fix typo in last commit

[wxWidgets.git] / docs / doxygen / overviews / eventhandling.h

/////////////////////////////////////////////////////////////////////////////\r
// Name:        eventhandling.h\r
// Purpose:     topic overview\r
// Author:      wxWidgets team\r
// RCS-ID:      $Id$\r
// Licence:     wxWindows license\r
/////////////////////////////////////////////////////////////////////////////\r
\r
/**\r
    @page overview_events Events and Event Handling\r
\r
    Related classes: wxEvtHandler, wxWindow, wxEvent\r
\r
@li @ref overview_events_introduction\r
@li @ref overview_events_eventhandling\r
@li @ref overview_events_processing\r
@li @ref overview_events_custom\r
@li @ref overview_events_misc\r
\r
\r
<hr>\r
\r
\r
@section overview_events_introduction Introduction to Events\r
\r
Like with all the other GUI frameworks, the control of flow in wxWidgets\r
applications is event-based: the program normally performs most of its actions\r
in response to the events generated by the user. These events can be triggered\r
by using the input devices (such as keyboard, mouse, joystick) directly or,\r
more commonly, by a standard control which synthesizes such input events into\r
higher level events: for example, a wxButton can generate a click event when\r
the user presses the left mouse button on it and then releases it without\r
pressing @c Esc in the meanwhile. There are also events which don't directly\r
correspond to the user actions, such as wxTimerEvent or wxSocketEvent.\r
\r
But in all cases wxWidgets represents these events in a uniform way and allows\r
you to handle them in the same way wherever they originate from. And while the\r
events are normally generated by wxWidgets itself, you can also do this, which\r
is especially useful when using custom events (see @ref overview_events_custom).\r
\r
To be more precise, each event is described by:\r
 - <em>Event type</em>: this is simply a value of type wxEventType which\r
 uniquely identifies the type of the event. For example, clicking on a button,\r
 selecting an item from a list box and pressing a key on the keyboard all\r
 generate events with different event types.\r
 - <em>Event class</em> carried by the event: each event has some information\r
 associated with it and this data is represented by an object of a class\r
 derived from wxEvent. Events of different types can use the same event class,\r
 for example both button click and listbox selection events use wxCommandEvent\r
 class (as do all the other simple control events), but the key press event\r
 uses wxKeyEvent as the information associated with it is different.\r
 - <em>Event source</em>: wxEvent stores the object which generated the event\r
 and, for windows, its identifier (see @ref overview_events_winid). As it is\r
 common to have more than one object generating events of the same type (e.g. a\r
 typical window contains several buttons, all generating the same button click\r
 event), checking the event source object or its id allows to distinguish\r
 between them.\r
\r
\r
@section overview_events_eventhandling Event Handling\r
\r
There are two principal ways to handle events in wxWidgets. One of them uses\r
<em>event table</em> macros and allows you to define the connection between events\r
and their handlers only statically, i.e., during program compilation. The other\r
one uses wxEvtHandler::Connect() call and can be used to connect, and\r
disconnect, the handlers dynamically, i.e., during run-time depending on some\r
conditions. It also allows the direct connection of the events of one object to a\r
handler method in another object. The static event tables can only handle\r
events in the object where they are defined so using Connect() is more flexible\r
than using the event tables. On the other hand, event tables are more succinct\r
and centralize all event handlers connection in one place. You can either\r
choose a single approach that you find preferable or freely combine both\r
methods in your program in different classes or even in one and the same class,\r
although this is probably sufficiently confusing to be a bad idea.\r
\r
But before you make this choice, let us discuss these two ways in more\r
detail. In the next section we provide a short introduction to handling the\r
events using the event tables. Please see @ref overview_events_connect\r
for the discussion of Connect().\r
\r
@subsection overview_events_eventtables Event Handling with Event Tables\r
