Merge branch 'bugfix/c++' into 'master'

Bugfix/c++

See merge request speech-recognition-framework/esp-sr!120

CHANGELOG.md

@@ -4,6 +4,10 @@
 - Available storage is less than the remaining flash space on IDF v5.0.   
 If you can not map model partition successfully, please check the left free storage by `spi_flash_mmap_get_free_pages(ESP_PARTITION_MMAP_DATA)` or update IDF to v5.1 or later.
 
+## 1.9.4
+- Fix bugs when using C++ to compile
+- Add more new wake words
+
 ## 1.9.3
 - Fix nsnet2 crash
 - Add esp32p4 ci test

Kconfig.projbuild

@@ -257,12 +257,12 @@ menu "Load Multiple Wake Words"
 
 
     config SR_WN_WN9_HEYWANDA_TTS_MULTI
-        bool "Hey,Wanda (wn9_heywanda_tts)"
+    bool "Hey,Wanda (wn9_heywanda_tts)"
-        depends on IDF_TARGET_ESP32S3 || IDF_TARGET_ESP32P4
+    depends on IDF_TARGET_ESP32S3 || IDF_TARGET_ESP32P4
 
     config SR_WN_WN9_HIMIAOMIAO_TTS_MULTI
-        bool "Hi,喵喵 (wn9_himiaomiao_tts)"
+    bool "Hi,喵喵 (wn9_himiaomiao_tts)"
-        depends on IDF_TARGET_ESP32S3 || IDF_TARGET_ESP32P4
+    depends on IDF_TARGET_ESP32S3 || IDF_TARGET_ESP32P4
 
 
     config SR_WN_WN9_MYCROFT_TTS_MULTI

esp-tts/esp_tts_chinese/include/esp_tts.h

@@ -40,7 +40,7 @@ typedef void * esp_tts_handle_t;
  *         - NULL: Init failed
  *         - Others: The instance of voice set
  */
-esp_tts_voice_t *esp_tts_voice_set_init(const esp_tts_voice_t *template, void *data);
+esp_tts_voice_t *esp_tts_voice_set_init(const esp_tts_voice_t *voice_template, void *data);
 
 /**
  * @brief Init an instance of the TTS voice set structure.

esp-tts/esp_tts_chinese/include/esp_tts_parser.h

@@ -4,7 +4,9 @@
 #include "stdlib.h"
 #include "esp_tts_voice.h"
 
-
+#ifdef __cplusplus
+extern "C" {
+#endif
 typedef struct {
 	int *syll_idx;
 	int syll_num;
@@ -22,4 +24,8 @@ esp_tts_utt_t* esp_tts_utt_alloc(int syll_num, esp_tts_voice_t *voice);
 
 void esp_tts_utt_free(esp_tts_utt_t *utt);
 
+#ifdef __cplusplus
+}
+#endif
+
 #endif

esp-tts/esp_tts_chinese/include/esp_tts_voice.h

@@ -1,6 +1,10 @@
 #ifndef _ESP_TTS_VOICE_H_
 #define _ESP_TTS_VOICE_H_
 
+#ifdef __cplusplus
+extern "C" {
+#endif
+
 typedef struct {
 	char *voice_name;       // voice set name 
 	char *format;           // the format of voice data, currently support pcm and amrwb
@@ -16,5 +20,7 @@ typedef struct {
 	unsigned char *data;    // the audio data of all syllables
 } esp_tts_voice_t;
 
-
+#ifdef __cplusplus
+}
+#endif
 #endif

idf_component.yml

@@ -1,4 +1,4 @@
-version: "1.9.3"
+version: "1.9.4"
 description: esp_sr provides basic algorithms for Speech Recognition applications
 url: https://github.com/espressif/esp-sr
 dependencies:

include/esp32/esp_afe_config.h

@@ -113,10 +113,12 @@ typedef struct {
     .memory_alloc_mode = AFE_MEMORY_ALLOC_INTERNAL_PSRAM_BALANCE, \
     .afe_linear_gain = 1.0, \
     .agc_mode = AFE_MN_PEAK_AGC_MODE_2, \
-    .pcm_config.total_ch_num = 2, \
+    .pcm_config = { \
-    .pcm_config.mic_num = 1, \
+        .total_ch_num = 2, \
-    .pcm_config.ref_num = 1, \
+        .mic_num = 1, \
-    .pcm_config.sample_rate = 16000, \
+        .ref_num = 1, \
+        .sample_rate = 16000, \
+    }, \
     .debug_init = false, \
     .debug_hook = {{AFE_DEBUG_HOOK_MASE_TASK_IN, NULL}, {AFE_DEBUG_HOOK_FETCH_TASK_IN, NULL}}, \
     .afe_ns_mode = NS_MODE_SSP, \
@@ -142,10 +144,12 @@ typedef struct {
     .memory_alloc_mode = AFE_MEMORY_ALLOC_MORE_PSRAM, \
     .afe_linear_gain = 1.0, \
     .agc_mode = AFE_MN_PEAK_AGC_MODE_2, \
-    .pcm_config.total_ch_num = 3, \
+    .pcm_config = { \
-    .pcm_config.mic_num = 2, \
+        .total_ch_num = 3, \
-    .pcm_config.ref_num = 1, \
+        .mic_num = 2, \
-    .pcm_config.sample_rate = 16000, \
+        .ref_num = 1, \
+        .sample_rate = 16000, \
+    }, \
     .debug_init = false, \
     .debug_hook = {{AFE_DEBUG_HOOK_MASE_TASK_IN, NULL}, {AFE_DEBUG_HOOK_FETCH_TASK_IN, NULL}}, \
     .afe_ns_mode = NS_MODE_SSP, \
@@ -155,4 +159,4 @@ typedef struct {
 
