在 c:
上,我有数以万计的 *.foobar
文件。它们在各种各样的地方(即子目录)。这些文件的大小大约为 1 - 64 kb,并且是纯文本。
我有一个 class Foobar(string fileContents)
可以强类型化这些 .foobar
文件。
我的挑战是获取 c:
上所有 *.foobar
文件的列表,表示为 Foobar
对象数组。最快的方法是什么?
我很想知道是否有比我的第一种方法更好的方法(毫无疑问),以及我的这种方法是否有任何潜在问题(例如 I/O 并发问题引发异常?):
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();
最佳答案
因为权限错误(或其他错误)显然会使枚举停止在其轨道上,您可能想要实现自己的枚举器,如下所示:
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();
}
}
关于c# - 最快/最安全的文件查找/解析?,我们在Stack Overflow上找到一个类似的问题: https://stackoverflow.com/questions/9746538/