Pr 1

.github/workflows/build-and-test.yml

@@ -34,12 +34,12 @@ jobs:
       if: runner.os == 'Linux'
       run: |
         sudo apt-get update
-        sudo apt-get install -y libsndfile1 ffmpeg
+        sudo apt-get install -y libsndfile1 ffmpeg portaudio19-dev
 
     - name: Install system dependencies (macOS)
       if: runner.os == 'macOS'
       run: |
-        brew install libsndfile ffmpeg
+        brew install libsndfile ffmpeg portaudio
 
     - name: Install system dependencies (Windows)
       if: runner.os == 'Windows'
@@ -84,6 +84,18 @@ jobs:
             print('[OK] All imports successful')
             asr = StreamingASR(debug=False)
             print(f'[OK] StreamingASR initialized: {asr.chunk_size_ms}ms chunks')
+
+            # Test microphone feature availability (should not fail in CI)
+            try:
+                import sounddevice as sd
+                print('[OK] sounddevice imported successfully')
+                # Don't test actual microphone functionality in CI
+                print('[INFO] Microphone functionality available (not tested in CI)')
+            except ImportError as e:
+                print(f'[WARNING] sounddevice not available: {e}')
+            except Exception as e:
+                print(f'[INFO] sounddevice available but no audio devices in CI: {e}')
+
         except Exception as e:
             print(f'[ERROR] Import failed: {e}')
             sys.exit(1)

README.md

@@ -53,6 +53,7 @@ for result in asr.stream_from_file("audio.wav"):
 
 ### Python API
 
+#### Streaming from File
 ```python
 from ViStreamASR import StreamingASR
 
@@ -67,6 +68,21 @@ for result in asr.stream_from_file("audio.wav"):
         print(f"Final: {result['text']}")
 ```
 
