MIDP1.0的float类实现代码

package henson.midp;

import java.lang.*;

/**
 * <p>Title: Class for float-point calculations in J2ME applications
 * (MIDP 1.0/CLDC 1.0 where float or double types are not available)</p>
 * <p>Description: It makes float-point calculations via integer numbers</p>
 * <p>Copyright: Nikolay Klimchuk Copyright (c) 2002-2005</p>
 * <p>Company: UNTEH</p>
 * <p>License: Free for use only for non-commercial purpose</p>
 * <p>If you want to use all or part of this class for commercial applications
 * then take into account these conditions:</p>
 * <p>1. I need a one copy of your product which includes my class with license
 * key and so on</p>
 * <p>2. Please append my copyright information henson.midp.Float (C) by Nikolay
 * Klimchuk on bout screen of your product</p>
 * <p>3. If you have web site please append link
 * <a href="http://henson.newmail.ru>Nikolay Klimchuk</a>
 * on the page with description of your product</p>
 * <p>That’s all, thank you!</p>
 * @author Nikolay Klimchuk http://henson.newmail.ru
 * @version 1.01
 */

public class Float {
   /** ERROR constant */
   final static private Float ERROR = new Float(Long.MAX_VALUE, Long.MAX_VALUE);
   /** Number of itterations in sqrt method, if you want to make calculations
    * more precise set ITNUM=6,7,… or ITMUN=4,3,… to make it faster */
   final static private int ITNUM = 5;
   /** Square root from 3 */
   final static public Float SQRT3 = new Float(1732050807568877294L, -18L);
   /** The Float value that is closer than any other to pi, the ratio of the
    * circumference of a circle to its diameter */
   final static public Float PI = new Float(3141592653589793238L, -18L);
   /** Zero constant */
   final static public Float ZERO = new Float();
   /** One constant */
   final static public Float ONE = new Float(1L);
   /** The Float value that is closer than any other to e, the base of the
    * natural logarithms */
   final static public Float E = new Float(271828182845904512L, -17);
   /** Log10 constant */
   final static public Float LOG10 = new Float(2302585092994045684L, -18);
   //
   /** Pi/2 constant */
   final static public Float PIdiv2 = PI.Div(2L);
   /** Pi/4 constant */
   final static public Float PIdiv4 = PIdiv2.Div(2L);
   /** Pi/6 constant */
   final static public Float PIdiv6 = PIdiv2.Div(3L);
   /** Pi/12 constant */
   final static public Float PIdiv12 = PIdiv6.Div(2L);
   /** Pi*2 constant */
   final static public Float PImul2 = PI.Mul(2L);
   /** Pi*4 constant */
   final static public Float PImul4 = PI.Mul(4L);
   /** ln(0.5) constant */
   final static public Float LOGdiv2 = new Float( -6931471805599453094L, -19);
   /** Mantissa */
   public long m_Val;
   /** Exponent */
   public long m_E;
   /** Limit of value */
   private long maxLimit = Long.MAX_VALUE / 100;
   //
   /**
    * Create object with zero inside
    */
   public Float() {
      m_Val = m_E = 0;
   }

   /**
    * Create object and makes initialization with value
    * @param value long – value
    */
   public Float(long value) {
      m_Val = value;
      m_E = 0;
   }

   /**
    * Create object and makes initialization both mantissa and exponent
    * @param value long – mantissa
    * @param e long – exponent
    */
   public Float(long value, long e) {
      m_Val = value;
      if (m_Val == 0) {
         m_E = 0;
      }
      else {
         m_E = e;
      }
   }

   /**
    * Create object and makes initialization by other Float object
    * @param value Float – source object
    */
   public Float(Float value) {
      m_Val = value.m_Val;
      if (m_Val == 0) {
         m_E = 0;
      }
      else {
         m_E = value.m_E;
      }
   }

   /**
    * Convert Float object to long number
    * @return long – number
    */
   public long toLong() {
      long tmpE = m_E;
      long tmpVal = m_Val;
      //
      while (tmpE != 0) {
         if (tmpE < 0) {
            tmpVal /= 10;
            tmpE++;
         }
         else {
            tmpVal *= 10;
            tmpE–;
         }
      }
      return tmpVal;
   }

   /**
    * Convert Float object to string without exponent
    * @return String – string
    */
   public String toShortString() {
      if (isError()) {
         return "NaN";
      }
      //
      StringBuffer sb = new StringBuffer();
      sb.append(m_Val);
      int len = (int) m_E;
      if (len > 0) {
         for (int k = 0; k < len; k++) {
            sb.append("0");
         }
         len = 0;
      }
      //
      String str = sb.toString();
      len += str.length();
      //
      if (m_Val < 0L) {
         if (len > 1) {
            return str.substring(0, len);
         }
      }
      else {
         if (len > 0) {
            return str.substring(0, len);
         }
      }
      //
      return "0";
   }

   /**
    * Check value of current Float object is NaN
    * @return boolean – true-if NaN, false-if not
    */
   public boolean isError() {
      return (this.m_Val == ERROR.m_Val && this.m_E == ERROR.m_E);
   }

