regex - JAVA: Regexp on user-defined tokens -

i need run regexp on set of user-defined tokens.

for example, i've string this: tok3 tok1 tok2 tok2 tok4 tok3 // example string

and using regexp this: (tok1|tok2)+ // regexp

i'd capture sequence of tokens: tok1 tok2 tok2 in example string.

now, regexp work on sequence of characters, problem different in sense tokens not characters strings. tokens composed 2 or more characters. furthermore, software should able detect regexp in example matches string @ position (1, 4).

for moment, solved problem mapping each token char in ascii alphabet , running regexp after removing spaces.

however, i'm not satsfied solution , wondering if there better one. thanks!

edit

spaces in regexp needed separate tokens. don't mean spaces mandatory between tokens.

how storing positions of spaces , using translate string position token position?

far elegant straight regex, it's idea.

treemap<integer, integer> spaces = new treemap<integer, integer>(); string regex = "(?<=^| )((tok1|tok2)( |$))+"; string str = "tok3 tok1 tok2 tok2 tok4 tok3"; int c = 0; spaces.put(0, 0); (int = 0; < str.length(); i++) {    if (str.charat(i) == ' ')      spaces.put(i, ++c); } pattern p = pattern.compile(regex); matcher m = p.matcher(str); while (m.find()) {    system.out.println(m.group());    system.out.println("start = " + spaces.floorentry(m.start()).getvalue());    system.out.println("finish = " + spaces.floorentry(m.end()).getvalue()); } 

another option string.split:

string str = "tok3 tok1 tok2 tok2 tok4 tok3"; string[] arr = str.split(" "); // maybe consider using \\s or \\s+ instead int start = -1; string match = ""; (int = 0; < arr.length; i++) {    if (arr[i].matches("(tok1|tok2)"))    {       if (start == -1)          start = i;       match += ((match.length() != 0) ? " " : "") + arr[i];    }    else if (start != -1)    {       system.out.println(match);       system.out.println("start = " + start);       system.out.println("finish = " + i);       match = "";       start = -1;    } }