Monday, August 22, 2011

Java Tips - Parsing Strings into Dates

The DateFormat class has some additional methods, notably parse( ) , which tries to parse a string according to the format stored in the given DateFormat object:
// DateParse1.java 

SimpleDateFormat formatter = new SimpleDateFormat ("yyyy-MM-dd"); 

String input = args.length == 0 ? "1818-11-11" : args[0]; 

System.out.print(input + " parses as "); 

Date t; 

try { 

    t = formatter.parse(input); 

    System.out.println(t); 

} catch (ParseException e) { 

    System.out.println("unparseable using " + formatter); 

}

This program parses any date back to Year Zero and well beyond Year 2000.
What if the date is embedded in an input string? You could, of course, use the string's substring( ) method to extract it, but there is an easier way. The ParsePosition object from java.text is designed to represent (and track) the position of an imaginary cursor in a string. Suppose we have genealogical data with input strings representing the times of a person's life:
BD: 1913-10-01 Vancouver, B.C. 

DD: 1983-06-06 Toronto, ON

This lists one person's birth date (BD) and place, and death date (DD) and place. We can parse these using String.indexOf(' ') to find the space after the : character, DateFormat parse( ) to parse the date, and String.substring( ) to get the city and other geographic information. Here's how:
// DateParse2.java 

SimpleDateFormat formatter = 

    new SimpleDateFormat ("yyyy-MM-dd"); 

String input[] = {  

    "BD: 1913-10-01 Vancouver, B.C.", 

    "MD: 1948-03-01 Ottawa, ON", 

    "DD: 1983-06-06 Toronto, ON" }; 

for (int i=0; i<input.length; i++) { 

    String aLine = input[i]; 

    String action; 

    switch(aLine.charAt(0)) { 

        case 'B': action = "Born"; break; 

        case 'M': action = "Married"; break; 

        case 'D': action = "Died"; break; 

        // others... 

        default: System.err.println("Invalid code in " + aLine); 

        continue; 

    } 

    int p = aLine.indexOf(' '); 

    ParsePosition pp = new ParsePosition(p); 

    Date d = formatter.parse(aLine, pp); 

    if (d == null) { 

        System.err.println("Invalid date in " + aLine); 

        continue; 

    } 

    String location = aLine.substring(pp.getIndex( )); 

    System.out.println( 

        action + " on " + d + " in " + location); 

}

This works like I said it would:
Born on Wed Oct 01 00:00:00 PDT 1913 in  Vancouver, B.C. 

Married on Mon Mar 01 00:00:00 PST 1948 in  Ottawa, ON 

Died on Mon Jun 06 00:00:00 PDT 1983 in  Toronto, ON

No comments:

Post a Comment