 #ifdef __cplusplus
 }
-#endif
+#endif

include/esp32/esp_mn_iface.h

@@ -198,7 +198,7 @@ typedef void (*esp_mn_iface_op_print_active_speech_commands)(model_iface_data_t
  * @param model_data     The model object to query
  * @param str            The input string
 */
-typedef int (*esp_mn_iface_op_check_speech_command)(model_iface_data_t *model_data, char *str);
+typedef int (*esp_mn_iface_op_check_speech_command)(model_iface_data_t *model_data, const char *str);
 
 typedef struct {
     esp_mn_iface_op_create_t create;

include/esp32/flite_g2p.h

@@ -15,6 +15,6 @@ void flite_g2p_result_print_string(flite_g2p_result *result, int map_phonemes);
 
 char *flite_g2p_result_get_string(flite_g2p_result *result, int map_phonemes);
 
-char *flite_g2p(char *graphemes, int map_phonemes);
+char *flite_g2p(const char *graphemes, int map_phonemes);
 
 #endif

include/esp32p4/esp_afe_config.h

@@ -115,10 +115,12 @@ typedef struct {
     .memory_alloc_mode = AFE_MEMORY_ALLOC_INTERNAL_PSRAM_BALANCE, \
     .afe_linear_gain = 1.0, \
     .agc_mode = AFE_MN_PEAK_AGC_MODE_2, \
-    .pcm_config.total_ch_num = 2, \
+    .pcm_config = { \
-    .pcm_config.mic_num = 1, \
+        .total_ch_num = 2, \
-    .pcm_config.ref_num = 1, \
+        .mic_num = 1, \
-    .pcm_config.sample_rate = 16000, \
+        .ref_num = 1, \
+        .sample_rate = 16000, \
+    }, \
     .debug_init = false, \
     .debug_hook = {{AFE_DEBUG_HOOK_MASE_TASK_IN, NULL}, {AFE_DEBUG_HOOK_FETCH_TASK_IN, NULL}}, \
     .afe_ns_mode = NS_MODE_SSP, \
@@ -145,10 +147,12 @@ typedef struct {
     .memory_alloc_mode = AFE_MEMORY_ALLOC_MORE_PSRAM, \
     .afe_linear_gain = 1.0, \
     .agc_mode = AFE_MN_PEAK_AGC_MODE_2, \
-    .pcm_config.total_ch_num = 2, \
+    .pcm_config = { \
-    .pcm_config.mic_num = 1, \
+        .total_ch_num = 2, \
-    .pcm_config.ref_num = 1, \
+        .mic_num = 1, \
-    .pcm_config.sample_rate = 16000, \
+        .ref_num = 1, \
+        .sample_rate = 16000, \
+    }, \
     .debug_init = false, \
     .debug_hook = {{AFE_DEBUG_HOOK_MASE_TASK_IN, NULL}, {AFE_DEBUG_HOOK_FETCH_TASK_IN, NULL}}, \
     .afe_ns_mode = NS_MODE_SSP, \
@@ -175,10 +179,12 @@ typedef struct {
     .memory_alloc_mode = AFE_MEMORY_ALLOC_MORE_PSRAM, \
     .afe_linear_gain = 1.0, \
     .agc_mode = AFE_MN_PEAK_AGC_MODE_2, \
-    .pcm_config.total_ch_num = 3, \
+    .pcm_config = { \
-    .pcm_config.mic_num = 2, \
+        .total_ch_num = 3, \
-    .pcm_config.ref_num = 1, \
+        .mic_num = 2, \
-    .pcm_config.sample_rate = 16000, \
+        .ref_num = 1, \
+        .sample_rate = 16000, \
+    }, \
     .debug_init = false, \
     .debug_hook = {{AFE_DEBUG_HOOK_MASE_TASK_IN, NULL}, {AFE_DEBUG_HOOK_FETCH_TASK_IN, NULL}}, \
     .afe_ns_mode = NS_MODE_SSP, \

include/esp32p4/esp_afe_sr_iface.h

@@ -112,7 +112,7 @@ typedef afe_fetch_result_t* (*esp_afe_sr_iface_op_fetch_t)(esp_afe_sr_data_t *af
  * @brief reset ringbuf of AFE.
  *
  * @param afe          The AFE_SR object to query
- * @return             -1: fail, 0: success
+ * @return             -1: fail, 1: success
  */
 typedef int (*esp_afe_sr_iface_op_reset_buffer_t)(esp_afe_sr_data_t *afe);
 
@@ -122,7 +122,7 @@ typedef int (*esp_afe_sr_iface_op_reset_buffer_t)(esp_afe_sr_data_t *afe);
  *
  * @param afe                The AFE_SR object to query
  * @param wakenet_word       The wakenet word, should be DEFAULT_WAKE_WORD or EXTRA_WAKE_WORD
- * @return             0: fail, 1: success
+ * @return             -1: fail, 1: success
  */
 typedef int (*esp_afe_sr_iface_op_set_wakenet_t)(esp_afe_sr_data_t *afe, char* model_name);
 