\r
To use an <em>event table</em> you must first decide in which class you wish to\r
handle the events. The only requirement imposed by wxWidgets is that this class\r
must derive from wxEvtHandler and so, considering that wxWindow derives from\r
it, any classes representing windows can handle events. Simple events such as\r
menu commands are usually processed at the level of a top-level window\r
containing the menu, so let's suppose that you need to handle some events in @c\r
MyFrame class deriving from wxFrame.\r
\r
First define one or more <em>event handlers</em>. They\r
are just simple (non-virtual) methods of the class that take as a parameter a\r
reference to an object of a wxEvent-derived class and have no return value (any\r
return information is passed via the argument, which is why it is non-const).\r
You also need to insert a macro\r
\r
@code\r
DECLARE_EVENT_TABLE()\r
@endcode\r
\r
somewhere in the class declaration. It doesn't matter where it appears but\r
it's customary to put it at the end because the macro changes the access\r
type internally so it's safest if nothing follows it. The\r
full class declaration might look like this:\r
\r
@code\r
class MyFrame : public wxFrame\r
{\r
public:\r
    MyFrame(...) : wxFrame(...) { }\r
\r
    ...\r
\r
protected:\r
    int m_whatever;\r
\r
private:\r
    // Notice that as the event handlers normally are not called from outside\r
    // the class, they normally are private. In particular they don't need\r
    // to be public.\r
    void OnExit(wxCommandEvent& event);\r
    void OnButton1(wxCommandEvent& event);\r
    void OnSize(wxSizeEvent& event);\r
\r
    // it's common to call the event handlers OnSomething() but there is no\r
    // obligation to do that; this one is an event handler too:\r
    void DoTest(wxCommandEvent& event);\r
\r
    DECLARE_EVENT_TABLE()\r
};\r
@endcode\r
\r
Next the event table must be defined and, as with any definition, it must be\r
placed in an implementation file. The event table tells wxWidgets how to map\r
events to member functions and in our example it could look like this:\r
\r
@code\r
BEGIN_EVENT_TABLE(MyFrame, wxFrame)\r
    EVT_MENU(wxID_EXIT, MyFrame::OnExit)\r
    EVT_MENU(DO_TEST, MyFrame::DoTest)\r
    EVT_SIZE(MyFrame::OnSize)\r
    EVT_BUTTON(BUTTON1, MyFrame::OnButton1)\r
END_EVENT_TABLE()\r
@endcode\r
\r
Notice that you must mention a method you want to use for the event handling in\r
the event table definition; just defining it in MyFrame class is @e not enough.\r
\r
Let us now look at the details of this definition: the first line means that we\r
are defining the event table for MyFrame class and that its base class is\r
wxFrame, so events not processed by MyFrame will, by default, be handled by\r
wxFrame. The next four lines define connections of individual events to their\r
handlers: the first two of them map menu commands from the items with the\r
identifiers specified as the first macro parameter to two different member\r
functions. In the next one, @c EVT_SIZE means that any changes in the size of\r
the frame will result in calling OnSize() method. Note that this macro doesn't\r
need a window identifier, since normally you are only interested in the current\r
window's size events.\r
\r
The @c EVT_BUTTON macro demonstrates that the originating event does not have to\r
come from the window class implementing the event table -- if the event source\r
is a button within a panel within a frame, this will still work, because event\r
tables are searched up through the hierarchy of windows for the command events.\r
(But only command events, so you can't catch mouse move events in a child\r
control in the parent window in the same way because wxMouseEvent doesn't\r
derive from wxCommandEvent. See below for how you can do it.) In this case, the\r
button's event table will be searched, then the parent panel's, then the\r
frame's.\r
\r
Finally, you need to implement the event handlers. As mentioned before, all\r
event handlers take a wxEvent-derived argument whose exact class differs\r
according to the type of event and the class of the originating window. For\r
size events, wxSizeEvent is used. For menu commands and most control commands\r
(such as button presses), wxCommandEvent is used. When controls get more\r
complicated, more specific wxCommandEvent-derived event classes providing\r
additional control-specific information can be used, such as wxTreeEvent for\r
events from wxTreeCtrl windows.\r
\r
In the simplest possible case an event handler may not use the @c event\r
parameter at all. For example,\r
\r
@code\r
void MyFrame::OnExit(wxCommandEvent& WXUNUSED(event))\r
{\r
    // when the user selects "Exit" from the menu we should close\r
    Close(true);\r
}\r
@endcode\r
\r
In other cases you may need some information carried by the @c event argument,\r
as in:\r
\r
@code\r
void MyFrame::OnSize(wxSizeEvent& event)\r
{\r
    wxSize size = event.GetSize();\r
\r
    ... update the frame using the new size ...\r
}\r
@endcode\r
\r
You will find the details about the event table macros and the corresponding\r
wxEvent-derived classes in the discussion of each control generating these\r
events.\r
\r
\r
@subsection overview_events_connect Dynamic Event Handling\r
\r
As with the event tables, decide in which class you intend to\r
handle the events first and, as before, this class must derive from\r
wxEvtHandler (usually indirectly via wxWindow). See the declaration of MyFrame\r
in the previous section. However the similarities end here and both the syntax\r
and the possibilities of handling events in this way are rather different.\r
\r
Let us start by looking at the syntax: the first obvious difference is that you\r
need not use DECLARE_EVENT_TABLE() nor BEGIN_EVENT_TABLE() and the\r
associated macros. Instead, in any place in your code, but usually in\r
the code of the class defining the handler itself (and definitely not in the\r
global scope as with the event tables), call its Connect() method like this:\r
\r
@code\r
MyFrame::MyFrame(...)\r
{\r
      Connect(wxID_EXIT, wxEVT_COMMAND_MENU_SELECTED,\r
                wxCommandEventHandler(MyFrame::OnExit));\r
}\r
@endcode\r
\r
This class should be self-explanatory except for wxCommandEventHandler part:\r
this is a macro that ensures that the method is of the correct type by using\r
static_cast in the same way as the event table macros.\r
\r
Now let us describe the semantic differences:\r
<ul>\r
    <li>\r
        Event handlers can be connected at any moment. For example, it's possible\r
        to do some initialization first and only connect the handlers if and when\r
        it succeeds. This can avoid the need to test that the object was properly\r
        initialized in the event handlers themselves. With Connect() they\r
        simply won't be called if it wasn't correctly initialized.\r
    </li>\r
\r
    <li>\r
        As a slight extension of the above, the handlers can also be\r
        Disconnect()-ed at any time and maybe later reconnected. Of course,\r
        it's also possible to emulate this behaviour with the classic\r
        static (i.e., connected via event tables) handlers by using an internal\r
        flag indicating whether the handler is currently enabled and returning\r
        from it if it isn't, but using dynamically connected handlers requires\r
        less code and is also usually more clear.\r
    </li>\r
\r
    <li>\r
        Also notice that you must derive a class inherited from, say,\r
        wxTextCtrl even if you don't want to modify the control behaviour at\r
        all but just want to handle some of its events. This is especially\r
        inconvenient when the control is loaded from the XRC. Connecting the\r
        event handler dynamically bypasses the need for this unwanted\r
        sub-classing.\r
    </li>\r
\r
    <li>\r
        Last but very, very far from least is the possibility to connect an\r
        event of some object to a method of another object. This is impossible\r
        to do with event tables because it is not possible to specify the\r
        object to dispatch the event to so it necessarily needs to be sent to\r
        the same object which generated the event. Not so with Connect() which\r
        has an optional @c eventSink parameter that can be used to specify the\r
        object which will handle the event. Of course, in this case the method\r
        being connected must belong to the class that is the type of the\r
        @c eventSink object! To give a quick example, people often want to catch\r
        mouse movement events happening when the mouse is in one of the frame\r
        children in the frame itself. Doing it in a naive way doesn't work:\r
        <ul>\r
            <li>\r
                A @c EVT_LEAVE_WINDOW(MyFrame::OnMouseLeave) line in the frame\r
                event table has no effect as mouse move (including entering and\r
