Add files via upload

Restructured code, replaced JSON definitions with DEFINEs, and improved flow.

ESP32-AI.ino

@@ -0,0 +1,24 @@
+#include "setup.h"
+
+void setup() {
+  Serial.begin(115200);
+  delay(1000);
+  Serial.println("---------- Starting ------------ ");
+  if(!start()){
+    return;
+  }
+  i2s_mic_install(); //Install I2S microphone driver
+  createWavAudio(samplesRecordedFilePath, SAMPLE_RATE*SECONDS);
+  WavToBase64inSDcard(samplesRecordedFilePath, samplesRecordedFilePath_base64);
+  String transcription = speechToText(); //transcribe audio to text
+  String answer = getAnswer(transcription);
+  i2s_driver_uninstall(I2S_NUM_0);
+  i2s_speaker_install(); //Install I2S Speaker Driver
+  textToSpeech(answer); //play the answer
+  Base64ToWavinSDcard(samplesGoogleFilePath_base64, samplesGoogleFilePath);
+  playAudioSamples(samplesGoogleFilePath);
+}
+
+void loop() {
+  // not used
+}

Google_Wrapper.cpp

@@ -0,0 +1,186 @@
+#include "Google_Wrapper.h"
+#include "global.h"
+#include "config.h"
+
+void textToSpeech(String text) {
+  Serial.print("Getting Text to Speech........ ");
+
+  WiFiClientSecure client;
+  client.setInsecure();
+  HTTPClient httpClient;
+  httpClient.begin(client, "https://texttospeech.googleapis.com/v1/text:synthesize?key=" + String(APIKEYGOOGLE));
+
+  DynamicJsonDocument doc(1024);
+  doc["input"]["text"] = text;
+  doc["voice"]["languageCode"] = "en-US";
+  doc["voice"]["name"] = "en-US-Wavenet-A";
+  doc["audioConfig"]["audioEncoding"] = "LINEAR16";
+  doc["audioConfig"]["sampleRateHertz"] = SAMPLE_RATE;
+  String requestBody;
+  serializeJson(doc, requestBody);
+
+  httpClient.addHeader("Content-Type", "application/json");
+  int httpCode = httpClient.POST(requestBody);
+  int contentLength = httpClient.getSize();
+
+  // Save the JSON response to a file on the SD card in chunks
+  File jsonResponseFile = SD.open(jsonResponseFilePath, FILE_WRITE);
+
+  const size_t chunkSize = 1024;
+  uint8_t buffer[chunkSize];
+  size_t bytesReceived;
+  while (httpClient.connected()) {
+    bytesReceived = httpClient.getStream().readBytesUntil('\0', buffer, chunkSize);
+    if (bytesReceived == 0) {
+      break;
+    }
+    jsonResponseFile.write(buffer, bytesReceived);
+  }
+
+  httpClient.end();
+
+  Serial.print("Status Code ");
+  Serial.println(httpCode);
+
+  // Find the position of the audioContent field in the JSON response
+  int audioContentPos = jsonResponseFile.find("\"audioContent\":");
+  if (audioContentPos == -1) {
+    Serial.println("Failed to find audioContent field in the JSON response");
+    jsonResponseFile.close();
+    return;
+  }
+
+  // Close and reopen as read, skip to audio content...
+  jsonResponseFile.close();
+  File jsonResponseFileRead = SD.open(jsonResponseFilePath, FILE_READ);
+  jsonResponseFileRead.seek(audioContentPos + 17);
+
+  //write base64 into sdcard
+  File audioSamplesFile_base64 = SD.open(samplesGoogleFilePath_base64, FILE_WRITE);
+  while (jsonResponseFileRead.available()) {
+    audioSamplesFile_base64.write(jsonResponseFileRead.read());  // Read one character from the file and write to the new file
+  }
+  jsonResponseFileRead.close();
+  audioSamplesFile_base64.close();
+
+  //Delete the JSON response file
+  //SD.remove(jsonResponseFilePath);
+}
+
+String speechToText(){
+  //USAGE: This function only has support for sending audio content from memory only.
+  //       In the future, an additional function with support for sending from a Google Storage URI will be added.
+  Serial.print("Getting Speech to Text........ ");  
+  WiFiClientSecure client;
+  client.setInsecure();
+  HTTPClient httpClient;
+  httpClient.begin(client, "https://speech.googleapis.com/v1p1beta1/speech:recognize?key=" + String(APIKEYGOOGLE));
+
+  File contentBase64 = SD.open(samplesRecordedFilePath_base64, FILE_READ);
+  Serial.println("PAYLOAD STATISTICS:");
