Java量词“？”如何工作？

Question

System.out.println(Pattern.matches("[amn]?", "a"));

This statement returns true.

But

System.out.println(Pattern.matches("[amn]?", "amn"));

System.out.println(Pattern.matches("[amn]?", "adef"));

These statements returns false.

Why ?

My understanding about regex quantifier "?" is this.

Regex: X?
Description: X occurs once or not at all

So the statement "[amn]?" "amn" should return true because a,m,n occurs once. And similarly in "[amn]?" "adef" a occurs only once and m and n do not occur at all.

Where am I going wrong ?

Answer 1

The regex [amn]? matches any string which consists of either a , m or n and nothing else. Such as "a" , which fulfills this condition.

amn and adef , however, start with one of these letters, but continue so that the "once or not at all" rule is not fulfilled.

Answer 2

The first returns true because a is one letter that is either a, m or n.

The others return false because there's not one letter, there's 3 and 4 letters.

Even though your letter group contains 3 letters, it will only check for the existence of one of them.

Answer 3

Function matches() matches the entire string against the regular expression, which means it will return true only if the complete string can be matched by the expression, not any sub-sequence. Refer this documentation .

[amn]? means that either a or m or n can exists once or not at all. Only cases for which matches() will return true :

1. "a"
2. "m"
3. "n"
4. ""

All other cases will be given as false.

If you want to find the regular expression in some string, then use find() as shown below.

    Pattern p = Pattern.compile("[amn]?");
    Matcher mat = p.matcher("");  //pass amn or adef
    boolean matches = false;
    while(mat.find()){
        matches = true;
        break;
    }
    System.out.println(matches);

Answer 4

[amn] is a group consisting of the characters "a", "m" and "n". [amn]? means "one of the characters from the group [amn] or no character at all".

Pattern.matches tries to match the entire pattern against the entire input string.

If you want the sequence of characters "amn" you could try (amn)? , which should mean "the sequence 'amn' or nothing".

Answer 5

? means match current regular expression not greedy