using System;
using System.Collections;
using System.Collections.Generic;
using System.IO;
using UnityEngine;
using UnityEngine.UI;

public class GetVoice : MonoBehaviour
{
    [SerializeField] 
    AudioSource listenSource;

    [SerializeField]
    Animator animator;

    //private fields
    private float gain = 8000.0f;
    float volume;
    float frequency;
    bool isSpeaking;
    bool isRecording;
    float bufferTimer;　//話した後の余韻の時間 or　言葉に詰まった時の時間　の計測時間
    float speakingTimer;　//実際に話している時間

    //constant values
    const int RECORD_SECONDS = 2;
    const int SAMPLING_FREQUENCY = 16000; // MSのSSTはサンプリングレートが16000Hz
    const int FFTSAMPLES = 2 << 8; //256bitのサンプルをFFTでは選んでとる。

    const float MIN_FREQ = 100.0f; //母音のFrequency は　100Hz 以上 1400Hz以下。　ただし、子音字は異なる。
    const float MAX_FREQ = 1400.0f;
    const float MIN_VOLUME = 1.0f;
    const float BUFFER_TIME = 1.6f;　//話した後の余韻の時間 or　言葉に詰まった時の時間。


    // Start is called before the first frame update
    void Start()
    {
        #if UNITY_EDITOR
        gain = 2000.0f;
        #endif

        GetMic();
    }


    void Update()
    {
        Waiting();
    }



    private void GetMic()
    {
        while (Microphone.devices.Length< 1) { }
        string device = Microphone.devices[0];
        listenSource.loop = true;
        listenSource.clip = Microphone.Start(device, true, RECORD_SECONDS, SAMPLING_FREQUENCY);
        while (!(Microphone.GetPosition(device) > 0)) { }
        listenSource.Play();
    }

    async private void Waiting()
    {
        CalculateVowel();
        if (MIN_FREQ < frequency && frequency < MAX_FREQ && MIN_VOLUME < volume) //しゃべり始めの時間
        {
            isSpeaking = true;

            bufferTimer = 0.0f;
            speakingTimer += Time.deltaTime;

            //始めてレコーディングを開始したとき
            if (!isRecording)
            {
                isRecording = true;

                bufferTimer = 0.0f;
                speakingTimer = 0.0f;

            }
        }
        else if (isSpeaking && volume > MIN_VOLUME)
        {
            //子音字をしゃべっていると判定
            bufferTimer = 0.0f;
            speakingTimer += Time.deltaTime;
        }
        else if (isSpeaking && bufferTimer < BUFFER_TIME) // 余韻の時間
        {
            bufferTimer += Time.deltaTime;
            speakingTimer += Time.deltaTime;
        }
        else
        {
            bufferTimer = 0.0f; // 後で、speakingTimerの条件処理あり！

            isSpeaking = false;
            if (isRecording)
            {
                isRecording = false;

                AizuchiAnimation(speakingTimer);
            }

            speakingTimer = 0.0f;
        }

    }

    private void CalculateVowel()
    {

        //ここが処理の重さ的にやばいかも？
        var max_volume = 0.0f;
        var max_index = 0;
        var total_volume = 0.0f;
        //録音時間＊サンプリング周波数の個数のデータがほしい！
        float[] temp = new float[FFTSAMPLES];
        listenSource.GetSpectrumData(temp, 0, FFTWindow.Blackman);
        for (int i = 0; i < temp.Length; i++)
        {
            if (max_volume < temp[i])
            {
                max_index = i;
                max_volume = temp[i];
            }
            total_volume += Mathf.Abs(temp[i]);
        }

        if (temp.Length > 0)
        {
            frequency = max_index * AudioSettings.outputSampleRate / 2 / temp.Length;
            volume = total_volume / temp.Length * gain;
        }
    }

    private void AizuchiAnimation(float speakingTime)
    {
        if (2.0f < speakingTime && speakingTime < 5.0f)
        {
            animator.SetInteger("aizuchi", 1);
        }

        else if (5.0f < speakingTime && speakingTime < 8.0f)
        {
            animator.SetInteger("aizuchi", 2);
        }
        else if (8.0f < speakingTime)
        {
            animator.SetInteger("aizuchi", 3);
        }
    }
}

animator.SetInteger("aizuchi", 0)

    [SerializeField] 
    public AudioClip[] aizuchiClips;

\\\
    void AizuchiPlayBack(int aizuchiIndex)
    {
        audioSource.PlayOneShot(aizuchiClips[aizuchiIndex-1], 1.0f);
    }