Update changelog

Optimize regex compilation and eliminate busy-wait loops

CHANGELOG.md

@@ -1,3 +1,21 @@
+# 1.2.4
+- **Subtitle generation is now available for all languages!** Abogen now supports subtitle generation for non-English languages using audio duration-based timing. Available modes include `Line`, `Sentence`, and `Sentence + Comma`. (Note: Word-level subtitle modes remain English-only due to Kokoro's timestamp token limitations.)
+- New option: **"Use spaCy for sentence segmentation"** You can now use [spaCy](https://spacy.io/) to automatically detect sentence boundaries and produce cleaner, more readable subtitles. Quick summary:
+    - **What it does:** Splits text into natural sentences so subtitle entries read better and align more naturally with speech.
+    - **Why this helps:** The previous punctuation-based splitting could break sentences incorrectly at common abbreviations (e.g. "Mr.", "Dr.", "Prof.") or initials, producing wrong subtitle breaks. spaCy avoids those false splits by using linguistic rules to detect real sentence boundaries.
+    - **For Non-English:** spaCy runs **before** audio generation to create better sentence chunks for TTS.
+    - **For English:** spaCy runs **during** subtitle generation to find accurate sentence breaks after TTS.
+    - **Note:** spaCy segmentation is only applied when subtitle mode is `Sentence` or `Sentence + Comma`. When turned off, it falls back to simple punctuation-based splitting.
+- New option: **Pre-download models and voices for offline use** You can now pre-download all required Kokoro models, voices, and spaCy language models using this option in the settings menu. Allowing you to use Abogen completely offline without any internet connection.
+- Added support for `.` separator in timestamps (e.g. `HH:MM:SS.ms`) for timestamp-based text files.
+- Optimized regex compilation and eliminated busy-wait loops.
+- Possibly fixed `Silent truncation of long paragraphs` issue mentioned in [#91](https://github.com/denizsafak/abogen/issues/91) by [@xklzlxr](https://github.com/xklzlxr)
+- Fixed unused regex patterns and variable naming conventions.
+- Improvements in code and documentation.
+
+# 1.2.3
+- Same as 1.2.2, re-released to fix an issue with subtitle timing when using timestamp-based text files.
+
 # 1.2.2
 - **You can now voice your subtitle files!** Simply add `.srt`, `.ass` or `.vtt` files to generate timed audio. Alternatively, add a text file with timestamps in `HH:MM:SS` or `HH:MM:SS,ms` format to generate audio that matches the timestamps. See [here](https://github.com/denizsafak/abogen?tab=readme-ov-file#about-timestamp-based-text-files) for detailed instructions.
     - New option: **"Use silent gaps between subtitles"**: Prevents unnecessary audio speed-up by letting speech continue into the silent gaps between subtitles.

README.md

@@ -171,8 +171,10 @@ Here’s Abogen in action: in this demo, it processes ∼3,000 characters of tex
 | **Clear cache files** | Deletes cache files created during the conversion or preview. |
 | **Use silent gaps between subtitles** | Prevents unnecessary audio speed-up by letting speech continue into the silent gaps between subtitle etries. In short, it ignores the end times in subtitle entries and uses the silent space until the beginning of the next subtitle entry. When disabled, it speeds up the audio to fit the exact time interval specified in the subtitle. (for subtitle files). |
 | **Subtitle speed adjustment method** | Choose how to speed up audio when needed: `TTS Regeneration (better quality)` re-generates the audio at a faster speed, while `FFmpeg Time-stretch (better speed)` quickly speeds up the generated audio. (for subtitle files). |
-| **Check for updates at startup** | Automatically checks for updates when the program starts. |
+| **Use spaCy for sentence segmentation** | When this option is enabled, Abogen uses [spaCy](https://spacy.io/) to detect sentence boundaries more accurately, instead of using punctuation marks (like periods, question marks, etc.) to split sentences, which could incorrectly cut off phrases like "Mr." or "Dr.". With spaCy, sentences are divided more accurately. For non-English text, spaCy runs **before** audio generation to create sentence chunks. For English text, spaCy runs **during** subtitle generation to improve timing and readability. spaCy is only used when subtitle mode is `Sentence` or `Sentence + Comma`. If you prefer the old punctuation splitting method, you can turn this option off. |
+| **Pre-download models and voices for offline use** | Opens a window that displays the available models and voices. Click `Download all` button to download all required models and voices, allowing you to use Abogen completely offline without any internet connection. |
 | **Disable Kokoro's internet access** | Prevents Kokoro from downloading models or voices from HuggingFace Hub, useful for offline use. |
+| **Check for updates at startup** | Automatically checks for updates when the program starts. |
 | **Reset to default settings** | Resets all settings to their default values. |
 
 > Special thanks to [@robmckinnon](https://github.com/robmckinnon) for adding Sentence + Highlighting feature in PR [#65](https://github.com/denizsafak/abogen/pull/65)
@@ -235,7 +237,7 @@ Similar to chapter markers, it is possible to add metadata tags for `M4B` files.
 > Note: `METADATA_COVER_PATH` is used to embed a cover image into the generated M4B file. Abogen automatically extracts the cover from EPUB and PDF files and adds this tag for you.
 
 ## `About Timestamp-based Text Files`
-Similar to converting subtitle files to audio, Abogen can automatically detect text files that contain timestamps in `HH:MM:SS` or `HH:MM:SS,ms` format. When timestamps are found inside your text file, Abogen will ask if you want to use them for audio timing. This is useful for creating timed narrations, scripts, or transcripts where you need exact control over when each segment is spoken.
+Similar to converting subtitle files to audio, Abogen can automatically detect text files that contain timestamps in `HH:MM:SS`, `HH:MM:SS,ms` or `HH:MM:SS.ms` format. When timestamps are found inside your text file, Abogen will ask if you want to use them for audio timing. This is useful for creating timed narrations, scripts, or transcripts where you need exact control over when each segment is spoken.
 
 Format your text file like this:
 ```
@@ -250,7 +252,7 @@ And this is the third segment, starting at 45 seconds.
 ```
 
 **Important notes:**
-- Timestamps must be in `HH:MM:SS` or `HH:MM:SS,ms` format (e.g., `00:05:30` for 5 minutes 30 seconds, or `00:05:30,500` for 5 minutes 30.5 seconds)
+- Timestamps must be in `HH:MM:SS`, `HH:MM:SS,ms` or `HH:MM:SS.ms` format (e.g., `00:05:30` for 5 minutes 30 seconds, or `00:05:30.500` for 5 minutes 30.5 seconds)
 - Milliseconds are optional and provide precision up to 1/1000th of a second
 - Text before the first timestamp (if any) will automatically start at `00:00:00`
 - When using timestamps, the subtitle generation mode setting is ignored
@@ -476,6 +478,7 @@ Feel free to explore the code and make any changes you like.
 
 ## `Credits`
 - Abogen uses [Kokoro](https://github.com/hexgrad/kokoro) for its high-quality, natural-sounding text-to-speech synthesis. Huge thanks to the Kokoro team for making this possible.
+- Thanks to the [spaCy](https://spacy.io/) project for its sentence-segmentation tools, which help Abogen produce cleaner, more natural sentence segmentation.
 - Thanks to [@wojiushixiaobai](https://github.com/wojiushixiaobai) for [Embedded Python](https://github.com/wojiushixiaobai/Python-Embed-Win64) packages. These modified packages include pip pre-installed, enabling Abogen to function as a standalone application without requiring users to separately install Python in Windows.
 - Thanks to creators of [EbookLib](https://github.com/aerkalov/ebooklib), a Python library for reading and writing ePub files, which is used for extracting text from ePub files.
 - Special thanks to the [PyQt](https://www.riverbankcomputing.com/software/pyqt/) team for providing the cross-platform GUI toolkit that powers Abogen's interface.
@@ -488,7 +491,7 @@ This project is available under the MIT License - see the [LICENSE](https://gith
 ## `Star History`
 [![Star History Chart](https://api.star-history.com/svg?repos=denizsafak/abogen&type=Date)](https://www.star-history.com/#denizsafak/abogen&Date)
 
-> [!IMPORTANT]
+> [!NOTE]
-> Subtitle generation currently works only for English. This is because Kokoro provides timestamp tokens only for English text. If you want subtitles in other languages, please request this feature in the [Kokoro project](https://github.com/hexgrad/kokoro). For more technical details, see [this line](https://github.com/hexgrad/kokoro/blob/6d87f4ae7abc2d14dbc4b3ef2e5f19852e861ac2/kokoro/pipeline.py#L383) in the Kokoro's code.
+> Abogen supports subtitle generation for all languages. However, word-level subtitle modes (e.g., "1 word", "2 words", "3 words", etc.) are only available for English because [Kokoro provides timestamp tokens only for English text](https://github.com/hexgrad/kokoro/blob/6d87f4ae7abc2d14dbc4b3ef2e5f19852e861ac2/kokoro/pipeline.py#L383). For non-English languages, Abogen uses a duration-based fallback that supports sentence-level and comma-based subtitle modes ("Line", "Sentence", "Sentence + Comma"). If you need word-level subtitles for other languages, please request that feature in the [Kokoro project](https://github.com/hexgrad/kokoro).
 
 > Tags: audiobook, kokoro, text-to-speech, TTS, audiobook generator, audiobooks, text to speech, audiobook maker, audiobook creator, audiobook generator, voice-synthesis, text to audio, text to audio converter, text to speech converter, text to speech generator, text to speech software, text to speech app, epub to audio, pdf to audio, markdown to audio, subtitle to audio, srt to audio, ass to audio, vtt to audio, webvtt to audio, content-creation, media-generation

WINDOWS_INSTALL.bat

@@ -278,7 +278,7 @@ if /I "%IS_NVIDIA%"=="true" (
     for /f %%i in ('%PYTHON_CONSOLE_PATH% -c "from torch.cuda import is_available; print(is_available())"') do set cuda_available=%%i
 
     if "%cuda_available%"=="False" (
-        echo Installing PyTorch with CUDA (12.8) support...
+        echo "Installing PyTorch with CUDA (12.8) support..."
         :: We need to use an older version of PyTorch (2.8.0) until this issue is fixed: https://github.com/pytorch/pytorch/issues/166628
         :: Solution mentioned by @mazenemam19 in #99:
         %PYTHON_CONSOLE_PATH% -m pip install torch==2.8.0+cu128 torchvision==0.23.0+cu128 torchaudio==2.8.0 --index-url https://download.pytorch.org/whl/cu128 --no-warn-script-location
@@ -304,7 +304,7 @@ if /I "%IS_NVIDIA%"=="true" (
     if errorlevel 2 (
         echo Skipping PyTorch installation.
     ) else (
-        echo Installing PyTorch with CUDA (12.8) support...
+        echo "Installing PyTorch with CUDA (12.8) support..."
         :: We need to use an older version of PyTorch (2.8.0) until this issue is fixed: https://github.com/pytorch/pytorch/issues/166628
         :: Solution mentioned by @mazenemam19 in #99:
         %PYTHON_CONSOLE_PATH% -m pip install torch==2.8.0+cu128 torchvision==0.23.0+cu128 torchaudio==2.8.0 --index-url https://download.pytorch.org/whl/cu128 --no-warn-script-location

abogen/VERSION

@@ -1 +1 @@
-1.2.2
+1.2.4

abogen/book_handler.py

@@ -44,6 +44,17 @@ logging.basicConfig(
     level=logging.INFO, format="%(asctime)s - %(levelname)s - %(message)s"
 )
 
