@j ddlZddlZddlmZddlmZmZmZmZddl m Z ddl m Z dddZ hd Zd eed ed eefd ZGdde ZdS)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|dr|d\}}|t vr;|rP|drd}n |drd}n|d}tj|d|}|d}|||f|S) Nz \([0-9]+\)z;;;z z...z--r )recompile startswithstripsplit _PUNCTUATIONSsubappend)rFrG_alt_recmudictlinewordphoness LC:\Projects\Historical\venv\Lib\site-packages\torchaudio/datasets/cmudict.py_parse_dictionaryrZJsj''G+-G'' tu--  zz||))$// f = # u%% && Aw vgr4((c""f~&&&& NceZdZdZ ddddddeeefded ed ed ed d f dZde d e ee effdZ d e fdZ ed e efdZd S)CMUDictaZ*CMU Pronouncing Dictionary* :cite:`cmudict` (CMUDict) dataset. Args: root (str or Path): Path to the directory where the dataset is found or downloaded. exclude_punctuations (bool, optional): When enabled, exclude the pronounciation of punctuations, such as `!EXCLAMATION-POINT` and `#HASH-MARK`. download (bool, optional): Whether to download the dataset if it is not found at root path. (default: ``False``). url (str, optional): The URL to download the dictionary from. (default: ``"http://svn.code.sf.net/p/cmusphinx/code/trunk/cmudict/cmudict-0.7b"``) url_symbols (str, optional): The URL to download the list of symbols from. (default: ``"http://svn.code.sf.net/p/cmusphinx/code/trunk/cmudict/cmudict-0.7b.symbols"``) TFr r )downloadurl url_symbolsrootrGr^r_r`rHNc||_t||_tj|jst d||jtj|z }|jtj|z }tj|s@|st d|t |d}t|||tj|s@|st d|t |d}t|||t|d5} d| D|_dddn #1swxYwYt|dd5} t| |j|_ddddS#1swxYwYdS) Nz#The root directory does not exist; z`The dictionary file is not found in the following location. Set `download=True` to download it. z\The symbol file is not found in the following location. Set `download=True` to download it. rc6g|]}|S)rO).0rVs rY z$CMUDict.__init__..s GGGdTZZ\\GGGr[zlatin-1)encoding)rG)rGr _root_pathospathisdir RuntimeErrorbasenameexists _CHECKSUMSgetr open readlines_symbolsrZ _dictionary) selfrarGr^r_r` dict_file symbol_filechecksumtexts rY__init__zCMUDict.__init__{s%9!t**w}}T_-- MKTKKLL LObg&6&6s&;&;; o(8(8(E(EE w~~i(( ; "G;DGG"~~c400H i : : :w~~k** E "I;FII"~~k488H k8 D D D +s # # HtGGdnn6F6FGGGDM H H H H H H H H H H H H H H H)S9 5 5 5 s01A1AX\XqrrrD  s s s s s s s s s s s s s s s s s ss$.$FF"%F";.G66G:=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 )ru)rvr|s rY __getitem__zCMUDict.__getitem__s""r[c*t|jS)N)lenrurvs rY__len__zCMUDict.__len__s4#$$$r[c4|jS)zLlist[str]: A list of phonemes symbols, such as ``"AA"``, ``"AE"``, ``"AH"``.)rtcopyrs rYsymbolszCMUDict.symbolss}!!###r[)T)__name__ __module__ __qualname____doc__rstrrboolr{intrrr~rpropertyrrer[rYr]r]is(&*'s Wg's's'sCI's#'s  's  's's 's's's'sR#S#U3S >%:#### %%%%%$c$$$X$$$r[r])rjrLpathlibrtypingrrrrtorch.utils.datartorchaudio._internalr rprQrrrZr]rer[rYrs  ////////////$$$$$$555555KMSU 999 xXc]$4PS9>Q$Q$Q$Q$Q$gQ$Q$Q$Q$Q$r[