o
    ôjú!  ã                
   @   s¸   d dl Z d dlmZ d dlmZmZ d dlZd dlmZ d dl	m
Z
 d dlmZ d dlmZ dd	d
ddg d¢ddœdddddg d¢ddœdddddg d¢ddœdœZG dd„ de
ƒZdS )é    N)ÚPath)ÚTupleÚUnion)ÚTensor)ÚDataset)Údownload_url_to_file)Ú_extract_tarZTEDLIUM_release1z:http://www.openslr.org/resources/7/TEDLIUM_release1.tar.gzZ@30301975fd8c5cac4040c261c0852f57cfa8adbbad2ce78e77e4986957445f27Ú Útrain)r
   ÚtestÚdevzTEDLIUM.150K.dic)Úfolder_in_archiveÚurlÚchecksumÚ	data_pathÚsubsetÚsupported_subsetsÚdictZTEDLIUM_release2z;http://www.openslr.org/resources/19/TEDLIUM_release2.tar.gzZ@93281b5fcaaae5c88671c9d000b443cb3c7ea3499ad12010b3934ca41a7b9c58zTEDLIUM.152k.diczTEDLIUM_release-3z9http://www.openslr.org/resources/51/TEDLIUM_release-3.tgzZ@ad1e454d14d1ad550bc2564c462d87c7a7ec83d4dc2b9210f22ab4973b9eccdbzdata/)Úrelease1Zrelease2Úrelease3c                   @   sÎ   e Zd ZdZ				d!deeef deded	ed
eddfdd„Zdede	dede
ee	ee	e	e	f fdd„Zd"dededede	dee	gf
dd„Zde	de
ee	ee	e	e	f fdd„Zde	fdd„Zedd „ ƒZdS )#ÚTEDLIUMaÎ  *Tedlium* :cite:`rousseau2012tedlium` dataset (releases 1,2 and 3).

    Args:
        root (str or Path): Path to the directory where the dataset is found or downloaded.
        release (str, optional): Release version.
            Allowed values are ``"release1"``, ``"release2"`` or ``"release3"``.
            (default: ``"release1"``).
        subset (str, optional): The subset of dataset to use. Valid options are ``"train"``, ``"dev"``,
            and ``"test"``. Defaults to ``"train"``.
        download (bool, optional):
            Whether to download the dataset if it is not found at root path. (default: ``False``).
        audio_ext (str, optional): extension for audio file (default: ``".sph"``)
    r   r
   Fú.sphÚrootÚreleaser   ÚdownloadÚ	audio_extÚreturnNc              	      s:  || _ |t ¡ v r t| d }t| d }|r|nt| d }n
td |t ¡ ¡ƒ‚|t| d vr>td |t| d ¡ƒ‚t |¡}tj |¡}tj 	||¡}	| 
d¡d }|d	krx|d
krmtj 	||t| d ¡| _ntj 	||d|¡| _ntj 	||t| d |¡| _|r§tj | j¡s¦tj |	¡s¢t| d }
t||	|
d� t|	ƒ ntj | j¡s·td| j› d�ƒ‚g | _tj 	| jd¡}tt |¡ƒD ]A‰ ˆ  d¡�r
tj 	| jdˆ ¡}t|ƒ�#}t| ¡ ƒ}ˆ  dd¡‰ | j ‡ fdd„t|ƒD ƒ¡ W d   ƒ n	1 �sw   Y  qÉtj 	||t| d ¡| _d | _d S )Nr   r   r   zFThe release {} does not match any of the supported tedlium releases{} r   zDThe subset {} does not match any of the supported tedlium subsets{} Ú.r   r   r
   r   Úlegacyr   )Úhash_prefixz	The path zT doesn't exist. Please check the ``root`` path or set `download=True` to download itÚstmú.stmr	   c                 3   s   � | ]}ˆ |fV  qd S )N© )Ú.0Úline©Úfiler"   úl/var/www/html/academy/tools/voices/clone/sv-venv/lib/python3.10/site-packages/torchaudio/datasets/tedlium.pyÚ	<genexpr>~   s   € z#TEDLIUM.__init__.<locals>.<genexpr>r   )Ú
_ext_audioÚ_RELEASE_CONFIGSÚkeysÚRuntimeErrorÚformatÚosÚfspathÚpathÚbasenameÚjoinÚsplitÚ_pathÚisdirÚisfiler   r   ÚexistsÚ	_filelistÚsortedÚlistdirÚendswithÚopenÚlenÚ	readlinesÚreplaceÚextendÚrangeÚ
_dict_pathÚ_phoneme_dict)Úselfr   r   r   r   r   r   r   r1   Úarchiver   Zstm_pathÚfÚlr"   r%   r'   Ú__init__:   sf   þÿ
þÿ
€ÿ
 ý€
zTEDLIUM.__init__Úfileidr$   r0   c                 C   sœ   t j |d|¡}t|d ƒ�}| ¡ | }| dd¡\}}}	}
}}}W d  ƒ n1 s,w   Y  t j |d|¡}| j|| j |
|d�\}}|||||	|fS )aÙ  Loads a TEDLIUM dataset sample given a file name and corresponding sentence name.