+#### Streaming from Microphone
+```python
+from ViStreamASR import StreamingASR
+
+# Initialize ASR
+asr = StreamingASR()
+
+# Process audio file
+for result in asr.stream_from_microphone(duration_seconds=10):
+    if result['partial']:
+        print(f"Partial: {result['text']}")
+    if result['final']:
+        print(f"Final: {result['text']}")
+```
+
 ### Command Line
 
 ```bash

example.py

@@ -1,47 +1,54 @@
 #!/usr/bin/env python3
 """
 Example script demonstrating ViStreamASR usage.
+This script tests all features using the local codebase before building.
 """
 
 import os
 import sys
+import argparse
 
-def main():
-    print("🎤 ViStreamASR Example")
-    print("=" * 40)
+# Add src directory to path to use local codebase instead of installed library
+sys.path.insert(0, 'src')
+
+def test_file_streaming(audio_file, chunk_size=640, debug=True):
+    """Test file streaming functionality."""
+    print(f"🎵 Testing File Streaming")
+    print(f"=" * 50)
 
     try:
-        from ViStreamASR import StreamingASR
-        print("✅ ViStreamASR imported successfully")
+        from streaming import StreamingASR
+        print("✅ Local StreamingASR imported successfully")
     except ImportError as e:
-        print(f"❌ Failed to import ViStreamASR: {e}")
-        print("💡 Try running: pip install -e .")
-        return 1
+        print(f"❌ Failed to import local StreamingASR: {e}")
+        print("💡 Make sure you're running from the project root directory")
+        return False
 
-    # Check if example audio file exists
-    audio_file = "resource/linh_ref_long.wav"
     if not os.path.exists(audio_file):
-        print(f"❌ Example audio file not found: {audio_file}")
-        return 1
+        print(f"❌ Audio file not found: {audio_file}")
+        return False
 
     print(f"📁 Using audio file: {audio_file}")
 
     # Initialize StreamingASR
     print(f"\n🔄 Initializing StreamingASR...")
-    asr = StreamingASR(chunk_size_ms=640, debug=True)
+    asr = StreamingASR(chunk_size_ms=chunk_size, debug=debug)
 
     # Run streaming transcription
-    print(f"\n🎵 Starting streaming transcription...")
+    print(f"\n🎵 Starting file streaming transcription...")
     print(f"=" * 60)
 
     final_segments = []
+    partial_count = 0
 
     try:
         for result in asr.stream_from_file(audio_file):
             chunk_info = result.get('chunk_info', {})
 
             if result.get('partial'):
-                print(f"📝 [PARTIAL {chunk_info.get('chunk_id', '?'):3d}] {result['text']}")
+                partial_count += 1
+                text = result['text'][:60] + "..." if len(result['text']) > 60 else result['text']
+                print(f"📝 [PARTIAL {chunk_info.get('chunk_id', '?'):3d}] {text}")
 
             if result.get('final'):
                 final_text = result['text']
@@ -50,26 +57,229 @@ def main():
                 print(f"-" * 60)
 
     except KeyboardInterrupt:
-        print(f"\n⏹️  Interrupted by user")
-        return 0
+        print(f"\n⏹️  File streaming interrupted by user")
+        return True
     except Exception as e:
-        print(f"\n❌ Error during streaming: {e}")
+        print(f"\n❌ Error during file streaming: {e}")
         import traceback
         traceback.print_exc()
-        return 1
+        return False
+
+    # Show results
+    print(f"\n📊 FILE STREAMING RESULTS")
+    print(f"=" * 40)
+    print(f"📝 Partial updates: {partial_count}")
+    print(f"📝 Final segments: {len(final_segments)}")
+
+    if final_segments:
+        print(f"\n📝 Complete Transcription:")
+        print(f"-" * 40)
+        complete_text = " ".join(final_segments)
+        print(f"{complete_text}")
+
+    print(f"\n✅ File streaming test completed successfully!")
+    return True
+
+def test_microphone_streaming(duration=10, chunk_size=640, debug=True):
+    """Test microphone streaming functionality."""
+    print(f"\n🎤 Testing Microphone Streaming")
+    print(f"=" * 50)
+
+    try:
+        from streaming import StreamingASR
+        print("✅ Local StreamingASR imported successfully")
+    except ImportError as e:
+        print(f"❌ Failed to import local StreamingASR: {e}")
+        return False
+
+    try:
+        import sounddevice as sd
+        print("✅ sounddevice library available")
+
+        # Test if microphone is available
+        devices = sd.query_devices()
+        input_devices = [d for d in devices if d['max_input_channels'] > 0]
+        if not input_devices:
+            print("❌ No input devices (microphones) found")
+            return False
+        print(f"🎤 Found {len(input_devices)} input device(s)")
+
+    except ImportError:
+        print("❌ sounddevice library not installed. Install with: pip install sounddevice")
+        return False
+    except Exception as e:
+        print(f"❌ Error checking audio devices: {e}")
+        return False
+
+    # Initialize StreamingASR
+    print(f"\n🔄 Initializing StreamingASR for microphone...")
+    asr = StreamingASR(chunk_size_ms=chunk_size, debug=debug)
+
+    # Run microphone streaming
+    print(f"\n🎤 Starting microphone streaming for {duration} seconds...")
+    print(f"🔊 Please speak into your microphone...")
+    print(f"=" * 60)
+
+    final_segments = []
+    partial_count = 0
+
+    try:
+        for result in asr.stream_from_microphone(duration_seconds=duration):
+            chunk_info = result.get('chunk_info', {})
+
+            if result.get('partial'):
+                partial_count += 1
+                text = result['text'][:60] + "..." if len(result['text']) > 60 else result['text']
+                print(f"🎙️  [PARTIAL {chunk_info.get('chunk_id', '?'):3d}] {text}")
+
+            if result.get('final'):
+                final_text = result['text']
+                final_segments.append(final_text)
+                print(f"✅ [FINAL   {chunk_info.get('chunk_id', '?'):3d}] {final_text}")