+# Pre-compile frequently used regex patterns for better performance
+_BRACKETED_NUMBERS_PATTERN = re.compile(r"\[\s*\d+\s*\]")
+_STANDALONE_PAGE_NUMBERS_PATTERN = re.compile(r"^\s*\d+\s*$", re.MULTILINE)
+_PAGE_NUMBERS_AT_END_PATTERN = re.compile(r"\s+\d+\s*$", re.MULTILINE)
+_PAGE_NUMBERS_WITH_DASH_PATTERN = re.compile(
+    r"\s+[-–—]\s*\d+\s*[-–—]?\s*$", re.MULTILINE
+)
+_HTML_TAG_PATTERN = re.compile(r"<[^>]+>")
+_LEADING_DASH_PATTERN = re.compile(r"^\s*[-–—]\s*")
+_LEADING_SIMPLE_DASH_PATTERN = re.compile(r"^\s*-\s*")
+
 
 class HandlerDialog(QDialog):
     # Class variables to remember checkbox states between dialog instances
@@ -428,19 +439,12 @@ class HandlerDialog(QDialog):
         """Pre-process all page contents from PDF document"""
         for page_num in range(len(self.pdf_doc)):
             text = clean_text(self.pdf_doc[page_num].get_text())
-            # Remove bracketed numbers (citations, footnotes)
+            # Remove bracketed numbers, page numbers, etc. using pre-compiled patterns
-            text = re.sub(r"\[\s*\d+\s*\]", "", text)
+            # Combine all regex operations for better performance
-
+            text = _BRACKETED_NUMBERS_PATTERN.sub("", text)
-            # Remove standalone page numbers (numbers alone on a line)
+            text = _STANDALONE_PAGE_NUMBERS_PATTERN.sub("", text)
-            text = re.sub(r"^\s*\d+\s*$", "", text, flags=re.MULTILINE)
+            text = _PAGE_NUMBERS_AT_END_PATTERN.sub("", text)
-
+            text = _PAGE_NUMBERS_WITH_DASH_PATTERN.sub("", text)
-            # Remove page numbers at the end of paragraphs
-            # This pattern looks for digits surrounded by whitespace at the end of paragraphs
-            text = re.sub(r"\s+\d+\s*$", "", text, flags=re.MULTILINE)
-
-            # Also remove page numbers followed by a hyphen or dash at paragraph end
-            # (common in headers/footers like "- 42 -")
-            text = re.sub(r"\s+[-–—]\s*\d+\s*[-–—]?\s*$", "", text, flags=re.MULTILINE)
 
             page_id = f"page_{page_num + 1}"
             self.content_texts[page_id] = text
@@ -569,7 +573,7 @@ class HandlerDialog(QDialog):
                 text = clean_text(soup.get_text()).strip()
                 if text:
                     self.content_texts[doc_href] = text
-                    self.content_lengths[doc_href] = len(text)
+                    self.content_lengths[doc_href] = calculate_text_length(text)
 
                     title = None
                     if soup.title and soup.title.string:
@@ -892,7 +896,7 @@ class HandlerDialog(QDialog):
                 text = clean_text(slice_soup.get_text()).strip()
                 if text:
                     self.content_texts[current_src] = text
-                    self.content_lengths[current_src] = len(text)
+                    self.content_lengths[current_src] = calculate_text_length(text)
                 else:
                     self.content_texts[current_src] = ""
                     self.content_lengths[current_src] = 0
@@ -2015,7 +2019,8 @@ class HandlerDialog(QDialog):
         html_content += "<hr/>"
 
         if self.book_metadata["description"]:
-            desc = re.sub(r"<[^>]+>", "", self.book_metadata["description"])
+            # Use pre-compiled pattern for better performance
+            desc = _HTML_TAG_PATTERN.sub("", self.book_metadata["description"])
             html_content += f"<h3>Description:</h3><p>{desc}</p>"
 
         if self.file_type == "pdf":
@@ -2296,8 +2301,8 @@ class HandlerDialog(QDialog):
             text = self.content_texts.get(identifier)
             if text and text.strip():
                 title = item.text(0)
-                # Remove leading dashes from title
+                # Remove leading dashes from title using pre-compiled pattern
-                title = re.sub(r"^\s*[-–—]\s*", "", title).strip()
+                title = _LEADING_DASH_PATTERN.sub("", title).strip()
                 marker = f"<<CHAPTER_MARKER:{title}>>"
                 chapter_texts.append(marker + "\n" + text)
 
@@ -2331,7 +2336,8 @@ class HandlerDialog(QDialog):
             text = self.content_texts.get(identifier)
             if text and text.strip():
                 title = item.text(0)
-                title = re.sub(r"^\s*[-–—]\s*", "", title).strip()
+                # Use pre-compiled pattern for better performance
+                title = _LEADING_DASH_PATTERN.sub("", title).strip()
                 marker = f"<<CHAPTER_MARKER:{title}>>"
                 chapter_texts.append(marker + "\n" + text)
 
@@ -2401,12 +2407,16 @@ class HandlerDialog(QDialog):
                             combined_text += "\n\n" + child_text
                         included_text_ids.add(child_id)
                     if combined_text.strip():
-                        title = re.sub(r"^\s*-\s*", "", parent_title).strip()
+                        # Use pre-compiled pattern for better performance
+                        title = _LEADING_SIMPLE_DASH_PATTERN.sub(
+                            "", parent_title
+                        ).strip()
                         marker = f"<<CHAPTER_MARKER:{title}>>"
                         section_titles.append((title, marker + "\n" + combined_text))
                         included_text_ids.add(parent_id)
                 elif not parent_checked and checked_children:
-                    title = re.sub(r"^\s*-\s*", "", parent_title).strip()
+                    # Use pre-compiled pattern for better performance
+                    title = _LEADING_SIMPLE_DASH_PATTERN.sub("", parent_title).strip()
                     marker = f"<<CHAPTER_MARKER:{title}>>"
                     for idx, (child, child_id) in enumerate(checked_children):
                         text = self.content_texts.get(child_id, "")
@@ -2426,7 +2436,8 @@ class HandlerDialog(QDialog):
                     text = self.content_texts.get(identifier, "")
                     if text:
                         title = item.text(0)
-                        title = re.sub(r"^\s*-\s*", "", title).strip()
+                        # Use pre-compiled pattern for better performance
+                        title = _LEADING_SIMPLE_DASH_PATTERN.sub("", title).strip()
                         marker = f"<<CHAPTER_MARKER:{title}>>"
                         section_titles.append((title, marker + "\n" + text))
                         included_text_ids.add(identifier)

abogen/constants.py

@@ -61,10 +61,7 @@ SUPPORTED_INPUT_FORMATS = [
 # Please refer to: https://github.com/hexgrad/kokoro/blob/6d87f4ae7abc2d14dbc4b3ef2e5f19852e861ac2/kokoro/pipeline.py
 # 383 English processing (unchanged)
 # 384 if self.lang_code in 'ab':
-SUPPORTED_LANGUAGES_FOR_SUBTITLE_GENERATION = [
+SUPPORTED_LANGUAGES_FOR_SUBTITLE_GENERATION = list(LANGUAGE_DESCRIPTIONS.keys())
-    "a",
-    "b",
-]
 
 # Voice and sample text constants
 VOICES_INTERNAL = [

abogen/conversion.py

@@ -28,13 +28,40 @@ import threading  # for efficient waiting
 import subprocess
 import platform
 
+# Configuration constants
+_USER_RESPONSE_TIMEOUT = (
+    0.1  # Timeout in seconds for checking user response/cancellation
+)
+
+# Pre-compile frequently used regex patterns for better performance
+_METADATA_TAG_PATTERN = re.compile(r"<<METADATA_[^:]+:[^>]*>>")
+_CHAPTER_MARKER_PATTERN = re.compile(r"<<CHAPTER_MARKER:[^>]*>>")
+_HTML_TAG_PATTERN = re.compile(r"<[^>]+>")
+_VOICE_TAG_PATTERN = re.compile(r"{[^}]+}")
+_ASS_STYLING_PATTERN = re.compile(r"\{[^}]+\}")
+_ASS_NEWLINE_N_PATTERN = re.compile(r"\\N")
+_ASS_NEWLINE_LOWER_N_PATTERN = re.compile(r"\\n")
+_CHAPTER_MARKER_SEARCH_PATTERN = re.compile(r"<<CHAPTER_MARKER:(.*?)>>")
+_WEBVTT_HEADER_PATTERN = re.compile(r"^WEBVTT.*?\n", re.MULTILINE)
+_VTT_STYLE_PATTERN = re.compile(r"STYLE\s*\n.*?(?=\n\n|$)", re.DOTALL)
+_VTT_NOTE_PATTERN = re.compile(r"NOTE\s*\n.*?(?=\n\n|$)", re.DOTALL)
+_DOUBLE_NEWLINE_SPLIT_PATTERN = re.compile(r"\n\s*\n")
+_VTT_TIMESTAMP_PATTERN = re.compile(r"([\d:.]+)\s*-->\s*([\d:.]+)")
+_TIMESTAMP_ONLY_PATTERN = re.compile(r"^(\d{1,2}:\d{2}:\d{2}(?:[.,]\d{1,3})?)$")
+_WINDOWS_ILLEGAL_CHARS_PATTERN = re.compile(r'[<>:"/\\|?*]')
+_CONTROL_CHARS_PATTERN = re.compile(r"[\x00-\x1f]")
+_LINUX_CONTROL_CHARS_PATTERN = re.compile(
+    r"[\x01-\x1f]"
+)  # Linux: exclude \x00 for separate handling
+_MACOS_ILLEGAL_CHARS_PATTERN = re.compile(r"[:]")
+_LINUX_ILLEGAL_CHARS_PATTERN = re.compile(r"[/\x00]")
+
 
 def clean_subtitle_text(text):
     """Remove chapter markers and metadata tags from subtitle text."""
-    # Remove metadata tags
+    # Use pre-compiled patterns for better performance
-    text = re.sub(r"<<METADATA_[^:]+:[^>]*>>", "", text)
+    text = _METADATA_TAG_PATTERN.sub("", text)
-    # Remove chapter markers
+    text = _CHAPTER_MARKER_PATTERN.sub("", text)
-    text = re.sub(r"<<CHAPTER_MARKER:[^>]*>>", "", text)
     return text.strip()
 
 
@@ -87,8 +114,8 @@ def parse_srt_file(file_path):
             start_sec = time_to_seconds(start_str)
             end_sec = time_to_seconds(end_str)
 
-            # Clean text of any styling tags
+            # Clean text of any styling tags using pre-compiled pattern
-            text = re.sub(r"<[^>]+>", "", text)
+            text = _HTML_TAG_PATTERN.sub("", text)
             # Remove chapter markers and metadata tags
             text = clean_subtitle_text(text)
 
@@ -114,13 +141,13 @@ def parse_vtt_file(file_path):
     with open(file_path, "r", encoding=encoding, errors="replace") as f:
         content = f.read()
 
-    # Remove WEBVTT header and any style/note blocks
+    # Remove WEBVTT header and any style/note blocks using pre-compiled patterns
-    content = re.sub(r"^WEBVTT.*?\n", "", content, flags=re.MULTILINE)
+    content = _WEBVTT_HEADER_PATTERN.sub("", content)
-    content = re.sub(r"STYLE\s*\n.*?(?=\n\n|$)", "", content, flags=re.DOTALL)
+    content = _VTT_STYLE_PATTERN.sub("", content)
-    content = re.sub(r"NOTE\s*\n.*?(?=\n\n|$)", "", content, flags=re.DOTALL)
+    content = _VTT_NOTE_PATTERN.sub("", content)
 
-    # Split by double newlines to get individual subtitle blocks
+    # Split by double newlines to get individual subtitle blocks using pre-compiled pattern
-    blocks = re.split(r"\n\s*\n", content.strip())
+    blocks = _DOUBLE_NEWLINE_SPLIT_PATTERN.split(content.strip())
 
     subtitles = []
     for block in blocks:
@@ -147,7 +174,8 @@ def parse_vtt_file(file_path):
 
         try:
             # VTT format: 00:00:00.000 --> 00:00:05.000 or 00:00.000 --> 00:05.000
-            match = re.match(r"([\d:.]+)\s*-->\s*([\d:.]+)", timestamp_line)
+            # Use pre-compiled pattern
+            match = _VTT_TIMESTAMP_PATTERN.match(timestamp_line)
             if not match:
                 continue
 
