remove: --chunk flag (always force client-side chunking)

3 files changed, 29 insertions, 111 deletions

diff --git a/app/converter/config.py b/app/converter/config.py
index 6a98136..5e35ee3 100644
--- a/app/converter/config.py
+++ b/app/converter/config.py
@@ -8,9 +8,6 @@ MAX_RETRIES = 3    # Attempts per chunk request
 HEARTBEAT_INTERVAL_SECONDS = 30  # Print "still working" in console logs every N seconds
 
 # Words per TTS generation request (client-side chunking).
-# The qwen and faster backends always chunk with this size
-# The audio.cpp backend chunks long text itself, so this is ignored
-# by default with that backend. Force chunking with --chunk
 CHUNK_SIZE = 250
 
 # Default TTS backend.
diff --git a/app/converter/converter.py b/app/converter/converter.py
index cef4808..4c1ffad 100644
--- a/app/converter/converter.py
+++ b/app/converter/converter.py
@@ -140,7 +140,7 @@ class AudiobookConverter:
                  speed: float = 1.0, single_file: bool = False, output_format: str = config.AUDIO_FORMAT,
                  language: Optional[str] = None, backend: str = config.BACKEND,
                  voice: Optional[str] = None, debug: bool = False,
-                 chunk: bool = False, model_id: Optional[str] = None,
+                 model_id: Optional[str] = None,
                  instructions: Optional[str] = None,
                  request_options: Optional[Dict[str, str]] = None):
         if speed <= 0:
@@ -162,12 +162,6 @@ class AudiobookConverter:
         self.backend = backend
         self.voice = voice
         self.debug = bool(debug)
-        # Client-side chunking: the qwen and faster backends always chunk
-        # (their servers do one generation per request and silently truncate
-        # long text). The audio.cpp server chunks long text itself, so it
-        # defaults to one request per chapter; --chunk forces client-side
-        # chunking on top (possible needless double-chunking).
-        self.client_chunks = bool(chunk) or backend != BACKEND_AUDIOCPP
         # Voice design / style instruction and free-form request options
         # (audio.cpp only): forwarded to AudioCppTTSClient, which validates
         # them against the server-hosted model at connect time.
@@ -185,7 +179,6 @@ class AudiobookConverter:
             # instructions describe or style the voice, request_options pass
             # per-model controls through to the server.
             self.tts = AudioCppTTSClient(voice=voice, language=self.language,
-                                         chunk_text=self.client_chunks,
                                          model_id=model_id,
                                          instructions=instructions,
                                          request_options=self.request_options)
@@ -437,10 +430,9 @@ class AudiobookConverter:
         and every request/response is logged.
         """
         total_chunks = len(chunks)
-        if self.client_chunks:
-            print(f"\n{'=' * 50}")
-            print(f"PROCESSING {total_chunks} CHUNKS")
-            print(f"{'=' * 50}")
+        print(f"\n{'=' * 50}")
+        print(f"PROCESSING {total_chunks} CHUNKS")
+        print(f"{'=' * 50}")
 
         results: Dict[int, Optional[Path]] = {}
         for chunk_num, chunk_text in enumerate(chunks, 1):
@@ -461,8 +453,7 @@ class AudiobookConverter:
                         destination = f" -> {copied.name}" if copied else ""
                         logger.debug("Chunk %d/%d response in %.1fs%s",
                                      chunk_num, total_chunks, elapsed, destination)
-                    if self.client_chunks:
-                        print(f"[OK] Chunk {chunk_num:3d}/{total_chunks} completed")
+                    print(f"[OK] Chunk {chunk_num:3d}/{total_chunks} completed")
                     logger.info("+ Chunk %d/%d completed", chunk_num, total_chunks)
                 else:
                     logger.error("Chunk %d/%d failed; aborting the remaining chunks",
@@ -476,25 +467,16 @@ class AudiobookConverter:
                 break
 
         successful_chunks = sum(1 for path in results.values() if path)
-        if self.client_chunks:
-            print(f"\n{'=' * 50}")
-            print("CHUNK PROCESSING COMPLETE")
-            print(f"Successful: {successful_chunks}/{total_chunks}")
-            print(f"{'=' * 50}")
+        print(f"\n{'=' * 50}")
+        print("CHUNK PROCESSING COMPLETE")
+        print(f"Successful: {successful_chunks}/{total_chunks}")
+        print(f"{'=' * 50}")
         logger.info("Chunk processing completed: %d/%d chunks", successful_chunks, total_chunks)
         return results
 
     def _chapter_chunks(self, text: str) -> List[str]:
-        """Split chapter text into TTS requests.
-
-        Client-side chunking splits into CHUNK_SIZE-word chunks (qwen and
-        faster always; audio.cpp only with --chunk). Otherwise (audio.cpp
-        default) the whole text is one request and the server does its own
-        long-form chunking.
-        """
-        if self.client_chunks:
-            return chunking.split_into_chunks(text)
-        return [text] if text.strip() else []
+        """Split chapter text into CHUNK_SIZE-word TTS requests."""
+        return chunking.split_into_chunks(text)
 
     def _convert_text(self, text: str, output_path: Path, start_time: float,
                       speed: Optional[float] = None,
@@ -530,31 +512,14 @@ class AudiobookConverter:
 
             chunk_sizes = [len(chunk.split()) for chunk in chunks]
             avg_chunk_size = sum(chunk_sizes) / len(chunk_sizes)
-            if len(chunks) == 1:
-                logger.info("Sending the whole text as one request (%d words; "
-                            "the server chunks long text itself)",
-                            chunk_sizes[0])
-            else:
-                logger.info("Split into %d chunks (avg %.0f words per chunk)",
-                            total_chunks, avg_chunk_size)
+            logger.info("Split into %d chunks (avg %.0f words per chunk)",
+                        total_chunks, avg_chunk_size)
             backend_labels = {
                 BACKEND_FASTER: "faster TTS API",
                 BACKEND_AUDIOCPP: "audio.cpp server",
             }
             backend = backend_labels.get(self.backend, "Qwen API")
-            if self.client_chunks:
-                print(f"[INFO] Processing {total_chunks} chunks via {backend}...")
-            else:
-                # The whole request is sent at once and the server does its
-                # own long-form chunking, so the chunk vocabulary does not
-                # apply; warn that this one request can take a very long time.
-                subject = (f"chapter {chapter[0]}/{chapter[1]}"
-                           if chapter is not None else "text")
-                print(f"[INFO] Sending the {subject} to the {backend} as a "
-                      "single request...")
-                print("[NOTE] It is expected for this to take a very long "
-                      "time: the server synthesizes the entire request before "
-                      "returning any audio.")
+            print(f"[INFO] Processing {total_chunks} chunks via {backend}...")
 
             results = self._synthesize_chunks(chunks, debug_dir=debug_dir)
             successful_chunks = sum(1 for path in results.values() if path)
@@ -578,11 +543,8 @@ class AudiobookConverter:
                     logger.info("Chapter %d/%d converted in %dm %ds (%d/%d chunks)",
                                 chapter[0], chapter[1], minutes, seconds,
                                 successful_chunks, total_chunks)
-                    if self.client_chunks:
-                        print(f"[INFO] Chapter {chapter[0]}/{chapter[1]} converted "
-                              f"({successful_chunks}/{total_chunks} chunks)")
-                    else:
-                        print(f"[INFO] Chapter {chapter[0]}/{chapter[1]} converted")
+                    print(f"[INFO] Chapter {chapter[0]}/{chapter[1]} converted "
+                          f"({successful_chunks}/{total_chunks} chunks)")
                 else:
                     logger.info("Conversion completed in %dm %ds: %s", minutes, seconds, output_path)
             else:
@@ -621,12 +583,6 @@ class AudiobookConverter:
                 print(f"Speaker: {config.SPEAKER}")
             if self.request_options:
                 print(f"Request options: {self.request_options}")
-            if self.client_chunks:
-                print("Chunking: client-side (--chunk; the server also chunks "
-                      "long text itself, so this may double-chunk)")
-            else:
-                print("Chunking: server-side (one request per chapter; "
-                      "--chunk forces client-side chunking)")
             print(f"Language: {self.language}")
         else:
             api_url = (config.CLONE_API_URL if self.voice_mode == VOICE_MODE_CLONE
diff --git a/app/converter/tts.py b/app/converter/tts.py
index 6ccef56..842cc0c 100644
--- a/app/converter/tts.py
+++ b/app/converter/tts.py
@@ -278,11 +278,6 @@ def whisper_backend_available() -> Optional[str]:
     return None
 
 
-# 150 wpm is a typical spoken pace; used only to size the HTTP request
-# timeout for long audio.cpp generations (not as a correctness check).
-_ESTIMATED_WORDS_PER_MINUTE = 150
-
-
 class _BaseTTSClient:
     """Shared chunk retry logic, heartbeat, and chunk file bookkeeping."""
 
