U
    ‹¾úg®,  ã                   @   s  d Z ddlZddlZddlmZmZ G dd„ dƒZG dd„ deƒZG d	d
„ d
eƒZ	e 
d¡Ze 
d¡Ze 
d¡Ze 
d¡Ze 
d¡Zdd„ Zdd„ ZG dd„ dƒZG dd„ deƒZG dd„ deƒZG dd„ deƒZG dd„ deƒZG dd„ deƒZG d d!„ d!eƒZd"d#„ ZdS )$u´  
A partial parser for WebVTT segments. Interprets enough of the WebVTT stream
to be able to assemble a single stand-alone subtitle file, suitably adjusting
timestamps on the way, while everything else is passed through unmodified.

Regular expressions based on the W3C WebVTT specification
<https://www.w3.org/TR/webvtt1/>. The X-TIMESTAMP-MAP extension is described
in RFC 8216 Â§3.5 <https://tools.ietf.org/html/rfc8216#section-3.5>.
é    Né   )Úint_or_noneÚtimetuple_from_msecc                   @   s8   e Zd ZdZdd„ Zdd„ Zdd„ Zdd	„ Zd
d„ ZdS )Ú_MatchParserz”
    An object that maintains the current parsing position and allows
    conveniently advancing it as syntax elements are successfully parsed.
    c                 C   s   || _ d| _d S ©Nr   )Ú_dataÚ_pos)ÚselfÚstring© r   ú1/tmp/pip-unpacked-wheel-q5ljy6pj/yt_dlp/webvtt.pyÚ__init__   s    z_MatchParser.__init__c                 C   sN   t |tjƒr| | j| j¡S t |tƒrB| j || j¡r>t|ƒS d S t	|ƒ‚d S ©N)
Ú
isinstanceÚreÚPatternÚmatchr   r   ÚstrÚ
startswithÚlenÚ
ValueError©r	   Úrr   r   r   r      s    
z_MatchParser.matchc                 C   sh   |d krd}nHt |tjƒr*t| d¡ƒ}n,t |tƒr>t|ƒ}nt |tƒrN|}nt|ƒ‚|  j|7  _|S r   )	r   r   ÚMatchr   Úgroupr   Úintr   r   )r	   ZbyÚamtr   r   r   Úadvance$   s    


z_MatchParser.advancec                 C   s   |   |  |¡¡S r   )r   r   r   r   r   r   Úconsume2   s    z_MatchParser.consumec                 C   s   t | ƒS r   )Ú_MatchChildParser©r	   r   r   r   Úchild5   s    z_MatchParser.childN)	Ú__name__Ú
__module__Ú__qualname__Ú__doc__r   r   r   r   r!   r   r   r   r   r      s   	r   c                       s(   e Zd ZdZ‡ fdd„Zdd„ Z‡  ZS )r   zÝ
    A child parser state, which advances through the same data as
    its parent, but has an independent position. This is useful when
    advancing through syntax elements we might later want to backtrack
    from.
    c                    s    t ƒ  |j¡ || _|j| _d S r   )Úsuperr   r   Ú_MatchChildParser__parentr   )r	   Úparent©Ú	__class__r   r   r   A   s    z_MatchChildParser.__init__c                 C   s   | j | j_ | jS )zW
        Advance the parent state to the current position of this child state.
        )r   r'   r    r   r   r   ÚcommitF   s    
z_MatchChildParser.commit)r"   r#   r$   r%   r   r+   Ú__classcell__r   r   r)   r   r   9   s   r   c                       s   e Zd Z‡ fdd„Z‡  ZS )Ú
ParseErrorc                    s6   |j |j|jd … }tƒ  d|j› d|›d¡ d S )Néd   zParse error at position z (near ú))r   r   r&   r   )r	   ÚparserÚdatar)   r   r   r   O   s    zParseError.__init__)r"   r#   r$   r   r,   r   r   r)   r   r-   N   s   r-   zL(?x)
    (?:([0-9]{1,}):)?
    ([0-9]{2}):
    ([0-9]{2})\.
    ([0-9]{3})?
z\Zz(?:\r\n|[\r\n]|$)z(?:\r\n|[\r\n])+z[ \t]*c                 C   s    dt dd„ t|  ¡ dƒD ƒƒ S )z•
    Convert a parsed WebVTT timestamp (a re.Match obtained from _REGEX_TS)
    into an MPEG PES timestamp: a tick counter at 90 kHz resolution.
    éZ   c                 s   s"   | ]\}}t |pd ƒ| V  qdS )r   N)r   )Ú.0ÚpartZmultr   r   r   Ú	<genexpr>i   s    z_parse_ts.<locals>.<genexpr>)i€î6 i`ê  iè  r   )ÚsumÚzipÚgroups©Útsr   r   r   Ú	_parse_tsd   s    
ÿr;   c                 C   s   dt t| d d ƒƒ S )zn
    Convert an MPEG PES timestamp into a WebVTT timestamp.
    This will lose sub-millisecond precision.
    z%02u:%02u:%02u.%03ué-   r2   )r   r   r9   r   r   r   Ú
_format_tsm   s    r=   c                   @   s,   e Zd ZdZdd„ Zedd„ ƒZdd„ ZdS )	ÚBlockz#
    An abstract WebVTT block.
    c                 K   s"   |  ¡ D ]\}}t| ||ƒ qd S r   )ÚitemsÚsetattr)r	   ÚkwargsÚkeyÚvalr   r   r   r   z   s    zBlock.__init__c                 C   s.   |  | j¡}|sd S | |¡ | | d¡dS )Nr   )Úraw)r   Ú_REGEXr   r   )Úclsr0   Úmr   r   r   Úparse~   s
    
