gtk2/docs/CODING-STYLE

642 lines
16 KiB
Plaintext
Raw Normal View History

GTK+ Coding Style
-------------------------------------------------------------------------------
This document is intended to be a short description of the preferred
coding style to be used for the GTK+ source code. It was strongly
inspired by Clutter's CODING_STYLE.
Coding style is a matter of consistency, readability and maintainance;
coding style is also completely arbitrary and a matter of taste. This
document will use examples at the very least to provide authoritative
and consistent answers to common questions regarding the coding style,
and will also try to identify the allowed exceptions.
The examples will show the preferred coding style; the negative examples
will be clearly identified. Please, don't submit code to GTK+ that
looks like any of these.
Part of the rationales for these coding style rules are available either
in the kernel CodingStyle document or in Cairo's CODING_STYLE one.
When in doubt, check the surrounding code and try to imitate it.
+ Line width
The maximum line width for source files is 80 characters, whenever possible.
Longer lines are usually an indication that you either need a function
or a pre-processor macro.
+ Indentation
Each new level is indented 2 or more spaces than the previous level:
if (condition)
single_statement ();
This can only be achieved using space characters. It may not be achieved
using tab characters alone, or using a combination of spaces and tabs.
Do not change the editor's configuration to change the meaning of a
tab character (see below); code using tabs to indent will not be accepted
into GTK+.
Even if two spaces for each indentation level allows deeper nesting than
8 spaces, GTK+ favours self-documenting function names that can take
quite some space. For this reason you should avoid deeply nested code.
+ Tab characters
The tab character must always be expanded to spaces. If a literal
tab must be used inside the source, the tab must always be interpreted
according to its traditional meaning:
Advance to the next column which is a multiple of 8.
[ these two lines should be aligned ]
+ Braces
Curly braces should not be used for single statement blocks:
if (condition)
single_statement ();
else
another_single_statement (arg1);
In case of multiple statements, curly braces should be put on another
indentation level:
if (condition)
{
statement_1 ();
statement_2 ();
statement_3 ();
}
The "no block for single statements" rule has only four exceptions:
① if the single statement covers multiple lines, e.g. for functions with
many arguments, and it is followed by else or else if:
/* valid */
if (condition)
{
a_single_statement_with_many_arguments (some_lengthy_argument,
another_lengthy_argument,
and_another_one,
plus_one);
}
else
another_single_statement (arg1, arg2);
② if the condition is composed of many lines:
/* valid */
if (condition1 ||
(condition2 && condition3) ||
condition4 ||
(condition5 && (condition6 || condition7)))
{
a_single_statement ();
}
③ Nested if's, in which case the block should be placed on the
outermost if:
/* valid */
if (condition)
{
if (another_condition)
single_statement ();
else
another_single_statement ();
}
/* invalid */
if (condition)
if (another_condition)
single_statement ();
else if (yet_another_condition)
another_single_statement ();
④ If either side of an if-else statement has braces, both sides
should, to match up indentation:
/* valid */
if (condition)
{
foo ();
bar ();
}
else
{
baz ();
}
/* invalid */
if (condition)
{
foo ();
bar ();
}
else
baz ();
In general, new blocks should be placed on a new indentation level,
like:
int retval = 0;
statement_1 ();
statement_2 ();
{
int var1 = 42;
gboolean res = FALSE;
res = statement_3 (var1);
retval = res ? -1 : 1;
}
While curly braces for function definitions should rest on a new line
they should not add an indentation level:
/* valid */
static void
my_function (int argument)
{
do_my_things ();
}
/* invalid */
static void
my_function (int argument) {
do_my_things ();
}
/* invalid */
static void
my_function (int argument)
{
do_my_things ();
}
Curly braces must not be placed on the same line as a condition:
/* invalid */
if (condition) {
statement_1 ();
statement_2 ();
}
+ Conditions
Do not check boolean values for equality:
/* invalid */
if (condition == TRUE)
do_foo ();
/* valid */
if (another_condition)
do_bar ();
Even if C handles NULL equality like a boolean, be explicit:
/* valid */
if (some_pointer == NULL)
do_blah ();
/* invalid */
if (some_other_pointer)
do_blurp ();
In case of conditions split over multiple lines, the logical operators should
always go at the end of the line:
/* invalid */
if (condition1
|| condition2
|| condition3)
{
do_foo ();
}
/* valid */
if (condition1 &&
condition2 &&
(condition3 || (condition4 && condition5)))
