v1.1.4 - Fixed extra metadata information not being saved to M4B files

CHANGELOG.md

@@ -1,3 +1,14 @@
+# 1.1.4
+- Fixed extra metadata information not being saved to M4B files, ensuring that all metadata is correctly written to the output file.
+- Reformatted the code using Black for better readability and consistency.
+
+# 1.1.3
+- `M4B (with chapters)` generation is faster now, as it directly generates `m4b` files instead of converting from `wav`, which significantly reduces processing time, fixes the issue mentioned by @Milor123 in #39 
+- Better sleep state handling for Linux.
+- The app window now tries to fit the screen if its height would exceed the available display area.
+- Fixed issue where the app would not restart properly on Windows.
+- Fixed last sentence/subtitle entry timing in generated subtitles, the end time of the final subtitle entry now correctly matches the end of the audio chunk, preventing zero or invalid timings at the end.
+
 # v1.1.2
 - Now you can play the audio files while they are processing.
 - Audio and subtitle files are now written directly to disk during generation, which significantly reduces memory usage.

abogen/VERSION

@@ -1 +1 @@
-1.1.2
+1.1.4

abogen/conversion.py

@@ -488,13 +488,33 @@ class ConversionThread(QThread):
                     )
                     ffmpeg_proc = None
                 elif self.output_format == "m4b":
-                    temp_wav_path = os.path.join(
-                        out_dir, f"temp_{base_name}{suffix}.wav"
-                    )
-                    merged_out_file = sf.SoundFile(
-                        temp_wav_path, "w", samplerate=24000, channels=1, format="wav"
-                    )
+                    # Real-time M4B generation using FFmpeg pipe
+                    static_ffmpeg.add_paths()
+                    merged_out_file = None
                     ffmpeg_proc = None
+                    # Prepare ffmpeg command for m4b output
+                    cmd = [
+                        "ffmpeg",
+                        "-y",
+                        "-thread_queue_size",
+                        "32768",
+                        "-f",
+                        "f32le",
+                        "-ar",
+                        "24000",
+                        "-ac",
+                        "1",
+                        "-i",
+                        "pipe:0",
+                        "-c:a",
+                        "aac",
+                        "-q:a",
+                        "2",
+                        "-movflags",
+                        "+faststart+use_metadata_tags",
+                    ]
+                    cmd.append(merged_out_path)
+                    ffmpeg_proc = create_process(cmd, stdin=subprocess.PIPE, text=False)
                 elif self.output_format == "opus":
                     static_ffmpeg.add_paths()
                     cmd = [
@@ -793,6 +813,7 @@ class ConversionThread(QThread):
                                 tokens_with_timestamps,
                                 new_entries,
                                 self.max_subtitle_words,
+                                fallback_end_time=chunk_start + chunk_dur,
                             )
                             if merged_subtitle_file:
                                 subtitle_format = getattr(
@@ -819,6 +840,7 @@ class ConversionThread(QThread):
                                 chapter_tokens_with_timestamps,
                                 new_chapter_entries,
                                 self.max_subtitle_words,
+                                fallback_end_time=chapter_current_time + chunk_dur,
                             )
                             if chapter_subtitle_file:
                                 subtitle_format = getattr(
@@ -910,8 +932,9 @@ class ConversionThread(QThread):
                 if self.output_format in ["wav", "mp3", "flac"]:
                     merged_out_file.close()
                 elif self.output_format == "m4b":
-                    merged_out_file.close()
-                    # Only generate chapters info if there are chapters
+                    ffmpeg_proc.stdin.close()
+                    ffmpeg_proc.wait()
+                    # Add chapters via fast post-processing
                     if total_chapters > 1:
                         chapters_info_path = f"{base_filepath_no_ext}_chapters.txt"
                         with open(chapters_info_path, "w", encoding="utf-8") as f:
@@ -923,15 +946,19 @@ class ConversionThread(QThread):
                                 f.write(f"START={int(chapter['start']*1000)}\n")
                                 f.write(f"END={int(chapter['end']*1000)}\n")
                                 f.write(f"title={chapter_title}\n\n")
+                        # Fast mux chapters into m4b (write to temp file, then replace original)
+                        static_ffmpeg.add_paths()
+                        orig_path = merged_out_path
+                        root, ext = os.path.splitext(orig_path)
+                        tmp_path = root + ".tmp" + ext
                         metadata_options = (
                             self._extract_and_add_metadata_tags_to_ffmpeg_cmd()
                         )
-                        static_ffmpeg.add_paths()
                         cmd = [
                             "ffmpeg",
                             "-y",
                             "-i",
-                            temp_wav_path,
+                            orig_path,
                             "-i",
                             chapters_info_path,
                             "-map",
@@ -940,46 +967,15 @@ class ConversionThread(QThread):
                             "1",
                             "-map_chapters",
                             "1",
-                            *metadata_options,
                             "-c:a",
-                            "aac",
-                            "-q:a",
-                            "2",
-                            "-movflags",
-                            "+faststart+use_metadata_tags",
-                            merged_out_path,
+                            "copy",
                         ]
+                        cmd += metadata_options
+                        cmd.append(tmp_path)
                         proc = create_process(cmd)
                         proc.wait()
-                        self.log_updated.emit(
-                            ("\nCleaning up temporary files...", "grey")
-                        )
-                        if os.path.exists(chapters_info_path):
+                        os.replace(tmp_path, orig_path)
                         os.remove(chapters_info_path)
-                    else:
-                        # No chapters: skip chapters info and related ffmpeg args
-                        metadata_options = (
-                            self._extract_and_add_metadata_tags_to_ffmpeg_cmd()
-                        )
-                        static_ffmpeg.add_paths()
-                        cmd = [
-                            "ffmpeg",
-                            "-y",
-                            "-i",
-                            temp_wav_path,
-                            *metadata_options,
-                            "-c:a",
-                            "aac",
-                            "-q:a",
-                            "2",
-                            "-movflags",
-                            "+faststart+use_metadata_tags",
-                            merged_out_path,
-                        ]
-                        proc = create_process(cmd)
-                        proc.wait()
-                    if os.path.exists(temp_wav_path):
-                        os.remove(temp_wav_path)
                 elif self.output_format in ["opus"]:
                     ffmpeg_proc.stdin.close()
                     ffmpeg_proc.wait()
@@ -1146,7 +1142,11 @@ class ConversionThread(QThread):
         return f"{h}:{m:02}:{s:02}.{cs:02}"
 
