qirdZddlZddlmZmZddlmZmZddlm Z ddl m Z ddl m Z mZmZmZmZmZmZmZddlmZdd lmZdd lmZmZmZmZdd lmZm Z dd l!m"Z"dd l#m$Z$ ddl%m&Z&m'Z'm(Z(m)Z)m*Z*m+Z+de0de*fdZ1eGddeZ2GddZ3Gdde e3Z4Gdde e3Z5y#e,$r7Z-e j\de-e j\de/de-dZ-[-wwxYw)z@Azure Cognitive Services Text-to-Speech service implementations.N) dataclassfield)AsyncGeneratorOptional)logger) BaseModel) CancelFrameEndFrame ErrorFrameFrameInterruptionFrame StartFrameTTSAudioRawFrameTTSStoppedFrame)FrameDirectionlanguage_to_azure_language) NOT_GIVEN TTSSettings _NotGiven_warn_deprecated_param)TextAggregationMode TTSService)Language) traced_tts)CancellationReason ResultReasonServicePropertyChannel SpeechConfigSpeechSynthesisOutputFormatSpeechSynthesizerz Exception: zCIn order to use Azure, you need to `pip install pipecat-ai[azure]`.zMissing module: sample_ratereturnctjtjtjtjtj tj d}|j|tjS)a Convert sample rate to Azure speech synthesis output format. Args: sample_rate: Sample rate in Hz. Returns: Corresponding Azure SpeechSynthesisOutputFormat enum value. Defaults to Raw24Khz16BitMonoPcm if sample rate not found. )@i>i"Vi]iDi)r Raw8Khz16BitMonoPcmRaw16Khz16BitMonoPcmRaw22050Hz16BitMonoPcmRaw24Khz16BitMonoPcmRaw44100Hz16BitMonoPcmRaw48Khz16BitMonoPcmget)r"sample_rate_maps L/opt/pipecat/venv/lib/python3.12/site-packages/pipecat/services/azure/tts.pysample_rate_to_output_formatr/0s]*==*??*AA*??*AA*?? O   {,G,\,\ ]]ceZdZUdZedZeezed<edZ eezed<edZ eezed<ed Z eezed <ed Z eezed <ed Z eezed<edZeezed<y)AzureTTSSettingsa8Settings for AzureTTSService and AzureHttpTTSService. Parameters: emphasis: Emphasis level for speech ("strong", "moderate", "reduced"). pitch: Voice pitch adjustment (e.g., "+10%", "-5Hz", "high"). rate: Speech rate adjustment (e.g., "1.0", "1.25", "slow", "fast"). role: Voice role for expression (e.g., "YoungAdultFemale"). style: Speaking style (e.g., "cheerful", "sad", "excited"). style_degree: Intensity of the speaking style (0.01 to 2.0). volume: Volume level (e.g., "+20%", "loud", "x-soft"). ctSNrr0r.zAzureTTSSettings.Ssir0)default_factoryemphasisctSr4r5r6r0r.r7zAzureTTSSettings.T9r0pitchctSr4r5r6r0r.r7zAzureTTSSettings.U)r0ratectSr4r5r6r0r.r7zAzureTTSSettings.Vr>r0rolectSr4r5r6r0r.r7zAzureTTSSettings.Wr;r0stylectSr4r5r6r0r.r7zAzureTTSSettings.Xs)r0 style_degreectSr4r5r6r0r.r7zAzureTTSSettings.YsIr0volumeN)__name__ __module__ __qualname____doc__rr9strr__annotations__r<r?rArCrErGr6r0r.r2r2Es !&6G HHcIoH"3DEE3?E!2CDD# /D!2CDD# /D"3DEE3?E$):K$LL# /L#4EFFC)OFr0r2ceZdZUdZeed<ddddddZGd d eZd e d e fd Z de de e fdZ de de fdZde de fdZy)AzureBaseTTSServicea1Base mixin class for Azure Cognitive Services text-to-speech implementations. Provides common functionality for Azure TTS services including SSML construction, voice configuration, and parameter management. This is a mixin class and should be used alongside TTSService or its subclasses. _settingsz&z<z>z"z')&<>"'ceZdZUdZdZeeed<ejZ eeed<dZ eeed<dZ eeed<dZ eeed<dZeeed<dZeeed <dZeeed <y) AzureBaseTTSService.InputParamsaInput parameters for Azure TTS voice configuration. .. deprecated:: 0.0.105 Use ``settings=AzureTTSSettings(...)