{
do_blah ();
}
+ Functions
Functions should be declared by placing the returned value on a separate
line from the function name:
void
my_function (void)
{
}
The arguments list must be broken into a new line for each argument,
with the argument names right aligned, taking into account pointers:
void
my_function (some_type_t type,
another_type_t *a_pointer,
final_type_t another_type)
{
}
The alignment also holds when invoking a function without breaking the
80 characters limit:
align_function_arguments (first_argument,
second_argument,
third_argument);
To respect the 80 characters limit do not break the function name from
the arguments:
/* invalid */
a_very_long_function_name_with_long_parameters
(argument_the_first, argument_the_second);
/* valid */
first_a = argument_the_first;
second_a = argument_the_second;
a_very_long_function_name_with_long_parameters (first_a, second_a);
+ Whitespace
Always put a space before a parenthesis but never after:
/* valid */
if (condition)
do_my_things ();
/* valid */
switch (condition)
{
}
/* invalid */
if(condition)
do_my_things();
/* invalid */
if ( condition )
do_my_things ( );
A switch() should open a block on a new indentation level, and each case
should start on the same indentation level as the curly braces, with the
case block on a new indentation level:
/* valid */
switch (condition)
{
case FOO:
do_foo ();
break;
case BAR:
do_bar ();
break;
}
/* invalid */
switch (condition) {
case FOO: do_foo (); break;
case BAR: do_bar (); break;
}
/* invalid */
switch (condition)
{
case FOO: do_foo ();
break;
case BAR: do_bar ();
break;
}
/* invalid */
switch (condition)
{
case FOO:
do_foo ();
break;
case BAR:
do_bar ();
break;
}
It is preferable, though not mandatory, to separate the various cases with
a newline:
switch (condition)
{
case FOO:
do_foo ();
break;
case BAR:
do_bar ();
break;
default:
do_default ();
}
The 'break' statement for the default: case is not mandatory.
If a case block needs to declare new variables, the same rules as the
inner blocks (see above) apply; the break statement should be placed
outside of the inner block:
switch (condition)
{
case FOO:
{
int foo;
foo = do_foo ();
}
break;
...
}
When declaring a structure type use newlines to separate logical sections
of the structure:
struct _GtkWrapBoxPrivate
{
GtkOrientation orientation;
GtkWrapAllocationMode mode;
GtkWrapBoxSpreading horizontal_spreading;
GtkWrapBoxSpreading vertical_spreading;
guint16 vertical_spacing;
guint16 horizontal_spacing;
guint16 minimum_line_children;
guint16 natural_line_children;
GList *children;
};
Do not eliminate whitespace and newlines just because something would
fit on 80 characters:
/* invalid */
if (condition) foo (); else bar ();
Do eliminate trailing whitespace on any line, preferably as a separate
patch or commit. Never use empty lines at the beginning or at the end of
a file.
Do enable the default git pre-commit hook that detect trailing
whitespace for you and help you to avoid corrupting GTK+'s tree with
it. Do that as follows:
chmod a+x .git/hooks/pre-commit
You might also find the git-stripspace utility helpful which acts as a
filter to remove trailing whitespace as well as initial, final, and
duplicate blank lines.
+ Headers
Headers are special, for GTK+, in that they don't have to obey the
80 characters limit. The only major rule for headers is that the function
definitions should be vertically aligned in three columns:
return value function_name (type argument,
type argument,
type argument);
The maximum width of each column is given by the longest element in the
column:
2011-06-06 18:07:23 +00:00
void gtk_type_set_property (GtkType *type,
const gchar *value,
GError **error);
const gchar *gtk_type_get_property (GtkType *type);
It is also possible to align the columns to the next tab:
void gtk_type_set_prop (GtkType *type,
gfloat value);
gfloat gtk_type_get_prop (GtkType *type);
gint gtk_type_update_foobar (GtkType *type);
Public headers should never be included directly:
#if !defined (__GTK_H_INSIDE__) && !defined (GTK_COMPILATION)
#error "Only <gtk/gtk.h> can be included directly."
#endif
All headers should have inclusion guards (for internal usage)
and C++ guards:
#ifndef __GTK_FOO_H__
#define __GTK_FOO_H__
#include <gtk/gtk-bar.h>
G_BEGIN_DECLS
...
G_END_DECLS
#endif /* __GTK_FOO_H__ */
+ Includes
GTK+ source files should never include the global gtk.h header, but
instead include the individual headers that are needed. Every file must
include config.h first, then its own header, then other GTK+ headers
that it needs, then system and third-party headers that it needs.
/* valid */
#include "config.h"
#include "gtkfoo.h"
#include "gtkwidget.h"
#include "gtkbutton.h"
...
#include <string.h>
+ GObject
GObject classes definition and implementation require some additional
coding style notices.
