`regexp`

Matches a regular expression against text and returns positions, matches, tokens or names.

Category: Language Fundamentals > Data Types

Syntax

[start, finish] = regexp(str, expr)

Start and end indices of every match.

out = regexp(str, expr, outkey)

Returns the piece selected by outkey: 'match', 'tokens', 'names', 'split', 'start', 'end', 'tokenExtents'.

[a, b, ...] = regexp(str, expr, key1, key2, ...)

Several outputs in the order requested.

___ = regexp(___, 'once')

Stops after the first match and returns it without the surrounding cell.

___ = regexp(___, 'ignorecase')

Ignores letter case.

Description

regexp scans the text from left to right and finds non-overlapping matches. Without output keys, the outputs come in a fixed order: start, end, token extents, match, tokens, names, split. Naming keys lets you pick only what you need.

If str is a cell array, the outputs are cell arrays with one entry per input element. If nothing matches, 'match' returns an empty cell (or an empty char with 'once').

Input arguments

str: Character vector, string, or cell array of them.
expr: Regular expression. Parentheses create tokens; (?<name>...) creates a named token.
outkey / option: Selector of outputs and options such as 'once', 'ignorecase', 'lineanchors'.

Output arguments

start, finish: Row vectors of indices.
match: Cell array of matched text (a single character vector with 'once').
tokens: For each match, a cell array of the text captured by each pair of parentheses.
names: Structure whose fields are the named tokens.
split: Cell array of the text between the matches.

Examples

Matches and tokens

[tok, mat] = regexp('x1y22', '(\w)(\d+)', 'tokens', 'match');
mat
tok{2}
regexp('key=val', '(?<k>\w+)=(?<v>\w+)', 'names')

Output

mat = 1×2 cell

    {'x1'}    {'y22'}
ans = 1×2 cell

    {'y'}    {'22'}
ans = struct with fields:

    k: 'key'
    v: 'val'

Positions and split

[s, e] = regexp('ab12cd345', '\d+')
regexp('a,b;c', '[,;]', 'split')