Namespace (computer science)

A namespace (sometimes also called a name scope) is an abstract container or environment created to hold a logical grouping of unique identifiers or symbols (i.e., names). An identifier defined in a namespace is associated only with that namespace. The same identifier can be independently defined in multiple namespaces. That is, the meaning associated with an identifier defined in one namespace may or may not have the same meaning as the same identifier defined in another namespace. Languages that support namespaces specify the rules that determine to which namespace an identifier (not its definition) belongs.^[1]

For example, Bill works for company X and his employee ID is 123. Jane works for company Y and her employee ID is also 123. The reason Bill and Jane can be identified by the same ID number is because they work for different companies. The different companies in this case would symbolize different namespaces. There would be serious confusion if the two people worked for the same company, and still had the same employee ID. For instance, a paycheck issued to employee ID 123 would not identify which person should receive the check.

In large computer programs or documents it is not uncommon to have hundreds or thousands of identifiers. Namespaces (or a similar technique, see Emulating namespaces) provide a mechanism for hiding local identifiers. They provide a means of grouping logically related identifiers into corresponding namespaces, thereby making the system more modular.

Data storage devices and many modern programming languages support namespaces. Storage devices use directories (or folders) as namespaces. This allows two files with the same name to be stored on the device so long as they are stored in different directories. In some programming languages (eg. C++, Python), the identifiers naming namespaces are themselves associated with an enclosing namespace. Thus, in these languages namespaces can nest, forming a namespace tree. At the root of this tree is the unnamed global namespace.

Use in common languages

C++

In C++, a namespace is defined with a namespace block.^[2]

namespace abc {
 int bar;
}

Within this block, identifiers can be used exactly as they are declared. Outside of this block, the namespace specifier must be prefixed. For example, outside of namespace abc, bar must be written abc::bar to be accessed. C++ includes another construct that makes this verbosity unnecessary. By adding the line

using namespace abc;

to a piece of code, the prefix abc:: is no longer needed.

Code that is not explicitly declared within a namespace is considered to be in the global namespace.

Namespace resolution in C++ is hierarchical. This means that within the hypothetical namespace food::soup, the identifier chicken refers to food::soup::chicken. If food::soup::chicken doesn't exist, it then refers to food::chicken. If neither food::soup::chicken nor food::chicken exist, chicken refers to ::chicken, an identifier in the global namespace.

Namespaces in C++ are most often used to avoid naming collisions. Although namespaces are used extensively in recent C++ code, most older code does not use this facility. For example, the entire C++ standard library is defined within namespace std, but before standardization many components were originally in the global namespace.

Java

In Java, the idea of a namespace is embodied in Java packages. All code belongs to a package, although that package need not be explicitly named. Code from other packages is accessed by prefixing the package name before the appropriate identifier, for example class String in package java.lang can be referred to as java.lang.String (this is known as the fully qualified class name). Like C++, Java offers a construct that makes it unnecessary to type the package name (import). However, certain features (such as reflection) require the programmer to use the fully qualified name.

Unlike C++, namespaces in Java are not hierarchical as far as the syntax of the language is concerned. However, packages are named in a hierarchical manner. For example, all packages beginning with java are a part of the Java platform—the package java.lang contains classes core to the language, and java.lang.reflect contains core classes specifically relating to reflection.

In Java (and Ada, C#, and others), namespaces/packages express semantic categories of code. For example, in C#, namespace System contains code provided by the system (the .NET Framework). How specific these categories are and how deep the hierarchies go differ from language to language.

Function and class scopes can be viewed as implicit namespaces that are inextricably linked with visibility, accessibility, and object lifetime.

Python

In Python, namespaces are defined by the individual modules, and since modules can be contained in hierarchical packages, then name spaces are hierarchical too.^[3][4] In general when a module is imported then the names defined in the module are defined via that module's name space, and are accessed in from the calling modules by using the fully qualified name.

