java：java.util.StringTokenizer实现字符串切割

java：java.util.StringTokenizer实现字符串切割

1 前言

java.util工具包提供了字符串切割的工具类StringTokenizer，Spring等常见框架的字符串工具类（如Spring的StringUtils），常见此类使用。

例如Spring的StringUtils下的方法：

public static String[] tokenizeToStringArray(
		@Nullable String str, String delimiters, boolean trimTokens, boolean ignoreEmptyTokens) {

	if (str == null) {
		return EMPTY_STRING_ARRAY;
	}

	StringTokenizer st = new StringTokenizer(str, delimiters);
	List<String> tokens = new ArrayList<>();
	while (st.hasMoreTokens()) {
		String token = st.nextToken();
		if (trimTokens) {
			token = token.trim();
		}
		if (!ignoreEmptyTokens || token.length() > 0) {
			tokens.add(token);
		}
	}
	return toStringArray(tokens);
}

又如定时任务框架Quartz中，cron表达式类CronExpression，其中的buildExpression方法是为了处理cron表达式的，cron表达式有7个子表达式，空格隔开，cron表达式字符串的切割也使用到了StringTokenizer类，方法如下：

protected void buildExpression(String expression) throws ParseException {
    this.expressionParsed = true;

    try {
        if (this.seconds == null) {
            this.seconds = new TreeSet();
        }

        if (this.minutes == null) {
            this.minutes = new TreeSet();
        }

        if (this.hours == null) {
            this.hours = new TreeSet();
        }

        if (this.daysOfMonth == null) {
            this.daysOfMonth = new TreeSet();
        }

        if (this.months == null) {
            this.months = new TreeSet();
        }

        if (this.daysOfWeek == null) {
            this.daysOfWeek = new TreeSet();
        }

        if (this.years == null) {
            this.years = new TreeSet();
        }

        int exprOn = 0;

        for(StringTokenizer exprsTok = new StringTokenizer(expression, " \t", false); exprsTok.hasMoreTokens() && exprOn <= 6; ++exprOn) {
            String expr = exprsTok.nextToken().trim();
            if (exprOn == 3 && expr.indexOf(76) != -1 && expr.length() > 1 && expr.contains(",")) {
                throw new ParseException("Support for specifying 'L' and 'LW' with other days of the month is not implemented", -1);
            }

            if (exprOn == 5 && expr.indexOf(76) != -1 && expr.length() > 1 && expr.contains(",")) {
                throw new ParseException("Support for specifying 'L' with other days of the week is not implemented", -1);
            }

            if (exprOn == 5 && expr.indexOf(35) != -1 && expr.indexOf(35, expr.indexOf(35) + 1) != -1) {
                throw new ParseException("Support for specifying multiple \"nth\" days is not implemented.", -1);
            }

            StringTokenizer vTok = new StringTokenizer(expr, ",");

            while(vTok.hasMoreTokens()) {
                String v = vTok.nextToken();
                this.storeExpressionVals(0, v, exprOn);
            }
        }

        if (exprOn <= 5) {
            throw new ParseException("Unexpected end of expression.", expression.length());
        } else {
            if (exprOn <= 6) {
                this.storeExpressionVals(0, "*", 6);
            }

            TreeSet<Integer> dow = this.getSet(5);
            TreeSet<Integer> dom = this.getSet(3);
            boolean dayOfMSpec = !dom.contains(NO_SPEC);
            boolean dayOfWSpec = !dow.contains(NO_SPEC);
            if ((!dayOfMSpec || dayOfWSpec) && (!dayOfWSpec || dayOfMSpec)) {
                throw new ParseException("Support for specifying both a day-of-week AND a day-of-month parameter is not implemented.", 0);
            }
        }
    } catch (ParseException var8) {
        throw var8;
    } catch (Exception var9) {
        throw new ParseException("Illegal cron expression format (" + var9.toString() + ")", 0);
    }
}

2 使用

import com.google.common.collect.Lists;

import java.util.List;
import java.util.StringTokenizer;

/**
 * @author xiaoxu
 * @date 2023-10-18
 * spring_boot:com.xiaoxu.boot.tokenizer.TestStringTokenizer
 */
public class TestStringTokenizer {

    public static void main(String[] args) {
        print("你 好 吗\t我是 \t你的\t 朋友 \t", " \t", false);
    }

    public static void print(String str, String delimiter, boolean isReturnDelims) {
        System.out.println("切割字符串：【" + str + "】；" + "分隔符：【" + delimiter + "】。");
        List<String> strs = Lists.newArrayList();
        String s;
        boolean x;
        for (StringTokenizer strToken = new StringTokenizer(str, delimiter, false); strToken.hasMoreTokens(); x = (s != null && strs.add(s))) {
            s = strToken.nextToken();
            System.out.println("切割：【" + s + "】");
            if(s.equals("吗"))
                s = null;
        }
        System.out.println("字符串数组：" + strs);
    }

}

执行结果：

切割字符串：【你 好 吗	我是 	你的	 朋友 	】；分隔符：【 	】。
切割：【你】
切割：【好】
切割：【吗】
切割：【我是】
切割：【你的】
切割：【朋友】
字符串数组：[你, 好, 我是, 你的, 朋友]

