Files
lucky-cat/components/control_plane/internal/control_plane_internal.h
T
chenjintao 0cfb10252c fix(multimodal): harden playback and client lifecycle handling
- retry the server connection prompt after transient playback failures
- serialize multimodal client and session start-stop operations
- tag downstream responses to ignore stale packets and playback events
- complete bounded teardown and clean callbacks after partial startup
2026-08-14 15:24:29 +08:00

165 lines
7.7 KiB
C

/*
* SPDX-FileCopyrightText: 2024-2026 Espressif Systems (Shanghai) CO LTD
*
* SPDX-License-Identifier: Apache-2.0
*/
#pragma once
#include <stdbool.h>
#include <stddef.h>
#include <stdint.h>
#include "control_plane.h"
#include "domain.h"
#include "esp_err.h"
#include "control_plane_events.h"
control_plane_direction_command_t control_plane_direction_parse_command_internal(
const char *input,
size_t *out_direction_index);
bool control_plane_extract_model_command(const char *text,
control_plane_model_command_type_t *out_command_type,
char *out_value,
size_t out_value_capacity,
size_t *out_value_index,
bool *out_value_index_valid);
const char *control_plane_multimodal_resolve_upstream_mode(void);
bool control_plane_multimodal_is_push2talk_mode(void);
const char *control_plane_multimodal_resolve_voice_id(const char *auth_voice_id);
uint32_t control_plane_multimodal_resolve_downstream_transmit_rate_limit(
uint32_t configured_limit);
uint32_t control_plane_multimodal_resolve_opus_frame_duration_ms(void);
void control_plane_multimodal_generate_uuid(char *out, size_t out_len);
esp_err_t control_plane_direction_calibrate_internal(
const domain_direction_calibration_t *config,
int timeout_ms);
/* Daily fortune fetch/apply policy (control_plane-owned). */
esp_err_t control_plane_fortune_fetch(const char *target_date, domain_fortune_result_t *out_result);
typedef void (*control_plane_multimodal_responding_text_callback_t)(const char *text,
bool finished,
void *user_data);
typedef enum {
CONTROL_PLANE_MULTIMODAL_EVENT_TRANSPORT_CONNECTED = 0,
CONTROL_PLANE_MULTIMODAL_EVENT_TRANSPORT_DISCONNECTED,
CONTROL_PLANE_MULTIMODAL_EVENT_TRANSPORT_ERROR,
CONTROL_PLANE_MULTIMODAL_EVENT_TASK_STARTED,
CONTROL_PLANE_MULTIMODAL_EVENT_TASK_FAILED,
CONTROL_PLANE_MULTIMODAL_EVENT_SERVER_LISTENING,
CONTROL_PLANE_MULTIMODAL_EVENT_SERVER_NOT_LISTENING,
CONTROL_PLANE_MULTIMODAL_EVENT_RESPONDING_STARTED,
CONTROL_PLANE_MULTIMODAL_EVENT_RESPONDING_CONTENT,
CONTROL_PLANE_MULTIMODAL_EVENT_RESPONDING_ENDED,
CONTROL_PLANE_MULTIMODAL_EVENT_SPEECH_STARTED,
CONTROL_PLANE_MULTIMODAL_EVENT_SPEECH_CONTENT,
CONTROL_PLANE_MULTIMODAL_EVENT_SPEECH_ENDED,
CONTROL_PLANE_MULTIMODAL_EVENT_OUTPUT_ERROR,
CONTROL_PLANE_MULTIMODAL_EVENT_DOWNSTREAM_AUDIO,
CONTROL_PLANE_MULTIMODAL_EVENT_DOWNSTREAM_STALLED,
} control_plane_multimodal_event_t;
typedef void (*control_plane_multimodal_event_callback_t)(control_plane_multimodal_event_t event,
void *user_data);
typedef void (*control_plane_multimodal_downstream_audio_callback_t)(const uint8_t *data,
size_t len,
void *user_data);
typedef struct {
bool connected;
bool starting;
bool stopping;
bool has_active_task;
bool server_listening;
bool remote_responding;
bool request_to_speak_pending;
bool client_speech_active;
bool ptt_release_required;
bool local_responding_started;
bool received_downstream_audio;
uint32_t last_heartbeat_ms;
uint32_t last_downstream_ms;
uint32_t last_downstream_audio_ms;
} control_plane_multimodal_runtime_state_t;
typedef enum {
CONTROL_PLANE_TRANSPORT_LOSS_WS_ERROR = 0,
CONTROL_PLANE_TRANSPORT_LOSS_WS_DISCONNECTED,
CONTROL_PLANE_TRANSPORT_LOSS_SEND_TEXT_FAILURE,
CONTROL_PLANE_TRANSPORT_LOSS_SEND_BINARY_FAILURE,
CONTROL_PLANE_TRANSPORT_LOSS_LISTENING_TIMEOUT,
CONTROL_PLANE_TRANSPORT_LOSS_NO_RESPONSE_TIMEOUT,