@@ -171,9 +199,9 @@ def parse_vtt_file(file_path):
             start_sec = time_to_seconds(start_str)
             end_sec = time_to_seconds(end_str)
 
-            # Clean text of any styling tags and cue settings
+            # Clean text of any styling tags and cue settings using pre-compiled patterns
-            text = re.sub(r"<[^>]+>", "", text)
+            text = _HTML_TAG_PATTERN.sub("", text)
-            text = re.sub(r"{[^}]+}", "", text)  # Remove voice tags
+            text = _VOICE_TAG_PATTERN.sub("", text)  # Remove voice tags
             # Remove chapter markers and metadata tags
             text = clean_subtitle_text(text)
 
@@ -196,8 +224,10 @@ def detect_timestamps_in_text(file_path):
 
         # Count lines that are ONLY timestamps (no other text)
         # Supports HH:MM:SS or HH:MM:SS,ms format
-        timestamp_pattern = r"^(\d{1,2}:\d{2}:\d{2}(?:,\d{1,3})?)$"
+        # Use pre-compiled pattern for better performance
-        timestamp_lines = sum(1 for line in lines if re.match(timestamp_pattern, line))
+        timestamp_lines = sum(
+            1 for line in lines if _TIMESTAMP_ONLY_PATTERN.match(line)
+        )
 
         # Must have at least 2 timestamp-only lines and they should be >5% of total lines
         return timestamp_lines >= 2 and (timestamp_lines / max(len(lines), 1)) > 0.05
@@ -213,20 +243,14 @@ def parse_timestamp_text_file(file_path):
         content = f.read()
 
     # Split by timestamp pattern (supports HH:MM:SS or HH:MM:SS,ms)
-    pattern = r"^(\d{1,2}:\d{2}:\d{2}(?:,\d{1,3})?)$"
+    pattern = r"^(\d{1,2}:\d{2}:\d{2}(?:[.,]\d{1,3})?)$"
     lines = content.split("\n")
 
     def parse_time(time_str):
         """Convert HH:MM:SS or HH:MM:SS,ms to seconds as float."""
-        if "," in time_str:
+        time_str = time_str.replace(",", ".")
-            time_part, ms_part = time_str.split(",")
-            parts = time_part.split(":")
-            seconds = int(parts[0]) * 3600 + int(parts[1]) * 60 + int(parts[2])
-            milliseconds = int(ms_part.ljust(3, "0"))  # Pad to 3 digits
-            return seconds + milliseconds / 1000.0
-        else:
         parts = time_str.split(":")
-            return float(int(parts[0]) * 3600 + int(parts[1]) * 60 + int(parts[2]))
+        return float(int(parts[0]) * 3600 + int(parts[1]) * 60 + float(parts[2]))
 
     entries = []
     current_time = None
@@ -348,10 +372,12 @@ def parse_ass_file(file_path):
                 start_sec = ass_time_to_seconds(start_str)
                 end_sec = ass_time_to_seconds(end_str)
 
-                # Clean text of ASS styling tags
+                # Clean text of ASS styling tags using pre-compiled patterns
-                text = re.sub(r"\{[^}]+\}", "", text)  # Remove {tags}
+                text = _ASS_STYLING_PATTERN.sub("", text)  # Remove {tags}
-                text = re.sub(r"\\N", "\n", text)  # Convert \N to newline
+                text = _ASS_NEWLINE_N_PATTERN.sub("\n", text)  # Convert \N to newline
-                text = re.sub(r"\\n", "\n", text)  # Convert \n to newline
+                text = _ASS_NEWLINE_LOWER_N_PATTERN.sub(
+                    "\n", text
+                )  # Convert \n to newline
                 # Remove chapter markers and metadata tags
                 text = clean_subtitle_text(text)
 
@@ -384,9 +410,10 @@ def sanitize_name_for_os(name, is_folder=True):
     if system == "Windows":
         # Windows illegal characters: < > : " / \ | ? *
         # Also can't end with space or dot
-        sanitized = re.sub(r'[<>:"/\\|?*]', "_", name)
+        # Use pre-compiled pattern for better performance
+        sanitized = _WINDOWS_ILLEGAL_CHARS_PATTERN.sub("_", name)
         # Remove control characters (0-31)
-        sanitized = re.sub(r"[\x00-\x1f]", "_", sanitized)
+        sanitized = _CONTROL_CHARS_PATTERN.sub("_", sanitized)
         # Remove trailing spaces and dots
         sanitized = sanitized.rstrip(". ")
         # Windows reserved names (CON, PRN, AUX, NUL, COM1-9, LPT1-9)
@@ -400,18 +427,20 @@ def sanitize_name_for_os(name, is_folder=True):
     elif system == "Darwin":  # macOS
         # macOS illegal characters: : (colon is converted to / by the system)
         # Also can't start with dot (hidden file) for folders typically
-        sanitized = re.sub(r"[:]", "_", name)
+        # Use pre-compiled pattern for better performance
+        sanitized = _MACOS_ILLEGAL_CHARS_PATTERN.sub("_", name)
         # Remove control characters
-        sanitized = re.sub(r"[\x00-\x1f]", "_", sanitized)
+        sanitized = _CONTROL_CHARS_PATTERN.sub("_", sanitized)
         # Avoid leading dot for folders (creates hidden folders)
         if is_folder and sanitized.startswith("."):
             sanitized = "_" + sanitized[1:]
     else:  # Linux and others
         # Linux illegal characters: / and null character
         # Though / is illegal, most other chars are technically allowed
-        sanitized = re.sub(r"[/\x00]", "_", name)
+        # Use pre-compiled pattern for better performance
-        # Remove other control characters for safety
+        sanitized = _LINUX_ILLEGAL_CHARS_PATTERN.sub("_", name)
-        sanitized = re.sub(r"[\x01-\x1f]", "_", sanitized)
+        # Remove other control characters for safety (excluding \x00 which is already handled)
+        sanitized = _LINUX_CONTROL_CHARS_PATTERN.sub("_", sanitized)
         # Avoid leading dot for folders (creates hidden folders)
         if is_folder and sanitized.startswith("."):
             sanitized = "_" + sanitized[1:]
@@ -589,24 +618,44 @@ class ConversionThread(QThread):
     log_updated = pyqtSignal(object)  # Updated signal for log updates
     chapters_detected = pyqtSignal(int)  # Signal for chapter detection
 
-    # Default split pattern for TTS processing
+    # Punctuation constants for unified handling across languages
-    DEFAULT_SPLIT_PATTERN = r"\n+"
+    PUNCTUATION_SENTENCE = ".!?।。！？"
+    PUNCTUATION_SENTENCE_COMMA = ".!?,।。！？、，"
+    PUNCTUATION_COMMAS = ",，、"
 
-    # Languages that should not use split pattern (better handled by Kokoro internally)
+    def _get_split_pattern(self, lang_code, subtitle_mode):
-    # These languages have different text segmentation rules (no spaces, character-based, etc.)
+        """
-    NO_SPLIT_LANGUAGES = {"z", "j"}  # Chinese, Japanese
+        Get the appropriate split pattern based on language and subtitle mode.
 
-    # Language-specific punctuation patterns for subtitle splitting
+        Args:
-    LANGUAGE_PUNCTUATION = {
+            lang_code: Language code (a, b, e, f, etc.)
-        "z": {
+            subtitle_mode: Subtitle mode ("Sentence", "Sentence + Comma", "Line", etc.)
-            "sentence": r"[。！？]",  # Chinese: period, exclamation, question
+
-            "comma": r"[。！？、，]",  # Chinese: includes enumeration comma and comma
+        Returns:
-        },
+            Split pattern string
-        "j": {
+        """
-            "sentence": r"[。！？]",  # Japanese: period, exclamation, question
+        # For English, always use newline splitting only
-            "comma": r"[。！？、，]",  # Japanese: includes enumeration comma and comma
+        if lang_code in ["a", "b"]:
-        },
+            return "\n"
-    }
+
+        # Determine spacing pattern based on language
+        spacing_pattern = r"\s*" if lang_code in ["z", "j"] else r"\s+"
+
+        # For Chinese/Japanese, when subtitle mode is Disabled or Line, prefer
+        # punctuation-based splitting instead of plain newline splitting.
+        if subtitle_mode in ("Disabled", "Line") and lang_code in ["z", "j"]:
+            return r"(?<=[{}]){}|\n+".format(self.PUNCTUATION_SENTENCE, spacing_pattern)
+
+        if subtitle_mode == "Line":
+            return "\n"
+        elif subtitle_mode == "Sentence":
+            return r"(?<=[{}]){}|\n+".format(self.PUNCTUATION_SENTENCE, spacing_pattern)
+        elif subtitle_mode == "Sentence + Comma":
+            return r"(?<=[{}]){}|\n+".format(
+                self.PUNCTUATION_SENTENCE_COMMA, spacing_pattern
+            )
+        else:
+            return r"\n+"  # Default to line breaks
 
     def __init__(
         self,
@@ -628,6 +677,7 @@ class ConversionThread(QThread):
     ):  # Add use_gpu parameter
         super().__init__()
         self._chapter_options_event = threading.Event()
+        self._timestamp_response_event = threading.Event()
         self.np = np_module
         self.KPipeline = kpipeline_class
         self.file_name = file_name
@@ -655,10 +705,9 @@ class ConversionThread(QThread):
         self.use_gpu = use_gpu  # Store the GPU setting
         self.max_subtitle_words = 50  # Default value, will be overridden from GUI
         self.silence_duration = 2.0  # Default value, will be overridden from GUI
-        # Set split pattern based on language - some languages handle splitting better internally
+        self.use_spacy_segmentation = True  # Default, will be overridden from GUI
-        self.split_pattern = (
+        # Set split pattern based on language and subtitle mode
-            None if lang_code in self.NO_SPLIT_LANGUAGES else self.DEFAULT_SPLIT_PATTERN
+        self.split_pattern = self._get_split_pattern(lang_code, subtitle_mode)
-        )
 
     def _stream_audio_in_chunks(
         self, segments, process_func, progress_prefix="Processing"
@@ -678,7 +727,7 @@ class ConversionThread(QThread):
         total_samples = sum(len(segment) for segment in segments)
         samples_processed = 0
 
-        self.log_updated.emit(f"\n{progress_prefix} segments...")
+        self.log_updated.emit((f"\n{progress_prefix} segments...", "grey"))
 
         # Stream each segment individually
         for i, segment in enumerate(segments):
@@ -706,7 +755,9 @@ class ConversionThread(QThread):
                 # Clear segment bytes from memory
                 del segment_bytes
             except Exception as e:
-                self.log_updated.emit(f"Error processing segment {i}: {str(e)}")
+                self.log_updated.emit(
+                    (f"Error processing segment {i}: {str(e)}", "red")
+                )
                 raise
 
         return samples_processed
@@ -764,6 +815,9 @@ class ConversionThread(QThread):
             self.log_updated.emit(
                 f"- Subtitle format: {next((label for value, label in SUBTITLE_FORMATS if value == getattr(self, 'subtitle_format', 'srt')), getattr(self, 'subtitle_format', 'srt'))}"
             )
+            self.log_updated.emit(
+                f"- Use spaCy for sentence segmentation: {'Yes' if getattr(self, 'use_spacy_segmentation', False) else 'No'}"
+            )
             self.log_updated.emit(f"- Save option: {self.save_option}")
             if self.replace_single_newlines:
                 self.log_updated.emit(f"- Replace single newlines: Yes")
@@ -817,7 +871,7 @@ class ConversionThread(QThread):
                     f"- Output folder: {self.output_folder or os.getcwd()}"
                 )
 
-            self.log_updated.emit("\nInitializing TTS pipeline...")
+            self.log_updated.emit(("\nInitializing TTS pipeline...", "grey"))
 
