一元多项式的存储和其算法运算

      线性结构可以存储一元多项式。一元多项式是由n + 1 个系数唯一确定，那么我可以将系数存放在线性表中，指数隐含在线性表的序号中。即线性表第一个元素存放指数0 的系数，线性表第二个元素存放指数1 的系数。。。以此类推线性表第n 个元素存放指数n - 1的系数。然而这样的存储方式存在一个空间浪费的问题，当指数有20 000 时但是多项式却只有三项，但我们却需要存储20 001项数据。

      那么我们该如何进行改进呢？我们可以通过两张线性表进行存储一元多项式即可。即一张线性表存放指数，另一张线性表存放系数。其实这很像Java Map容器，它是以键值对存储的，我们可以看做指数是键值，系数是值。由于指数和系数都是整型如果我们使用Map 作为存储使用起来并不方便，因此我们可以创建一个Item 类，代码如下：

public class Item {
		
		private int exponential;
		
		private int coefficient;

		public Item(int exponential, int coefficient) {
			this.exponential = exponential;
			this.coefficient = coefficient;
		}
		
		public int getCoefficient() {
			return coefficient;
		}

		public void setCoefficient(int coefficient) {
			this.coefficient = coefficient;
		}

		public int getExponential() {
			return exponential;
		}

		public void setExponential(int exponential) {
			this.exponential = exponential;
		}
		
		public String toString() {
			return "{" + exponential + ", " + coefficient + "}";
		}
	}

这样在使用起来就比较方便了。接下来我们需要解决的是一元多项式的算法问题，由于算法比较简单这里不做解释，直接将代码附上，请自行理解：）。代码如下：

public class Polynmial {

	public class Item {
		
		private int exponential;
		
		private int coefficient;

		public Item(int exponential, int coefficient) {
			this.exponential = exponential;
			this.coefficient = coefficient;
		}
		
		public int getCoefficient() {
			return coefficient;
		}

		public void setCoefficient(int coefficient) {
			this.coefficient = coefficient;
		}

		public int getExponential() {
			return exponential;
		}

		public void setExponential(int exponential) {
			this.exponential = exponential;
		}
		
		public String toString() {
			return "{" + exponential + ", " + coefficient + "}";
		}
	}
	
	private List<Polynmial.Item> elems = new ArrayList<Polynmial.Item>();
	
	public int size() {
		return elems.size();
	}
	
	public String toString() {
		StringBuffer buf = new StringBuffer();
		for(int i = 0; i < elems.size(); i++) {
			if(i == 0) {
				buf.append("{");
			}else {
				buf.append(", ");
			}
			buf.append(elems.get(i));
		}
		buf.append("}");
		return buf.toString();		
	}
	
	public Polynmial add(int exponential, int coefficient) {
		Item elem = new Item(exponential, coefficient);
		add(elem);
		return this;
	}
	
	public Polynmial remove(int exponential) {
		for(Item elem: elems) {
			if(elem.getExponential() == exponential) {
				elems.remove(elem);
				return this;
			}
		}
		return this;
	}
	
	private void add(Item elem) {
		if(elems.size() == 0) {
			elems.add(0, elem);
			return;
		}
		for(int i = 0; i < elems.size(); i++) {
			Item e = elems.get(i);
			if(elem.getExponential() < e.getExponential()) {
				elems.add(i, elem);
				return;
			}
		}
		elems.add(elem);
	}
	
	public Polynmial plus(Polynmial p) {
		int 
			idx1 = 0,
			idx2 = 0;
		while(idx1 < elems.size() && idx2 < p.elems.size()) {
			Item 
				e1 = elems.get(idx1),
				e2 = p.elems.get(idx2);
			if(e1.getExponential() < e2.getExponential()) {
				++idx1;
			}else if(e1.getExponential() == e2.getExponential()) {
				int coefficient = e1.getCoefficient() + e2.getCoefficient();
				if(coefficient == 0) {
					remove(idx1);
				}else {
					e1.setCoefficient(coefficient);
				}
				++idx1;
				++idx2;
			}else {
				add(e2);
				++idx2;
			}	
		}
		while(idx2 < p.elems.size()) {
			Item e2 = p.elems.get(idx2);
			add(e2);
			++idx2;
		}
		return this;
	}
	
	public Polynmial minus(Polynmial p) {
		int 
			idx1 = 0,
			idx2 = 0;
		while(idx1 < elems.size() && idx2 < p.elems.size()) {
			Item 
				e1 = elems.get(idx1),
				e2 = p.elems.get(idx2);
			if(e1.getExponential() < e2.getExponential()) {
				++idx1;
			}else if(e1.getExponential() == e2.getExponential()) {
				int coefficient = e1.getCoefficient() - e2.getCoefficient();
				if(coefficient == 0) {
					remove(idx1);
				}else {
					e1.setCoefficient(coefficient);
				}
				++idx1;
				++idx2;
			}else {
				e2.setCoefficient(-e2.getCoefficient());
				add(e2);
				++idx2;
			}	
		}
		while(idx2 < p.elems.size()) {
			Item e2 = p.elems.get(idx2);
			e2.setCoefficient(-e2.getCoefficient());
			add(e2);
			++idx2;
		}
		return this;
	}
}