B
    \ Çc9  ã               @   sV   d Z ddlZe d¡jZe d¡jZe d¡Ze d¡Ze d¡Z[G dd	„ d	ƒZ	dS )
zÎShared support for scanning document type declarations in HTML and XHTML.

This module is used as a foundation for the html.parser module.  It has no
documented public API and should not be used directly.

é    Nz[a-zA-Z][-_.a-zA-Z0-9]*\s*z(\'[^\']*\'|"[^"]*")\s*z--\s*>z	]\s*]\s*>z]\s*>c               @   s   e Zd ZdZdd„ Zdd„ Zdd„ Zdd	„ Zd
d„ ZdZ	dd„ Z
d#dd„Zd$dd„Zdd„ Zdd„ Zdd„ Zdd„ Zdd„ Zdd„ Zd d!„ Zd"S )%Ú
ParserBaseziParser base class which provides some common support methods used
    by the SGML/HTML and XHTML parsers.c             C   s   | j tkrtdƒ‚d S )Nz)_markupbase.ParserBase must be subclassed)Ú	__class__r   ÚRuntimeError)Úself© r   úlib/python3.7/_markupbase.pyÚ__init__   s    
zParserBase.__init__c             C   s   t dƒ‚d S )Nz.subclasses of ParserBase must override error())ÚNotImplementedError)r   Úmessager   r   r   Úerror    s    zParserBase.errorc             C   s   d| _ d| _d S )Né   r   )ÚlinenoÚoffset)r   r   r   r   Úreset$   s    zParserBase.resetc             C   s   | j | jfS )z&Return current line number and offset.)r   r   )r   r   r   r   Úgetpos(   s    zParserBase.getposc             C   sb   ||kr|S | j }| d||¡}|rN| j| | _| d||¡}||d  | _n| j| | | _|S )NÚ
r   )ÚrawdataÚcountr   Úrindexr   )r   ÚiÚjr   ZnlinesÚposr   r   r   Ú	updatepos0   s    zParserBase.updateposÚ c       
      C   sä  | j }|d }|||… dks&tdƒ‚|||d … dkrB|d S |||d … dkrZdS t|ƒ}|||d … dkr€|  |¡S || d	kr–|  |¡S |  ||¡\}}|d
k r²|S |dkrÀd| _x||k rÞ|| }|dkr||d |… }|dkr|  |¡ n
|  |¡ |d S |dkr@t	||ƒ}|s6dS | 
¡ }n|dkr\|  ||¡\}	}nt|| jkrr|d }n^|d	kr¾|dkr˜|  |d |¡}n$|dkr²|  d| ¡ n
|  d¡ n|  d||  ¡ |d
k rÄ|S qÄW dS )Né   z<!z$unexpected call to parse_declarationr   ú>)ú-r   éÿÿÿÿz--ú[r   Zdoctyper   z"'Z4abcdefghijklmnopqrstuvwxyzABCDEFGHIJKLMNOPQRSTUVWXYZ>   ÚlinkÚattlistÚelementÚlinktypez&unsupported '[' char in %s declarationz"unexpected '[' char in declarationz!unexpected %r char in declaration)r   ÚAssertionErrorÚlenÚparse_commentÚparse_marked_sectionÚ
_scan_nameÚ_decl_othercharsZhandle_declÚunknown_declÚ_declstringlit_matchÚendÚ_parse_doctype_subsetr   )
r   r   r   r   ÚnZdecltypeÚcÚdataÚmÚnamer   r   r   Úparse_declaration@   sZ    












zParserBase.parse_declarationr   c             C   sÆ   | j }|||d … dks"tdƒ‚|  |d |¡\}}|dk rB|S |dkr\t ||d ¡}n4|dkrvt ||d ¡}n|  d||d |…  ¡ |s˜dS |r¼| d¡}|  ||d |… ¡ | 	d¡S )	Né   z<![z)unexpected call to parse_marked_section()r   >   ÚtempÚcdataÚignoreÚincludeÚrcdata>   ÚelseÚendifÚifz+unknown status keyword %r in marked sectionr   )
r   r#   r'   Ú_markedsectioncloseÚsearchÚ_msmarkedsectioncloser   Ústartr)   r+   )r   r   Úreportr   ZsectNamer   Úmatchr   r   r   r&   ’   s     
zParserBase.parse_marked_sectionc             C   sj   | j }|||d … dkr$|  d¡ t ||d ¡}|s<dS |r`| d¡}|  ||d |… ¡ | d¡S )Né   z<!--z"unexpected call to parse_comment()r   r   )r   r   Ú_commentcloser=   r?   Zhandle_commentr+   )r   r   r@   r   rA   r   r   r   r   r%   ¨   s    

