Java regex does not match as expected

Question

I'm starting with regex in Java recently, and I cant wrap my head around this problem.

        Pattern p = Pattern.compile("[^A-Z]+");
        Matcher matcher = p.matcher("GETs");
        if (matcher.matches()) {
            System.out.println("Matched.");
        } else {
            System.out.println("Did not match.");
        }

Result: Did not Match(Unexpected result) Explain this

I get the output "Did not match." This is strange to me, while reading https://docs.oracle.com/javase/7/docs/api/java/util/regex/Pattern.html, I'm using the X+, which matches "One, or more times".

I thought my code in words would go something like this:

"Check if there is one or more characters in the string "GETs" which does not belong in A to Z."

So I'm expecting the following result:

"Yes, there is one character that does not belong to A-Z in "GETs", the regex was a match."

However this is not the case, I'm confused to why this is. I tried the following:

        Pattern p = Pattern.compile("[A-Z]+");
        Matcher matcher = p.matcher("GETs");
        if (matcher.matches()) {
            System.out.println("Matched.");
        } else {
            System.out.println("Did not match.");
        }

Result: Did not match. (Expected result)

        Pattern p = Pattern.compile("[A-Z]+");
        Matcher matcher = p.matcher("GET");
        if (matcher.matches()) {
            System.out.println("Matched.");
        } else {
            System.out.println("Did not match.");
        }

Result: Matched. (Expected result)

Please, explain why my first example did not work.

Not sure about the down-votes. I've tried to provide a suitable answer for you though... :) — Pineda, Apr 14 '17 at 15:56
It is a dupe of [Difference between matches() and find() in Java Regex](http://stackoverflow.com/questions/4450045/difference-between-matches-and-find-in-java-regex). — Wiktor Stribiżew, Apr 14 '17 at 16:16

score 10 · Accepted Answer · edited Jan 22 '23 at 17:56

10

Matcher.matches returns true only if the ENTIRE region matches the pattern.

For the output you are looking for, use Matches.find instead

Explanation of each case:

Pattern p = Pattern.compile("[^A-Z]+");
Matcher matcher = p.matcher("GETs");
  if (matcher.matches()) {

Fails because the ENTIRE region 'GETs' isn't lowercase

Pattern p = Pattern.compile("[A-Z]+");
Matcher matcher = p.matcher("GETs");
  if (matcher.matches()) {

This fails because the ENTIRE region 'GETs' isn't uppercase

Pattern p = Pattern.compile("[A-Z]+");
Matcher matcher = p.matcher("GET");
  if (matcher.matches()) {

The ENTIRE region 'GET' is uppercase, the pattern matches.

edited Jan 22 '23 at 17:56

Marc Bannout

388
4
15

answered Apr 14 '17 at 15:34

Pineda

7,435
3
30
45

Thats not true, it clearly sais matching one or more times? – Ludvig W Apr 14 '17 at 15:46
That applies to `Matcher.matches`, NOT `Pattern.matches` – Pineda Apr 14 '17 at 15:52
@Pineda actually, [Pattern#matches](https://docs.oracle.com/javase/7/docs/api/java/util/regex/Pattern.html#matches(java.lang.String,%20java.lang.CharSequence)) also matches the regex, against whole String. – dumbPotato21 Apr 14 '17 at 16:06
I still dont understand, according to the website: "X+ X, one or more times". – Ludvig W Apr 14 '17 at 16:08
@Shashwat: so it does :-) – Pineda Apr 14 '17 at 16:08
@Lurr: `Matcher.find` is what you are looking for. AGAIN: `Matcher.matches` matches against the ENTIRE target region... – Pineda Apr 14 '17 at 16:29
I totally forgot, i should add .* before and .* after.. yes ur totally correct – Ludvig W Apr 14 '17 at 16:37
No worries. If you found this answer helpful I'd greatly appreciate an up-vote and an accepted answer :) – Pineda Apr 14 '17 at 16:49

score 1 · Answer 2 · answered Apr 14 '17 at 15:33

1

You're very first regex asks to match any character that is not in an uppercase range of A-Z. The match is on the lowercase "s" in GETs.

answered Apr 14 '17 at 15:33

Jay Kravetz

233
2
10

Yea but why doesnt it match then? – Ludvig W Apr 14 '17 at 15:46
Because you're attempting to match the whole word. – Jay Kravetz Apr 14 '17 at 15:50
Yes, but why did the 2nd not match then? Its the same. – Ludvig W Apr 14 '17 at 15:54
The 2nd pattern is asking if the whole word GETs only _has_ chracters in the upper case A-Z range [^A-Z]+ : Anything _NOT_ in uppercase A-Z [^A-Za-z]+ : Anything _NOT_ in A-Z or a-z – Jay Kravetz Apr 14 '17 at 16:02

score 0 · Answer 3 · answered Feb 14 '19 at 14:56

0

if you want a regex to match either in UPPERCASE and lowercase, you can use this:
String test = "yes"; String test2= "YEs"; test.matches("(?i).*\\byes\\b.*"); test2.matches("(?i).*\\byes\\b.*");

will return true in the two cases

answered Feb 14 '19 at 14:56

Frigg0

125
3
9

Java regex does not match as expected

3 Answers3

Explanation of each case: