为什么string.matches [“ +-* /”]将报告模式异常？

Question

I have this code:

public static void main(String[] args) {
       String et1 = "test";
       String et2 = "test";
       et1.matches("[-+*/]"); //works fine
       et2.matches("[+-*/]"); //java.util.regex.PatternSyntaxException, why?
}

Because '-' is escape character? But why it will works fine, if '-' switchs with '+' ?

Answer 1

it is because - is used to define a range of characters in a character class. Since + is after * in the ascii table, the range has no sense, and you obtain an error.

To have a literal - in the middle of a character class, you must escape it. There is no problem if the - is at the begining or at the end of the class because it's unambigous.

An other situation where you don't need to escape the - is when you have a character class shortcut before, example:

[\\d-abc]  

(other regex engines like pcre allows the same when the character class shortcut is placed after [abc-\\d] , but Java doesn't seem to allow this.)

Answer 2

- inside a character class (the [xxx] ) is used to define a range, for example: [az] for all lower case characters. If you want to actually mean "dash", it has to be in first or last position. I generally place it first to avoid any confusions.

Alternatively you can escape it: [+\\\\-*/] .

Answer 3

Just FYI, the Java regular expression meta characters are defined here :

The metacharacters supported by this API are: <([{\^-=$!|]})?*+.>

As a general rule, to save myself from regexp debugging headaches, if I want to use any of these characters as a literal then I precede them with a \\ (Or \\\\ inside of a Java String expression).

Either:

et2.matches("[\\+\\-\\*/]");

Or:

et2.matches("[\\-\\+\\*/]");

Will work regardless of order.

Answer 4

我认为您应该使用：[\\-\\ + \\ * /]因为：'-'定义范围，例如：[ad]的意思是：a，b，c，d