@@ -130,7 +130,7 @@ typedef int (*esp_afe_sr_iface_op_set_wakenet_t)(esp_afe_sr_data_t *afe, char* m
  * @brief Disable wakenet model.
  *
  * @param afe          The AFE_SR object to query
- * @return             0: fail, 1: success
+ * @return             -1: fail, 0: disabled, 1: enabled
  */
 typedef int (*esp_afe_sr_iface_op_disable_wakenet_t)(esp_afe_sr_data_t *afe);
 
@@ -138,7 +138,7 @@ typedef int (*esp_afe_sr_iface_op_disable_wakenet_t)(esp_afe_sr_data_t *afe);
  * @brief Enable wakenet model.
  *
  * @param afe          The AFE_SR object to query
- * @return             0: fail, 1: success
+ * @return             -1: fail, 0: disabled, 1: enabled
  */
 typedef int (*esp_afe_sr_iface_op_enable_wakenet_t)(esp_afe_sr_data_t *afe);
 
@@ -146,7 +146,7 @@ typedef int (*esp_afe_sr_iface_op_enable_wakenet_t)(esp_afe_sr_data_t *afe);
  * @brief Disable AEC algorithm.
  *
  * @param afe          The AFE_SR object to query
- * @return             0: fail, 1: success
+ * @return             -1: fail, 0: disabled, 1: enabled
  */
 typedef int (*esp_afe_sr_iface_op_disable_aec_t)(esp_afe_sr_data_t *afe);
 
@@ -154,7 +154,7 @@ typedef int (*esp_afe_sr_iface_op_disable_aec_t)(esp_afe_sr_data_t *afe);
  * @brief Enable AEC algorithm.
  *
  * @param afe          The AFE_SR object to query
- * @return             0: fail, 1: success
+ * @return             -1: fail, 0: disabled, 1: enabled
  */
 typedef int (*esp_afe_sr_iface_op_enable_aec_t)(esp_afe_sr_data_t *afe);
 
@@ -162,7 +162,7 @@ typedef int (*esp_afe_sr_iface_op_enable_aec_t)(esp_afe_sr_data_t *afe);
  * @brief Disable SE algorithm.
  *
  * @param afe          The AFE_SR object to query
- * @return             0: fail, 1: success
+ * @return             -1: fail, 0: disabled, 1: enabled
  */
 typedef int (*esp_afe_sr_iface_op_disable_se_t)(esp_afe_sr_data_t *afe);
 
@@ -170,7 +170,7 @@ typedef int (*esp_afe_sr_iface_op_disable_se_t)(esp_afe_sr_data_t *afe);
  * @brief Enable SE algorithm.
  *
  * @param afe          The AFE_SR object to query
- * @return             0: fail, 1: success
+ * @return             -1: fail, 0: disabled, 1: enabled
  */
 typedef int (*esp_afe_sr_iface_op_enable_se_t)(esp_afe_sr_data_t *afe);
 

include/esp32p4/esp_mn_iface.h

@@ -198,7 +198,7 @@ typedef void (*esp_mn_iface_op_print_active_speech_commands)(model_iface_data_t
  * @param model_data     The model object to query
  * @param str            The input string
 */
-typedef int (*esp_mn_iface_op_check_speech_command)(model_iface_data_t *model_data, char *str);
+typedef int (*esp_mn_iface_op_check_speech_command)(model_iface_data_t *model_data, const char *str);
 
 typedef struct {
     esp_mn_iface_op_create_t create;

include/esp32p4/flite_g2p.h

@@ -9,12 +9,12 @@ typedef struct {
 
 void flite_g2p_result_free(flite_g2p_result *result);
 
-flite_g2p_result *flite_g2p_get_result(char *grapheme);
+flite_g2p_result *flite_g2p_get_result(const char *grapheme);
 
 void flite_g2p_result_print_string(flite_g2p_result *result, int map_phonemes);
 
 char *flite_g2p_result_get_string(flite_g2p_result *result, int map_phonemes);
 
-char *flite_g2p(char *graphemes, int map_phonemes);
+char *flite_g2p(const char *graphemes, int map_phonemes);
 
 #endif

include/esp32s3/esp_afe_config.h

@@ -115,10 +115,12 @@ typedef struct {
     .memory_alloc_mode = AFE_MEMORY_ALLOC_INTERNAL_PSRAM_BALANCE, \
     .afe_linear_gain = 1.0, \
     .agc_mode = AFE_MN_PEAK_AGC_MODE_2, \
-    .pcm_config.total_ch_num = 2, \
+    .pcm_config = { \
-    .pcm_config.mic_num = 1, \
+        .total_ch_num = 2, \
-    .pcm_config.ref_num = 1, \
+        .mic_num = 1, \
-    .pcm_config.sample_rate = 16000, \
+        .ref_num = 1, \
+        .sample_rate = 16000, \
+    }, \
     .debug_init = false, \
     .debug_hook = {{AFE_DEBUG_HOOK_MASE_TASK_IN, NULL}, {AFE_DEBUG_HOOK_FETCH_TASK_IN, NULL}}, \
     .afe_ns_mode = NS_MODE_SSP, \
@@ -145,10 +147,12 @@ typedef struct {
     .memory_alloc_mode = AFE_MEMORY_ALLOC_MORE_PSRAM, \
     .afe_linear_gain = 1.0, \
     .agc_mode = AFE_MN_PEAK_AGC_MODE_2, \
-    .pcm_config.total_ch_num = 2, \
+    .pcm_config = { \
-    .pcm_config.mic_num = 1, \
+        .total_ch_num = 2, \
-    .pcm_config.ref_num = 1, \
+        .mic_num = 1, \
-    .pcm_config.sample_rate = 16000, \
+        .ref_num = 1, \
+        .sample_rate = 16000, \
+    }, \
     .debug_init = false, \
     .debug_hook = {{AFE_DEBUG_HOOK_MASE_TASK_IN, NULL}, {AFE_DEBUG_HOOK_FETCH_TASK_IN, NULL}}, \
     .afe_ns_mode = NS_MODE_SSP, \
@@ -175,10 +179,12 @@ typedef struct {
     .memory_alloc_mode = AFE_MEMORY_ALLOC_MORE_PSRAM, \
     .afe_linear_gain = 1.0, \
     .agc_mode = AFE_MN_PEAK_AGC_MODE_2, \
-    .pcm_config.total_ch_num = 3, \
+    .pcm_config = { \
-    .pcm_config.mic_num = 2, \
+        .total_ch_num = 3, \
-    .pcm_config.ref_num = 1, \
+        .mic_num = 2, \
-    .pcm_config.sample_rate = 16000, \
+        .ref_num = 1, \
+        .sample_rate = 16000, \
+    }, \
     .debug_init = false, \
     .debug_hook = {{AFE_DEBUG_HOOK_MASE_TASK_IN, NULL}, {AFE_DEBUG_HOOK_FETCH_TASK_IN, NULL}}, \
     .afe_ns_mode = NS_MODE_SSP, \

