为什么关闭方括号“]”不需要在Java正则表达式中转义？

Question

Consider the array:

new Pattern[] {Pattern.compile("\\["),Pattern.compile("\\]") };

Intellij IDEA tells me that \\\\ is redundant and tells me to replace this with ] eg the result is:

new Pattern[] {Pattern.compile("\\["),Pattern.compile("]") };

Why in the first Pattern.compile("\\\\[") is the \\\\ OK, but for the second it is redundant?

Answer 1

The ] symbol is not a special regex operator outside the character class if there is no corresponding unescaped [ before it. Only special characters require escaping. A [ is a special regex operator outside a character class (as it may mark the starting point of a character class). Once the Java regular expression engine sees an unescaped [ in the pattern, it knows there must be a ] to close the character class ahead. Whether it is escaped or not, it does not matter for the engine. If there is no opening [ in the expression, the ] is treated as a mere literal ] symbol. So, [abc] will match a , b or c , and \\[abc] or \\[abc\\] will match [abc] literal character sequence.

So, the [ should be escaped always, and ] does not have to be escaped outside a character class .

When used inside a character class , both [ and ] must be escaped inside a Java regular expression as they may form intersection/subtraction patterns.

Answer 2

] is only considered metacharacter if it is part of [...] . But if your regex doesn't contain unescaped opening [ , which ] would be closing, such bracket is considered simple literal and doesn't require escaping.