v1.1.6

CHANGELOG.md

@@ -1,3 +1,10 @@
+# 1.1.6
+- Improved EPUB chapter detection: Now reliably detects chapters from NAV HTML (TOC) files, even in non-standard EPUBs, fixes the issue mentioned by @jefro108 in #33
+- Fixed SRT subtitle numbering issue, mentioned by @page-muncher in #41
+- Fixed missing chapter contents issue in some EPUB files.
+- Windows installer script now prompts the user to install the CUDA version of PyTorch even if no NVIDIA GPU is detected.
+- Abogen now includes Mandarin Chinese (misaki[zh]) by default; manual installation is no longer required.
+
 # 1.1.5
 - Changed the temporary directory path to user's cache directory, which is more appropriate for storing cache files and avoids issues with unintended cleanup.
 - Fixed the isssue where extra metadata information was not being saved to M4B files when they have no chapters, ensuring that all metadata is correctly written to the output file.
@@ -23,7 +30,7 @@
 - Added a new option: `Reset to default settings`, allowing users to reset all settings to their default values.
 - Added a new option: `Disable Kokoro's internet access`. This lets you prevent Kokoro from downloading models or voices from HuggingFace Hub, which can help avoid long waiting times if your computer is offline.
 - HuggingFace Hub telemetry is now disabled by default for improved privacy. (HuggingFace Hub is used by Kokoro to download its models)
-- Potential fix for #37 and #38, where the program was becoming slow while processing large files.
+- cPotential fix for #37 and #38, where the program was becoming slow while processing large files.
 - Fixed `Open folder` and `Open file` buttons in the queue manager GUI.
 - Improvements in code structure.
 

WINDOWS_INSTALL.bat

@@ -1,5 +1,5 @@
 @echo off
-setlocal
+setlocal EnableDelayedExpansion
 cd /d "%~dp0"
 
 :: Set misaki language
@@ -23,7 +23,6 @@ set CUDA_VERSION=128
 for /f "delims=: tokens=*" %%A in ('findstr /b ::: "%~f0"') do @echo(%%A
 
 set CURRENT_DIR="%CD%"
-setlocal enabledelayedexpansion
 set NAME=abogen
 set PROJECTFOLDER=abogen
 set RUN=python_embedded\Scripts\abogen.exe
@@ -276,6 +275,7 @@ if "%MISAKI_LANG%" NEQ "en" (
 for /f %%i in ('%PYTHON_CONSOLE_PATH% -c "from abogen.is_nvidia import check; print(check())"') do set IS_NVIDIA=%%i
 
 :: Check if torch is installed with CUDA support
+echo.
 echo Checking CUDA availability...
 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
@@ -293,7 +293,26 @@ if /I "%IS_NVIDIA%"=="true" (
         echo CUDA is available on NVIDIA GPU.
     )
 ) else (
-    echo GPU is not NVIDIA. Skipping PyTorch CUDA installation.
+    echo.
+    echo Unable to detect an NVIDIA GPU in your system.
+    echo.
+    echo Do you want to install PyTorch anyway?
+    echo.
+    echo If you DO have an NVIDIA GPU, please press Y.
+    echo If you DO NOT have an NVIDIA GPU, please press N.
+    echo.
+    choice /C YN /M "Y=Yes, N=No"
+    if errorlevel 2 (
+        echo Skipping PyTorch installation.
+    ) else (
+        echo Installing PyTorch with CUDA %CUDA_VERSION% support...
+        %PYTHON_CONSOLE_PATH% -m pip install torch torchvision torchaudio --extra-index-url https://download.pytorch.org/whl/cu%CUDA_VERSION% --no-warn-script-location
+        if errorlevel 1 (
+            echo Failed to install PyTorch.
+            pause
+            exit /b
+        )
+    )
 )
 
 :: Ask user if they want to create a desktop shortcut

abogen/VERSION

@@ -1 +1 @@
-1.1.5
+1.1.6

abogen/book_handler.py

@@ -324,14 +324,31 @@ class HandlerDialog(QDialog):
                     f"Found NCX item via ITEM_NAVIGATION: {ncx_in_nav.get_name()}"
                 )
 
