pygpt-net 2.4.53__py3-none-any.whl → 2.4.55__py3-none-any.whl

pygpt_net/plugin/audio_output/worker.py

@@ -6,18 +6,26 @@
 # GitHub:  https://github.com/szczyglis-dev/py-gpt   #
 # MIT License                                        #
 # Created By  : Marcin Szczygliński                  #
-# Updated Date: 2024.11.26 19:00:00                  #
+# Updated Date: 2025.01.18 03:00:00                  #
 # ================================================== #
 
 import time
+import pyaudio
+import wave
+import numpy as np
+import io
+from pydub import AudioSegment
 
 from PySide6.QtCore import Slot, Signal
+
 from pygpt_net.plugin.base.worker import BaseWorker, BaseSignals
 
 
 class WorkerSignals(BaseSignals):
-    playback = Signal(object, str)
+    volume_changed = Signal(float)  # Emits the volume level
+    playback = Signal(str)
     stop = Signal()
+    error_playback = Signal(object)
 
 
 class Worker(BaseWorker):
@@ -32,6 +40,9 @@ class Worker(BaseWorker):
         self.cache_file = None  # path to cache file
         self.mode = "generate"  # generate|playback
         self.audio_file = None
+        self.is_stopped = False
+        self.p = None
+        self.stream = None
 
     @Slot()
     def run(self):
@@ -44,35 +55,118 @@ class Worker(BaseWorker):
             self.error(e)
 
     def generate(self):
-        """
-        Generate and play audio file
-        """
+        """Generate and play audio file"""
         if self.text is None or self.text == "":
             time.sleep(0.2)  # wait
             return
         path = self.plugin.get_provider().speech(self.text)
         if path:
-            from pygame import mixer
-            mixer.init()
-            playback = mixer.Sound(path)
-            self.stop_playback()  # stop previous playback
-            playback.play()
-            self.send(playback)  # send playback object to main thread to allow force stop
-
+            self.play_audio(path)
             # store in cache if enabled
             if self.cache_file:
                 self.cache_audio_file(path, self.cache_file)
 
     def play(self):
+        """Play audio file only"""
+        if self.audio_file:
+            self.play_audio(self.audio_file)
+
+    def play_audio(self, audio_file: str):
         """
-        Play audio file only
+        Play audio file using PyAudio
+
+        :param audio_file: audio file path
         """
-        if self.audio_file:
-            from pygame import mixer
-            mixer.init()
-            playback = mixer.Sound(self.audio_file)
-            playback.play()
-            self.send(playback)  # send playback object to main thread to allow force stop
+        try:
+            self.signals.playback.emit(self.event)
+            audio = AudioSegment.from_file(audio_file)
+            audio = audio.set_frame_rate(44100)  # resample to 44.1 kHz
+            wav_io = io.BytesIO()
+            audio.export(wav_io, format='wav')
+            wav_io.seek(0)
+            wf = wave.open(wav_io, 'rb')
+            self.p = pyaudio.PyAudio()
+            self.stream = self.p.open(format=self.p.get_format_from_width(wf.getsampwidth()),
+                                      channels=wf.getnchannels(),
+                                      rate=wf.getframerate(),
+                                      output=True)
+
+            sample_width = wf.getsampwidth()
+            format = self.p.get_format_from_width(sample_width)
+
+            if format == pyaudio.paInt8:
+                dtype = np.int8
+                max_value = 2 ** 7 - 1  # 127
+                offset = 0
+            elif format == pyaudio.paInt16:
+                dtype = np.int16
+                max_value = 2 ** 15 - 1  # 32767
+                offset = 0
+            elif format == pyaudio.paInt32:
+                dtype = np.int32
+                max_value = 2 ** 31 - 1  # 2147483647
+                offset = 0
+            elif format == pyaudio.paUInt8:
+                dtype = np.uint8
+                max_value = 2 ** 8 - 1  # 255
+                offset = 128  # center unsigned data
+            else:
+                raise ValueError(f"Unsupported format: {format}")
+
+            chunk_size = 512
+            data = wf.readframes(chunk_size)
+
+            while data != b'' and not self.is_stopped:
+                self.stream.write(data)
+
+                audio_data = np.frombuffer(data, dtype=dtype)
+                if len(audio_data) > 0:
+                    audio_data = audio_data.astype(np.float32)
+                    if dtype == np.uint8:
+                        audio_data -= offset
+
+                    # compute RMS
+                    rms = np.sqrt(np.mean(audio_data ** 2))
+
+                    if rms > 0:
+                        # RMS to decibels
+                        db = 20 * np.log10(rms / max_value)
+
+                        # define minimum and maximum dB levels
+                        min_db = -60  # adjust as needed
+                        max_db = 0
+
+                        # clamp the db value to the range [min_db, max_db]
+                        if db < min_db:
+                            db = min_db
+                        elif db > max_db:
+                            db = max_db
+
+                        # map decibel value to volume percentage
+                        volume_percentage = ((db - min_db) / (max_db - min_db)) * 100
+                    else:
+                        volume_percentage = 0
+
+                    # emit volume signal
+                    self.signals.volume_changed.emit(volume_percentage)
+                else:
+                    # if empty audio_data
+                    self.signals.volume_changed.emit(0)
+
+                data = wf.readframes(chunk_size)
+
+            # close the stream
+            if self.stream is not None:
+                if self.stream.is_active():
+                    self.stream.stop_stream()
+                self.stream.close()
+            if self.p is not None:
+                self.p.terminate()
+            wf.close()
+            self.signals.volume_changed.emit(0)
+        except Exception as e:
+            self.signals.volume_changed.emit(0)
+            self.signals.error_playback.emit(e)
 
