refactor: voice resolution on ConversionRequest, PipelinePool without job param

- Added speakers field to ConversionRequest
- Rewrote collect_required_voice_ids(), initialize_voice_cache(),
  job_voice_fallback(), chapter_voice_spec(), chunk_voice_spec()
  to accept ConversionRequest instead of job
- PipelinePool.get() now takes request= instead of job=
- Updated all tests to use ConversionRequest interface
- 1493 tests passing

abogen/application/conversion_request.py

@@ -95,6 +95,9 @@ class ConversionRequest:
     # --- Metadata ---
     metadata_tags: Dict[str, Any] = field(default_factory=dict)
 
+    # --- Voice profiles (loaded by UI, used by app for voice resolution) ---
+    speakers: Dict[str, Any] = field(default_factory=dict)
+
     # --- Artifacts ---
     cover_image_path: Optional[Path] = None
     cover_image_mime: Optional[str] = None

abogen/domain/pipeline_factory.py

@@ -95,7 +95,8 @@ class PipelinePool:
         language: str,
         use_gpu: bool,
         *,
-        job: Any = None,
+        request: Any = None,
+        events: Any = None,
     ) -> Any:
         """Get or create a cached pipeline for the given provider.
 
@@ -103,7 +104,8 @@ class PipelinePool:
             provider: TTS provider name ("kokoro" or "supertonic").
             language: Language code (for kokoro).
             use_gpu: Whether GPU acceleration is requested.
-            job: Optional job object for voice cache initialization.
+            request: ConversionRequest for voice cache initialization.
+            events: ConversionEvents for logging during cache init.
         """
         provider = str(provider or "kokoro").strip().lower() or "kokoro"
         if not is_plugin_registered(provider):
@@ -116,8 +118,8 @@ class PipelinePool:
         pipeline = create_pipeline_for_job(provider, language, use_gpu)
         self._pipelines[provider] = pipeline
 
-        if provider == "kokoro" and not self._voice_cache_initialized and job is not None:
+        if provider == "kokoro" and not self._voice_cache_initialized and request is not None:
-            initialize_voice_cache(job)
+            initialize_voice_cache(request, events=events)
             self._voice_cache_initialized = True
 
         return pipeline

abogen/domain/voice_resolution.py

@@ -2,6 +2,9 @@
 
 Functions for resolving voice specifications, collecting required voice IDs,
 and determining the voice to use for chapters and chunks.
+
+All functions accept ConversionRequest (the app-layer contract) instead of
+UI-specific objects. This keeps the domain layer UI-agnostic.
 """
 
 from __future__ import annotations
@@ -29,12 +32,28 @@ def spec_to_voice_ids(spec: Any) -> Set[str]:
     return set()
 
 
-def job_voice_fallback(job: Any) -> str:
+def _get_chapter_overrides(request: Any) -> list:
-    base = str(getattr(job, "voice", "") or "").strip()
+    """Extract chapter overrides from ConversionRequest."""
+    cc = getattr(request, "chapter_chunk", None)
+    if cc is not None:
+        return getattr(cc, "chapter_overrides", []) or []
+    return []
+
+
+def _get_chunks(request: Any) -> list:
+    """Extract chunks from ConversionRequest."""
+    cc = getattr(request, "chapter_chunk", None)
+    if cc is not None:
+        return getattr(cc, "chunks", []) or []
+    return []
+
+
+def job_voice_fallback(request: Any) -> str:
+    base = str(getattr(request, "voice", "") or "").strip()
     if base and base != "__custom_mix":
         return base
 
-    speakers = getattr(job, "speakers", None)
+    speakers = getattr(request, "speakers", None)
     if isinstance(speakers, dict):
         narrator = speakers.get("narrator")
         if isinstance(narrator, dict):
@@ -52,7 +71,7 @@ def job_voice_fallback(job: Any) -> str:
                 if candidate and candidate != "__custom_mix":
                     return candidate
 
-    for chapter in getattr(job, "chapters", []) or []:
+    for chapter in _get_chapter_overrides(request):
         if not isinstance(chapter, dict):
             continue
         for key in ("resolved_voice", "voice_formula", "voice"):
@@ -63,24 +82,24 @@ def job_voice_fallback(job: Any) -> str:
     return ""
 
 
-def collect_required_voice_ids(job: Any) -> Set[str]:
+def collect_required_voice_ids(request: Any) -> Set[str]:
     voices: Set[str] = set()
-    voices.update(spec_to_voice_ids(job.voice))
+    voices.update(spec_to_voice_ids(request.voice))
-    voices.update(spec_to_voice_ids(job_voice_fallback(job)))
+    voices.update(spec_to_voice_ids(job_voice_fallback(request)))
 
