.convert_commentc             S   s>   |  dr| d d } tj| dd }|d k	r:|| |S )N?r,   rL   r   )endswithr   r
   splitr?   )rJ   rK   r\   r   r   r   
convert_pi  s    

z)_init_node_converters.<locals>.convert_pic                s   |d k	r |t |  d S )N)rM   )rJ   rK   )r[   r   r   convert_text  s    z+_init_node_converters.<locals>.convert_text)N)r   r"   r	   r
   r   )r   rF   r]   r^   rb   rc   r   )r[   r<   rB   rG   r   rW   rC   r   r2      s    
r2   )name2codepointz&(\w+);c             C   s   | sdS dd }t || S )NrX   c             S   s2   yt t| d S  tk
r,   | dS X d S )Nr   r   )unichrrd   grouprI   )mr   r   r   unescape_entity5  s    z!unescape.<locals>.unescape_entity)handle_entities)r5   rh   r   r   r   rM   1  s    rM   )NN)NN)N)$__doc____all__reZlxmlr   r   Zbs4r   r   r	   r
   r   r   r   r0   ImportErrorr   r   r   r   compile
IGNORECASEr>   r6   r"   r   r2   Zhtml.entitiesrd   Zhtmlentitydefssubri   re   	NameErrorchrrM   r   r   r   r   <module>   s:   
$ 




Uc