                leaving) events are not propagated up to the parent window\r
                (at least not by default).\r
            </li>\r
\r
            <li>\r
                Putting the same line in a child event table will crash during\r
                run-time because the MyFrame method will be called on a wrong\r
                object -- it's easy to convince oneself that the only object\r
                that can be used here is the pointer to the child, as\r
                wxWidgets has nothing else. But calling a frame method with the\r
                child window pointer instead of the pointer to the frame is, of\r
                course, disastrous.\r
            </li>\r
        </ul>\r
\r
        However writing\r
        @code\r
            MyFrame::MyFrame(...)\r
            {\r
              m_child->Connect(wxID_ANY, wxEVT_LEAVE_WINDOW,\r
                               wxMouseEventHandler(MyFrame::OnMouseLeave),\r
                               NULL,  // unused extra data parameter\r
                               this); // this indicates the object to connect to\r
            }\r
        @endcode\r
        will work exactly as expected. Note that you can get the object that\r
        generated the event -- and that is not the same as the frame -- via\r
        wxEvent::GetEventObject() method of @c event argument passed to the\r
        event handler.\r
    </li>\r
</ul>\r
\r
To summarize, using Connect() requires slightly more typing but is much more\r
flexible than using static event tables so don't hesitate to use it when you\r
need this extra power. On the other hand, event tables are still perfectly fine\r
in simple situations where this extra flexibility is not needed.\r
\r
\r
@section overview_events_processing How Events are Processed\r
\r
The previous sections explain how to define event handlers but don't address\r
the question of how exactly wxWidgets finds the handler to call for the\r
given event. This section describes the algorithm used in detail.\r
\r
When an event is received from the windowing system, wxWidgets calls\r
wxEvtHandler::ProcessEvent() on the first event handler object belonging to the\r
window generating the event. The normal order of event table searching by\r
ProcessEvent() is as follows, with the event processing stopping as soon as a\r
handler is found (unless the handler calls wxEvent::Skip() in which case it\r
doesn't count as having handled the event and the search continues):\r
<ol>\r
    <li value="0">\r
    Before anything else happens, wxApp::FilterEvent() is called. If it returns\r
    anything but -1 (default), the event handling stops immediately.\r
    </li>\r
\r
    <li value="1">\r
    If this event handler is disabled via a call to\r
    wxEvtHandler::SetEvtHandlerEnabled() the next three steps are skipped and\r
    the event handler resumes at step (5).\r
    </li>\r
\r
    <li value="2">\r
    If the object is a wxWindow and has an associated validator, wxValidator\r
    gets a chance to process the event.\r
    </li>\r
\r
    <li value="3">\r
    The list of dynamically connected event handlers, i.e., those for which\r
    Connect() was called, is consulted. Notice that this is done before\r
    checking the static event table entries, so if both a dynamic and a static\r
    event handler match the same event, the static one is never going to be\r
    used.\r
    </li>\r
\r
    <li value="4">\r
    The event table containing all the handlers defined using the event table\r
    macros in this class and its base classes is examined. Notice that this\r
    means that any event handler defined in a base class will be executed at\r
    this step.\r
    </li>\r
\r
    <li value="5">\r
    The event is passed to the next event handler, if any, in the event handler\r
    chain, i.e., the steps (1) to (4) are done for it. This chain can be formed\r
    using wxEvtHandler::SetNextHandler():\r
        @image html overview_events_chain.png\r
    (referring to the image, if @c A->ProcessEvent is called and it doesn't handle\r
     the event, @c B->ProcessEvent will be called and so on...).\r
    In the case of wxWindow you can build a stack (implemented using wxEvtHandler\r
    double-linked list) using wxWindow::PushEventHandler():\r
        @image html overview_events_winstack.png\r
    (referring to the image, if @c W->ProcessEvent is called, it immediately calls\r
