U %dx @sddlZddlmZddlmZmZddlZddlmZddl m Z ddl m Z ddl mZdZd Zd d d d dddgZddddddddZddZeeeeeeeeeeefdddZGddde ZdS)N)Path)TupleUnion)Tensor)download_url_to_file)Dataset)extract_archiveztrain-clean-100Z LibriSpeechz dev-cleanz dev-otherz test-cleanz test-otherztrain-clean-360ztrain-other-500Z@76f87d090650617fca0cac8f88b9416e0ebf80350acb97b343a85fa903728ab3Z@12661c48e8c3fe1de2c1caa4c3e135193bfb1811584f11f569dd12645aa84365Z@39fde525e59672dc6d1551919b1478f724438a95aa55f874b576be21967e6c23Z@d09c181bba5cf717b3dee7d4d592af11a3ee3a09e08ae025c5506f6ebe961c29Z@d4ddd1d5a6ab303066f14971d768ee43278a5f2a0aa43dc716b0e64ecbbbf6e2Z@146a56496217e96c14334a160df97fffedd6e0a04e66b9c5af0d40be3c792ecfZ@ddb22f27f96ec163645d53215559df6aa36515f26e01dd70798188350adcb6d2)z4http://www.openslr.org/resources/12/dev-clean.tar.gzz4http://www.openslr.org/resources/12/dev-other.tar.gzz5http://www.openslr.org/resources/12/test-clean.tar.gzz5http://www.openslr.org/resources/12/test-other.tar.gzz:http://www.openslr.org/resources/12/train-clean-100.tar.gzz:http://www.openslr.org/resources/12/train-clean-360.tar.gzz:http://www.openslr.org/resources/12/train-other-500.tar.gzcCs^d}d}||}tj||}tj||}tj|sRt|d}t|||dt|dS)Nz$http://www.openslr.org/resources/12/z.tar.gz)Z hash_prefix)ospathjoinisfile _CHECKSUMSgetrr)rooturlbase_urlZ ext_archivefilenamearchive download_urlZchecksumrC/tmp/pip-unpacked-wheel-lbdmvq91/torchaudio/datasets/librispeech.pydownload_librispeech!s  r)fileidr ext_audioext_txtreturnc Cs|d\}}}|d|d|}||}tj||||}t|\} } |d||} tj|||| } t| >} | D]$} | dd\}}||kr~qq~td|W5QRX| | |t |t |t |fS)N- zTranslation not found for ) splitr r r torchaudioloadopenstripFileNotFoundErrorint)rr rrZ speaker_idZ chapter_idZ utterance_idZ fileid_audioZ file_audioZwaveformZ sample_rateZ file_textftlineZ fileid_textZ transcriptrrrload_librispeech_item.s( r(c@sleZdZdZdZdZeedfee e fe e e ddddZ e eee e e e e fd d d Ze d d dZdS) LIBRISPEECHaCreate a Dataset for *LibriSpeech* [:footcite:`7178964`]. Args: root (str or Path): Path to the directory where the dataset is found or downloaded. url (str, optional): The URL to download the dataset from, or the type of the dataset to dowload. Allowed type values are ``"dev-clean"``, ``"dev-other"``, ``"test-clean"``, ``"test-other"``, ``"train-clean-100"``, ``"train-clean-360"`` and ``"train-other-500"``. (default: ``"train-clean-100"``) folder_in_archive (str, optional): The top-level directory of the dataset. (default: ``"LibriSpeech"``) download (bool, optional): Whether to download the dataset if it is not found at root path. (default: ``False``). z .trans.txtz.flacFN)rrfolder_in_archivedownloadrcCs|tkrtd|dtdt|}tj||||_tj|jsj|rXt||nt d|jdt ddt |j d|j D|_dS) Nz Invalid url 'z' given; please provide one of .zDataset not found at z5. Please set `download=True` to download the dataset.css|]}t|jVqdSN)strstem).0prrr wsz'LIBRISPEECH.__init__..z*/*/*) _DATA_SUBSETS ValueErrorr fspathr r _pathisdirr RuntimeErrorsortedrglob _ext_audio_walker)selfrrr*r+rrr__init__bs   zLIBRISPEECH.__init__)nrcCs|j|}t||j|j|jS)aLoad the n-th sample from the dataset. Args: n (int): The index of the sample to be loaded Returns: (Tensor, int, str, int, int, int): ``(waveform, sample_rate, transcript, speaker_id, chapter_id, utterance_id)`` )r<r(r6r;_ext_txt)r=r?rrrr __getitem__ys zLIBRISPEECH.__getitem__)rcCs t|jSr-)lenr<)r=rrr__len__szLIBRISPEECH.__len__)__name__ __module__ __qualname____doc__r@r;URLFOLDER_IN_ARCHIVErr.rboolr>r%rrrArCrrrrr)Os   r))r pathlibrtypingrrr ZtorchrZ torch.hubrZtorch.utils.datarZtorchaudio.datasets.utilsrrHrIr3r rr.r%r(r)rrrrsB        !