             # Set device based on use_gpu setting and platform
             if self.use_gpu:
@@ -844,18 +898,26 @@ class ConversionThread(QThread):
                     )
                 elif file_ext == ".txt" and detect_timestamps_in_text(self.file_name):
                     is_timestamp_text = True
-                    self.log_updated.emit("\nDetected timestamps in text file")
+                    self.log_updated.emit(
+                        ("\nDetected timestamps in text file", "grey")
+                    )
                     # Signal to ask user (-1 indicates timestamp detection)
                     self.chapters_detected.emit(-1)
-                    # Wait for user response
+                    # Wait for user response using event with timeout for responsive cancellation
-                    while not hasattr(self, "_timestamp_response"):
+                    while not self._timestamp_response_event.wait(
+                        timeout=_USER_RESPONSE_TIMEOUT
+                    ):
+                        if self.cancel_requested:
+                            self.conversion_finished.emit("Cancelled", None)
+                            return
+                    # Check cancellation one more time after event is set
                     if self.cancel_requested:
                         self.conversion_finished.emit("Cancelled", None)
                         return
-                        time.sleep(0.1)
                     if not self._timestamp_response:
                         is_timestamp_text = False
                     delattr(self, "_timestamp_response")
+                    self._timestamp_response_event.clear()
 
             # Process subtitle files separately
             if is_subtitle_file or is_timestamp_text:
@@ -875,12 +937,12 @@ class ConversionThread(QThread):
             text = clean_text(text)
 
             # Remove metadata markers from the text to be processed
-            metadata_pattern = r"<<METADATA_[^:]+:[^>]*>>"
+            # Use pre-compiled pattern for better performance
-            text = re.sub(metadata_pattern, "", text)
+            text = _METADATA_TAG_PATTERN.sub("", text)
 
             # --- Chapter splitting logic ---
-            chapter_pattern = r"<<CHAPTER_MARKER:(.*?)>>"
+            # Use pre-compiled pattern for better performance
-            chapter_splits = list(re.finditer(chapter_pattern, text))
+            chapter_splits = list(_CHAPTER_MARKER_SEARCH_PATTERN.finditer(text))
             chapters = []
             if chapter_splits:
                 # prepend Introduction for content before first marker
@@ -936,7 +998,7 @@ class ConversionThread(QThread):
                     (f"\nDetected chapters ({total_chapters}):\n" + chapter_list)
                 )
             else:
-                self.log_updated.emit((f"\nProcessing {chapters[0][0]}..."))
+                self.log_updated.emit((f"\nProcessing {chapters[0][0]}...", "grey"))
 
             # If save_chapters_separately is enabled, find a unique suffix ONCE and use for both folder and merged file
             save_chapters_separately = getattr(self, "save_chapters_separately", False)
@@ -984,10 +1046,14 @@ class ConversionThread(QThread):
                     parent_dir, f"{sanitized_base_name}{suffix}_chapters"
                 )
                 # Only check for files with allowed extensions (extension without dot, case-insensitive)
+                # Use generator expression to avoid processing all files upfront
+                file_parts = (
+                    os.path.splitext(fname) for fname in os.listdir(parent_dir)
+                )
                 clash = any(
-                    os.path.splitext(fname)[0] == f"{sanitized_base_name}{suffix}"
+                    name == f"{sanitized_base_name}{suffix}"
-                    and os.path.splitext(fname)[1][1:].lower() in allowed_exts
+                    and ext[1:].lower() in allowed_exts
-                    for fname in os.listdir(parent_dir)
+                    for name, ext in file_parts
                 )
                 if not os.path.exists(chapters_out_dir_candidate) and not clash:
                     break
@@ -1334,11 +1400,96 @@ class ConversionThread(QThread):
                 else:
                     chapter_subtitle_path = None
                     chapter_subtitle_file = None
-                for result in tts(
+
+                # Determine if spaCy segmentation should be used for PRE-TTS segmentation
+                # Only non-English languages use spaCy for pre-segmentation
+                # English uses spaCy only for subtitle generation (post-TTS)
+                # spaCy is disabled when subtitle mode is "Disabled" or "Line"
+                # spaCy is also disabled when input is a subtitle file
+                is_subtitle_input = (
+                    not self.is_direct_text
+                    and self.file_name
+                    and os.path.splitext(self.file_name)[1].lower()
+                    in [".srt", ".ass", ".vtt"]
+                )
+                use_spacy = (
+                    getattr(self, "use_spacy_segmentation", False)
+                    and self.subtitle_mode not in ["Disabled", "Line"]
+                    and not is_subtitle_input
+                )
+                spacy_sentences = None
+                active_split_pattern = self.split_pattern
+                spacing_pattern = r"\s*" if self.lang_code in ["z", "j"] else r"\s+"
+
+                # Pre-load spaCy model for English if it will be needed for subtitle generation
+                if (
+                    use_spacy
+                    and self.lang_code in ["a", "b"]
+                    and self.subtitle_mode in ["Sentence", "Sentence + Comma"]
+                ):
+                    from abogen.spacy_utils import get_spacy_model
+
+                    nlp = get_spacy_model(
+                        self.lang_code,
+                        log_callback=lambda msg: self.log_updated.emit(msg),
+                    )
+                    if nlp:
+                        self.log_updated.emit(
+                            (
+                                "\nUsing spaCy for sentence segmentation (only for subtitles)...",
+                                "grey",
+                            )
+                        )
+
+                if use_spacy and self.lang_code not in ["a", "b"]:
+                    # Non-English: use spaCy for pre-TTS segmentation
+                    self.log_updated.emit(
+                        ("\nUsing spaCy for sentence segmentation (pre-TTS)...", "grey")
+                    )
+                    from abogen.spacy_utils import segment_sentences
+
+                    spacy_sentences = segment_sentences(
                         chapter_text,
+                        self.lang_code,
+                        log_callback=lambda msg: self.log_updated.emit(msg),
+                    )
+                    if spacy_sentences:
+                        self.log_updated.emit(
+                            (
+                                f"\nspaCy: Text segmented into {len(spacy_sentences)} sentences...",
+                                "grey",
+                            )
+                        )
+                        # For Sentence + Comma mode, still split on commas within spaCy sentences
+                        if self.subtitle_mode == "Sentence + Comma":
+                            active_split_pattern = r"(?<=[{}]){}|\n+".format(
+                                self.PUNCTUATION_COMMAS, spacing_pattern
+                            )
+                        else:
+                            active_split_pattern = (
+                                "\n"  # Use newline splitting for Sentence mode
+                            )
+                    else:
+                        self.log_updated.emit(
+                            ("\nspaCy: Fallback to default segmentation...", "grey")
+                        )
+
+                # Process text - either as spaCy sentences or as single text
+                text_segments = spacy_sentences if spacy_sentences else [chapter_text]
+
+                # Print active split pattern used by the TTS engine once for this batch
+                try:
+                    print(f"Using split pattern: {active_split_pattern!r}")
+                except Exception:
+                    # Print must never break processing
+                    print("Using split pattern: (unprintable)")
+
+                for text_segment in text_segments:
+                    for result in tts(
+                        text_segment,
                         voice=loaded_voice,
                         speed=self.speed,
-                    split_pattern=self.split_pattern,
+                        split_pattern=active_split_pattern,
                     ):
                         # Print the result for debugging
                         # print(f"Result: {result}")
@@ -1383,6 +1534,22 @@ class ConversionThread(QThread):
                         # Subtitle logic
                         if self.subtitle_mode != "Disabled":
                             tokens_list = getattr(result, "tokens", [])
+
+                            # Fallback for languages without token support (non-English)
+                            # Create a single token representing the entire segment duration
+                            if not tokens_list and result.graphemes:
+
+                                class FakeToken:
+                                    def __init__(self, text, start, end):
+                                        self.text = text
+                                        self.start_ts = start
+                                        self.end_ts = end
+                                        self.whitespace = ""
+
+                                tokens_list = [
+                                    FakeToken(result.graphemes, 0, chunk_dur)
+                                ]
+
                             tokens_with_timestamps = []
                             chapter_tokens_with_timestamps = []
 
@@ -1401,7 +1568,8 @@ class ConversionThread(QThread):
                                         {
                                             "start": chapter_current_time
                                             + (tok.start_ts or 0),
-                                        "end": chapter_current_time + (tok.end_ts or 0),
+                                            "end": chapter_current_time
+                                            + (tok.end_ts or 0),
                                             "text": tok.text,
                                             "whitespace": tok.whitespace,
                                         }
@@ -1485,7 +1653,10 @@ class ConversionThread(QThread):
                                 chapter_current_time += chunk_dur
                         # Calculate percentage based on characters processed
                         percent = min(
-                        int(self.processed_char_count / self.total_char_count * 100), 99
+                            int(
+                                self.processed_char_count / self.total_char_count * 100
+                            ),
+                            99,
                         )
 
                         # Calculate ETR based on characters processed
@@ -1498,7 +1669,9 @@ class ConversionThread(QThread):
                             chars_done > 0 and elapsed > 0.5
                         ):  # Check elapsed > 0.5 to avoid instability
                             avg_time_per_char = elapsed / chars_done
-                        remaining = self.total_char_count - self.processed_char_count
+                            remaining = (
+                                self.total_char_count - self.processed_char_count
+                            )
                             if remaining > 0:
                                 secs = avg_time_per_char * remaining
                                 h = int(secs // 3600)
@@ -1642,14 +1815,14 @@ class ConversionThread(QThread):
                 if self.subtitle_mode != "Disabled":
                     self.conversion_finished.emit(
                         (
-                            f"\nAudiobook saved to: {merged_out_path}\n\nSubtitle saved to: {merged_subtitle_path}",
+                            f"\nAudio saved to: {merged_out_path}\n\nSubtitle saved to: {merged_subtitle_path}",
                             "green",
                         ),
                         merged_out_path,
                     )
                 else:
                     self.conversion_finished.emit(
-                        (f"\nAudiobook saved to: {merged_out_path}", "green"),
+                        (f"\nAudio saved to: {merged_out_path}", "green"),
                         merged_out_path,
                     )
             else:
@@ -1701,7 +1874,9 @@ class ConversionThread(QThread):
                 )
                 return
 
-            self.log_updated.emit(f"\nFound {len(subtitles)} subtitle entries")
+            self.log_updated.emit(
+                (f"\nFound {len(subtitles)} subtitle entries", "grey")
+            )
 
             # Setup output paths
             base_name = os.path.splitext(os.path.basename(base_path))[0]
@@ -1727,10 +1902,12 @@ class ConversionThread(QThread):
             allowed_exts = set(SUPPORTED_SOUND_FORMATS + SUPPORTED_SUBTITLE_FORMATS)
             while True:
                 suffix = f"_{counter}" if counter > 1 else ""
+                # Use generator expression to avoid processing all files upfront
+                file_parts = (os.path.splitext(f) for f in os.listdir(parent_dir))
                 if not any(
-                    os.path.splitext(f)[0] == f"{sanitized_base_name}{suffix}"
+                    name == f"{sanitized_base_name}{suffix}"
-                    and os.path.splitext(f)[1][1:].lower() in allowed_exts
+                    and ext[1:].lower() in allowed_exts
-                    for f in os.listdir(parent_dir)
+                    for name, ext in file_parts
                 ):
                     break
                 counter += 1
@@ -1881,7 +2058,11 @@ class ConversionThread(QThread):
                     int(start_time % 60),
                 )
                 ms1 = int((start_time - int(start_time)) * 1000)
-                is_last = use_gaps and idx == len(subtitles)
+                is_last = (
+                    is_timestamp_text
+                    or (use_gaps and idx == len(subtitles))
+                    or end_time is None
+                )
                 if is_last:
                     time_str = (
                         f"{h1:02d}:{m1:02d}:{s1:02d}"
@@ -1938,7 +2119,10 @@ class ConversionThread(QThread):
                 audio_duration = len(full_audio) / rate
 
