How to convert HH:mm:ss.SSS to milliseconds?

Question

I have a String 00:01:30.500 which is equivalent to 90500 milliseconds. I tried using SimpleDateFormat which give milliseconds including current date. I just need that String representation to milliseconds. Do I have to write custom method, which will split and calculate milliseconds? or Is there any other way to do this? Thanks.

I have tried as follows:

        String startAfter = "00:01:30.555";         SimpleDateFormat dateFormat = new SimpleDateFormat("HH:mm:ss.SSS");         Date date = dateFormat.parse(startAfter);         System.out.println(date.getTime()); 

Answer 1

You can use SimpleDateFormat to do it. You just have to know 2 things.

1. All dates are internally represented in UTC
2. .getTime() returns the number of milliseconds since 1970-01-01 00:00:00 UTC.

package se.wederbrand.milliseconds;  import java.text.SimpleDateFormat; import java.util.Date; import java.util.TimeZone;  public class Main {             public static void main(String[] args) throws Exception {         SimpleDateFormat sdf = new SimpleDateFormat("yyyy-MM-dd HH:mm:ss.SSS");         sdf.setTimeZone(TimeZone.getTimeZone("UTC"));          String inputString = "00:01:30.500";          Date date = sdf.parse("1970-01-01 " + inputString);         System.out.println("in milliseconds: " + date.getTime());             } } 

Answer 2

If you want to parse the format yourself you could do it easily with a regex such as

private static Pattern pattern = Pattern.compile("(\\d{2}):(\\d{2}):(\\d{2}).(\\d{3})");  public static long dateParseRegExp(String period) {     Matcher matcher = pattern.matcher(period);     if (matcher.matches()) {         return Long.parseLong(matcher.group(1)) * 3600000L              + Long.parseLong(matcher.group(2)) * 60000              + Long.parseLong(matcher.group(3)) * 1000              + Long.parseLong(matcher.group(4));      } else {         throw new IllegalArgumentException("Invalid format " + period);     } } 

However, this parsing is quite lenient and would accept 99:99:99.999 and just let the values overflow. This could be a drawback or a feature.