In computing, source code is any collection of computer instructions, possibly with comments, written using a human-readable programming language, usually as ordinary text. The source code of a program is specially designed to facilitate the work of computer programmers, who specify the actions to be performed by a computer mostly by writing source code. The source code is often transformed by an assembler or compiler into binary machine code understood by the computer. The machine code might then be stored for execution at a later time. Alternatively, source code may be interpreted and thus immediately executed.
Most application software is distributed in a form that includes only executable files. If the source code were included it would be useful to a user, programmer or a system administrator, any of whom might wish to study or modify the program.
The Linux Information Project defines source code as:
Source code (also referred to as source or code) is the version of software as it is originally written (i.e., typed into a computer) by a human in plain text (i.e., human readable alphanumeric characters).
The notion of source code may also be taken more broadly, to include machine code and notations in graphical languages, neither of which are textual in nature. An example from an article presented on the annual IEEE conference and on Source Code Analysis and Manipulation:
For the purpose of clarity ‘source code’ is taken to mean any fully executable description of a software system. It is therefore so construed as to include machine code, very high level languages and executable graphical representations of systems.
Often there are several steps of program translation or minification between the original source code typed by a human and an executable program. While some, like the FSF, argue that an intermediate file "is not real source code and does not count as source code", others find it convenient to refer to each intermediate file as the source code for the next steps.
The earliest programs for stored-program computers were entered in binary through the front panel switches of the computer. This first-generation programming language had no distinction between source code and machine code.
When IBM first offered software to work with its machine, the source code was provided at no additional charge. At that time, the cost of developing and supporting software was included in the price of the hardware. For decades, IBM distributed source code with its software product licenses, until 1983.
Most early computer magazines published source code as type-in programs.
Occasionally the entire source code to a large program is published as a hardback book, such as "Computers & Typesetting, Volume B: TeX: The Program", "PGP Source Code and Internals", "PC SpeedScript", and "uC/OS: The Real-Time Kernel".
The source code which constitutes a program is usually held in one or more text files stored on a computer's hard disk; usually these files are carefully arranged into a directory tree, known as a source tree. Source code can also be stored in a database (as is common for stored procedures) or elsewhere.
The source code for a particular piece of software may be contained in a single file or many files. Though the practice is uncommon, a program's source code can be written in different programming languages. For example, a program written primarily in the C programming language, might have portions written in assembly language for optimization purposes. It is also possible for some components of a piece of software to be written and compiled separately, in an arbitrary programming language, and later integrated into the software using a technique called library linking. In some languages, such as Java, this can be done at run time (each class is compiled into a separate file that is linked by the interpreter at runtime).
Yet another method is to make the main program an interpreter for a programming language, either designed specifically for the application in question or general-purpose, and then write the bulk of the actual user functionality as macros or other forms of add-ins in this language, an approach taken for example by the GNU Emacs text editor.
The code base of a computer programming project is the larger collection of all the source code of all the computer programs which make up the project. It has become common practice to maintain code bases in version control systems. Moderately complex software customarily requires the compilation or assembly of several, sometimes dozens or even hundreds, of different source code files. In these cases, instructions for compilations, such as a Makefile, are included with the source code. These describe the programming relationships among the source code files, and contain information about how they are to be compiled.
The revision control system is another tool frequently used by developers for source code maintenance.
Source code is primarily used as input to the process that produces an executable program (i.e., it is compiled or interpreted). It is also used as a method of communicating algorithms between people (e.g., code snippets in books).
Computer programmers often find it helpful to review existing source code to learn about programming techniques. The sharing of source code between developers is frequently cited as a contributing factor to the maturation of their programming skills. Some people consider source code an expressive artistic medium.
Porting software to other computer platforms is usually prohibitively difficult without source code. Without the source code for a particular piece of software, portability is generally computationally expensive. Possible porting options include binary translation and emulation of the original platform.
Decompilation of an executable program can be used to generate source code, either in assembly code or in a high-level language.
Programmers frequently adapt source code from one piece of software to use in other projects, a concept known as software reusability.
Before 1974, when the US Commission on New Technological Uses of Copyrighted Works (CONTU) decided that "computer programs, to the extent that they embody an author's original creation, are proper subject matter of copyright", software and its source code was not copyrightable and therefore always public domain software. The 1983 Apple v. Franklin legislation clarified also for object code that the Copyright Act gave computer programs the copyright status of literary works.
In 1999, in the United States court case Bernstein v. United States it was decided if source code should be considered a constitutionally protected form of free speech. Proponents of free speech argued that because source code conveys information to programmers, is written in a language, and can be used to share humor and other artistic pursuits, it is a protected form of communication. University of California mathematics professor and cryptology specialist Dan Bernstein brought the case to court as at the time, encryption algorithms were classified as munitions by the United States government. Exporting encryption to other countries was considered an issue of national security, and had to be approved by the State Department. On Bernstein's behalf the Electronic Frontier Foundation sued the U.S. government. The court ruling in the case declared software as protected speech under the First Amendment, and national restrictions on encryption software were overturned.
An author of a non-trivial work like software, has several exclusive rights, among them the copyright for the source code and object code. The author has the right and possibility to grant customers and users of his software some of his exclusive rights in form of software licensing. Software, and its accompanying source code, can be associated with several licensing paradigms; the most important distinction is open source vs proprietary software.
Generally speaking, software is open source if the source code is free to use, distribute, modify and study, and proprietary if the source code is kept secret, or is privately owned and restricted. One of the first software license to be published and to explicitly grant these freedoms was the GNU General Public License in 1989, the BSD license is another early example from 1990.
For proprietary software, the provisions of the various copyright laws, trade secrecy and patents are used to keep the source code closed. Additionally, many pieces of retail software come with an end-user license agreement (EULA) which typically prohibits decompilation, reverse engineering, analysis, modification, or circumventing of copy protection. Types of source code protection – beyond traditional compilation to object code – include code encryption, code obfuscation or code morphing.
The way a program is written can have important consequences for its maintainers. Coding conventions, which stress readability and some language-specific conventions, are aimed at the maintenance of the software source code, which involves debugging and updating. Other priorities, such as the speed of the program's execution, or the ability to compile the program for multiple architectures, often make code readability a less important consideration, since code quality generally depends on its purpose.