\chapter{Building C and \Cpp{} Extensions on Windows%     \label{building-on-windows}}  This chapter briefly explains how to create a Windows extension modulefor Python using Microsoft Visual \Cpp, and follows with moredetailed background information on how it works.  The explanatorymaterial is useful for both the Windows programmer learning to buildPython extensions and the \UNIX{} programmer interested in producingsoftware which can be successfully built on both \UNIX{} and Windows. Module authors are encouraged to use the distutils approach forbuilding extension modules, instead of the one described in thissection. You will still need the C compiler that was used to buildPython; typically Microsoft Visual \Cpp. \begin{notice}  This chapter mentions a number of filenames that include an encoded  Python version number.  These filenames are represented with the  version number shown as \samp{XY}; in practive, \character{X} will  be the major version number and \character{Y} will be the minor  version number of the Python release you're working with.  For  example, if you are using Python 2.2.1, \samp{XY} will actually be  \samp{22}.\end{notice}  \section{A Cookbook Approach \label{win-cookbook}} There are two approaches to building extension modules on Windows,just as there are on \UNIX: use the\ulink{\module{distutils}}{../lib/module-distutils.html} package tocontrol the build process, or do things manually.  The distutilsapproach works well for most extensions; documentation on using\ulink{\module{distutils}}{../lib/module-distutils.html} to build andpackage extension modules is available in\citetitle[../dist/dist.html]{Distributing Python Modules}.  Thissection describes the manual approach to building Python extensionswritten in C or \Cpp. To build extensions using these instructions, you need to have a copyof the Python sources of the same version as your installed Python.You will need Microsoft Visual \Cpp{} ``Developer Studio''; projectfiles are supplied for V\Cpp{} version 7.1, but you can use olderversions of V\Cpp.  Notice that you should use the same version ofV\Cpp that was used to build Python itself. The example filesdescribed here are distributed with the Python sources in the\file{PC\textbackslash example_nt\textbackslash} directory. \begin{enumerate}  \item  \strong{Copy the example files}\\    The \file{example_nt} directory is a subdirectory of the \file{PC}    directory, in order to keep all the PC-specific files under the    same directory in the source distribution.  However, the    \file{example_nt} directory can't actually be used from this    location.  You first need to copy or move it up one level, so that    \file{example_nt} is a sibling of the \file{PC} and \file{Include}    directories.  Do all your work from within this new location.   \item  \strong{Open the project}\\    From V\Cpp, use the \menuselection{File \sub Open Solution}    dialog (not \menuselection{File \sub Open}!).  Navigate to and    select the file \file{example.sln}, in the \emph{copy} of the    \file{example_nt} directory you made above.  Click Open.   \item  \strong{Build the example DLL}\\    In order to check that everything is set up right, try building:     \begin{enumerate}      \item        Select a configuration.  This step is optional.  Choose        \menuselection{Build \sub Configuration Manager \sub Active         Solution Configuration} and select either \guilabel{Release}         or\guilabel{Debug}.  If you skip this step,        V\Cpp{} will use the Debug configuration by default.       \item        Build the DLL.  Choose \menuselection{Build \sub Build        Solution}.  This creates all intermediate and result files in        a subdirectory called either \file{Debug} or \file{Release},        depending on which configuration you selected in the preceding        step.    \end{enumerate}   \item  \strong{Testing the debug-mode DLL}\\    Once the Debug build has succeeded, bring up a DOS box, and change    to the \file{example_nt\textbackslash Debug} directory.  You    should now be able to repeat the following session (\code{C>} is    the DOS prompt, \code{>>>} is the Python prompt; note that    build information and various debug output from Python may not    match this screen dump exactly): \begin{verbatim}C>..\..\PCbuild\python_dAdding parser accelerators ...Done.Python 2.2 (#28, Dec 19 2001, 23:26:37) [MSC 32 bit (Intel)] on win32Type "copyright", "credits" or "license" for more information.>>> import example[4897 refs]>>> example.foo()Hello, world[4903 refs]>>>\end{verbatim}     Congratulations!  