Module lessons (4/4)
Escaping meta-characters
Some characters in regex have a special meaning: they are called
meta-characters. If you want to match the literal character (e.g. a real
., a real +, a real parenthesis) you need to prefix them with a
backslash \, which escapes them.
The meta-characters to remember are twelve:
. * + ? ^ $ | \ ( ) [ ] { } /(The / is not really an engine meta-character, but it must be escaped in
the JavaScript literal notation /.../.)
Pattern: \$\d+\.\d\d
Sample: Prezzi: $9.99, $12.50, totale $22.49.
^^^^^ ^^^^^^ ^^^^^^Here we escape $ (end-of-string anchor) and . (wildcard) to match the
literal characters in the price format "dollar + number + dot + two digits".
Double backslash in JS strings
When you write a pattern inside a JavaScript string (like you do in our
exercises) the backslash must be doubled. The JS string '\\d' contains
two characters (\ and d) and is exactly the regex pattern \d.
const re1 = /\d+/; // literal notation: a single backslash
const re2 = new RegExp('\\d+'); // string: double backslashThe course editor shows the pattern already decoded (you see a single
\d); you do not have to double anything yourself.
The 12 special meta-characters
The characters with special meanings in the engine are exactly 12: \\\ ^ $ . | ? * + ( ) [ {. If you want to match them as literal text, you must precede them with a backslash. In some environments (e.g. non-raw JS strings), you might need to double the backslash (\\\\).
Try it
Find every price in the `$N.NN` format (dollar, one or more digits, dot, two digits). Remember to escape `$` and `.`.
Show hint
$ must be escaped as \\$, the dot as \\. - otherwise the dot matches anything.
Solution available after 3 attempts
Review exercise
Find every literal question mark `?` in the text. `?` is a meta-character: it must be escaped.
Show hint
Without an escape, ? would be a quantifier (module 2). Put a \\ in front of it.
Solution available after 3 attempts
Additional challenge
Find all occurrences of `.tar.gz` in the text, ensuring that the dots are matched literally.
Show hint
Insert a backslash before each dot to escape it.
Solution available after 3 attempts