+  Serial.printf("File Size:       %d Bytes\n", contentBase64.size());
+
+  int docSize = contentBase64.size() + 2048;
+  int availSpace = esp_get_free_heap_size();
+  Serial.print("Free heap size at start: ");
+  Serial.println(availSpace);
+
+  if (docSize > availSpace){
+    Serial.println("Payload requires more than available memory. Handle this.");
+  }
+
+  //construct parameter string for API call
+  String requestBody = "{\"config\":{\"encoding\":\"LINEAR16\",\"sampleRateHertz\":16000,\"languageCode\":\"en-US\"},\"audio\":{\"content\":\"";
+  const int BUFFER_SIZE = 1;
+  char buffer[BUFFER_SIZE];
+  String base64Str = "";
+  //read bytes of Base64 encoded recorded audio content from SD card
+  while (contentBase64.available()) {
+    int bytesRead = contentBase64.readBytes(buffer, BUFFER_SIZE);
+    base64Str += String(buffer).substring(0, bytesRead);
+  }
+  requestBody += base64Str;
+  requestBody += "\"}}";
+
+  httpClient.addHeader("Content-Type", "application/json");
+  httpClient.addHeader("Content-Length", String(requestBody.length()));
+  int httpCode = httpClient.POST(requestBody);
+  Serial.println(requestBody);
+  String responseString = httpClient.getString(); // get the whole response at once
+  DynamicJsonDocument respDoc(4096);
+  deserializeJson(respDoc, responseString);
+  Serial.println(responseString);
+
+  // Extract the content field
+  String content = respDoc["response"]["results"]["alternatives"]["transcript"];
+  int confidence = respDoc["response"]["results"]["alternatives"]["confidence"];
+
+  httpClient.end();
+  contentBase64.close();
+
+  Serial.print("Status Code ");
+  Serial.println(httpCode);
+
+  Serial.print("Transcripted Audio: ");
+  Serial.println(content);
+  Serial.print("Transcription Confidence: ");
+  Serial.println(confidence);
+
+  return content;
+}
+
+//Use Google refresh token to obtain your Oauth 2.0 token
+String getOAuthToken(){
+  Serial.print("Requesting Auth 2.0 Token.....");
+
+  WiFiClientSecure client;
+  client.setInsecure();
+  HTTPClient httpClient;
+
+  httpClient.begin("https://oauth2.googleapis.com/token");
+  httpClient.addHeader("Content-Type", "application/x-www-form-urlencoded");
+  String requestBody = "grant_type=refresh_token&response_type=token&refresh_token=" + String(REFRESHTOKENGOOGLE) + "&client_secret=" + String(CLIENTSECRETGOOGLE) + "&client_id="+ String(CLIENTIDGOOGLE);
+  int httpCode = httpClient.POST(requestBody);
+
+  String responseString = httpClient.getString(); // get the whole response at once
+  DynamicJsonDocument respDoc(1024);
+  deserializeJson(respDoc, responseString); //deserialize into JSON
+
+  String accessToken = respDoc["access_token"]; //obtain oauth2.0 token
+  if (httpCode == 200 && accessToken.length() > 30){  Serial.println(" Success!"); }
+
+  httpClient.end();
+  return accessToken; //return token
+}
+
+//not finished, only required if audio we are sending to speech-to-text is over 10mb or 1 minute
+void uploadToStorage(String token, String fileName){
+  Serial.print("Storing on Cloud.............. ");
+
+  WiFiClientSecure client;
+  client.setInsecure();
+  HTTPClient httpClient;
+  httpClient.begin(client, "https://storage.googleapis.com/upload/storage/v1/b/hara-esp32/o");
+  httpClient.addHeader("Authorization", "Bearer " + String(token));  // Your OAuth 2.0 token
+
+  DynamicJsonDocument doc(1024);
+  doc["documentId"] = "LINEAR16";
+  doc["title"] = fileName;
+  String requestBody;
+  serializeJson(doc, requestBody);
+
+  int httpCode = httpClient.POST(requestBody);
+
+  String responseString = httpClient.getString(); // get the whole response at once
+  DynamicJsonDocument respDoc(1024);
+  deserializeJson(respDoc, responseString);
+
+  Serial.print("Status Code ");
+  Serial.println(httpCode);
+  Serial.print("Response: ");
+  Serial.println(responseString);
+
+  httpClient.end();
+}
+
+// ... other function implementations