Merge pull request #11 from hugobloem/refactor-mqtt-topic-handling

Refactor mqtt topic handling, add convenience script, and siteId separation.

.gitignore

@@ -86,3 +86,4 @@ managed_components
 
 # Secrets file
 secrets.h
+sites.yaml

README.md

@@ -17,6 +17,8 @@ https://user-images.githubusercontent.com/42470993/226731674-cff14709-fd51-44b7-
 To get started please copy secrets_template.h to secrets.h and edit the variables in there. After that you can flash your ESP-BOX using esptool. I recommend using the Visual Studio Code ESP-IDF plugin as it installs all the required programs for you and flashed the device seamlessly. 
 
 ## Managing voice commands
-As of now voice commands can be sending MQTT messages to the `esp-ha/config/add_cmd` topic. As data you should provide a json like this: `{"text": "<your voice command>", "phonetic": "<phonetic voice command"}`. The `text` entry is the command you would like to send to Home Assistant/Rhasspy for recognition. The `phonetic` entry is the phonetic version of it. This phonetic version can be generated using the following python command `python esp-ha\managed_components\espressif__esp-sr\tool\multinet_g2p.py -t <your voice command>`.
+As of now voice commands can be sending MQTT messages to the `esp-ha-speech/config/add_cmd` topic. As data you should provide a json like this: `{"text": "<your voice command>", "phonetic": "<phonetic voice command", "siteId": "<your-siteId>"}`. The `text` entry is the command you would like to send to Home Assistant/Rhasspy for recognition. The `phonetic` entry is the phonetic version of it. This phonetic version can be generated using the following python command `python esp-ha\managed_components\espressif__esp-sr\tool\multinet_g2p.py -t <your voice command>`. `siteId` is used to seperate different esp32s, so you can for example have one in the living room and one in the kitchen and they will only listen to the messages meant for that device.
 
-To delete all existing commands send an MQTT message to `esp-ha/config/rm_all` with payload `{"confirm": "yes"}`. Note that there are now no voice commands in the system, thus trying to invoke the wake word will result in a crash.
+Another option to add commands is to use the convenience script [`configure_sites.py`](./configure_sites.py). To get started create a `sites.yaml` file from the [`sites_template.yaml`](./sites_template.yaml) file. Most options are straightforward but note that under the `sites` tag multiple sites (or satellites) can be configured, each with their own set of devices. The Python script will fetch intent templates from the [Home Assistant intents repo](https://github.com/home-assistant/intents), it will then create some sentences and phonemes for the given entities and send to each site. At the moment this only supports turning on and off entities under the 'lights' tag. 
+
+To delete all existing commands send an MQTT message to `esp-ha-speech/config/rm_all` with payload `{"confirm": "yes", "siteId": "<your-siteId>"}`. Note that there are now no voice commands in the system, thus trying to invoke the wake word will result in a crash.

