MSDN Magazine - October 2007 - (Page 17)

IronPython and the Dynamic Language Runtime BILL CHILES T he Microsoft® .NET Framework was designed to support a broad range of programming languages on a single runtime—the common language runtime (CLR). The CLR provides shared services to these languages, including garbage collection, just-in-time (JIT) compilation, a sandboxed security model, and support for tools integration. Sharing these features has two huge benefits for language implementers. It is easier to implement a language because much of the lower-level engineering work is already done. And languages running on the CLR can integrate seamlessly. Sharing libraries and frameworks allows languages new to the CLR to build on the work of others. The CLR also includes support for dynamic languages, as IronPython 1.0 demonstrates (www.codeplex.com/ironpython). Dynamic programming languages deliver a powerful and productive experience for programmers, as James Schementi explained in the October 2006 installment of “CLR Inside Out” (msdn.microsoft.com/msdnmag/ issues/06/10/CLRInsideOut). Now Microsoft is building the Dynamic Language Runtime (DLR), which adds to the CLR a set of services designed specifically to meet the needs of dynamic languages. The DLR adds functionality such as a shared dynamic type system, a standard hosting model, and support to make it easy to generate fast dynamic code and fast symbol tables. With these additional features, it is much easier to build high-quality dynamic language implementations for .NET. Furthermore, these features enable dynamic languages built on the DLR to share libraries written in other dynamic languages or in CLR-based static languages. The DLR lets you build the best experience for your favorite dynamic language. This means your experience coding in your language should feel like it always does. You should also benefit from excellent tools, great performance, and shared libraries. Because using the DLR means you can share standard functionality, you have more time to focus on the features that make a language unique. For example, you can avoid having to implement garbage collection, code generation, and method caching. The DLR uses adaptive method caching to produce fast dynamic programs, and all language implementations on the DLR benefit from this shared work. Code runs faster by avoiding repeated searching of class precedence lists. It also avoids resolving overloads on .NET methods each time you call on an object (for example, call a method, foo, on some object, o). The source code for IronPython and the DLR is available on Codeplex under the Berkeley Software Distribution (BSD)-style Microsoft Permissive License (see the 2.0 alpha release). This is a very early release, and there is still significant work to be done. There are low-level mechanisms to design, performance enhancements to implement, and documentation to write. By the time you read this, there will be Codeplex projects available for the DLR code itself and for IronRuby, in addition to IronPython. Though we are working on DLR JScript® and VBX implementations, these may be available only in binary form. IronPython Hello World I’ll walk through a simple Python program (a twist on Hello World) and discuss what happens when it executes in ipy.exe. The program I’ll examine is msdnmag.py: def yo(yourname): text = “hello, “ return text + yourname print yo(“bill”) I’ve added something a little beyond the classic “print hello” functionality, so I can discuss more interesting aspects of the IronPython implementation. For instance, I’ll be able to demonstrate name binding for identifier references and caching the meaning of “+” in the yo function. I’m using IronPython 2.0 alpha, so take a few minutes now to download it so you can walk through it with me. To begin, open the ironpython.sln file in Visual Studio® 2005 and build the solution with the Debug solution configuration. Then right-click on the IronPythonConsole project and select the Debug tab in the left navigation bar. Set this project as the start project (first set of radio buttons under Start Action). Next, under Start Options, enter msdnmag.py as the command-line argument. Finally, copy the msdnmag.py Python code listed above into a file called msdnmag.py in the \bin\debug\ directory that was created when you built the solution. The DLR adds a shared dynamic type system, a standard hosting model, and support to make it easy to generate fast dynamic code and fast symbol tables. Launching Ipy.exe Now let’s see how ipy.exe starts up. The DLR provides some default console classes so that language implementers can easily get This column is based on a prerelease version of the Dynamic Language Runtime. Details contained herein are subject to change. october2007 17 http://www.codeplex.com/ironpython http://msdn.microsoft.com/msdnmag/issues/06/10/CLRInsideOut http://msdn.microsoft.com/msdnmag/issues/06/10/CLRInsideOut

Table of Contents Feed for the Digital Edition of MSDN Magazine - October 2007

Cover Contents Toolbox CLR Inside Out Basic Instincts Data Points Cutting Edge Pooled Threads WPF Threads Parallel Linq Parallel Performance Mobile Apps Test Run Foundations Windows with C++ Netting C++ .NET Matters { End Bracket } Net Nuptials

MSDN Magazine - October 2007

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.