:jfddlZddlZddlmZddlmZmZmZmZddl m Z ddl m Z dddZ hd Zd eed ed eefd ZGdde Zy)N)Path)IterableListTupleUnion)Dataset)download_url_to_file@209a8b4cd265013e96f4658632a9878103b0c5abf62b50d4ef3ae1be226b29e4@408ccaae803641c6d7b626b6299949320c2dbca96b2220fd3fb17887b023b027)Bhttp://svn.code.sf.net/p/cmusphinx/code/trunk/cmudict/cmudict-0.7bJhttp://svn.code.sf.net/p/cmusphinx/code/trunk/cmudict/cmudict-0.7b.symbols>8.DOT+PLUS-DASH"QUOTE'QUOTE(PAREN)PAREN,COMMA--DASH.POINT/SLASH:COLON{BRACE(PARENS)PARENS-HYPHEN.PERIOD"UNQUOTE%PERCENT.DECIMAL "END-QUOTE "IN-QUOTES #HASH-MARK &ERSAND 'END-QUOTE )END-PAREN .FULL-STOP #POUND-SIGN #SHARP-SIGN (LEFT-PAREN )END-PARENS ...ELLIPSIS ;SEMI-COLON {LEFT-BRACE {OPEN-BRACE "CLOSE-QUOTE 'INNER-QUOTE (PARENTHESES )CLOSE-PAREN )RIGHT-PAREN }CLOSE-BRACE }RIGHT-BRACE "DOUBLE-QUOTE "END-OF-QUOTE 'SINGLE-QUOTE (BEGIN-PARENS)END-THE-PAREN;SEMI-COLON(1)?QUESTION-MARK(IN-PARENTHESES)UN-PARENTHESES'END-INNER-QUOTE)END-PARENTHESES(OPEN-PARENTHESES!EXCLAMATION-POINT)CLOSE-PARENTHESESlinesexclude_punctuationsreturnctjd}g}|D]}|r|jdr|jj d\}}|t vr0|rD|jdrd}n|jdrd}n|d}tj |d|}|j d}|j||f|S) Nz \([0-9]+\)z;;;z z...z--r )recompile startswithstripsplit _PUNCTUATIONSsubappend)rFrG_alt_recmudictlinewordphoness p/home/mcse/projects/srt_converter/srt-converter-venv/lib/python3.12/site-packages/torchaudio/datasets/cmudict.py_parse_dictionaryrZJsjj'G+-G'tu- zz|))$/ f = #u%&Aw vvgr4(c"f~&/'2 NceZdZdZ ddddddeeefdeded ed ed d f d Zde d e ee effdZ d e fdZ ed e efdZy )CMUDictaZ*CMU Pronouncing Dictionary* :cite:`cmudict` (CMUDict) dataset. Args: root (str or Path): Path to the directory where the dataset is found or downloaded. exclude_punctuations (bool, optional): When enabled, exclude the pronounciation of punctuations, such as `!EXCLAMATION-POINT` and `#HASH-MARK`. download (bool, optional): Whether to download the dataset if it is not found at root path. (default: ``False``). url (str, optional): The URL to download the dictionary from. (default: ``"http://svn.code.sf.net/p/cmusphinx/code/trunk/cmudict/cmudict-0.7b"``) url_symbols (str, optional): The URL to download the list of symbols from. (default: ``"http://svn.code.sf.net/p/cmusphinx/code/trunk/cmudict/cmudict-0.7b.symbols"``) Fr r )downloadurl url_symbolsrootrGr^r_r`rHNc||_t||_tjj |jst d||jtjj|z }|jtjj|z }tjj|s3|st d|tj|d}t|||tjj|s3|st d|tj|d}t|||t|d5} | jD cgc]} | jc} |_dddt|dd5} t!| j|j|_dddycc} w#1swYPxYw#1swYyxYw)Nz#The root directory does not exist; z`The dictionary file is not found in the following location. Set `download=True` to download it. z\The symbol file is not found in the following location. Set `download=True` to download it. rzlatin-1)encoding)rG)rGr _root_pathospathisdir RuntimeErrorbasenameexists _CHECKSUMSgetr open readlinesrO_symbolsrZ _dictionary) selfrarGr^r_r` dict_file symbol_filechecksumtextrVs rY__init__zCMUDict.__init__{s%9!t*ww}}T__-!DTFKL LOObgg&6&6s&;; oo(8(8(EE ww~~i(";;D+G"~~c40H i :ww~~k*";;F-I"~~k48H k8 D +s # Ht6:nn6FGdTZZ\GDM H)S9 5 s01AX\XqXqrD  s sH H H s ss*G*GG+G$GG!$G-nc |j|S)aLoad the n-th sample from the dataset. Args: n (int): The index of the sample to be loaded. Returns: Tuple of a word and its phonemes str: Word List[str]: Phonemes )rq)rrrxs rY __getitem__zCMUDict.__getitem__s""r[c,t|jS)N)lenrqrrs rY__len__zCMUDict.__len__s4##$$r[c6|jjS)zLlist[str]: A list of phonemes symbols, such as ``"AA"``, ``"AE"``, ``"AH"``.)rpcopyr}s rYsymbolszCMUDict.symbolss}}!!##r[)T)__name__ __module__ __qualname____doc__rstrrboolrwintrrrzr~propertyrr[rYr]r]is(&*'s Wg'sCI's#'s  's  's's 'sR#S#U3S >%:# %%$c$$r[r])rfrLpathlibrtypingrrrrtorch.utils.datartorchaudio._internalr rlrQrrrZr]rr[rYrsh //$5KMSU 9 xXc]$4PS9>Q$gQ$r[