qi=FxdZddlZddlmZddlmZmZmZddlZddl m Z ddl m Z m Z mZmZmZmZmZddlmZmZddlmZmZdd lmZ dd lmZdd lmZeGddeZ$GddeZ%GddeZ&y#e $r7Z!e jDd e!e jDd e#de!dZ![!wwxYw)zDeepgram text-to-speech service implementation. This module provides integration with Deepgram's text-to-speech API for generating speech from text using various voice models. N) dataclass)AnyAsyncGeneratorOptional)logger) CancelFrameEndFrame ErrorFrameFrame StartFrameTTSAudioRawFrameTTSStoppedFrame) TTSSettings_warn_deprecated_param) TTSServiceWebsocketTTSService) traced_tts)connect)Statez Exception: z\In order to use DeepgramWebsocketTTSService, you need to `pip install pipecat-ai[deepgram]`.zMissing module: ceZdZdZy)DeepgramTTSSettingsz;Settings for DeepgramTTSService and DeepgramHttpTTSService.N)__name__ __module__ __qualname____doc__O/opt/pipecat/venv/lib/python3.12/site-packages/pipecat/services/deepgram/tts.pyrr,sErrc>eZdZUdZeZeed<dZddddddded e ed ed e e d ed e ef fdZ de fdZ deffd Zdeffd Zdeffd ZfdZfdZdedeeefffd ZdZdZdZdefdZdZd"de efdZe d edede!e"dffd!Z#xZ$S)#DeepgramTTSServiceaDeepgram WebSocket-based text-to-speech service. Provides real-time text-to-speech synthesis using Deepgram's WebSocket API. Supports streaming audio generation with interruption handling via the Clear message for conversational AI use cases. _settings)linear16mulawalawNzwss://api.deepgram.comr"voicebase_url sample_rateencodingsettingsapi_keyr&r'r(r)r*c ||j|jvr+td|ddj|jdt ddd}|t dtd||_||_||j|t |(d |d d d d |d |||_ ||_ ||_ d|_y) aInitialize the Deepgram WebSocket TTS service. Args: api_key: Deepgram API key for authentication. voice: Voice model to use for synthesis. .. deprecated:: 0.0.105 Use ``settings=DeepgramTTSSettings(voice=...)`` instead. base_url: WebSocket base URL for Deepgram API. Defaults to "wss://api.deepgram.com". sample_rate: Audio sample rate in Hz. If None, uses service default. encoding: Audio encoding format. Defaults to "linear16". Must be one of SUPPORTED_ENCODINGS. settings: Runtime-updatable settings. When provided alongside deprecated parameters, ``settings`` values take precedence. **kwargs: Additional arguments passed to parent InterruptibleTTSService class. Raises: ValueError: If encoding is not in SUPPORTED_ENCODINGS. zUnsupported encoding 'z'. Must be one of z, z for WebSocket TTS.Naura-2-helena-enmodelr&languager&TF)r(pause_frame_processingpush_stop_framespush_start_frameappend_trailing_spacer*r)lowerSUPPORTED_ENCODINGS ValueErrorjoinrrr/r& apply_updatesuper__init___api_key _base_url _encoding _receive_task) selfr+r&r'r(r)r*kwargsdefault_settings __class__s rr;zDeepgramTTSService.__init__@s< >> 4#;#; ;( 2DTYYtOgOgEhDii|}  /$    "7,? I%*  "%*  "    ) )( 3  ##'"!"