clean: dead code removal + unused imports

- executor: remove dead subtitle_writer stub (lines 121-125)
- executor: replace getattr with direct field access on ResolvedVoice
- service: remove duplicate split_pattern import
- service: remove unused import time
- planner: remove unused import os
- adapters: remove unused import threading, time

abogen/application/conversion_executor.py

@@ -118,13 +118,8 @@ def execute_conversion(
             )
             result.audio_path = audio_path
 
-        # Open subtitle writer for merged output
-        subtitle_writer: Optional[SubtitleWriter] = None
-        if subtitle_writer and audio_sink:
-            # This will be set below when we know subtitle mode
-            pass
-
         # Open subtitle writer if needed
+        subtitle_writer: Optional[SubtitleWriter] = None
         if request.subtitle_mode != "Disabled" and audio_sink:
             subtitle_writer = make_subtitle_writer(
                 audio_path,
@@ -407,8 +402,8 @@ def _resolve_voice(
         return (
             resolved.provider,
             resolved.voice,
-            getattr(resolved, "speed", request.speed),
+            resolved.speed,
-            getattr(resolved, "supertonic_steps", None),
+            resolved.supertonic_steps,
         )
     except Exception:
         # Fallback to base voice
@@ -416,8 +411,8 @@ def _resolve_voice(
         return (
             resolved.provider,
             resolved.voice,
-            getattr(resolved, "speed", request.speed),
+            resolved.speed,
-            getattr(resolved, "supertonic_steps", None),
+            resolved.supertonic_steps,
         )
 
 

abogen/application/conversion_planner.py

@@ -8,7 +8,6 @@ This is Stage 2 of the conversion flow unification plan.
 
 from __future__ import annotations
 
-import os
 from pathlib import Path
 from typing import Any, Dict, List, Optional, Tuple
 

abogen/application/conversion_service.py

@@ -15,7 +15,6 @@ The service NEVER imports from PyQt or WebUI.
 
 from __future__ import annotations
 
-import time
 from typing import Any, Callable, Dict, Optional
 
 from abogen.application.conversion_executor import execute_conversion
@@ -113,7 +112,6 @@ def _prepare_tts_context(
         compile_pronunciation_rules,
         merge_pronunciation_overrides,
     )
-    from abogen.domain.split_pattern import get_split_pattern
 
     # Get runtime normalization settings
     normalization_settings = get_runtime_settings()

abogen/pyqt/conversion_adapter.py

@@ -14,8 +14,6 @@ Subtitle file/timestamp special paths remain in ConversionThread.run() early ret
 from __future__ import annotations
 
 import os
-import threading
-import time
 from pathlib import Path
 from typing import Any, Callable, Dict, List, Optional