     @c A->ProcessEvent; if nor @c A nor @c B handle the event, then the wxWindow\r
     itself is used - i.e. the dynamically connected event handlers and static\r
     event table entries of wxWindow are looked as the last possibility, after\r
     all pushed event handlers were tested).\r
    Note however that usually there are no wxEvtHandler chains nor wxWindows stacks\r
    so this step will usually do anything.\r
    </li>\r
\r
    <li value="6">\r
    If the object is a wxWindow and the event is set to propagate (by default\r
    only wxCommandEvent-derived events are set to propagate), then the\r
    processing restarts from the step (1) (and excluding the step (7)) for the\r
    parent window. If this object is not a window but the next handler exists,\r
    the event is passed to its parent if it is a window. This ensures that in a\r
    common case of (possibly several) non-window event handlers pushed on top\r
    of a window, the event eventually reaches the window parent.\r
    </li>\r
\r
    <li value="7">\r
    Finally, i.e., if the event is still not processed, the wxApp object itself\r
    (which derives from wxEvtHandler) gets a last chance to process it.\r
    </li>\r
</ol>\r
\r
<em>Please pay close attention to step 6!</em> People often overlook or get\r
confused by this powerful feature of the wxWidgets event processing system. The\r
details of event propagation up the window hierarchy are described in the\r
next section.\r
\r
Also please notice that there are additional steps in the event handling for\r
the windows-making part of wxWidgets document-view framework, i.e.,\r
wxDocParentFrame, wxDocChildFrame and their MDI equivalents wxDocMDIParentFrame\r
and wxDocMDIChildFrame. The parent frame classes modify step (2) above to\r
send the events received by them to wxDocManager object first. This object, in\r
turn, sends the event to the current view and the view itself lets its\r
associated document process the event first. The child frame classes send\r
the event directly to the associated view which still forwards it to its\r
document object. Notice that to avoid remembering the exact order in which the\r
events are processed in the document-view frame, the simplest, and recommended,\r
solution is to only handle the events at the view classes level, and not in the\r
document or document manager classes\r
\r
\r
@subsection overview_events_propagation How Events Propagate Upwards\r
\r
As mentioned above, the events of the classes deriving from wxCommandEvent are\r
propagated by default to the parent window if they are not processed in this\r
window itself. But although by default only the command events are propagated\r
like this, other events can be propagated as well because the event handling\r
code uses wxEvent::ShouldPropagate() to check whether an event should be\r
propagated. It is also possible to propagate the event only a limited number of\r
times and not until it is processed (or a top level parent window is reached).\r
\r
Finally, there is another additional complication (which, in fact, simplifies\r
life of wxWidgets programmers significantly): when propagating the command\r
events up to the parent window, the event propagation stops when it\r
reaches the parent dialog, if any. This means that you don't risk getting\r
unexpected events from the dialog controls (which might be left unprocessed by\r
the dialog itself because it doesn't care about them) when a modal dialog is\r
popped up. The events do propagate beyond the frames, however. The rationale\r
for this choice is that there are only a few frames in a typical application\r
and their parent-child relation are well understood by the programmer while it\r
may be difficult, if not impossible, to track down all the dialogs that\r
may be popped up in a complex program (remember that some are created\r
automatically by wxWidgets). If you need to specify a different behaviour for\r
some reason, you can use wxWindow::SetExtraStyle(wxWS_EX_BLOCK_EVENTS)\r
explicitly to prevent the events from being propagated beyond the given window\r
or unset this flag for the dialogs that have it on by default.\r
\r
Typically events that deal with a window as a window (size, motion,\r
paint, mouse, keyboard, etc.) are sent only to the window.  Events\r
that have a higher level of meaning or are generated by the window\r
itself, (button click, menu select, tree expand, etc.) are command\r
events and are sent up to the parent to see if it is interested in the event.\r
