各种形式的日期格式标准化

冰a冰 2021-03-30 03:08:40

在近期的文本解析过程中需要提取文本内的日期并规范格式，常用格式为“yyyy-MM-dd”。
由于写法各异，无法用Date函数或SimpleDateFormat函数格式化，所以自己实现了一个方法。
经过测试后凡是带有分隔符的日期都能准确提取，现在不知道是否完善，代码如下：

// 个位数的月份正则

static Pattern month = Pattern.compile("(?<=-)\\d(?=-)");

// 个位数的日/号正则

static Pattern day = Pattern.compile("(?<=-)\\d$");

public static String formatDate(String sourceDate) {

    sourceDate = sourceDate.replaceAll("[年月.．/—]","-")

            .replaceAll("[号日]$", "")

            .replaceAll("[〇零ＯOo０]","0")

            .replace("一","1")

            .replace("二","2")

            .replace("三","3")

            .replace("四","4")

            .replace("五","5")

            .replace("六","6")

            .replace("七","7")

            .replace("八","8")

            .replace("九","9")

            .replace("廿","2");

    // 十月的十转换为 10

    sourceDate = sourceDate.replaceAll("(?<=-)十(?=-)","10");

    // 十几的十转换为 1

    sourceDate = sourceDate.replaceAll("(?<=-)十(?=\\d)","1");

    // 整十的十转换为 0

    sourceDate = sourceDate.replaceAll("(?<=\\d)十$","0");

    // 几十几的十删掉

    sourceDate = sourceDate.replaceAll("(?<=\\d)十(?=\\d)","");

    // 个位数月份前面加 0

    Matcher matcher = month.matcher(sourceDate);

    if (matcher.find()) {

        sourceDate = sourceDate.substring(0,matcher.start()) + "0" + sourceDate.substring(matcher.start());

    }

    // 个位数日期前面加 0

    matcher = day.matcher(sourceDate);

    if (matcher.find()) {

        sourceDate = sourceDate.substring(0,matcher.start()) + "0" + sourceDate.substring(matcher.start());

    }

    return sourceDate;

}