-    for chapter in getattr(job, "chapters", []) or []:
+    for chapter in _get_chapter_overrides(request):
         if not isinstance(chapter, dict):
             continue
         for key in ("resolved_voice", "voice_formula", "voice"):
             voices.update(spec_to_voice_ids(chapter.get(key)))
 
-    for chunk in getattr(job, "chunks", []) or []:
+    for chunk in _get_chunks(request):
         if not isinstance(chunk, dict):
             continue
         for key in ("resolved_voice", "voice_formula", "voice"):
             voices.update(spec_to_voice_ids(chunk.get(key)))
 
-    speakers = getattr(job, "speakers", {})
+    speakers = getattr(request, "speakers", {})
     if isinstance(speakers, dict):
         for payload in speakers.values() or []:
             if not isinstance(payload, dict):
@@ -92,30 +111,38 @@ def collect_required_voice_ids(job: Any) -> Set[str]:
     return voices
 
 
-def initialize_voice_cache(job: Any) -> None:
+def initialize_voice_cache(request: Any, events: Any = None) -> None:
+    """Initialize voice cache by downloading required voice assets.
+
+    Args:
+        request: ConversionRequest with voice/chapter/chunk/speaker info.
+        events: ConversionEvents for logging (optional, for backward compat).
+    """
+    log = (lambda msg, level="info": events.log(msg, level=level)) if events else (lambda msg, level="info": None)
+
     try:
-        targets = collect_required_voice_ids(job)
+        targets = collect_required_voice_ids(request)
         downloaded, errors = ensure_voice_assets(
             targets,
-            on_progress=lambda message: job.add_log(message, level="debug"),
+            on_progress=lambda message: log(message, level="debug"),
         )
     except RuntimeError as exc:
-        job.add_log(f"Voice cache unavailable: {exc}", level="warning")
+        log(f"Voice cache unavailable: {exc}", level="warning")
         return
 
     if downloaded:
-        job.add_log(
+        log(
             f"Cached {len(downloaded)} voice asset{'s' if len(downloaded) != 1 else ''} locally.",
             level="info",
         )
 
     for voice_id, error in errors.items():
-        job.add_log(f"Failed to cache voice '{voice_id}': {error}", level="warning")
+        log(f"Failed to cache voice '{voice_id}': {error}", level="warning")
 
 
-def chapter_voice_spec(job: Any, override: Optional[Dict[str, Any]]) -> str:
+def chapter_voice_spec(request: Any, override: Optional[Dict[str, Any]]) -> str:
     if not override:
-        return job_voice_fallback(job)
+        return job_voice_fallback(request)
 
     resolved = str(override.get("resolved_voice", "")).strip()
     if resolved:
@@ -129,17 +156,17 @@ def chapter_voice_spec(job: Any, override: Optional[Dict[str, Any]]) -> str:
     if voice:
         return voice
 
-    return job_voice_fallback(job)
+    return job_voice_fallback(request)
 
 
-def chunk_voice_spec(job: Any, chunk: Dict[str, Any], fallback: str) -> str:
+def chunk_voice_spec(request: Any, chunk: Dict[str, Any], fallback: str) -> str:
     for key in ("resolved_voice", "voice_formula", "voice"):
         value = chunk.get(key)
         if value:
             return str(value)
 
     speaker_id = chunk.get("speaker_id")
-    speakers = getattr(job, "speakers", None)
+    speakers = getattr(request, "speakers", None)
     if isinstance(speakers, dict) and speaker_id in speakers:
         speaker_entry = speakers.get(speaker_id) or {}
         if isinstance(speaker_entry, dict):
@@ -163,7 +190,7 @@ def chunk_voice_spec(job: Any, chunk: Dict[str, Any], fallback: str) -> str:
 
     if fallback:
         return fallback
-    return job_voice_fallback(job)
+    return job_voice_fallback(request)
 
 
 def resolve_fallback_voice_spec(

tests/test_domain_pipeline_factory.py

@@ -124,11 +124,11 @@ class TestPipelinePool:
         mock_create.return_value = MagicMock()
         pool = PipelinePool()
 
-        job = MagicMock()
+        request = MagicMock()
-        pool.get("kokoro", "en", use_gpu=True, job=job)
+        pool.get("kokoro", "en", use_gpu=True, request=request)
         assert mock_cache.call_count == 1
 