`` instead. Parameters: emphasis: Emphasis level for speech ("strong", "moderate", "reduced"). language: Language for synthesis. Defaults to English (US). pitch: Voice pitch adjustment (e.g., "+10%", "-5Hz", "high"). rate: Speech rate adjustment (e.g., "1.0", "1.25", "slow", "fast"). role: Voice role for expression (e.g., "YoungAdultFemale"). style: Speaking style (e.g., "cheerful", "sad", "excited"). style_degree: Intensity of the speaking style (0.01 to 2.0). volume: Volume level (e.g., "+20%", "loud", "x-soft"). Nr9languager<r?rArCrErG)rHrIrJrKr9rrLrMrEN_USrXr<r?rArCrErGr6r0r. InputParamsrWps~ #'(3-&'/~~(8$5#x}#"hsm""hsm"#x}#&* hsm* $ $r0rZapi_keyregionc.||_||_d|_y)a)Initialize Azure-specific configuration. This method should be called by subclasses after initializing their TTSService parent. Args: api_key: Azure Cognitive Services subscription key. region: Azure region identifier (e.g., "eastus", "westus2"). N)_api_key_region_speech_synthesizer)selfr[r\s r._init_azure_basez$AzureBaseTTSService._init_azure_bases   #' r0rXr#ct|S)zConvert a Language enum to Azure language format. Args: language: The language to convert. Returns: The Azure-specific language code, or None if not supported. rrarXs r.language_to_service_languagez0AzureBaseTTSService.language_to_service_languages*(33r0textcp|jj}|j|}d|d|jjd}|jjr|d|jjdz }|jj r|d|jj dz }|jj r|d|jj dz }|dz }g}|jjr)|jd |jjd|jjr)|jd |jjd|jjr)|jd |jjd|r|d d j|dz }|jjr|d|jjdz }||z }|jjr|dz }|r|dz }|jjr|dz }|dz }|S)Nzzz z zz) rPrX _escape_textvoicerCrErAr?appendr<rGjoinr9)rarfrX escaped_textssml prosody_attrss r._construct_ssmlz#AzureBaseTTSService._construct_ssmls>>**((. .hZ8!NN0012E E  >>   /0D0D/EQG GD~~**.)D)D(EQGG~~""'$.."5"5!6a88 CKD >>    6$..*=*=)>a!@ A >>    74>>+?+?*@!B C >>  8DNN,A,A+B!!D E  i 78: :D >> " " '(?(?'@C CD  >> " " M !D  L D >>   ) )D "" r0cz|}tjjD]\}}|j||}|S)aqEscapes XML/SSML reserved characters according to Microsoft documentation. This method escapes the following characters: - & becomes & - < becomes < - > becomes > - " becomes " - ' becomes ' Args: text: The text to escape. Returns: The escaped text. )rOSSML_ESCAPE_CHARSitemsreplace)rarfrmchar escape_codes r.riz AzureBaseTTSService._escape_textsF  !4!F!F!L!L!N C D+'//kBL Cr0N)rHrIrJrKr2rMrrrrZrLrbrrrerprir6r0r.rOrO\s       %i%4(( ($ 4X 4(3- 42C2C2hr0rOceZdZdZeZddddddddededeedeedee jd eed ee d ee ffd Z d e fdZdeffd Zdeffd Zdeffd Zd e fdZded e fdZdZdZdedefdZdedefdZdZdZdZdZe jBfde"d e ffd! Z#d"Z$d'd#eefd$Z%de&d e ffd% Z'e(ded#ed e)e"dffd&Z*xZ+S)(AzureTTSServicea:Azure Cognitive Services streaming TTS