解压缩适用于单线程，但不适用于多线程 [英] Unzipping works on singlethread, but not multithread

问题描述

我正在尝试使用PowerShell解压缩大量文件.我认为这是并行化的好地方.但是，即使它在单线程模式下工作，我进行并行化的尝试似乎也无法使它解压缩.

I'm trying to unzip a ton of files using PowerShell. I figured this is a great place to parallelize. However, my attempt to parallelize seems to make the unzip have no effect, even though it worked in the single threaded mode.

$unzip = {
    param([string]$sourceFile, [string]$destinationDir)
    #Clean out the destination if it exists
    rmdir $destination -Force -Recurse -ErrorAction SilentlyContinue
    mkdir $destination -Force

    #Actual unzip
    $shell = new-object -com shell.application
    $zipFile = $shell.NameSpace($sourceFile)
    $destinationDir = $shell.NameSpace($destination)
    $destinationDir.copyhere($zipFile.items())
}

foreach($file in $files){
    $args = ($file.FullName, $destinationDir)
    Start-Job $unzip -ArgumentList $args
}

#Cleanup
While (Get-Job -State "Running") { Start-Sleep 2 }
Remove-Job *

当我在没有多线程代码的情况下运行此程序时，它可以正常工作，但实际上没有文件解压缩.为什么会这样?

When I run this without the multithreaded code, it works fine, but with it none of the files actually get unzipped. Why is this?

推荐答案

不确定示例是否复制粘贴，但参数为$ destinationDir，但您引用$ destination，然后使用$ destination创建$ destinationDir.我假设这是一个错字.我修复了您的功能，它可以按预期工作.

Not sure if your sample is copy pasted or not, but your param is $destinationDir but you reference $destination and then use $destination to create $destinationDir. I'm assuming this is a typo. I fixed your function and it works as you would expect.

$unzip = {
    param([string]$sourceFile, [string]$destination)
    #Clean out the destination if it exists
    rmdir $destination -Force -Recurse -ErrorAction SilentlyContinue
    mkdir $destination -Force

    #Actual unzip
    $shell = new-object -com shell.application
    $zipFile = $shell.NameSpace($sourceFile)
    $destinationDir = $shell.NameSpace($destination)
    $destinationDir.copyhere($zipFile.items())
}

使用接收工作会向您显示以下错误，指示您朝着正确的方向:

Using receive-job would have shown you the following errors pointing you in the right direction:

Cannot bind argument to parameter 'Path' because it is null.
    + CategoryInfo          : InvalidData: (:) [mkdir], ParameterBindingValidationException
    + FullyQualifiedErrorId : ParameterArgumentValidationErrorNullNotAllowed,mkdir
    + PSComputerName        : localhost

Method invocation failed because [System.String] doesn't contain a method named 'copyhere'.
    + CategoryInfo          : InvalidOperation: (:) [], RuntimeException
    + FullyQualifiedErrorId : MethodNotFound
    + PSComputerName        : localhost

如果可能的话，我仍然建议从shell.application的comobject转移到使用System.IO.Compression的.Net解决方案.还要注意，PowerShell作业的硬上限为5个同时运行的作业.我不确定这在v5中是否已解决. CookieMonster撰写了优秀文章和功能使用基于某些

I would still recommend moving away from the comobject for shell.application and to a .Net solution using System.IO.Compression if possible however. Also note that powershell jobs have a hard cap of 5 concurrently running jobs. I'm not sure if this is fixed in v5 or not. CookieMonster wrote an excellent post and a function using runspaces based off some work by Boe Prox as a better way of handling concurrency with improved performance.

这篇关于解压缩适用于单线程，但不适用于多线程的文章就介绍到这了，希望我们推荐的答案对大家有所帮助，也希望大家多多支持IT屋！