Phase 3: Speaker diarization and full transcription pipeline

- Implement DiarizeService with pyannote.audio speaker detection
- Build PipelineService combining transcribe → diarize → merge with
  overlap-based speaker assignment per segment
- Add pipeline.start and diarize.start IPC handlers
- Add run_pipeline Tauri command for full pipeline execution
- Wire frontend to use pipeline: speakers auto-created with colors,
  segments assigned to detected speakers
- Build SpeakerManager with rename support (double-click or edit button)
- Add speaker color coding throughout transcript display
- Add pyannote.audio dependency
- Tests: 24 Python (including merge logic), 6 Rust, 0 Svelte errors

Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>

python/tests/test_pipeline.py

@@ -0,0 +1,90 @@
+"""Tests for pipeline service data structures and merge logic."""
+
+from voice_to_notes.services.diarize import SpeakerSegment
+from voice_to_notes.services.pipeline import (
+    PipelineResult,
+    PipelineSegment,
+    PipelineService,
+    pipeline_result_to_payload,
+)
+from voice_to_notes.services.transcribe import (
+    SegmentResult,
+    TranscriptionResult,
+    WordResult,
+)
+
+
+def test_pipeline_result_to_payload():
+    result = PipelineResult(
+        segments=[
+            PipelineSegment(
+                text="Hello world",
+                start_ms=0,
+                end_ms=2000,
+                speaker="SPEAKER_00",
+                words=[
+                    WordResult(word="Hello", start_ms=0, end_ms=800, confidence=0.95),
+                    WordResult(word="world", start_ms=900, end_ms=2000, confidence=0.88),
+                ],
+            ),
+        ],
+        language="en",
+        language_probability=0.98,
+        duration_ms=10000,
+        speakers=["SPEAKER_00", "SPEAKER_01"],
+        num_speakers=2,
+    )
+    payload = pipeline_result_to_payload(result)
+    assert payload["language"] == "en"
+    assert payload["num_speakers"] == 2
+    assert len(payload["segments"]) == 1
+    assert payload["segments"][0]["speaker"] == "SPEAKER_00"
+    assert len(payload["segments"][0]["words"]) == 2
+
+
+def test_pipeline_result_to_payload_empty():
+    result = PipelineResult()
+    payload = pipeline_result_to_payload(result)
+    assert payload["segments"] == []
+    assert payload["speakers"] == []
+    assert payload["num_speakers"] == 0
+
+
+def test_merge_results_assigns_speakers():
+    """Test that _merge_results correctly assigns speakers based on overlap."""
+    service = PipelineService()
+
+    transcription = TranscriptionResult(
+        segments=[
+            SegmentResult(text="Hello there", start_ms=0, end_ms=3000, words=[]),
+            SegmentResult(text="How are you", start_ms=4000, end_ms=7000, words=[]),
+        ],
+        language="en",
+        language_probability=0.99,
+        duration_ms=10000,
+    )
+
+    speaker_segments = [
+        SpeakerSegment(speaker="SPEAKER_00", start_ms=0, end_ms=3500),
+        SpeakerSegment(speaker="SPEAKER_01", start_ms=3500, end_ms=8000),
+    ]
+
+    result = service._merge_results(transcription, speaker_segments)
+    assert len(result.segments) == 2
+    assert result.segments[0].speaker == "SPEAKER_00"
+    assert result.segments[1].speaker == "SPEAKER_01"
+
+
+def test_merge_results_no_speaker_segments():
+    """With no speaker segments, all speakers should be None."""
+    service = PipelineService()
+
+    transcription = TranscriptionResult(
+        segments=[SegmentResult(text="Hello", start_ms=0, end_ms=1000, words=[])],
+        language="en",
+        language_probability=0.99,
+        duration_ms=1000,
+    )
+
+    result = service._merge_results(transcription, [])
+    assert result.segments[0].speaker is None