about_Regular_Expressions

Short description

Describes regular expressions in PowerShell.

Long description

A regular expression is a pattern used to match text. It can be made up of literal characters, operators, and other constructs.

This article demonstrates regular expression syntax in PowerShell. PowerShell has several operators and cmdlets that use regular expressions. You can read more about their syntax and usage at the links below.

• Select-String
• -match and -replace operators
• -split operator
• switch statement with -regex option

PowerShell regular expressions are case-insensitive by default. Each method shown above has a different way to force case sensitivity.

• For Select-String, use the CaseSensitive parameter.
• For operators that use regular expressions, use the case-sensitive version: -cmatch, -creplace, or -csplit
• For the switch statement, use the -casesensitive option

Character literals

A regular expression can be a literal character or a string. The expression causes the engine to match the text specified exactly.

# This statement returns true because book contains the string "oo"
'book' -match 'oo'

Character classes

While character literals work if you know the exact pattern, character classes allow you to be less specific.

Character groups

[character group] allows you to match any number of characters one time, while [^character group] only matches characters NOT in the group.

# This expression returns true if the pattern matches big, bog, or bug.
'big' -match 'b[iou]g'

If your list of characters to match includes the hyphen character (-), it must be at the beginning or end of the list to distinguish it from a character range expression.

Character ranges

A pattern can also be a range of characters. The characters can be alphabetic [A-Z], numeric [0-9], or even ASCII-based [ -~] (all printable characters).

# This expression returns true if the pattern matches any 2 digit number.
42 -match '[0-9][0-9]'

Numbers

The \d character class will match any decimal digit. Conversely, \D will match any character except decimal digits.

# This expression returns true if it matches a server name.
# (Server-01 - Server-99).
'Server-01' -match 'Server-\d\d'

Word characters

The \w character class will match any word character [a-zA-Z_0-9]. To match any non-word character, use \W.

# This expression returns true.
# The pattern matches the first word character 'B'.
'Book' -match '\w'

Wildcards

The period (.) is a wildcard character in regular expressions. It will match any character except a newline (\n).

# This expression returns true.
# The pattern matches any 4 characters except the newline.
'a1\ ' -match '....'

Whitespace

You can match any whitespace character with the \s character class. You can match any non-whitespace character with \S. You can match literal space characters with .

# This expression returns true.
# The pattern uses the whitespace character class to match the leading
# space and a literal space to matching the trailing space.
' - ' -match '\s- '

Quantifiers

Quantifiers control how many instances of each element should be present in the input string.

The following are a few of the quantifiers available in PowerShell:

The asterisk (*) matches the previous element zero or more times. The result is that even an input string without the element would be a match.

# This returns true for all account name strings even if the name is absent.
'ACCOUNT NAME:    Administrator' -match 'ACCOUNT NAME:\s*\w*'

The plus sign (+) matches the previous element one or more times.

# This returns true if it matches any server name.
'DC-01' -match '[A-Z]+-\d\d'

The question mark ? matches the previous element zero or one time. Like asterisk *, it will even match strings where the element is absent.

# This returns true for any server name, even server names without dashes.
'SERVER01' -match '[A-Z]+-?\d\d'

The {n, m} quantifier can be used several different ways to allow granular control over the quantifier. The second element m and the comma , are optional.

# This returns true if it matches any phone number.
'111-222-3333' -match '\d{3}-\d{3}-\d{4}'

Anchors

Anchors allow you to cause a match to succeed or fail based on the matches position within the input string.

The two commonly used anchors are ^ and $. The caret ^ matches the start of a string, and $, which matches the end of a string. The anchors allow you to match your text at a specific position while also discarding unwanted characters.

# The pattern expects the string 'fish' to be the only thing on the line.
# This returns FALSE.
'fishing' -match '^fish$'

When using anchors in PowerShell, you should understand the difference between Singleline and Multiline regular expression options.

• Multiline: Multiline mode forces ^ and $ to match the beginning and end of every LINE instead of the beginning and end of the input string.
• Singleline: Singleline mode treats the input string as a SingleLine. It forces the . character to match every character (including newlines), instead of matching every character EXCEPT the newline \n.

To read more about these options and how to use them, visit the Regular Expression Language - Quick Reference.

Escaping characters

The backslash (\) is used to escape characters so they aren't parsed by the regular expression engine.

The following characters are reserved: [().\^$|?*+{.