service with word timestamps. Provides real-time text-to-speech synthesis using Azure's WebSocket-based streaming API. Audio chunks and word boundaries are streamed as they become available for lower latency playback and accurate word-level synchronization. N)rjr"paramssettingsaggregate_sentencestext_aggregation_moder[r\rjr"ryrzr{r|c Btdddddddddd } |tdtd|| _|tdt|s|j| _|jr|j |jnd| _|j | _|j| _|j| _|j| _ |j| _ |j| _ || j|t |8d ||dddd|| d | |j|| d|_d|_t%j&|_t%j&|_d|_d |_d |_d |_d |_d|_d|_d|_y) aWInitialize the Azure streaming TTS service. Args: api_key: Azure Cognitive Services subscription key. region: Azure region identifier (e.g., "eastus", "westus2"). voice: Voice name to use for synthesis. .. deprecated:: 0.0.105 Use ``settings=AzureTTSSettings(voice=...)`` instead. sample_rate: Audio sample rate in Hz. If None, uses service default. params: Voice and synthesis parameters configuration. .. deprecated:: 0.0.105 Use ``settings=AzureTTSSettings(...)`` instead. settings: Runtime-updatable settings. When provided alongside deprecated parameters, ``settings`` values take precedence. aggregate_sentences: Deprecated. Use text_aggregation_mode instead. .. deprecated:: 0.0.104 Use ``text_aggregation_mode`` instead. text_aggregation_mode: How to aggregate text before synthesis. **kwargs: Additional arguments passed to parent WordTTSService. Nen-US-SaraNeuralen-US modelrjrXr9r<r?rArCrErGrjryFT)r{r|push_text_framespush_stop_framespush_start_framepause_frame_processingr"rzr[r\r6)r2rrjr9rXrer<r?rArCrErG apply_updatesuper__init__rb_speech_configr`asyncioQueue _audio_queue_word_boundary_queue_word_processor_task_cumulative_audio_offset_current_sentence_base_offset_current_sentence_duration!_current_sentence_max_word_offset _last_word_last_timestamp_current_context_id) rar[r\rjr"ryrzr{r|kwargsdefault_settings __class__s r.rzAzureTTSService.__init__sP,$     "7,,2OO )55fooF !) *0 &(.  %(.  %)/ &060C0C -*0-- '    ) )( 3  3"7"!!#'#%   gf="#' #MMO$+MMO!$(!/2%47*14'  .*.04   r0r#cyzCheck if this service can generate processing metrics. Returns: True, as Azure TTS service supports metrics generation. Tr6ras r.can_generate_metricsz$AzureTTSService.can_generate_metricsj r0framecKt||d{|jryt|j|j |_|j j|j_|jjt|j|jjddtjt|jd|_|j j"j%|j&|j j(j%|j*|j j,j%|j.|j j0j%|j2|j4s%|j7|j9|_yy7w)zStart the Azure TTS service and initialize speech synthesizer. Args: frame: Start frame containing initialization parameters. N subscriptionr\z8synthesizer.synthesis.connection.synthesisConnectionImpl websocket speech_config audio_config)rstartrrr^r_rPrXspeech_synthesis_language"set_speech_synthesis_output_formatr/r"set_service_propertyrUriQueryParameterr!r` synthesizingconnect_handle_synthesizingsynthesis_completed_handle_completedsynthesis_canceled_handle_canceledsynthesis_word_boundary_handle_word_boundaryr