You've successfully built your first Python    extension module.   \item  \strong{Creating your own project}\\    Choose a name and create a directory for it.  Copy your C sources    into it.  Note that the module source file name does not    necessarily have to match the module name, but the name of the    initialization function should match the module name --- you can    only import a module \module{spam} if its initialization function    is called \cfunction{initspam()}, and it should call    \cfunction{Py_InitModule()} with the string \code{"spam"} as its    first argument (use the minimal \file{example.c} in this directory    as a guide).  By convention, it lives in a file called    \file{spam.c} or \file{spammodule.c}.  The output file should be    called \file{spam.dll} or \file{spam.pyd} (the latter is supported    to avoid confusion with a system library \file{spam.dll} to which    your module could be a Python interface) in Release mode, or    \file{spam_d.dll} or \file{spam_d.pyd} in Debug mode.     Now your options are:     \begin{enumerate}      \item  Copy \file{example.sln} and \file{example.vcproj}, rename             them to \file{spam.*}, and edit them by hand, or      \item  Create a brand new project; instructions are below.    \end{enumerate}     In either case, copy \file{example_nt\textbackslash example.def}    to \file{spam\textbackslash spam.def}, and edit the new    \file{spam.def} so its second line contains the string    `\code{initspam}'.  If you created a new project yourself, add the    file \file{spam.def} to the project now.  (This is an annoying    little file with only two lines.  An alternative approach is to    forget about the \file{.def} file, and add the option    \programopt{/export:initspam} somewhere to the Link settings, by    manually editing the setting in Project Properties dialog).   \item  \strong{Creating a brand new project}\\    Use the \menuselection{File \sub New \sub Project} dialog to    create a new Project Workspace.  Select \guilabel{Visual C++    Projects/Win32/ Win32 Project}, enter the name (\samp{spam}), and    make sure the Location is set to parent of the \file{spam}    directory you have created (which should be a direct subdirectory    of the Python build tree, a sibling of \file{Include} and    \file{PC}).  Select Win32 as the platform (in my version, this is    the only choice).  Make sure the Create new workspace radio button    is selected.  Click OK.     You should now create the file \file{spam.def} as instructed in    the previous section. Add the source files to the project, using    \menuselection{Project \sub Add Existing Item}. Set the pattern to    \code{*.*} and select both \file{spam.c} and \file{spam.def} and    click OK.  (Inserting them one by one is fine too.)     Now open the \menuselection{Project \sub spam properties} dialog.    You only need to change a few settings.  Make sure \guilabel{All    Configurations} is selected from the \guilabel{Settings for:}    dropdown list.  Select the C/\Cpp{} tab.  Choose the General    category in the popup menu at the top.  Type the following text in    the entry box labeled \guilabel{Additional Include Directories}: \begin{verbatim}..\Include,..\PC\end{verbatim}     Then, choose the General category in the Linker tab, and enter \begin{verbatim}..\PCbuild\end{verbatim}     in the text box labelled \guilabel{Additional library Directories}.     Now you need to add some mode-specific settings:     Select \guilabel{Release} in the \guilabel{Configuration}    dropdown list.  Choose the \guilabel{Link} tab, choose the    \guilabel{Input} category, and append \code{pythonXY.lib} to the    list in the \guilabel{Additional Dependencies} box.     Select \guilabel{Debug} in the \guilabel{Configuration} dropdown    list, and append \code{pythonXY_d.lib} to the list in the    \guilabel{Additional Dependencies} box.  Then click the C/\Cpp{}    tab, select \guilabel{Code Generation}, and select    \guilabel{Multi-threaded Debug DLL} from the \guilabel{Runtime    library} dropdown list.     Select \guilabel{Release} again from the \guilabel{Configuration}    dropdown list.  Select \guilabel{Multi-threaded DLL} from the    \guilabel{Runtime library} dropdown list.\end{enumerate}  If your module creates a new type, you may have trouble with this line: \begin{verbatim}    PyObject_HEAD_INIT(&PyType_Type)\end{verbatim} Change it to: \begin{verbatim}    PyObject_HEAD_INIT(NULL)\end{verbatim} and add the following to the module initialization function: \begin{verbatim}    MyObject_Type.ob_type = &PyType_Type;\end{verbatim} Refer to section~3 of the\citetitle[http://www.python.org/doc/FAQ.html]{Python FAQ} for detailson why you must do this.  \section{Differences Between \UNIX{} and Windows     \label{dynamic-linking}}\sectionauthor{Chris Phoenix}{cphoenix@best.com}  \UNIX{} and Windows use completely different paradigms for run-timeloading of code.  Before you try to build a module that can bedynamically loaded, be aware of how your system works. In \UNIX, a shared object (\file{.so}) file contains code to be used by theprogram, and also the names of functions and data that it expects tofind in the program.  When the file is joined to the program, allreferences to those functions and data in the file's code are changedto point to the actual locations in the program where the functionsand data are placed in memory.  This is basically a link operation. In Windows, a dynamic-link library (\file{.dll}) file has no danglingreferences.  Instead, an access to functions or data goes through alookup table.  So the DLL code does not have to be fixed up at runtimeto refer to the program's memory; instead, the code already uses theDLL's lookup table, and the lookup table is modified at runtime topoint to the functions and data. In \UNIX, there is only one type of library file (\file{.a}) whichcontains code from several object files (\file{.o}).  During the linkstep to create a shared object file (\file{.so}), the linker may findthat it doesn't know where an identifier is defined.  The linker willlook for it in the object files in the libraries; if it finds it, itwill include all the code from that object file. In Windows, there are two types of library, a static library and animport library (both called \file{.lib}).  A static library is like a\UNIX{} \file{.a} file; it contains code to be included as necessary.An import library is basically used only to reassure the linker that acertain identifier is legal, and will be present in the program whenthe DLL is loaded.  So the linker uses the information from theimport library to build the lookup table for using identifiers thatare not included in the DLL.  When an application or a DLL is linked,an import library may be generated, which will need to be used for allfuture DLLs that depend on the symbols in the application or DLL. Suppose you are building two dynamic-load modules, B and C, which shouldshare another block of code A.  On \UNIX, you would \emph{not} pass\file{A.a} to the linker for \file{B.so} and \file{C.so}; that wouldcause it to be included twice, so that B and C would each have theirown copy.  In Windows, building \file{A.dll} will also build\file{A.lib}.  You \emph{do} pass \file{A.lib} to the linker for B andC.  \file{A.lib} does not contain code; it just contains informationwhich will be used at runtime to access A's code.   In Windows, using an import library is sort of like using \samp{importspam}; it gives you access to spam's names, but does not create aseparate copy.  On \UNIX, linking with a library is more like\samp{from spam import *}; it does create a separate copy.  \section{Using DLLs in Practice \label{win-dlls}}\sectionauthor{Chris Phoenix}{cphoenix@best.com} Windows Python is built in Microsoft Visual \Cpp; using othercompilers may or may not work (though Borland seems to).  The rest ofthis section is MSV\Cpp{} specific. When creating DLLs in Windows, you must pass \file{pythonXY.lib} tothe linker.  To build two DLLs, spam and ni (which uses C functionsfound in spam), you could use these commands: \begin{verbatim}cl /LD /I/python/include spam.c ../libs/pythonXY.libcl /LD /I/python/include ni.c spam.lib ../libs/pythonXY.lib\end{verbatim} The first command created three files: \file{spam.obj},\file{spam.dll} and \file{spam.lib}.  \file{Spam.dll} does not containany Python functions (such as \cfunction{PyArg_ParseTuple()}), but itdoes know how to find the Python code thanks to \file{pythonXY.lib}. The second command created \file{ni.dll} (and \file{.obj} and\file{.lib}), which knows how to find the necessary functions fromspam, and also from the Python executable. Not every identifier is exported to the lookup table.  If you want anyother modules (including Python) to be able to see your identifiers,you have to say \samp{_declspec(dllexport)}, as in \samp{void_declspec(dllexport) initspam(void)} or \samp{PyObject_declspec(dllexport) *NiGetSpamData(void)}. Developer Studio will throw in a lot of import libraries that you donot really need, adding about 100K to your executable.  To get rid ofthem, use the Project Settings dialog, Link tab, to specify\emph{ignore default libraries}.  Add the correct\file{msvcrt\var{xx}.lib} to the list of libraries.