     def cache_audio_file(self, src: str, dst: str):
         """
@@ -102,4 +196,4 @@ class Worker(BaseWorker):
 
     def stop(self):
         """Send stop signal to main thread"""
-        self.signals.stop.emit()
+        self.is_stopped = True

pygpt_net/provider/core/config/patch.py

@@ -6,7 +6,7 @@
 # GitHub:  https://github.com/szczyglis-dev/py-gpt   #
 # MIT License                                        #
 # Created By  : Marcin Szczygliński                  #
-# Updated Date: 2025.01.17 02:00:00                  #
+# Updated Date: 2025.01.18 16:00:00                  #
 # ================================================== #
 
 import copy
@@ -1811,6 +1811,14 @@ class Patch:
                 if 'audio.input.continuous' not in data:
                     data["audio.input.continuous"] = False
 
+            # < 2.4.55
+            if old < parse_version("2.4.55"):
+                print("Migrating config from < 2.4.55...")
+                if 'audio.input.timeout' not in data:
+                    data["audio.input.timeout"] = 120
+                if 'audio.input.timeout.continuous' not in data:
+                    data["audio.input.timeout.continuous"] = False
+
         # update file
         migrated = False
         if updated:

pygpt_net/ui/__init__.py

@@ -91,7 +91,8 @@ class UI:
 
         # FIRST RUN: initial sizes if not set yet
         if not self.window.core.config.has("layout.splitters") \
-            or self.window.core.config.get("layout.splitters") == {}:
+            or self.window.core.config.get("layout.splitters") == {}\
+            or self.window.core.config.get("license.accepted") == False:
             self.set_initial_size()
 
         # menus

pygpt_net/ui/dialog/snap.py

@@ -6,10 +6,10 @@
 # GitHub:  https://github.com/szczyglis-dev/py-gpt   #
 # MIT License                                        #
 # Created By  : Marcin Szczygliński                  #
-# Updated Date: 2024.02.26 22:00:00                  #
+# Updated Date: 2025.01.18 05:00:00                  #
 # ================================================== #
 
-from pygpt_net.ui.widget.dialog.snap import SnapDialogCamera, SnapDialogAudioInput
+from pygpt_net.ui.widget.dialog.snap import SnapDialogCamera, SnapDialogAudioInput, SnapDialogAudioOutput
 
 
 class Snap:
@@ -25,3 +25,4 @@ class Snap:
         """Setup snap dialog"""
         self.window.ui.dialog['snap_camera'] = SnapDialogCamera(self.window)
         self.window.ui.dialog['snap_audio_input'] = SnapDialogAudioInput(self.window)
+        self.window.ui.dialog['snap_audio_output'] = SnapDialogAudioOutput(self.window)

pygpt_net/ui/layout/chat/input.py

@@ -6,7 +6,7 @@
 # GitHub:  https://github.com/szczyglis-dev/py-gpt   #
 # MIT License                                        #
 # Created By  : Marcin Szczygliński                  #
-# Updated Date: 2024.12.09 03:00:00                  #
+# Updated Date: 2025.01.18 03:00:00                  #
 # ================================================== #
 
 from PySide6.QtCore import Qt
@@ -18,6 +18,7 @@ from pygpt_net.ui.layout.chat.attachments import Attachments
 from pygpt_net.ui.layout.chat.attachments_uploaded import AttachmentsUploaded
 from pygpt_net.ui.layout.chat.attachments_ctx import AttachmentsCtx
 from pygpt_net.ui.layout.status import Status
+from pygpt_net.ui.widget.audio.bar import OutputBar
 from pygpt_net.ui.widget.audio.input import AudioInput
 from pygpt_net.ui.widget.audio.input_button import AudioInputButton
 from pygpt_net.ui.widget.element.labels import HelpLabel
@@ -176,8 +177,10 @@ class Input:
 
