Java regular expressions

[DBQ]

I have a question about Java regular expressions. Suppose, I specify this string:

HelloHelloHelloHelloHelloHelloHelloHelloHelloHello HelloHelloHelloHelloHello

If I use the regular expression (Hello)* it captures just the last instance of Hello. Is there a way I can capture ALL instances? In reality my string is more complicated. Suppose, I want to find ALL instances that match the pattern H*o and find out what they are?

Thank You.

[slavik]

Are you sure that Java does not return a list of the matches?

to grab all the H*o = ((?:H*o)*), I am not sure what Java RE's non extracting operator is (in Perl, it is ?:

[DBQ]

Java RegEx does not support the conditionals.

[mssever]

I have a question about Java regular expressions. Suppose, I specify this string:

HelloHelloHelloHelloHelloHelloHelloHelloHelloHello HelloHelloHelloHelloHello

If I use the regular expression (Hello)* it captures just the last instance of Hello. Is there a way I can capture ALL instances? In reality my string is more complicated. Suppose, I want to find ALL instances that match the pattern H*o and find out what they are?

Thank You.
Your example pattern, H*o doesn't match your example string at all, unless Java's regexp syntax is quite different from the syntaxes I know. Do you mean you want to match H.*o or (Hello)* ? H.*o matches the entire string, so there'll be only one match.

[DBQ]

I want to match ALL substrings which begin with H and end with o.

[mssever]

Perhaps this Ruby example might push you in the right direction (I don't know Java regexps):

str = "HelloHelloHelloHiHello"
a = [] # array to collect the matches
offset = 0
while offset < str.length
# str[offset..str.length] returns a substring beginning at offset and extending to the
# end of the string
#
# *? is a non-greedy match operator in Ruby
match = str[offset..str.length].match /H.*?o/
if match
a.push match[0]
offset += match.offset(0)[1]
else
break
end
end

puts a.inspect # prints ["Hello", "Hello", "Hello", "HiHello"]

[DBQ]

Perhaps this Ruby example might push you in the right direction (I don't know Java regexps):

str = "HelloHelloHelloHiHello"
a = [] # array to collect the matches
offset = 0
while offset < str.length
# str[offset..str.length] returns a substring beginning at offset and extending to the
# end of the string
#
# *? is a non-greedy match operator in Ruby
match = str[offset..str.length].match /H.*?o/
if match
a.push match[0]
offset += match[0].length
else
break
end
end


*SIGH*. Yes, I know this is very easy in Perl (and aparently in Ruby). However, I am using Java - the regular expressions here are not as powerful it seems. But thanks for the help though.

[mssever]

*SIGH*. Yes, I know this is very easy in Perl (and aparently in Ruby). However, I am using Java - the regular expressions here are not as powerful it seems. But thanks for the help though.
But if you can get one match, you can write a loop to operate on a substring and iteratively collect the matches. Or does Java not have a non-greedy match operator?

EDIT: I fixed a bug in my earlier post related to how I was computing offsets.

[bobrocks]

I have gotten a bit confused over this thread, if you want all occurances of the substring H<something>o then


import java.util.regex.*;

public class RegExTest {

public static void main(String[] args) {

String r = "h(.*?)o";

Pattern p = Pattern.compile(r);
Matcher m = p.matcher("helloheffpdlohellohellohello");

while (m.find()) {
System.out.println("outer : " + m.group(0));
System.out.println("inner : " + m.group(1));
}
}
}



output


outer : hello
inner : ell
outer : heffpdlo
inner : effpdl
outer : hello
inner : ell
outer : hello
inner : ell
outer : hello
inner : ell


should be fine. Java's regex is not bad, it's pretty complete and pretty fast, it's just a little verbose

[DBQ]

Is there a way to generalize this? What if I want to get ALL occurrences of a substring in a string which matches a certain regex?

[DBQ]

What I am trying to do is this. Lets say I have a list like this:

A(a,b),m,H(b,v),I(m,l),l

Each element is either a tuple (ie A(a,b), H(b,v)...) or a non-tuple (ie m and l). I am trying to develop a regular expression to find and extract each tuple and non-tuple, and know which one I extracted.