\r
As mentioned above, only command events are recursively applied to the parents\r
event handler in the library itself. As this quite often causes confusion for\r
users, here is a list of system events that will @em not get sent to the\r
parent's event handler:\r
\r
@li wxEvent: The event base class\r
@li wxActivateEvent: A window or application activation event\r
@li wxCloseEvent: A close window or end session event\r
@li wxEraseEvent: An erase background event\r
@li wxFocusEvent: A window focus event\r
@li wxKeyEvent: A keypress event\r
@li wxIdleEvent: An idle event\r
@li wxInitDialogEvent: A dialog initialisation event\r
@li wxJoystickEvent: A joystick event\r
@li wxMenuEvent: A menu event\r
@li wxMouseEvent: A mouse event\r
@li wxMoveEvent: A move event\r
@li wxPaintEvent: A paint event\r
@li wxQueryLayoutInfoEvent: Used to query layout information\r
@li wxSetCursorEvent: Used for special cursor processing based on current mouse position\r
@li wxSizeEvent: A size event\r
@li wxScrollWinEvent: A scroll event sent by a scrolled window (not a scroll bar)\r
@li wxSysColourChangedEvent: A system colour change event\r
\r
In some cases, it might be desired by the programmer to get a certain number\r
of system events in a parent window, for example all key events sent to, but not\r
used by, the native controls in a dialog. In this case, a special event handler\r
will have to be written that will override ProcessEvent() in order to pass\r
all events (or any selection of them) to the parent window.\r
\r
\r
@section overview_events_custom Custom Event Summary\r
\r
@subsection overview_events_custom_general General approach\r
\r
Since version 2.2.x of wxWidgets, each event type is identified by an ID\r
given to the event type @e at runtime that makes it possible to add\r
new event types to the library or application without risking ID clashes\r
(two different event types mistakingly getting the same event ID).\r
This event type ID is stored in a struct of type <b>const wxEventType</b>.\r
\r
In order to define a new event type, there are principally two choices.\r
One is to define an entirely new event class (typically deriving from\r
wxEvent or wxCommandEvent).\r
\r
The other is to use the existing event classes and give them a new event\r
type. You'll have to define and declare a new event type either way\r
using the following macros:\r
\r
@code\r
// in the header of the source file\r
extern const wxEventType wxEVT_YOUR_EVENT_NAME;\r
\r
// in the implementation\r
DEFINE_EVENT_TYPE(wxEVT_YOUR_EVENT_NAME)\r
@endcode\r
\r
See also the @ref page_samples_event for an example of code\r
defining and working with the custom event types.\r
\r
\r
@subsection overview_events_custom_existing Using Existing Event Classes\r
\r
If you just want to use a wxCommandEvent with a new event type, use\r
one of the generic event table macros listed below, without having to define a\r
new event class yourself. This also has the advantage that you won't have to define a\r
new wxEvent::Clone() method for posting events between threads etc.\r
\r
Example:\r
\r
@code\r
extern const wxEventType wxEVT_MY_EVENT;\r
DEFINE_EVENT_TYPE(wxEVT_MY_EVENT)\r
\r
// user code intercepting the event\r
\r
BEGIN_EVENT_TABLE(MyFrame, wxFrame)\r
EVT_MENU    (wxID_EXIT, MyFrame::OnExit)\r
// ....\r
EVT_COMMAND (ID_MY_WINDOW, wxEVT_MY_EVENT, MyFrame::OnMyEvent)\r
END_EVENT_TABLE()\r
\r
void MyFrame::OnMyEvent( wxCommandEvent& event )\r
{\r
    // do something\r
    wxString text = event.GetText();\r
}\r
\r
\r
// user code sending the event\r
\r
void MyWindow::SendEvent()\r
{\r
    wxCommandEvent event( wxEVT_MY_EVENT, GetId() );\r
    event.SetEventObject( this );\r
\r
    // Give it some contents\r
    event.SetText( wxT("Hallo") );\r
\r
    // Send it\r
    GetEventHandler()->ProcessEvent( event );\r
}\r
@endcode\r
\r
\r
@subsection overview_events_custom_generic Generic Event Table Macros\r
\r
@beginTable\r
@row2col{EVT_CUSTOM(event\, id\, func),\r
        Allows you to add a custom event table\r
        entry by specifying the event identifier (such as wxEVT_SIZE),\r
        the window identifier, and a member function to call.}\r
@row2col{EVT_CUSTOM_RANGE(event\, id1\, id2\, func),\r
        The same as EVT_CUSTOM, but responds to a range of window identifiers.}\r
@row2col{EVT_COMMAND(id\, event\, func),\r
        The same as EVT_CUSTOM, but expects a member function with a\r
