regex string between two strings

Try with:

/<rate[^>]*>(.*?)<\/rate>/

However it is better NOT TO USE REGEX WITH HTML.

The way I do it is using the match all

matched = Regex.Matches(result, @"(?<=<rate curr=\"KRW\" unit=\"100\">)(.*?)(?=</rate>)");

Then get one by one using match[i].Groups[1].value

If you're analyzing HTML, you're probably better off going with javascript and .innerHTML(). Regex is a bit overkill.

If you want a generic solution, i.e to find a string between two strings You may use Pattern.quote() [or wrap string with \Q and \E around] to quote start and end strings and use (.*?) for a non greedy match.

See an example of its use in below snippet

@Test
public void quoteText(){
    String str1 = "<rate curr=\"KRW\" unit=\"100\">";
    String str2 = "</rate>";

    String input = "<rate curr=\"KRW\" unit=\"100\">19,94</rate>"
                      +"<rate curr=\"KRW\" unit=\"100\"></rate>"
                      +"<rate curr=\"KRW\" unit=\"100\">19,96</rate>";

    String regex = Pattern.quote(str1)+"(.*?)"+Pattern.quote(str2);
    System.out.println("regex:"+regex);

    Pattern p = Pattern.compile(regex);
    Matcher m = p.matcher(input);
    while(m.find()){
        String group = m.group(1);
        System.out.println("--"+group);
    }

Output

regex:\Q<rate curr="KRW" unit="100">\E(.*?)\Q</rate>\E
--19,94
--
--19,96

Note:Though its not recommended to use regex to parse entire HTML, I think there is no harm in conscious use of regex while treating HTML as plain text

The simple regex matching string you're looking for is:

(?<=<rate curr=\"KRW\" unit=\"100\">)(.*?)(?=</rate>)

In Ruby, for example, this would translate to:

string = '<rate curr="KRW" unit="100">19,94</rate>'

string.match("(?<=<rate curr=\"KRW\" unit=\"100\">)(.*?)(?=</rate>)").to_s
# => "19,94"

Thanks to Will Yu.

I suggest that you use an HTML parser. The grammar that defines HTML is a context-free grammar, which is fundamentally too complex to be parsed by regular expressions. Even if you manage to write a regular expression that will achieve what you want, but will probably fail on some corner cases.

For instance, what if you are expected to parse the following HTML?

<rate curr="KRW" unit="100"><rate curr="KRW" unit="100">19,94</rate></rate>

A regular expression may not handle this corner case properly.