Fast and simple binary concatenate files in Powershell

前端 未结 4 1431
耶瑟儿~
耶瑟儿~ 2020-12-05 10:32

What\'s the best way of concatenating binary files using Powershell? I\'d prefer a one-liner that simple to remember and fast to execute.

The best I\'ve come up with

4条回答
  •  悲哀的现实
    2020-12-05 10:55

    Performance is very much dependent on the buffer size used. Those are fairly small by default. Concatenating 2x2GB files I'd take a buffersize of about 256kb. Going larger might sometimes fail, smaller and you'll get less throughput than your drive is capable of.

    With gc that'd be with -ReadCount not simply -Read (PowerShell 5.0):

    gc -ReadCount 256KB -Path $infile -Encoding Byte | ...
    

    Plus I found Add-Content to be better and going file-by-file for a lot of small files, because piping only a moderate amount of data (200MB) I found my computer going oom, PowerShell freezing and CPU at full.

    Although Add-Content randomly fails a few times for a few hundred files with an error about the destination file being in use, so I added a while loop and a try catch:

    # Empty the file first
    sc -Path "$path\video.ts" -Value @() -Encoding Byte 
    $tsfiles | foreach {    
        while ($true) {
            try { # I had -ReadCount 0 because the files are smaller than 256KB
                gc -ReadCount 0 -Path "$path\$_" -Encoding Byte | `
                    Add-Content -Path "$path\video.ts" -Encoding Byte -ErrorAction Stop
                break;
            } catch {
            }
        }
    }
    

    Using a file stream is much faster still. You cannot specify a buffer size with [System.IO.File]::Open but you can with new [System.IO.FileStream] like so:

    # $path = "C:\"
    $ins = @("a.ts", "b.ts")
    $outfile = "$path\out.mp4"
    $out = New-Object -TypeName "System.IO.FileStream" -ArgumentList @(
        $outfile, 
        [System.IO.FileMode]::Create,
        [System.IO.FileAccess]::Write,
        [System.IO.FileShare]::None,
        256KB,
        [System.IO.FileOptions]::None)
    try {
        foreach ($in in $ins) {
            $fs = New-Object -TypeName "System.IO.FileStream" -ArgumentList @(
                "$path\$in", 
                [System.IO.FileMode]::Open,
                [System.IO.FileAccess]::Read,
                [System.IO.FileShare]::Read,
                256KB,
                [System.IO.FileOptions]::SequentialScan)
            try {
                $fs.CopyTo($out)
            } finally {
                $fs.Dispose()
            }
        }
    } finally {
        $out.Dispose()
    }
    

提交回复
热议问题