zBlock.parsec                 C   s   |  | j¡ d S r   )ÚwriterD   ©r	   Ústreamr   r   r   Ú
write_into†   s    zBlock.write_intoN)r"   r#   r$   r%   r   ÚclassmethodrH   rL   r   r   r   r   r>   u   s
   
r>   c                   @   s   e Zd ZdZdS )ÚHeaderBlockzm
    A WebVTT block that may only appear in the header part of the file,
    i.e. before any cue blocks.
    N)r"   r#   r$   r%   r   r   r   r   rN   Š   s   rN   c                   @   sh   e Zd Ze d¡Ze d¡Ze d¡Ze d¡Ze d¡Z	e d¡Z
edd„ ƒZed	d
„ ƒZdd„ ZdS )ÚMagicz,\ufeff?WEBVTT([ \t][^\r\n]*)?(?:\r\n|[\r\n])zX-TIMESTAMP-MAP=zLOCAL:zMPEGTS:([0-9]+)z[ \t]*,[ \t]*z6(?:(?!-->)[^\r\n])+:(?:(?!-->)[^\r\n])+(?:\r\n|[\r\n])c                 C   s¸   |  ¡ }| | j¡}|rL| t¡}|d kr2t|ƒ‚t|ƒ}|d kr„t|ƒ‚n8| | j¡}|r|t| d¡ƒ}|d kr„t|ƒ‚nt|ƒ‚| | j	¡r’q| t
¡ržq¨t|ƒ‚q| ¡  ||fS )Nr   )r!   r   Ú_REGEX_TSMAP_LOCALÚ	_REGEX_TSr-   r;   Ú_REGEX_TSMAP_MPEGTSr   r   Ú_REGEX_TSMAP_SEPÚ	_REGEX_NLr+   )rF   r0   rG   ÚlocalÚmpegtsr   r   r   Z__parse_tsmap©   s,    




zMagic.__parse_tsmapc                 C   sœ   |  ¡ }| | j¡}|s t|ƒ‚| d¡}d\}}}| t¡s„| | j¡rZ|  |¡\}}q4| | j¡}|rz|| d¡7 }q4t|ƒ‚q4| 	¡  | ||||dS )Nr   )NNÚ r   )ÚextrarV   rU   Úmeta)
r!   r   rE   r-   r   rT   Ú_REGEX_TSMAPÚ_Magic__parse_tsmapÚ_REGEX_METAr+   )rF   r0   rG   rX   rU   rV   rY   r   r   r   rH   Ç   s"    



zMagic.parsec                 C   s°   |  d¡ | jd k	r |  | j¡ |  d¡ | js6| jr|  d¡ |  t| jd k	rV| jndƒ¡ |  d¡ |  t| jd k	r~| jndƒ¡ |  d¡ | jr¢|  | j¡ |  d¡ d S )NZWEBVTTÚ
zX-TIMESTAMP-MAP=LOCAL:r   z,MPEGTS:)rI   rX   rU   rV   r=   r   rY   rJ   r   r   r   rL   Ý   s    





zMagic.write_intoN)r"   r#   r$   r   ÚcompilerE   rZ   rP   rR   rS   r\   rM   r[   rH   rL   r   r   r   r   rO   ’   s   







