Merge pull request #45 from Integration-Automation/dev

Dev

dev.toml

@@ -6,14 +6,12 @@ build-backend = "setuptools.build_meta"
 
 [project]
 name = "frontengine_dev"
-version = "0.0.26"
+version = "0.0.28"
 authors = [
     { name = "JE-Chen", email = "[email protected]" },
 ]
 dependencies = [
-    "PySide6",
-    "qt-material",
-    "EdgeGPT"
+    "PySide6", "qt-material", "EdgeGPT", "pyttsx3", "SpeechRecognition", "PyAudio"
 ]
 description = "FrontEngine is BingGPT that can use-define front end or only use like screen saver"
 requires-python = ">=3.8"

frontengine/ui/chat/chat_scene_input.py

@@ -1,6 +1,5 @@
 import pyttsx3
 from PySide6.QtCore import QTimer
-from PySide6.QtGui import QScreen
 from PySide6.QtWidgets import QBoxLayout, QWidget, QPushButton, QHBoxLayout, QTextEdit, QMessageBox
 
 from frontengine.show.chat.chat_toast import ChatToast

frontengine/ui/chat/chatthread.py

@@ -57,7 +57,6 @@ async def send_chat_async():
 
             asyncio.run(send_chat_async())
             self.current_message = chat_response
-            print(json.dumps(self.current_message, indent=2))
             for text_dict in self.current_message.get("item").get("messages"):
                 if text_dict.get("author") == "bot":
                     response_text: str = text_dict.get("text")

frontengine/ui/chat/speech_to_text.py

@@ -0,0 +1,54 @@
+import sys
+import time
+
+from PySide6.QtWidgets import QWidget, QPushButton, QBoxLayout, QLineEdit
+from speech_recognition import Microphone
+from speech_recognition import Recognizer
+from speech_recognition import RequestError, UnknownValueError
+from threading import Thread
+
+from frontengine.utils.multi_language.language_wrapper import language_wrapper
+
+
+def callback(recognizer: Recognizer, audio):
+    try:
+        print(recognizer.recognize_google(audio))
+    except (RequestError, UnknownValueError) as error:
+        print(repr(error), file=sys.stderr)
+
+
+class ChatSpeechToText(QWidget):
+
+    def __init__(self):
+        super().__init__()
+        # Recognize
+        self.recognizer = Recognizer()
+        try:
+            self.microphone = Microphone()
+        except IOError as error:
+            print(repr(error), file=sys.stderr)
+        # UI
+        self.box_layout = QBoxLayout(QBoxLayout.Direction.LeftToRight)
+        self.voice_text_edit = QLineEdit()
+        self.start_listen_button = QPushButton(
+            language_wrapper.language_word_dict.get("chat_recognizer_voice_button"))
+        self.start_listen_button.clicked.connect(self.start_listener_thread)
+        self.send_text_button = QPushButton(
+            language_wrapper.language_word_dict.get("chat_scene_send_chat"))
+        self.box_layout.addWidget(self.voice_text_edit)
+        self.box_layout.addWidget(self.start_listen_button)
+        self.box_layout.addWidget(self.send_text_button)
+        self.setLayout(self.box_layout)
+
+    def start_listener_thread(self):
+        listener_thread = Thread(target=self.start_listener)
+        listener_thread.daemon = True
+        listener_thread.start()
+
+    def start_listener(self):
+        with self.microphone as source:
+            self.recognizer.adjust_for_ambient_noise(source, duration=0.3)
+        stop_listening = self.recognizer.listen_in_background(self.microphone, callback)
+        for receive_sound_time in range(50):
+            time.sleep(0.1)
+        stop_listening(wait_for_stop=False)

frontengine/ui/setting/chat_scene/chat_scene_setting.py

@@ -7,6 +7,7 @@
 from frontengine.ui.chat.chat_model import load_scene_json, chat_model
 from frontengine.ui.chat.chat_scene_input import ChatInputDialog
 from frontengine.ui.chat.chatthread import ChatThread, DELEGATE_CHAT
+from frontengine.ui.chat.speech_to_text import ChatSpeechToText
 from frontengine.utils.logging.loggin_instance import front_engine_logger
 from frontengine.utils.multi_language.language_wrapper import language_wrapper
 
@@ -15,6 +16,7 @@ class ChatSceneUI(QWidget):
 
     def __init__(self):
         super().__init__()
+        self.voice_input = None
         self.grid_layout = QGridLayout()
         self.grid_layout = QGridLayout(self)
         self.grid_layout.setContentsMargins(0, 0, 0, 0)
@@ -79,6 +81,10 @@ def __init__(self):
         self.local_box = QBoxLayout(QBoxLayout.Direction.LeftToRight)
         self.local_box.addWidget(self.locale_label)
         self.local_box.addWidget(self.locale_input)
+        # Start voice input
+        self.start_voice_input_button = QPushButton(
+            language_wrapper.language_word_dict.get("start_chat_voice_input_ui"))
+        self.start_voice_input_button.clicked.connect(self.start_voice_input)
         # Add to layout
         self.grid_layout.addWidget(self.choose_style_combobox, 0, 0)
         self.grid_layout.addWidget(self.close_delay_label, 0, 1)
@@ -88,7 +94,8 @@ def __init__(self):
         self.grid_layout.addLayout(self.local_box, 0, 5)
         self.grid_layout.addWidget(self.new_topic_button, 0, 6)
         self.grid_layout.addWidget(self.scene_input_button, 0, 7)
-        self.grid_layout.addWidget(self.start_button, 0, 8)
+        self.grid_layout.addWidget(self.start_voice_input_button, 0, 8)
+        self.grid_layout.addWidget(self.start_button, 0, 9)
         self.grid_layout.addWidget(self.chat_panel_scroll_area, 1, 0, -1, -1)
         self.setLayout(self.grid_layout)
 
@@ -102,6 +109,16 @@ def start_chat(self) -> None:
         if chat_model.rowCount() > 0:
             self.start_scene()
 
+    def start_voice_input(self):
+        self.voice_input = ChatSpeechToText()
+        self.voice_input.show()
+        self.voice_input.send_text_button.clicked.connect(self.send_voice_chat)
+
+    def send_voice_chat(self):
+        chat_thread = ChatThread(
+            self.chat_panel, self.voice_input.voice_text_edit.text(), self.locale_input.text())
+        chat_thread.start()
+
     def send_chat(self):
         chat_thread = ChatThread(
             self.chat_panel, self.chat_input.chat_input.toPlainText(), self.locale_input.text())

frontengine/utils/multi_language/english.py

@@ -107,4 +107,7 @@
     "chat_scene_precise": "precise",
     "chat_scene_balanced": "balanced",
     "chat_gpt_exception": "GPT error, pls retry or new topic",
+    # Chat voice input
+    "start_chat_voice_input_ui": "Open voice input ui",
+    "chat_recognizer_voice_button": "Start listen voice"
 }

frontengine/utils/multi_language/traditional_chinese.py

@@ -107,4 +107,7 @@
     "chat_scene_precise": "精準",
     "chat_scene_balanced": "平衡",
     "chat_gpt_exception": "GPT 錯誤，請重試或開始一個新話題",
+    # Chat voice input
+    "start_chat_voice_input_ui": "開啟聲音輸入介面",
+    "chat_recognizer_voice_button": "開始錄製聲音"
 }

pyproject.toml

@@ -6,14 +6,12 @@ build-backend = "setuptools.build_meta"
 
 [project]
 name = "frontengine"
-version = "0.0.25"
+version = "0.0.27"
 authors = [
     { name = "JE-Chen", email = "[email protected]" },
 ]
 dependencies = [
-    "PySide6",
-    "qt-material",
-    "EdgeGPT"
+    "PySide6", "qt-material", "EdgeGPT", "pyttsx3", "SpeechRecognition", "PyAudio"
 ]
 description = "FrontEngine is BingGPT that can use-define front end or only use like screen saver"
 requires-python = ">=3.8"

requirements.txt

@@ -0,0 +1,7 @@
+PySide6
+qt-material
+EdgeGPT
+pyttsx3
+SpeechRecognition
+frontengine
+PyAudio