Question

Is it possible to use regex character class subtraction in some way in GNU sed, GNU awk, GNU grep, or anywhere else from bash?

Is it supported in C++11?

Where this feature is supported? I was unable to find any references except this one: http://msdn.microsoft.com/en-us/library/20bw873z.aspx (at the bottom).

To be more specific: what’s the best way to convert all the digits except 2 and 5 to asterisks, something like

sed -re 's/[0-9-[25]]/*/g'

Answer 1

Is it possible to use regex character class subtraction in some way in GNU sed, GNU awk, GNU grep, or anywhere else from bash?

No, not with those tools/regex-implementations.

Most (popular) languages do not support this. Java has a way to do this: [0-9&&[^25]] (matching any (ASCII) digit except '2' and '5'), but I know of no other implementations that support this.

Either use negative look-ahead (?![25])[0-9], as Ignacio already suggested, or do it “the hard way”: [1346-9]

See:

• http://www.regular-expressions.info/refflavors.html
• http://docs.oracle.com/javase/6/docs/api/java/util/regex/Pattern.html