我今天刚刚在博客上写了这个。它是INotifyPropertyChanged属性周围的强类型响应式包装器。

GetPropertyValues返回一个IObservable<T>的值，当它们改变时，从当前值开始。如果忽略当前值，可以对结果调用Skip(1)。

用法如下:

IObservable<int> values = viewModel.GetPropertyValues(x => x.IntProperty);

实现:

public static class NotifyPropertyChangeReactiveExtensions
{
    // Returns the values of property (an Expression) as they change, 
    // starting with the current value
    public static IObservable<TValue> GetPropertyValues<TSource, TValue>(
        this TSource source, Expression<Func<TSource, TValue>> property)
        where TSource : INotifyPropertyChanged
    {
        MemberExpression memberExpression = property.Body as MemberExpression;

        if (memberExpression == null)
        {
            throw new ArgumentException(
                "property must directly access a property of the source");
        }

        string propertyName = memberExpression.Member.Name;

        Func<TSource, TValue> accessor = property.Compile();

        return source.GetPropertyChangedEvents()
            .Where(x => x.EventArgs.PropertyName == propertyName)
            .Select(x => accessor(source))
            .StartWith(accessor(source));
    }

    // This is a wrapper around FromEvent(PropertyChanged)
    public static IObservable<IEvent<PropertyChangedEventArgs>>
        GetPropertyChangedEvents(this INotifyPropertyChanged source)
    {
        return Observable.FromEvent<PropertyChangedEventHandler, 
            PropertyChangedEventArgs>(
            h => new PropertyChangedEventHandler(h),
            h => source.PropertyChanged += h,
            h => source.PropertyChanged -= h);
    }
}

我们有一个部署工具可以在不同环境之间进行部署。由于文件可以标记为修改，但实际上没有不同，我想出了这个:

/// <summary>
/// Compares the files to see if they are different. 
/// First checks file size
/// Then modified if the file is larger than the specified size
/// Then compares the bytes
/// </summary>
/// <param name="file1">The source file</param>
/// <param name="file2">The destination file</param>
/// <param name="mb">Skip the smart check if the file is larger than this many megabytes. Default is 10.</param>
/// <returns></returns>
public static bool IsDifferentThan(this FileInfo file1, FileInfo file2, int mb = 10)
{
  var ret = false;

  // different size is a different file
  if(file1.Length != file2.Length) return true;

  // if the file times are different and the file is bigger than 10mb flag it for updating
  if(file1.LastWriteTimeUtc > file2.LastWriteTimeUtc && file1.Length > ((mb*1024)*1024)) return true;

  var f1 = File.ReadAllBytes(file1.FullName);
  var f2 = File.ReadAllBytes(file2.FullName);

  // loop through backwards because if they are different
  // it is more likely that the last few bytes will be different
  // than the first few
  for(var i = file1.Length - 1; i > 0; i--)
  {
    if(f1[i] != f2[i])
    {
      ret = true;
      break;
    }
  }

  return ret;
}

一般的尝试:

class Program
{
    static void Main(string[] args)
    {
        var z = 0;
        var a = 0.AsDefaultFor(() => 1 / z);
        Console.WriteLine(a);
        Console.ReadLine();
    }
}

public static class TryExtensions
{
    public static T AsDefaultFor<T>(this T @this, Func<T> operation)
    {
        try
        {
            return operation();
        }
        catch
        {
            return @this;
        }
    }
}

如果你愿意，把它放到CodePlex项目上。

我总是使用格式，希望与StringBuilder的新行，所以下面这个非常简单的扩展节省了几行代码:

public static class Extensions
{
    public static void AppendLine(this StringBuilder builder,string format, params object[] args)
    {
        builder.AppendLine(string.Format(format, args));
    }
}

替代方案是StringBuilder中带\n或Environment.NewLine的AppendFormat。

在单元测试中很有用:

public static IList<T> Clone<T>(this IList<T> list) where T : ICloneable
{
    var ret = new List<T>(list.Count);
    foreach (var item in list)
        ret.Add((T)item.Clone());

    // done
    return ret;
}

像TWith2Sugars这样的一系列，交替缩短语法:

public static long? ToNullableInt64(this string val)
{
    long ret;
    return Int64.TryParse(val, out ret) ? ret : new long?();
}

最后，在BCL中是否已经有一些东西做了下面的事情?

public static void Split<T>(this T[] array, 
    Func<T,bool> determinator, 
    IList<T> onTrue, 
    IList<T> onFalse)
{
    if (onTrue == null)
        onTrue = new List<T>();
    else
        onTrue.Clear();

    if (onFalse == null)
        onFalse = new List<T>();
    else
        onFalse.Clear();

    if (determinator == null)
        return;

    foreach (var item in array)
    {
        if (determinator(item))
            onTrue.Add(item);
        else
            onFalse.Add(item);
    }
}

The Substring method on the string class has always felt inadequate to me. Usually when you do a substring, you know the character(s) from where you want to start, and the charachter(s) where you want to end. Thus, I've always felt that have to specify length as the second parameter is stupid. Therefore, I've written my own extension methods. One that takes a startIndex and an endIndex. And one, that takes a startText (string) and endText (string) so you can just specify the text from where to start the substring, and the text for where to end it.

注意:我不能将方法命名为。net中的Substring，因为我的第一个重载采用了与。net重载中的一个相同的参数类型。因此我将它们命名为Subsetstring。请随意添加到CodePlex…

public static class StringExtensions
{
    /// <summary>
    /// Returns a Subset string starting at the specified start index and ending and the specified end
    /// index.
    /// </summary>
    /// <param name="s">The string to retrieve the subset from.</param>
    /// <param name="startIndex">The specified start index for the subset.</param>
    /// <param name="endIndex">The specified end index for the subset.</param>
    /// <returns>A Subset string starting at the specified start index and ending and the specified end
    /// index.</returns>
    public static string Subsetstring(this string s, int startIndex, int endIndex)
    {
        if (startIndex > endIndex)
        {
            throw new InvalidOperationException("End Index must be after Start Index.");
        }

        if (startIndex < 0)
        {
            throw new InvalidOperationException("Start Index must be a positive number.");
        }

        if(endIndex <0)
        {
            throw new InvalidOperationException("End Index must be a positive number.");
        }

        return s.Substring(startIndex, (endIndex - startIndex));
    }

    /// <summary>
    /// Finds the specified Start Text and the End Text in this string instance, and returns a string
    /// containing all the text starting from startText, to the begining of endText. (endText is not
    /// included.)
    /// </summary>
    /// <param name="s">The string to retrieve the subset from.</param>
    /// <param name="startText">The Start Text to begin the Subset from.</param>
    /// <param name="endText">The End Text to where the Subset goes to.</param>
    /// <param name="ignoreCase">Whether or not to ignore case when comparing startText/endText to the string.</param>
    /// <returns>A string containing all the text starting from startText, to the begining of endText.</returns>
    public static string Subsetstring(this string s, string startText, string endText, bool ignoreCase)
    {
        if (string.IsNullOrEmpty(startText) || string.IsNullOrEmpty(endText))
        {
            throw new ArgumentException("Start Text and End Text cannot be empty.");
        }
        string temp = s;
        if (ignoreCase)
        {
            temp = s.ToUpperInvariant();
            startText = startText.ToUpperInvariant();
            endText = endText.ToUpperInvariant();
        }
        int start = temp.IndexOf(startText);
        int end = temp.IndexOf(endText, start);
        return Subsetstring(s, start, end);
    }
}

用法:

string s = "This is a tester for my cool extension method!!";
       s = s.Subsetstring("tester", "cool",true);

输出:"test for my "