         :return: QHBoxLayout
         """
+        self.window.ui.plugin_addon['audio.output.bar'] = OutputBar(self.window)
         layout = QHBoxLayout()
         layout.addLayout(self.status.setup())
+        layout.addWidget(self.window.ui.plugin_addon['audio.output.bar'], alignment=Qt.AlignCenter)
         layout.addLayout(self.setup_buttons())
         layout.setContentsMargins(2, 0, 2, 0)
 

pygpt_net/ui/widget/audio/bar.py

@@ -0,0 +1,103 @@
+#!/usr/bin/env python3
+# -*- coding: utf-8 -*-
+# ================================================== #
+# This file is a part of PYGPT package               #
+# Website: https://pygpt.net                         #
+# GitHub:  https://github.com/szczyglis-dev/py-gpt   #
+# MIT License                                        #
+# Created By  : Marcin Szczygliński                  #
+# Updated Date: 2025.01.18 03:00:00                  #
+# ================================================== #
+
+from PySide6.QtCore import Qt
+from PySide6.QtGui import QPainter, QPalette
+from PySide6.QtWidgets import QWidget
+
+import pygpt_net.icons_rc
+
+class InputBar(QWidget):
+    def __init__(self, parent=None):
+        super().__init__(parent)
+        self._level = 0.0  # level from 0.0 to 100.0
+        self.setFixedSize(200, 5)  # bar size
+
+    def setLevel(self, level):
+        """
+        Set volume level
+
+        :param level: level
+        """
+        self._level = level
+        self.update()
+
+    def paintEvent(self, event):
+        """
+        Paint event
+
+        :param event: event
+        """
+        palette = self.palette()
+        painter = QPainter(self)
+        painter.fillRect(self.rect(), Qt.transparent)
+        level_width = (self._level / 100.0) * self.width()
+        painter.setBrush(palette.color(QPalette.ButtonText))
+        painter.setPen(Qt.NoPen)
+        painter.drawRect(0, 0, level_width, self.height())
+
+    """
+        # --- bar from center ---
+        def paintEvent(self, event):
+        painter = QPainter(self)
+        painter.fillRect(self.rect(), Qt.transparent)
+        level_width = (self._level / 100.0) * self.width()
+        half_level_width = level_width / 2
+        center_x = self.width() / 2
+        rect_x = center_x - half_level_width
+        painter.setBrush(Qt.green)
+        painter.setPen(Qt.NoPen)
+        painter.drawRect(rect_x, 0, level_width, self.height())
+    """
+
+
+class OutputBar(QWidget):
+    def __init__(self, parent=None):
+        super().__init__(parent)
+        self._level = 0.0  # level from 0.0 to 100.0
+        self.setFixedSize(200, 5)  # bar size
+
+    def setLevel(self, level):
+        """
+        Set volume level
+
+        :param level: level
+        """
+        self._level = level
+        self.update()
+
+    def paintEvent(self, event):
+        """
+        Paint event
+
+        :param event: event
+        """
+        palette = self.palette()
+        painter = QPainter(self)
+        painter.fillRect(self.rect(), Qt.transparent)
+        level_width = (self._level / 100.0) * self.width()
+        painter.setBrush(palette.color(QPalette.ButtonText))
+        painter.setPen(Qt.NoPen)
+        painter.drawRect(0, 0, level_width, self.height())
+
+    """
+        # --- bar from center ---
+        def paintEvent(self, event):
+        painter = QPainter(self)
+        painter.fillRect(self.rect(), Qt.transparent)
+        level_width = (self._level / 100.0) * self.width()
+        half_level_width = level_width / 2
+        center_x = self.width() / 2
+        rect_x = center_x - half_level_width
+        painter.setBrush(Qt.green)
+        painter.setPen(Qt.NoPen)
+        painter.drawRect(rect_x, 0, level_width, self.height())
+    """

