[[TracNav(TracNav/TOC_Development)]]
[[TOC]]
= Style Guide for Orxonox C++ Code =

'Coding style' refers to the way source code is formatted. For C++, this involves things like brace placement, indentation, and the way parentheses are used. The most important thing is for the code to be consistent within a program or library - code with sloppy formatting is not acceptable, since it is hard to read.

When writing a new program or library, please follow a consistent style of brace placement and indentation. We recommend the following code style.

Remember that any violation to the guide of course is allowed if it enhances readability.

== Files ==
=== General ===
Keep your classes/files short, don't exceed 2000 LOC (if it get's longer you may separate functions into different modules). Put every class in a separate file and name the file like the class name (use !CamelCase names). Create a separate header (ends with .h) and source file (ends with .cc).
Example for class {{{MyExampleClass}}}.
{{{
file names:
MyExampleClass.cc
MyExampleClass.h
}}}

=== Headers ===
Try to minimize header file dependencies by making use of forward declaration and careful dependency studies. Whenever a class member is just a pointer or reference, the forward declaration is sufficient. To make that easier there exists a prerequisites file for each library, containing all the necessary forward declarations. These have to be kept up to date of course.
Example for a header file include sequence.
{{{
#include "MyLibraryPrereqs.h"

#include <string>              // std includes
#include <OgrePrerequisites.h> // most external libraries have forward declaration files as well
#include "OrxonoxClass.h"      // only include this if you inherit from it or you don't use an object as pointer/reference
}}}
Note: Never ever (!!!) write {{{using namespace foo;}}} in header files! (think about what that would mean)

=== Source files ===
Again, keep the dependencies short. However it is not that severe in source files. [[br]]
To ensure that every header file can be compiled without additional header dependencies, include the class header file first.
Include sequence is for {{{MyClass}}}:
{{{
#include "MyClass.h"

#include <vector>              // std headers
#include <map>
#include <OgreSceneManager.h>  // external library headers
#include "BaseObject.h"        // our own header files
}}}

=== Namespaces ===
Use namespaces for separate libraries like network, 'orxonox' otherwise. Names representing namespaces should be all lowercase.[[br]]
Don't write "using namespace ..;" in header files because otherwise if someone included that header file, the "using namespace .." would automatically be included as well. That may lead to unwanted errors. 

=== Inlining and Templates ===
Inline functions and template structures should be declared in a .h header file as described above and implemented below the class declaration if the definition exceeds one line. [[br]]
Using the 'inline' keyword within the class declaration doesn't give advantages (it's inline anyway), so let it be. It is necessary outside however. [[br]]
An Example:
{{{
class FooBar
{
    void funcA()
        { return this->myVar_; }
    void funcB();
};

inline void FooBar::funcB()
{
    doSomething();
    doOtherStuff();
}
}}}


=== Machine-Dependent Code ===
Place machine-dependent code in a special file so that it may be easily located when porting code from one machine to another.

=== Include Guard ===
Header files must contain an include guard.
{{{
#ifndef _ClassName_H__
#define _ClassName_H__


#endif /* _ClassName_H__ */
}}}

== Comments ==

=== Introductory Comment ===
Every file that contains source code must be documented with an introductory comment that provides information on the file name and its contents:
{{{
/**
    @file
    @brief A brief description.

    More description...
*/
}}}
Note: try not to fill in the filename. It is obvious and the comment compiler (Doxygen) will do it automatically and you don't have to adjust it when renaming the file.

== Naming Conventions ==

=== Classes and Types ===
Names representing types must be in mixed case (CamelCase) starting with upper case.
{{{
class SavingsAccount
{
};

struct SimpleStruct
{
};
}}}

=== Variables ===
Variable names must be in mixed case starting with lower case. Each variable declaration on a new line. Avoid abbreviations in names.
{{{
string testString;
int numberOfTimes;
}}}
Memeber variables of classes all end with an underline:
{{{
class TestClass
{
private:
    int numberOfTimes_;
};
}}}
Static variables of classes have to end with '_s':
{{{
class TestClass2
{
protected:
    static float someList_s;
};
}}}
Use useful names:
{{{
// wrong
ProgressBar *prbar;
string prtxt, errstr;
 
// correct
ProgressBar* downloadProgressBar;
string progressText;
string errorString;
}}}
''Comment: Apart from its name and its type, the scope of a variable is its most important feature. Indicating class scope by using underscore makes it easy to distinguish class variables from local scratch variables. This is important because class variables are considered to have higher significance than method variables, and should be treated with special care by the programmer.''