        Args:
            fileid (str): File id to identify both text and audio files corresponding to the sample
            line (int): Line identifier for the sample inside the text file
            path (str): Dataset root path

        Returns:
            (Tensor, int, str, int, int, int):
            ``(waveform, sample_rate, transcript, talk_id, speaker_id, identifier)``
        r    r!   ú é   NÚsph)Ú
start_timeÚend_time)r.   r0   r2   r<   r>   r3   Ú_load_audior)   )rD   rI   r$   r0   Útranscript_pathrF   Ú
transcriptZtalk_idÚ_Ú
speaker_idrM   rN   Ú
identifierZ	wave_pathÚwaveformÚsample_rater"   r"   r'   Ú_load_tedlium_itemƒ   s   þzTEDLIUM._load_tedlium_itemé€>  rM   rN   rV   c                 C   s@   t t|ƒ| ƒ}t t|ƒ| ƒ}||| dœ}tj|fi |¤ŽS )a"  Default load function used in TEDLIUM dataset, you can overwrite this function to customize functionality
        and load individual sentences from a full ted audio talk file.

        Args:
            path (str): Path to audio file
            start_time (int): Time in seconds where the sample sentence stars
            end_time (int): Time in seconds where the sample sentence finishes
            sample_rate (float, optional): Sampling rate

        Returns:
            [Tensor, int]: Audio tensor representation and sample rate
        )Úframe_offsetÚ
num_frames)ÚintÚfloatÚ
torchaudioÚload)rD   r0   rM   rN   rV   Úkwargsr"   r"   r'   rO   ™   s   zTEDLIUM._load_audioÚnc                 C   s   | j | \}}|  ||| j¡S )a¼  Load the n-th sample from the dataset.

        Args:
            n (int): The index of the sample to be loaded

        Returns:
            Tuple of the following items;

            Tensor:
                Waveform
            int:
                Sample rate
            str:
                Transcript
            int:
                Talk ID
            int:
                Speaker ID
            int:
                Identifier
        )r8   rW   r4   )rD   r`   rI   r$   r"   r"   r'   Ú__getitem__­   s   zTEDLIUM.__getitem__c                 C   s
   t | jƒS )z†TEDLIUM dataset custom function overwritting len default behaviour.

        Returns:
            int: TEDLIUM dataset length
        )r=   r8   )rD   r"   r"   r'   Ú__len__Æ   s   
zTEDLIUM.__len__c                 C   sz   | j s8i | _ t| jddd��"}| ¡ D ]}| ¡  ¡ }t|dd… ƒ| j |d < qW d  ƒ n1 s3w   Y  | j  ¡ S )zƒdict[str, tuple[str]]: Phonemes. Mapping from word to tuple of phonemes.
        Note that some words have empty phonemes.
        Úrzutf-8)Úencodingé   Nr   )rC   r<   rB   r>   Ústripr3   ÚtupleÚcopy)rD   rF   r$   Úcontentr"   r"   r'   Úphoneme_dictÎ   s   þÿ
zTEDLIUM.phoneme_dict)r   r
   Fr   )rX   )Ú__name__Ú
__module__Ú__qualname__Ú__doc__r   Ústrr   ÚboolrH   r[   r   r   rW   r\   rO   ra   rb   Úpropertyrj   r"   r"   r"   r'   r   +   s2    ú
þýüûú
ù*I$"r   )r.   Úpathlibr   Útypingr   r   r]   Útorchr   Útorch.utils.datar   Ztorchaudio._internalr   Útorchaudio.datasets.utilsr   r*   r   r"   r"   r"   r'   Ú<module>   sD    ù
ù
ùí