                 # Use actual audio length for timing
-                if use_gaps:
+                if is_timestamp_text:
+                    end_time = start_time + audio_duration
+                    subtitle_duration = audio_duration
+                elif use_gaps:
                     end_time = min(start_time + audio_duration, next_start)
                     subtitle_duration = end_time - start_time
                 elif subtitle_duration is None:
@@ -2127,7 +2311,7 @@ class ConversionThread(QThread):
                 subtitle_file.close()
 
             self.progress_updated.emit(100, "00:00:00")
-            result_msg = f"\nAudiobook saved to: {merged_out_path}" + (
+            result_msg = f"\nAudio saved to: {merged_out_path}" + (
                 f"\n\nSubtitle saved to: {subtitle_path}" if subtitle_path else ""
             )
             self.conversion_finished.emit((result_msg, "green"), merged_out_path)
@@ -2155,6 +2339,7 @@ class ConversionThread(QThread):
     def set_timestamp_response(self, treat_as_subtitle):
         """Set whether to treat timestamp text file as subtitle."""
         self._timestamp_response = treat_as_subtitle
+        self._timestamp_response_event.set()
 
     def _extract_and_add_metadata_tags_to_ffmpeg_cmd(self):
         """Extract metadata tags from text content and add them to ffmpeg command"""
@@ -2281,16 +2466,19 @@ class ConversionThread(QThread):
 
         processed_tokens = tokens_with_timestamps  # Use tokens directly
 
+        # For English with spaCy enabled and sentence-based modes, use spaCy for sentence boundaries
+        # spaCy is disabled when subtitle mode is "Disabled" or "Line"
+        use_spacy_for_english = (
+            getattr(self, "use_spacy_segmentation", False)
+            and self.subtitle_mode not in ["Disabled", "Line"]
+            and self.lang_code in ["a", "b"]
+            and self.subtitle_mode in ["Sentence", "Sentence + Comma"]
+        )
         # Use processed_tokens instead of tokens_with_timestamps for the rest of the method
         if self.subtitle_mode == "Sentence + Highlighting":
             # Sentence-based processing with karaoke highlighting
-            # Use language-specific punctuation for CJK languages (without comma)
+            # Use punctuation without comma
-            lang_punct = self.LANGUAGE_PUNCTUATION.get(self.lang_code, {})
+            separator = r"[{}]".format(self.PUNCTUATION_SENTENCE)
-            separator = (
-                lang_punct.get("sentence", r"[.!?]")
-                if isinstance(lang_punct, dict)
-                else r"[.!?]"
-            )
             current_sentence = []
             word_count = 0
 
@@ -2347,25 +2535,102 @@ class ConversionThread(QThread):
                     subtitle_entries[-1] = (start, fallback_end_time, text)
 
         elif self.subtitle_mode in ["Sentence", "Sentence + Comma", "Line"]:
+            # Check if we should use spaCy for English sentence boundaries
+            if use_spacy_for_english and self.subtitle_mode != "Line":
+                # Use spaCy for English sentence boundary detection (model already loaded)
+                from abogen.spacy_utils import get_spacy_model
+
+                nlp = get_spacy_model(
+                    self.lang_code
+                )  # No log_callback since model is already loaded
+                if nlp:
+                    # Build full text and track character positions to token indices
+                    full_text = ""
+                    char_to_token = []  # Maps character index to token index
+                    for idx, token in enumerate(processed_tokens):
+                        start_char = len(full_text)
+                        text_part = token["text"] + (token.get("whitespace", "") or "")
+                        full_text += text_part
+                        char_to_token.extend([idx] * len(text_part))
+
+                    # Get sentence boundaries from spaCy
+                    doc = nlp(full_text)
+                    sentence_boundaries = [sent.end_char for sent in doc.sents]
+
+                    # For "Sentence + Comma" mode, also split on commas
+                    if self.subtitle_mode == "Sentence + Comma":
+                        comma_positions = [
+                            i + 1 for i, c in enumerate(full_text) if c == ","
+                        ]
+                        sentence_boundaries = sorted(
+                            set(sentence_boundaries + comma_positions)
+                        )
+
+                    # Group tokens by sentence boundaries
+                    current_sentence = []
+                    word_count = 0
+                    current_char_pos = 0
+                    boundary_idx = 0
+
+                    for idx, token in enumerate(processed_tokens):
+                        current_sentence.append(token)
+                        word_count += 1
+                        text_len = len(token["text"]) + len(
+                            token.get("whitespace", "") or ""
+                        )
+                        current_char_pos += text_len
+
+                        # Check if we've hit a sentence boundary or max words
+                        at_boundary = (
+                            boundary_idx < len(sentence_boundaries)
+                            and current_char_pos >= sentence_boundaries[boundary_idx]
+                        )
+                        if at_boundary or word_count >= max_subtitle_words:
+                            if current_sentence:
+                                start_time = current_sentence[0]["start"]
+                                end_time = current_sentence[-1]["end"]
+                                sentence_text = "".join(
+                                    t["text"] + (t.get("whitespace", "") or "")
+                                    for t in current_sentence
+                                )
+                                subtitle_entries.append(
+                                    (start_time, end_time, sentence_text.strip())
+                                )
+                                current_sentence = []
+                                word_count = 0
+                            if at_boundary:
+                                boundary_idx += 1
+
+                    # Add remaining tokens
+                    if current_sentence:
+                        start_time = current_sentence[0]["start"]
+                        end_time = current_sentence[-1]["end"]
+                        sentence_text = "".join(
+                            t["text"] + (t.get("whitespace", "") or "")
+                            for t in current_sentence
+                        )
+                        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)
+                    return  # Exit early, spaCy processing complete
+
+            # Default regex-based processing (non-English or spaCy unavailable)
             # Define separator pattern based on mode
             if self.subtitle_mode == "Line":
                 separator = r"\n"
             elif self.subtitle_mode == "Sentence":
-                # Use language-specific punctuation for CJK languages (without comma)
+                # Use punctuation without comma
-                lang_punct = self.LANGUAGE_PUNCTUATION.get(self.lang_code, {})
+                separator = r"[{}]".format(self.PUNCTUATION_SENTENCE)
-                separator = (
-                    lang_punct.get("sentence", r"[.!?]")
-                    if isinstance(lang_punct, dict)
-                    else r"[.!?]"
-                )
             else:  # Sentence + Comma
-                # Use language-specific punctuation for CJK languages (with comma)
+                # Use punctuation with comma
-                lang_punct = self.LANGUAGE_PUNCTUATION.get(self.lang_code, {})
+                separator = r"[{}]".format(self.PUNCTUATION_SENTENCE_COMMA)
-                separator = (
-                    lang_punct.get("comma", r"[.!?,]")
-                    if isinstance(lang_punct, dict)
-                    else r"[.!?,]"
-                )
             current_sentence = []
             word_count = 0
 

abogen/gui.py

@@ -1,5 +1,6 @@
 import os
 import time
+import sys
 import tempfile
 import platform
 import base64
@@ -390,7 +391,7 @@ class InputBox(QLabel):
         window = self.window()
         if hasattr(window, "subtitle_combo"):
             # Only enable if language supports it
-            current_lang = getattr(window, "lang_code", "a")
+            current_lang = getattr(window, "selected_lang", "a")
             window.subtitle_combo.setEnabled(
                 current_lang in SUPPORTED_LANGUAGES_FOR_SUBTITLE_GENERATION
             )
@@ -824,9 +825,10 @@ class abogen(QWidget):
         self.use_gpu = self.config.get(
             "use_gpu", True  # Load GPU setting with default True
         )
-        self.replace_single_newlines = self.config.get("replace_single_newlines", False)
+        self.replace_single_newlines = self.config.get("replace_single_newlines", True)
-        self.use_silent_gaps = self.config.get("use_silent_gaps", False)
+        self.use_silent_gaps = self.config.get("use_silent_gaps", True)
         self.subtitle_speed_method = self.config.get("subtitle_speed_method", "tts")
+        self.use_spacy_segmentation = self.config.get("use_spacy_segmentation", True)
         self._pending_close_event = None
         self.gpu_ok = False  # Initialize GPU availability status
 
@@ -1062,9 +1064,6 @@ class abogen(QWidget):
         )
         self.subtitle_combo.setCurrentText(self.subtitle_mode)
         self.subtitle_combo.currentTextChanged.connect(self.on_subtitle_mode_changed)
-        # Enable/disable subtitle options based on selected language (profile or voice)
-        enable = self.selected_lang in SUPPORTED_LANGUAGES_FOR_SUBTITLE_GENERATION
-        self.subtitle_combo.setEnabled(enable)
         subtitle_layout.addWidget(self.subtitle_combo)
         controls_layout.addLayout(subtitle_layout)
 
@@ -1146,10 +1145,10 @@ class abogen(QWidget):
         except Exception:
             # Fail-safe: don't crash UI if model manipulation isn't supported on some platforms
             pass
-        # Enable/disable subtitle format based on selected language
+
-        self.subtitle_format_combo.setEnabled(
+        # Enable/disable subtitle options based on selected language (profile or voice)
-            self.selected_lang in SUPPORTED_LANGUAGES_FOR_SUBTITLE_GENERATION
+        self.update_subtitle_options_availability()
-        )
+
         controls_layout.addLayout(subtitle_format_layout)
 
         # Replace single newlines dropdown (acts like checkbox)
@@ -1590,19 +1589,69 @@ class abogen(QWidget):
         self.config["speed"] = s
         save_config(self.config)
 
+    def update_subtitle_options_availability(self):
+        """
+        Update the enabled state of subtitle options based on the selected language.
+        For non-English languages, only sentence-based and line-based modes are supported.
+        """
+        # Check if current file is a subtitle file
+        is_subtitle_input = False
+        if self.selected_file and self.selected_file.lower().endswith(
+            (".srt", ".ass", ".vtt")
+        ):
+            is_subtitle_input = True
+
+        if self.selected_lang not in SUPPORTED_LANGUAGES_FOR_SUBTITLE_GENERATION:
+            self.subtitle_combo.setEnabled(False)
+            self.subtitle_format_combo.setEnabled(False)
+            return
+
+        # Only enable subtitle_combo if it's NOT a subtitle input
+        self.subtitle_combo.setEnabled(not is_subtitle_input)
+        self.subtitle_format_combo.setEnabled(True)
+
+        is_english = self.selected_lang in ["a", "b"]
+
+        # Items to keep enabled for non-English
+        allowed_modes = ["Disabled", "Line", "Sentence", "Sentence + Comma"]
+
+        model = self.subtitle_combo.model()
+        for i in range(self.subtitle_combo.count()):
+            text = self.subtitle_combo.itemText(i)
+            item = model.item(i)
+            if not item:
+                continue
+
+            if is_english:
+                item.setEnabled(True)
+            else:
+                if text in allowed_modes:
+                    item.setEnabled(True)
+                else:
+                    item.setEnabled(False)
+
+        # If current selection is disabled, switch to a valid one
+        current_text = self.subtitle_combo.currentText()
+        current_idx = self.subtitle_combo.currentIndex()
+        current_item = model.item(current_idx)
+
+        if current_item and not current_item.isEnabled():
+            # Switch to "Sentence" if available, else "Disabled"
+            sentence_idx = self.subtitle_combo.findText("Sentence")
+            if sentence_idx >= 0:
+                self.subtitle_combo.setCurrentIndex(sentence_idx)
+            else:
+                self.subtitle_combo.setCurrentIndex(0)  # Disabled
+
+        self.subtitle_mode = self.subtitle_combo.currentText()
+
     def on_voice_changed(self, index):
         voice = self.voice_combo.itemData(index)
         self.selected_voice, self.selected_lang = voice, voice[0]
         self.config["selected_voice"] = voice
         save_config(self.config)
         # Enable/disable subtitle options based on language
-        if self.selected_lang in SUPPORTED_LANGUAGES_FOR_SUBTITLE_GENERATION:
+        self.update_subtitle_options_availability()
-            self.subtitle_combo.setEnabled(True)
-            self.subtitle_format_combo.setEnabled(True)
-            self.subtitle_mode = self.subtitle_combo.currentText()
-        else:
-            self.subtitle_combo.setEnabled(False)
-            self.subtitle_format_combo.setEnabled(False)
 
     def on_voice_combo_changed(self, index):
         data = self.voice_combo.itemData(index)
@@ -1624,12 +1673,7 @@ class abogen(QWidget):
             self.config.pop("selected_voice", None)
             save_config(self.config)
             # enable subtitles based on profile language
-            self.subtitle_combo.setEnabled(
+            self.update_subtitle_options_availability()
-                self.selected_lang in SUPPORTED_LANGUAGES_FOR_SUBTITLE_GENERATION
-            )
-            self.subtitle_format_combo.setEnabled(
-                self.selected_lang in SUPPORTED_LANGUAGES_FOR_SUBTITLE_GENERATION
-            )
         else:
             self.mixed_voice_state = None
             self.selected_profile_name = None
@@ -1638,13 +1682,7 @@ class abogen(QWidget):
             if "selected_profile_name" in self.config:
                 del self.config["selected_profile_name"]
             save_config(self.config)
-            if self.selected_lang in SUPPORTED_LANGUAGES_FOR_SUBTITLE_GENERATION:
+            self.update_subtitle_options_availability()
-                self.subtitle_combo.setEnabled(True)
-                self.subtitle_format_combo.setEnabled(True)
-                self.subtitle_mode = self.subtitle_combo.currentText()
-            else:
-                self.subtitle_combo.setEnabled(False)
-                self.subtitle_format_combo.setEnabled(False)
 
     def update_subtitle_combo_for_profile(self, profile_name):
         from abogen.voice_profiles import load_profiles
@@ -2044,7 +2082,7 @@ class abogen(QWidget):
         # pipeline_loaded_callback remains unchanged
         def pipeline_loaded_callback(np_module, kpipeline_class, error):
             if error:
-                self.update_log((f"Error loading numpy or KPipeline: {error}", False))
+                self.update_log((f"Error loading numpy or KPipeline: {error}", "red"))
                 prevent_sleep_end()
                 return
 
@@ -2091,6 +2129,8 @@ class abogen(QWidget):
             self.conversion_thread.use_silent_gaps = self.use_silent_gaps
             # Pass subtitle_speed_method setting
             self.conversion_thread.subtitle_speed_method = self.subtitle_speed_method
+            # Pass use_spacy_segmentation setting
+            self.conversion_thread.use_spacy_segmentation = self.use_spacy_segmentation
             # Pass separate_chapters_format setting
             self.conversion_thread.separate_chapters_format = (
                 self.separate_chapters_format
@@ -2227,7 +2267,7 @@ class abogen(QWidget):
         self.is_converting = False
         elapsed = int(time.time() - self.start_time)
         h, m, s = elapsed // 3600, (elapsed % 3600) // 60, elapsed % 60
-        self.update_log(f"\nTime elapsed: {h:02d}:{m:02d}:{s:02d}")
+        self.update_log((f"\nTime elapsed: {h:02d}:{m:02d}:{s:02d}", "grey"))
 
         # Default to showing the button
         show_open_file_button = True
@@ -2790,6 +2830,33 @@ class abogen(QWidget):
             self.conversion_thread.cancel()
             self.conversion_thread.wait()
 
+    def cleanup_preview_threads(self):
+        # Stop preview generation thread
+        if (
+            hasattr(self, "preview_thread")
+            and self.preview_thread is not None
+            and self.preview_thread.isRunning()
+        ):
+            self.preview_thread.terminate()
+            self.preview_thread.wait()
+
+        # Stop audio playback thread
+        if (
+            hasattr(self, "play_audio_thread")
+            and self.play_audio_thread is not None
+            and self.play_audio_thread.isRunning()
+        ):
+            self.play_audio_thread.stop()
+            self.play_audio_thread.wait()
+
+        # Cleanup pygame mixer if initialized
+        try:
+            pygame = sys.modules.get("pygame")
+            if pygame and pygame.mixer.get_init():
+                pygame.mixer.quit()
+        except Exception:
+            pass
+
     def closeEvent(self, event):
         if self.is_converting:
             box = QMessageBox(self)
@@ -2804,11 +2871,13 @@ class abogen(QWidget):
             box.setDefaultButton(QMessageBox.StandardButton.No)
             if box.exec() == QMessageBox.StandardButton.Yes:
                 self.cleanup_conversion_thread()
+                self.cleanup_preview_threads()
                 event.accept()
             else:
                 event.ignore()
         else:
             self.cleanup_conversion_thread()
+            self.cleanup_preview_threads()
             event.accept()
 
     def show_chapter_options_dialog(self, chapter_count):
@@ -3140,6 +3209,25 @@ class abogen(QWidget):
         # Add separator
         menu.addSeparator()
 
+        # Add spaCy sentence segmentation option
+        spacy_action = QAction("Use spaCy for sentence segmentation", self)
+        spacy_action.setCheckable(True)
+        spacy_action.setChecked(self.use_spacy_segmentation)
+        spacy_action.triggered.connect(
+            lambda checked: self.toggle_spacy_segmentation(checked)
+        )
+        menu.addAction(spacy_action)
+
+        # Add separator
+        menu.addSeparator()
+
+        # Add "Pre-download models and voices for offline use" option
+        predownload_action = QAction(
+            "Pre-download models and voices for offline use", self
+        )
+        predownload_action.triggered.connect(self.show_predownload_dialog)
+        menu.addAction(predownload_action)
+
         # Add "Disable Kokoro's internet access" option
         disable_kokoro_action = QAction("Disable Kokoro's internet access", self)
         disable_kokoro_action.setCheckable(True)
@@ -3204,6 +3292,11 @@ class abogen(QWidget):
         self.config["subtitle_speed_method"] = method
         save_config(self.config)
 
+    def toggle_spacy_segmentation(self, enabled):
+        self.use_spacy_segmentation = enabled
+        self.config["use_spacy_segmentation"] = enabled
+        save_config(self.config)
+
     def restart_app(self):
 
         import sys
@@ -3499,6 +3592,13 @@ Categories=AudioVideo;Audio;Utility;
                     self.voice_combo.setCurrentIndex(idx)
             self.mixed_voice_state = dialog.get_selected_voices()
 
+    def show_predownload_dialog(self):
+        """Show the pre-download models and voices dialog."""
+        from abogen.predownload_gui import PreDownloadDialog
+
+        dialog = PreDownloadDialog(self)
+        dialog.exec()
+
     def show_about_dialog(self):
         """Show an About dialog with program information including GitHub link."""
         # Get application icon for dialog

