refactor: heading transforms + dedup moved to shared layer

- conversion_planner.py: caps normalization in _build_chapters()
- conversion_executor.py: heading dedup + state machine via headings_equivalent()
- Fixed seg_start_time → chapter_body_start bug in executor
- Removed getattr fallback defaults in both adapters
- Added 4 tests for caps normalization and heading dedup
- Updated ARCHITECTURE_REFACTOR_PLAN.md with deferred WebUI cleanup

tests/test_conversion_executor_unified.py

@@ -511,3 +511,114 @@ class TestExecuteConversion:
 
             assert result.metadata["title"] == "Test Book"
             assert result.metadata["author"] == "Author"
+
+
+class TestHeadingDedup:
+    """Tests for heading dedup in executor."""
+
+    def test_heading_dedup_strips_matching_first_line(self):
+        """When first segment matches heading, it should be stripped."""
+        with tempfile.TemporaryDirectory() as tmpdir:
+            req = ConversionRequest(
+                direct_text="Hello",
+                voice="M1",
+                auto_prefix_chapter_titles=True,
+            )
+            # Simulate: heading = "Chapter 1", first segment = "Chapter 1: The Beginning"
+            # headings_equivalent should match these
+            plan = ConversionPlan(
+                request=req,
+                metadata={},
+                chapters=[
+                    ChapterPlan(
+                        index=1,
+                        title="Chapter 1",
+                        original_title="Chapter 1",
+                        body_text="Chapter 1: The Beginning\nBody text here",
+                        segments=[
+                            SegmentPlan(
+                                text="Chapter 1: The Beginning",
+                                voice_spec="M1",
+                                kind="body",
+                                source="chapter",
+                            ),
+                            SegmentPlan(
+                                text="Body text here",
+                                voice_spec="M1",
+                                kind="body",
+                                source="chapter",
+                            ),
+                        ],
+                        voice_spec="M1",
+                    )
+                ],
+                output_layout=OutputLayout(
+                    parent_dir=Path(tmpdir),
+                    audio_dir=Path(tmpdir),
+                ),
+            )
+
+            events = FakeEvents()
+            pipeline = FakePipelineProvider()
+            resolver = FakeVoiceResolver()
+            tts_context = TTSContext()
+
+            result = execute_conversion(
+                plan, events, pipeline, resolver, tts_context
+            )
+
+            # The executor should have logged the heading
+            log_messages = [m for m, _ in events.logs if "Title:" in m]
+            assert len(log_messages) >= 1
+
+    def test_heading_dedup_no_match_preserves_all(self):
+        """When first segment doesn't match heading, nothing is stripped."""
+        with tempfile.TemporaryDirectory() as tmpdir:
+            req = ConversionRequest(
+                direct_text="Hello",
+                voice="M1",
+                auto_prefix_chapter_titles=True,
+            )
+            plan = ConversionPlan(
+                request=req,
+                metadata={},
+                chapters=[
+                    ChapterPlan(
+                        index=1,
+                        title="Chapter 1",
+                        original_title="Chapter 1",
+                        body_text="Completely different text\nMore text",
+                        segments=[
+                            SegmentPlan(
+                                text="Completely different text",
+                                voice_spec="M1",
+                                kind="body",
+                                source="chapter",
+                            ),
+                            SegmentPlan(
+                                text="More text",
+                                voice_spec="M1",
+                                kind="body",
+                                source="chapter",
+                            ),
+                        ],
+                        voice_spec="M1",
+                    )
+                ],
+                output_layout=OutputLayout(
+                    parent_dir=Path(tmpdir),
+                    audio_dir=Path(tmpdir),
+                ),
+            )
+
+            events = FakeEvents()
+            pipeline = FakePipelineProvider()
+            resolver = FakeVoiceResolver()
+            tts_context = TTSContext()
+
+            result = execute_conversion(
+                plan, events, pipeline, resolver, tts_context
+            )
+
+            # Both segments should be synthesized (heading + 2 body segments)
+            assert result.total_segments >= 2

tests/test_conversion_planner.py

@@ -641,3 +641,31 @@ class TestFeatureParity:
         if output_format.lower() == "m4b":
             merge_chapters_at_end = True
         assert merge_chapters_at_end is True
+
+
+class TestCapsNormalization:
+    """Tests for caps normalization in planner."""
+
+    def test_caps_normalization_applied_when_enabled(self):
+        """When normalize_chapter_opening_caps=True, body text is normalized."""
+        req = ConversionRequest(
+            direct_text="<<CHAPTER_MARKER:Chapter 1>>\nALL CAPS OPENING TEXT here",
+            voice="M1",
+            normalize_chapter_opening_caps=True,
+        )
+        plan = build_conversion_plan(req)
+        body = plan.chapters[0].body_text
+        # ALL CAPS should be normalized to Title Case
+        assert body != "ALL CAPS OPENING TEXT here"
+        assert "ALL CAPS" not in body
+
+    def test_caps_normalization_skipped_when_disabled(self):
+        """When normalize_chapter_opening_caps=False, body text is unchanged."""
+        req = ConversionRequest(
+            direct_text="<<CHAPTER_MARKER:Chapter 1>>\nALL CAPS OPENING TEXT here",
+            voice="M1",
+            normalize_chapter_opening_caps=False,
+        )
+        plan = build_conversion_plan(req)
+        body = plan.chapters[0].body_text
+        assert "ALL CAPS OPENING TEXT" in body