&%     !!!rreturncy)zCheck if the service can generate metrics. Returns: True, as Deepgram WebSocket TTS service supports metrics generation. Trr@s rcan_generate_metricsz'DeepgramTTSService.can_generate_metrics rframectKt||d{|jd{y77w)zStart the Deepgram WebSocket TTS service. Args: frame: The start frame containing initialization parameters. N)r:start_connectr@rIrCs rrKzDeepgramTTSService.starts3 gmE"""mmo #848688ctKt||d{|jd{y77w)zbStop the Deepgram WebSocket TTS service. Args: frame: The end frame. N)r:stop _disconnectrMs rrPzDeepgramTTSService.stops6 gl5!!!    " rNctKt||d{|jd{y77w)zgCancel the Deepgram WebSocket TTS service. Args: frame: The cancel frame. N)r:cancelrQrMs rrSzDeepgramTTSService.cancels6 gnU###    $ rNcKt|d{|jd{|jr=|js0|j |j |j|_yyy7f7Pw)z5Connect to Deepgram WebSocket and start receive task.N)r:rL_connect_websocket _websocketr? create_task_receive_task_handler _report_errorr@rCs rrLzDeepgramTTSService._connectsmg   %%''' ??4#5#5!%!1!1$2L2LTM_M_2`!aD $6? !'s BA<BA>AB>BcKt|d{|jr*|j|jd{d|_|j d{y7S7&7 w)z6Disconnect from Deepgram WebSocket and clean up tasks.N)r:rQr? cancel_task_disconnect_websocketrZs rrQzDeepgramTTSService._disconnectsfg!###   ""4#5#56 6 6!%D ((*** $ 7 +s3A/A).A/A+A/#A-$A/+A/-A/deltac Kt||d{}d|vr5|jj|j_|j |r0|j d{|jd{|S7q7 7 w)zApply a settings delta. Args: delta: A :class:`TTSSettings` (or ``DeepgramTTSSettings``) delta. Returns: Dict mapping changed field names to their previous values. Nr&)r:_update_settingsr!r&r/_sync_model_name_to_metricsrQrL)r@r^changedrCs rr`z#DeepgramTTSService._update_settingss077 g #'>>#7#7DNN  , , . ""$ $ $--/ ! !8 % !s4BBAB)B *BB B B BcK |jr'|jjtjuryt j dg}|j d|jj|j d|j|j d|j|jddj|}dd |ji}t|| d{|_|jjj j#Dcic]\}}|j%d s||}}}t j |d |d |j'dd{y7cc}}w7 #t($rp}t j*|d||j-t/|d|d{7d|_|j'd|d{7Yd}~yd}~wwxYww)z;Connect to Deepgram WebSocket API with configured settings.Nz Connecting to Deepgram WebSocketzmodel=z encoding=z sample_rate=z /v1/speak?& AuthorizationToken )additional_headerszdg-z0: Websocket connection initialized: {"headers": } on_connected exception:  error: erroron_connection_error)rVstaterOPENrdebugappendr!r&r>r(r=r8r<websocket_connectresponseheadersitems startswith_call_event_handler Exceptionrm push_errorr )r@paramsurlrukves rrUz%DeepgramTTSService._connect_websockets J4??#8#8EJJ#F LL; <F MMF4>>#7#7"89 : MMIdnn%56 7 MML)9)9(:; <^^$Jsxx/?.@AC&&(@AG$5cg$VVDO"&!9!9!A!A!G!G!IAQ\\Z_M`1G LLD6!RSZR[[]^ _**>: : :W ; J LLD6aS1 2//*dV8A33G"HI I I"DO**+@QCI I I  JsH 2FH BH HH cK |jd{|jrotjd|jj t j ddid{|jjd{d|_|jdd{y77I7)#t$rM}tj|d||jt|d|d{7Yd}~xd}~wwxYw7d#d|_|jdd{7wxYww) z+Close WebSocket connection and reset state.Nz%Disconnecting from Deepgram WebSockettypeCloserjrkrlon_disconnected) stop_all_metricsrVrrqsendjsondumpscloseryrmrzr rx)r@rs rr]z(DeepgramTTSService._disconnect_websockets  >'') ) ) DEoo**4::vw6G+HIIIoo++--- #DO**+<= = = * J- J LLD6aS1 2//*dV8A33G"HI I I J >#DO**+<= = =sD?B>B8AB>0B:1!B>B<B>D?2D3D?8B>:B><B>> D=DDD DDDD?D<5D86D<<D?cH|jr |jStd)z3Get active websocket connection or raise exception.zWebsocket not connected)rVryrFs r_get_websocketz!DeepgramTTSService._get_websockets ???? "122r context_idc0K|jd{|jr: |jjtjddid{yy7K7#t $r$}t j|d|Yd}~yd}~wwxYww)a9Send Clear message to Deepgram when an audio context is interrupted. The Clear message will clear Deepgram's internal text buffer and stop sending audio, allowing for a new response to be generated. Args: context_id: The ID of the audio context that was interrupted. NrClearz error sending Clear message: )rrVrrrryrrm)r@rrs ron_audio_context_interruptedz/DeepgramTTSService.on_audio_context_interrupted s##%%% ?? Ioo**4::vw6G+HIII  &J I v%CA3GHH IsJBA"B3A&A$A& B$A&& B/B BBBc K|j23d{}t|trD|j}t ||j d|}|j ||d{]t|tsn tj|}|jd}|dk(rtjd|n|dk(rftjd||j}|j |t|d{|j|d{nf|d k(rtjd |nH|d k(r+tj|d |jd dntj d|77277u#tj"$rtj$d|YwxYw6yw)z5Receive and process messages from Deepgram WebSocket.N)rrMetadatazReceived metadata: FlushedzReceived Flushed: ClearedzReceived Cleared: Warningz warning: descriptionzUnknown warningzReceived unknown message type: zInvalid JSON message: )r isinstancebytesget_active_audio_context_idr r(append_to_audio_contextstrrloadsgetrtracerremove_audio_contextwarningrqJSONDecodeErrorrm)r@messagectx_idrImsgmsg_types r_receive_messagesz$DeepgramTTSService._receive_messagess!002 E E''5)99;($2B2BARXY2265AAAGS)E**W-C"wwvH:- ':3%%@A!Y. '9#%?@!%!A!A!C"::"Ov$F#77???!Y. '9#%?@"Y.#fJsww}FW/X.YZ 'Fse%LM= E B@++ELL#9'!CDE?3s|GGFGAG*F+GBF#FF#+F!,A+F#GGGF#!F##+GGGGcK|jr< ddi}|jjtj|d{yy7#t$r$}t j |d|Yd}~yd}~wwxYww)zFlush any pending audio synthesis by sending Flush command. This should be called when the LLM finishes a complete response to force generation of audio from Deepgram's internal text buffer. rFlushNz error sending Flush message: )rVrrrryrrm)r@r flush_msgrs r flush_audiozDeepgramTTSService.flush_audio=sw ?? I#W- oo**4::i+@AAA B I v%CA3GHH Is@ A>5AA A A> A A;A61A>6A;;A>textcKtj|d|d |jr&|jjtj ur|j d{d|d}|jjtj|d{dy7I7 #t$r}td|Yd}~yd}~wwxYww)a9Generate speech from text using Deepgram's WebSocket TTS API. Args: text: The text to synthesize into speech. context_id: The context ID for tracking audio frames. Yields: Frame: Audio frames containing the synthesized speech, plus start/stop frames. : Generating TTS []NSpeak)rrzUnknown error occurred: rl) rrqrVrorCLOSEDrLrrrrryr )r@rr speak_msgrs rrun_ttszDeepgramTTSService.run_ttsJs  v/vQ78 C??doo&;&;u||&Kmmo%% ")$7I%%',,TZZ -BC C CJ& D  C%=aS#AB B B CsSCAB2$B.%>B2#B0$ B2-C.B20B22 C;C CCC)N)%rrrrrSettings__annotations__r6rrintr;boolrGr rKr rPrrSrLrQrdictrr`rUr]rrrrrrr r __classcell__rCs@rr r 3sI#H""7 $0%)"26D"D"} D"  D" c] D"D"./D"Ld!!!