U
    :g'                     @   sT  d Z dddgZddlZddlmZmZ z0d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zddlmZ W n" ek
r   ddlmZ Y nX edjZz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   8/tmp/pip-unpacked-wheel-m1sofc2u/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   ElementTree)filer   r   r   rootr   r   r   r   $   s    	
c                 C   s*   t | |}| }|D ]}|| q|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_treegetchildrenremove)beautiful_soup_treer   r   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_FEATURESfeatureszhtml.parser   r   )r   r   r   lentag)sourcer   r   r   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  }}t| D ]Z\}}t|trl|d krF|}|}|d kr|jr|j dkr|}q(|d kr(t|tr(|}q(|d krg  }}	| j	}
n2| j	d | }| j	||d  }
| j	|d d  }	|d k	r|

|}|
d | |j	 |
|d d   |_	nt|
}t|}||}|}t|D ]&}||}|d k	r,|| |}q,|}|	D ]&}||}|d k	r\|| |}q\|d k	rz| }W n tk
r   |j}Y nX t|}|sn>| \}}| j}|o|dd |_|o|dd |_|S )Nr   r"   )r   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docinfo	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_urir?   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    D ]}|  |<  | q| S r-   )append)handlert)
convertersordered_node_typestypesr   r   add   s    z5_init_node_converters.<locals>.converter.<locals>.addr   )rL   rM   rJ   rK   )rL   r   	converter   s    z(_init_node_converters.<locals>.converterc                    s$   D ]}t | |r |   S qd S r-   )r5   )noderI   rN   r   r   find_best_converter   s    
z2_init_node_converters.<locals>.find_best_converterc                    sP   z t |  }W n( tk
r8   |  } t | < Y nX |d krFd S || |S r-   )typeKeyError)bs_nodeparentrH   )rJ   rQ   r   r   rD      s    z+_init_node_converters.<locals>.convert_nodec                 S   sT   t | trBi }|  D ](\}}t |tr2d|}t|||< qndd | D }|S )N c                 S   s   i | ]\}}|t |qS r   unescape).0kvr   r   r   
<dictcomp>   s      z<_init_node_converters.<locals>.map_attrs.<locals>.<dictcomp>)r5   dictitemslistjoinrX   )Zbs_attrsattribsrZ   r[   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    r2   )r#   texttail)rU   rd   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}| D ]H}zt| }W n tk
r|   Y nX |d k	rT||| qT || qT|S )N)attrib)r)   r   
SubElementr(   rR   rS   )rT   rU   r)   ra   resr    rH   )rD   rJ   r   rb   r   r   convert_tag   s     
z*_init_node_converters.<locals>.convert_tagc                 S   s    t | }|d k	r|| |S r-   )r   ZHtmlCommentrG   rT   rU   ri   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?r2   rV   r"   )endswithr   r
   splitrG   rk   r   r   r   
convert_pi  s    

z)_init_node_converters.<locals>.convert_pic                    s   |d k	r |t |  d S r-   rW   )rT   rU   )rf   r   r   convert_text  s    z+_init_node_converters.<locals>.convert_text)N)r   r'   r	   r
   r   )r   rO   rj   rl   rp   rq   r   )rf   rD   rJ   rQ   r   rb   rK   r   r9      s     



r9   )name2codepointz&(\w+);c                 C   s   | sdS dd }t || S )Nrc   c                 S   s8   zt t| d W S  tk
r2   | d Y S X d S )Nr"   r   )unichrrr   grouprS   )mr   r   r   unescape_entity5  s    z!unescape.<locals>.unescape_entity)handle_entities)r<   rv   r   r   r   rX   1  s    rX   )NN)NN)N)$__doc____all__relxmlr   r   Zbs4r   r   r	   r
   r   r   r   r7   ImportErrorr   r   r   r   compile
IGNORECASErF   r=   r'   r   r9   html.entitiesrr   htmlentitydefssubrw   rs   	NameErrorchrrX   r   r   r   r   <module>   s<   
$ 



Uc
