o i/: @sdZddlZddlZddlmZmZddlZddlm Z ddl m Z m Z m Z ddlmZmZmZddlmZmZzddlZWn ey^Zze dee d ed edZ[wwGd d d eZGd ddeZdS)aKrisp turn analyzer for end-of-turn detection using Krisp VIVA SDK. This module provides a turn analyzer implementation using Krisp's turn detection (Tt) API to determine when a user has finished speaking in a conversation. Note: This analyzer uses a different model than KrispVivaFilter. The model path can be specified via the KRISP_VIVA_TURN_MODEL_PATH environment variable or passed directly to the constructor. N)OptionalTuple)logger)KrispVivaSDKManagerint_to_krisp_frame_durationint_to_krisp_sample_rate)BaseTurnAnalyzerBaseTurnParamsEndOfTurnState) MetricsDataTurnMetricsDataz Exception: z?In order to use KrispVivaTurn, you need to install krisp_audio.zMissing module: c@s*eZdZUdZdZeed<dZeed<dS)KrispTurnParamsaeConfiguration parameters for Krisp turn analysis. Parameters: threshold: Probability threshold for turn completion (0.0 to 1.0). Higher values require more confidence before marking turn as complete. frame_duration_ms: Frame duration in milliseconds for turn detection. Supported values: 10, 15, 20, 30, 32. g? thresholdframe_duration_msN) __name__ __module__ __qualname____doc__rfloat__annotations__rintrrV/home/ubuntu/.local/lib/python3.10/site-packages/pipecat/audio/turn/krisp_viva_turn.pyr (s  r c seZdZdZddddddeedeedeeded df fd d Zd d Z defddZ deffdd Z e d e fddZe d eefddZe d efddZe d efddZdeded efddZd eeeeffddZd d!ZZS)" KrispVivaTurnzTurn analyzer using Krisp VIVA SDK for end-of-turn detection. Uses Krisp's turn detection (Tt) API to determine when a user has finished speaking. This analyzer requires a valid Krisp model file to operate. N) model_path sample_rateparamsapi_keyrrrrreturnc stj|dz tj|dd|_Wnty(}z d|_td|d}~wwz|p0td|_ |j s>t dt d |j d sHtd tj|j sWtd |j |p[t|_d|_d|_d|_t|_d|_d|_g|_tj|_d|_d|_d|_ |r|nd }z |!||_WWdSty}zt j d|ddd|_td||d}~wwty|jrt"d|_w)ahInitialize the Krisp turn analyzer. Args: model_path: Path to the Krisp turn detection model file (.kef extension). If None, uses KRISP_VIVA_TURN_MODEL_PATH environment variable. sample_rate: Optional initial sample rate for audio processing. If provided, this will be used as the fixed sample rate. params: Configuration parameters for turn analysis behavior. api_key: Krisp SDK API key. If empty, falls back to the KRISP_VIVA_API_KEY environment variable. Raises: ValueError: If model_path is not provided and KRISP_VIVA_TURN_MODEL_PATH is not set. Exception: If model file doesn't have .kef extension. FileNotFoundError: If model file doesn't exist. RuntimeError: If Krisp SDK initialization fails. )r)rTFz Failed to initialize Krisp SDK: NKRISP_VIVA_TURN_MODEL_PATHzEModel path is not provided and KRISP_VIVA_TURN_MODEL_PATH is not set.z.Model path for KrispVivaTurn must be provided.z.kefz%Model is expected with .kef extensionzModel file not found: i>)Failed to create turn detection session: exc_info)#super__init__racquire _sdk_acquired Exception RuntimeErrorosgetenv _model_pathrerror ValueErrorendswithpathisfileFileNotFoundErrorr _params _tt_session_preload_tt_session_samples_per_frame bytearray _audio_buffer_speech_triggered_last_probability_frame_probabilitiesr INCOMPLETE _last_state_speech_stopped_time_e2e_processing_time_ms _last_metrics_create_tt_sessionrelease)selfrrrrepreload_sample_rate __class__rrr&=s^      zKrispVivaTurn.