From 80ff1832cd3f847bbf9e0bc060a251b949e88704 Mon Sep 17 00:00:00 2001
From: Dobromir Popov <d-popov@abv.bg>
Date: Tue, 10 Sep 2024 02:31:14 +0300
Subject: [PATCH] win py agent maid - transcription service

---
 agent-mAId/main.py | 93 ++++++++++++++++++++++++++++++++++++++++++++++
 1 file changed, 93 insertions(+)
 create mode 100644 agent-mAId/main.py

diff --git a/agent-mAId/main.py b/agent-mAId/main.py
new file mode 100644
index 0000000..707f8b6
--- /dev/null
+++ b/agent-mAId/main.py
@@ -0,0 +1,93 @@
+# let's create a new application which will run in the background at startup, will listen to the microphone while mouse or other (configurable) button is pressed and will send the audio to Whisper and will enter (emulate keypress) of the transcribed text result on key release 
+# https://console.groq.com/docs/speech-text
+
+# import system_hooks
+# import audio_recorder
+# import whisper_api
+# import keyboard_emulator
+
+# def on_button_press():
+#     audio_recorder.start_recording()
+
+# def on_button_release():
+#     audio = audio_recorder.stop_recording()
+#     text = whisper_api.transcribe(audio)
+#     keyboard_emulator.type_text(text)
+
+# def main():
+#     system_hooks.set_startup_run()
+#     system_hooks.register_button_callback(on_button_press, on_button_release)
+#     system_hooks.run_event_loop()
+
+# if __name__ == "__main__":
+#     main()
+
+
+import pyaudio
+import wave
+import pyautogui
+import requests
+import keyboard
+
+# Constants
+API_URL = "https://api.openai.com/v1/whisper"
+API_KEY = "your_openai_api_key"
+BUTTON = 'ctrl'  # The button to listen for
+
+def record_audio(filename):
+    # Setup audio recording
+    audio = pyaudio.PyAudio()
+    stream = audio.open(format=pyaudio.paInt16, channels=1, rate=44100, input=True, frames_per_buffer=1024)
+    
+    frames = []
+    print("Recording...")
+
+    # Record while button is pressed
+    while keyboard.is_pressed(BUTTON):
+        data = stream.read(1024)
+        frames.append(data)
+    
+    print("Recording stopped.")
+    stream.stop_stream()
+    stream.close()
+    audio.terminate()
+
+    # Save the recorded audio
+    wave_file = wave.open(filename, 'wb')
+    wave_file.setnchannels(1)
+    wave_file.setsampwidth(audio.get_sample_size(pyaudio.paInt16))
+    wave_file.setframerate(44100)
+    wave_file.writeframes(b''.join(frames))
+    wave_file.close()
+
+def transcribe_audio(filename):
+    # Transcribe audio using Whisper API
+    with open(filename, 'rb') as audio_file:
+        response = requests.post(
+            API_URL,
+            headers={"Authorization": f"Bearer {API_KEY}"},
+            files={"file": audio_file}
+        )
+    return response.json().get('text', '')
+
+def simulate_keypress(text):
+    # Simulate keypress for each character in text
+    for char in text:
+        pyautogui.typewrite(char)
+    pyautogui.press('enter')
+
+def main():
+    filename = "output.wav"
+    
+    print("Press and hold the button to record...")
+    keyboard.wait(BUTTON)  # Wait for button press
+    record_audio(filename)
+    
+    print("Transcribing audio...")
+    transcribed_text = transcribe_audio(filename)
+    
+    print("Entering text...")
+    simulate_keypress(transcribed_text)
+
+if __name__ == "__main__":
+    main()
\ No newline at end of file