You'll need to escape these characters in your patterns to match them in your input strings.

# This returns true and matches numbers with at least 2 digits of precision.
# The decimal point is escaped using the backslash.
'3.141' -match '3\.\d{2,}'

There's a static method of the regex class that can escape text for you.

[regex]::escape('3.\d{2,}')

3\.\\d\{2,}

Other character escapes

There are also reserved character escapes that you can use to match special character types.

The following are a few commonly used character escapes:

Groups, Captures, and Substitutions

Grouping constructs separate an input string into substrings that can be captured or ignored. Grouped substrings are called subexpressions. By default subexpressions are captured in numbered groups, though you can assign names to them as well.

A grouping construct is a regular expression surrounded by parentheses. Any text matched by the enclosed regular expression is captured. The following example breaks the input text into two capturing groups.

'The last logged on user was CONTOSO\jsmith' -match '(.+was )(.+)'

True

Use the $Matches Hashtable automatic variable to retrieve captured text. The text representing the entire match is stored at key 0. It's important to note that the $Matches hashtable contains only the first occurrence of any matching pattern.

$Matches.0

The last logged on user was CONTOSO\jsmith

Captures are stored in numeric Integer keys that increase from left to right. Capture 1 contains all the text until the username, capture 2 contains just the username.

$Matches

Name           Value
----           -----
2              CONTOSO\jsmith
1              The last logged on user was
0              The last logged on user was CONTOSO\jsmith

PS> 'Good Dog' -match 'Dog'
True

PS> $Matches[0]
Dog

PS> $Matches.Item(0)
Dog

PS> $Matches.0
Dog

Named Captures

By default, captures are stored in ascending numeric order, from left to right. You can also assign a name to a capturing group. This name becomes a key on the $Matches Hashtable automatic variable.

Inside a capturing group, use ?<keyname> to store captured data under a named key.

PS> $string = 'The last logged on user was CONTOSO\jsmith'
PS> $string -match 'was (?<domain>.+)\\(?<user>.+)'
True

PS> $Matches

Name                           Value
----                           -----
domain                         CONTOSO
user                           jsmith
0                              was CONTOSO\jsmith

PS> $Matches.domain
CONTOSO

PS> $Matches.user
jsmith

The following example stores the newest log entry in the Windows Security Log. The provided regular expression extracts the username and domain from the message and stores them under the keys:N for name and D for domain.

$log = (Get-EventLog -LogName Security -Newest 1 -InstanceId 4689).message
$r = '(?s).*Account Name:\s*(?<N>.*).*Account Domain:\s*(?<D>[A-Z,0-9]*)'
$log -match $r

True

$Matches

Name                           Value
----                           -----
D                              CONTOSO
N                              jsmith
0                              A process has exited....

For more information, see Grouping Constructs in Regular Expressions.

Substitutions in Regular Expressions

Using the regular expressions (regex) with the -replace operator allows you to dynamically replace text using captured text.

<input> -replace <original>, <substitute>

• <input>: The string to be searched
• <original>: A regular expression used to search the input string
• <substitute>: A regex substitution expression to replace matches found in the input string.

The <original> and <substitute> operands are subject to rules of the regular expression engine such as character escaping or substitution expressions. The replacement pattern can consist of one or more substitutions along with literal characters.

Capture groups can be referenced in the <substitute> string using the $ character before the group identifier.

Two ways to reference capturing groups are by Number and by Name.

• By Number - Capturing Groups are numbered from left to right.

  'John D. Smith' -replace '(\w+) (\w+)\. (\w+)', '$1.$2.$3@contoso.com'
  

  John.D.Smith@contoso.com
  

• By Name - Capture Groups can also be referenced by name.

  'CONTOSO\Administrator' -replace '\w+\\(?<user>\w+)', 'FABRIKAM\${user}'
  

  FABRIKAM\Administrator
  

The $& expression represents all the text matched.

'Gobble' -replace 'Gobble', '$& $&'

Gobble Gobble

'Hello World' -replace '(\w+) \w+', '$1 Universe'
"Hello World" -replace "(\w+) \w+", "`$1 Universe"

Hello Universe
Hello Universe

'5.72' -replace '(.+)', '$$$1'
"5.72" -replace "(.+)", "`$`$`$1"

$5.72
$5.72

For detailed information on substitution expressions, see Substitutions in Regular Expressions.

See also

• about_Operators
• about_Comparison_Operators