-        pool.get("kokoro", "en", use_gpu=True, job=job)
+        pool.get("kokoro", "en", use_gpu=True, request=request)
         assert mock_cache.call_count == 1
 
     @patch("abogen.domain.pipeline_factory.initialize_voice_cache")

tests/test_regression_webui_domain.py

@@ -122,10 +122,10 @@ class TestPipelinePoolRegression:
     def test_voice_cache_initialized_only_once(self, mock_cache, mock_create):
         mock_create.return_value = MagicMock()
         pool = PipelinePool()
-        job = MagicMock()
+        request = MagicMock()
 
-        pool.get("kokoro", "en", use_gpu=True, job=job)
+        pool.get("kokoro", "en", use_gpu=True, request=request)
-        pool.get("kokoro", "en", use_gpu=True, job=job)
+        pool.get("kokoro", "en", use_gpu=True, request=request)
         assert mock_cache.call_count == 1
 
     @patch("abogen.domain.pipeline_factory.create_pipeline_for_job")
@@ -133,13 +133,13 @@ class TestPipelinePoolRegression:
     def test_after_dispose_voice_cache_can_reinitialize(self, mock_cache, mock_create):
         mock_create.return_value = MagicMock()
         pool = PipelinePool()
-        job = MagicMock()
+        request = MagicMock()
 
-        pool.get("kokoro", "en", use_gpu=True, job=job)
+        pool.get("kokoro", "en", use_gpu=True, request=request)
         assert mock_cache.call_count == 1
 
         pool.dispose_all()
         assert pool._voice_cache_initialized is False
 
-        pool.get("kokoro", "en", use_gpu=True, job=job)
+        pool.get("kokoro", "en", use_gpu=True, request=request)
         assert mock_cache.call_count == 2

tests/test_voice_resolution.py

@@ -1,6 +1,11 @@
 """Tests for voice resolution helpers.
 
 Tests import from domain.voice_resolution (new location).
+All test objects match the ConversionRequest interface:
+- voice: str
+- speakers: dict
+- chapter_chunk.chapter_overrides: list
+- chapter_chunk.chunks: list
 """
 
 from __future__ import annotations
@@ -11,6 +16,25 @@ from unittest.mock import patch
 
 import pytest
 
+from abogen.application.conversion_config import ChapterChunkConfig
+
+
+def _make_request(
+    voice: str = "",
+    speakers: dict | None = None,
+    chapter_overrides: list | None = None,
+    chunks: list | None = None,
+) -> SimpleNamespace:
+    """Create a ConversionRequest-like object for testing."""
+    return SimpleNamespace(
+        voice=voice,
+        speakers=speakers,
+        chapter_chunk=ChapterChunkConfig(
+            chapter_overrides=chapter_overrides or [],
+            chunks=chunks or [],
+        ),
+    )
+
 
 # ---------------------------------------------------------------------------
 # spec_to_voice_ids
@@ -72,65 +96,62 @@ class TestSpecToVoiceIds:
 
 
 class TestJobVoiceFallback:
-    """job_voice_fallback resolves a fallback voice from job attributes."""
+    """job_voice_fallback resolves a fallback voice from request attributes."""
 
     def test_direct_voice(self):
         from abogen.domain.voice_resolution import job_voice_fallback
 
-        job = SimpleNamespace(voice="af_heart", speakers=None, chapters=[])
+        request = _make_request(voice="af_heart")
-        assert job_voice_fallback(job) == "af_heart"
+        assert job_voice_fallback(request) == "af_heart"
 
     def test_custom_mix_ignored(self):
         from abogen.domain.voice_resolution import job_voice_fallback
 
-        job = SimpleNamespace(voice="__custom_mix", speakers=None, chapters=[])
+        request = _make_request(voice="__custom_mix")
-        assert job_voice_fallback(job) == ""
+        assert job_voice_fallback(request) == ""
 
     def test_narrator_speaker(self):
         from abogen.domain.voice_resolution import job_voice_fallback
 
-        job = SimpleNamespace(
+        request = _make_request(
             voice="__custom_mix",
             speakers={"narrator": {"resolved_voice": "af_heart"}},
-            chapters=[],
         )
-        assert job_voice_fallback(job) == "af_heart"
+        assert job_voice_fallback(request) == "af_heart"
 
     def test_speaker_voice_formula(self):
         from abogen.domain.voice_resolution import job_voice_fallback
 