   /** Convert Float object to string */
   public String toString() {
      if (isError()) {
         return "NaN";
      }
      //
      RemoveZero();
      //
      Long l = new Long(m_Val);
      String str = l.toString();
      int len = str.length();
      boolean neg = false;
      if (m_Val < 0L) {
         neg = true;
         str = str.substring(1, len);
         len–;
      }
      //
      StringBuffer sb = new StringBuffer();
      //
      if (m_E < 0L) {
         int absE = (int) Math.abs(m_E);
         if (absE < len) {
            sb.append(str.substring(0, len – absE));
            sb.append(".");
            sb.append(str.substring(len – absE));
         }
         else {
            sb.append(str);
            for (int i = 0; i < (absE – len); i++) {
               sb.insert(0, "0");
            }
            sb.insert(0, "0.");
         }
      }
      else {
         if (len + m_E > 6) {
            sb.append(str.charAt(0));
            if (str.length() > 1) {
               sb.append(".");
               sb.append(str.substring(1));
            }
            else {
               sb.append(".0");
            }
            sb.append("E" + (len – 1 + m_E));
         }
         else {
            sb.append(str);
            for (int i = 0; i < m_E; i++) {
               sb.append("0");
            }
         }
      }
      //
      str = sb.toString();
      sb = null;
      if (neg) {
         str = "-" + str;
      }
      //
      return str;
   }

   /**
    * Append value of argument to value of current Float object and return as
    * new Float object
    * @param value Float – argument
    * @return Float – current+value
    */
   public Float Add(Float value) {
      if (value.Equal(ZERO)) {
         return new Float(this);
      }
      //
      long e1 = m_E;
      long e2 = value.m_E;
      long v1 = m_Val;
      long v2 = value.m_Val;
      // E must be equal in both operators
      while (e1 != e2) {
         if (e1 > e2) {
            if (Math.abs(v1) < maxLimit) {
               v1 *= 10;
               e1–;
            }
            else {
               v2 /= 10;
               e2++;
            }
         }
         else
         if (e1 < e2) {
            if (Math.abs(v2) < maxLimit) {
               v2 *= 10;
               e2–;
            }
            else {
               v1 /= 10;
               e1++;
            }
         }
      }
      //
      if ( (v1 > 0 && v2 > Long.MAX_VALUE – v1) ||
          (v1 < 0 && v2 < Long.MIN_VALUE – v1)) {
         v1 /= 10;
         e1++;
         v2 /= 10;
         e2++;
      }
      //
      if (v1 > 0 && v2 > Long.MAX_VALUE – v1) {
         return new Float(ERROR);
      }
      else
      if (v1 < 0 && v2 < Long.MIN_VALUE – v1) {
         return new Float(ERROR);
      }
      //
      return new Float(v1 + v2, e1);
   }

   /**
    * Subtract value of argument from value of current Float object and return
    * as new Float object
    * @param value Float – argument
    * @return Float – current-value
    */
   public Float Sub(Float value) {
      if (value.Equal(ZERO)) {
         return new Float(m_Val, m_E);
      }
      return Add(new Float( -value.m_Val, value.m_E));
   }

   /**
    * Divide value of current Float object on argument and return as new Float object
    * @param value Float – argument
    * @return Float – current/value
    */
   public Float Mul(long value) {
      return Mul(new Float(value, 0));
   }

   /**
    * Multiply value of current Float object on argument and return as new Float object
    * @param value Float – argument
    * @return Float – current*value
    */
   public Float Mul(Float value) {
      if (value.Equal(ZERO) || this.Equal(ZERO)) {
         return new Float(ZERO);
      }
      if (value.Equal(ONE)) {
         return new Float(this);
      }
      //
      boolean negative1 = (m_Val < 0);
      if (negative1) {
         m_Val = -m_Val;
      }
      boolean negative2 = (value.m_Val < 0);
      if (negative2) {
         value.m_Val = -value.m_Val;
      }
      // Check overflow and underflow
      do {
         if (value.m_Val > m_Val) {
            if (Long.MAX_VALUE / m_Val < value.m_Val) {
               value.m_Val /= 10;
               value.m_E++;
            }
            else {
               break;
            }
         }
         else {
            if (Long.MAX_VALUE / value.m_Val < m_Val) {
               m_Val /= 10;
               m_E++;
            }
            else {
               break;
            }
         }
      }
      while (true);
      //
      if (negative1) {
         m_Val = -m_Val;
      }
      if (negative2) {
         value.m_Val = -value.m_Val;
      }
      //
      long e = m_E + value.m_E;
      long v = m_Val * value.m_Val;
      return new Float(v, e);
   }

   /**
    * Divide value of current Float object on argument and return as new Float object
    * @param value Float – argument
    * @return Float – current/value
    */
   public Float Div(long value) {
      return Div(new Float(value, 0));
   }

