/* JOrbis
 * Copyright (C) 2000 ymnk, JCraft,Inc.
 *  
 * Written by: 2000 ymnk<ymnk@jcraft.com>
 *   
 * Many thanks to 
 *   Monty <monty@xiph.org> and 
 *   The XIPHOPHORUS Company http://www.xiph.org/ .
 * JOrbis has been based on their awesome works, Vorbis codec.
 *   
 * This program is free software; you can redistribute it and/or
 * modify it under the terms of the GNU Library General Public License
 * as published by the Free Software Foundation; either version 2 of
 * the License, or (at your option) any later version.
   
 * This program is distributed in the hope that it will be useful,
 * but WITHOUT ANY WARRANTY; without even the implied warranty of
 * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
 * GNU Library General Public License for more details.
 * 
 * You should have received a copy of the GNU Library General Public
 * License along with this program; if not, write to the Free Software
 * Foundation, Inc., 675 Mass Ave, Cambridge, MA 02139, USA.
 */

package com.jcraft.jorbis;

class Mdct{

  static private final float cPI3_8=0.38268343236508977175f;
  static private final float cPI2_8=0.70710678118654752441f;
  static private final float cPI1_8=0.92387953251128675613f;

  int n;
  int log2n;
  
  float[] trig;
  int[] bitrev;

  float scale;

  void init(int n){
    bitrev=new int[n/4];
    trig=new float[n+n/4];

    int n2=n>>>1;
    log2n=(int)Math.rint(Math.log(n)/Math.log(2));
    this.n=n;


    int AE=0;
    int AO=1;
    int BE=AE+n/2;
    int BO=BE+1;
    int CE=BE+n/2;
    int CO=CE+1;
    // trig lookups...
    for(int i=0;i<n/4;i++){
      trig[AE+i*2]=(float)Math.cos((Math.PI/n)*(4*i));
      trig[AO+i*2]=(float)-Math.sin((Math.PI/n)*(4*i));
      trig[BE+i*2]=(float)Math.cos((Math.PI/(2*n))*(2*i+1));
      trig[BO+i*2]=(float)Math.sin((Math.PI/(2*n))*(2*i+1));
    }
    for(int i=0;i<n/8;i++){
      trig[CE+i*2]=(float)Math.cos((Math.PI/n)*(4*i+2));
      trig[CO+i*2]=(float)-Math.sin((Math.PI/n)*(4*i+2));
    }

    {
      int mask=(1<<(log2n-1))-1;
      int msb=1<<(log2n-2);
      for(int i=0;i<n/8;i++){
	int acc=0;
	for(int j=0;msb>>>j!=0;j++)
	  if(((msb>>>j)&i)!=0)acc|=1<<j;
	bitrev[i*2]=((~acc)&mask);
//	bitrev[i*2]=((~acc)&mask)-1;
	bitrev[i*2+1]=acc;
      }
    }
    scale=4.f/n;
  }

  void clear(){
  }

  void forward(float[] in, float[] out){
  }

  float[] _x=new float[1024];
  float[] _w=new float[1024];

  synchronized void backward(float[] in, float[] out){
    if(_x.length<n/2){_x=new float[n/2];}
    if(_w.length<n/2){_w=new float[n/2];}
    float[] x=_x;
    float[] w=_w;
    int n2=n>>>1;
    int n4=n>>>2;
    int n8=n>>>3;

    // rotate + step 1
    {
      int inO=1;
      int xO=0;
      int A=n2;

      int i;
      for(i=0;i<n8;i++){
	A-=2;
	x[xO++]=-in[inO+2]*trig[A+1] - in[inO]*trig[A];
	x[xO++]= in[inO]*trig[A+1] - in[inO+2]*trig[A];
	inO+=4;
      }

      inO=n2-4;

      for(i=0;i<n8;i++){
	A-=2;
	x[xO++]=in[inO]*trig[A+1] + in[inO+2]*trig[A];
	x[xO++]=in[inO]*trig[A] - in[inO+2]*trig[A+1];
	inO-=4;
      }
    }

    float[] xxx=mdct_kernel(x,w,n,n2,n4,n8);
    int xx=0;

    // step 8

    {
      int B=n2;
      int o1=n4,o2=o1-1;
      int o3=n4+n2,o4=o3-1;
    
      for(int i=0;i<n4;i++){
	float temp1= (xxx[xx] * trig[B+1] - xxx[xx+1] * trig[B]);
	float temp2=-(xxx[xx] * trig[B] + xxx[xx+1] * trig[B+1]);
    
	out[o1]=-temp1;
	out[o2]= temp1;
	out[o3]= temp2;
	out[o4]= temp2;

	o1++;
	o2--;
	o3++;
	o4--;
	xx+=2;
	B+=2;
      }
    }
  }
  private float[] mdct_kernel(float[] x, float[] w,
	                       int n, int n2, int n4, int n8){
    // step 2

    int xA=n4;
    int xB=0;
    int w2=n4;
    int A=n2;

    for(int i=0;i<n4;){
      float x0=x[xA] - x[xB];
      float x1;
      w[w2+i]=x[xA++]+x[xB++];

      x1=x[xA]-x[xB];
      A-=4;

      w[i++]=   x0 * trig[A] + x1 * trig[A+1];
      w[i]=     x1 * trig[A] - x0 * trig[A+1];

      w[w2+i]=x[xA++]+x[xB++];
      i++;
    }

    // step 3

    {
      for(int i=0;i<log2n-3;i++){
        int k0=n>>>(i+2);
	int k1=1<<(i+3);
	int wbase=n2-2;

	A=0;
	float[] temp;

	for(int r=0;r<(k0>>>2);r++){
	  int w1=wbase;
	  w2=w1-(k0>>1);
	  float AEv= trig[A],wA;
	  float AOv= trig[A+1],wB;
	  wbase-=2;
		      
	  k0++;
	  for(int s=0;s<(2<<i);s++){
	    wB     =w[w1]   -w[w2];
	    x[w1]  =w[w1]   +w[w2];

	    wA     =w[++w1] -w[++w2];
	    x[w1]  =w[w1]   +w[w2];
	    
	    x[w2]  =wA*AEv  - wB*AOv;
	    x[w2-1]=wB*AEv  + wA*AOv;

	    w1-=k0;
	    w2-=k0;
	  }
	  k0--;
	  A+=k1;
	}

	temp=w;
	w=x;
	x=temp;
      }
    }

    // step 4, 5, 6, 7
    {
      int C=n;
      int bit=0;
      int x1=0;
      int x2=n2-1;

      for(int i=0;i<n8;i++){
	int t1=bitrev[bit++];
	int t2=bitrev[bit++];

	float wA=w[t1]-w[t2+1];
	float wB=w[t1-1]+w[t2];
	float wC=w[t1]+w[t2+1];
	float wD=w[t1-1]-w[t2];

	float wACE=wA* trig[C];
	float wBCE=wB* trig[C++];
	float wACO=wA* trig[C];
	float wBCO=wB* trig[C++];
      
	x[x1++]=( wC+wACO+wBCE)*.5f;
	x[x2--]=(-wD+wBCO-wACE)*.5f;
	x[x1++]=( wD+wBCO-wACE)*.5f; 
	x[x2--]=( wC-wACO-wBCE)*.5f;
      }
    }
    return(x);
  }
}