rO   c                   @   s   e Zd Ze d¡ZdS )Ú
StyleBlockzs(?x)
        STYLE[\ \t]*(?:\r\n|[\r\n])
        ((?:(?!-->)[^\r\n])+(?:\r\n|[\r\n]))*
        (?:\r\n|[\r\n])
    N©r"   r#   r$   r   r^   rE   r   r   r   r   r_   í   s   r_   c                   @   s   e Zd Ze d¡ZdS )ÚRegionBlockze(?x)
        REGION[\ \t]*
        ((?:(?!-->)[^\r\n])+(?:\r\n|[\r\n]))*
        (?:\r\n|[\r\n])
    Nr`   r   r   r   r   ra   õ   s   ra   c                   @   s   e Zd Ze d¡ZdS )ÚCommentBlockzo(?x)
        NOTE(?:\r\n|[\ \t\r\n])
        ((?:(?!-->)[^\r\n])+(?:\r\n|[\r\n]))*
        (?:\r\n|[\r\n])
    Nr`   r   r   r   r   rb   ý   s   rb   c                   @   st   e Zd ZdZe d¡Ze d¡Ze d¡Ze d¡Z	e
dd„ ƒZdd	„ Zed
d„ ƒZdd„ Ze
dd„ ƒZdd„ ZdS )ÚCueBlockz6
    A cue block. The payload is not interpreted.
    z$((?:(?!-->)[^\r\n])+)(?:\r\n|[\r\n])z[ \t]+-->[ \t]+z[ \t]+((?:(?!-->)[^\r\n])+)z[^\r\n]+(?:\r\n|[\r\n])?c                 C   sî   |  ¡ }d }| | j¡}|r&| d¡}| t¡}|s8d S | | j¡sHd S | t¡}|sZd S | | j¡}| t¡ | t¡s~d S t	|ƒ}t	|ƒ}|d k	r | d¡nd }	t
 ¡ }
| | j¡}|s¾qÐ|
 | d¡¡ q¬| ¡  | ||||	|
 ¡ dS )Nr   r   )ÚidÚstartÚendÚsettingsÚtext)r!   r   Ú	_REGEX_IDr   rQ   Ú_REGEX_ARROWÚ_REGEX_SETTINGSÚ_REGEX_OPTIONAL_WHITESPACErT   r;   ÚioÚStringIOÚ_REGEX_PAYLOADrI   r+   Úgetvalue)rF   r0   Zid_rG   Zm0Úm1Úm2re   rf   rg   rh   r   r   r   rH     sB    




  ýzCueBlock.parsec                 C   sŽ   | j d k	r | | j ¡ | d¡ | t| jƒ¡ | d¡ | t| jƒ¡ | jd k	rj| d¡ | | j¡ | d¡ | | j¡ | d¡ d S )Nr]   z --> ú )rd   rI   r=   re   rf   rg   rh   rJ   r   r   r   rL   7  s    





zCueBlock.write_intoc                 C   s   | j | j| j| j| jdœS )N©rd   re   rf   rh   rg   rt   r    r   r   r   Úas_jsonE  s    ûzCueBlock.as_jsonc                 C   s   | j |j kS r   )ru   ©r	   Úotherr   r   r   Ú__eq__O  s    zCueBlock.__eq__c                 C   s&   | |d |d |d |d |d dS )Nrd   re   rf   rh   rg   rt   r   )rF   Újsonr   r   r   Ú	from_jsonR  s    ûzCueBlock.from_jsonc                 C   sJ   | j |j krdS | j|jkr dS | j| j  koD|j  koD|jkS   S )NF)rh   rg   re   rf   rv   r   r   r   Úhinges\  s
    zCueBlock.hingesN)r"   r#   r$   r%   r   r^   ri   rj   rk   ro   rM   rH   rL   Úpropertyru   rx   rz   r{   r   r   r   r   rc     s   




'
	
	rc   c                 c   sÄ   t |  ¡ ƒ}t |¡V  | t¡st| t¡r.qt |¡}|rD|V  qt	 |¡}|rZ|V  qt
 |¡}|rt|V  qqtq| t¡sÀ| t¡rŠqtt
 |¡}|r |V  qtt |¡}|r¶|V  qtt|ƒ‚qtdS )zŽ
    A generator that yields (partially) parsed WebVTT blocks when given
    a bytes object containing the raw contents of a WebVTT file.
    N)r   ÚdecoderO   rH   r   Ú
_REGEX_EOFr   Ú_REGEX_BLANKra   r_   rb   rc   r-   )Zfrag_contentr0   Úblockr   r   r   Úparse_fragmentd  s<    








r   )r%   rm   r   Úutilsr   r   r   r   Ú	Exceptionr-   r^   rQ   r~   rT   r   rl   r;   r=   r>   rN   rO   r_   ra   rb   rc   r   r   r   r   r   Ú<module>   s*   
(





	[_