Pattern matching

[Mary Cole]

Hi,

I have a string lets say "ABC@^*NNNNN@^*sdsdsds@^*dsdsds"
I have to split this string matching @^*. I can use String Tokenizer to split it.
But if the string is "AB*C@^*NNNNN@^*sdsdsds@^*dsdsds" (note the *), it fails.I need the result as

AB*C
NNNNN
sdsdsds
dsdsds

How do I do it....Pls advice

[Mary Cole]

any luck guys
[ July 14, 2006: Message edited by: Mary Cole ]

[Jim Yingst]

[Mary]: I have to split this string matching @^*. I can use String Tokenizer to split it.

No, StringTokenizer doesn't work the way you think it does. If you tell StringTokenizer to use "@^*" for the delimiter, that means it will use @, ^, or * as delimiters. There's no way to tell StringTokenizer to use the whole string as a single delimiter - it's always a list of possible single-character delimiters. This is typically not what users expect, and that's one reason by StringTokenizer is not widely recommended these days. An alternative is to use the split() method in String (available since JDK 1.4). Unfortunately this has some issues too, as you need to know about regular expressions to use this correctly. The problem is that ^ and * have special meaning in regular expressions, and to refer to literal ^ and *, you need special escape sequences. Here are three options:

String[] parts = input.split("@\\^\\*");

String[] parts = input.split(Pattern.quote("@^*"));

String[] parts = input.split("\\Q@^*\\E");

Or if regular expressions aren't to your taste, you could just use String's indexOf() method to find the position of each @^*, then use substring() to extract the other bits.
[ July 14, 2006: Message edited by: Jim Yingst ]