meeting-assistant/MeetingAssistant/Speakers/WavSpeakerSnippetExtractor.cs

using MeetingAssistant.Transcription;
using NAudio.Wave;

namespace MeetingAssistant.Speakers;

public sealed class WavSpeakerSnippetExtractor : ISpeakerSnippetExtractor
{
    public Task<byte[]> ExtractSnippetAsync(
        string audioPath,
        IReadOnlyList<TranscriptionSegment> speakerSegments,
        CancellationToken cancellationToken)
    {
        if (speakerSegments.Count == 0 || !File.Exists(audioPath))
        {
            return Task.FromResult<byte[]>([]);
        }

        var start = speakerSegments.Min(segment => segment.Start);
        var end = speakerSegments.Max(segment => segment.End);
        if (end <= start)
        {
            return Task.FromResult<byte[]>([]);
        }

        using var reader = new WaveFileReader(audioPath);
        using var output = new MemoryStream();
        using (var writer = new WaveFileWriter(output, reader.WaveFormat))
        {
            reader.CurrentTime = start;
            var bytesRemaining = reader.WaveFormat.AverageBytesPerSecond * (end - start).TotalSeconds;
            var buffer = new byte[reader.WaveFormat.AverageBytesPerSecond];
            while (bytesRemaining > 0)
            {
                cancellationToken.ThrowIfCancellationRequested();
                var bytesToRead = (int)Math.Min(buffer.Length, bytesRemaining);
                var bytesRead = reader.Read(buffer, 0, bytesToRead);
                if (bytesRead == 0)
                {
                    break;
                }

                writer.Write(buffer, 0, bytesRead);
                bytesRemaining -= bytesRead;
            }
        }

        return Task.FromResult(output.ToArray());
    }
}