Powershell是否从路径名中选择字符串？

我有一个带有目录路径的CSV文件，我需要从中提取一些信息来生成日志文件名。不知道怎么做？ 选择字符串

CSV文件如下所示

User,Computer,Directory
john.doe,CAD-12,C:\Cimatron\CimatronE11_64bit\Program\IT\var\post
john.doe,CAD-12,C:\Cimatron\CimatronE11_64bit\Program\IT\var\post2
john.doe,CAD-12,C:\Cimatron\CimatronE11_64bit\Data\templates
john.doe,CAD-12,C:\Program Files\CGTech\VERICUT 7.3.3\library
john.doe,CAD-12,C:\Program Files\CGTech\VERICUT 7.4\library

我知道我必须导入CSV文件，以下是我所拥有的

$UsersCSV = import-csv C:\test.csv
foreach($Computer in $UsersCSV)
{$Logfile = Select-String "not sure how to search the string for what I want"
$LogFile
}

需要有关选择字符串的帮助。下面是我想要结束的内容，或者是给$Logfile一个唯一名称的最简单的内容

CimE11_post
CimE11_post2
CimE11_templates
VERI_7.3.3_lib
VERI_7.4_lib

---

感谢您的帮助

-replace是您的朋友，下面是一个示例，说明如何从您提供的场景中获得您想要的字符串

希望这有帮助，祝你好运

$file=导入csv$filename

$file.directory中的foreach$dir{ 如果$dir-匹配VERI{ $dir.replace'C:\Program Files\CGTech\，.replace'rary'，.replace'CUT'，.replace'\， }

如果$dir-匹配Cima{ $dir.replace'C:\Cimatron\，.replace'atron'，.replace'64bit\Data\，.replace'64bit\Program\IT\var\， }

}

在对以下问题的评论中提出了一个实用的解决方案：将输入路径中的实例替换为文件名中合法的字符，以获得用作文件名的每个路径的表示形式：

这将产生以下文件名：

C_Cimatron_CimatronE11_64bit_Program_IT_var_post.log
C_Cimatron_CimatronE11_64bit_Program_IT_var_post2.log
C_Cimatron_CimatronE11_64bit_Data_templates.log
C_Program_Files_CGTech_VERICUT_7.3.3_library.log
C_Program_Files_CGTech_VERICUT_7.4_library.log

正如在评论中所说的，在你想要的文件名背后没有明确的算法逻辑

这里有一个解决方案，可以生成所需的文件名，并尽量保持通用性。 这在现实生活中可能不值得做，但也许可以激发现实生活中的解决方案：

# Helper function that takes a directory path and converts it to a log filename
# representing that dir.
function pathToLogFilename([string] $path) {

  # Ordered hashtable defining regex-based transformations (replacements).
  # Entries are processed in order, and processing stops as soon as 
  # a match is found and the transformation has been performed.
  # Entry format: <regex> = <replacement>, to be passed to the -replace operator.
  $ohtTansform = [ordered] @{
    '^(.{3}).+?(E\d+).*$' = '$1$2' # 'CimatronE11_64bit' -> 'CimE11', for instance. 
    '^(.{4}).+? (\d.*)$' = '$1_$2' # 'VERICUT 7.3.3' -> 'VER_7.3.3', for instance
    '^(lib)rary$' = '$1'           # 'library' -> 'lib'
    '^(post.*|templates)$' = '$&'  #  preserve, if prefixed with 'post' or equal to 'templates' 
    '.+' = ''                      #  remove all other tokens
  }

  $logFilename = ''; $i = 0
  # Split the .Directory path into tokens (path components) and synthesize
  # the log filename from a subset of the tokens, with transformations applied.
  foreach($token in ($path -split '\\')) {
    # Apply transformation.
    $transformedToken = $token
    foreach($regex in $ohtTansform.Keys) {
      if ($token -match $regex) {
        $transformedToken = $token -replace $regex, $ohtTansform[$regex]
        break
      }
    }
    # Add transformed token to filename, separated with "_".
    if ($transformedToken) {
      $logFilename += $(if ($i++) { '_' } else { '' }) + $transformedToken 
    }
  }
  # Output the synthesized filename.
  $logFilename

}


Import-Csv C:\test.csv | % {
  # Call the helper function to transform the directory path to a log filename.
  $LogFile = pathToLogFilename $_.Directory
}

---

为日志文件指定唯一名称的最简单方法是计数器1.log、2.log、3.log、4.log。。。n、 日志。这样行吗？你所描述的看起来不可思议的复杂-脚本怎么知道“VERICUT”很重要，而“CGTech”不重要？如果将“VERICUT”缩短为“VERI”，那么“VERICUT”是有意义的，但是“templates”不应该缩短吗？如果您只是交换\\and:for-并获取C程序文件CGTech VERICUT 7.4-library.log呢？@tessellingheckler您可能是对的。我只是希望缩短日志名。我可能会使用这个方法，并保持它的简单。请。一个只使用示例输入的硬编码解决方案是没有帮助的。

# Helper function that takes a directory path and converts it to a log filename
# representing that dir.
function pathToLogFilename([string] $path) {

  # Ordered hashtable defining regex-based transformations (replacements).
  # Entries are processed in order, and processing stops as soon as 
  # a match is found and the transformation has been performed.
  # Entry format: <regex> = <replacement>, to be passed to the -replace operator.
  $ohtTansform = [ordered] @{
    '^(.{3}).+?(E\d+).*$' = '$1$2' # 'CimatronE11_64bit' -> 'CimE11', for instance. 
    '^(.{4}).+? (\d.*)$' = '$1_$2' # 'VERICUT 7.3.3' -> 'VER_7.3.3', for instance
    '^(lib)rary$' = '$1'           # 'library' -> 'lib'
    '^(post.*|templates)$' = '$&'  #  preserve, if prefixed with 'post' or equal to 'templates' 
    '.+' = ''                      #  remove all other tokens
  }

  $logFilename = ''; $i = 0
  # Split the .Directory path into tokens (path components) and synthesize
  # the log filename from a subset of the tokens, with transformations applied.
  foreach($token in ($path -split '\\')) {
    # Apply transformation.
    $transformedToken = $token
    foreach($regex in $ohtTansform.Keys) {
      if ($token -match $regex) {
        $transformedToken = $token -replace $regex, $ohtTansform[$regex]
        break
      }
    }
    # Add transformed token to filename, separated with "_".
    if ($transformedToken) {
      $logFilename += $(if ($i++) { '_' } else { '' }) + $transformedToken 
    }
  }
  # Output the synthesized filename.
  $logFilename

}


Import-Csv C:\test.csv | % {
  # Call the helper function to transform the directory path to a log filename.
  $LogFile = pathToLogFilename $_.Directory
}