include/esp32s3/esp_afe_sr_iface.h

@@ -112,7 +112,7 @@ typedef afe_fetch_result_t* (*esp_afe_sr_iface_op_fetch_t)(esp_afe_sr_data_t *af
  * @brief reset ringbuf of AFE.
  *
  * @param afe          The AFE_SR object to query
- * @return             -1: fail, 0: success
+ * @return             -1: fail, 1: success
  */
 typedef int (*esp_afe_sr_iface_op_reset_buffer_t)(esp_afe_sr_data_t *afe);
 
@@ -122,7 +122,7 @@ typedef int (*esp_afe_sr_iface_op_reset_buffer_t)(esp_afe_sr_data_t *afe);
  *
  * @param afe                The AFE_SR object to query
  * @param wakenet_word       The wakenet word, should be DEFAULT_WAKE_WORD or EXTRA_WAKE_WORD
- * @return             0: fail, 1: success
+ * @return             -1: fail, 1: success
  */
 typedef int (*esp_afe_sr_iface_op_set_wakenet_t)(esp_afe_sr_data_t *afe, char* model_name);
 
@@ -130,7 +130,7 @@ typedef int (*esp_afe_sr_iface_op_set_wakenet_t)(esp_afe_sr_data_t *afe, char* m
  * @brief Disable wakenet model.
  *
  * @param afe          The AFE_SR object to query
- * @return             0: fail, 1: success
+ * @return             -1: fail, 0: disabled, 1: enabled
  */
 typedef int (*esp_afe_sr_iface_op_disable_wakenet_t)(esp_afe_sr_data_t *afe);
 
@@ -138,7 +138,7 @@ typedef int (*esp_afe_sr_iface_op_disable_wakenet_t)(esp_afe_sr_data_t *afe);
  * @brief Enable wakenet model.
  *
  * @param afe          The AFE_SR object to query
- * @return             0: fail, 1: success
+ * @return             -1: fail, 0: disabled, 1: enabled
  */
 typedef int (*esp_afe_sr_iface_op_enable_wakenet_t)(esp_afe_sr_data_t *afe);
 
@@ -146,7 +146,7 @@ typedef int (*esp_afe_sr_iface_op_enable_wakenet_t)(esp_afe_sr_data_t *afe);
  * @brief Disable AEC algorithm.
  *
  * @param afe          The AFE_SR object to query
- * @return             0: fail, 1: success
+ * @return             -1: fail, 0: disabled, 1: enabled
  */
 typedef int (*esp_afe_sr_iface_op_disable_aec_t)(esp_afe_sr_data_t *afe);
 
@@ -154,7 +154,7 @@ typedef int (*esp_afe_sr_iface_op_disable_aec_t)(esp_afe_sr_data_t *afe);
  * @brief Enable AEC algorithm.
  *
  * @param afe          The AFE_SR object to query
- * @return             0: fail, 1: success
+ * @return             -1: fail, 0: disabled, 1: enabled
  */
 typedef int (*esp_afe_sr_iface_op_enable_aec_t)(esp_afe_sr_data_t *afe);
 
@@ -162,7 +162,7 @@ typedef int (*esp_afe_sr_iface_op_enable_aec_t)(esp_afe_sr_data_t *afe);
  * @brief Disable SE algorithm.
  *
  * @param afe          The AFE_SR object to query
- * @return             0: fail, 1: success
+ * @return             -1: fail, 0: disabled, 1: enabled
  */
 typedef int (*esp_afe_sr_iface_op_disable_se_t)(esp_afe_sr_data_t *afe);
 
@@ -170,7 +170,7 @@ typedef int (*esp_afe_sr_iface_op_disable_se_t)(esp_afe_sr_data_t *afe);
  * @brief Enable SE algorithm.
  *
  * @param afe          The AFE_SR object to query
- * @return             0: fail, 1: success
+ * @return             -1: fail, 0: disabled, 1: enabled
  */
 typedef int (*esp_afe_sr_iface_op_enable_se_t)(esp_afe_sr_data_t *afe);
 

include/esp32s3/esp_mn_iface.h

@@ -198,7 +198,7 @@ typedef void (*esp_mn_iface_op_print_active_speech_commands)(model_iface_data_t
  * @param model_data     The model object to query
  * @param str            The input string
 */
-typedef int (*esp_mn_iface_op_check_speech_command)(model_iface_data_t *model_data, char *str);
+typedef int (*esp_mn_iface_op_check_speech_command)(model_iface_data_t *model_data, const char *str);
 
 typedef struct {
     esp_mn_iface_op_create_t create;

include/esp32s3/flite_g2p.h

@@ -9,12 +9,12 @@ typedef struct {
 
 void flite_g2p_result_free(flite_g2p_result *result);
 
