release esp-sr v1.9.4

.gitlab-ci.yml

@@ -99,8 +99,6 @@ before_script:
   artifacts:
     when: always
     paths:
-      - "**/build*/size.json"
-      - "**/build*/build_log.txt"
       - "**/build*/*.bin"
       # upload to s3 server to save the artifacts size
       - "**/build*/*.map"

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

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.mic_num = 1, \
-    .pcm_config.ref_num = 1, \
-    .pcm_config.sample_rate = 16000, \
+    .pcm_config = { \
+        .total_ch_num = 2, \
+        .mic_num = 1, \
+        .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.mic_num = 2, \
-    .pcm_config.ref_num = 1, \
-    .pcm_config.sample_rate = 16000, \
+    .pcm_config = { \
+        .total_ch_num = 3, \
+        .mic_num = 2, \
+        .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.mic_num = 1, \
-    .pcm_config.ref_num = 1, \
-    .pcm_config.sample_rate = 16000, \
+    .pcm_config = { \
+        .total_ch_num = 2, \
+        .mic_num = 1, \
+        .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.mic_num = 1, \
-    .pcm_config.ref_num = 1, \
-    .pcm_config.sample_rate = 16000, \
+    .pcm_config = { \
+        .total_ch_num = 2, \
+        .mic_num = 1, \
+        .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.mic_num = 2, \
-    .pcm_config.ref_num = 1, \
-    .pcm_config.sample_rate = 16000, \
+    .pcm_config = { \
+        .total_ch_num = 3, \
+        .mic_num = 2, \
+        .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_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.