Directory.GetFiles 使用 SearchPattern 返回意外结果

Question

I'm working on a batch program that process a big amount of files (more than 50 000 files) and I'm facing weird behavior with the Directory.GetFiles method.

In the process I move files that matches the following search pattern "*.pdf", and get the files thanks to the Directory.GetFiles method : I was very surprised to see that sometimes I have .pdfa files moved.

So I've checked the doc and it clearly states that if the search pattern contains an extension with 3 letters every files that have an extension that begins with the extension will be returned.

I've tested with a simple program and It does not behave like stated in the doc, it only behaves like this in very rare occasion.

With this code :

static void Main(string[] args) {
    var directory = @"E:\Test\";
    var files = Directory.GetFiles(directory, "*.pdf");
    foreach(var file in files)
        Console.WriteLine(file);
}

I have this result :

Do you have any explanation about this behavior ?

Answer 1

This is the expected behavior of the GetFiles method and it's same on Windows as well, if you search in directory with .pdf it will pick files with extensions.pdfa or *.pdfaaa , you would need to put a Where() yourself like:

Directory.GetFiles(directory, "*.pdf").Where(item => item.EndsWith(".pdf"));

As you can see that when we search in windows It is giving the same result as your code was giving:

For the reason of why the GetFiles is behaving that way please have a look here and you might also want to look at this post as well

Answer 2

As explained by @luaan and by @hans-passant (thanks a lot !) I do not found the file with the .pdfa extension, because the 8.3 format is disabled on my hard drive.

On a hard drive with the 8.3 format enabled, the method behaves like stated in the doc.

The GetFiles has a different behavior with the setting enabled or not.