Does a dot have to be escaped in a character class (square brackets) of a regular expression?

Question

A dot . in a regular expression matches any single character. In order for regex to match a dot, the dot has to be escaped: \.

It has been pointed out to me that inside square brackets [] a dot does not have to be escaped. For example, the expression: [.]{3} would match ... string.

Doesn't it, really? And if so, is it true for all regex standards?

Yes that is true that DOT (and most other special characters) don't need to be escaped in character class. — anubhava, Nov 14 '13 at 11:05
@BoltClock there are some: posix, posix extended, perl. See http://en.wikipedia.org/wiki/Regular_expression#Standards — Dariusz, Nov 14 '13 at 12:31
Worth noting that modern regex lets you escape any symbol even if not required, so you can just escape stuff when unsure. Only really need to worry about brevity when a super-concise one-liner is your goal, or your expression has too many slashes that it hurts readability. Over time you'll naturally learn how to write more cleanly. And sometimes it actually makes things easier to read when you consistently escape things like `(` or `?` so that at a glance you can see "this is a literal" without thinking about the context of where that token is (esp. with any crazy nesting). — Beejor, Feb 10 '19 at 04:33
Suppose you were designing a regex engine and faced the question of whether a period within a character class must be escaped. The only reason for doing so would be to allow a non-escaped period in the character class to represent the meta-character `.`, which matches any character. But wait, what use is a character class that matches any character? — Cary Swoveland, Mar 14 '20 at 02:05

score 207 · Accepted Answer · edited Dec 27 '16 at 17:56

207

In a character class (square brackets) any character except ^, -, ] or \ is a literal.

This website is a brilliant reference and has lots of info on the nuances of different regex flavours. http://www.regular-expressions.info/refcharclass.html

edited Dec 27 '16 at 17:56

MrWhite

43,179
8
60
84

answered Nov 14 '13 at 11:18

lilactiger89

2,308
1
13
14

7

It really depends on how the language handles it, but for most languages this is true. – Patrick Oscity Nov 14 '13 at 11:26
56

`-` is also literal if it's the last value – Pedro Lobito May 10 '16 at 12:38
In ICU and Java regexps, both the `[` and `]` must be escaped inside a character class (and `{` and `}` must be escaped outside the character class). – Wiktor Stribiżew Oct 11 '16 at 07:03
26

And ^ is literal if it's not the first character – Dec 17 '16 at 02:52
1

If $ looks like a variable, it also needs to be escaped. E.g.: [$.] – W3Coder Mar 31 '17 at 13:55
1

@PedroLobito And if it's the first, I think – Tim Malone Aug 07 '18 at 12:05
In languages with a start and end character, the end character must be escaped. For instance escape `/` if meant as a literal in the character class `/[\/]/` _(That becomes less obvious when doing all special characters)_. – WEBjuju Dec 27 '18 at 16:55
Don't forget `:` used for POSIX named character classes. – Beejor Feb 10 '19 at 04:28
1

does anybody know if this is supported by all browsers consistently? – YakovL Sep 14 '19 at 10:32
@TimMalone for sure! – Vasilii Suricov Jan 18 '20 at 17:03
1

@PedroLobito `-` is also a literal if it's the first value. For example: `re.search(r'[-asdf]+', 'a-s').group()` yields the same results as `re.search(r'[asdf-]+', 'a-s').group()`. – kimbo Feb 26 '20 at 16:57
2

@kimbo Yes, `-` doesn't need to be escaped if it's the first or last character inside a character class. – Pedro Lobito Feb 26 '20 at 17:17

Does a dot have to be escaped in a character class (square brackets) of a regular expression?

1 Answers1

Linked

Related