Regular expressions without the pain — a 5-minute practical guide

The four characters that matter most

The character class `[abc]` (matches one of the listed characters), the quantifier `+` (one or more of the previous element), the anchor `^` (start of the line) and `$` (end of the line). With those four, you can write a useful regex. Everything else is precision.

Anatomy of a regex

A regex is a pattern, optionally followed by flags. The pattern is a sequence of literal characters and metacharacters. Metacharacters are the things that mean something other than themselves: `.` (any single character), `*` (zero or more), `+` (one or more), `?` (zero or one), `[...]` (character class), `(...)` (group), `|` (alternation), `\` (escape). The flags are single letters after the closing `/` that modify behaviour: `g` (global), `i` (case-insensitive), `m` (multiline).

A worked example

Look at the regex this site uses to detect a YouTube video ID in a URL: `[a-zA-Z0-9_-]{11}`. That is a character class containing the 26 letters in both cases, the 10 digits, the underscore, and the hyphen, followed by `{11}` (exactly 11 of the previous element). The whole thing says: "find 11 consecutive characters, each of which is a letter, digit, underscore, or hyphen" — which is what a YouTube video ID looks like.

Common pitfalls

Greedy quantifiers: `*` and `+` match as much as they can, which is rarely what you want. Use `*?` and `+?` (the lazy versions) when you want the shortest match. Catastrophic backtracking: a pattern like `(a+)+b` can take exponential time on input that does not match; modern engines protect against this, but it is worth knowing about. Anchoring: a regex without `^` and `$` will match anywhere in the string, which is sometimes the bug.

Testing and explaining

A regex tester shows the matches inline and explains what each part of the pattern does. Use it before you ship a regex. The free regex tester on this site does both: live match highlighting and a plain-English explanation of every group.