If you're looking to collapse multiple consecutive whitespace characters into a single space then you can do this using the -replace
operator. Given...
PS> $beforeReplace = ' [ Hello, World! ] '
PS> $beforeReplace
[ Hello, World! ]
PS> $beforeReplace.Length
29
...you would call the -replace
operator like this...
PS> $afterReplace = $beforeReplace -replace '\s+', ' '
PS> $afterReplace
[ Hello, World! ]
PS> $afterReplace.Length
19
The first parameter to -replace
is a regular expression pattern to match, and the second parameter is the text that will replace any matches. \s
will match a whitespace character, and +
indicates to match one or more occurrences, so, in other words, one or more adjacent whitespace characters will be replaced with a single space.
Replacement without whitespace normalization
If you don't need to normalize all whitespace characters to spaces and, thus, it's ok for standalone whitespace characters to be left untouched, then for long strings you might see better performance with this variation...
PS> $afterReplace = $beforeReplace -replace '\s{2,}', ' '
PS> $afterReplace
[ Hello, World! ]
PS> $afterReplace.Length
19
The \s{2,}
uses a quantifier meaning "match the preceding element at least two times"; therefore, standalone whitespace characters will not be replaced. When the input string contains a mix of whitespace characters...
PS> $beforeReplace = "1Space: ;2Space: ;1Tab:`t;2Tab:`t`t;1Newline:`n;2Newline:`n`n;"
PS> $beforeReplace
1Space: ;2Space: ;1Tab: ;2Tab: ;1Newline:
;2Newline:
;
PS> $beforeReplace.Length
57
...note how the results for the two approaches differ...
PS> $afterReplaceNormalized = $beforeReplace -replace '\s+', ' '
PS> $afterReplaceNormalized
1Space: ;2Space: ;1Tab: ;2Tab: ;1Newline: ;2Newline: ;
PS> $afterReplaceNormalized.Length
54
PS> $afterReplaceUnnormalized = $beforeReplace -replace '\s{2,}', ' '
PS> $afterReplaceUnnormalized
1Space: ;2Space: ;1Tab: ;2Tab: ;1Newline:
;2Newline: ;
PS> $afterReplaceUnnormalized.Length
54
While both yield strings of the same length, the unnormalized replacement leaves the single space, single tab, and single newline whitespace runs unmodified. This would work just the same whether adjacent whitespace characters are identical or not.
Additional documentation