www.digitalmars.com

D Programming Language 2.0


Last update Mon Dec 26 20:17:03 2011

Comparison

To D, or not to D. -- Willeam NerdSpeare

This table is a quick and rough list of various features of that can be used to compare with other languages. While many capabilities are available with standard libraries, this table is for features built in to the core language itself. Rationale.

D Language Feature Comparison Table
Feature
Garbage Collection Yes
Functions
Function delegates Yes
Function overloading Yes
Out function parameters Yes
Nested functions Yes
Function literals Yes
Closures Yes
Typesafe variadic arguments Yes
Lazy function argument evaluation Yes
Compile time function evaluation Yes
Arrays
Lightweight arrays Yes
Resizeable arrays Yes
Built-in strings Yes
Array slicing Yes
Array bounds checking Yes
Array literals Yes
Associative arrays Yes
Strong typedefs Yes
String switches Yes
Aliases Yes
OOP
Object Oriented Yes
Multiple Inheritance No
Interfaces Yes
Operator overloading Yes
Modules Yes
Dynamic class loading No
Nested classes Yes
Inner (adaptor) classes Yes
Covariant return types Yes
Properties Yes
Performance
Inline assembler Yes
Direct access to hardware Yes
Lightweight objects Yes
Explicit memory allocation control Yes
Independent of VM Yes
Direct native code gen Yes
Generic Programming
Class Templates Yes
Function Templates Yes
Implicit Function Template Instantiation Yes
Partial and Explicit Specialization Yes
Value Template Parameters Yes
Template Template Parameters Yes
Variadic Template Parameters Yes
Template Constraints Yes
Mixins Yes
static if Yes
is expressions Yes
typeof Yes
foreach Yes
Implicit Type Inference Yes
Reliability
Contract Programming Yes
Unit testing Yes
Static construction order Yes
Guaranteed initialization Yes
RAII (automatic destructors) Yes
Exception handling Yes
Scope guards Yes
try-catch-finally blocks Yes
Thread synchronization primitives Yes
Compatibility
C-style syntax Yes
Enumerated types Yes
Support all C types Yes
80 bit floating point Yes
Complex and Imaginary Yes
Direct access to C Yes
Use existing debuggers Yes
Struct member alignment control Yes
Generates standard object files Yes
Macro text preprocessor No
Other
Conditional compilation Yes
Unicode source text Yes
Documentation comments Yes

Notes

Object Oriented
This means support for classes, member functions, inheritance, and virtual function dispatch.

Inline assembler
Many C and C++ compilers support an inline assembler, but this is not a standard part of the language, and implementations vary widely in syntax and quality.

Interfaces
Support in C++ for interfaces is weak enough that an IDL (Interface Description Language) was invented to compensate.

Modules
Many correctly argue that C++ doesn't really have modules. But C++ namespaces coupled with header files share many features with modules.

Garbage Collection
The Hans-Boehm garbage collector can be successfully used with C and C++, but it is not a standard part of the language.

Implicit Type Inference
This refers to the ability to pick up the type of a declaration from its initializer.

Contract Programming
The Digital Mars C++ compiler supports Contract Programming as an extension. Compare some C++ techniques for doing Contract Programming with D.

Resizeable arrays
Part of the standard library for C++ implements resizeable arrays, however, they are not part of the core language. A conforming freestanding implementation of C++ (C++98 17.4.1.3) does not need to provide these libraries.

Built-in Strings
Part of the standard library for C++ implements strings, however, they are not part of the core language. A conforming freestanding implementation of C++ (C++98 17.4.1.3) does not need to provide these libraries. Here's a comparison of C++ strings and D built-in strings.

Strong typedefs
Strong typedefs can be emulated in C/C++ by wrapping a type in a struct. Getting this to work right requires much tedious programming, and so is considered as not supported.

Use existing debuggers
By this is meant using common debuggers that can operate using debug data in common formats embedded in the executable. A specialized debugger useful only with that language is not required.

Struct member alignment control
Although many C/C++ compilers contain pragmas to specify struct alignment, these are nonstandard and incompatible from compiler to compiler.
The C# standard ECMA-334 25.5.8 says only this about struct member alignment: "The order in which members are packed into a struct is unspecified. For alignment purposes, there may be unnamed padding at the beginning of a struct, within a struct, and at the end of the struct. The contents of the bits used as padding are indeterminate." Therefore, although Microsoft may have extensions to support specific member alignment, they are not an official part of standard C#.

Support all C types
C99 adds many new types not supported by C++.

80 bit floating point
While the standards for C and C++ specify long doubles, few compilers (besides Digital Mars C/C++) actually implement 80 bit (or longer) floating point types.

Mixins
Mixins have many different meanings in different programming languages. D mixins mean taking an arbitrary sequence of declarations and inserting (mixing) them into the current scope. Mixins can be done at the global, class, struct, or local level.

C++ Mixins
C++ mixins refer to a couple different techniques. The first is analogous to D's interface classes. The second is to create a template of the form:
template <class Base> class Mixin : public Base
{
    ... mixin body ...
}
D mixins are different.

Static If
The C and C++ preprocessor directive #if would appear to be equivalent to the D static if. But there are major and crucial differences - the #if does not have access to any of the constants, types, or symbols of the program. It can only access preprocessor macros. See this example.

Is Expressions
Is expressions enable conditional compilation based on the characteristics of a type. This is done after a fashion in C++ using template parameter pattern matching. See this example for a comparison of the different approaches.

Comparison with Ada
James S. Rogers has written a comparison chart with Ada.

Inner (adaptor) classes
A nested class is one whose definition is within the scope of another class. An inner class is a nested class that can also reference the members and fields of the lexically enclosing class; one can think of it as if it contained a 'this' pointer to the enclosing class.

Documentation comments
Documentation comments refer to a standardized way to produce documentation from the source code file using specialized comments.

Errors

If I've made any errors in this table, please contact me so I can correct them.





Forums | Comments |  D  | Search | Downloads | Home