删除 C# 中的零件文件名重复项

Question

I have a directory filled with files that have part of the name as duplicate,

eg

"afilename[a].txt"

"afilename[f].txt"

"afilename[j].txt"

I would like to delete all files, but the first file, that contain "afilename", so I will be left with "afilename[a].txt"

I have the following code,

var duplicateNames = files.GroupBy(file => file.Name)
                          .Where(group => group.Count() > 1)
                          .Select(group => group.Key);

But is it possible to compare the filenames upto the "[" and then return a list of the full file name, but skipping the first file found?

Many thanks.

----------Solution I used was: -

using System;

public class Program
{
    public static void Main()
    {

        // get all files
        string directoryPath = "C:\\temp\\dups";
        var allFiles = Directory.GetFiles(directoryPath, "*.mp3", SearchOption.AllDirectories);

        var toKeep = allFiles.Where(file => file.Contains('['))
                 .GroupBy(file => file.Remove(file.IndexOf('[')))
                 .Select(group => group.Min())
                 .ToHashSet();

        var whatToDelete = allFiles.Except(toKeep);

        foreach (var fileToDelete in whatToDelete)
        {
            System.IO.File.Delete(fileToDelete);
        }
    }
}

Answer 1

Sure. Let's take only files with [ in the name, group them by only the part of the string up to that, order the result, skip the first one and expand the group back into a list of names that should be removed

var duplicateNames = files.Where(file => file.Contains('[')
                      .GroupBy(file => file.Name.Remove(file.IndexOf('[')))
                      .Where(group => group.Count() > 1)
                      .SelectMany(group => group.OrderBy(file => file).Skip(1));

The only thing you might need to think about, is that filenames on windows are not case sensitive, but strings in c# are, so aFileName[a].txt and aFILEname[b].txt are probably a "b should be deleted" but this won't pick them up. Perhaps lowercase the result of the Remove when you group, and when you OrderBy..

Note; you've said "i want to remove all except the first", but you haven't directly said anything about what you consider to be a first. If your lists are already in the order you want them, then you could ditch the ordering before the Skip because generally LINQ preserves order, but I'd advise you to impose some order so that you can be sure of what will be deleted, rather than rely on some innate ordering of the input.

Another approach, incidentally, might be easier:

var toKeep = files.Where(file => file.Contains('[')
                 .GroupBy(file => file.Name.Remove(file.IndexOf('[')))
                 .Select(group => group.Min())
                 .ToHashSet()

This will generate a hashset of the files you want to keep. You could either then do files.Except(toKeep) to generate the list to remove, or just loop over files issuing delete commands for any file not in toKeep . The same case-sens considerations apply to Min() as they do to GroupBy , in this latter example just like the first