configure_sites.py

@@ -0,0 +1,114 @@
+'''
+Load sites/sattelites from a file and configure the commands.
+
+The site configuration is loaded from sites.yaml, the intents are loaded from the github.com/home-assistant/intents repo.
+'''
+
+import re
+import time
+
+import requests
+import yaml
+from g2p_en import G2p
+from paho.mqtt import client as mqtt_client
+
+base_repo = 'https://raw.githubusercontent.com/home-assistant/intents/main/sentences/en/'
+intent_scripts = ['homeassistant_HassTurnOff.yaml', 'homeassistant_HassTurnOn.yaml']
+common_script = '_common.yaml'
+
+# Load sites
+with open('sites.yaml', 'r') as f:
+    conf = yaml.safe_load(f)
+    sites = conf['sites']
+
+# Load intents from repo
+intents = {}
+for script in intent_scripts:
+    r = requests.get(base_repo + script)
+    if r.status_code == 200:
+        ryaml = yaml.safe_load(r.text)
+        for intent, data in ryaml['intents'].items():
+            intents[intent] = data['data'][0]['sentences']
+    else:
+        SystemError('Failed to load intent script: ' + script)
+
+
+# Load expansion rules
+r = requests.get(base_repo + common_script)
+if r.status_code == 200:
+    ryaml = yaml.safe_load(r.text)
+    expansions = ryaml['expansion_rules']
+    for k, v in expansions.items():
+        expansions[k] = v[1:-1].split('|') # Remove quotes and split
+
+
+# Construct sentences
+sentences = [
+    'turn on the <name>',
+    'turn off the <name>',
+    'turn on <name>',
+    'turn off <name>',
+    'switch on the <name>',
+    'switch off the <name>',
+    'switch on <name>',
+    'switch off <name>',
+    'activate the <name>',
+    'deactivate the <name>',
+    'activate <name>',
+    'deactivate <name>',
+]
+
+def english_g2p(text_list, alphabet=None):
+    g2p = G2p()
+    outs = []
+    out = ''
+    if alphabet is None:
+        alphabet={"AE1": "a", "N": "N", " ": " ", "OW1": "b", "V": "V", "AH0": "c", "L": "L", "F": "F", "EY1": "d", "S": "S", "B": "B", "R": "R", "AO1": "e", "D": "D", "AH1": "c", "EH1": "f", "OW0": "b", "IH0": "g", "G": "G", "HH": "h", "K": "K", "IH1": "g", "W": "W", "AY1": "i", "T": "T", "M": "M", "Z": "Z", "DH": "j", "ER0": "k", "P": "P", "NG": "l", "IY1": "m", "AA1": "n", "Y": "Y", "UW1": "o", "IY0": "m", "EH2": "f", "CH": "p", "AE0": "a", "JH": "q", "ZH": "r", "AA2": "n", "SH": "s", "AW1": "t", "OY1": "u", "AW2": "t", "IH2": "g", "AE2": "a", "EY2": "d", "ER1": "k", "TH": "v", "UH1": "w", "UW2": "o", "OW2": "b", "AY2": "i", "UW0": "o", "AH2": "c", "EH0": "f", "AW0": "t", "AO2": "e", "AO0": "e", "UH0": "w", "UH2": "w", "AA0": "n", "AY0": "i", "IY2": "m", "EY0": "d", "ER2": "k", "OY2": "u", "OY0": "u"}
+
+    for item in text_list:
+        item = item.split(",")
+        for phrase in item:
+            labels = g2p(phrase)
+            for char in labels:
+                if char not in alphabet:
+                    print("skip %s, not found in alphabet")
+                    continue
+                else:
+                    out += alphabet[char]
+            if phrase != item[-1]:
+                out += ','
+        outs.append(out)
+        out = ''
+
+    return outs
+
+site_sentences = {}
+for siteId, entities in sites.items():
+    site_sentences[siteId] = {'text': []}
+    for entity in entities['lights']: # TODO: Add other entities
+        for sentence in sentences:
+            site_sentences[siteId]['text'].append(sentence.replace('<name>', entity))
+    site_sentences[siteId]['phonetic'] = english_g2p(site_sentences[siteId]['text'])
+
+    assert len(site_sentences[siteId]['text']) == len(site_sentences[siteId]['phonetic'])
+    assert len(site_sentences[siteId]['text']) <= 200
+
+# Connect to MQTT
+def on_connect(client, userdata, flags, rc):
+    if rc == 0:
+        print("Connected to MQTT")
+    else:
+        print("Failed to connect, return code %d\n", rc)
+
+client = mqtt_client.Client()
+client.username_pw_set(conf['mqtt']['username'], conf['mqtt']['password'])
+client.on_connect = on_connect
+client.connect(conf['mqtt']['host'], conf['mqtt']['port'], 60)
+
+# Send intents
+for siteId, data in site_sentences.items():
+    for i, (text, phonetic) in enumerate(zip(data['text'], data['phonetic'])):
+        message = f'{{"text": "{text}", "phonetic": "{phonetic}", "siteId": "{siteId}"}}'
+        client.publish(f'{conf["mqtt"]["topic"]}/add_cmd', message)
+        print(f'Sent {i+1}/{len(data["text"])}: {message}')
+        time.sleep(0.5)

main/app/app_api_mqtt.c

@@ -1,6 +1,8 @@
 #include <string.h>
 #include "esp_system.h"
 #include "esp_event.h"
+#include "esp_check.h"
+#include "esp_err.h"
 
 #include "freertos/FreeRTOS.h"
 #include "freertos/task.h"
@@ -10,6 +12,8 @@
 #include "esp_log.h"
 #include "mqtt_client.h"
 
+#include "cJSON.h"
+
 #include "app_api_mqtt.h"
 #include "app_hass.h"
 #include "ui_net_config.h"
@@ -33,37 +37,53 @@ esp_err_t data_handler(char *topic_, char *data, int topic_len, int data_len)
     char topic[512];
     strncpy(topic, topic_, topic_len);
     topic[topic_len] = '\0';
+    char *subtopic = strtok(topic, "/");
 
     cJSON *jData = cJSON_ParseWithLength(data, data_len);
     if (jData == NULL) {
         ESP_LOGE(TAG, "Error parsing json");
         cJSON_Delete(jData);
         return ESP_FAIL;
     }
+
+    cJSON *siteId = cJSON_GetObjectItemCaseSensitive(jData, "siteId");
+    if (cJSON_IsString(siteId) == 0) {
+        ESP_LOGI(TAG, "No siteId.");
+        cJSON_Delete(jData);
+        return ESP_OK;
+    } else if (strcmp(siteId->valuestring, MQTT_SITE_ID) != 0) {
+        ESP_LOGI(TAG, "siteId does not match.");
+        cJSON_Delete(jData);
+        return ESP_OK;
+    }
 