   /**
    * Divide value of current Float object on argument and return as new Float object
    * @param value Float – argument
    * @return Float – current/value
    */
   public Float Div(Float value) {
      if (value.Equal(ONE)) {
         return new Float(this);
      }
      //
      long e1 = m_E;
      long e2 = value.m_E;
      long v2 = value.m_Val;
      if (v2 == 0L) {
         return new Float(ERROR);
      }
      long v1 = m_Val;
      if (v1 == 0L) {
         return new Float(ZERO);
      }
      //
      long val = 0L;
      while (true) {
         val += (v1 / v2);
         v1 %= v2;
         if (v1 == 0L || Math.abs(val) > (Long.MAX_VALUE / 10L)) {
            break;
         }
         if (Math.abs(v1) > (Long.MAX_VALUE / 10L)) {
            v2 /= 10L;
            e2++;
         }
         else {
            v1 *= 10L;
            e1–;
         }
         val *= 10L;
      }
      //
      Float f = new Float(val, e1 – e2);
      f.RemoveZero();
      return f;
   }

   public void RemoveZero() {
      if (m_Val == 0) {
         return;
      }
      while (m_Val % 10 == 0) {
         m_Val /= 10;
         m_E++;
      }
   }

   /**
    * Is value of current Float object greater?
    * @param x Float – argument
    * @return boolean – true-if current value is greater x, false-if not
    */
   public boolean Great(Float x) {
      long e1 = m_E;
      long e2 = x.m_E;
      long v1 = m_Val;
      long v2 = x.m_Val;
      //
      while (e1 != e2) {
         if (e1 > e2) {
            if (Math.abs(v1) < maxLimit) {
               v1 *= 10;
               e1–;
            }
            else {
               v2 /= 10;
               e2++;
            }
         }
         else
         if (e1 < e2) {
            if (Math.abs(v2) < maxLimit) {
               v2 *= 10;
               e2–;
            }
            else {
               v1 /= 10;
               e1++;
            }
         }
      }
      //
      return v1 > v2;
   }

   /**
    * Is value of current Float object less?
    * @param x Float – argument
    * @return boolean – true-if current value is less x, false-if not
    */
   public boolean Less(long x) {
      return Less(new Float(x, 0));
   }

   /**
    * Is value of current Float object less?
    * @param x Float – argument
    * @return boolean – true-if current value is less x, false-if not
    */
   public boolean Less(Float x) {
      long e1 = m_E;
      long e2 = x.m_E;
      long v1 = m_Val;
      long v2 = x.m_Val;
      //
      while (e1 != e2) {
         if (e1 > e2) {
            if (Math.abs(v1) < maxLimit) {
               v1 *= 10;
               e1–;
            }
            else {
               v2 /= 10;
               e2++;
            }
         }
         else
         if (e1 < e2) {
            if (Math.abs(v2) < maxLimit) {
               v2 *= 10;
               e2–;
            }
            else {
               v1 /= 10;
               e1++;
            }
         }
      }
      //
      return v1 < v2;
   }

   /**
    * Equal with current Float object?
    * @param x Float – argument
    * @return boolean – true-if equal, false-if not
    */
   public boolean Equal(Float x) {
      long e1 = m_E;
      long e2 = x.m_E;
      long v1 = m_Val;
      long v2 = x.m_Val;
      //
      if ( (v1 == 0 && v2 == 0) || (v1 == v2 && e1 == e2)) {
         return true;
      }
      // Values with exponent differences more than 20 times never could be equal
      /*
           if(Math.abs(e1-e2)>20)
        return false;
       */
      long diff = e1 – e2;
      if (diff < -20 || diff > 20) {
         return false;
      }
      //
      while (e1 != e2) {
         if (e1 > e2) {
            if (Math.abs(v1) < maxLimit) {
               v1 *= 10;
               e1–;
            }
            else {
               v2 /= 10;
               e2++;
            }
         }
         else
         if (e1 < e2) {
            if (Math.abs(v2) < maxLimit) {
               v2 *= 10;
               e2–;
            }
            else {
               v1 /= 10;
               e1++;
            }
         }
      }
      //
      return (v1 == v2);
   }

   /**
    * Reverse sign of value in current Float object and return as new Float object
    * @return Float – new Float object
    */
   public Float Neg() {
      return new Float( -m_Val, m_E);
   }

   /**
    * Returns the trigonometric sine of an angle. Special cases: If the argument
    * is NaN or an infinity, then the result is NaN. If the argument is zero,
    * then the result is a zero with the same sign as the argument. A result must
    * be within 1 ulp of the correctly rounded result. Results must be semi-monotonic
    * @param x Float – an angle, in radians
    * @return Float – the sine of the argument
    */
   static public Float sin(Float x) {
      while (x.Great(PI)) {
         x = x.Sub(PImul2);
      }
      while (x.Less(PI.Neg())) {
         x = x.Add(PImul2);
      }
      // x*x*x
      Float m1 = x.Mul(x.Mul(x));
      Float q1 = m1.Div(6L);
      // x*x*x*x*x
      Float m2 = x.Mul(x.Mul(m1));
      Float q2 = m2.Div(120L);
      // x*x*x*x*x*x*x
      Float m3 = x.Mul(x.Mul(m2));
      Float q3 = m3.Div(5040L);
      // x*x*x*x*x*x*x*x*x
      Float m4 = x.Mul(x.Mul(m3));
      Float q4 = m4.Div(362880L);
      // x*x*x*x*x*x*x*x*x*x*x
      Float m5 = x.Mul(x.Mul(m4));
      Float q5 = m5.Div(39916800L);
      //
      Float result = x.Sub(q1).Add(q2).Sub(q3).Add(q4).Sub(q5);
      // 1e-6
      if (result.Less(new Float( -999999, -6))) {
         return new Float( -1L);
      }
      // 1e-6
      if (result.Great(new Float(999999, -6))) {
         return new Float(1L);
      }
      // 5e-4
      if (result.Great(new Float( -5, -4)) && result.Less(new Float(5, -4))) {
         return new Float(0L);
      }
      //
      return result;
   }

