o i0@sdZddlZddlZddlZddlZddlZddlZddlZddlZddl m Z ddlm Z ddl m Z mZmZmZmZmZddlmZddlmZddlmZdd lmZmZGd d d eZdS) #Festival backend for the phonemizerN)Logger)Path)OptionalDictListIOUnionPattern) BaseBackend)lispy) Separator)get_package_resourceversion_as_tuplec seZdZdZdZ   d6dedeeeefde dee ffdd Z e d d Z ed efd dZedefddZeddZeddZe deeeffddZdeededede deef ddZe dedefdd Ze defd!d"Zedeefd#d$Zdefd%d&Ze d'ed(edefd)d*Ze d+eedede defd,d-Z ed.eeedede defd/d0Z!edede defd1d2Z"ed3edede deefd4d5Z#Z$S)7FestivalBackendrNFlanguagepunctuation_markspreserve_punctuationloggercsttj||||d|jd|td}t|d }||_Wdn1s,wY|jd|dS)N)rrrzfestival executable is %szfestival/phonemize.scmrz loaded %s) super__init__rdebug executableropenread_script)selfrrrr script_filefscript __class__X/home/ubuntu/.local/lib/python3.10/site-packages/phonemizer/backend/festival/festival.pyr*s  zFestivalBackend.__init__cCsdS)Nfestivalr"r"r"r"r#name<szFestivalBackend.namercCsN|dur d|_dSt|}|rt|tjs t|d||_dS)aSets the festival backend to use `executable` If this is not set, the backend uses the default festival executable from the system installation. Parameters ---------- executable (str) : the path to the festival executable to use as backend. Set `executable` to None to restore the default. Raises ------ RuntimeError if `executable` is not an executable file. N is not an executable file) _FESTIVAL_EXECUTABLEpathlibris_fileosaccessX_OK RuntimeErrorresolveclsrr"r"r#set_executable@s zFestivalBackend.set_executablereturncCsx|jr|jSdtjvr+ttjd}|rtj|tjds'td|d| St d}|s6tdt| S)avReturns the absolute path to the festival executable used as backend The following precedence rule applies for executable lookup: 1. As specified by FestivalBackend.set_executable() 2. Or as specified by the environment variable PHONEMIZER_FESTIVAL_EXECUTABLE 3. Or the default 'festival' binary found on the system with ``shutil.which('festival')`` Raises ------ RuntimeError if the festival executable cannot be found or if the environment variable PHONEMIZER_FESTIVAL_EXECUTABLE is set to a non-executable file PHONEMIZER_FESTIVAL_EXECUTABLE)modezPHONEMIZER_FESTIVAL_EXECUTABLE=r&r$z"failed to find festival executable) r'r*environr(rr)r+r,r-r.shutilwhichr/r"r"r#r\s(    zFestivalBackend.executablecCs$z|WdStyYdSw)z=True if the festival executable is available, False otherwiseFT)rr-r0r"r"r# is_availables   zFestivalBackend.is_availablecCs`|}t|dgd}d}zt||d}Wt |Sty/t d|dw)zFestival version as a tupe of integers (major, minor, patch) Raises ------ RuntimeError if FestivalBackend.is_available() is False or if the version cannot be extracted for some reason. z --versionlatin1z.* ([0-9\.]+[0-9]):z%cannot extract festival version from N) r subprocess check_outputdecodestriprematchgroupAttributeErrorr-r)r0r$ long_versionfestival_version_reversionr"r"r#rFs"   zFestivalBackend.versioncCsddiS)zA dictionnary of language codes -> name supported by festival Actually only en-us (American English) is supported. zen-usz english-usr"r"r"r"r#supported_languagessz#FestivalBackend.supported_languagestextoffset separatorr?cCs6||}t|dkr gS||}||||}|S)aReturn a phonemized version of `text` with festival This function is a wrapper on festival, a text to speech program, allowing simple phonemization of some English text. The US phoneset we use is the default one in festival, as described