+                print(f"-" * 60)
+
+    except KeyboardInterrupt:
+        print(f"\n⏹️  Microphone streaming interrupted by user")
+        return True
+    except Exception as e:
+        print(f"\n❌ Error during microphone streaming: {e}")
+        import traceback
+        traceback.print_exc()
+        return False
 
     # Show results
-    print(f"\n📊 RESULTS")
+    print(f"\n📊 MICROPHONE STREAMING RESULTS")
     print(f"=" * 40)
+    print(f"📝 Partial updates: {partial_count}")
     print(f"📝 Final segments: {len(final_segments)}")
 
-    print(f"\n📝 Complete Transcription:")
-    print(f"-" * 40)
-    complete_text = " ".join(final_segments)
-    print(f"{complete_text}")
+    if final_segments:
+        print(f"\n📝 Transcribed from microphone:")
+        print(f"-" * 40)
+        complete_text = " ".join(final_segments)
+        print(f"{complete_text}")
+    else:
+        print(f"ℹ️  No speech detected or transcribed during recording")
 
-    print(f"\n✅ Example completed successfully!")
-    return 0
+    print(f"\n✅ Microphone streaming test completed successfully!")
+    return True
+
+def test_asr_engine():
+    """Test low-level ASREngine functionality."""
+    print(f"\n🔧 Testing ASREngine (Low-level API)")
+    print(f"=" * 50)
+
+    try:
+        from core import ASREngine
+        print("✅ Local ASREngine imported successfully")
+    except ImportError as e:
+        print(f"❌ Failed to import local ASREngine: {e}")
+        return False
+
+    try:
+        # Test ASREngine initialization
+        engine = ASREngine(chunk_size_ms=640, debug_mode=True)
+        print("✅ ASREngine initialized successfully")
+
+        # Test model initialization
+        print("🔄 Initializing models (this may take a moment)...")
+        engine.initialize_models()
+        print("✅ Models initialized successfully")
+
+        # Test RTF calculation
+        rtf = engine.get_asr_rtf()
+        print(f"📊 Current RTF: {rtf:.3f}x")
+
+        # Test state reset
+        engine.reset_state()
+        print("✅ State reset successful")
+
+        print(f"\n✅ ASREngine test completed successfully!")
+        return True
+
+    except Exception as e:
+        print(f"\n❌ Error during ASREngine testing: {e}")
+        import traceback
+        traceback.print_exc()
+        return False
+
+def main():
+    parser = argparse.ArgumentParser(
+        description="ViStreamASR Feature Test Script",
+        formatter_class=argparse.RawDescriptionHelpFormatter,
+        epilog="""
+Examples:
+  python example.py                                    # Test all features
+  python example.py --file-only                       # Test only file streaming
+  python example.py --mic-only                        # Test only microphone streaming
+  python example.py --mic-duration 15                 # Test microphone for 15 seconds
+  python example.py --show-debug                        # Enable debug output
+        """
+    )
+
+    parser.add_argument('--file-only', action='store_true',
+                       help='Test only file streaming functionality')
+    parser.add_argument('--mic-only', action='store_true',
+                       help='Test only microphone streaming functionality')
+    parser.add_argument('--engine-only', action='store_true',
+                       help='Test only ASREngine low-level functionality')
+    parser.add_argument('--audio-file', default="resource/linh_ref_long.wav",
+                       help='Audio file for testing (default: resource/linh_ref_long.wav)')
+    parser.add_argument('--mic-duration', type=int, default=10,
+                       help='Duration for microphone test in seconds (default: 10)')
+    parser.add_argument('--chunk-size', type=int, default=640,
+                       help='Chunk size in milliseconds (default: 640)')
+    parser.add_argument('--show-debug', action='store_true', default=False,
+                       help='Enable debug output')
+
+    args = parser.parse_args()
+
+    print("🎤 ViStreamASR Feature Test")
+    print("=" * 50)
+    print("🧪 Testing local codebase (src/) instead of installed library")
+    print()
+
+    debug_mode = args.show_debug
+    test_results = []
+
+    # Test file streaming
+    if not args.mic_only and not args.engine_only:
+        file_result = test_file_streaming(args.audio_file, args.chunk_size, debug_mode)
+        test_results.append(("File Streaming", file_result))
+
+    # Test microphone streaming
+    if not args.file_only and not args.engine_only:
+        mic_result = test_microphone_streaming(args.mic_duration, args.chunk_size, debug_mode)
+        test_results.append(("Microphone Streaming", mic_result))
+
+    # Test ASREngine
+    if not args.file_only and not args.mic_only:
+        engine_result = test_asr_engine()
+        test_results.append(("ASREngine", engine_result))
+
+    # Summary
+    print(f"\n🏁 TEST SUMMARY")
+    print(f"=" * 50)
+
+    passed = 0
+    failed = 0
+
+    for test_name, result in test_results:
+        status = "✅ PASSED" if result else "❌ FAILED"
+        print(f"{status} {test_name}")
+        if result:
+            passed += 1
+        else:
+            failed += 1
+
+    print(f"\n📊 Results: {passed} passed, {failed} failed")
+
+    if failed == 0:
+        print(f"🎉 All tests passed! The codebase is ready for building.")
+        return 0
+    else:
+        print(f"⚠️  Some tests failed. Please review the issues above.")
+        return 1
 
 if __name__ == "__main__":
     sys.exit(main()) 

requirements.txt

@@ -3,4 +3,5 @@ torchaudio>=2.5.0
 numpy>=1.19.0
 requests>=2.25.0
 flashlight-text
-librosa 
+librosa
+sounddevice>=0.5.2

setup.py

@@ -65,6 +65,7 @@ def get_author():
         "requests>=2.25.0",
         "flashlight-text",
         "librosa",
+        "sounddevice>=0.4.0",
     ],
     extras_require={
         "dev": [