   /**
    * Returns the trigonometric cosine of an angle. Special cases: If the
    * argument is NaN or an infinity, then the result is NaN. A result must
    * be within 1 ulp of the correctly rounded result. Results must be semi-monotonic
    * @param x Float – an angle, in radians
    * @return Float – the cosine of the argument
    */
   static public Float cos(Float x) {
      while (x.Great(PI)) {
         x = x.Sub(PImul2);
      }
      while (x.Less(PI.Neg())) {
         x = x.Add(PImul2);
      }
      // x*x
      Float m1 = x.Mul(x);
      Float q1 = m1.Div(2L);
      // x*x*x*x
      Float m2 = m1.Mul(m1);
      Float q2 = m2.Div(24L);
      // x*x*x*x*x*x
      Float m3 = m1.Mul(m2);
      Float q3 = m3.Div(720L);
      // x*x*x*x*x*x*x*x
      Float m4 = m2.Mul(m2);
      Float q4 = m4.Div(40320L);
      // x*x*x*x*x*x*x*x*x*x
      Float m5 = m4.Mul(m1);
      Float q5 = m5.Div(3628800L);
      Float result = ONE.Sub(q1).Add(q2).Sub(q3).Add(q4).Sub(q5);
      // 1e-6
      if (result.Less(new Float( -999999, -6))) {
         return new Float( -1L);
      }
      // 1e-6
      if (result.Great(new Float(999999, -6))) {
         return new Float(1L);
      }
      // 5e-4
      if (result.Great(new Float( -5, -4)) && result.Less(new Float(5, -4))) {
         return new Float(0L);
      }
      //
      return result;
   }

   /**
    * Returns the correctly rounded positive square root of a double value.
    * Special cases: If the argument is NaN or less than zero, then the result
    * is NaN. If the argument is positive infinity, then the result is positive
    * infinity. If the argument is positive zero or negative zero, then the
    * result is the same as the argument. Otherwise, the result is the double
    * value closest to the true mathematical square root of the argument value
    * @param x Float – a value
    * @return Float – the positive square root of a. If the argument is NaN or
    * less than zero, the result is NaN
    */
   static public Float sqrt(Float x) {
      int sp = 0;
      boolean inv = false;
      Float a, b;
      //
      if (x.Less(ZERO)) {
         return new Float(ERROR);
      }
      if (x.Equal(ZERO)) {
         return new Float(ZERO);
      }
      if (x.Equal(ONE)) {
         return new Float(ONE);
      }
      // argument less than 1 : invert it
      if (x.Less(ONE)) {
         x = ONE.Div(x);
         inv = true;
      }
      //
      long e = x.m_E / 2;
      // exponent compensation
      Float tmp = new Float(x.m_Val, x.m_E – e * 2);
      // process series of division by 16 until argument is <16
      while (tmp.Great(new Float(16L))) {
         sp++;
         tmp = tmp.Div(16L);
      }
      // initial approximation
      a = new Float(2L);
      // Newtonian algorithm
      for (int i = ITNUM; i > 0; i–) {
         b = tmp.Div(a);
         a = a.Add(b);
         a = a.Div(2L);
      }
      // multiply result by 4 : as much times as divisions by 16 took place
      while (sp > 0) {
         sp–;
         a = a.Mul(4L);
      }
      // exponent compensation
      a.m_E += e;
      // invert result for inverted argument
      if (inv) {
         a = ONE.Div(a);
      }
      return a;
   }

   /**
    * Returns the trigonometric tangent of an angle. Special cases: If the
    * argument is NaN or an infinity, then the result is NaN. If the argument
    * is zero, then the result is a zero with the same sign as the argument.
    * A result must be within 1 ulp of the correctly rounded result. Results
    * must be semi-monotonic
    * @param x Float – an angle, in radians
    * @return Float – the tangent of the argument
    */
   static public Float tan(Float x) {
      Float c = cos(x);
      if (c.Equal(ZERO)) {
         return new Float(ERROR);
      }
      return (sin(x).Div(c));
   }