-flite_g2p_result *flite_g2p_get_result(char *grapheme);
+flite_g2p_result *flite_g2p_get_result(const char *grapheme);
 
 void flite_g2p_result_print_string(flite_g2p_result *result, int map_phonemes);
 
 char *flite_g2p_result_get_string(flite_g2p_result *result, int map_phonemes);
 
-char *flite_g2p(char *graphemes, int map_phonemes);
+char *flite_g2p(const char *graphemes, int map_phonemes);
 
 #endif

src/esp_mn_speech_commands.c

@@ -79,7 +79,7 @@ esp_err_t esp_mn_commands_clear(void)
     return ESP_OK;
 }
 
-esp_mn_node_t *esp_mn_command_search(char *string)
+esp_mn_node_t *esp_mn_command_search(const char *string)
 {
     esp_mn_node_t *temp = esp_mn_root;
     if(NULL == esp_mn_root) {
@@ -95,7 +95,7 @@ esp_mn_node_t *esp_mn_command_search(char *string)
     return NULL;
 }
 
-esp_err_t esp_mn_commands_add(int command_id, char *string)
+esp_err_t esp_mn_commands_add(int command_id, const char *string)
 {
     if (NULL == esp_mn_root || esp_mn_model_handle == NULL || esp_mn_model_data == NULL) {
         ESP_LOGE(TAG, "Please create mn model first.\n");
@@ -156,7 +156,7 @@ esp_err_t esp_mn_commands_add(int command_id, char *string)
     return ESP_OK;
 }
 
-esp_err_t esp_mn_commands_phoneme_add(int command_id, char *string, char *phonemes)
+esp_err_t esp_mn_commands_phoneme_add(int command_id, const char *string, const char *phonemes)
 {
     if (NULL == esp_mn_root || esp_mn_model_handle == NULL || esp_mn_model_data == NULL) {
         ESP_LOGE(TAG, "Please create mn model first.\n");
@@ -214,7 +214,7 @@ esp_err_t esp_mn_commands_phoneme_add(int command_id, char *string, char *phonem
     return ESP_OK;
 }
 
-esp_err_t esp_mn_commands_modify(char *old_string, char *new_string)
+esp_err_t esp_mn_commands_modify(const char *old_string, const char *new_string)
 {
 #ifdef CONFIG_SR_MN_EN_MULTINET7_QUANT
     char *phonemes = flite_g2p(new_string, 1);
@@ -257,7 +257,7 @@ esp_err_t esp_mn_commands_modify(char *old_string, char *new_string)
     return ESP_OK;
 }
 
-esp_err_t esp_mn_commands_remove(char *string)
+esp_err_t esp_mn_commands_remove(const char *string)
 {
     esp_mn_node_t *temp = esp_mn_root;
     ESP_RETURN_ON_FALSE(NULL != esp_mn_root, ESP_ERR_INVALID_STATE, TAG, "The mn commands is not initialized");
@@ -374,7 +374,7 @@ void esp_mn_active_commands_print(void)
     ESP_LOGI(TAG, "---------------------------------------------------------\n");
 }
 
-esp_mn_phrase_t *esp_mn_phrase_alloc(int command_id, char *string)
+esp_mn_phrase_t *esp_mn_phrase_alloc(int command_id, const char *string)
 {
 
     int string_len = strlen(string);

src/include/esp_mn_speech_commands.h

@@ -57,7 +57,7 @@ esp_err_t esp_mn_commands_free(void);
  *     - ESP_OK                  Success
  *     - ESP_ERR_INVALID_STATE   Fail
  */
-esp_err_t esp_mn_commands_add(int command_id, char *string);
+esp_err_t esp_mn_commands_add(int command_id, const char *string);
 
 /**
  * @brief Add one speech commands with command string, command phonemes and command ID
@@ -71,7 +71,7 @@ esp_err_t esp_mn_commands_add(int command_id, char *string);
  *     - ESP_OK                  Success
  *     - ESP_ERR_INVALID_STATE   Fail
  */
-esp_err_t esp_mn_commands_phoneme_add(int command_id, char *string, char *phonemes);
+esp_err_t esp_mn_commands_phoneme_add(int command_id, const char *string, const char *phonemes);
 
 /**
  * @brief Modify one speech commands with new command string
@@ -83,7 +83,7 @@ esp_err_t esp_mn_commands_phoneme_add(int command_id, char *string, char *phonem
  *     - ESP_OK                  Success
  *     - ESP_ERR_INVALID_STATE   Fail
  */
-esp_err_t esp_mn_commands_modify(char *old_string, char *new_string);
+esp_err_t esp_mn_commands_modify(const char *old_string, const char *new_string);
 
 /**
  * @brief Remove one speech commands by command string
@@ -94,7 +94,7 @@ esp_err_t esp_mn_commands_modify(char *old_string, char *new_string);
  *     - ESP_OK                  Success
  *     - ESP_ERR_INVALID_STATE   Fail
  */
-esp_err_t esp_mn_commands_remove(char *string);
+esp_err_t esp_mn_commands_remove(const char *string);
 