源码片段分析：

public StringTokenizer(String str, String delim, boolean returnDelims) {
    currentPosition = 0;
    newPosition = -1;
    delimsChanged = false;
    this.str = str;
    maxPosition = str.length();
    delimiters = delim;
    retDelims = returnDelims;
    setMaxDelimCodePoint();
}

private void setMaxDelimCodePoint() {
    if (delimiters == null) {
        maxDelimCodePoint = 0;
        return;
    }

    int m = 0;
    int c;
    int count = 0;
    for (int i = 0; i < delimiters.length(); i += Character.charCount(c)) {
        c = delimiters.charAt(i);
        if (c >= Character.MIN_HIGH_SURROGATE && c <= Character.MAX_LOW_SURROGATE) {
            c = delimiters.codePointAt(i);
            hasSurrogates = true;
        }
        if (m < c)
            m = c;
        count++;
    }
    maxDelimCodePoint = m;

    if (hasSurrogates) {
        delimiterCodePoints = new int[count];
        for (int i = 0, j = 0; i < count; i++, j += Character.charCount(c)) {
            c = delimiters.codePointAt(j);
            delimiterCodePoints[i] = c;
        }
    }
}

调用setMaxDelimCodePoint()方法，源码可知，切割时设置int maxDelimCodePoint，是为了优化分隔符的检测（取的是分隔字符串中char的ASCII码值最大的字符的ASCII值，存入maxDelimCodePoint中。在方法int scanToken(int startPos)中，若满足条件(c <= maxDelimCodePoint) && (delimiters.indexOf© >= 0)，意即该字符的ASCII码值小于等于最大的maxDelimCodePoint，那么这个字符可能存在于分隔字符串中，再检测delimiters分隔字符串中是否包含该字符，反之，若ASCII码值大于分隔字符串中最大的maxDelimCodePoint，也就是说该字符一定不存在于分隔字符串里，&&直接跳过delimiters.indexOf的检测，也就达到了优化分隔符检测的效果了）。

private int scanToken(int startPos) {
    int position = startPos;
    while (position < maxPosition) {
        if (!hasSurrogates) {
            char c = str.charAt(position);
            if ((c <= maxDelimCodePoint) && (delimiters.indexOf(c) >= 0))
                break;
            position++;
        } else {
            int c = str.codePointAt(position);
            if ((c <= maxDelimCodePoint) && isDelimiter(c))
                break;
            position += Character.charCount(c);
        }
    }
    if (retDelims && (startPos == position)) {
        if (!hasSurrogates) {
            char c = str.charAt(position);
            if ((c <= maxDelimCodePoint) && (delimiters.indexOf(c) >= 0))
                position++;
        } else {
            int c = str.codePointAt(position);
            if ((c <= maxDelimCodePoint) && isDelimiter(c))
                position += Character.charCount(c);
        }
    }
    return position;
}

scanToken方法即跳过分隔字符串，只要某此循环时，该字符包含在分隔字符串里，那么position不再自增，以此时的position值作为实际切割获取字符串的末索引， 因为subString方法是左闭右开的，该值是实际获取字符串的末索引值+1，所以可以截取到完整的不包含分隔符的字符串片段。

skipDelimiters方法类似，即过滤连续包含于分隔字符串中的字符，获取实际需要切割获取的字符串的开始索引值。

private int skipDelimiters(int startPos) {
    if (delimiters == null)
        throw new NullPointerException();

    int position = startPos;
    while (!retDelims && position < maxPosition) {
        if (!hasSurrogates) {
            char c = str.charAt(position);
            if ((c > maxDelimCodePoint) || (delimiters.indexOf(c) < 0))
                break;
            position++;
        } else {
            int c = str.codePointAt(position);
            if ((c > maxDelimCodePoint) || !isDelimiter(c)) {
                break;
            }
            position += Character.charCount(c);
        }
    }
    return position;
}

上述分析可知，只要待切割字符串中的字符，在分隔字符串中出现，那么就会做一次切割（也就是不论分隔字符串中的每个char或字符串片段的顺序，只要连续包含在分隔字符串里，就切割）。

演示如下（注意countTokens()方法不要在循环中和nextToken()一同使用）：

public static void print2(String str, String delimiter, boolean isReturnDelims) {
    StringTokenizer strTokenizer = new StringTokenizer(str, delimiter);
    System.out.println("总数目:" + strTokenizer.countTokens());
    int count;
    String[] strs = new String[count = strTokenizer.countTokens()];
    // 注意：不要在循环里写 int i = 0; i < strTokenizer.countTokens();
    // 因为  countTokens方法需要使用currentPosition，而每次执行nextToken方法时，currentPosition会一直往下偏移计算，
    // 会导致循环中， i < strTokenizer.countTokens();发生改变，这里应该是常量总数目
    for (int i = 0; i < count; i++) {
        String s = strTokenizer.nextToken();
        strs[i] = s;
    }
    System.out.println(Arrays.toString(strs));
}

countTokens源码如下：

public int countTokens() {
    int count = 0;
    int currpos = currentPosition;
    while (currpos < maxPosition) {
        currpos = skipDelimiters(currpos);
        if (currpos >= maxPosition)
            break;
        currpos = scanToken(currpos);
        count++;
    }
    return count;
}

执行：

print2("1a2b3c4ca5bc6ba7abc8acbbaba9", "abc", false);

结果如下所示：

总数目:9
[1, 2, 3, 4, 5, 6, 7, 8, 9]