# assume modulea defines two functions : func1() and func2() and one class : class1
import modulea
 
modulea.func1()
modulea.func2()
a = modulea.class1()

The "from ... import ..." can be used to insert the relevant names directly into the calling module's namespace, and those names can be accessed from the calling module without the qualified name :

# assume modulea defines two functions : func1() and func2() and one class : class1
from modulea import func1
 
func1()
func2() # this will fail as an undefined name, as will the full name modulea.func2()
a = class1() # this will fail as an undefined name, as will the full name modulea.class1()

Since this directly imports names (without qualification) it can overwrite existing names with no warnings.

A special form is "from ... import *", which imports all names defined in the named package directly in the calling modules namespace. Use of this form of import, although supported within the language, is generally discouraged as it pollutes the namespace of the calling module and will cause already defined names to be overwritten in the case of name clashes.

Python also supports "from ... import ... as ...." as a way of providing an alias or alternative name for use by the calling module.

XML namespace

In XML, the XML namespace specification enables the names of elements and attributes in an XML document to be unique, similar to the role of namespaces in programming languages. Using XML namespaces, XML documents may contain element or attribute names from more than one XML vocabulary.

PHP

Namespaces were introduced into PHP from version 5.3 onwards. Naming collision of classes, functions and variables can be avoided. In PHP, a namespace is defined with a namespace block.

# assume this is a class file defines two functions : foo() and bar()
# location of the file phpstar/foobar.php
namespace phpstar;
class fooBar
{
    public function foo()
    {
        echo 'hello world, from function foo';
    }
 
    public function bar()
    {
        echo 'hello world, from function bar';
    }
}

We can reference a PHP namespace with following different ways:

# location of the file index.php
# Include the file
include "phpstar/foobar.php";
 
# Option 1: directly prefix the class name with the namespace
$obj_foobar = new \phpstar\fooBar();
 
# Option 2: import the namespace
use phpstar\fooBar;
$obj_foobar = new fooBar();
 
# Option 2a: import & alise the namespace
use phpstar\fooBar as FB;
$obj_foobar = new FB();
 
# Access the properties and methods with regular way
$obj_foobar->foo();
$obj_foobar->bar();

Emulating namespaces

In programming languages lacking language support for namespaces, namespaces can be emulated to some extent by using an identifier naming convention. For example, C libraries e.g. Libpng often use a fixed prefix for all functions and variables that are part of their exposed interface. Libpng exposes identifiers such as:

png_create_write_struct
png_get_signature
png_read_row
png_set_invalid

This provides reasonable assurance that the identifiers are unique and can therefore be used in larger programs without fear of identifier naming collisions. Likewise, many packages originally written in Fortran (e.g., BLAS, LAPACK) reserve the first few letters of a function's name to indicate which group it belongs to.

Unfortunately, this technique has several drawbacks:

• It doesn't scale well to nested namespaces; identifiers become excessively long.
• Individuals or organizations may use dramatically inconsistent naming conventions, potentially introducing unwanted obfuscation.
• Compound or 'query-based' operations on groups of identifiers, based on the namespaces in which they are declared, is rendered unwieldy or unfeasible.
• All uses of the identifiers must, in effect, be fully namespace-qualified. Languages with direct support for namespaces usually provide ways for the programmer to declare up front that they wish to use some or all identifiers from a specific namespace, which they can then use without qualification for the remainder of the block.
• In languages in which identifier length is restricted, the use of prefixes limits the number of characters that can be used to identify what the function does. This is a particular problem for packages originally written in FORTRAN 77 that offered only 6 characters per identifier. For example, the BLAS function DGEMM function indicates that it operates on double precision numbers (the 'D'), general matrices (the 'GE') and only the last two characters show what it actually does: a matrix-matrix multiplication (the 'MM').

See also

• Identity column in a Table (database)
• Identity (object-oriented programming)
• Java package
• Name resolution
• Namespace (general concept)
• Plan 9, extends everything is a file metaphor beyond files to names
• Primary key in a database table
• Reiser4, naming systems should reflect rather than mold structure
• Scope (programming)

References