How do I retrieve only one specific file from a 7z instead of unzipping and dumping to a folder?

Question

Currently, I am using 7ZipCLI to unzip my .7z folder to a specified folder destPath like this:

private void ExtractFile(string archivePath, string destPath)
{
    string zpath = @"C:\Program Files\7-Zip\x64\7za.exe";
    try
    {
        ProcessStartInfo pro = new ProcessStartInfo();
        pro.WindowStyle = ProcessWindowStyle.Hidden;
        pro.FileName = zpath;
        pro.Arguments = string.Format("x \"{0}\" -y -o\"{1}\"", archivePath, destPath);
        Process x = Process.Start(pro);
        x.WaitForExit();
    }
    catch (System.Exception Ex)
    {
        Console.WriteLine("{0} Exception: ", Ex)
    }
} 

This takes a long time because the application unzips the folder, unloads it into destPath then does a search within destPath for the specified file. How would I instead look into the .7z, find the specified file and copy just that file to destPath?

Answer 1

If you want to do interesting things with archives I would suggest that you use a library instead of rolling your own process execution solution. There are NuGet packages for just about every task, and this is no exception.

For instance, SharpCompress is a fairly common library that handles decompressing 7z archives well enough for most uses. Add that to your project and try something like this:

// Usings:
//  SharpCompress.Archives;
//  SharpCompress.Common;
//  System.Linq;

private static bool ExtractFile(string archivePath, string destPath, string fileSubstring)
{
    using (var archive = ArchiveFactory.Open(archivePath))
    {
        var entry = archive.Entries.FirstOrDefault(e => e.Key.Contains(fileSubstring));
        if (entry != null)
        {
            var opt = new ExtractionOptions
            {
                ExtractFullPath = false,
                Overwrite = true
            };
            try
            {
                entry.WriteToDirectory(destPath, opt);
                return true;
            }
            catch { }
        }
    }
    return false;
}

That's a simple example. You could pass in a filter predicate and process multiple results, whatever fits your requirements.

Running a few tests here using SysInternals ProcMon to confirm, this does not create extraneous files and works quickly pulling little files out of big archives.

And as a bonus it doesn't care what type of archive you give it, as long as it's one that's supported by the library. It'll read RAR, ZIP, 7z and a host of others, and you can use the same library to do compression for a few common formats if needed.