   /**
    * Returns a new Float object initialized to the value represented by the
    * specified String
    * @param str String – the string to be parsed
    * @param radix int – basement of number
    * @return Float – the Float object represented by the string argument
    */
   static public Float parse(String str, int radix) {
      // Abs
      boolean neg = false;
      if (str.charAt(0) == ‘-‘) {
         str = str.substring(1);
         neg = true;
      }
      //
      int pos = str.indexOf(".");
      long exp = 0;
      // Find exponent position
      int pos2 = str.indexOf(‘E’);
      if (pos2 == -1) {
         pos2 = str.indexOf(‘e’);
      }
      //
      if (pos2 != -1) {
         String tmp = new String(str.substring(pos2 + 1));
         exp = Long.parseLong(tmp);
         str = str.substring(0, pos2);
      }
      //
      if (pos != -1) {
         for (int m = pos + 1; m < str.length(); m++) {
            if (Character.isDigit(str.charAt(m))) {
               exp–;
            }
            else {
               break;
            }
         }
         str = str.substring(0, pos) + str.substring(pos + 1);
         while (str.length() > 1 && str.charAt(0) == ‘0’ &&
                str.charAt(1) != ‘.’) {
            str = str.substring(1);
         }
      }
      //
      long result = 0L;
      int len = str.length();
      //
      StringBuffer sb = new StringBuffer(str);
      while (true) {
         // Long value can’t have length more than 20
         while (len > 20) {
            // Very large number for Long
            sb = sb.deleteCharAt(len – 1);
            // Compensation of removed zeros
            if (len < pos || pos == -1) {
               exp++;
            }
            //
            len–;
         }
         //
         try {
            result = Long.parseLong(sb.toString(), radix);
            if (neg) {
               result = -result;
            }
            break;
         }
         catch (Exception e) {
            // Very large number for Long
            sb = sb.deleteCharAt(len – 1);
            // Compensation of removed zeros
            if (len < pos || pos == -1) {
               exp++;
            }
            //
            len–;
         }
      }
      sb = null;
      //
      Float newValue = new Float(result, exp);
      newValue.RemoveZero();
      return newValue;
   }

   /**
    * Returns the arc cosine of an angle, in the range of 0.0 through pi.
    * Special case: If the argument is NaN or its absolute value is greater than 1,
    * then the result is NaN. A result must be within 1 ulp of the correctly
    * rounded result. Results must be semi-monotonic
    * @param x Float – the value whose arc cosine is to be returned
    * @return Float – the arc cosine of the argument
    */
   static public Float acos(Float x) {
      Float f = asin(x);
      if (f.isError()) {
         return f;
      }
      return PIdiv2.Sub(f);
   }

   /**
    * Returns the arc sine of an angle, in the range of -pi/2 through pi/2.
    * Special cases: If the argument is NaN or its absolute value is greater
    * than 1, then the result is NaN. If the argument is zero, then the result
    * is a zero with the same sign as the argument. A result must be within 1
    * ulp of the correctly rounded result. Results must be semi-monotonic
    * @param x Float – the value whose arc sine is to be returned
    * @return Float – the arc sine of the argument
    */
   static public Float asin(Float x) {
      if (x.Less(ONE.Neg()) || x.Great(ONE)) {
         return new Float(ERROR);
      }
      if (x.Equal(ONE.Neg())) {
         return PIdiv2.Neg();
      }
      if (x.Equal(ONE)) {
         return PIdiv2;
      }
      return atan(x.Div(sqrt(ONE.Sub(x.Mul(x)))));
   }

   /**
    * Returns the arc tangent of an angle, in the range of -pi/2 through pi/2.
    * Special cases: If the argument is NaN, then the result is NaN. If the
    * argument is zero, then the result is a zero with the same sign as the
    * argument. A result must be within 1 ulp of the correctly rounded result.
    * Results must be semi-monotonic
    * @param x Float – the value whose arc tangent is to be returned
    * @return Float – the arc tangent of the argument
    */
   static public Float atan(Float x) {
      boolean signChange = false;
      boolean Invert = false;
      int sp = 0;
      Float x2, a;
      // check up the sign change
      if (x.Less(ZERO)) {
         x = x.Neg();
         signChange = true;
      }
      // check up the invertation
      if (x.Great(ONE)) {
         x = ONE.Div(x);
         Invert = true;
      }
      // process shrinking the domain until x<PI/12
      while (x.Great(PIdiv12)) {
         sp++;
         a = x.Add(SQRT3);
         a = ONE.Div(a);
         x = x.Mul(SQRT3);
         x = x.Sub(ONE);
         x = x.Mul(a);
      }
      // calculation core
      x2 = x.Mul(x);
      a = x2.Add(new Float(14087812, -7));
      a = new Float(55913709, -8).Div(a);
      a = a.Add(new Float(60310579, -8));
      a = a.Sub(x2.Mul(new Float(5160454, -8)));
      a = a.Mul(x);
      // process until sp=0
      while (sp > 0) {
         a = a.Add(PIdiv6);
         sp–;
      }
      // invertation took place
      if (Invert) {
         a = PIdiv2.Sub(a);
      }
      // sign change took place
      if (signChange) {
         a = a.Neg();
      }
      //
      return a;
   }