-        # 3. If still no nav_item, check for ITEM_NCX directly
+        # 3. If still no nav_item, check for NCX or fallback to NAV HTML in all ITEM_DOCUMENTs
-        if not nav_item:
+        ncx_constant = getattr(epub, "ITEM_NCX", None)
-            ncx_items = list(self.book.get_items_of_type(ebooklib.ITEM_NCX))
+        if not nav_item and ncx_constant is not None:
+            ncx_items = list(self.book.get_items_of_type(ncx_constant))
             if ncx_items:
-                nav_item = ncx_items[0]  # Take the first one
+                nav_item = ncx_items[0]
                 nav_type = "ncx"
-                logging.info(f"Found NCX item via ITEM_NCX: {ncx_items[0].get_name()}")
+                logging.info(f"Found NCX item via ITEM_NCX: {nav_item.get_name()}")
-
+        # Fallback: search all ITEM_DOCUMENTs for a NAV HTML with <nav epub:type="toc">
+        if not nav_item:
+            for item in self.book.get_items_of_type(ebooklib.ITEM_DOCUMENT):
+                try:
+                    html_content = item.get_content().decode("utf-8", errors="ignore")
+                    if "<nav" in html_content and 'epub:type="toc"' in html_content:
+                        soup = BeautifulSoup(html_content, "html.parser")
+                        nav_tag = soup.find("nav", attrs={"epub:type": "toc"})
+                        if nav_tag:
+                            nav_item = item
+                            nav_type = "html"
+                            logging.info(
+                                f"Found NAV HTML with TOC in: {item.get_name()}"
+                            )
+                            break
+                except Exception as e:
+                    continue
         # 4. If no navigation item found by any method, trigger fallback
         if not nav_item or not nav_type:
             logging.warning(
@@ -516,10 +533,14 @@ class HandlerDialog(QDialog):
                         slice_html += self.doc_content.get(intermediate_doc_href, "")
                 except Exception:
                     pass
-
+            # Fallback: if slice_html is empty, try to get the whole file's text
+            if not slice_html.strip() and current_doc_html:
+                logging.warning(
+                    f"No content found for src '{current_src}', using full file as fallback."
+                )
+                slice_html = current_doc_html
             if slice_html.strip():
                 slice_soup = BeautifulSoup(slice_html, "html.parser")
-                # Add double newlines after <p> and <div> tags
                 for tag in slice_soup.find_all(["p", "div"]):
                     tag.append("\n\n")
                 for tag in slice_soup.find_all(["sup", "sub"]):
@@ -581,6 +602,23 @@ class HandlerDialog(QDialog):
             f"Finished processing EPUB navigation. Found {len(self.content_texts)} content sections linked to TOC."
         )
 
+    def _find_doc_key(self, base_href, doc_order, doc_order_decoded):
+        """Find the best matching doc_key for a given base_href using robust matching."""
+        candidates = [
+            base_href,
+            urllib.parse.unquote(base_href),
+        ]
+        base_name = os.path.basename(base_href).lower()
+        for k in list(doc_order.keys()) + list(doc_order_decoded.keys()):
+            if os.path.basename(k).lower() == base_name:
+                candidates.append(k)
+        for candidate in candidates:
+            if candidate in doc_order:
+                return candidate, doc_order[candidate]
+            elif candidate in doc_order_decoded:
+                return candidate, doc_order_decoded[candidate]
+        return None, None
+
     def _parse_ncx_navpoint(
         self,
         nav_point,
@@ -603,27 +641,15 @@ class HandlerDialog(QDialog):
 
         if src:
             base_href, fragment = src.split("#", 1) if "#" in src else (src, None)
-            # Try both original and decoded hrefs
+            doc_key, doc_idx = self._find_doc_key(
-            doc_key = None
+                base_href, doc_order, doc_order_decoded
-            if base_href in doc_order:
+            )
-                doc_key = base_href
+            if not doc_key:
-                doc_idx = doc_order[base_href]
-            elif urllib.parse.unquote(base_href) in doc_order:
-                doc_key = urllib.parse.unquote(base_href)
-                doc_idx = doc_order[doc_key]
-            elif base_href in doc_order_decoded:
-                doc_key = base_href
-                doc_idx = doc_order_decoded[base_href]
-            elif urllib.parse.unquote(base_href) in doc_order_decoded:
-                doc_key = urllib.parse.unquote(base_href)
-                doc_idx = doc_order_decoded[doc_key]
-            else:
                 logging.warning(
-                    f"Navigation entry '{title}' points to '{base_href}', which is not in the spine or document list."
+                    f"Navigation entry '{title}' points to '{base_href}', which is not in the spine or document list (even after basename fallback)."
                 )
                 current_entry_node["has_content"] = False
-                doc_key = None
+            else:
-            if doc_key is not None:
                 position = find_position_func(doc_key, fragment)
                 entry_data = {
                     "src": src,
@@ -698,28 +724,15 @@ class HandlerDialog(QDialog):
         current_entry_node["title"] = title
         current_entry_node["src"] = src
 