        wxCommandEvent argument.}\r
@row2col{EVT_COMMAND_RANGE(id1\, id2\, event\, func),\r
        The same as EVT_CUSTOM_RANGE, but\r
        expects a member function with a wxCommandEvent argument.}\r
@row2col{EVT_NOTIFY(event\, id\, func),\r
        The same as EVT_CUSTOM, but\r
        expects a member function with a wxNotifyEvent argument.}\r
@row2col{EVT_NOTIFY_RANGE(event\, id1\, id2\, func),\r
        The same as EVT_CUSTOM_RANGE, but\r
        expects a member function with a wxNotifyEvent argument.}\r
@endTable\r
\r
\r
@subsection overview_events_custom_ownclass Defining Your Own Event Class\r
\r
Under certain circumstances, you must define your own event\r
class e.g., for sending more complex data from one place to another. Apart\r
from defining your event class, you will also need to define your own\r
event table macro (which is quite long). Watch out to put in enough\r
casts to the inherited event function. Here is an example:\r
\r
@code\r
// code defining event\r
\r
class wxPlotEvent: public wxNotifyEvent\r
{\r
public:\r
    wxPlotEvent( wxEventType commandType = wxEVT_NULL, int id = 0 );\r
\r
    // accessors\r
    wxPlotCurve *GetCurve()\r
        { return m_curve; }\r
\r
    // required for sending with wxPostEvent()\r
    virtual wxEvent *Clone() const;\r
\r
private:\r
    wxPlotCurve   *m_curve;\r
};\r
\r
extern const wxEventType wxEVT_PLOT_ACTION;\r
typedef void (wxEvtHandler::*wxPlotEventFunction)(wxPlotEvent&);\r
\r
#define wxPlotEventHandler(func) \\r
    (wxObjectEventFunction)(wxEventFunction)wxStaticCastEvent(wxPlotEventFunction, &func)\r
#define EVT_PLOT(id, fn) \\r
    wx__DECLARE_EVT1(wxEVT_PLOT_ACTION, id, wxPlotEventHandler(fn))\r
\r
\r
// code implementing the event type and the event class\r
\r
DEFINE_EVENT_TYPE( wxEVT_PLOT_ACTION )\r
\r
wxPlotEvent::wxPlotEvent( ... )\r
{\r
    ...\r
}\r
\r
\r
// user code intercepting the event\r
\r
BEGIN_EVENT_TABLE(MyFrame, wxFrame)\r
EVT_PLOT  (ID_MY_WINDOW, MyFrame::OnPlot)\r
END_EVENT_TABLE()\r
\r
void MyFrame::OnPlot( wxPlotEvent &event )\r
{\r
    wxPlotCurve *curve = event.GetCurve();\r
}\r
\r
\r
// user code sending the event\r
\r
void MyWindow::SendEvent()\r
{\r
    wxPlotEvent event( wxEVT_PLOT_ACTION, GetId() );\r
    event.SetEventObject( this );\r
    event.SetCurve( m_curve );\r
    GetEventHandler()->ProcessEvent( event );\r
}\r
@endcode\r
\r
\r
@section overview_events_misc Miscellaneous Notes\r
\r
@subsection overview_events_virtual Event Handlers vs Virtual Methods\r
\r
It may be noted that wxWidgets' event processing system implements something\r
close to virtual methods in normal C++ in spirit: both of these mechanisms\r
allow you to alter the behaviour of the base class by defining the event handling\r
functions in the derived classes.\r
\r
There is however an important difference between the two mechanisms when you\r
want to invoke the default behaviour, as implemented by the base class, from a\r
derived class handler. With the virtual functions, you need to call the base\r
class function directly and you can do it either in the beginning of the\r
derived class handler function (to post-process the event) or at its end (to\r
pre-process the event). With the event handlers, you only have the option of\r
pre-processing the events and in order to still let the default behaviour\r
happen you must call wxEvent::Skip() and @em not call the base class event\r
handler directly. In fact, the event handler probably doesn't even exist in the\r
base class as the default behaviour is often implemented in platform-specific\r
code by the underlying toolkit or OS itself. But even if it does exist at\r
wxWidgets level, it should never be called directly as the event handlers are\r
not part of wxWidgets API and should never be called directly.\r
\r
Finally, please notice that the event handlers themselves shouldn't be virtual.\r
They should always be non-virtual and usually private (as there is no need to\r
make them public) methods of a wxEvtHandler-derived class.\r
\r
\r
@subsection overview_events_prog User Generated Events vs Programmatically Generated Events\r
\r
While generically wxEvents can be generated both by user\r
actions (e.g., resize of a wxWindow) and by calls to functions\r
(e.g., wxWindow::SetSize), wxWidgets controls normally send wxCommandEvent-derived\r
events only for the user-generated events. The only @b exceptions to this rule are:\r
\r
@li wxNotebook::AddPage: No event-free alternatives\r