-    // TODO: change to strtok
-    // topic starts with "hermes/"
-    if (strncmp(topic, "hermes/", 7) == 0) {
-        // Handle hermes messages
-        ESP_LOGI(TAG, "hermes message: %s", jData);
-        // app_hass_handle_hermes(topic, jData);
+    ESP_LOGI(TAG, "subtopic: %s", subtopic);
 
+    // HERMES MESSAGES"
+    if (strcmp(subtopic, "hermes") == 0) {
+        // Handle hermes messages
+        ESP_LOGI(TAG, "hermes message: %s", cJSON_Print(jData));
+
 
-    } else if (strncmp(topic, "esp-ha-speech/", 7) == 0) {
+    // ESP-HA MESSAGES"
+    } else if (strcmp(subtopic, "esp-ha-speech") == 0) {
         // Handle esp-ha messages
-        ESP_LOGI(TAG, "esp-ha message: %s", jData);
+        ESP_LOGI(TAG, "esp-ha-speech message: %s", cJSON_Print(jData));
 
-        if (strncmp(topic, "esp-ha/config/add_cmd", 21) == 0) {
+        subtopic = strtok(NULL, "/");
+        if (strcmp(subtopic, "add_cmd") == 0) {
             // Handle config messages
-            ESP_LOGI(TAG, "adding command");
             app_hass_add_cmd_from_msg(jData);
 
-        } else if (strncmp(topic, "esp-ha/config/rm_all", 24) == 0) {
+        } else if (strcmp(subtopic, "rm_all") == 0) {
             // Handle config messages
-            ESP_LOGI(TAG, "removing all commands");
             app_hass_rm_all_cmd(jData);
+
         }
+    } else {
+        ESP_LOGW(TAG, "Unknown subtopic: %s", subtopic);
     }
+    cJSON_Delete(jData);
     return ESP_OK;
 }
 
@@ -149,7 +169,7 @@ void app_api_mqtt_start(void)
 
     // Subscribe to hermes and configuration topics
     esp_mqtt_client_subscribe(client, "hermes/#", 0);
-    esp_mqtt_client_subscribe(client, "esp-ha/#", 0);
+    esp_mqtt_client_subscribe(client, "esp-ha-speech/#", 0);
 
 }
 

main/app/app_hass.c

@@ -192,15 +192,13 @@ void app_hass_add_cmd_from_msg(cJSON *root) {
     cJSON *sr_phn = cJSON_GetObjectItemCaseSensitive(root, "phonetic");
     if (sr_txt == NULL || sr_phn == NULL) {
         ESP_LOGE(TAG, "Error parsing text");
-        cJSON_Delete(root);
         return;
     } else {
         // Add sr command to speech recognition
         app_hass_add_cmd(sr_txt->valuestring, sr_phn->valuestring, true);
 
         app_hass_write_cmd_to_nvs(keynum, sr_txt->valuestring, sr_phn->valuestring);
         ESP_LOGI(TAG, "Added command: %s; %s", sr_txt->valuestring, sr_phn->valuestring);
-        cJSON_Delete(root);
         return;
     }
 }
@@ -210,7 +208,6 @@ void app_hass_rm_all_cmd(cJSON *root) {
     cJSON *sr_txt = cJSON_GetObjectItemCaseSensitive(root, "confirm");
     if (sr_txt == NULL) {
         ESP_LOGE(TAG, "Error parsing text");
-        cJSON_Delete(root);
         return;
     } else if (strcmp(sr_txt->valuestring, "yes") == 0) {
         // remove sr command from speech recognition
@@ -223,7 +220,6 @@ void app_hass_rm_all_cmd(cJSON *root) {
         app_hass_rm_cmds_from_nvs();
 
         ESP_LOGI(TAG, "Removed all commands");
-        cJSON_Delete(root);
         return;
     }
 }

main/app/app_hass.h

@@ -1,5 +1,6 @@
 #pragma once
 #include <esp_err.h>
+#include "cJSON.h"
 
 #ifdef __cplusplus
 extern "C" {
@@ -11,8 +12,8 @@ bool app_hass_is_connected(void);
 void app_hass_send_cmd(char *cmd);
 
 void app_hass_add_cmd(char *cmd, char *phoneme, bool commit);
-void app_hass_add_cmd_from_msg(char *msg);
-void app_hass_rm_all_cmd();
+void app_hass_add_cmd_from_msg(cJSON *root);
+void app_hass_rm_all_cmd(cJSON *root);
 
 #ifdef __cplusplus
 }

sites_template.yaml

@@ -0,0 +1,12 @@
+mqtt:
+  host: homeassistant.local
+  port: 1883
+  username: "your_mqtt_username"
+  password: "your_mqtt_password"
+  topic: esp-ha-speech
+
+sites:
+  esp32: # This is the name of the site/satellite
+    lights:
+      - "Light 1"
+      - "Light 2"