a
    åzeÐ$  ã                   @   sX   d dl Z d dlZd dlZdgZe  dd¡ZG dd„ dƒZG dd„ dƒZG dd	„ d	ƒZ	dS )
é    NÚRobotFileParserÚRequestRatezrequests secondsc                   @   sn   e 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d„ Zdd„ Zdd„ ZdS )r   Ú c                 C   s2   g | _ g | _d | _d| _d| _|  |¡ d| _d S )NFr   )ÚentriesÚsitemapsÚdefault_entryÚdisallow_allÚ	allow_allÚset_urlÚlast_checked©ÚselfÚurl© r   ú(/usr/lib/python3.9/urllib/robotparser.pyÚ__init__   s    
zRobotFileParser.__init__c                 C   s   | j S ©N)r   ©r   r   r   r   Úmtime%   s    zRobotFileParser.mtimec                 C   s   dd l }|  ¡ | _d S )Nr   )Útimer   )r   r   r   r   r   Úmodified.   s    zRobotFileParser.modifiedc                 C   s&   || _ tj |¡dd… \| _| _d S )Né   é   )r   ÚurllibÚparseÚurlparseÚhostÚpathr   r   r   r   r
   6   s    zRobotFileParser.set_urlc              
   C   sˆ   zt j | j¡}W nT t jjyf } z8|jdv r8d| _n|jdkrR|jdk rRd| _W Y d }~n&d }~0 0 | 	¡ }|  
| d¡ ¡ ¡ d S )N)i‘  i“  Ti  iô  zutf-8)r   ZrequestÚurlopenr   ÚerrorÚ	HTTPErrorÚcoder   r	   Úreadr   ÚdecodeÚ
splitlines)r   ÚfÚerrÚrawr   r   r   r"   ;   s    
zRobotFileParser.readc                 C   s,   d|j v r| jd u r(|| _n| j |¡ d S ©NÚ*)Ú
useragentsr   r   Úappend)r   Úentryr   r   r   Ú
_add_entryH   s    

zRobotFileParser._add_entryc                 C   sP  d}t ƒ }|  ¡  |D ]}|sP|dkr4t ƒ }d}n|dkrP|  |¡ t ƒ }d}| d¡}|dkrn|d |… }| ¡ }|s|q| dd¡}t|ƒdkr|d  ¡  ¡ |d< tj	 
|d  ¡ ¡|d< |d dkrú|dkrä|  |¡ t ƒ }|j |d ¡ d}q|d dkr.|dkr6|j t|d dƒ¡ d}q|d d	krb|dkr6|j t|d d
ƒ¡ d}q|d dkr |dkr6|d  ¡  ¡ ršt|d ƒ|_d}q|d dkr|dkr6|d  d¡}t|ƒdkr|d  ¡  ¡ r|d  ¡  ¡ rtt|d ƒt|d ƒƒ|_d}q|d dkr| j |d ¡ q|dkrL|  |¡ d S )Nr   r   é   ú#ú:z
user-agentZdisallowFZallowTzcrawl-delayzrequest-rateú/Zsitemap)ÚEntryr   r-   ÚfindÚstripÚsplitÚlenÚlowerr   r   Úunquoter*   r+   Ú	rulelinesÚRuleLineÚisdigitÚintÚdelayr   Úreq_rater   )r   ÚlinesÚstater,   ÚlineÚiZnumbersr   r   r   r   Q   sj    








 ÿ
zRobotFileParser.parsec                 C   s    | j r
dS | jrdS | jsdS tj tj |¡¡}tj dd|j|j	|j
|jf¡}tj |¡}|sfd}| jD ]}| |¡rl| |¡  S ql| jrœ| j |¡S dS )NFTr   r1   )r   r	   r   r   r   r   r8   Ú
urlunparser   ÚparamsÚqueryÚfragmentÚquoter   Ú
applies_toÚ	allowancer   )r   Ú	useragentr   Z
parsed_urlr,   r   r   r   Ú	can_fetchš   s&    ÿ

zRobotFileParser.can_fetchc                 C   s>   |   ¡ sd S | jD ]}| |¡r|j  S q| jr:| jjS d S r   )r   r   rH   r=   r   ©r   rJ   r,   r   r   r   Úcrawl_delay·   s    

zRobotFileParser.crawl_delayc                 C   s>   |   ¡ sd S | jD ]}| |¡r|j  S q| jr:| jjS d S r   )r   r   rH   r>   r   rL   r   r   r   Úrequest_rateÁ   s    

zRobotFileParser.request_ratec                 C   s   | j s
d S | j S r   )r   r   r   r   r   Ú	site_mapsË   s    zRobotFileParser.site_mapsc                 C   s,   | j }| jd ur|| jg }d tt|ƒ¡S )Nz

)r   r   ÚjoinÚmapÚstr)r   r   r   r   r   Ú__str__Ð   s    
zRobotFileParser.__str__N)r   )Ú__name__Ú
__module__Ú__qualname__r   r   r   r
   r"   r-   r   rK   rM   rN   rO   rS   r   r   r   r   r      s   
			I

c                   @   s$   e Zd Zdd„ Zdd„ Zdd„ ZdS )r:   c                 C   s<   |dkr|sd}t j t j |¡¡}t j |¡| _|| _d S )Nr   T)r   r   rC   r   rG   r   rI   )r   r   rI   r   r   r   r   Ú   s
    zRuleLine.__init__c                 C   s   | j dkp| | j ¡S r(   )r   Ú
startswith)r   Úfilenamer   r   r   rH   â   s    zRuleLine.applies_toc                 C   s   | j r
dndd | j S )NZAllowZDisallowz: )rI   r   r   r   r   r   rS   å   s    zRuleLine.__str__N)rT   rU   rV   r   rH   rS   r   r   r   r   r:   ×   s   r:   c                   @   s,   e Zd Zdd„ Zdd„ Zdd„ Zdd„ Zd	S )
r2   c                 C   s   g | _ g | _d | _d | _d S r   )r*   r9   r=   r>   r   r   r   r   r   ë   s    zEntry.__init__c                 C   s‚   g }| j D ]}| d|› ¡ q
| jd ur<| d| j› ¡ | jd urf| j}| d|j› d|j› ¡ | tt| j	ƒ¡ d 
|¡S )NzUser-agent: zCrawl-delay: zRequest-rate: r1   Ú
)r*   r+   r=   r>   ZrequestsZsecondsÚextendrQ   rR   r9   rP   )r   ÚretÚagentZrater   r   r   rS   ñ   s    


zEntry.__str__c                 C   sF   |  d¡d  ¡ }| jD ](}|dkr* dS | ¡ }||v r dS qdS )Nr1   r   r)   TF)r5   r7   r*   )r   rJ   r\   r   r   r   rH   ý   s    
zEntry.applies_toc                 C   s$   | j D ]}| |¡r|j  S qdS )NT)r9   rH   rI   )r   rX   rA   r   r   r   rI   
  s    

zEntry.allowanceN)rT   rU   rV   r   rS   rH   rI   r   r   r   r   r2   é   s   r2   )
ÚcollectionsÚurllib.parser   Zurllib.requestÚ__all__Ú
namedtupler   r   r:   r2   r   r   r   r   Ú<module>   s    B