pygpt_net/ui/widget/audio/input_button.py

@@ -6,7 +6,7 @@
 # GitHub:  https://github.com/szczyglis-dev/py-gpt   #
 # MIT License                                        #
 # Created By  : Marcin Szczygliński                  #
-# Updated Date: 2025.01.17 02:00:00                  #
+# Updated Date: 2025.01.18 03:00:00                  #
 # ================================================== #
 
 from PySide6.QtCore import Qt
@@ -16,6 +16,8 @@ from PySide6.QtWidgets import QLabel, QHBoxLayout, QWidget, QPushButton, QVBoxLa
 from pygpt_net.core.events import Event, AppEvent
 from pygpt_net.ui.widget.option.toggle_label import ToggleLabel
 from pygpt_net.utils import trans
+from .bar import InputBar
+
 import pygpt_net.icons_rc
 
 class VoiceControlButton(QWidget):
@@ -34,7 +36,7 @@ class VoiceControlButton(QWidget):
         self.btn_toggle.setCursor(Qt.PointingHandCursor)
         self.btn_toggle.setMinimumWidth(200)
 
-        self.bar = LevelBar(self)
+        self.bar = InputBar(self)
         self.bar.setLevel(0)
 
         # status
@@ -89,7 +91,7 @@ class AudioInputButton(QWidget):
         self.btn_toggle.setCursor(Qt.PointingHandCursor)
         self.btn_toggle.setMinimumWidth(200)
 
-        self.bar = LevelBar(self)
+        self.bar = InputBar(self)
         self.bar.setLevel(0)
 
         btn_layout = QVBoxLayout()
@@ -133,46 +135,3 @@ class AudioInputButton(QWidget):
         """Toggle recording"""
         event = Event(Event.AUDIO_INPUT_RECORD_TOGGLE)
         self.window.dispatch(event)
-
-
-class LevelBar(QWidget):
-    def __init__(self, parent=None):
-        super().__init__(parent)
-        self._level = 0.0  # level from 0.0 to 100.0
-        self.setFixedSize(200, 5)  # bar size
-
-    def setLevel(self, level):
-        """
-        Set volume level
-
-        :param level: level
-        """
-        self._level = level
-        self.update()
-
-    def paintEvent(self, event):
-        """
-        Paint event
-
-        :param event: event
-        """
-        painter = QPainter(self)
-        painter.fillRect(self.rect(), Qt.transparent)
-        level_width = (self._level / 100.0) * self.width()
-        painter.setBrush(Qt.green)
-        painter.setPen(Qt.NoPen)
-        painter.drawRect(0, 0, level_width, self.height())
-
-    """
-        # --- bar from center ---
-        def paintEvent(self, event):
-        painter = QPainter(self)
-        painter.fillRect(self.rect(), Qt.transparent)
-        level_width = (self._level / 100.0) * self.width()
-        half_level_width = level_width / 2
-        center_x = self.width() / 2
-        rect_x = center_x - half_level_width
-        painter.setBrush(Qt.green)
-        painter.setPen(Qt.NoPen)
-        painter.drawRect(rect_x, 0, level_width, self.height())
-    """

pygpt_net/ui/widget/dialog/snap.py

@@ -24,7 +24,7 @@ class SnapDialogCamera(QDialog):
         super(SnapDialogCamera, self).__init__(window)
         self.window = window
         self.setParent(window)
-        self.setWindowTitle("Snap detected")
+        self.setWindowTitle("Snap version detected")
         self.cmd = CmdLabel(self.window, "sudo snap connect pygpt:camera")
 
         self.btn = QPushButton("OK")
@@ -34,8 +34,8 @@ class SnapDialogCamera(QDialog):
         # layout
         self.layout = QVBoxLayout()
         self.message = QLabel(
-            "Camera is not connected! It must be connected in Snap environment.\n"
-            "Run the following command to enable the camera:")
+            "Camera not connected? It must be connected in the Snap environment.\n"
+            "Run the following command to enable the camera and restart the application:")
         self.message.setStyleSheet("margin: 10px 0px 10px 0px;")
         self.layout.addWidget(self.message)
         self.layout.addWidget(self.cmd)
@@ -54,7 +54,7 @@ class SnapDialogAudioInput(QDialog):
         super(SnapDialogAudioInput, self).__init__(window)
         self.window = window
         self.setParent(window)
-        self.setWindowTitle("Snap is detected")
+        self.setWindowTitle("Snap version is detected")
         self.cmd = CmdLabel(self.window, "sudo snap connect pygpt:alsa && sudo snap connect pygpt:audio-record :audio-record")
 