CONTROL_PLANE_TRANSPORT_LOSS_DOWNSTREAM_STALL,
CONTROL_PLANE_TRANSPORT_LOSS_PROTOCOL_ERROR,
} control_plane_transport_loss_reason_t;
esp_err_t control_plane_multimodal_start(void);
esp_err_t control_plane_multimodal_stop(void);
bool control_plane_multimodal_is_connected(void);
esp_err_t control_plane_multimodal_send_start(void);
esp_err_t control_plane_multimodal_send_stop(const char *reason);
esp_err_t control_plane_multimodal_request_downstream_stall_stop(void);
esp_err_t control_plane_multimodal_send_request_to_speak(void);
esp_err_t control_plane_multimodal_send_speech_start(void);
esp_err_t control_plane_multimodal_send_speech_stop(void);
esp_err_t control_plane_multimodal_send_cancel_speech(void);
esp_err_t control_plane_multimodal_send_request_to_respond(
const control_plane_multimodal_respond_request_t *request);
esp_err_t control_plane_request_cloud_chouqian(void);
esp_err_t control_plane_multimodal_send_update_info(
const control_plane_multimodal_update_info_request_t *request);
esp_err_t control_plane_multimodal_send_local_responding_started(void);
esp_err_t control_plane_multimodal_send_local_responding_ended(void);
esp_err_t control_plane_multimodal_send_heartbeat(void);
esp_err_t control_plane_multimodal_send_audio_realtime(const uint8_t *data, size_t len);
void control_plane_multimodal_mark_transport_unavailable(
control_plane_transport_loss_reason_t reason,
const char *caller);
esp_err_t control_plane_multimodal_get_runtime_state(
control_plane_multimodal_runtime_state_t *out_state);
esp_err_t control_plane_multimodal_register_event_callback(
control_plane_multimodal_event_callback_t cb,
void *user_data);
esp_err_t control_plane_multimodal_register_downstream_audio_callback(
control_plane_multimodal_downstream_audio_callback_t cb,
void *user_data);
esp_err_t control_plane_multimodal_register_responding_text_callback(
control_plane_multimodal_responding_text_callback_t cb,
void *user_data);
esp_err_t control_plane_multimodal_start_capture(void);
esp_err_t control_plane_multimodal_start_uplink(void);
void control_plane_multimodal_stop_uplink(void);
bool control_plane_multimodal_is_uplink_running(void);
esp_err_t control_plane_multimodal_play_downstream_audio(const uint8_t *data, size_t len);
esp_err_t control_plane_multimodal_handle_ptt_press_step(void);
esp_err_t control_plane_multimodal_handle_ptt_release(void);
void control_plane_multimodal_reset_uplink_failure_window(void);
esp_err_t control_plane_local_wake_start(void);
esp_err_t control_plane_local_wake_stop(int timeout_ms);
void control_plane_local_wake_pause_for_foreground(void);
void control_plane_local_wake_resume_after_foreground(void);
bool control_plane_local_wake_prompt_gate_active(void);
void control_plane_ptt_reset_capture_window(void);
bool control_plane_voice_turn_is_idle(void);
esp_err_t control_plane_start_system_event_task(void);
void control_plane_stop_system_event_task(void);
esp_err_t control_plane_post_system_event(const control_plane_system_event_t *event, uint32_t timeout_ms);
int64_t control_plane_event_timestamp_ms(void);
const char *control_plane_key_id_name(control_plane_key_id_t key_id);
const char *control_plane_key_event_type_name(control_plane_key_event_type_t type);
esp_err_t control_plane_start_key_input_task(void);
void control_plane_stop_key_input_task(void);
esp_err_t control_plane_schedule_websocket_wait_task(void);
void control_plane_set_ptt_active(bool active);
esp_err_t control_plane_enter_mini_program_reprovisioning(const char *reason);
esp_err_t control_plane_remote_log_start(void);
void control_plane_remote_log_stop(int timeout_ms);
esp_err_t control_plane_remote_log_flush_now(void);
esp_err_t control_plane_remote_log_clear(void);
esp_err_t control_plane_remote_log_format_status(char *out, size_t out_len);