Programming Perl, 3rd Edition

Embedding Perl (Using Perl from C)

You can access a Perl interpreter from C by embedding Perl inside your C program. Since Perl is itself a C program, embedding consists of taking the important chunks of Perl and integrating them into yours.

Note that embedding isn't necessary if your only goal is to use a standalone Perl program and you don't mind launching a separate process to do so. You can use a function like C's popen (3) to exchange data between your C program and any external Perl program, just like you can use Perl's open(PIPE, "| program") or the IPC::Open2 and IPC::Open3 modules to exchange data between your Perl program and any other program. But if you want to avoid the overhead of launching a separate process, you can embed an interpreter into your C program.

When developing long-running applications (say, for embedding in a web server), it's a good idea to maintain a single persistent interpreter rather than creating and destroying interpreters over and over again. The major reason is speed, since Perl will only be loaded into memory once. By using a persistent Perl interpreter, Apache's mod_perl module avoids loading Perl into memory anew every time someone hits an Apache web page. The perlembed manpage provides an example of a persistent interpreter, as well as an example of how a Perl program can manage multiple simultaneous interpreters (another big plus for web servers).

Compiling Embedded Programs

When you embed Perl in C, your C program will usually allocate, ...