create_task_word_processor_task_handlerrarrs r.rzAzureTTSService.startrsr gmE"""    +<< 9=8O8O5 >> ()9)9 :  00 F  " 4 4 $5--D$    --55d6O6OP   44<K|]}|j ywr4)isalnum).0cs r. z7AzureTTSService._is_punctuation_only..s: O:s)all)rarfs r._is_punctuation_onlyz$AzureTTSService._is_punctuation_onlys::T:::r0c|j}|jdz }|j|z}||jkDr||_|sy|j r|j ||y|j ||y)aHandle word boundary events from Azure SDK. Azure sends punctuation as separate word boundaries, and breaks CJK text into individual characters/particles. This method routes to language-specific handlers to properly merge and emit word boundaries. Args: evt: SpeechSynthesisWordBoundaryEventArgs from Azure Speech SDK containing word text and audio offset timing. gcAN)rf audio_offsetrrr_handle_cjk_word_boundary_handle_non_cjk_word_boundary)raevtwordsentence_relative_secondsabsolute_secondss r.rz%AzureTTSService._handle_word_boundarysxx$'$4$4|$C! ==@YY %t'M'M M5ND 2   "  * *41A B  . .t5E Fr0c|j@|jj|j|jfd|_d|_yy)z/Emit the currently buffered word if one exists.N)rr put_nowaitrrs r._emit_pending_wordz"AzureTTSService._emit_pending_wordsC ?? &  % % 0 0$//4CWCW1X Y"DO#'D  'r0r timestampc6|j||_||_y|j|r&|xj|z c_|jy|j |k7r|j||_||_y|xj|z c_y)aHandle word boundaries for CJK languages (Chinese, Japanese, Korean). CJK languages don't use spaces between words, so we merge characters together and only emit at natural break points (punctuation or whitespace boundaries). Without this logic, we don't get word output for CJK languages. Args: word: The word/character from Azure. timestamp: Timestamp in seconds. N)rrrrstriprarrs r.rz)AzureTTSService._handle_cjk_word_boundarys ?? ""DO#,D   $ $T * OOt #O  # # %  ::<4   # # %"DO#,D  4r0c|j|r"|j|xj|z c_y|j1|jj|j|jf||_||_y)aHandle word boundaries for non-CJK languages. Non-CJK languages use spaces between words, so we emit each word separately after merging any trailing punctuation. Args: word: The word from Azure. timestamp: Timestamp in seconds. N)rrrrrrs r.rz-AzureTTSService._handle_non_cjk_word_boundarysj  $ $T *t/J OOt #O  ?? &  % % 0 0$//4CWCW1X Y(r0cK |jjd{\}}|jr'|j||fg|jd{|jj t7U7!#t j $rYyt$r)}|jd||d{7Yd}~Fd}~wwxYww)zDProcess word timestamps from the queue and call add_word_timestamps.NUnknown error occurred:  error_msg exception) rr,radd_word_timestamps task_donerCancelledError Exception push_error)rartimestamp_secondses r.rz,AzureTTSService._word_processor_task_handler#s ]040I0I0M0M0O*O''++22 123T5M5M))335*O))  ]oo2J1#0NZ[o\\\ ]shCA<A85A<A:A<7C8A<:A<<CCCB=2B53B=8C=CCc|jrG|jjr0|jj|jjyyy)zsHandle audio chunks as they arrive. Args: evt: Synthesis event containing audio data. N)result audio_datarrrars r.rz$AzureTTSService._handle_synthesizing2s< ::#**//    ( ()>)> ?0:r0cf|j?|jj|j|jfd|_d|_|jr?