-        job = SimpleNamespace(
+        request = _make_request(
             voice="",
             speakers={"speaker1": {"voice_formula": "v1*v2"}},
-            chapters=[],
         )
-        assert job_voice_fallback(job) == "v1*v2"
+        assert job_voice_fallback(request) == "v1*v2"
 
     def test_chapter_voice(self):
         from abogen.domain.voice_resolution import job_voice_fallback
 
-        job = SimpleNamespace(
+        request = _make_request(
             voice="",
-            speakers=None,
+            chapter_overrides=[{"resolved_voice": "af_bella"}],
-            chapters=[{"resolved_voice": "af_bella"}],
         )
-        assert job_voice_fallback(job) == "af_bella"
+        assert job_voice_fallback(request) == "af_bella"
 
-    def test_empty_job(self):
+    def test_empty_request(self):
         from abogen.domain.voice_resolution import job_voice_fallback
 
-        job = SimpleNamespace(voice="", speakers=None, chapters=[])
+        request = _make_request(voice="")
-        assert job_voice_fallback(job) == ""
+        assert job_voice_fallback(request) == ""
 
     def test_narrator_custom_mix_falls_through(self):
         from abogen.domain.voice_resolution import job_voice_fallback
 
-        job = SimpleNamespace(
+        request = _make_request(
             voice="",
             speakers={"narrator": {"voice": "__custom_mix"}},
-            chapters=[{"voice": "af_heart"}],
+            chapter_overrides=[{"voice": "af_heart"}],
         )
-        assert job_voice_fallback(job) == "af_heart"
+        assert job_voice_fallback(request) == "af_heart"
 
 
 # ---------------------------------------------------------------------------
@@ -144,35 +165,35 @@ class TestChapterVoiceSpec:
     def test_no_override_uses_fallback(self):
         from abogen.domain.voice_resolution import chapter_voice_spec
 
-        job = SimpleNamespace(voice="af_heart", speakers=None, chapters=[])
+        request = _make_request(voice="af_heart")
-        assert chapter_voice_spec(job, None) == "af_heart"
+        assert chapter_voice_spec(request, None) == "af_heart"
 
     def test_resolved_voice_wins(self):
         from abogen.domain.voice_resolution import chapter_voice_spec
 
-        job = SimpleNamespace(voice="af_heart", speakers=None, chapters=[])
+        request = _make_request(voice="af_heart")
         override = {"resolved_voice": "af_bella", "voice_formula": "x", "voice": "y"}
-        assert chapter_voice_spec(job, override) == "af_bella"
+        assert chapter_voice_spec(request, override) == "af_bella"
 
     def test_formula_second(self):
         from abogen.domain.voice_resolution import chapter_voice_spec
 
-        job = SimpleNamespace(voice="", speakers=None, chapters=[])
+        request = _make_request(voice="")
         override = {"voice_formula": "v1*v2", "voice": "y"}
-        assert chapter_voice_spec(job, override) == "v1*v2"
+        assert chapter_voice_spec(request, override) == "v1*v2"
 
     def test_voice_third(self):
         from abogen.domain.voice_resolution import chapter_voice_spec
 
-        job = SimpleNamespace(voice="", speakers=None, chapters=[])
+        request = _make_request(voice="")
         override = {"voice": "af_nicole"}
-        assert chapter_voice_spec(job, override) == "af_nicole"
+        assert chapter_voice_spec(request, override) == "af_nicole"
 
     def test_empty_override_falls_to_fallback(self):
         from abogen.domain.voice_resolution import chapter_voice_spec
 
-        job = SimpleNamespace(voice="af_heart", speakers=None, chapters=[])
+        request = _make_request(voice="af_heart")
-        assert chapter_voice_spec(job, {}) == "af_heart"
+        assert chapter_voice_spec(request, {}) == "af_heart"
 
 
 # ---------------------------------------------------------------------------
@@ -186,37 +207,37 @@ class TestChunkVoiceSpec:
     def test_chunk_direct_voice(self):
         from abogen.domain.voice_resolution import chunk_voice_spec
 
-        job = SimpleNamespace(speakers=None)
+        request = _make_request()
         chunk = {"resolved_voice": "af_heart"}
-        assert chunk_voice_spec(job, chunk, "fallback") == "af_heart"
+        assert chunk_voice_spec(request, chunk, "fallback") == "af_heart"
 
     def test_chunk_speaker_lookup(self):
         from abogen.domain.voice_resolution import chunk_voice_spec
 