 /**
  * @brief Clear all speech commands in linked list
@@ -157,7 +157,7 @@ esp_mn_error_t *esp_mn_commands_update();
  *
  * @return the pointer of esp_mn_phrase_t
  */
-esp_mn_phrase_t *esp_mn_phrase_alloc(int command_id, char *string);
+esp_mn_phrase_t *esp_mn_phrase_alloc(int command_id, const char *string);
 
 /**
  * @brief Free esp_mn_phrase_t pointer.

test_apps/esp-sr/main/CMakeLists.txt

@@ -1,9 +1,9 @@
 
 set(srcs
-    "test_app_main.c"
+    "app_main.cpp"
-    "test_wakenet.c"
+    "test_wakenet.cpp"
-    "test_multinet.c"
+    "test_multinet.cpp"
-    "test_afe.c"
+    "test_afe.cpp"
 )
 
 idf_component_register(SRCS ${srcs}

test_apps/esp-sr/main/app_main.cpp

@@ -7,12 +7,11 @@
    CONDITIONS OF ANY KIND, either express or implied.
 */
 
+#include "unity.h"
 #include <stdio.h>
 #include <string.h>
-#include "unity.h"
 
-
+extern "C" void app_main(void)
-void app_main(void)
 {
     /* These are the different ways of running registered tests.
      * In practice, only one of them is usually needed.
@@ -45,4 +44,4 @@ void app_main(void)
      * Make sure that task watchdog is disabled if you use this function.
      */
     unity_run_menu();
-}
+}

test_apps/esp-sr/main/test_afe.cpp

@@ -112,7 +112,7 @@ TEST_CASE(">>>>>>>> audio_front_end SR create/destroy API & memory leak <<<<<<<<
                         afe_data = afe_handle->create_from_config(&afe_config);
 
                         audio_chunksize = afe_handle->get_feed_chunksize(afe_data);
-                        feed_buff = malloc(audio_chunksize * sizeof(int16_t) * afe_config.pcm_config.total_ch_num);
+                        feed_buff = (int16_t *) malloc(audio_chunksize * sizeof(int16_t) * afe_config.pcm_config.total_ch_num);
                         assert(feed_buff);
 
                         afe_handle->feed(afe_data, feed_buff);
@@ -145,7 +145,7 @@ void test_feed_Task(void *arg)
     esp_afe_sr_iface_t *afe_handle = (esp_afe_sr_iface_t *)arg;
     int feed_chunksize = afe_handle->get_feed_chunksize(afe_data);
     int total_nch = afe_handle->get_total_channel_num(afe_data);
-    int16_t *i2s_buff = malloc(feed_chunksize * sizeof(int16_t) * total_nch);
+    int16_t *i2s_buff = (int16_t *) malloc(feed_chunksize * sizeof(int16_t) * total_nch);
     assert(i2s_buff);
     ESP_LOGI(TAG, "feed task start\n");
     // FILE *fp = fopen("/sdcard/out", "w");
@@ -169,7 +169,7 @@ void test_detect_Task(void *arg)
     // esp_afe_sr_iface_t *afe_handle = &ESP_AFE_SR_HANDLE;
     esp_afe_sr_iface_t *afe_handle = (esp_afe_sr_iface_t *)arg;
     int fetch_chunksize = afe_handle->get_fetch_chunksize(afe_data);
-    int16_t *buff = malloc(fetch_chunksize * sizeof(int16_t));
+    int16_t *buff = (int16_t *) malloc(fetch_chunksize * sizeof(int16_t));
     assert(buff);
     ESP_LOGI(TAG, "------------detect start------------\n");
 
@@ -208,21 +208,29 @@ esp_err_t audio_sys_get_real_time_stats(void)
 
     // Allocate array to store current task states
     start_array_size = uxTaskGetNumberOfTasks() + ARRAY_SIZE_OFFSET;
-    start_array = malloc(sizeof(TaskStatus_t) * start_array_size);
+    start_array = (TaskStatus_t*) malloc(sizeof(TaskStatus_t) * start_array_size);
     assert(start_array);
     // Get current task states
     start_array_size = uxTaskGetSystemState(start_array, start_array_size, &start_run_time);
     if (start_array_size == 0) {
         ESP_LOGE(TAG, "Insufficient array size for uxTaskGetSystemState. Trying increasing ARRAY_SIZE_OFFSET");
         ret = ESP_FAIL;
-        goto exit;
+        if (start_array) {
+            free(start_array);
+            start_array = NULL;
+        }
+        if (end_array) {
+            free(end_array);
+            end_array = NULL;
+        }
+        return ret;
     }
 
     vTaskDelay(pdMS_TO_TICKS(AUDIO_SYS_TASKS_ELAPSED_TIME_MS));
 
     // Allocate array to store tasks states post delay
     end_array_size = uxTaskGetNumberOfTasks() + ARRAY_SIZE_OFFSET;
-    end_array = malloc(sizeof(TaskStatus_t) * end_array_size);
+    end_array = (TaskStatus_t*) malloc(sizeof(TaskStatus_t) * end_array_size);
     assert(end_array);
 
     // Get post delay task states
@@ -230,7 +238,15 @@ esp_err_t audio_sys_get_real_time_stats(void)
     if (end_array_size == 0) {
         ESP_LOGE(TAG, "Insufficient array size for uxTaskGetSystemState. Trying increasing ARRAY_SIZE_OFFSET");
         ret = ESP_FAIL;
-        goto exit;
+        if (start_array) {
+            free(start_array);
+            start_array = NULL;
+        }
+        if (end_array) {
+            free(end_array);
+            end_array = NULL;
+        }
+        return ret;
     }
 