zParserBase.parse_commentc       
      C   s4  | j }t|ƒ}|}x||k r.|| }|dkr4|||d … }|dkrNdS |dkrt|  ||d ¡ |  d| ¡ |d |kr„dS |d |kr”dS |||d … dkrÄ| j|d	d
}|d	k r|S q|  |d |¡\}}|dkrädS |dkr|  ||d ¡ |  d| ¡ t| d| ƒ}	|	||ƒ}|d	k r,|S q|dkrŠ|d |krPdS |  |d |¡\}}|d	k rr|S || dkr,|d }q|dkrþ|d }x&||k rÂ||  ¡ rÂ|d }qžW ||k rø|| dkrà|S |  ||¡ |  d¡ ndS q| ¡ r|d }q|  ||¡ |  d| ¡ qW dS )Nú<r   r   z<!r   z*unexpected char in internal subset (in %r)rB   z<!--r   )r@   >   r    r!   ÚentityÚnotationz)unknown declaration %r in internal subsetZ_parse_doctype_ú%ú;ú]r   z%unexpected char after internal subsetz%unexpected char %r in internal subset)r   r$   r   r   r%   r'   ÚgetattrÚisspace)
r   r   Údeclstartposr   r-   r   r.   Úsr1   Úmethr   r   r   r,   ¶   sn    











z ParserBase._parse_doctype_subsetc             C   sF   |   ||¡\}}|dkrdS | j}d||d … krB| d|¡d S dS )Nr   r   r   )r'   r   Úfind)r   r   rL   r1   r   r   r   r   r   Ú_parse_doctype_elementù   s    z!ParserBase._parse_doctype_elementc             C   sÀ  | j }|  ||¡\}}|||d … }|dkr2dS |dkrB|d S xv|  ||¡\}}|dk rb|S |||d … }|dkr~dS |dkràd||d … kr¨| d|¡d }ndS x |||d …  ¡ rÌ|d }q®W ||d … sðdS n|  ||¡\}}|||d … }|s
dS |dkrLt||ƒ}|r.| ¡ }ndS |||d … }|sLdS |d	kr¨||d … d	krldS |  |d |¡\}}|dk rŽ|S |||d … }|s¨dS |dkrF|d S qFW d S )
Nr   r   r   r   r   ú(ú)z'"ú#)r   r'   rO   rK   r*   r+   )r   r   rL   r   r1   r   r.   r0   r   r   r   Ú_parse_doctype_attlist  sZ    




z!ParserBase._parse_doctype_attlistc             C   s”   |   ||¡\}}|dk r|S | j}xl|||d … }|s<dS |dkrL|d S |dkrpt||ƒ}|sfdS | ¡ }q$|   ||¡\}}|dk r$|S q$W d S )Nr   r   r   r   z'")r'   r   r*   r+   )r   r   rL   r1   r   r   r.   r0   r   r   r   Ú_parse_doctype_notation=  s$    

z"ParserBase._parse_doctype_notationc             C   sì   | j }|||d … dkrV|d }x6|||d … }|s<dS | ¡ rN|d }q$P q$W n|}|  ||¡\}}|dk rv|S xp| j ||d … }|s’dS |dkr¸t||ƒ}|r²| ¡ }qädS qx|dkrÈ|d S |  ||¡\}}|dk rx|S qxW d S )Nr   rG   r   r   z'"r   )r   rK   r'   r*   r+   )r   r   rL   r   r   r.   r1   r0   r   r   r   Ú_parse_doctype_entityT  s8    


z ParserBase._parse_doctype_entityc             C   s†   | j }t|ƒ}||krdS t||ƒ}|r\| ¡ }| ¡ }|t|ƒ |krLdS | ¡ | ¡ fS |  ||¡ |  d|||d …  ¡ d S )N)Nr   zexpected name token at %ré   )	r   r$   Ú_declname_matchÚgroupÚstripÚlowerr+   r   r   )r   r   rL   r   r-   r0   rM   r1   r   r   r   r'   x  s    
zParserBase._scan_namec             C   s   d S )Nr   )r   r/   r   r   r   r)   Š  s    zParserBase.unknown_declN)r   )r   )Ú__name__Ú
__module__Ú__qualname__Ú__doc__r   r   r   r   r   r(   r2   r&   r%   r,   rP   rT   rU   rV   r'   r)   r   r   r   r   r      s"   R

C9$r   )
r_   ÚreÚcompilerA   rX   r*   rC   r<   r>   r   r   r   r   r   Ú<module>   s   