     def _process_subtitle_tokens(
-        self, tokens_with_timestamps, subtitle_entries, max_subtitle_words
+        self,
+        tokens_with_timestamps,
+        subtitle_entries,
+        max_subtitle_words,
+        fallback_end_time=None,
     ):
         """Helper function to process subtitle tokens according to the subtitle mode"""
         if not tokens_with_timestamps:
@@ -1191,9 +1191,15 @@ class ConversionThread(QThread):
                 sentence_text = ""
                 for t in current_sentence:
                     sentence_text += t["text"] + (t.get("whitespace", "") or "")
-
                 subtitle_entries.append((start_time, end_time, sentence_text.strip()))
 
+            # Fallback for last entry
+            if subtitle_entries and fallback_end_time is not None:
+                last_entry = subtitle_entries[-1]
+                start, end, text = last_entry
+                if end is None or end <= start or end <= 0:
+                    subtitle_entries[-1] = (start, fallback_end_time, text)
+
         else:
             # Word count-based grouping
             try:
@@ -1230,36 +1236,12 @@ class ConversionThread(QThread):
                     subtitle_entries.append(
                         (group[0]["start"], group[-1]["end"], text.strip())
                     )
-
-    def _write_ass_subtitle(
-        self, file_path, subtitle_entries, is_centered=False, is_narrow=False
-    ):
-        with open(file_path, "w", encoding="utf-8", errors="replace") as f:
-            # Minimal ASS header
-            f.write("[Script Info]\n")
-            f.write("Title: Generated by Abogen\n")
-            f.write("ScriptType: v4.00+\n\n")
-
-            # Only events section, use override tags for positioning
-            f.write("[Events]\n")
-            f.write(
-                "Format: Layer, Start, End, Style, Name, MarginL, MarginR, MarginV, Effect, Text\n"
-            )
-
-            # Set margin based on is_narrow parameter
-            margin = "90" if is_narrow else ""
-            alignment_tag = ""
-            if is_centered:
-                alignment = 5
-                alignment_tag = f"{{\\an{alignment}}}"
-
-            # Write each subtitle with override tag for alignment and margins
-            for i, (start, end, text) in enumerate(subtitle_entries, 1):
-                start_time = self._ass_time(start)
-                end_time = self._ass_time(end)
-                f.write(
-                    f"Dialogue: 0,{start_time},{end_time},Default,,{margin},{margin},0,,{alignment_tag}{text}\n"
-                )
+            # Fallback for last entry
+            if subtitle_entries and fallback_end_time is not None:
+                last_entry = subtitle_entries[-1]
+                start, end, text = last_entry
+                if end is None or end <= start or end <= 0:
+                    subtitle_entries[-1] = (start, fallback_end_time, text)
 
