java.util.regex.PatternSyntaxException：\\ d *的未封闭字符类

Question

My regular expression:

java.util.regex.PatternSyntaxException: Unclosed character class near index 14
[^]*?￥(\d*.\d*)
              ^

But I can't find any flaw in my expression and it works well on http://regexr.com/ .

My intention for [^] :

Use [^] to match any character because . only match any character except line break.

Call code:

Pattern pattern = compile(regex);

StackTrace:

0 = {StackTraceElement@12872} "java.util.regex.Pattern.error(Pattern.java:1955)"
1 = {StackTraceElement@12873} "java.util.regex.Pattern.clazz(Pattern.java:2548)"
2 = {StackTraceElement@12874} "java.util.regex.Pattern.sequence(Pattern.java:2063)"
3 = {StackTraceElement@12875} "java.util.regex.Pattern.expr(Pattern.java:1996)"
4 = {StackTraceElement@12876} "java.util.regex.Pattern.compile(Pattern.java:1696)"
5 = {StackTraceElement@12877} "java.util.regex.Pattern.<init>(Pattern.java:1351)"
6 = {StackTraceElement@12878} "java.util.regex.Pattern.compile(Pattern.java:1028)" 
...

EDIT:

[^.]*?￥(\\d*.\\d*) works. It seems java.util.regex doesn't support [^] to match any character.

If any one is willing to write some explanation for that, I will accept it as answer.

Answer 1

The problem is that you are using the wrong tool to test your regex . RegExr v2 uses JavaScript regex, while RegExr v1 uses Flex regex, both have different parsing rule compared to Java regex.

While [^] is valid in JavaScript, it is parsed as an (incomplete) negated character class in Java, which contain literal ] (and whatever follows).

Java does not accept empty character class (or negated empty character class), so the first ] in character class (ie [] or [^] ) does not close the character class, and ] is treated as literal character instead.

To make . match any character without exception, you might want to use Pattern.DOTALL flag:

Pattern pattern = Pattern.compile(".*?￥(\\d*.\\d*)", Pattern.DOTALL);