You compose a predefined set by using the backslash character together with an alphabetical character. Lowercase letters indicate inclusive sets, character classes composed of the characters having a special feature; capital letters denote the negated class of their corresponding lowercase set, composed of all the characters not having that special feature.

• \d Numeric digit. Equivalent to [0-9].
• \D Anything but numeric digit. Equivalent to [^0-9].
• \w Alphanumerical character, including the underscore: [a-zA-Z0-9_].
• \W Non-alphanumerical character, nor the underscore: [^a-zA-Z0-9_].
• \l Lowercase letter. It also includes the accented characters as defined by the current language selection.
• \L Not a lowercase letter.
• \u Uppercase letter. It also includes the accented characters as defined by the current language selection.
• \U Not an uppercase letter.
• \s White-space, like space, tabs, newlines, etc.
• \S A non-spacing character.

Predefined character sets represented by lowercase characters can be included within user-defined sets. For example '[\da-fA-F]' describes hexadecimal characters. Of course, you cannot define a range with either end being a predefined set.