     // Calculate total_elapsed_time in units of run time stats clock period.
@@ -238,7 +254,15 @@ esp_err_t audio_sys_get_real_time_stats(void)
     if (total_elapsed_time == 0) {
         ESP_LOGE(TAG, "Delay duration too short. Trying increasing AUDIO_SYS_TASKS_ELAPSED_TIME_MS");
         ret = ESP_FAIL;
-        goto exit;
+        if (start_array) {
+            free(start_array);
+            start_array = NULL;
+        }
+        if (end_array) {
+            free(end_array);
+            end_array = NULL;
+        }
+        return ret;
     }
 
     ESP_LOGI(TAG, "| Task              | Run Time    | Per | Prio | HWM       | State   | CoreId   | Stack ");
@@ -277,15 +301,6 @@ esp_err_t audio_sys_get_real_time_stats(void)
     printf("\n");
     ret = ESP_OK;
 
-exit:    // Common return path
-    if (start_array) {
-        free(start_array);
-        start_array = NULL;
-    }
-    if (end_array) {
-        free(end_array);
-        end_array = NULL;
-    }
     return ret;
 #else
     ESP_LOGW(TAG, "Please enbale `CONFIG_FREERTOS_VTASKLIST_INCLUDE_COREID` and `CONFIG_FREERTOS_GENERATE_RUN_TIME_STATS` in menuconfig");
@@ -386,7 +401,7 @@ TEST_CASE("audio_front_end VC create/destroy API & memory leak", "[afe_vc]")
                             afe_config.se_init = se_init;
                             afe_config.vad_init = vad_init;
                             afe_config.voice_communication_agc_init = voice_communication_agc_init;
-                            afe_config.afe_ns_mode = afe_ns_mode;
+                            afe_config.afe_ns_mode = (afe_ns_mode_t)afe_ns_mode;
 
                             //start_total_mem_size = heap_caps_get_free_size(MALLOC_CAP_8BIT);
                             //start_internal_mem_size = heap_caps_get_free_size(MALLOC_CAP_8BIT | MALLOC_CAP_INTERNAL);
@@ -412,7 +427,7 @@ TEST_CASE("audio_front_end VC create/destroy API & memory leak", "[afe_vc]")
                                 }
 
                                 audio_chunksize = afe_handle->get_feed_chunksize(afe_data);
-                                feed_buff = malloc(audio_chunksize * sizeof(int16_t) * afe_config.pcm_config.total_ch_num);
+                                feed_buff = (int16_t *) malloc(audio_chunksize * sizeof(int16_t) * afe_config.pcm_config.total_ch_num);
                                 assert(feed_buff);
 
                                 afe_handle->feed(afe_data, feed_buff);

test_apps/esp-sr/main/test_multinet.cpp

@@ -88,7 +88,7 @@ TEST_CASE("multinet cpu loading", "[mn]")
         printf("commands: da kai kong tiao, size:%d\n", data_size);
     }
     
-    int16_t *buffer = malloc(audio_chunksize);
+    int16_t *buffer = (int16_t *) malloc(audio_chunksize);
     int chunks = 0;
     struct timeval tv_start, tv_end;
     gettimeofday(&tv_start, NULL);

test_apps/esp-sr/main/test_wakenet.cpp

@@ -65,7 +65,7 @@ TEST_CASE("wakenet create/destroy API & memory leak", "[wn]")
         //     DET_MODE_3CH_90 = 4,
         //     DET_MODE_3CH_95 = 5,
         // } det_mode_t;
-        model_data = wakenet->create(model_name, i);
+        model_data = wakenet->create(model_name, (det_mode_t)i);
 
         printf("destroy ...\n");
         wakenet->destroy(model_data);
@@ -89,7 +89,7 @@ TEST_CASE("wakenet detect API & cpu loading", "[wn]")
     model_iface_data_t *model_data = wakenet->create(model_name, DET_MODE_95);
     int frequency = wakenet->get_samp_rate(model_data);
     int audio_chunksize = wakenet->get_samp_chunksize(model_data) * sizeof(int16_t);
-    int16_t *buffer = malloc(audio_chunksize);
+    int16_t *buffer = (int16_t *) malloc(audio_chunksize);
     int chunks = 0;
     int detected = 0;
     struct timeval tv_start, tv_end;

test_apps/esp-tts/main/CMakeLists.txt

@@ -1,7 +1,7 @@
 
 set(srcs
-    "test_app_main.c"
+    "app_main.cpp"
-    "test_chinese_tts.c"
+    "test_chinese_tts.cpp"
 )
 
 idf_component_register(SRCS ${srcs}

test_apps/esp-tts/main/app_main.cpp

@@ -7,12 +7,11 @@
    CONDITIONS OF ANY KIND, either express or implied.
 */
 
+#include "unity.h"
 #include <stdio.h>
 #include <string.h>
-#include "unity.h"
 