   /**
    * Converts rectangular coordinates (x,爕) to polar (r,爐heta). This method
    * computes the phase theta by computing an arc tangent of y/x in the range
    * of -pi to pi. Special cases: If either argument is NaN, then the result is
    * NaN. If the first argument is positive zero and the second argument is
    * positive, or the first argument is positive and finite and the second
    * argument is positive infinity, then the result is positive zero. If the
    * first argument is negative zero and the second argument is positive, or
    * the first argument is negative and finite and the second argument is positive
    * infinity, then the result is negative zero. If the first argument is positive
    * zero and the second argument is negative, or the first argument is positive
    * and finite and the second argument is negative infinity, then the result is
    * the double value closest to pi. If the first argument is negative zero and
    * the second argument is negative, or the first argument is negative and finite
    * and the second argument is negative infinity, then the result is the double
    * value closest to -pi. If the first argument is positive and the second
    * argument is positive zero or negative zero, or the first argument is positive
    * infinity and the second argument is finite, then the result is the double
    * value closest to pi/2. If the first argument is negative and the second
    * argument is positive zero or negative zero, or the first argument is negative
    * infinity and the second argument is finite, then the result is the double
    * value closest to -pi/2. If both arguments are positive infinity, then the
    * result is the double value closest to pi/4. If the first argument is positive
    * infinity and the second argument is negative infinity, then the result is
    * the double value closest to 3*pi/4. If the first argument is negative
    * infinity and the second argument is positive infinity, then the result
    * is the double value closest to -pi/4. If both arguments are negative infinity,
    * then the result is the double value closest to -3*pi/4. A result must be
    * within 2 ulps of the correctly rounded result. Results must be semi-monotonic
    * @param y Float – the ordinate coordinate
    * @param x Float – the abscissa coordinate
    * @return Float – the theta component of the point (r,爐heta) in polar
    * coordinates that corresponds to the point (x,爕) in Cartesian coordinates
    */
   static public Float atan2(Float y, Float x) {
      // if x=y=0
      if (y.Equal(ZERO) && x.Equal(ZERO)) {
         return new Float(ZERO);
      }
      // if x>0 atan(y/x)
      if (x.Great(ZERO)) {
         return atan(y.Div(x));
      }
      // if x<0 sign(y)*(pi – atan(|y/x|))
      if (x.Less(ZERO)) {
         if (y.Less(ZERO)) {
            return Float.PI.Sub(atan(y.Div(x))).Neg();
         }
         else {
            return Float.PI.Sub(atan(y.Div(x).Neg()));
         }
      }
      // if x=0 y!=0 sign(y)*pi/2
      if (y.Less(ZERO)) {
         return PIdiv2.Neg();
      }
      else {
         return new Float(PIdiv2);
      }
   }

   // precise
   // x=-35 diff=1.48%
   // x=-30 diff=0.09%
   // x=30 diff=0.09%
   // x=31 diff=0.17%
   // x=32 diff=0.31%
   // x=33 diff=0.54%
   // x=34 diff=0.91%
   // x=35 diff=1.46%
   /**
    * Returns Euler’s number e raised to the power of a double value. Special
    * cases: If the argument is NaN, the result is NaN. If the argument is
    * positive infinity, then the result is positive infinity. If the argument
    * is negative infinity, then the result is positive zero. A result must be
    * within 1 ulp of the correctly rounded result. Results must be semi-monotonic
    * @param x Float – the exponent to raise e to
    * @return Float – the value e^x, where e is the base of the natural logarithms
    */
   static public Float exp(Float x) {
      if (x.Equal(ZERO)) {
         return new Float(ONE);
      }
      //
      Float f = new Float(ONE);
      long d = 1;
      Float k = null;
      boolean isless = x.Less(ZERO);
      if (isless) {
         x = x.Neg();
      }
      k = new Float(x).Div(d);
      //
      for (long i = 2; i < 50; i++) {
         f = f.Add(k);
         k = k.Mul(x).Div(i);
      }
      //
      if (isless) {
         return ONE.Div(f);
      }
      else {
         return f;
      }
   }

   // precise
   // x=25 diff=0.12%
   // x=30 diff=0.25%
   // x=35 diff=0.44%
   // x=40 diff=0.67%
   /**
    * Internal log subroutine
    * @param x Float
    * @return Float
    */
   static private Float _log(Float x) {
      if (!x.Great(ZERO)) {
         return new Float(ERROR);
      }
      //
      Float f = new Float(ZERO);
      // Make x to close at 1
      int appendix = 0;
      while (x.Great(ZERO) && x.Less(ONE)) {
         x = x.Mul(2);
         appendix++;
      }
      //
      x = x.Div(2);
      appendix–;
      //
      Float y1 = x.Sub(ONE);
      Float y2 = x.Add(ONE);
      Float y = y1.Div(y2);
      //
      Float k = new Float(y);
      y2 = k.Mul(y);
      //
      for (long i = 1; i < 50; i += 2) {
         f = f.Add(k.Div(i));
         k = k.Mul(y2);
      }
      //
      f = f.Mul(2);
      for (int i = 0; i < appendix; i++) {
         f = f.Add(LOGdiv2);
      }
      //
      return f;
   }