at http://www.festvox.org/bsv/c4711.html Any opening and closing parenthesis in `text` are removed, as they interfer with the Scheme expression syntax. Moreover double quotes are replaced by simple quotes because double quotes denotes utterances boundaries in festival. Parsing a ill-formed Scheme expression during post-processing (typically with unbalanced parenthesis) raises an IndexError. r) _preprocesslen_process _postprocess)rrHrIrJr?r"r"r#_phonemize_auxs   zFestivalBackend._phonemize_auxlinecCs d|dS)z4Return the string `line` surrounded by double quotes"r"rPr"r"r#_double_quoteds zFestivalBackend._double_quotedcCs4t|tdkr d}|ddddddS)z+Remove 'forbidden' characters from the line'rQ())setreplacer?rRr"r"r#_cleaneds zFestivalBackend._cleanedcs*fdd|D}dfdd|DS)aReturns the contents of `text` formatted for festival input This function adds double quotes to begining and end of each line in text, if not already presents. The returned result is a multiline string. Empty lines in inputs are ignored. c3"|] }|dkr|VqdSrUN)rZ.0rPr8r"r# z.FestivalBackend._preprocess.. c3r[r\)rSr]r8r"r#r_r`)join)r0rH cleaned_textr"r8r#rKs zFestivalBackend._preprocessc Cstjddd}z||||j}tjdkr |dd}tjdddp}z`||j ||| d|j}|j rJ|j d|t d*}|||Wd Wt|jWd Wt|jWd S1s|wYWt|jnt|jwWd n1swYWt|jnt|jwWd d S1swYd S) a Return the raw phonemization of `text` This function delegates to festival the text analysis and syllabic structure extraction. Return a string containing the "SylStructure" relation tree of the text, as a scheme expression. zw+F)deletewin32\z\\z -b z running %sN)tempfileNamedTemporaryFilewritecloser%sysplatformrYrformatrrr TemporaryFile _run_festivalr*remove)rrHdatar%scmcmdfstderrr"r"r#rMs6        "zFestivalBackend._processrsrtc Csvztjtj|dd|d}tdd|dWStjy:}z|dt d|d |j d | d d }~ww) zRuns the festival command for phonemization Returns the raw phonemized output (need to be postprocesses). Raises a RuntimeError if festival fails. F)posix)stderrz + r:rz Command "z" returned exit status z , output is: N) r<r=shlexsplitr@subr>CalledProcessErrorseekr- returncoder)rsrtoutputerrr"r"r#ros  zFestivalBackend._run_festivalsyllcCs@|j}dd|ddD}|dd|D}|r|S||S)z3Parse a syllable from festival to phonemized outputcss$|] }|ddddVqdS)rrQrUN)rY)r^phoner"r"r#r_/s"z4FestivalBackend._postprocess_syll..r;Ncss|] }|dkr|VqdSr\r")r^or"r"r#r_0s)rrb)rrJr?sepoutr"r"r#_postprocess_syll+sz!FestivalBackend._postprocess_syllwordcs:j}|fdd|ddD}r|S||S)z/Parse a word from festival to phonemized outputc3s|] }|VqdS)N)r)r^rr0rJr?r"r#r_7s  z4FestivalBackend._postprocess_word..r;N)syllablerb)r0rrJr?rrr"rr#_postprocess_word3s  z!FestivalBackend._postprocess_wordcCsT|j}g}t|D]}||||}|dkr||q ||}|r&|S||S)z/Parse a line from festival to phonemized outputrU)rr parserappendrb)r0rPrJr?rrrr"r"r#_postprocess_line<s  z!FestivalBackend._postprocess_linetreecsfdd|dDS)z8Conversion from festival syllable tree to desired formatcs"g|] }|dvr|qS))rUz (nil nil nil))rr]rr"r# Ls  z0FestivalBackend._postprocess..ra)ry)r0rrJr?r"rr#rNIszFestivalBackend._postprocess)NFN)%__name__ __module__ __qualname____doc__r'strrr r boolrr staticmethodr% classmethodr1rrr9rFrrGrintr rOrSrZrKrMrrorrrrN __classcell__r"r"r r#r#sV (  &  * $ (r)rr*r(r@rxr6r<rkrgloggingrrtypingrrrrr r phonemizer.backend.baser phonemizer.backend.festivalr phonemizer.separatorr phonemizer.utilsrrrr"r"r"r#s"