+        doc_key = None
+        doc_idx = None
+        position = 0
+        fragment = None
         if src:
             base_href, fragment = src.split("#", 1) if "#" in src else (src, None)
-            # Try both original and decoded hrefs
+            doc_key, doc_idx = self._find_doc_key(
-            doc_key = None
+                base_href, doc_order, doc_order_decoded
-            if base_href in doc_order:
+            )
-                doc_key = base_href
-                doc_idx = doc_order[base_href]
-            elif urllib.parse.unquote(base_href) in doc_order:
-                doc_key = urllib.parse.unquote(base_href)
-                doc_idx = doc_order[doc_key]
-            elif base_href in doc_order_decoded:
-                doc_key = base_href
-                doc_idx = doc_order_decoded[base_href]
-            elif urllib.parse.unquote(base_href) in doc_order_decoded:
-                doc_key = urllib.parse.unquote(base_href)
-                doc_idx = doc_order_decoded[doc_key]
-            else:
-                logging.warning(
-                    f"Navigation entry '{title}' points to '{base_href}', which is not in the spine or document list."
-                )
-                current_entry_node["has_content"] = False
-                doc_key = None
             if doc_key is not None:
                 position = find_position_func(doc_key, fragment)
                 entry_data = {
@@ -731,13 +744,16 @@ class HandlerDialog(QDialog):
                 }
                 ordered_entries.append(entry_data)
                 current_entry_node["has_content"] = True
+            else:
+                logging.warning(
+                    f"Navigation entry '{title}' points to '{base_href}', which is not in the spine or document list (even after basename fallback)."
+                )
+                current_entry_node["has_content"] = False
         else:
             current_entry_node["has_content"] = False
 
-        child_ol = li_element.find("ol", recursive=False)
+        for child_ol in li_element.find_all("ol", recursive=False):
-        if child_ol:
             for child_li in child_ol.find_all("li", recursive=False):
-                # Pass find_position_func down recursively
                 self._parse_html_nav_li(
                     child_li,
                     ordered_entries,
@@ -746,12 +762,7 @@ class HandlerDialog(QDialog):
                     current_entry_node["children"],
                     find_position_func,
                 )
-
+        tree_structure_list.append(current_entry_node)
-        if title and (
-            current_entry_node.get("has_content", False)
-            or current_entry_node["children"]
-        ):
-            tree_structure_list.append(current_entry_node)
 
     def _find_position_robust(self, doc_href, fragment_id):
         if doc_href not in self.doc_content:

abogen/conversion.py

@@ -434,8 +434,8 @@ class ConversionThread(QThread):
                 )
                 # Only check for files with allowed extensions (extension without dot, case-insensitive)
                 clash = any(
-                    os.path.splitext(fname)[0] == f"{base_name}{suffix}" and
+                    os.path.splitext(fname)[0] == f"{base_name}{suffix}"
-                    os.path.splitext(fname)[1][1:].lower() in allowed_exts
+                    and os.path.splitext(fname)[1][1:].lower() in allowed_exts
                     for fname in os.listdir(parent_dir)
                 )
                 if not os.path.exists(chapters_out_dir_candidate) and not clash:
@@ -467,6 +467,8 @@ class ConversionThread(QThread):
                     {"chapter": chapter[0], "start": 0.0, "end": 0.0}
                     for chapter in chapters
                 ]
+                # SRT numbering fix: use a global counter
+                merged_srt_index = 1  # SRT numbering for merged file
                 # Prepare output file/ffmpeg process for merged output
                 if self.output_format in ["wav", "mp3", "flac"]:
                     merged_out_file = sf.SoundFile(
@@ -483,8 +485,8 @@ class ConversionThread(QThread):
                     merged_out_file = None
                     ffmpeg_proc = None
                     metadata_options = (
-                            self._extract_and_add_metadata_tags_to_ffmpeg_cmd()
+                        self._extract_and_add_metadata_tags_to_ffmpeg_cmd()
-                        )
+                    )
                     # Prepare ffmpeg command for m4b output
                     cmd = [
                         "ffmpeg",
@@ -599,6 +601,7 @@ class ConversionThread(QThread):
                     {"chapter": chapter[0], "start": 0.0, "end": 0.0}
                     for chapter in chapters
                 ]
+                srt_index = 1  # SRT numbering fix for chapter-only mode
             # Instead of processing the whole text, process by chapter
             for chapter_idx, (chapter_name, chapter_text) in enumerate(chapters, 1):
                 chapter_out_path = None
@@ -677,6 +680,9 @@ class ConversionThread(QThread):
                         continue
                     # Open chapter subtitle file for incremental writing if needed
                     chapter_subtitle_file = None
+                    chapter_srt_index = (
+                        1  # Initialize SRT numbering for this chapter file
+                    )
                     if self.subtitle_mode != "Disabled":
                         subtitle_format = getattr(self, "subtitle_format", "srt")
                         file_extension = "ass" if "ass" in subtitle_format else "srt"
@@ -821,12 +827,12 @@ class ConversionThread(QThread):
                                             f"Dialogue: 0,{start_time},{end_time},Default,,{merged_subtitle_margin},{merged_subtitle_margin},0,,{merged_subtitle_alignment_tag}{text}\n"
                                         )
                                 else:
-                                    for i, (start, end, text) in enumerate(
+                                    for entry in new_entries:
-                                        new_entries, 1
+                                        start, end, text = entry
-                                    ):
                                         merged_subtitle_file.write(
-                                            f"{i}\n{self._srt_time(start)} --> {self._srt_time(end)}\n{text}\n\n"
+                                            f"{merged_srt_index}\n{self._srt_time(start)} --> {self._srt_time(end)}\n{text}\n\n"
                                         )
+                                        merged_srt_index += 1
                         # Per-chapter subtitle processing for both file and ffmpeg_proc
                         if chapter_out_file or chapter_ffmpeg_proc:
                             new_chapter_entries = []
@@ -848,12 +854,12 @@ class ConversionThread(QThread):
                                             f"Dialogue: 0,{start_time},{end_time},Default,,{chapter_subtitle_margin},{chapter_subtitle_margin},0,,{chapter_subtitle_alignment_tag}{text}\n"
                                         )
                                 else:
-                                    for i, (start, end, text) in enumerate(
+                                    for entry in new_chapter_entries:
-                                        new_chapter_entries, 1
+                                        start, end, text = entry
-                                    ):
                                         chapter_subtitle_file.write(
-                                            f"{i}\n{self._srt_time(start)} --> {self._srt_time(end)}\n{text}\n\n"
+                                            f"{chapter_srt_index}\n{self._srt_time(start)} --> {self._srt_time(end)}\n{text}\n\n"
                                         )
+                                        chapter_srt_index += 1
                     if merge_chapters_at_end:
                         current_time += chunk_dur
                         if chapter_out_file or chapter_ffmpeg_proc:

abogen/main.py

@@ -32,11 +32,13 @@ os.environ["MIOPEN_CONV_PRECISE_ROCM_TUNING"] = "0"
 # Reset sleep states
 atexit.register(prevent_sleep_end)
 
+
 # Also handle signals (Ctrl+C, kill, etc.)
 def _cleanup_sleep(signum, frame):
     prevent_sleep_end()
     sys.exit(0)
 
+
 signal.signal(signal.SIGINT, _cleanup_sleep)
 signal.signal(signal.SIGTERM, _cleanup_sleep)
 
@@ -50,6 +52,7 @@ if sys.stderr is None:
 if platform.system() == "Darwin" and platform.processor() == "arm":
     os.environ["PYTORCH_ENABLE_MPS_FALLBACK"] = "1"
 
+
 # Custom message handler to filter out specific Qt warnings
 def qt_message_handler(mode, context, message):
     if "Wayland does not support QWindow::requestActivate()" in message:

abogen/utils.py

@@ -74,24 +74,32 @@ def get_user_config_path():
         if os.path.exists(custom_dir):
             config_dir = custom_dir
         else:
-            config_dir = user_config_dir("abogen", appauthor=False, roaming=True, ensure_exists=True)
+            config_dir = user_config_dir(
+                "abogen", appauthor=False, roaming=True, ensure_exists=True
+            )
     else:
         # Windows and fallback case
-        config_dir = user_config_dir("abogen", appauthor=False, roaming=True, ensure_exists=True)
+        config_dir = user_config_dir(
+            "abogen", appauthor=False, roaming=True, ensure_exists=True
+        )
 
     return os.path.join(config_dir, "config.json")
 
+
 # Define cache path
 def get_user_cache_path(folder=None):
     from platformdirs import user_cache_dir
 
-    cache_dir = user_cache_dir("abogen", appauthor=False, opinion=True, ensure_exists=True)
+    cache_dir = user_cache_dir(
+        "abogen", appauthor=False, opinion=True, ensure_exists=True
+    )
     if folder:
         cache_dir = os.path.join(cache_dir, folder)
         # Ensure the directory exists
         os.makedirs(cache_dir, exist_ok=True)
     return cache_dir
 
+
 _sleep_procs = {"Darwin": None, "Linux": None}  # Store sleep prevention processes
 
 

pyproject.toml

@@ -15,6 +15,7 @@ keywords = ["audiobook", "epub", "pdf", "text-to-speech", "subtitle", "tts", "ko
 dependencies = [
     "PyQt5>=5.15.11",
     "kokoro>=0.9.4",
+    "misaki[zh]>=0.9.4",
     "ebooklib>=0.19",
     "beautifulsoup4>=4.13.4",
     "PyMuPDF>=1.25.5",