What is the difference between using #include<filename> and #include<filename.h> in c++

Question

What is the difference between using #include<filename> and #include<filename.h> in c++ which of the two is used and why it is used

Answer 1

Those without the .h is C++ header files while those with .h are C header files. This only applies to the standard header files in C++.

If you are including your own files or files that is not part of standard C++ you need to always write the complete file name (which can be anything).

Answer 2

If you are talking about the standard libraries (because otherwise it wont work at all), the difference is

#include< header >
//my code

is the correct way to call the header according the the C++ standard, while

#include< header.h >
//my code

is deprecated (in the C++ standard, but still necessary according to the C99 standard) and came along with the rest of the C baggage.

So you should use:

#include< cmath >
//my code

not:

#include< math.h >
//my code

Answer 3

The #include <foo.h> was common in C++ code prior to the C++ standard. The standard changed it to #include <foo> with everything from the header placed in the std namespace. (Thanks to litb for pointing out that the standard has never allowed .h headers.)

There is no magic going on, the first looks for a file called 'foo.h' and the second for a file called 'foo'. They are two different files in the file system. The standard just changed the name of the file that should be included.

In most compilers the old headers are still there for backwards compatibility (and compatibility with C), but modern C++ programs that want to follow the standard should not use them.

In the case of standard C headers, the C++ versions have a c at the beginning, so the C header

#include <stdio.h>

becomes

#include <cstdio>

Answer 4

The old standard used the #include <filename.h> syntax. When namespaces and templates were added to the language, the standard was changed to #include <filename>.

This was done so that the standard library stuff could all be placed in the std namespace. Older code, which had no concept of namespaces would still work since the #include <filename.h> files don't use namespaces.

New code should always use the #include <filename> format. If you use the older format, all the symbols they define will be placed in the global namespace rather than std.

Answer 5

C++ only include-files not found in the C standard never used filename.h . Since the very first C++ Standard came out (1998) they have used filename for their own headers.

Files inherited by the C Standard became cfilename instead of filename.h. The C files inherited used like filename.h are deprecated, but still part of the C++ standard.

The difference is that names not defined as macros in C are found within namespace std:: in cfilename in C++, while names in filename.h are within the global namespace scope. So you will find ::size_t in stddef.h, and std::size_t in cstddef. Both are Standard C++, but use of ::size_t is deprecated (See Annex D of the C++ Standard).

Now those were the difference.

Why would you use `filename.h` ?

• Compatibility with C compilers
• Compatibility with very old C++ compilers

Why should you use `cfilename` ?

• Names are within namespace std:: . No name-clashes anymore.
• New C++ features (e.g. overloaded math functions for float, long)
• C Compatibility Headers (filename.h) could disappear in future.