After comparing and trying a number of lightweight web servers (Apache was not an option for this kind of small app) I picked SHTTPD, mainly because:

1. It was one of the few I managed to embed in my C++ code (I’m not a C expert…)
2. It offered the possibility of registering call back methods when a particular URL is called, which makes it a quite good candidate for a REST interface.

Here is how I did proceed to integrate it to my Microsoft Visual C++ project:

1. Compile the project (the core, not the example), this should create a shttpd.lib file. (This post might in case you do not succeed this step ).
2. Copy the shttpd.lib, shttpd.h to your src folder (the one of your Visual Studio project).
3. Download shttpd.pem and copy it to your src folder as well.
4. Add the shttpd.h file to your project.
5. Add the following lib references to your project: shttpd.lib ws2_32.lib, see to get details on how to add it.
6. Add the code to start and setup the server. Snippets can be found in the example folder of the SHTTPD distrib. That in my case:


// TestSHTTPD.cpp : Defines the entry point for the console application.
#include
#include
#include
#include
#include

#include “stdafx.h”
#include “shttpd.h”

#define ALIAS_URI “/my_c”
#define ALIAS_DIR “c:\\”

static void show_index(struct shttpd_arg *arg) {
shttpd_printf(arg, “%s”,
“HTTP/1.1 200 OK\r\nContent-Type: text/html\r\n\r\n”
“Welcome to embedded example of SHTTPD”);
arg->flags |= SHTTPD_END_OF_OUTPUT;
}

int _tmain(int argc, char* argv[])
{
/*
* Initialize SHTTPD context.
* Attach folder c:\ to the URL /my_c (for windows), and
* /etc/ to URL /my_etc (for UNIX). These are Apache-like aliases.
* Set WWW root to current directory.
* Start listening on ports 8080 and 8081
*/
int data = 1234567;
struct shttpd_ctx *ctx;
ctx = shttpd_init(argc, argv);
shttpd_set_option(ctx, “ssl_cert”, “shttpd.pem”);
shttpd_set_option(ctx, “aliases”, ALIAS_URI “=” ALIAS_DIR);
shttpd_set_option(ctx, “ports”, “8080,8081s”);

/* Register an index page under two URIs */
shttpd_register_uri(ctx, “/”, &show_index, (void *) &data);
shttpd_register_uri(ctx, “/abc.html”, &show_index, (void *) &data);

/* Serve connections infinitely until someone kills us */
for (;;)
shttpd_poll(ctx, 1000);

/* Probably unreached, because we will be killed by a signal */
shttpd_fini(ctx);

return 0;
}

You should now be able to use the web server within your application.

To use the dll of Visual Studio (Express 2008) project A in project B we do what microsoft calls an implicit linking (http://msdn.microsoft.com/en-us/library/d14wsce5(VS.80).aspx). In more concrete terms here are the steps to follow:

1) Generate the dll of project A. When you do this, the linker (after-compiler thingy) creates both a dll and a lib file.
2) Copy the dll and the lib file to project B, namely to a place accessible by the compiler and the linker, a place that’s on the classpath to use the Java terminology.
3) Right-click on the project B in Visual Studio and select Configuration Properties -> Linker -> Input -> Additional Dependencies
4) There you can select the .lib counterpart of the dll (e.g. swill.lib in my case).

Add additional dependencies

5) Add the header file (.h) of the dll to your project.

That’s it you should now be able to use the dll of A from project B, at least it worked fine for me.
Btw, being mainly a Java developer, I’m rather new to C++ so if that’s a silly way of doing it please do comment the post.