MSDN Magazine - February 2009 - (Page 16)

as function parameters or return values. In the absence of exception handling, every caller needs to check for the success of the function that it called (or just assume everything is always OK). Most Win32 APIs return a non-zero value to indicate failure because Windows doesn’t use exception handling. The programmer has to wrap every function call with code that checks the return dwError = GetLastError(); if (dwError != ERROR_NO_MORE_FILES) { ErrorHandler(TEXT("FindFirstFile")); } FindClose(hFind); return dwError; Like motorcycles and chain saws, the raw power of exceptions makes them useful. value of the called function. For example, this code from MSDN documentation on how to list files in a directory explicitly checks each call for success. The call to FindNextFile( ) is wrapped in a check to see if the return is non-zero. If the call is not successful then a separate function call—GetLastError()—provides details of the exceptional condition. Note that every call must be checked for success on the next frame because return values are necessarily limited to the local function scope: // FindNextFile requires checking for success of each call while (FindNextFile(hFind, &ffd) != 0); An error condition can only pass from the function containing the unexpected condition to that function’s caller. Exceptions have the power to pass the results of a function’s execution out of the current function’s scope to every frame up the stack until it reaches the frame that knows how to handle the unexpected condition. The CLR’s exception system (called a two-pass exception system) delivers the exception to every predecessor on the thread’s call stack, beginning with the caller and proceeding until some function says it will handle the exception (this is known as the first pass). The exception system will then unwind the state of each frame on the call stack between where the exception is raised and where it will be handled (known as the second pass). As the stack unwinds, the CLR will run both finally clauses and fault clauses in each frame as it is unwound. Then, the catch clause in the handling frame is executed. Because the CLR checks every predecessor on the call stack there’s no need for the caller to have a catch block—the exception can be caught anywhere up the stack. Instead of having code to immediately check the result of every function call, a programmer can handle errors in a place far away from where the exception NCache TM Distributed Object Cache for .NET Boost .NET application scalability in server farms! NCache is an extremely fast & scalable in-memory distributed cache for .NET. With an API very similar to ASP.NET Cache, NCache synchronizes the cache across multiple servers to let you scale out. Enterprise Distributed Data Cache Scale with Replicated, Partitioned, Client Cache topologies 100% up time with dynamic cache clustering Use in ASP.NET, SOA & Grid Computing Distributed ASP.NET Sessions Eliminate session storage bottlenecks Use fast & scalable in-memory session storage Incorporate without any code change Ph: +1-925-236-3830 16 msdn magazine Download FREE Trial www.alachisoft.com CLR Inside Out http://www.alachisoft.com http://www.alachisoft.com

Table of Contents Feed for the Digital Edition of MSDN Magazine - February 2009

MSDN Magazine - February 2009 Contents Toolbox CLR Inside Out Data Points Cutting Edge Patterns In Practice Best Practices .Net Interop "Oslo" Basics Patterns Silverlight Under The Table Foundations Windows With C++ .NET Matters Going Places { End Bracket }

MSDN Magazine - February 2009

For optimal viewing of this digital publication, please enable JavaScript and then refresh the page. If you would like to try to load the digital publication without using Flash Player detection, please click here.