__init__c s|jrFz$t|dr|jdurd|_t|dr|jdurd|_td|_WdStyE}ztjd|ddWYd}~dSd}~wwdS)z1Release SDK reference when analyzer is destroyed.r5Nr6FzError in __del__: Tr#) r(hasattrr5r6rrCr)rr.)rDrErrrcleanups "zKrispVivaTurn.cleanupc Csz/t}|j|_t}t||_t|jj |_ ||_ t ||jj d|_ tj|}|WStyM}ztjd|ddtd||d}~ww)aVCreate a turn detection session with the specified sample rate. Args: sample_rate: Sample rate for the session Returns: krisp_audio.TtFloat instance Raises: ValueError: If sample rate or frame duration is not supported RuntimeError: If session creation fails z/Failed to create Krisp turn detection session: Tr#N) krisp_audio ModelInfor-r1TtSessionConfigrinputSampleRaterr4rinputFrameDuration modelInforr7TtFloatcreater)rr.r*)rDr model_infott_cfg tt_instancerErrrrBs   z KrispVivaTurn._create_tt_sessionc sx|j|krdSt|z||j|_|WdSty;}ztjd|ddd|_WYd}~dSd}~ww)zSet the sample rate and create/update the turn detection session. Args: sample_rate: The sample rate to set. Nr"Tr#) _sample_rater%set_sample_raterBr5clearr)rr.)rDrrErGrrrXs  zKrispVivaTurn.set_sample_ratecC|jS)zGet all probabilities from the last append_audio call. Returns: List of probability values for each frame processed in the last append_audio call. )r<rDrrrframe_probabilitiesz!KrispVivaTurn.frame_probabilitiescCrZ)zGet the last turn probability value computed. Returns: Last probability value, or None if no frames have been processed yet. )r;r[rrrlast_probabilityr]zKrispVivaTurn.last_probabilitycCrZ)zCheck if speech has been detected and triggered analysis. Returns: True if speech has been detected and turn analysis is active. )r:r[rrrspeech_triggeredr]zKrispVivaTurn.speech_triggeredcCrZ)zGet the current turn analyzer parameters. Returns: Current turn analyzer configuration parameters. )r4r[rrrrr]zKrispVivaTurn.paramsbuffer is_speechc Cs|jdurtdtj|_tjS|jdur"tdtj|_tjSz|j|g|_ t |jd}||j}|dkrDtj|_tjWS||j}|d}t |jd|}|j|d|_t j |t jd}|t jd} | d|j} tj} | D]i} |r|jstd d|_d |_d|_n |jr|jdurt|_|j| } | dkrqz| |_|j | |jr| |jjkr|jdurt|jd |_td d | |jd |_t dtj!} |"nqz| |_| WSt#y}ztj$d|d dtj}||_|WYd}~Sd}~ww)aAppend audio data for turn analysis. Args: buffer: Raw audio data bytes to append for analysis. is_speech: Whether the audio buffer contains detected speech. Returns: Current end-of-turn state after processing the audio. Nzr7r9extendr<lenbytesnp frombufferint16astypefloat32reshaper:tracer@r?time perf_counterprocesstolistr;appendr4rr rAdebugCOMPLETErYr)r.)rDr`ra total_samplesnum_complete_framescomplete_samples_countbytes_to_processaudio_to_process audio_int16 audio_float32framesstateframeprobrE error_staterrr append_audios~            zKrispVivaTurn.append_audiocs|j}d|_|j|fS)zAnalyze the current audio state to determine if turn has ended. Returns: Tuple containing the end-of-turn state and optional metrics data. Returns the last state determined by append_audio(). N)rAr>)rDmetricsrrranalyze_end_of_turn_s  z!KrispVivaTurn.analyze_end_of_turncCs"d|_|jtj|_d|_dS)z-Reset the turn analyzer to its initial state.FN)r:r9rYr r=r>r?r[rrrrYls  zKrispVivaTurn.clear)rrrrrstrrr r&rJrBrXpropertylistr\rr^boolr_rrlr rrr rrY __classcell__rrrGrr6s> Rl r)rr+rttypingrrnumpyrmlogururpipecat.audio.krisp_instancerrr%pipecat.audio.turn.base_turn_analyzerrr r pipecat.metrics.metricsr r rLModuleNotFoundErrorrEr.r)r rrrrrs&