abogen/predownload_gui.py

@@ -0,0 +1,590 @@
+"""
+Pre-download dialog and worker for Abogen
+
+This module consolidates pre-download logic for Kokoro voices and model
+and spaCy language models. The code favors clarity, avoids duplication,
+and handles optional dependencies gracefully.
+"""
+
+from typing import List, Optional, Tuple
+import importlib
+import importlib.util
+
+from PyQt6.QtWidgets import (
+    QDialog,
+    QVBoxLayout,
+    QHBoxLayout,
+    QLabel,
+    QPushButton,
+    QSpacerItem,
+    QSizePolicy,
+)
+from PyQt6.QtCore import QThread, pyqtSignal
+
+from abogen.constants import COLORS, VOICES_INTERNAL
+from abogen.spacy_utils import SPACY_MODELS
+import abogen.hf_tracker
+
+
+# Helpers
+def _unique_sorted_models() -> List[str]:
+    """Return a sorted list of unique spaCy model package names."""
+    return sorted(set(SPACY_MODELS.values()))
+
+
+def _is_package_installed(pkg_name: str) -> bool:
+    """Return True if a package with the given name can be imported (site-packages)."""
+    try:
+        return importlib.util.find_spec(pkg_name) is not None
+    except Exception:
+        return False
+
+
+# NOTE: explicit HF cache helper removed; we use try_to_load_from_cache in-scope where needed
+
+
+class PreDownloadWorker(QThread):
+    """Worker thread to download required models/voices.
+
+    Emits human-readable messages via `progress`. Uses `category_done` to indicate
+    a category (voices/model/spacy) finished successfully. Emits `error` on exception
+    and `finished` after all work completes.
+    """
+
+    # Emit (category, status, message)
+    progress = pyqtSignal(str, str, str)
+    category_done = pyqtSignal(str)
+    finished = pyqtSignal()
+    error = pyqtSignal(str)
+
+    def __init__(self, parent=None):
+        super().__init__(parent)
+        self._cancelled = False
+        # repo and filenames used for Kokoro model
+        self._repo_id = "hexgrad/Kokoro-82M"
+        self._model_files = ["kokoro-v1_0.pth", "config.json"]
+        # Track download success per category
+        self._voices_success = False
+        self._model_success = False
+        self._spacy_success = False
+        # Suppress HF tracker warnings during downloads
+        self._original_emitter = abogen.hf_tracker.show_warning_signal_emitter
+
+    def cancel(self) -> None:
+        self._cancelled = True
+
+    def run(self) -> None:
+        # Suppress HF tracker warnings during downloads
+        abogen.hf_tracker.show_warning_signal_emitter = None
+        try:
+            self._download_kokoro_voices()
+            if self._cancelled:
+                return
+            if self._voices_success:
+                self.category_done.emit("voices")
+
+            self._download_kokoro_model()
+            if self._cancelled:
+                return
+            if self._model_success:
+                self.category_done.emit("model")
+
+            self._download_spacy_models()
+            if self._cancelled:
+                return
+            if self._spacy_success:
+                self.category_done.emit("spacy")
+
+            self.finished.emit()
+        except Exception as exc:  # pragma: no cover - best-effort reporting
+            self.error.emit(str(exc))
+        finally:
+            # Restore original emitter
+            abogen.hf_tracker.show_warning_signal_emitter = self._original_emitter
+
+    # Kokoro voices
+    def _download_kokoro_voices(self) -> None:
+        self._voices_success = True
+        try:
+            from huggingface_hub import hf_hub_download, try_to_load_from_cache
+        except Exception:
+            self.progress.emit(
+                "voice", "warning", "huggingface_hub not installed, skipping voices..."
+            )
+            self._voices_success = False
+            return
+
+        voice_list = VOICES_INTERNAL
+        for idx, voice in enumerate(voice_list, start=1):
+            if self._cancelled:
+                self._voices_success = False
+                return
+            filename = f"voices/{voice}.pt"
+            if try_to_load_from_cache(repo_id=self._repo_id, filename=filename):
+                self.progress.emit(
+                    "voice",
+                    "installed",
+                    f"{idx}/{len(voice_list)}: {voice} already present",
+                )
+                continue
+            self.progress.emit(
+                "voice", "downloading", f"{idx}/{len(voice_list)}: {voice}..."
+            )
+            try:
+                hf_hub_download(repo_id=self._repo_id, filename=filename)
+                self.progress.emit("voice", "downloaded", f"{voice} downloaded")
+            except Exception as exc:
+                self.progress.emit(
+                    "voice", "warning", f"could not download {voice}: {exc}"
+                )
+                self._voices_success = False
+
+    # Kokoro model
+    def _download_kokoro_model(self) -> None:
+        self._model_success = True
+        try:
+            from huggingface_hub import hf_hub_download, try_to_load_from_cache
+        except Exception:
+            self.progress.emit(
+                "model", "warning", "huggingface_hub not installed, skipping model..."
+            )
+            self._model_success = False
+            return
+        for fname in self._model_files:
+            if self._cancelled:
+                self._model_success = False
+                return
+            category = "config" if fname == "config.json" else "model"
+            if try_to_load_from_cache(repo_id=self._repo_id, filename=fname):
+                self.progress.emit(
+                    category, "installed", f"file {fname} already present"
+                )
+                continue
+            self.progress.emit(category, "downloading", f"file {fname}...")
+            try:
+                hf_hub_download(repo_id=self._repo_id, filename=fname)
+                self.progress.emit(category, "downloaded", f"file {fname} downloaded")
+            except Exception as exc:
+                self.progress.emit(
+                    category, "warning", f"could not download file {fname}: {exc}"
+                )
+                self._model_success = False
+
+    # spaCy models
+    def _download_spacy_models(self) -> None:
+        """Download spaCy models. Prefer missing models provided by parent.
+
+        Parent dialog will populate _spacy_models_missing during checking.
+        """
+        self._spacy_success = True
+        # Determine which models to process: prefer parent-provided missing list to avoid
+        # re-checking everything; otherwise use the full unique list.
+        parent = self.parent()
+        models_to_process: List[str] = _unique_sorted_models()
+        try:
+            if (
+                parent is not None
+                and hasattr(parent, "_spacy_models_missing")
+                and parent._spacy_models_missing
+            ):
+                models_to_process = list(dict.fromkeys(parent._spacy_models_missing))
+        except Exception:
+            pass
+
+        # If spaCy is not available to run the CLI, skip gracefully
+        try:
+            import spacy.cli as _spacy_cli
+        except Exception:
+            self.progress.emit(
+                "spacy", "warning", "spaCy not available, skipping spaCy models..."
+            )
+            self._spacy_success = False
+            return
+
+        for idx, model_name in enumerate(models_to_process, start=1):
+            if self._cancelled:
+                self._spacy_success = False
+                return
+            if _is_package_installed(model_name):
+                self.progress.emit(
+                    "spacy",
+                    "installed",
+                    f"{idx}/{len(models_to_process)}: {model_name} already installed",
+                )
+                continue
+            self.progress.emit(
+                "spacy",
+                "downloading",
+                f"{idx}/{len(models_to_process)}: {model_name}...",
+            )
+            try:
+                _spacy_cli.download(model_name)
+                self.progress.emit("spacy", "downloaded", f"{model_name} downloaded")
+            except Exception as exc:
+                self.progress.emit(
+                    "spacy", "warning", f"could not download {model_name}: {exc}"
+                )
+                self._spacy_success = False
+
+
+class PreDownloadDialog(QDialog):
+    """Dialog to show and control pre-download process."""
+
+    VOICE_PREFIX = "Kokoro voices: "
+    MODEL_PREFIX = "Kokoro model: "
+    CONFIG_PREFIX = "Kokoro config: "
+    SPACY_PREFIX = "spaCy models: "
+
+    def __init__(self, parent=None):
+        super().__init__(parent)
+        self.setWindowTitle("Pre-download Models and Voices")
+        self.setMinimumWidth(500)
+        self.worker: Optional[PreDownloadWorker] = None
+        self.has_missing = False
+        self._spacy_models_checked: List[tuple] = []
+        self._spacy_models_missing: List[str] = []
+        self._status_worker = None
+
+        # Map keywords to (label, prefix) - labels filled after UI creation
+        self.status_map = {
+            "voice": (None, self.VOICE_PREFIX),
+            "spacy": (None, self.SPACY_PREFIX),
+            "model": (None, self.MODEL_PREFIX),
+            "config": (None, self.CONFIG_PREFIX),
+        }
+
+        self.category_map = {
+            "voices": ["voice"],
+            "model": ["model", "config"],
+            "spacy": ["spacy"],
+        }
+
+        self._setup_ui()
+        self._start_status_check()
+
+    def _setup_ui(self) -> None:
+        layout = QVBoxLayout(self)
+        layout.setSpacing(0)
+        layout.setContentsMargins(15, 0, 15, 15)
+
+        desc = QLabel(
+            "You can pre-download all required models and voices for offline use.\n"
+            "This includes Kokoro voices, Kokoro model (and config), and spaCy models."
+        )
+        desc.setWordWrap(True)
+        layout.addWidget(desc)
+
+        # Status rows
+        status_layout = QVBoxLayout()
+        status_title = QLabel("<b>Current Status:</b>")
+        status_layout.addWidget(status_title)
+
+        self.voices_status = QLabel(self.VOICE_PREFIX + "⏳ Checking...")
+        row = QHBoxLayout()
+        row.addWidget(self.voices_status)
+        row.addStretch()
+        status_layout.addLayout(row)
+
+        self.model_status = QLabel(self.MODEL_PREFIX + "⏳ Checking...")
+        row = QHBoxLayout()
+        row.addWidget(self.model_status)
+        row.addStretch()
+        status_layout.addLayout(row)
+
+        self.config_status = QLabel(self.CONFIG_PREFIX + "⏳ Checking...")
+        row = QHBoxLayout()
+        row.addWidget(self.config_status)
+        row.addStretch()
+        status_layout.addLayout(row)
+
+        self.spacy_status = QLabel(self.SPACY_PREFIX + "⏳ Checking...")
+        row = QHBoxLayout()
+        row.addWidget(self.spacy_status)
+        row.addStretch()
+        status_layout.addLayout(row)
+
+        # register labels
+        self.status_map["voice"] = (self.voices_status, self.VOICE_PREFIX)
+        self.status_map["model"] = (self.model_status, self.MODEL_PREFIX)
+        self.status_map["config"] = (self.config_status, self.CONFIG_PREFIX)
+        self.status_map["spacy"] = (self.spacy_status, self.SPACY_PREFIX)
+
+        layout.addLayout(status_layout)
+
+        layout.addItem(
+            QSpacerItem(0, 20, QSizePolicy.Policy.Minimum, QSizePolicy.Policy.Fixed)
+        )
+
+        # Buttons
+        button_row = QHBoxLayout()
+        button_row.setSpacing(10)
+        self.download_btn = QPushButton("Download all")
+        self.download_btn.setMinimumWidth(100)
+        self.download_btn.setMinimumHeight(35)
+        self.download_btn.setEnabled(False)
+        self.download_btn.clicked.connect(self._start_download)
+        button_row.addWidget(self.download_btn)
+
+        self.close_btn = QPushButton("Close")
+        self.close_btn.setMinimumWidth(100)
+        self.close_btn.setMinimumHeight(35)
+        self.close_btn.clicked.connect(self._handle_close)
+        button_row.addWidget(self.close_btn)
+
+        layout.addLayout(button_row)
+        self.adjustSize()
+
+    # Status checking worker
+    class StatusCheckWorker(QThread):
+        voices_checked = pyqtSignal(bool, list)
+        model_checked = pyqtSignal(bool)
+        config_checked = pyqtSignal(bool)
+        spacy_model_checking = pyqtSignal(str)
+        spacy_model_result = pyqtSignal(str, bool)
+        spacy_checked = pyqtSignal(bool, list)
+
+        def run(self):
+            parent = self.parent()
+            if parent is None:
+                return
+
+            voices_ok, missing_voices = parent._check_kokoro_voices()
+            self.voices_checked.emit(voices_ok, missing_voices)
+
+            model_ok = parent._check_kokoro_model()
+            self.model_checked.emit(model_ok)
+
+            config_ok = parent._check_kokoro_config()
+            self.config_checked.emit(config_ok)
+
+            # Check spaCy models by package name to detect site-package installs
+            unique = _unique_sorted_models()
+            missing: List[str] = []
+            for name in unique:
+                self.spacy_model_checking.emit(name)
+                ok = _is_package_installed(name)
+                self.spacy_model_result.emit(name, ok)
+                if not ok:
+                    missing.append(name)
+            parent._spacy_models_missing = missing
+            self.spacy_checked.emit(len(missing) == 0, missing)
+
+    def _start_status_check(self) -> None:
+        self._status_worker = self.StatusCheckWorker(self)
+        self._status_worker.voices_checked.connect(self._update_voices_status)
+        self._status_worker.model_checked.connect(self._update_model_status)
+        self._status_worker.config_checked.connect(self._update_config_status)
+        self._status_worker.spacy_model_checking.connect(self._spacy_model_checking)
+        self._status_worker.spacy_model_result.connect(self._spacy_model_result)
+        self._status_worker.spacy_checked.connect(self._update_spacy_status)
+
+        # These are initialized in __init__ to keep consistent object state
+
+        # Set checking visual state
+        for lbl in (
+            self.voices_status,
+            self.model_status,
+            self.config_status,
+            self.spacy_status,
+        ):
+            lbl.setStyleSheet(f"color: {COLORS['ORANGE']};")
+
+        self.spacy_status.setText(self.SPACY_PREFIX + "⏳ Checking...")
+        self._status_worker.start()
+
+    # UI update callbacks
+    def _spacy_model_checking(self, name: str) -> None:
+        self.spacy_status.setText(f"{self.SPACY_PREFIX}Checking {name}...")
+
+    def _spacy_model_result(self, name: str, ok: bool) -> None:
+        self._spacy_models_checked.append((name, ok))
+        if not ok and name not in self._spacy_models_missing:
+            self._spacy_models_missing.append(name)
+        checked = len(self._spacy_models_checked)
+        missing_count = len(self._spacy_models_missing)
+        if missing_count:
+            self.spacy_status.setText(
+                f"{self.SPACY_PREFIX}{checked} checked, {missing_count} missing..."
+            )
+        else:
+            self.spacy_status.setText(f"{self.SPACY_PREFIX}{checked} checked...")
+
+    def _update_voices_status(self, ok: bool, missing: List[str]) -> None:
+        if ok:
+            self._set_status("voice", "✓ Downloaded", COLORS["GREEN"])
+        else:
+            self.has_missing = True
+            if missing:
+                self._set_status(
+                    "voice", f"✗ Missing {len(missing)} voices", COLORS["RED"]
+                )
+            else:
+                self._set_status("voice", "✗ Not downloaded", COLORS["RED"])
+
+    def _update_model_status(self, ok: bool) -> None:
+        if ok:
+            self._set_status("model", "✓ Downloaded", COLORS["GREEN"])
+        else:
+            self.has_missing = True
+            self._set_status("model", "✗ Not downloaded", COLORS["RED"])
+
+    def _update_config_status(self, ok: bool) -> None:
+        if ok:
+            self._set_status("config", "✓ Downloaded", COLORS["GREEN"])
+        else:
+            self.has_missing = True
+            self._set_status("config", "✗ Not downloaded", COLORS["RED"])
+
+    def _update_spacy_status(self, ok: bool, missing: List[str]) -> None:
+        if ok:
+            self._set_status("spacy", "✓ Downloaded", COLORS["GREEN"])
+        else:
+            self.has_missing = True
+            if missing:
+                self._set_status(
+                    "spacy", f"✗ Missing {len(missing)} model(s)", COLORS["RED"]
+                )
+            else:
+                self._set_status("spacy", "✗ Not downloaded", COLORS["RED"])
+        self.download_btn.setEnabled(self.has_missing)
+
+    def _set_status(self, key: str, text: str, color: str) -> None:
+        lbl, prefix = self.status_map.get(key, (None, ""))
+        if not lbl:
+            return
+        lbl.setText(prefix + text)
+        lbl.setStyleSheet(f"color: {color};")
+
+    # Helper checks
+    def _check_kokoro_voices(self) -> Tuple[bool, List[str]]:
+        """Return (ok, missing_list) for Kokoro voices check."""
+        missing = []
+        try:
+            from huggingface_hub import try_to_load_from_cache
+
+            for voice in VOICES_INTERNAL:
+                if not try_to_load_from_cache(
+                    repo_id="hexgrad/Kokoro-82M", filename=f"voices/{voice}.pt"
+                ):
+                    missing.append(voice)
+        except Exception:
+            # If HF missing, report all as missing
+            return False, list(VOICES_INTERNAL)
+        return (len(missing) == 0), missing
+
+    def _check_kokoro_model(self) -> bool:
+        try:
+            from huggingface_hub import try_to_load_from_cache
+
+            return (
+                try_to_load_from_cache(
+                    repo_id="hexgrad/Kokoro-82M", filename="kokoro-v1_0.pth"
+                )
+                is not None
+            )
+        except Exception:
+            return False
+
+    def _check_kokoro_config(self) -> bool:
+        try:
+            from huggingface_hub import try_to_load_from_cache
+
+            return (
+                try_to_load_from_cache(
+                    repo_id="hexgrad/Kokoro-82M", filename="config.json"
+                )
+                is not None
+            )
+        except Exception:
+            return False
+
+    def _check_spacy_models(self) -> bool:
+        unique = _unique_sorted_models()
+        missing = [m for m in unique if not _is_package_installed(m)]
+        self._spacy_models_missing = missing
+        return len(missing) == 0
+
+    # Download control
+    def _start_download(self) -> None:
+        self.download_btn.setEnabled(False)
+        self.download_btn.setText("Downloading...")
+        # mark the start of downloads; this triggers the labels
+        self._on_progress("system", "starting", "Processing, please wait...")
+        self.worker = PreDownloadWorker(self)
+        self.worker.progress.connect(self._on_progress)
+        self.worker.category_done.connect(self._on_category_done)
+        self.worker.finished.connect(self._on_download_finished)
+        self.worker.error.connect(self._on_download_error)
+        self.worker.start()
+
+    def _on_progress(self, category: str, status: str, message: str) -> None:
+        """Map worker (category, status, message) to UI label updates.
+
+        Status is one of: 'downloading', 'installed', 'downloaded', 'warning', 'starting'.
+        Category is one of: 'voice', 'model', 'spacy', 'config', or 'system'.
+        """
+        try:
+            # If the category targets a specific label, update directly
+            if category in self.status_map:
+                lbl, prefix = self.status_map[category]
+                if not lbl:
+                    return
+                # Compose message and set color based on status token
+                full_text = prefix + message
+                if len(full_text) > 60:
+                    display_text = full_text[:57] + "..."
+                    lbl.setText(display_text)
+                    lbl.setToolTip(full_text)
+                else:
+                    lbl.setText(full_text)
+                    lbl.setToolTip("")  # Clear tooltip if not needed
+                if status == "downloading":
+                    lbl.setStyleSheet(f"color: {COLORS['ORANGE']};")
+                elif status in ("installed", "downloaded"):
+                    lbl.setStyleSheet(f"color: {COLORS['GREEN']};")
+                elif status == "warning":
+                    lbl.setStyleSheet(f"color: {COLORS['RED']};")
+                elif status == "error":
+                    lbl.setStyleSheet(f"color: {COLORS['RED']};")
+                return
+
+            # System-level messages
+            if category == "system":
+                if status == "starting":
+                    for k in self.status_map:
+                        lbl, prefix = self.status_map[k]
+                        if lbl:
+                            lbl.setText(prefix + "Processing, please wait...")
+                            lbl.setStyleSheet(f"color: {COLORS['ORANGE']};")
+                # other system statuses don't require action
+                return
+        except Exception:
+            # Do not let UI thread crash on unexpected worker message
+            pass
+
+    def _on_category_done(self, category: str) -> None:
+        for key in self.category_map.get(category, []):
+            self._set_status(key, "✓ Downloaded", COLORS["GREEN"])
+
+    def _on_download_finished(self) -> None:
+        self.has_missing = False
+        self.download_btn.setText("Download all")
+        self.download_btn.setEnabled(False)
+
+    def _on_download_error(self, error_msg: str) -> None:
+        self.download_btn.setText("Download all")
+        self.download_btn.setEnabled(True)
+        for key in self.status_map:
+            self._set_status(key, f"✗ Error - {error_msg}", COLORS["RED"])
+
+    def _handle_close(self) -> None:
+        if self.worker and self.worker.isRunning():
+            self.worker.cancel()
+            self.worker.wait(2000)
+        self.accept()
+
+    def closeEvent(self, event) -> None:
+        if self.worker and self.worker.isRunning():
+            self.worker.cancel()
+            self.worker.wait(2000)
+        super().closeEvent(event)