-        job = SimpleNamespace(speakers={"narrator": {"resolved_voice": "af_bella"}})
+        request = _make_request(speakers={"narrator": {"resolved_voice": "af_bella"}})
         chunk = {"speaker_id": "narrator"}
-        assert chunk_voice_spec(job, chunk, "") == "af_bella"
+        assert chunk_voice_spec(request, chunk, "") == "af_bella"
 
     def test_chunk_voice_profile_lookup(self):
         from abogen.domain.voice_resolution import chunk_voice_spec
 
-        job = SimpleNamespace(speakers={"角色A": {"voice": "af_nicole"}})
+        request = _make_request(speakers={"角色A": {"voice": "af_nicole"}})
         chunk = {"voice_profile": "角色A"}
-        assert chunk_voice_spec(job, chunk, "") == "af_nicole"
+        assert chunk_voice_spec(request, chunk, "") == "af_nicole"
 
     def test_uses_fallback_string(self):
         from abogen.domain.voice_resolution import chunk_voice_spec
 
-        job = SimpleNamespace(speakers=None)
+        request = _make_request()
         chunk = {}
-        assert chunk_voice_spec(job, chunk, "my_fallback") == "my_fallback"
+        assert chunk_voice_spec(request, chunk, "my_fallback") == "my_fallback"
 
-    def test_fallback_to_job(self):
+    def test_fallback_to_request(self):
         from abogen.domain.voice_resolution import chunk_voice_spec
 
-        job = SimpleNamespace(voice="af_heart", speakers=None, chapters=[])
+        request = _make_request(voice="af_heart")
         chunk = {}
-        assert chunk_voice_spec(job, chunk, "") == "af_heart"
+        assert chunk_voice_spec(request, chunk, "") == "af_heart"
 
 
 # ---------------------------------------------------------------------------
@@ -225,50 +246,46 @@ class TestChunkVoiceSpec:
 
 
 class TestCollectRequiredVoiceIds:
-    """collect_required_voice_ids gathers all voice IDs from a job."""
+    """collect_required_voice_ids gathers all voice IDs from a request."""
 
-    def test_includes_job_voice(self):
+    def test_includes_request_voice(self):
         from abogen.domain.voice_resolution import collect_required_voice_ids
 
-        job = SimpleNamespace(voice="af_heart", chapters=[], chunks=[], speakers={})
+        request = _make_request(voice="af_heart")
         with patch("abogen.domain.voice_resolution.get_voices", return_value={"af_heart"}), \
              patch("abogen.domain.voice_resolution.job_voice_fallback", return_value=""):
-            result = collect_required_voice_ids(job)
+            result = collect_required_voice_ids(request)
         assert "af_heart" in result
 
     def test_includes_chapter_voices(self):
         from abogen.domain.voice_resolution import collect_required_voice_ids
 
-        job = SimpleNamespace(
+        request = _make_request(
             voice="",
-            chapters=[{"resolved_voice": "af_bella"}],
+            chapter_overrides=[{"resolved_voice": "af_bella"}],
-            chunks=[],
-            speakers={},
         )
         with patch("abogen.domain.voice_resolution.get_voices", return_value={"af_bella"}), \
              patch("abogen.domain.voice_resolution.job_voice_fallback", return_value=""):
-            result = collect_required_voice_ids(job)
+            result = collect_required_voice_ids(request)
         assert "af_bella" in result
 
     def test_includes_chunk_voices(self):
         from abogen.domain.voice_resolution import collect_required_voice_ids
 
-        job = SimpleNamespace(
+        request = _make_request(
             voice="",
-            chapters=[],
             chunks=[{"voice": "af_nicole"}],
-            speakers={},
         )
         with patch("abogen.domain.voice_resolution.get_voices", return_value={"af_nicole"}), \
              patch("abogen.domain.voice_resolution.job_voice_fallback", return_value=""):
-            result = collect_required_voice_ids(job)
+            result = collect_required_voice_ids(request)
         assert "af_nicole" in result
 
     def test_always_includes_kokoro_voices(self):
         from abogen.domain.voice_resolution import collect_required_voice_ids
 
-        job = SimpleNamespace(voice="", chapters=[], chunks=[], speakers={})
+        request = _make_request(voice="")
         with patch("abogen.domain.voice_resolution.get_voices", return_value={"af_heart", "af_bella"}), \
              patch("abogen.domain.voice_resolution.job_voice_fallback", return_value=""):
-            result = collect_required_voice_ids(job)
+            result = collect_required_voice_ids(request)
         assert {"af_heart", "af_bella"}.issubset(result)