Questions tagged [header-files]

Header files are used in some programming languages to hold source code in a single container, the header file, that may be needed in multiple source files during the processing of the source files. The goal of a header file is to reduce duplication of the same source lines needed in multiple files. Instead of cloning the source lines into each of the files where they are needed, the source lines are put into a header file. The header file is then pulled into a source file by the compiler or script engine during processing of a source file.

The result is that only a single copy of the source lines needs to be maintained and the single copy of source can be reused in multiple files.

The primary purposes of header files is to make it easier to reuse functionality and to reduce errors caused by duplicated source code which becomes slightly different due to maintenance changes.

Typically a special directive such as include or require is used to insert a copy of the contents of the header file at a particular line of a source file during source processing by compiler or script engine. The concept is the include or require directive is replaced by the contents of the header file. Header files are used in a number of programming languages such as C and C++ as well as PHP and Ruby and Lua. The semantics of these directives will vary from programming language to programming language.

The import directive used in languages such as Java and Go is different from the include directive of C and C++ being more of a way to establish linkages and references between classes and packages as a part of combining software components (Java and Go) rather than the including of lines of source code (C and C++ and PHP).

Source lines typically put into header files are definitions and declarations that are needed in multiple files. Information may include class definitions or declarations, function prototypes or structure definitions/declarations or constants of various kinds.

Considerations for header files

When determining what source to put into a header file the primary rule of thumb is they should be lines of source that can be included in multiple files without causing a problem. For instance with the C programming language, a definition of a struct would be appropriate in a header file.

However putting the definition of a standard (not in-line) function into a C header file would not be appropriate. The problem with putting a standard C function definition into a header file is the function will be duplicated at each point where the header file is pulled into a source file. The result would be multiple definitions of the same function causing a linker error.

Some programming languages require care be taken that a header file is not included more than once. This can happen when several header files that are included into a source file themselves include still another header file. In the C and C++ programming languages macros or pragmas are used to enforce a once only rule. In other cases the source lines in the header file may not require a once only rule due to the nature of the source in the header file.

Care must also be taken that a dependency cycle is not created causing an infinite loop during the processing of header files. A header file may include other header files which end up including the original header file again causing a cycle or infinite loop until the compiler runs out of resources during header file processing.

PHP has the require_once directive which will perform a check if the header file has already been included or not.

In C and C++ an external utility, the Preprocessor, handles macro processing and other directives such as the include directive for header files generating a temporary file that is then compiled. In other languages the include directive processing is built into the compiler or script engine.

Typically a file extension of .h is used for header files. However for many programming languages this naming convention is not a requirement for correct processing.