Efficient use of 'Directory.GetFiles' with 'StartsWith', 'Contains' and 'EndsWith' to filter file names

Question

Is there a more efficient way to filter file names using Directory.GetFiles and 'StartsWith', 'Contains' and 'EndsWith' with rather than the way I am currently doing it?

        _files = Directory.GetFiles(_path);

        if (!String.IsNullOrEmpty(_startsWith))
        {
            _files = _files.Where(x => x.StartsWith(_startsWith)).ToArray();
        }

        if (!String.IsNullOrEmpty(_contains))
        {
            _files = _files.Where(x => x.Contains(_contains)).ToArray();
        }

        if (!String.IsNullOrEmpty(_endsWith))
        {
            _files = _files.Where(x => x.EndsWith(_endsWith)).ToArray();
        }

The very slowest part will always be `Directory.GetFiles`, which is _almost_ unavoidable. Searching for the strings is comparatively negligible, that's not what you want to improve, is it? — Grant Thomas, Mar 05 '13 at 12:02
There is an overload of GetFiles() where you can provide a searchpattern, have you tried that? — Jobo, Mar 05 '13 at 12:04

score 2 · Accepted Answer · answered Mar 05 '13 at 12:30

2

You should switch to Directory.EnumerateFiles() cause it is lazy and doesn't need to built up the complete list in the first place.

answered Mar 05 '13 at 12:30

Oliver

43,366
8
94
151

score 0 · Answer 2 · answered Mar 05 '13 at 12:10

I think following overloads will help you:

Directory.GetFiles(strPath, "*" + strFilter, SearchOption.TopDirectoryOnly); // to get behaviour of EndsWith

Directory.GetFiles(strPath, "*" + strFilter + "*", SearchOption.TopDirectoryOnly); // to get behaviour of Contains

Directory.GetFiles(strPath, strFilter + "*", SearchOption.TopDirectoryOnly); // to get behaviour of StartsWith

score 0 · Answer 3 · answered Mar 05 '13 at 12:11

If you are trying to cram everything into a single line of Linq, you can do something like the following:

_files = _files.Where(_ => !String.IsNullOrEmpty(_startsWith)).Where(x => x.StartsWith(_startsWith)).ToArray();

However, as mentioned above, the Directory.GetFiles will almost certainly be the slowest part of the code here and you are better off sticking with something which is easier to read.

Efficient use of 'Directory.GetFiles' with 'StartsWith', 'Contains' and 'EndsWith' to filter file names

3 Answers3