+!b+KDcN,J@>"3 ISI "EH IHSM IC#C3C>%QU+;VCCrr ceZdZUdZeZeed<dddddddedeed e jd ed ee d ed eeffdZ de fdZedededeedffdZxZS)DeepgramHttpTTSServicezDeepgram HTTP text-to-speech service. Provides text-to-speech synthesis using Deepgram's HTTP TTS API. Supports various voice models and audio encoding formats with configurable sample rates and quality settings. r!Nzhttps://api.deepgram.comr"r%r+r&aiohttp_sessionr'r(r)r*c tddd} |tdtd|| _|| _|| j |t |d|dd| d|||_||_||_ ||_ y)a[Initialize the Deepgram TTS service. Args: api_key: Deepgram API key for authentication. voice: Voice model to use for synthesis. .. deprecated:: 0.0.105 Use ``settings=DeepgramTTSSettings(voice=...)`` instead. aiohttp_session: Shared aiohttp session for HTTP requests with connection pooling. base_url: Custom base URL for Deepgram API. Defaults to "https://api.deepgram.com". sample_rate: Audio sample rate in Hz. If None, uses service default. encoding: Audio encoding format. Defaults to "linear16". settings: Runtime-updatable settings. When provided alongside deprecated parameters, ``settings`` values take precedence. **kwargs: Additional arguments passed to parent TTSService class. Nr-r.r&T)r(r3r2r*r) rrr/r&r9r:r;r<_sessionr=r>) r@r+r&rr'r(r)r*rArBrCs rr;zDeepgramHttpTTSService.__init__ts</$    "7,? I%*  "%*  "    ) )( 3  #!!%      ' !!rrDcy)zCheck if the service can generate metrics. Returns: True, as Deepgram TTS service supports metrics generation. TrrFs rrGz+DeepgramHttpTTSService.can_generate_metricsrHrrrcKtj|d|d|jd}d|jdd}|jj |j |jdd}d |i} |jd {|jj|||| 4d {}|jd k7r3|jd {}td |jd||j|d {|j} d} |j j#| 23d {} | r|j%d {d} | s(t'| |jd|E7777x7H706d d d d {7y #1d {7swYy xYw#t$r$} t)dt+| Yd } ~ y d } ~ wwxYww)a/Generate speech from text using Deepgram's TTS API. Args: text: The text to synthesize into speech. context_id: The context ID for tracking audio frames. Yields: Frame: Audio frames containing the synthesized speech, plus start/stop frames. rrz /v1/speakrfzapplication/json)rez Content-Typenone)r/r)r( containerrN)rurr{zHTTP z: TFr)audior( num_channelsrzError getting audio: )rrqr=r<r!r&r>r(start_ttfb_metricsrpoststatusrrystart_tts_usage_metrics chunk_sizecontent iter_chunkedstop_ttfb_metricsr r r) r@rrr|rur{payloadrt error_text CHUNK_SIZE first_chunkchunkrs rrzDeepgramHttpTTSService.run_ttss  v/vQ78 *&,T]]O$ > > ?sA0G 3F0E<'F0.E>/F02#FF3F F -F7F;F<F?FFFF<F0>F0FFFFF F0FF0G F-!F$ "F-)F0,G -F00 G9GG GG )rrrrrrrrraiohttp ClientSessionrr;rrGrrr rrrs@rrris#H"" $2%)"26;";"} ;" !.. ;"  ;"c];";"./;"zd9?#9?39?>%QU+;V9?9?rr)'rr dataclassesrtypingrrrrlogururpipecat.frames.framesrr r r r r rpipecat.services.settingsrrpipecat.services.tts_servicerr(pipecat.utils.tracing.service_decoratorsrwebsockets.asyncio.clientrrswebsockets.protocolrModuleNotFoundErrorrrmryrr rrrrrs  !00JH?,F)  +   sC,sCl J?ZJ?K ,FLL;qc"#FLLf &qc* ++ ,s A==B92B44B9