|jj r)|jj j |_|jjdy)zmHandle synthesis completion. Args: evt: Completion event from Azure Speech SDK. N) rrrrraudio_duration total_secondsrrrs r.rz!AzureTTSService._handle_completed;s ?? &  % % 0 0$//4CWCW1X Y"DO#'D  ::#**33.1jj.G.G.U.U.WD + $$T*r0c|jjj}|tjk(r4t j |d|jjdy|jj}d|}|jr|d|jz }|jjt|y)z[Handle synthesis cancellation. Args: evt: Cancellation event. z2: Speech synthesis canceled by user (interruption)NzAzure TTS synthesis canceled: z - ) rcancellation_detailsreasonrCancelledByUserrdebugrr error_detailsr)rarrdetailsrs r.rz AzureTTSService._handle_canceledMs 0077 '77 7 LLD6!ST U    ( ( .jj55G8AI$$s7#8#8"9::    ( (9)= >r0 directioncKt|||d{t|ttfrT|j t|tr3|j r&|jdg|j d{yyyy7o7 w)zPush a frame and handle state changes. Args: frame: The frame to push. direction: The direction to push the frame. N)Resetr)r push_frame isinstancerr _reset_staterr)rarrrs r.rzAzureTTSService.push_frame_sg  222 eo/@A B    %1d6N6N.. ~t?W?WXXX7O1 C 3Ys"B BA&B >B ? B  B cfd|_d|_d|_d|_d|_d|_d|_y)zReset TTS state between turns.rN)rrrrrrrrs r.rzAzureTTSService._reset_statels9(+%-0**-'14.##' r0 context_idc<Ktj|dyw)zFlush any pending audio data.z: flushing audioN)rtrace)rars r. flush_audiozAzureTTSService.flush_audiovs v-./scKt|||d{|jd{|jrB |jj }t j |jd{|j|jjsP |jj|jj|jjsP|j jsQ |j j|j j|j jsPyy7Q7<7#t$r*}|jd||d{7Yd}~d}~wwxYw#t j$rYwxYw#t j$rYywxYww)zHandle interruption by stopping current synthesis. Args: frame: The interruption frame. direction: Frame processing direction. Nrr)r_handle_interruptionstop_all_metricsr`stop_speaking_asyncr to_threadr,rrrrempty get_nowaitr QueueEmptyr)rarr result_futurerrs r.rz$AzureTTSService._handle_interruptionzsg*5)<<<##%%%  # # ]!% 8 8 L L N '' (9(9::: ##))+ !!,,.!!++-##))+++113 ))446))335++113/ =%; ]oo2J1#0NZ[o\\\ ]%%  %%  sGE(GE+GE0*G-4F&!G<G4F? G&G+G.E00 F#9FFFGF##G&F<9G;F<<G?GGGGchKtj|d|d|jjsO|jj |jj |jjsO |j y ||_|j|_ d|_ d|_ |j|}|j j||j|d{ |jjd{}|nHt!|t"rt%t'|nt)||j*d|}|m|j*dk(r |xj|jz c_n |xj|jz c_yy77#t"$r;}t%d |t-| |j/Yd}~yd}~wwxYw#t"$r}t%d |Yd}~yd}~wwxYww) a!Generate speech from text using Azure's streaming synthesis. Args: text: The text to synthesize into speech. context_id: The context ID for tracking audio frames. Yields: Frame: Audio frames containing synthesized speech data. : Generating TTS []Nrerroraudior" num_channelsrr%r)r)rrrr r rr`rrrrrrpspeak_ssml_asyncstart_tts_usage_metricsr,rrr rLrr"rr)rarfrrnchunkrrs r.run_ttszAzureTTSService.run_ttss  v/vQ78 ##))+    ( ( *    ' ' )##))+4 C''/- +5(6:5R5R225/9<6++D1((99$?224888"&"3"3"7"7"99E}!%3(s5z::,#$($4$4%&#- E  K*##t+11T5[5[[111T5T5TT1259 :2  )A!'EFF%<<!!#    C%=aS#AB B B CsBH2 H H2A&G<F?="GG BG=H2?GG H 1H=H H2HH H/H*%H2*H//H2r4),rHrIrJrKr2SettingsrLrintrOrZboolrrrrrr rr rrrrrfloatrrrrrrr DOWNSTREAMr rrrr rrrr __classcell__rs@r.rxrxs H $%)<@/3.2?Cm m  m } m c] m ,889m +,m &d^m  ((;<m ^d&^&^P)))+)L$L ; ; ;!GF( c e B)#)%)* ]@+$?