-
+extern "C" void app_main(void)
-void app_main(void)
 {
     /* These are the different ways of running registered tests.
      * In practice, only one of them is usually needed.
@@ -45,4 +44,4 @@ void app_main(void)
      * Make sure that task watchdog is disabled if you use this function.
      */
     unity_run_menu();
-}
+}

test_apps/esp-tts/main/test_chinese_tts.cpp

@@ -0,0 +1,96 @@
+/* test_mean.c: Implementation of a testable component.
+
+   This example code is in the Public Domain (or CC0 licensed, at your option.)
+
+   Unless required by applicable law or agreed to in writing, this
+   software is distributed on an "AS IS" BASIS, WITHOUT WARRANTIES OR
+   CONDITIONS OF ANY KIND, either express or implied.
+*/
+#include <stdio.h>
+#include <stdlib.h>
+#include "string.h"
+#include <limits.h>
+#include "unity.h"
+#include <sys/time.h>
+#include "freertos/FreeRTOS.h"
+#include "freertos/task.h"
+#include "esp_system.h"
+#include "esp_tts.h"
+#include "esp_tts_voice_xiaole.h"
+#include "esp_tts_voice_template.h"
+#include "esp_tts_player.h"
+#include "esp_partition.h"
+#include "esp_idf_version.h"
+
+TEST_CASE("chinese tts create/destroy API & memory leak", "[tts]")
+{
+    int start_size = heap_caps_get_free_size(MALLOC_CAP_8BIT);
+    int start_internal_size = heap_caps_get_free_size(MALLOC_CAP_INTERNAL);
+    /*** 1. create esp tts handle ***/
+    const void* voicedata;
+    const esp_partition_t* partition=esp_partition_find_first(ESP_PARTITION_TYPE_DATA, ESP_PARTITION_SUBTYPE_ANY, "voice_data");
+    assert(partition != NULL);
+#if ESP_IDF_VERSION >= ESP_IDF_VERSION_VAL(5, 0, 0)
+    esp_partition_mmap_handle_t mmap_handle;
+    ESP_ERROR_CHECK(esp_partition_mmap(partition, 0, partition->size, ESP_PARTITION_MMAP_DATA, &voicedata, &mmap_handle));
+#else
+    spi_flash_mmap_handle_t mmap_handle;
+    ESP_ERROR_CHECK(esp_partition_mmap(partition, 0, partition->size, SPI_FLASH_MMAP_DATA, &voicedata, &mmap_handle));
+#endif 
+
+    // test model loading time
+    struct timeval tv_start, tv_end;
+    gettimeofday(&tv_start, NULL);
+    esp_tts_voice_t *voice=esp_tts_voice_set_init(&esp_tts_voice_template, (int16_t*)voicedata); 
+    esp_tts_handle_t *tts_handle=(esp_tts_handle_t *)esp_tts_create(voice);
+    gettimeofday(&tv_end, NULL);
+    int tv_ms = (tv_end.tv_sec - tv_start.tv_sec) * 1000 + (tv_end.tv_usec - tv_start.tv_usec) / 1000;
+    printf("create latency:%d ms\n", tv_ms);
+
+    // test model memory concumption
+    int create_size = start_size - heap_caps_get_free_size(MALLOC_CAP_8BIT);
+    int create_internal_size = start_internal_size - heap_caps_get_free_size(MALLOC_CAP_INTERNAL);
+    printf("Internal RAM: %d, PSRAM:%d\n", create_internal_size, create_size - create_internal_size);
+
+    // test memory leak
+    esp_tts_voice_set_free(voice);
+    esp_tts_destroy(tts_handle);
+#if ESP_IDF_VERSION >= ESP_IDF_VERSION_VAL(5, 0, 0)
+        esp_partition_munmap(mmap_handle); // support esp-idf v5
+#else
+        spi_flash_munmap(mmap_handle);  // support esp-idf v4
+#endif
+    int first_end_size = heap_caps_get_free_size(MALLOC_CAP_8BIT);
+    int last_end_size = first_end_size;
+    int mem_leak = start_size - last_end_size;
+    printf("create&destroy times:%d, memory leak:%d\n", 1, mem_leak);
+
+    for (int i = 0; i < 6; i++) {
+        printf("init partition ...\n");
+    partition=esp_partition_find_first(ESP_PARTITION_TYPE_DATA, ESP_PARTITION_SUBTYPE_ANY, "voice_data");
+    assert(partition != NULL);
+#if ESP_IDF_VERSION >= ESP_IDF_VERSION_VAL(5, 0, 0)
+    ESP_ERROR_CHECK(esp_partition_mmap(partition, 0, partition->size, ESP_PARTITION_MMAP_DATA, &voicedata, &mmap_handle));
+#else
+    ESP_ERROR_CHECK(esp_partition_mmap(partition, 0, partition->size, SPI_FLASH_MMAP_DATA, &voicedata, &mmap_handle));
+#endif 
+
+        printf("create ...\n");
+        voice=esp_tts_voice_set_init(&esp_tts_voice_template, (int16_t*)voicedata); 
+        tts_handle=(esp_tts_handle_t *)esp_tts_create(voice);
+
+        printf("destroy ...\n");
+        esp_tts_voice_set_free(voice);
+        esp_tts_destroy(tts_handle);
+#if ESP_IDF_VERSION >= ESP_IDF_VERSION_VAL(5, 0, 0)
+        esp_partition_munmap(mmap_handle); // support esp-idf v5
+#else
+        spi_flash_munmap(mmap_handle);  // support esp-idf v4
+#endif
+        last_end_size = heap_caps_get_free_size(MALLOC_CAP_8BIT);
+        mem_leak = start_size - last_end_size;
+        printf("create&destroy times:%d, memory leak:%d\n", i + 2, mem_leak);
+    }
+
+    TEST_ASSERT_EQUAL(true, (mem_leak) < 1000 && last_end_size == first_end_size);
+}