#include question

From: Dan Moos (dan.moos_at_verizon.net)
Date: 02/29/04 

Date: Sun, 29 Feb 2004 19:26:59 GMT

hello group.

Okay, this is a very basic question.

Say I have a program who's source spans , oh say , 5 source files
(translation units as I found them called in Bjarne Stroustrup's book).

Lets say they all use some common header, say <vector>.

So they all #include <vector>

so if I understand it, the code in the header <vector> gets "inserted" into
5 different places in the toatal source code (all 5 translation units)

Is this what really happens, or does the linker somehow account for this and
only #include it once. I always thought that translation units were compiled
as seperate entities, and before linking, have no knowledge of each other,
and so I assume that, yes, they all get that code gets included. It just
seems uneccesary and inefficient to have multiple copies of the same code in
a project. I mean, in the end the linker makes it one big block of code
anyway.

I know about multiple inclusion guards, but doesn't that only protect
against multiple inclusion in the same translation unit? or are they
specifically for what I am asking about.