HashiTalks 2025 Learn about unique use cases, homelab setups, and best practices at scale at our 24-hour virtual knowledge sharing event. Register
Terraform
Terraform Home

regex Function

regex applies a regular expression to a string and returns the matching substrings.

regex(pattern, string)

The return type of regex depends on the capture groups, if any, in the pattern:

  • If the pattern has no capture groups at all, the result is a single string covering the substring matched by the pattern as a whole.
  • If the pattern has one or more unnamed capture groups, the result is a list of the captured substrings in the same order as the definition of the capture groups.
  • If the pattern has one or more named capture groups, the result is a map of the captured substrings, using the capture group names as map keys.

It's not valid to mix both named and unnamed capture groups in the same pattern.

If the given pattern does not match at all, the regex raises an error. To test whether a given pattern matches a string, use regexall and test that the result has length greater than zero.

The pattern is a string containing a mixture of literal characters and special matching operators as described in the following table. Note that when giving a regular expression pattern as a literal quoted string in the Terraform language, the quoted string itself already uses backslash \ as an escape character for the string, so any backslashes intended to be recognized as part of the pattern must be escaped as \\.

SequenceMatches
.Any character except newline
[xyz]Any character listed between the brackets (x, y, and z in this example)
[a-z]Any character between a and z, inclusive
[^xyz]The opposite of [xyz]
\dASCII digits (0 through 9, inclusive)
\DAnything except ASCII digits
\sASCII spaces (space, tab, newline, carriage return, form feed)
\SAnything except ASCII spaces
\wThe same as [0-9A-Za-z_]
\WAnything except the characters matched by \w
[[:alnum:]]The same as [0-9A-Za-z]
[[:alpha:]]The same as [A-Za-z]
[[:ascii:]]Any ASCII character
[[:blank:]]ASCII tab or space
[[:cntrl:]]ASCII/Unicode control characters
[[:digit:]]The same as [0-9]
[[:graph:]]All "graphical" (printable) ASCII characters
[[:lower:]]The same as [a-z]
[[:print:]]The same as [[:graph:]]
[[:punct:]]The same as [!-/:-@[-`{-~]
[[:space:]]The same as [\t\n\v\f\r ]
[[:upper:]]The same as [A-Z]
[[:word:]]The same as \w
[[:xdigit:]]The same as [0-9A-Fa-f]
\pNUnicode character class by using single-letter class names ("N" in this example)
\p{Greek}Unicode character class by unicode name ("Greek" in this example)
\PNThe opposite of \pN
\P{Greek}The opposite of \p{Greek}
xyx followed immediately by y
x|yeither x or y, preferring x
x*zero or more x, preferring more
x*?zero or more x, preferring fewer
x+one or more x, preferring more
x+?one or more x, preferring fewer
x?zero or one x, preferring one
x??zero or one x, preferring zero
x{n,m}between n and m repetitions of x, preferring more
x{n,m}?between n and m repetitions of x, preferring fewer
x{n,}at least n repetitions of x, preferring more
x{n,}?at least n repetitions of x, preferring fewer
x{n}exactly n repetitions of x
(x)unnamed capture group for sub-pattern x
(?P<name>x)named capture group, named name, for sub-pattern x
(?:x)non-capturing sub-pattern x
\*Literal * for any punctuation character *
\Q...\ELiteral ... for any text ... as long as it does not include literally \E

In addition to the above matching operators that consume the characters they match, there are some additional operators that only match, but consume no characters. These are "zero-width" matching operators:

SequenceMatches
^At the beginning of the given string
$At the end of the given string
\AAt the beginning of the given string
\zAt the end of the given string
\bAt an ASCII word boundary (transition between \w and either \W, \A or \z, or vice-versa)
\BNot at an ASCII word boundary

Terraform uses the RE2 regular expression language. This engine does not support all of the features found in some other regular expression engines; in particular, it does not support backreferences.

Matching Flags

Some of the matching behaviors described above can be modified by setting matching flags, activated using either the (?flags) operator (to activate within the current sub-pattern) or the (?flags:x) operator (to match x with the modified flags). Each flag is a single letter, and multiple flags can be set at once by listing multiple letters in the flags position. The available flags are listed in the table below:

FlagMeaning
iCase insensitive: a literal letter in the pattern matches both lowercase and uppercase versions of that letter
mThe ^ and $ operators also match the beginning and end of lines within the string, marked by newline characters; behavior of \A and \z is unchanged
sThe . operator also matches newline
UThe meaning of presence or absense ? after a repetition operator is inverted. For example, x* is interpreted like x*? and vice-versa.

Examples

> regex("[a-z]+", "53453453.345345aaabbbccc23454")
aaabbbccc
 
> regex("(\\d\\d\\d\\d)-(\\d\\d)-(\\d\\d)", "2019-02-01")
[
  "2019",
  "02",
  "01",
]
 
> regex("^(?:(?P<scheme>[^:/?#]+):)?(?://(?P<authority>[^/?#]*))?", "https://terraform.io/docs/")
{
  "authority" = "terraform.io"
  "scheme" = "https"
}
 
> regex("[a-z]+", "53453453.34534523454")
 
Error: Error in function call
 
Call to function "regex" failed: pattern did not match any part of the given
string.

Related Functions

  • regexall searches for potentially multiple matches of a given pattern in a string.
  • replace replaces a substring of a string with another string, optionally matching using the same regular expression syntax as regex.

If Terraform already has a more specialized function to parse the syntax you are trying to match, prefer to use that function instead. Regular expressions can be hard to read and can obscure your intent, making a configuration harder to read and understand.

Edit this page on GitHub