   /**
    * Returns the natural logarithm (base e) of a double value. Special cases:
    * If the argument is NaN or less than zero, then the result is NaN. If the
    * argument is positive infinity, then the result is positive infinity. If
    * the argument is positive zero or negative zero, then the result is negative
    * infinity. A result must be within 1 ulp of the correctly rounded result.
    * Results must be semi-monotonic
    * @param x Float – a number greater than 0.0
    * @return Float – the value ln(x), the natural logarithm of x
    */
   static public Float log(Float x) {
      if (!x.Great(ZERO)) {
         return new Float(ERROR);
      }
      //
      if (x.Equal(ONE)) {
         return new Float(ZERO);
      }
      //
      if (x.Great(Float.ONE)) {
         x = ONE.Div(x);
         return _log(x).Neg();
      }
      return _log(x);
   }

   static public Float log10(Float x) {
      if (!x.Great(ZERO)) {
         return new Float(ERROR);
      }
      //
      if (x.Equal(ONE)) {
         return new Float(ZERO);
      }
      //
      Float f = log(x);
      if (f.isError()) {
         return f;
      }
      return f.Div(LOG10);
   }

   /*
     static public Float log10(Float x)
     {
       if(!x.Great(ZERO))
         return new Float(ERROR);
       //
       boolean neg=false;
       if(x.m_Val<0)
       {
         neg=true;
         x.m_Val=-x.m_Val;
       }
       //
       int index=0;
       if(x.Great(Float.ONE))
       {
         // 令朦 1
         while(x.Great(Float.ONE))
         {
           x=x.Div(10);
           index++;
         }
       }
       else
       {
         // 体睃 桦?疣忭?1
         while(x.Less(Float.ONE))
         {
           x=x.Mul(10);
           index–;
         }
       }
       //
       Float res=new Float(index);
       if(!x.Equal(ONE))
         res=res.Add(log(x).Div(LOG10));
       //
       if(neg)
         return Float.ONE.Div(res);
       else
         return res;
     }
    */

   // precise y=3.5
   // x=15 diff=0.06%
   // x=20 diff=0.40%
   // x=25 diff=1.31%
   // x=30 diff=2.95%
   // if x negative y must be integer value
   /**
    * Returns the value of the first argument raised to the power of the
    * second argument. Special cases: If the second argument is positive
    * or negative zero, then the result is 1.0. If the second argument is 1.0,
    * then the result is the same as the first argument. If the second argument is NaN,
    * then the result is NaN. If the first argument is NaN and the second argument
    * is nonzero, then the result is NaN. If the absolute value of the first
    * argument is greater than 1 and the second argument is positive infinity,
    * or the absolute value of the first argument is less than 1 and the second
    * argument is negative infinity, then the result is positive infinity.
    * If the absolute value of the first argument is greater than 1 and the
    * second argument is negative infinity, or the absolute value of the first
    * argument is less than 1 and the second argument is positive infinity,
    * then the result is positive zero. If the absolute value of the first
    * argument equals 1 and the second argument is infinite, then the result is NaN.
    *  If the first argument is positive zero and the second argument is greater
    * than zero, or the first argument is positive infinity and the second argument
    * is less than zero, then the result is positive zero. If the first argument
    * is positive zero and the second argument is less than zero, or the first argument
    * is positive infinity and the second argument is greater than zero, then the
    * result is positive infinity. If the first argument is negative zero and the
    * second argument is greater than zero but not a finite odd integer, or the
    * first argument is negative infinity and the second argument is less than
    * zero but not a finite odd integer, then the result is positive zero.
    * If the first argument is negative zero and the second argument is a positive
    *  finite odd integer, or the first argument is negative infinity and the
    * second argument is a negative finite odd integer, then the result is negative zero.
    * If the first argument is negative zero and the second argument is less than
    * zero but not a finite odd integer, or the first argument is negative infinity
    * and the second argument is greater than zero but not a finite odd integer,
    *  then the result is positive infinity. If the first argument is negative
    * zero and the second argument is a negative finite odd integer, or the first
    * argument is negative infinity and the second argument is a positive finite
    * odd integer, then the result is negative infinity. If the first argument
    * is finite and less than zero if the second argument is a finite even integer,
    *  the result is equal to the result of raising the absolute value of the first
    * argument to the power of the second argument if the second argument is
    * a finite odd integer, the result is equal to the negative of the result of
    * raising the absolute value of the first argument to the power of the second
    * argument if the second argument is finite and not an integer, then the result
    * is NaN. If both arguments are integers, then the result is exactly equal to
    * the mathematical result of raising the first argument to the power of the
    * second argument if that result can in fact be represented exactly as a
    * double value. (In the foregoing descriptions, a floating-point value
    * is considered to be an integer if and only if it is finite and a fixed
    * point of the method ceil or, equivalently, a fixed point of the method floor.
    * A value is a fixed point of a one-argument method if and only if the result
    * of applying the method to the value is equal to the value.) A result must be
    * within 1 ulp of the correctly rounded result. Results must be semi-monotonic.
    * @param x Float – the base
    * @param y Float – the exponent
    * @return Float – the value a^b
    */
   static public Float pow(Float x, Float y) {
      if (x.Equal(ZERO)) {
         return new Float(ZERO);
      }
      if (x.Equal(ONE)) {
         return new Float(ONE);
      }
      if (y.Equal(ZERO)) {
         return new Float(ONE);
      }
      if (y.Equal(ONE)) {
         return new Float(x);
      }
      //
      long l = y.toLong();
      boolean integerValue = y.Equal(new Float(l));
      //
      if (integerValue) {
         boolean neg = false;
         if (y.Less(0)) {
            neg = true;
         }
         //
         Float result = new Float(x);
         for (long i = 1; i < (neg ? -l : l); i++) {
            result = result.Mul(x);
         }
         //
         if (neg) {
            return ONE.Div(result);
         }
         else {
            return result;
         }
      }
      else {
         if (x.Great(ZERO)) {
            return exp(y.Mul(log(x)));
         }
         else {
            return new Float(ERROR);
         }
      }
   }

