U
    ,º]Ñ  ã                   @   s‚   d dl mZmZmZ d dlmZ d dlmZ d dlm	Z
mZ e
e_	ee_dgZG dd„ deƒZG dd„ deƒZG d	d
„ d
eƒZdS )é    )Úabsolute_importÚdivisionÚunicode_literals©Ústr)Úurllib)ÚparseÚrequestÚRobotFileParserc                   @   sZ   e Zd ZdZddd„Zdd„ Zdd„ Zd	d
„ Zdd„ Zdd„ Z	dd„ Z
dd„ Zdd„ ZdS )r
   zs This class provides a set of methods to read, parse and answer
    questions about a single robots.txt file.

    Ú c                 C   s,   g | _ d | _d| _d| _|  |¡ d| _d S )NFr   )ÚentriesÚdefault_entryÚdisallow_allÚ	allow_allÚset_urlÚlast_checked©ÚselfÚurl© r   úE/usr/lib/python3/dist-packages/future/backports/urllib/robotparser.pyÚ__init__   s    
zRobotFileParser.__init__c                 C   s   | j S )z·Returns the time the robots.txt file was last fetched.

        This is useful for long-running web spiders that need to
        check for new robots.txt files periodically.

        )r   ©r   r   r   r   Úmtime&   s    zRobotFileParser.mtimec                 C   s   ddl }|  ¡ | _dS )zYSets the time the robots.txt file was last fetched to the
        current time.

        r   N)Útimer   )r   r   r   r   r   Úmodified/   s    zRobotFileParser.modifiedc                 C   s&   || _ tj |¡dd… \| _| _dS )z,Sets the URL referring to a robots.txt file.é   é   N)r   r   r   ÚurlparseZhostÚpathr   r   r   r   r   7   s    zRobotFileParser.set_urlc              
   C   s|   zt j | j¡}W nH t jjk
rZ } z&|jdkr:d| _n|jdkrJd| _W 5 d}~X Y nX | 	¡ }|  
| d¡ ¡ ¡ dS )z4Reads the robots.txt URL and feeds it to the parser.)i‘  i“  Ti  Nzutf-8)r   r	   Zurlopenr   ÚerrorZ	HTTPErrorÚcoder   r   Úreadr   ÚdecodeÚ
splitlines)r   ÚfÚerrÚrawr   r   r   r"   <   s    

zRobotFileParser.readc                 C   s,   d|j kr| jd kr(|| _n| j |¡ d S ©NÚ*)Ú
useragentsr   r   Úappend)r   Úentryr   r   r   Ú
_add_entryI   s    

zRobotFileParser._add_entryc                 C   sn  d}t ƒ }|D ]D}|sH|dkr,t ƒ }d}n|dkrH|  |¡ t ƒ }d}| d¡}|dkrf|d|… }| ¡ }|stq| dd¡}t|ƒdkr|d  ¡  ¡ |d< tj 	|d  ¡ ¡|d< |d dkrò|dkrÜ|  |¡ t ƒ }|j
 |d ¡ d}q|d dkr&|dkrT|j t|d d	ƒ¡ d}q|d d
kr|dkr|j t|d dƒ¡ d}q|dkrj|  |¡ dS )z”Parse the input lines from a robots.txt file.

        We allow that a user-agent: line is not preceded by
        one or more blank lines.
        r   r   é   ú#Nú:z
user-agentZdisallowFZallowT)ÚEntryr-   ÚfindÚstripÚsplitÚlenÚlowerr   r   Úunquoter*   r+   Ú	rulelinesÚRuleLine)r   ÚlinesÚstater,   ÚlineÚir   r   r   r   R   sJ    






zRobotFileParser.parsec                 C   s–   | j r
dS | jrdS tj tj |¡¡}tj dd|j|j|j	|j
f¡}tj |¡}|s\d}| jD ]}| |¡rb| |¡  S qb| jr’| j |¡S dS )z=using the parsed robots.txt decide if useragent can fetch urlFTr   ú/)r   r   r   r   r   r7   Z
urlunparser   ZparamsZqueryZfragmentÚquoter   Ú
applies_toÚ	allowancer   )r   Ú	useragentr   Z
parsed_urlr,   r   r   r   Ú	can_fetch…   s&      ÿ

zRobotFileParser.can_fetchc                 C   s   d  dd„ | jD ƒ¡S )Nr   c                 S   s   g | ]}t |ƒd  ‘qS )Ú
r   )Ú.0r,   r   r   r   Ú
<listcomp>   s     z+RobotFileParser.__str__.<locals>.<listcomp>)Újoinr   r   r   r   r   Ú__str__œ   s    zRobotFileParser.__str__N)r   )Ú__name__Ú
__module__Ú__qualname__Ú__doc__r   r   r   r   r"   r-   r   rC   rH   r   r   r   r   r
      s   
		3c                   @   s(   e Zd ZdZdd„ Zdd„ Zdd„ ZdS )	r9   zoA rule line is a single "Allow:" (allowance==True) or "Disallow:"
       (allowance==False) followed by a path.c                 C   s(   |dkr|sd}t j |¡| _|| _d S )Nr   T)r   r   r?   r   rA   )r   r   rA   r   r   r   r   £   s    zRuleLine.__init__c                 C   s   | j dkp| | j ¡S r(   )r   Ú
startswith)r   Úfilenamer   r   r   r@   ª   s    zRuleLine.applies_toc                 C   s   | j r
dpdd | j S )NZAllowZDisallowz: )rA   r   r   r   r   r   rH   ­   s    zRuleLine.__str__N)rI   rJ   rK   rL   r   r@   rH   r   r   r   r   r9       s   r9   c                   @   s0   e Zd ZdZdd„ Zdd„ Zdd„ Zdd	„ Zd
S )r1   z?An entry has one or more user-agents and zero or more rulelinesc                 C   s   g | _ g | _d S )N)r*   r8   r   r   r   r   r   ³   s    zEntry.__init__c                 C   sH   g }| j D ]}| d|dg¡ q
| jD ]}| t|ƒdg¡ q&d |¡S )NzUser-agent: rD   r   )r*   Úextendr8   r   rG   )r   ZretÚagentr<   r   r   r   rH   ·   s    

zEntry.__str__c                 C   sF   |  d¡d  ¡ }| jD ](}|dkr* dS | ¡ }||kr dS qdS )z2check if this entry applies to the specified agentr>   r   r)   TF)r4   r6   r*   )r   rB   rP   r   r   r   r@   ¿   s    
zEntry.applies_toc                 C   s$   | j D ]}| |¡r|j  S qdS )zZPreconditions:
        - our agent applies to this entry
        - filename is URL decodedT)r8   r@   rA   )r   rN   r<   r   r   r   rA   Ì   s    

zEntry.allowanceN)rI   rJ   rK   rL   r   rH   r@   rA   r   r   r   r   r1   ±   s
   r1   N)Z
__future__r   r   r   Zfuture.builtinsr   Zfuture.backportsr   Zfuture.backports.urllibr   Ú_parser	   Z_requestÚ__all__Úobjectr
   r9   r1   r   r   r   r   Ú<module>   s    	