Logo Questions Linux Laravel Mysql Ubuntu Git Menu
 

Fastest/safest file finding/parsing?

On c:, I have tens of thousands of *.foobar files. They're in all sorts of places (i.e. subdirs). These files are roughly 1 - 64 kb in size, and plaintext.

I have a class Foobar(string fileContents) that strongly types these .foobar files.

My challenge to is get a list of all the *.foobar files on c:, represented as an array of Foobar objects. What's the quickest way to do this?

I'm interested to find out if there's a better way (undoubtedly) than my first approach, which follows, and if this approach of mine has any potential problems (e.g. I/O concurrency issues throwing exceptions?):

var files = Directory.EnumerateFiles
                (rootPath, "*.foobar", SearchOption.AllDirectories);

Foobar[] foobars = 
(
    from filePath in files.AsParallel()
    let contents = File.ReadAllText(filePath)
    select new Foobar(contents)
)
.ToArray();
like image 758
user979672 Avatar asked Mar 17 '12 01:03

user979672


1 Answers

Because permission errors (or other errors) can apparently stop the enumeration dead in its tracks, you may want to implement your own enumerator something like this:

class SafeFileEnumerator : IEnumerable<string>
{
  private string root;
  private string pattern;
  private IList<Exception> errors;
  public SafeFileEnumerator(string root, string pattern)
  {
     this.root = root;
     this.pattern = pattern;
     this.errors = new List<Exception>();
  }

  public SafeFileEnumerator(string root, string pattern, IList<Exception> errors)
  {
     this.root = root;
     this.pattern = pattern;
     this.errors = errors;
  }

  public Exception[] Errors()
  {
     return errors.ToArray();
  }
  class Enumerator : IEnumerator<string>
  {
     IEnumerator<string> fileEnumerator;
     IEnumerator<string> directoryEnumerator;
     string root;
     string pattern;
     IList<Exception> errors;

     public Enumerator(string root, string pattern, IList<Exception> errors)
     {
        this.root = root;
        this.pattern = pattern;
        this.errors = errors;
        fileEnumerator = System.IO.Directory.EnumerateFiles(root, pattern).GetEnumerator();
        directoryEnumerator = System.IO.Directory.EnumerateDirectories(root).GetEnumerator();
     }
     public string Current
     {
        get
        {
           if (fileEnumerator == null) throw new ObjectDisposedException("FileEnumerator");
           return fileEnumerator.Current;
        }
     }

     public void Dispose()
     {
        if (fileEnumerator != null)
           fileEnumerator.Dispose();
        fileEnumerator = null;
        if (directoryEnumerator != null)
           directoryEnumerator.Dispose();
        directoryEnumerator = null;
     }

     object System.Collections.IEnumerator.Current
     {
        get { return Current; }
     }

     public bool MoveNext()
     {
        if ((fileEnumerator != null) && (fileEnumerator.MoveNext()))
           return true;
        while ((directoryEnumerator != null) && (directoryEnumerator.MoveNext()))
        {
           if (fileEnumerator != null)
              fileEnumerator.Dispose();
           try
           {
              fileEnumerator = new SafeFileEnumerator(directoryEnumerator.Current, pattern, errors).GetEnumerator();
           }
           catch (Exception ex)
           {
              errors.Add(ex);
              continue;
           }
           if (fileEnumerator.MoveNext())
              return true;
        }
        if (fileEnumerator != null)
           fileEnumerator.Dispose();
        fileEnumerator = null;
        if (directoryEnumerator != null)
           directoryEnumerator.Dispose();
        directoryEnumerator = null;
        return false;
     }

     public void Reset()
     {
        Dispose();
        fileEnumerator = System.IO.Directory.EnumerateFiles(root, pattern).GetEnumerator();
        directoryEnumerator = System.IO.Directory.EnumerateDirectories(root).GetEnumerator();
     }
  }
  public IEnumerator<string> GetEnumerator()
  {
     return new Enumerator(root, pattern, errors);
  }

  System.Collections.IEnumerator System.Collections.IEnumerable.GetEnumerator()
  {
     return GetEnumerator();
  }
}
like image 151
BlueMonkMN Avatar answered Oct 05 '22 02:10

BlueMonkMN