@@ -327,15 +322,10 @@ class _BaseTTSClient:
         return None
 
     @contextlib.contextmanager
-    def _chunk_heartbeat(self, chunk_num: int, label: Optional[str] = None):
-        """Print a periodic "still working" message while a request generates.
-
-        ``label`` overrides the default "Chunk {chunk_num}" subject, for
-        backends that send one request per chapter without client-side
-        chunking (the audio.cpp default) where "chunk" would be misleading.
-        """
+    def _chunk_heartbeat(self, chunk_num: int):
+        """Print a periodic "still working" message while a request generates."""
         stop = threading.Event()
-        subject = label if label is not None else f"Chunk {chunk_num}"
+        subject = f"Chunk {chunk_num}"
 
         def _beat():
             start = time.time()
@@ -796,20 +786,14 @@ class AudioCppTTSClient(_BaseTTSClient):
     the request's "options" object, which is the server's generic
     pass-through for per-model controls.
 
-    Chunking: the server does its own long-form text chunking for every
-    family (its ``text_chunk_size`` option, with a per-family default), so
-    by default each chapter is sent as a single request and the audio
-    comes back already stitched. With ``chunk_text=True`` (the --chunk CLI
-    flag), text is instead split client-side into CHUNK_SIZE-word
-    sub-requests, which may needlessly double-chunk — the warning is
-    printed by the CLI.
-
-    Each response is a complete WAV file, so sub-request audio is
-    concatenated with the same lossless path used for the Qwen client.
+    Chunking: text is split client-side into sub-requests of at most
+    config.CHUNK_SIZE words each; each sub-request returns a complete
+    WAV file and the parts are concatenated with the same lossless path
+    used for the Qwen client.
     """
 
     def __init__(self, voice: Optional[str] = None, language: Optional[str] = None,
-                 api_url: Optional[str] = None, chunk_text: bool = False,
+                 api_url: Optional[str] = None,
                  model_id: Optional[str] = None,
                  instructions: Optional[str] = None,
                  request_options: Optional[Dict[str, str]] = None):
@@ -844,10 +828,6 @@ class AudioCppTTSClient(_BaseTTSClient):
         # speakers gets its voice from the instruction alone (no voice field).
         self.design_mode = False
         self.instruction_voice = False
-        # When False (default), each chapter is sent as one request and the
-        # server does its own long-form chunking (text_chunk_size); when True,
-        # text is split client-side into CHUNK_SIZE-word sub-requests first.
-        self.chunk_text = bool(chunk_text)
         # Family and task of the selected model entry and the family's request
         # profile; all are resolved from GET /v1/models during _connect.
         self.family = ""
@@ -1217,11 +1197,6 @@ class AudioCppTTSClient(_BaseTTSClient):
             url, data=json.dumps(payload).encode("utf-8"),
             headers={"Content-Type": "application/json"}, method="POST")
         timeout = config.API_TIMEOUT
-        if not self.chunk_text:
-            # Estimated audio duration at 150 wpm, doubled plus a minute of
-            # slack, bounded below by the configured per-request timeout.
-            estimated_seconds = 60.0 * len(text.split()) / _ESTIMATED_WORDS_PER_MINUTE
-            timeout = max(timeout, int(estimated_seconds * 2) + 60)
         try:
             with urllib.request.urlopen(request, timeout=timeout) as response:
                 wav = response.read()
@@ -1259,28 +1234,18 @@ class AudioCppTTSClient(_BaseTTSClient):
     def generate_chunk(self, text: str, chunk_num: int) -> Optional[str]:
         """Generate one audio chunk; returns its path in the chunks folder.
 
-        By default the whole text goes out as a single request and the
-        server does its own long-form chunking (see the class docstring).
-        With ``chunk_text=True`` (--chunk), the text is split into
-        sub-requests of at most ``config.CHUNK_SIZE`` words each; each
-        sub-request returns a complete WAV file and the parts are
-        concatenated into one chunk file.
+        The text is split into sub-requests of at most ``config.CHUNK_SIZE``
+        words each; each sub-request returns a complete WAV file and the
+        parts are concatenated into one chunk file.
         """
         try:
-            if self.chunk_text:
-                sub_texts = split_into_chunks(text, max_words=config.CHUNK_SIZE)
-            elif text.strip():
-                sub_texts = [text]
-            else:
-                sub_texts = []
+            sub_texts = split_into_chunks(text, max_words=config.CHUNK_SIZE)
             if not sub_texts:
                 raise RuntimeError("No text to synthesize")
 
             output_path: Optional[Path] = None
             with tempfile.TemporaryDirectory(prefix="tts_parts_") as parts_dir, \
-                    self._chunk_heartbeat(
-                        chunk_num,
-                        label=None if self.chunk_text else "Request"):
+                    self._chunk_heartbeat(chunk_num):
                 part_paths = []
                 for sub_num, sub_text in enumerate(sub_texts, 1):
                     wav = self._request_wav_with_retry(