Matching '_' and '-' in java regexes

Question

I had this regex in java that matched either an alphanumeric character or the tilde (~)

^([a-z0-9])+|~$

Now I have to add also the characters - and _ I've tried a few combinations, neither of which work, for example:

^([a-zA-Z0-9_-])+|~$ ^([a-zA-Z0-9]|-|_)+|~$

Sample input strings that must match:

woZOQNVddd

00000

ncnW0mL14-

dEowBO_Eu7

7MyG4XqFz-

A8ft-y6hDu ~

Any clues / suggestion?

Answer 1

You need to escape the -, like \-, since it is a special character (the range operator). _ is ok.

So ^([a-z0-9_\-])+|~$.

Edit: your last input String will not match because the regular expression you are using matches a string of alphanumeric characters (plus - and _) OR a tilde (because of the pipe). But not both. If you want to allow an optional tilde on the end, change to:

^([a-z0-9_\-])+(~?)$

Answer 2

If you put the - first, it won't be interpreted as the range indicator.

^([-a-zA-Z0-9_])+|~$

This matches all of your examples except the last one using the following code:

String str = "A8ft-y6hDu ~";
System.out.println("Result: " + str.matches("^([-a-zA-Z0-9_])+|~$"));

That last example won't match because it doesn't fit your description. The regex will match any combination of alphanumerics, -, and _, OR a ~ character.

Answer 3

- is a special character within square brackets. It indicates a range. If it's not at either end of the regex it needs to be escaped by putting a \ before it.

It's worth pointing out a shortcut: \w is equivalent to [0-9a-zA-Z_] so I think this is more readable:

^([\w-]+|~$