abogen/spacy_utils.py

@@ -0,0 +1,155 @@
+"""
+Lazy-loaded spaCy utilities for sentence segmentation.
+"""
+
+# Cached spaCy module and models (lazy loaded)
+_spacy = None
+_nlp_cache = {}
+
+# Language code to spaCy model mapping
+SPACY_MODELS = {
+    "a": "en_core_web_sm",  # American English
+    "b": "en_core_web_sm",  # British English
+    "e": "es_core_news_sm",  # Spanish
+    "f": "fr_core_news_sm",  # French
+    "i": "it_core_news_sm",  # Italian
+    "p": "pt_core_news_sm",  # Brazilian Portuguese
+    "z": "zh_core_web_sm",  # Mandarin Chinese
+    "j": "ja_core_news_sm",  # Japanese
+    "h": "xx_sent_ud_sm",  # Hindi (multi-language model)
+}
+
+
+def _load_spacy():
+    """Lazy load spaCy module."""
+    global _spacy
+    if _spacy is None:
+        try:
+            import spacy
+
+            _spacy = spacy
+        except ImportError:
+            return None
+    return _spacy
+
+
+def get_spacy_model(lang_code, log_callback=None):
+    """
+    Get or load a spaCy model for the given language code.
+    Downloads the model automatically if not available.
+
+    Args:
+        lang_code: Language code (a, b, e, f, etc.)
+        log_callback: Optional function to log messages
+
+    Returns:
+        Loaded spaCy model or None if unavailable
+    """
+
+    def log(msg, is_error=False):
+        # Prefer GUI log callback when provided to avoid spamming stdout.
+        if log_callback:
+            color = "red" if is_error else "grey"
+            try:
+                log_callback((msg, color))
+            except Exception:
+                # Fallback to printing if callback misbehaves
+                print(msg)
+        else:
+            print(msg)
+
+    # Check if model is cached
+    if lang_code in _nlp_cache:
+        return _nlp_cache[lang_code]
+
+    # Check if language is supported
+    model_name = SPACY_MODELS.get(lang_code)
+    if not model_name:
+        log(f"\nspaCy: No model mapping for language '{lang_code}'...")
+        return None
+
+    # Lazy load spaCy
+    spacy = _load_spacy()
+    if spacy is None:
+        log("\nspaCy: Module not installed, falling back to default segmentation...")
+        return None
+
+    # Try to load the model
+    try:
+        log(f"\nLoading spaCy model '{model_name}'...")
+        nlp = spacy.load(
+            model_name,
+            disable=["ner", "parser", "tagger", "lemmatizer", "attribute_ruler"],
+        )
+        # Enable sentence segmentation only
+        if "sentencizer" not in nlp.pipe_names:
+            nlp.add_pipe("sentencizer")
+        _nlp_cache[lang_code] = nlp
+        return nlp
+    except OSError:
+        # Model not found, attempt download
+        log(f"\nspaCy: Downloading model '{model_name}'...")
+        try:
+            from spacy.cli import download
+
+            download(model_name)
+            # Retry loading
+            nlp = spacy.load(
+                model_name,
+                disable=["ner", "parser", "tagger", "lemmatizer", "attribute_ruler"],
+            )
+            if "sentencizer" not in nlp.pipe_names:
+                nlp.add_pipe("sentencizer")
+            _nlp_cache[lang_code] = nlp
+            log(f"spaCy model '{model_name}' downloaded and loaded")
+            return nlp
+        except Exception as e:
+            log(
+                f"\nspaCy: Failed to download model '{model_name}': {e}...",
+                is_error=True,
+            )
+            return None
+    except Exception as e:
+        log(f"\nspaCy: Error loading model '{model_name}': {e}...", is_error=True)
+        return None
+
+
+def segment_sentences(text, lang_code, log_callback=None):
+    """
+    Segment text into sentences using spaCy.
+
+    Args:
+        text: Text to segment
+        lang_code: Language code
+        log_callback: Optional function to log messages
+
+    Returns:
+        List of sentence strings, or None if spaCy unavailable
+    """
+    nlp = get_spacy_model(lang_code, log_callback)
+    if nlp is None:
+        return None
+
+    # Ensure spaCy can handle large texts by adjusting max_length if necessary
+    try:
+        text_len = len(text or "")
+        if text_len and hasattr(nlp, "max_length") and text_len > nlp.max_length:
+            # increase a bit beyond the text length to be safe
+            nlp.max_length = text_len + 1000
+    except Exception:
+        pass
+
+    # Process text and extract sentences
+    doc = nlp(text)
+    return [sent.text.strip() for sent in doc.sents if sent.text.strip()]
+
+
+def is_spacy_available():
+    """Check if spaCy can be imported."""
+    return _load_spacy() is not None
+
+
+def clear_cache():
+    """Clear the model cache to free memory."""
+    global _nlp_cache
+    _nlp_cache.clear()

abogen/utils.py

@@ -10,6 +10,13 @@ from threading import Thread
 
 warnings.filterwarnings("ignore")
 
+# Pre-compile frequently used regex patterns for better performance
+_WHITESPACE_PATTERN = re.compile(r"[^\S\n]+")
+_MULTIPLE_NEWLINES_PATTERN = re.compile(r"\n{3,}")
+_SINGLE_NEWLINE_PATTERN = re.compile(r"(?<!\n)\n(?!\n)")
+_CHAPTER_MARKER_PATTERN = re.compile(r"<<CHAPTER_MARKER:.*?>>")
+_METADATA_PATTERN = re.compile(r"<<METADATA_[^:]+:[^>]*>>")
+
 
 def detect_encoding(file_path):
     import chardet
@@ -128,13 +135,16 @@ def clean_text(text, *args, **kwargs):
     cfg = load_config()
     replace_single_newlines = cfg.get("replace_single_newlines", False)
     # Collapse all whitespace (excluding newlines) into single spaces per line and trim edges
-    lines = [re.sub(r"[^\S\n]+", " ", line).strip() for line in text.splitlines()]
+    # Use pre-compiled pattern for better performance
+    lines = [_WHITESPACE_PATTERN.sub(" ", line).strip() for line in text.splitlines()]
     text = "\n".join(lines)
     # Standardize paragraph breaks (multiple newlines become exactly two) and trim overall whitespace
-    text = re.sub(r"\n{3,}", "\n\n", text).strip()
+    # Use pre-compiled pattern for better performance
+    text = _MULTIPLE_NEWLINES_PATTERN.sub("\n\n", text).strip()
     # Optionally replace single newlines with spaces, but preserve double newlines
     if replace_single_newlines:
-        text = re.sub(r"(?<!\n)\n(?!\n)", " ", text)
+        # Use pre-compiled pattern for better performance
+        text = _SINGLE_NEWLINE_PATTERN.sub(" ", text)
     return text
 
 
@@ -243,14 +253,12 @@ def save_config(config):
 
 
 def calculate_text_length(text):
-    # Ignore chapter markers
+    # Use pre-compiled patterns for better performance
-    text = re.sub(r"<<CHAPTER_MARKER:.*?>>", "", text)
+    # Ignore chapter markers and metadata patterns in a single pass
-    # Ignore metadata patterns
+    text = _CHAPTER_MARKER_PATTERN.sub("", text)
-    text = re.sub(r"<<METADATA_[^:]+:[^>]*>>", "", text)
+    text = _METADATA_PATTERN.sub("", text)
-    # Ignore newlines
+    # Ignore newlines and leading/trailing spaces
-    text = text.replace("\n", "")
+    text = text.replace("\n", "").strip()
-    # Ignore leading/trailing spaces
-    text = text.strip()
     # Calculate character count
     char_count = len(text)
     return char_count

pyproject.toml

@@ -25,7 +25,8 @@ dependencies = [
     "charset_normalizer>=3.4.1",
     "chardet>=5.2.0",
     "static_ffmpeg>=2.13",
-    "Markdown>=3.9"
+    "Markdown>=3.9",
+    "spacy>=3.8.7"
 ]
 
 classifiers = [