Get count of exact word match within a Text file-CodePudding

The requirement is to get the exact match count for a word "test". So in the following example it should be 1:

testing 1 2 3 "test" testing
Tester testing 2345 tes testers testings testing
test

I tried the below code :

(Get-Content "C:\Users\abc\Desktop\POC\Findstring.txt" | 
    Select-String -Pattern "test" -AllMatches).matches.count

But it provides me the value as 9 since it provides a like functionality (it is also considering tester,testing etc in the count).

How should we ensure that we get the count for exact match and not for a LIKE operator scenario (similar to in SQL).

CodePudding user response：

tl;dr

Use regex \btest\b as the -Pattern argument so as to match test as a whole word only.
Pass your input file path directly to Select-String's -LiteralPath parameter, which is much faster and more efficient than streaming the individual lines from the file via Get-Content.

(
  Select-String -AllMatches `
                -Pattern '\btest\b' `
                -LiteralPath C:\Users\abc\Desktop\POC\Findstring.txt 
).Matches.Count

^{Note: The command is spread across multiple lines for readability. To convert it to a single-line form, also remove the line-ending ` (backtick) characters, which act as line continuations.}

Your intent is to limit matching test substrings to whole words.

Since Select-String uses regexes (regular expressions), you can do so by enclosing the substring in word-boundary assertions, \b, as Theo advises, i.e. '\btest\b'

For a detailed explanation of this regex and the ability to interact with it, see this regex101.com page

Also note that Select-String - like PowerShell in general - is case-insensitive by default; to match case-sensitively, add the -CaseSensitive switch.

Variation with also ignoring the word test when enclosed in "..."

If you additionally want to ignore "test" substrings (i.e. double-quoted instances of the word), you must amend your regex to also include a negative look-behind assertion, (?!...) in order to preclude a " preceding the word:

(
  Select-String -AllMatches `
                -Pattern '(?<!")\btest\b' `
                -LiteralPath C:\Users\abc\Desktop\POC\Findstring.txt 
).Matches.Count

See this regex101.com page.

CodePudding user response：

Currently, you search for the pattern test which is also true in case of testing, testers, etc. The following should do the trick:

((Get-Content "C:\tmp\testdata.txt") -split " " | Select-String -Pattern '^(test)$' -AllMatches).count