Typedef declarations should be placed at the beginning of the file:
typedef struct _GtkFoo GtkFoo;
typedef struct _GtkFooClass GtkFooClass;
typedef struct _GtkFooPrivate GtkFooPrivate;
This includes enumeration types:
typedef enum
{
GTK_SIZE_REQUEST_WIDTH_FOR_HEIGHT,
GTK_SIZE_REQUEST_HEIGHT_FOR_WIDTH
} GtkSizeRequestMode;
And callback types:
typedef void (* GtkCallback) (GtkWidget *widget,
gpointer user_data);
Instance structures should only contain the parent type, and optionally a
pointer to a private data structure, and they should be annotated as
"private" using the gtk-doc trigraph:
struct _GtkFoo
{
/*< private >*/
GtkWidget parent_instance;
GtkFooPrivate *priv;
};
The private data pointer is optional and should be omitted in newly
written classes.
Always use the G_DEFINE_TYPE(), G_DEFINE_TYPE_WITH_PRIVATE(), and
G_DEFINE_TYPE_WITH_CODE() macros, or their abstract variants
G_DEFINE_ABSTRACT_TYPE(), G_DEFINE_ABSTRACT_TYPE_WITH_PRIVATE(), and
G_DEFINE_ABSTRACT_TYPE_WITH_CODE(); also, use the similar macros for
defining interfaces and boxed types.
All the properties should be stored inside the private data structure, which
is defined inside the source file - or, if needed, inside a private header
file; the private header filename must end with "private.h" and must not be
installed.
The private data structure should only be accessed internally either using the
pointer inside the instance structure, if one is available, or the generated
instance private data getter function for your type. You should never use the
G_TYPE_INSTANCE_GET_PRIVATE() macro or the g_type_instance_get_private()
function.
Interface types should always have the dummy typedef for cast purposes:
typedef struct _GtkFoo GtkFoo;
The interface structure should have "Interface" postfixed to the dummy typedef:
typedef struct _GtkFooInterface GtkFooInterface;
Interfaces must have the following macros:
- Macro: - Expands to:
• GTK_TYPE_<iface_name> <iface_name>_get_type
• GTK_<iface_name> G_TYPE_CHECK_INSTANCE_CAST
• GTK_IS_<iface_name> G_TYPE_CHECK_INSTANCE_TYPE
• GTK_<iface_name>_GET_IFACE G_TYPE_INSTANCE_GET_INTERFACE
+ Memory allocation
When dynamically allocating data on the heap either use g_new() or,
if allocating multiple small data structures, g_slice_new().
Public structure types should always be returned after being zero-ed,
either explicitly for each member, or by using g_new0() or g_slice_new0().
+ Macros
Try to avoid private macros unless strictly necessary. Remember to #undef
them at the end of a block or a series of functions needing them.
Inline functions are usually preferable to private macros.
Public macros should not be used unless they evaluate to a constant.
+ Public API
Avoid exporting variables as public API, since this is cumbersome on some
platforms. It is always preferable to add getters and setters instead.
All public functions must be listed in the gtk.symbols file.
+ Private API
Non-exported functions that are needed in more than one source file
should be named "_gtk_...", and declared in a private header file.
Underscore-prefixed functions are never exported.
Non-exported functions that are only needed in one source file
should be declared static.
+ Documentation
All public APIs must have gtk-doc comments. For functions, these should
be placed in the source file, directly above the function.
/* valid */
/**
* gtk_get_flow:
* @widget: a #GtkWidget
*
* Gets the flow of a widget.
*
* Note that flows may be laminar or turbulent...
*
* Returns: (transfer none): the flow of @widget
*/
GtkFlow *
gtk_get_flow (GtkWidget *widget)
{
...
}
Doc comments for macros, function types, class structs, etc should be
placed next to the definitions, typically in headers.
Section introductions should be placed in the source file they describe,
after the license header:
/* valid */
/**
* SECTION:gtksizerequest
* @Short_Description: Height-for-width geometry management
* @Title: GtkSizeRequest
*
* The GtkSizeRequest interface is GTK+'s height-for-width (and
* width-for-height) geometry management system.
* ...
*/
To properly document a new function, macro, function type or struct,
it needs to be listed in the gtk3-sections.txt file.
To properly document a new class, it needs to be given its own section
in gtk3-sections.txt, needs to be included in gtk-docs.sgml, and the
get_type function needs to listed in gtk3.types.
+ Old code
New code that is being added to GTK+ should adhere to the style
explained above. Existing GTK+ code does largely follow these
conventions, but there are some differences, e.g. occurrences
of tabs, etc.
It is ok to update the style of a code block or function when you
are touching it anyway, but sweeping whitespace changes obscure the
git history and should be avoided.