     def cancel(self):
         self.cancel_requested = True

abogen/gui.py

@@ -763,7 +763,11 @@ class abogen(QWidget):
         self.setWindowTitle(f"{PROGRAM_NAME} v{VERSION}")
         screen = QApplication.primaryScreen().geometry()
         width, height = 500, 800
-        x, y = (screen.width() - width) // 2, (screen.height() - height) // 2
+        x = (screen.width() - width) // 2
+        # If desired height is larger than screen, fit to screen height
+        if height > screen.height() - 65:
+            height = screen.height() - 100  # Leave a margin for window borders
+        y = max((screen.height() - height) // 2, 0)
         self.setGeometry(x, y, width, height)
         outer_layout = QVBoxLayout()
         outer_layout.setContentsMargins(15, 15, 15, 15)
@@ -2815,6 +2819,24 @@ class abogen(QWidget):
         self.config["replace_single_newlines"] = enabled
         save_config(self.config)
 
+    def restart_app(self):
+        from PyQt5.QtCore import QProcess
+        import sys
+
+        exe = sys.executable
+        args = sys.argv
+
+        # On Windows, use .exe if available
+        if platform.system() == "Windows":
+            script_path = args[0]
+            if not script_path.lower().endswith(".exe"):
+                exe_path = os.path.splitext(script_path)[0] + ".exe"
+                if os.path.exists(exe_path):
+                    args[0] = exe_path
+
+        QProcess.startDetached(exe, args)
+        QApplication.quit()
+
     def toggle_kokoro_internet_access(self, disabled):
         if disabled:
             message = (
@@ -2838,11 +2860,7 @@ class abogen(QWidget):
             self.config["disable_kokoro_internet"] = disabled
             save_config(self.config)
             try:
-                from PyQt5.QtCore import QProcess
-                import sys
-
-                QProcess.startDetached(sys.executable, sys.argv)
-                QApplication.quit()
+                self.restart_app()
             except Exception as e:
                 QMessageBox.critical(
                     self, "Restart Failed", f"Failed to restart the application:\n{e}"
@@ -2858,16 +2876,12 @@ class abogen(QWidget):
         )
         if reply == QMessageBox.Yes:
             from abogen.utils import get_user_config_path
-            import sys
 
             config_path = get_user_config_path()
             try:
                 if os.path.exists(config_path):
                     os.remove(config_path)
-                from PyQt5.QtCore import QProcess
-
-                QProcess.startDetached(sys.executable, sys.argv)
-                QApplication.quit()
+                self.restart_app()
             except Exception as e:
                 QMessageBox.critical(
                     self, "Reset Error", f"Could not reset settings:\n{e}"

abogen/main.py

@@ -2,6 +2,7 @@ from json import load
 import os
 import sys
 import platform
+import atexit
 from PyQt5.QtWidgets import QApplication
 from PyQt5.QtGui import QIcon
 from PyQt5.QtCore import qInstallMessageHandler, QtMsgType
@@ -9,7 +10,7 @@ from PyQt5.QtCore import qInstallMessageHandler, QtMsgType
 # Add the directory to Python path
 sys.path.insert(0, os.path.join(os.path.dirname(__file__)))
 
-from abogen.utils import get_resource_path, load_config
+from abogen.utils import get_resource_path, load_config, prevent_sleep_end
 
 # Set Hugging Face Hub environment variables
 os.environ["HF_HUB_DISABLE_TELEMETRY"] = "1"  # Disable Hugging Face telemetry
@@ -27,6 +28,9 @@ from abogen.constants import PROGRAM_NAME, VERSION
 os.environ["MIOPEN_FIND_MODE"] = "FAST"
 os.environ["MIOPEN_CONV_PRECISE_ROCM_TUNING"] = "0"
 
+# Reset sleep states
+atexit.register(prevent_sleep_end)
+
 # Ensure sys.stdout and sys.stderr are valid in GUI mode
 if sys.stdout is None:
     sys.stdout = open(os.devnull, "w")

abogen/utils.py

@@ -244,6 +244,8 @@ def get_gpu_acceleration(enabled):
 
 
 def prevent_sleep_start():
+    from abogen.constants import PROGRAM_NAME
+
     system = platform.system()
     if system == "Windows":
         import ctypes
@@ -254,11 +256,15 @@ def prevent_sleep_start():
     elif system == "Darwin":
         _sleep_procs["Darwin"] = create_process(["caffeinate"])
     elif system == "Linux":
-        # use a sleep that never exits so inhibition stays active
+        # Add program name and reason for inhibition
+        program_name = PROGRAM_NAME
+        reason = "Prevent sleep during abogen process"
         _sleep_procs["Linux"] = create_process(
             [
                 "systemd-inhibit",
-                "--what=handle-lid-switch:sleep",
+                f"--who={program_name}",
+                f"--why={reason}",
+                "--what=sleep",
                 "--mode=block",
                 "sleep",
                 "infinity",