Java replaceAll but the specified regex

For your particular task, I think

getStringFromDom(doc).replaceAll(".*?(contract=\"\\d*\"|name=\"\\p{L}*\").*", "$1");

should do what you need.

You want to remove everything that does not match the pattern. This is the same as simply filtering the pattern matches. Use the regex to find matches for that pattern, then collect the matches in a stringbuilder.

Matcher m = Pattern.compile(your pattern).matcher(your input);
StringBuilder sb = new StringBuilder();
while (m.find()) sb.append (m.group()).append('\n');
String result = sb.toString();

I also think that removing what your are not looking for is a double negative. Concentrate on what you are looking for and use a pattern matching for that. This example searches your document for any name attributes:

import java.util.regex.Matcher;
import java.util.regex.Pattern;

public class Test {
    public static void main(String[] args) {
        String input = "<AnotherDoc accNum=\"1111\" docDate=\"2017-09-26\" docNum=\"2222\" name=\"foo\"> <anotherTag>some date</anotherTag>";

        Pattern pattern = Pattern.compile("name=\"[^\\\"]*\""); // value are all characters but "
        Matcher matcher = pattern.matcher(input);

        while (matcher.find())
            System.out.println(matcher.group());
    }
}

This prints:

name="foo"