I have a text file of 3 name entries:

# dot_test.txt
001 AALTON, Alan .....25 Every Street
006 JOHNS, Jason .... 3 Steep Street
002 BROWN. James .... 101 Browns Road

My task is to find instances of NAME. when it should be NAME, using the following:

Select-String -AllMatches -Path $input_path -Pattern '(?s)[A-Z]{3}.*?\D(?=\s|$)' -CaseSensitive |
    ForEach-Object { if($_.Matches.Value -match '\.$'){$_.Matches.Value -replace '\,$'} }

The output is:

BROWN.

The conclusion is this script block identifies the instance of NAME. but fails to make the replacement.

Any suggestions on how to achieve this would be appreciated.

CodePudding user response：

$_.Matches.Value -replace '\,$'

This attempts to replace a , (which you needn't escape as \,) at the end of ($) your match with the empty string (due to the absence of a second, replacement operand), i.e. it would effectively remove a trailing ,.

However, given that your match contains no , and that you instead want to replace its trailing . with ,, use the following:

$_.Matches.Value -replace '\.$', ','  # -> 'BROWN,'

CodePudding user response：

You can use -replace directly, and if you need to replace both a comma and dot at the end of the string, use [.,]$ regex:

Select-String -AllMatches -Path $input_path -Pattern '(?s)[A-Z]{3}.*?\D(?=\s|$)' -CaseSensitive | % {$_.Matches.Value -replace '\.$', ','}

Details:

• (?s)[A-Z]{3}.*?\D(?=\s|$) - matches
  • (?s) - RegexOptions.Singleline mode on and . can match line breaks
  • [A-Z]{3} - three uppercase ASCII letters
  • .*? - any zero or more chars as few as possible
  • \D - any non-digit char
  • (?=\s|$) - a positive lookahead that matches a location either immediately followed with a whitespace or end of string.

The \.$ pattern matches a . at the end of string.