用于替换除下划线之外的所有特殊字符的Java正则表达式

Question

I would like a regular expression for removing all special characters except an underscore. I can replace all special characters, but I don't know how to keep the underscores. Here's the code for removing all speial characters,

String myname= "!john_smith@#-".replaceAll("\\p{Punct}+", "");

Any ideas how to modify the regex so that I keep underscores?

Thanks, Patrick

Answer 1

Well \\\\p... is the same as [^\\\\P...] (negated character class of characters not having the property), and then you could put the underscore in there as well:

"[^\\P{Punct}_]+"

Alternatively, use a negative lookahead

"(?:(?!_)\\p{Punct})+"

Also, seeing your example, maybe something as simple as this will be enough for you:

"[^\\w\\s]+"

Will remove everything except for letters, digits, underscores and whitespace.

Answer 2

Use set difference:

System.out.println("#%a#%^$^_#$%b#$".replaceAll("[\\p{Punct}&&[^_]]", ""));

prints

a_b

Reference: Character Classes