         self.btn = QPushButton("OK")
@@ -64,8 +64,39 @@ class SnapDialogAudioInput(QDialog):
         # layout
         self.layout = QVBoxLayout()
         self.message = QLabel(
-            "Microphone is not connected! It must be connected in Snap environment.\n"
-            "Run the following command to enable the microphone:")
+            "Tip: Microphone must be manually connected in the Snap environment.\n"
+            "If it is connected, click on the OK button, and this warning will not be displayed again.\n"
+            "If it is NOT connected yet, run the following command and restart the application:")
+        self.message.setStyleSheet("margin: 10px 0px 10px 0px;")
+        self.layout.addWidget(self.message)
+        self.layout.addWidget(self.cmd)
+        self.layout.addWidget(self.btn)
+        self.layout.addStretch()
+        self.setLayout(self.layout)
+
+
+class SnapDialogAudioOutput(QDialog):
+    def __init__(self, window=None):
+        """
+        Snap dialog for audio output
+
+        :param window: main window
+        """
+        super(SnapDialogAudioOutput, self).__init__(window)
+        self.window = window
+        self.setParent(window)
+        self.setWindowTitle("Snap version is detected")
+        self.cmd = CmdLabel(self.window, "sudo snap connect pygpt:alsa && sudo snap connect pygpt:audio-playback")
+
+        self.btn = QPushButton("OK")
+        self.btn.clicked.connect(self.accept)
+        self.btn.setStyleSheet("margin: 10px 0px 0px 0px;")
+
+        # layout
+        self.layout = QVBoxLayout()
+        self.message = QLabel(
+            "Audio Device not connected? It must be connected in the Snap environment.\n"
+            "Run the following command to enable the audio output and restart the application:")
         self.message.setStyleSheet("margin: 10px 0px 10px 0px;")
         self.layout.addWidget(self.message)
         self.layout.addWidget(self.cmd)

{pygpt_net-2.4.53.dist-info → pygpt_net-2.4.55.dist-info}/METADATA

@@ -1,6 +1,6 @@
 Metadata-Version: 2.1
 Name: pygpt-net
-Version: 2.4.53
+Version: 2.4.55
 Summary: Desktop AI Assistant powered by models: OpenAI o1, GPT-4o, GPT-4, GPT-4 Vision, GPT-3.5, DALL-E 3, Llama 3, Mistral, Gemini, Claude, Bielik, and other models supported by Langchain, Llama Index, and Ollama. Features include chatbot, text completion, image generation, vision analysis, speech-to-text, internet access, file handling, command execution and more.
 Home-page: https://pygpt.net
 License: MIT
@@ -93,7 +93,7 @@ Description-Content-Type: text/markdown
 
 [![pygpt](https://snapcraft.io/pygpt/badge.svg)](https://snapcraft.io/pygpt)
 
-Release: **2.4.53** | build: **2025.01.17** | Python: **>=3.10, <3.13**
+Release: **2.4.55** | build: **2025.01.18** | Python: **>=3.10, <3.13**
 
 > Official website: https://pygpt.net | Documentation: https://pygpt.readthedocs.io
 > 
@@ -212,6 +212,13 @@ sudo snap connect pygpt:audio-record :audio-record
 sudo snap connect pygpt:alsa
 ```
 
+**Using audio output:** to use audio output in Snap version you must connect the audio with:
+
+```commandline
+sudo snap connect pygpt:audio-playback
+sudo snap connect pygpt:alsa
+```
+
 **Connecting IPython in Docker in Snap version**:
 
 To use IPython in the Snap version, you must connect PyGPT to the Docker daemon:
@@ -4044,6 +4051,16 @@ may consume additional tokens that are not displayed in the main window.
 
 ## Recent changes:
 
+**2.4.55 (2025-01-18)**
+
+- Added a new option in settings: Audio -> Recording timeout.
+- Added a new option in settings: Audio -> Enable timeout in continuous mode.
+
+**2.4.54 (2025-01-18)**
+
+- Audio output switched from PyGame to PyAudio. It may be necessary to manually connect Alsa in Snap version with: "sudo snap connect pygpt:alsa".
+- Added audio output volume progress bar.
+
 **2.4.53 (2025-01-17)**
 
 - Fix: issue #89