a
    Rîh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$dd„Zdd„ Zdd„ Zdd„ Zdd„ Zdd„ Zdd„ Zd d!„ Zd"S )%Ú
ParserBaseziParser base class which provides some common support methods used
    by the SGML/HTML and XHTML parsers.c                 C   s   | j tu rtdƒ‚d S )Nz)_markupbase.ParserBase must be subclassed)Ú	__class__r   ÚRuntimeError©Úself© r   ú3/opt/python-3.9.24/usr/lib/python3.9/_markupbase.pyÚ__init__   s    
ÿzParserBase.__init__c                 C   s   t dƒ‚d S )Nz.subclasses of ParserBase must override error())ÚNotImplementedError)r   Úmessager   r   r   Úerror    s    ÿzParserBase.errorc                 C   s   d| _ d| _d S )Né   r   ©ÚlinenoÚoffsetr   r   r   r   Úreset$   s    zParserBase.resetc                 C   s   | j | jfS )z&Return current line number and offset.r   r   r   r   r   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   sÂ  | j }|d }|||d … dkr*|d S |||d … dv rBdS t|ƒ}|||d … dkrh|  |¡S || dkr~|  |¡S |  ||¡\}}|dk rš|S |d	kr¨d
| _||k r¾|| }|dkrø||d |… }|d	kræ|  |¡ n
|  |¡ |d S |dv r t||ƒ}|sdS | 	¡ }n|dv r<|  ||¡\}	}nt|| jv rR|d }n^|dkrž|d	krx|  
|d |¡}n$|dv r’|  d| ¡ n
|  d¡ n|  d||  ¡ |dk r¨|S q¨dS )Né   r   ú>)ú-r   éÿÿÿÿz--ú[r   Zdoctyper   z"'Z4abcdefghijklmnopqrstuvwxyzABCDEFGHIJKLMNOPQRSTUVWXYZ>   ÚlinkZlinktypeÚattlistÚelementz&unsupported '[' char in %s declarationz"unexpected '[' char in declarationz!unexpected %r char in declaration)r   ÚlenÚparse_commentÚparse_marked_sectionÚ
_scan_nameÚ_decl_othercharsZhandle_declÚunknown_declÚ_declstringlit_matchÚendÚ_parse_doctype_subsetr   )
r   r   r   r   ÚnZdecltypeÚcÚdataÚmÚnamer   r   r   Úparse_declaration@   sZ    












ÿzParserBase.parse_declarationr   c                 C   s®   | j }|  |d |¡\}}|dk r&|S |dv r@t ||d ¡}n8|dv rZt ||d ¡}n|  d||d |…  ¡ d }|s€dS |r¤| d¡}|  ||d |… ¡ | d¡S )Né   r   >   ZrcdataÚignoreZincludeZtempZcdata>   ÚifÚelseZendifz+unknown status keyword %r in marked sectionr   )	r   r'   Ú_markedsectioncloseÚsearchÚ_msmarkedsectioncloser   Ústartr)   r+   )r   r   Úreportr   ZsectNamer   Úmatchr   r   r   r&   ’   s     
zParserBase.parse_marked_sectionc                 C   sj   | j }|||d … dkr$|  d¡ t ||d ¡}|s<dS |r`| d¡}|  ||d |… ¡ | d¡S )Né   ú<!--z"unexpected call to parse_comment()r   r   )r   r   Ú_commentcloser8   r:   Zhandle_commentr+   )r   r   r;   r   r<   r   r   r   r   r%   ©   s    

zParserBase.parse_commentc           
      C   s*  | j }t|ƒ}|}||k r&|| }|dkr0|||d … }|dkrJdS |dkrp|  ||d ¡ |  d| ¡ |d |kr€dS |d |krdS |||d … dkrÀ| j|d	d
}|d	k r|S q|  |d |¡\}}|dkràdS |dvr|  ||d ¡ |  d| ¡ t| d| ƒ}	|	||ƒ}|d	k r$|S q|dkr†|d |krLdS |  |d |¡\}}|d	k rn|S || dkr$|d }q|dkrö|d }||k r¼||  ¡ r¼|d }q˜||k rð|| dkrØ|S |  ||¡ |  d¡ ndS q| ¡ r
|d }q|  ||¡ |  d| ¡ qdS )Nú<r   r   z<!r   z*unexpected char in internal subset (in %r)r=   r>   r   )r;   >   ZnotationZentityr"   r#   z)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,   ·   sp    


ÿ








z ParserBase._parse_doctype_subsetc                 C   sF   |   ||¡\}}|dkrdS | j}d||d … v rB| d|¡d S dS )Nr   r   r   )r'   r   Úfind)r   r   rF   r1   r   r   r   r   r   Ú_parse_doctype_elementú   s    z!ParserBase._parse_doctype_elementc                 C   s¶  | j }|  ||¡\}}|||d … }|dkr2dS |dkrB|d S |  ||¡\}}|dk r^|S |||d … }|dkrzdS |dkrØd||d … v r¤| d|¡d }ndS |||d …  ¡ rÆ|d }q¨||d … sèdS n|  ||¡\}}|||d … }|sdS |dv rDt||ƒ}|r&| ¡ }ndS |||d … }|sDdS |d	kr ||d … d	krddS |  |d |¡\}}|dk r†|S |||d … }|s dS |dkrB|d S qBd S )
Nr   r   r   r   r   ú(ú)ú'"ú#)r   r'   rI   rE   r*   r+   )r   r   rF   r   r1   r   r.   r0   r   r   r   Ú_parse_doctype_attlist  sX    





z!ParserBase._parse_doctype_attlistc                 C   s   |   ||¡\}}|dk r|S | j}|||d … }|s:dS |dkrJ|d S |dv rnt||ƒ}|sddS | ¡ }q"|   ||¡\}}|dk r"|S q"d S )Nr   r   r   r   rM   )r'   r   r*   r+   )r   r   rF   r1   r   r   r.   r0   r   r   r   Ú_parse_doctype_notation>  s"    

z"ParserBase._parse_doctype_notationc                 C   sä   | j }|||d … dkrR|d }|||d … }|s:dS | ¡ rP|d }q"qVq"n|}|  ||¡\}}|dk rr|S | j ||d … }|sŒdS |dv r²t||ƒ}|r¬| ¡ }qÞdS qr|dkrÂ|d S |  ||¡\}}|dk rr|S qrd S )Nr   rA   r   r   rM   r   )r   rE   r'   r*   r+   )r   r   rF   r   r   r.   r1   r0   r   r   r   Ú_parse_doctype_entityU  s4    


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   rF   r   r-   r0   rG   r1   r   r   r   r'   y  s    
ÿzParserBase._scan_namec                 C   s   d S )Nr   )r   r/   r   r   r   r)   ‹  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,   rJ   rO   rP   rQ   r'   r)   r   r   r   r   r      s"   R

C9$r   )
rZ   ÚreÚcompiler<   rS   r*   r?   r7   r9   r   r   r   r   r   Ú<module>   s   


