Consider a practical task – we have a phone number
"+7(903)-123-45-67", and we need to turn it into pure numbers:
To do so, we can find and remove anything that’s not a number. Character classes can help with that.
A character class is a special notation that matches any symbol from a certain set.
For the start, let’s explore a “digit” class. It’s written as
\d. We put it in the pattern, that means “any single digit”.
For instance, the let’s find the first digit in the phone number:
Without the flag
g, the regular expression only looks for the first match, that is the first digit
Let’s add the
g flag to find all digits:
That was a character class for digits. There are other character classes as well.
Most used are:
\d(“d” is from “digit”)
- A digit: a character from
\s(“s” is from “space”)
- A space symbol: that includes spaces, tabs, newlines.
\w(“w” is from “word”)
- A “wordly” character: either a letter of English alphabet or a digit or an underscore. Non-Latin letters (like cyrillic or hindi) do not belong to
\d\s\w means a “digit” followed by a “space character” followed by a “wordly character”, like
A regexp may contain both regular symbols and character classes.
CSS\d matches a string
CSS with a digit after it:
Also we can use many character classes:
The match (each character class corresponds to one result character):
A word boundary
\b – is a special character class.
It does not denote a character, but rather a boundary between characters.
Java in the string
Hello, Java!, but not in the script
The boundary has “zero width” in a sense that usually a character class means a character in the result (like a wordly character or a digit), but not in this case.
The boundary is a test.
When regular expression engine is doing the search, it’s moving along the string in an attempt to find the match. At each string position it tries to find the pattern.
When the pattern contains
\b, it tests that the position in string is a word boundary, that is one of three variants:
- Immediately before is
\w, and immediately after – not
\w, or vise versa.
- At string start, and the first string character is
- At string end, and the last string character is
For instance, in the string
Hello, Java! the following positions match
So it matches
- At the beginning of the string the first
- Then the word
\bmatches, as we’re between
oand a space.
\bJava\b also matches. But not
\bHell\b (because there’s no word boundary after
l) and not
Java!\b (because the exclamation sign is not a wordly character, so there’s no word boundary after it).
Once again let’s note that
\b makes the searching engine to test for the boundary, so that
Java only when followed by a word boundary, but it does not add a letter to the result.
Usually we use
\b to find standalone English words. So that if we want
"Java" language then
\bJava\b finds exactly a standalone word and ignores it when it’s a part of another word, e.g. it won’t match
Another example: a regexp
\b\d\d\b looks for standalone two-digit numbers. In other words, it requires that before and after
\d\d must be a symbol different from
\w (or beginning/end of the string).
The word boundary check
\b tests for a boundary between
\w and something else. But
\w means an English letter (or a digit or an underscore), so the test won’t work for other characters (like cyrillic or hieroglyphs).
Later we’ll come by Unicode character classes that allow to solve the similar task for different languages.
For every character class there exists an “inverse class”, denoted with the same letter, but uppercased.
The “reverse” means that it matches all other characters, for instance:
- Non-digit: any character except
\d, for instance a letter.
- Non-space: any character except
\s, for instance a letter.
- Non-wordly character: anything but
- Non-boundary: a test reverse to
In the beginning of the chapter we saw how to get all digits from the phone
One way was to match all digits and join them:
An alternative, shorter way is to find non-digits
\D and remove them from the string:
Usually we pay little attention to spaces. For us strings
1 - 5 are nearly identical.
But if a regexp doesn’t take spaces into account, it may fail to work.
Let’s try to find digits separated by a dash:
Here we fix it by adding spaces into the regexp
\d - \d:
A space is a character. Equal in importance with any other character.
Of course, spaces in a regexp are needed only if we look for them. Extra spaces (just like any other extra characters) may prevent a match:
In other words, in a regular expression all characters matter, spaces too.
"." is a special character class that matches “any character except a newline”.
Or in the middle of a regexp:
Please note that the dot means “any character”, but not the “absense of a character”. There must be a character to match it:
Usually a dot doesn’t match a newline character.
A, and then
B with any character between them, except a newline.
This doesn’t match:
Sometimes it’s inconvenient, we really want “any character”, newline included.
s flag does. If a regexp has it, then the dot
"." match literally any character:
There exist following character classes:
\s– space symbols, tabs, newlines.
\S– all but
\w– English letters, digits, underscore
\W– all but
.– any character if with the regexp
's'flag, otherwise any except a newline.
…But that’s not all!
- A cyrillic letter is:
- A dash (be it a small hyphen
-or a long dash
- A currency symbol, such as
- …And much more. Unicode has a lot of character categories that we can select from.
These patterns require
'u' regexp flag to work. More about that in the chapter Unicode: flag "u".