feat: Implement segment update and regeneration features in Audiobook API and frontend

qwen3-tts-backend/core/audiobook_service.py

@@ -672,6 +672,118 @@ async def generate_project(project_id: int, user: User, db: Session, chapter_ind
             crud.update_audiobook_project_status(db, project_id, "error", error_message=str(e))
 
 
+async def generate_single_segment(segment_id: int, user: User, db: Session) -> None:
+    from db.models import AudiobookSegment as _Seg
+    seg = db.query(_Seg).filter(_Seg.id == segment_id).first()
+    if not seg:
+        return
+
+    output_base = Path(settings.OUTPUT_DIR) / "audiobook" / str(seg.project_id) / "segments"
+    output_base.mkdir(parents=True, exist_ok=True)
+
+    crud.update_audiobook_segment_status(db, segment_id, "generating")
+    try:
+        from core.tts_service import TTSServiceFactory
+        from core.security import decrypt_api_key
+
+        backend_type = user.user_preferences.get("default_backend", "aliyun") if user.user_preferences else "aliyun"
+        user_api_key = None
+        if backend_type == "aliyun" and user.aliyun_api_key:
+            user_api_key = decrypt_api_key(user.aliyun_api_key)
+
+        backend = await TTSServiceFactory.get_backend(backend_type, user_api_key)
+
+        char = crud.get_audiobook_character(db, seg.character_id)
+        if not char or not char.voice_design_id:
+            crud.update_audiobook_segment_status(db, segment_id, "error")
+            return
+
+        design = crud.get_voice_design(db, char.voice_design_id, user.id)
+        if not design:
+            crud.update_audiobook_segment_status(db, segment_id, "error")
+            return
+
+        await _bootstrap_character_voices([seg], user, backend, backend_type, db)
+        db.refresh(design)
+
+        audio_filename = f"ch{seg.chapter_index:03d}_seg{seg.segment_index:04d}.wav"
+        audio_path = output_base / audio_filename
+
+        if backend_type == "aliyun":
+            if design.aliyun_voice_id:
+                audio_bytes, _ = await backend.generate_voice_design(
+                    {"text": seg.text, "language": "zh"},
+                    saved_voice_id=design.aliyun_voice_id
+                )
+            else:
+                audio_bytes, _ = await backend.generate_voice_design({
+                    "text": seg.text,
+                    "language": "zh",
+                    "instruct": _get_gendered_instruct(char.gender, design.instruct),
+                })
+        elif char.use_indextts2 and design.ref_audio_path and Path(design.ref_audio_path).exists():
+            from core.tts_service import IndexTTS2Backend
+            indextts2 = IndexTTS2Backend()
+            audio_bytes = await indextts2.generate(
+                text=seg.text,
+                spk_audio_prompt=design.ref_audio_path,
+                output_path=str(audio_path),
+                emo_text=seg.emo_text or None,
+                emo_alpha=seg.emo_alpha if seg.emo_text else 0.5,
+            )
+        else:
+            if design.voice_cache_id:
+                from core.cache_manager import VoiceCacheManager
+                cache_manager = await VoiceCacheManager.get_instance()
+                cache_result = await cache_manager.get_cache_by_id(design.voice_cache_id, db)
+                x_vector = cache_result['data'] if cache_result else None
+                if x_vector:
+                    audio_bytes, _ = await backend.generate_voice_clone(
+                        {
+                            "text": seg.text,
+                            "language": "Auto",
+                            "max_new_tokens": 2048,
+                            "temperature": 0.3,
+                            "top_k": 10,
+                            "top_p": 0.9,
+                            "repetition_penalty": 1.05,
+                        },
+                        x_vector=x_vector
+                    )
+                else:
+                    audio_bytes, _ = await backend.generate_voice_design({
+                        "text": seg.text,
+                        "language": "Auto",
+                        "instruct": _get_gendered_instruct(char.gender, design.instruct),
+                        "max_new_tokens": 2048,
+                        "temperature": 0.3,
+                        "top_k": 10,
+                        "top_p": 0.9,
+                        "repetition_penalty": 1.05,
+                    })
+            else:
+                audio_bytes, _ = await backend.generate_voice_design({
+                    "text": seg.text,
+                    "language": "Auto",
+                    "instruct": _get_gendered_instruct(char.gender, design.instruct),
+                    "max_new_tokens": 2048,
+                    "temperature": 0.3,
+                    "top_k": 10,
+                    "top_p": 0.9,
+                    "repetition_penalty": 1.05,
+                })
+
+        with open(audio_path, "wb") as f:
+            f.write(audio_bytes)
+
+        crud.update_audiobook_segment_status(db, segment_id, "done", audio_path=str(audio_path))
+        logger.info(f"Single segment {segment_id} generated: {audio_path}")
+
+    except Exception as e:
+        logger.error(f"Single segment {segment_id} generation failed: {e}", exc_info=True)
+        crud.update_audiobook_segment_status(db, segment_id, "error")
+
+
 def merge_audio_files(audio_paths: list[str], output_path: str) -> None:
     from pydub import AudioSegment