=== Constants ===
Named constants (including enumeration values) must be all uppercase using underscore to separate words. Always use constants instead of numbers and macros. 
{{{
const int MAX_ITERATIONS = 5;

for (int i = 0; i < MAX_ITERATIONS; i++) { } // Instead of (... i < 5; ...)
}}}

=== Functions ===
Names representing methods or functions must be verbs and written in mixed case starting with lower case. Avoid abbreviations in names.
{{{
void calculateTheLowerBoundary() {}
int howManyItemsLeft() {}
}}}
For interface functions to local member variables use short functions (setter and getter functions):
{{{
class WorldEntity
{
public:
    std::string getName()
        { return name_; }

    void setName(const std::string& name)
        { name_ = name; }

private:
    std::string name_;
};
}}}

The name of the object is implicit, and should be avoided in a method name.
{{{
class Line
{
public:
    float getLength(); // Not getLineLength() since it's clear that it is a line.
};
}}}

=== Enumerations ===
Enumeration constants can be prefixed by a common type name.
{{{
enum Color
{ 
    COLOR_RED, 
    COLOR_GREEN, 
    COLOR_BLUE 
};
}}}
However it is better to place an enumeration in a separate namespace. Use 'Enum' as the name of the enum.
{{{
namespace Colour
{
    enum Enum
    {
        Red,
        Green,
        Blue
    };
}
}}}

=== Namespaces ===
Names representing namespaces should be all lowercase and consist of one word.
{{{
namespace util
{
}
}}}


== Indentation ==
=== Spaces ===
Use 4 spaces for an indentation level. Never use tabs, as it is common in windows IDEs.

=== Split Lines ===
The incompleteness of split lines must be made obvious:
{{{
totalSum = a + b + c + 
           d + e; 
function(param1, param2, 
         param3); 

setText("Long line split" 
        "into two parts."); 

for (int tableNo = 0; tableNo < nTables; 
     tableNo += tableStep) { ... }
}}}

== Whitespaces ==
Use one space after each keyword and enclose operators with 2 of them: {{{ if (a + b < c) }}}


=== Include Statements ===
Include statements should be sorted and grouped. Sorted by their hierarchical position in the system with low level files included first. Leave an empty line between groups of include statements.
{{{
#include <fstream> 
#include <iomanip> 

#include <qt/qbutton.h> 
#include <qt/qtextfield.h> 

#include "com/company/ui/PropertiesDialog.h" 
#include "com/company/ui/MainWindow.h"
}}}


=== Constants ===
The use of magic numbers in the code should be avoided. Numbers other than 0 and 1 should be considered declared as named constants instead.


== Statements ==


=== Access Modifiers for Classes ===
The parts of a class must be sorted public, protected and private. All sections must be identified explicitly. Not applicable sections should be left out.


=== Type Casts ===
Type conversions must always be done explicitly. Never rely on implicit type conversion.
{{{
floatValue = static_cast<float>(intValue); // NOT: floatValue = intValue;
// or:
floatValue = float(intValue);
}}}
By this, the programmer indicates that he is aware of the different types involved and that the mix is intentional. [[br]]
Note: Never, ever use static_cast or C-Style cast when dealing with multiple or virtual inheritance!



=== Data Encapsulation ===
Class variables should never be declared public. The concept of C++ information hiding and encapsulation is violated by public variables. Use private variables and access functions instead. One exception to this rule is when the class is essentially a data structure, with no behavior (equivalent to a C struct). In this case it is appropriate to make the class' instance variables public.


=== Loops ===
Only loop control statements must be included in the for() construction.


== Layout ==
=== Indentation ===
Basic indentation should be 4.
{{{
for (i = 0; i < numberOfElements; ++i) 
    a[i] = 0;
}}}


=== Block Layout ===
Block layout should be as illustrated in this example:
{{{
while (!done)
{
    doSomething();
    done = moreToDo();
}
}}}

=== Class Declaration ===
The class declarations should have the following form:
{{{
class SomeClass : public BaseClass 
{ 
    public: 
        void functionBla();

    protected:
        ...

    private:
        ...
};
}}}


=== Function Definitions ===
Method definitions should have the following form:
{{{
void functionBla()
{
    ...
}
}}}

=== If-Statement ===
The if-else class of statements should have the following form:
{{{
if (condition)
{
    ...
}
else
{
  ...
}
}}}
Never use the token below as it does not, what you expect:
{{{
if (condition1)
    if (condition2)
        doSomething();
else                    // This else corresponds to the second if!
    doSomethingElse();
}}}

== Special Implementations ==
=== Use STL Libraries ===
Use STL (standard template library) implementations of common abstract data structures like lists, arrays, stacks, queues, never implement your own!

== General Topics ==
 * different build targets possible (for subprojects)
 * different make options (make-dbg, make-opt)
 * unit test supported (in debug mode)
 * compiler mode: warnings being treated as errors

== Code Reviews ==
Code review is a pretty simple process: at its heart, it is little more than reading some code written by someone else. Nevertheless, it can be useful to have a set of things on which to focus during a review:

    * Does the branch merge or the diff apply cleanly?
    * Are there unit tests for the code being changed or added?
    * Do the unit tests pass for you?
    * Do the unit tests pass for buildbot?
    * Is there documentation for new code?
    * Where appropriate, has existing documentation been updated (including ChangeLog?/NEWS files)?
    * Does the code adhere to the coding standard? 

There's the easy list. Most are mechanical checks. Don't feel bad about rejecting a branch if the answer to any of these questions is no: the problems may seem minor in isolation, but each contributes to overall poor code quality. Moreover, sometimes apparently minor problems can be hiding larger issues.