@li wxNotebook::AdvanceSelection: No event-free alternatives\r
@li wxNotebook::DeletePage: No event-free alternatives\r
@li wxNotebook::SetSelection: Use wxNotebook::ChangeSelection instead, as\r
    wxNotebook::SetSelection is deprecated\r
@li wxTreeCtrl::Delete: No event-free alternatives\r
@li wxTreeCtrl::DeleteAllItems: No event-free alternatives\r
@li wxTreeCtrl::EditLabel: No event-free alternatives\r
@li All wxTextCtrl methods\r
\r
wxTextCtrl::ChangeValue can be used instead of wxTextCtrl::SetValue but the other\r
functions, such as wxTextCtrl::Replace or wxTextCtrl::WriteText don't have event-free\r
equivalents.\r
\r
\r
\r
@subsection overview_events_pluggable Pluggable Event Handlers\r
\r
<em>TODO: Probably deprecated, Connect() provides a better way to do this</em>\r
\r
In fact, you don't have to derive a new class from a window class\r
if you don't want to. You can derive a new class from wxEvtHandler instead,\r
defining the appropriate event table, and then call wxWindow::SetEventHandler\r
(or, preferably, wxWindow::PushEventHandler) to make this\r
event handler the object that responds to events. This way, you can avoid\r
a lot of class derivation, and use instances of the same event handler class (but different\r
objects as the same event handler object shouldn't be used more than once) to\r
handle events from instances of different widget classes.\r
\r
If you ever have to call a window's event handler\r
manually, use the GetEventHandler function to retrieve the window's event handler and use that\r
to call the member function. By default, GetEventHandler returns a pointer to the window itself\r
unless an application has redirected event handling using SetEventHandler or PushEventHandler.\r
\r
One use of PushEventHandler is to temporarily or permanently change the\r
behaviour of the GUI. For example, you might want to invoke a dialog editor\r
in your application that changes aspects of dialog boxes. You can\r
grab all the input for an existing dialog box, and edit it 'in situ',\r
before restoring its behaviour to normal. So even if the application\r
has derived new classes to customize behaviour, your utility can indulge\r
in a spot of body-snatching. It could be a useful technique for on-line\r
tutorials, too, where you take a user through a serious of steps and\r
don't want them to diverge from the lesson. Here, you can examine the events\r
coming from buttons and windows, and if acceptable, pass them through to\r
the original event handler. Use PushEventHandler/PopEventHandler\r
to form a chain of event handlers, where each handler processes a different\r
range of events independently from the other handlers.\r
\r
\r
\r
@subsection overview_events_winid Window Identifiers\r
\r
Window identifiers are integers, and are used to\r
uniquely determine window identity in the event system (though you can use it\r
for other purposes). In fact, identifiers do not need to be unique\r
across your entire application as long they are unique within the\r
particular context you're interested in, such as a frame and its children. You\r
may use the @c wxID_OK identifier, for example, on any number of dialogs\r
as long as you don't have several within the same dialog.\r
\r
If you pass @c wxID_ANY to a window constructor, an identifier will be\r
generated for you automatically by wxWidgets. This is useful when you don't\r
care about the exact identifier either because you're not going to process the\r
events from the control being created or because you process the events\r
from all controls in one place (in which case you should specify @c wxID_ANY\r
in the event table or wxEvtHandler::Connect call\r
as well). The automatically generated identifiers are always negative and so\r
will never conflict with the user-specified identifiers which must be always\r
positive.\r
\r
See @ref page_stdevtid for the list of standard identifiers available.\r
You can use wxID_HIGHEST to determine the number above which it is safe to\r
define your own identifiers. Or, you can use identifiers below wxID_LOWEST.\r
Finally, you can allocate identifiers dynamically using wxNewId() function too.\r
If you use wxNewId() consistently in your application, you can be sure that\r
your identifiers don't conflict accidentally.\r
\r
\r
@subsection overview_events_macros Event Handling Summary\r
\r
For the full list of event classes, please see the\r
@ref group_class_events "event classes group page".\r
\r
\r
*/\r
\r