Documentation

    regexp_extract_all

    Find every non-overlapping match of a regular expression in text and return them as a list. An optional group number picks a capturing group, and regex options can follow.

    Try an expression

    In Formula, enter the expression and set Output type to VARCHAR.

    Expression

    regexp_extract_all('Ada: 12, Lin: 34', '[0-9]+')

    Result

    '[12, 34]'

    The expression returns VARCHAR[]. Formula stores it as VARCHAR in this example. To use your own data, replace the example values with column names. Keep format strings and other options as written.

    Arguments and types

    regexp_extract_all(string, regex, group, options)
    string
    Text
    regex
    Text
    group
    Number, List
    options
    Text
    Supported signatures (5)
    regexp_extract_all(string: VARCHAR, regex: VARCHAR) → VARCHAR[]
    regexp_extract_all(string: VARCHAR, regex: VARCHAR, group: INTEGER) → VARCHAR[]
    regexp_extract_all(string: VARCHAR, regex: VARCHAR, group: INTEGER, options: VARCHAR) → VARCHAR[]
    regexp_extract_all(string: VARCHAR, regex: VARCHAR, group: VARCHAR[]) → VARCHAR[]
    regexp_extract_all(string: VARCHAR, regex: VARCHAR, group: VARCHAR[], options: VARCHAR) → VARCHAR[]

    The result type can depend on the input types. See DuckDB for the full argument rules.

    Things to know

    If the first argument is missing, the result is NULL.

    Browse patternsDuckDB documentation

    Function metadata from DuckDB, used under the MIT license.