   /**
    * Returns the smallest (closest to negative infinity) double value that is not
    * less than the argument and is equal to a mathematical integer. Special cases:
    * If the argument value is already equal to a mathematical integer, then the
    * result is the same as the argument. If the argument is NaN or an infinity
    * or positive zero or negative zero, then the result is the same as the
    * argument. If the argument value is less than zero but greater than -1.0,
    * then the result is negative zero
    * @param x Float – a value
    * @return Float – the smallest (closest to negative infinity) floating-point
    * value that is not less than the argument and is equal to a mathematical integer
    */
   static public Float ceil(Float x) {
      long tmpVal = x.m_Val;
      //
      if (x.m_E < 0) {
         long coeff = 1;
         //
         if (x.m_E > -19) {
            for (long i = 0; i < -x.m_E; i++) {
               coeff *= 10;
            }
            tmpVal /= coeff;
            tmpVal *= coeff;
            if (x.m_Val – tmpVal > 0) {
               tmpVal += coeff;
            }
         }
         else
         if (tmpVal > 0) {
            return ONE;
         }
         else {
            return ZERO;
         }
      }
      //
      return new Float(tmpVal, x.m_E);
   }

   /**
    * Returns the largest (closest to positive infinity) double value that is not
    * greater than the argument and is equal to a mathematical integer. Special
    * cases: If the argument value is already equal to a mathematical integer,
    * then the result is the same as the argument. If the argument is NaN or
    * an infinity or positive zero or negative zero, then the result is the same
    * as the argument
    * @param x Float – a value
    * @return Float – the largest (closest to positive infinity) floating-point
    * value that is not greater than the argument and is equal to a mathematical integer
    */
   static public Float floor(Float x) {
      long tmpVal = x.m_Val;
      //
      if (x.m_E < 0) {
         long coeff = 1;
         //
         if (x.m_E > -19) {
            for (long i = 0; i < -x.m_E; i++) {
               coeff *= 10;
            }
            tmpVal /= coeff;
            tmpVal *= coeff;
            if (x.m_Val – tmpVal < 0) {
               tmpVal -= coeff;
            }
         }
         else
         if (tmpVal < 0) {
            return ONE.Neg();
         }
         else {
            return ZERO;
         }
      }
      //
      return new Float(tmpVal, x.m_E);
   }

   /**
    * Returns the absolute value of a Float object. If the argument is not negative,
    * the argument is returned. If the argument is negative, the negation of the
    * argument is returned. Special cases: If the argument is positive zero or
    * negative zero, the result is positive zero. If the argument is infinite,
    * the result is positive infinity. If the argument is NaN, the result is NaN
    * @param x Float – the argument whose absolute value is to be determined
    * @return Float – the absolute value of the argument
    */
   static public Float abs(Float x) {
      if (x.m_Val < 0) {
         return x.Neg();
      }
      return new Float(x);
   }

   /**
    * Integer part of Float object
    * @param x Float – source Float object
    * @return Float – result Float object
    */
   static public Float Int(Float x) {
      long tmpVal = x.m_Val;
      //
      if (x.m_E < 0) {
         long coeff = 1;
         //
         if (x.m_E > -19) {
            for (long i = 0; i < -x.m_E; i++) {
               coeff *= 10;
            }
            tmpVal /= coeff;
            tmpVal *= coeff;
         }
         else {
            return Float.ZERO;
         }
      }
      //
      return new Float(tmpVal, x.m_E);
   }

   /**
    * Fractional part of Float object
    * @param x Float – source Float object
    * @return Float – result Float object
    */
   static public Float Frac(Float x) {
      return x.Sub(Int(x));
   }

   /**
    * Converts an angle measured in degrees to an approximately equivalent
    * angle measured in radians. The conversion from degrees to radians is generally inexact
    * @param x Float – an angle, in degrees
    * @return Float – the measurement of the angle x in radians
    */
   static public Float toRadians(Float x) {
      return x.Mul(PI).Div(180L);
   }

   /**
    * Converts an angle measured in radians to an approximately equivalent
    * angle measured in degrees. The conversion from radians to degrees is generally inexact;
    * users should not expect cos(toRadians(90.0)) to exactly equal 0.0
    * @param x Float – an angle, in radians
    * @return Float – the measurement of the angle angrad in degrees
    */
   static public Float toDegrees(Float x) {
      return x.Mul(180L).Div(PI);
   }
}