RegEx find text within / surrounded by string(s) | Pattern.matches() == false

Question

I have a the following json string, which I have to filter to get the so called raw sentence.

{"subscribe":["message","sentence"," ANYTEXTWITHANYLENGTHHEREINCLUDINGNUMBERSDOTS ;\\r\\n\\"]}

Deserialization of the json string is not allowed!

I did filter the searched text with the following regular expression :

((?<=\{\"subscribe\":\[\"message\",\"sentence\",")(.*)(?=;))

or (since it has to be escaped in the java code):

String pattern = "(?<=\\{\"subscribe\":\\[\"message\",\"sentence\",\")(.*)(?=;)";

This is working perfectly if I use Matcher.find() and iterate within a while statement:

while (_regexMatcher.find()) {
    matches.add(_regexMatcher.group()); //Add to List<String>
}

Unfortunately the whole code is given by our professor. He did surround the whole code with:

if (Pattern.matches(pattern, json)) {
     ... 
}

Since Pattern.matches() uses the pattern with ^ in the beginnen of the whole pattern and $ at the ending, so it won't find any matches.
Is there another way to get the ANYTEXTWITHANYLENGTHHEREINCLUDINGNUMBERSDOTS so Pattern.matches() returns true?

Answer 1

You can try with:

[\s\S]*(?<=\{"subscribe":\["message","sentence",")(.*)(?=;)[\s\S]*

the [\\s\\S]* will match anything before and after, so the whole string will be matched, but only if there will be (?<=\\{"subscribe":\\["message","sentence",")(.*)(?=;) inside.

Answer 2

while (in_file.hasNextLine()) 
    {
        one_line = in_file.nextLine();

        tokens = one_line.split("\\P{Alpha}+");
        for (i = 0; i < tokens.length; i++) 
        {
            //do whatever;
        }

more detailed info here http://docs.oracle.com/javase/7/docs/api/java/util/regex/Pattern.html

Answer 3

Just add .*? in front of your regular expression. It will match reluctantly (by opposition with the greedy .*), which means it will stop matching as soon as it reaches your previous json pattern, which should work as previously