Question

如何使用C＃遍历文件夹结构而不会陷入junction points的陷阱？

Answer 1

对于那些不知道的人：连接点的行为类似于linux上文件夹的符号链接。设置递归文件夹结构时会发生提到的陷阱，如下所示：

given folder /a/b
let /a/b/c point to /a
then
/a/b/c/b/c/b becomes valid folder locations.

我建议像这样的策略。在Windows上，您被限制为路径字符串上的最大长度，因此递归解决方案可能不会破坏堆栈。

private void FindFilesRec(
    string newRootFolder,
    Predicate<FileInfo> fileMustBeProcessedP,
    Action<FileInfo> processFile)
{
    var rootDir = new DirectoryInfo(newRootFolder);
    foreach (var file in from f in rootDir.GetFiles()
                         where fileMustBeProcessedP(f)
                         select f)
    {
        processFile(file);
    }

    foreach (var dir in from d in rootDir.GetDirectories()
                        where (d.Attributes & FileAttributes.ReparsePoint) != FileAttributes.ReparsePoint
                        select d)
    {
        FindFilesRec(
            dir.FullName,
            fileMustBeProcessedP,
            processFile);
    }
}

Answer 2

您可以使用以下代码：

private void processing(string directory)
        {
            cmbFilesTypesSelectedIndex = cmbFilesTypes.SelectedIndex;
            CheckForProjectFile(directory);
            DirectoryInfo dInfo = new DirectoryInfo(directory);
            DirectoryInfo[] dirs = dInfo.GetDirectories() ;
            foreach (DirectoryInfo subDir in dirs)
            {
                CheckForProjectFile(subDir.FullName);
                processing(subDir.FullName);
            }
        }

        private void CheckForProjectFile(string directory)
        {
            Boolean flag = false; 
            DirectoryInfo dirInfo = new DirectoryInfo(directory);
            FileInfo[] files = dirInfo.GetFiles();
            //You can also traverse in files also
            foreach (FileInfo subfile in files)
            {
                //Do you want

            }
        }

C＃中的目录遍历

2 个答案: