java 解析html标签_Java解析HTML标签

#基于状态机思想实现解析HTML TAG ，后续有时间的话尝试写一个JSON解析工具。

import java.util.HashMap;

import java.util.Map;

public class TagParser {

private Integer index = 0;

private char[] tagChar;

private int position = 0;

/* 解析符号 */

private char symbol = '"';

public static final String START_SCRIPT = "

public static final String END_SCRIPT = ">";

public static final String END_SCRIPT_1 = "/>";

public static final String EQ = "=";

public static final char SPACE = ' ';

public static final String MUST_SPACE = " ";

public static final String TAB = "";

public static final String SYMBOL = "'";

private DFAStatus status;

private Map map = new HashMap();

private Map result = new HashMap();

private String tagName;

public String getTagName() {

return this.tagName;

}

public static void main(String[] args) throws SymbolError {

String tag = "";

int i = 0;

long start = System.currentTimeMillis();

while (i < 2) {

TagParser token = new TagParser(tag);

token.parser();

System.out.println(token.getAttr("src") + "|" + token.getTagName());

i++;

}

System.out.println("use time:" + (System.currentTimeMillis() - start));

}

private boolean startsWith(String str) {

char[] c