8
3
In an effort to reduce code duplication, I often use and have used this style to capture handling of exceptions on a boundary of an application: Given the following extension methods:
public static class FuncExtenstion
{
[DebuggerStepThrough]
public static Action<A> AsAction<A, T>(this Func<A, T> f)
{
return (a) => f(a);
}
[DebuggerStepThrough]
public static Func<bool> AsFunc(this Action a)
{
return () => { a(); return true; };
}
}
public static class IlogExtensions
{
public static TResult TryCatchLogThrow<TResult>(this ILog logger, Func<TResult> f)
{
try
{
return f();
}
catch (Exception ex)
{
logger.Error(ex.Message, ex);
throw;
}
}
public static void TryCatchLogThrow(this ILog logger, Action f)
{
logger.TryCatchLogThrow(f.AsFunc());
}
public static TResult TryCatchLog<TResult>(this ILog logger, Func<TResult> f)
{
try
{
return f();
}
catch (Exception ex)
{
logger.Error(ex.Message, ex);
return default(TResult);
}
}
public static void TryCatchLog(this ILog logger, Action f)
{
logger.TryCatchLog(f.AsFunc());
}
}
In the code I will use these (e.g. in a Windows/WCF/Web Service) to either silently catch the error and let the code continue to run (used in a polling service) or to at least log the error locally on the server and then rethrow the full exception again. I have used several other variations where the exception gets wrapped in more 'user friendly' exceptions or in WCF
FaultExceptions
So typically this code is used as follows:
Response ISomeService.SomeOperation(Request request)
{
return _logger.TryCatchLogThrow(() => _domainImplementation.SomeOperation(request));
}
OtherResponse ISomeService.OtherOperation(OtherRequest request)
{
return _logger.TryCatchLogThrow(() => _domainImplementation.OtherOperation(request));
}
I have not seen this style of code anywhere else. So I was wondering if there is an other pattern I should be using, or if this is ok to use.
from : https://codereview.stackexchange.com/questions/11999/encapsulating-common-try-catch-code-is-this-a-known-pattern-is-it-good-or-bad
沒有留言:
張貼留言