$JXIbIb Ye Y Y(0HSM0#0A#n#JGC#GC3GC>%QU+;VGCGCr0rxceZdZdZeZddddddededeedeedee jd eef fd Z d e fd Z d effd Zededed eedffdZxZS)AzureHttpTTSServicezAzure Cognitive Services HTTP-based TTS service. Provides text-to-speech synthesis using Azure's HTTP API for simpler, non-streaming synthesis. Suitable for use cases where streaming is not required and simpler integration is preferred. N)rjr"ryrzr[r\rjr"ryrzc ftdddddddddd }|tdtd||_|tdt|s|j|_|jr|j |jnd|_|j |_|j|_|j|_|j|_ |j|_ |j|_ ||j|t |8d |dd|d||j|| d|_d|_y) aNInitialize the Azure HTTP TTS service. Args: api_key: Azure Cognitive Services subscription key. region: Azure region identifier (e.g., "eastus", "westus2"). voice: Voice name to use for synthesis. .. deprecated:: 0.0.105 Use ``settings=AzureTTSSettings(voice=...)`` instead. sample_rate: Audio sample rate in Hz. If None, uses service default. params: Voice and synthesis parameters configuration. .. deprecated:: 0.0.105 Use ``settings=AzureTTSSettings(...)`` instead. settings: Runtime-updatable settings. When provided alongside deprecated parameters, ``settings`` values take precedence. **kwargs: Additional arguments passed to parent TTSService. Nr~rrrjryT)r"rrrzrr6)r2rrjr9rXrer<r?rArCrErGrrrrbrr`) rar[r\rjr"ryrzrrrs r.rzAzureHttpTTSService.__init__sM@,$     "7,,2OO )55fooF !) *0 &(.  %(.  %)/ &060C0C -*0-- '    ) )( 3  #!!%     gf="#' r0r#cyrr6rs r.rz(AzureHttpTTSService.can_generate_metricsIrr0rcKt||d{|jryt|j|j |_|j j|j_|jjt|jt|jd|_ y7w)zStart the Azure HTTP TTS service and initialize speech synthesizer. Args: frame: Start frame containing initialization parameters. Nrr)rrrrr^r_rPrXrrr/r"r!r`rs r.rzAzureHttpTTSService.startQs gmE"""    *<< 9=8O8O5 >> ()9)9 : $5--D$   #sCB>B(CrfrcKtj|d|d|j|}tj|j j |d{}|jtjk(r[|j|d{|jd{t|jdd|jd|y|jtjk(ri|j }tj"d|j|jt$j&k(rt)d|j* yyy777w) a(Generate speech from text using Azure's HTTP synthesis API. Args: text: The text to synthesize into speech. context_id: The context ID for tracking audio frames. Yields: Frame: Audio frames containing the complete synthesized speech. rrN,rrzSpeech synthesis canceled: rr)rrrprr r` speak_ssmlrrSynthesizingAudioCompletedrstop_ttfb_metricsrrr"CanceledrwarningrErrorr r)rarfrrnrrs r.rzAzureHttpTTSService.run_ttshsB  v/vQ78##D)(()A)A)L)LdSS ==LCC C..t4 4 4((* * *"'', ,,%   ]]l33 3#)#>#> NN89M9T9T8UV W#**.@.F.FF 45I5W5W4XYG4T 5 *s7AE'E 5E'E#E'*E%+B6E'#E'%E')rHrIrJrKr2rrLrrrOrZrrrrrrrr rr"r#s@r.r%r%s H $%)<@/3S(S( S( } S( c] S(,889S(+,S(jd  . # 3 >%QU+;V  r0r%)6rKr dataclassesrrtypingrrlogururpydanticrpipecat.frames.framesr r r r r rrr"pipecat.processors.frame_processorrpipecat.services.azure.commonrpipecat.services.settingsrrrrpipecat.services.tts_servicerrpipecat.transcriptions.languager(pipecat.utils.tracing.service_decoratorsrazure.cognitiveservices.speechrrrrr r!ModuleNotFoundErrorrrrrr/r2rOrxr%r6r0r.r>sG(+   >D__H4? ,^c^6Q^* G{G G,RRjvCj"5vCr_*&9_A,FLL;qc"#FLLVW &qc* ++,s$B//C+42C&&C+