Table of Contents
Except under unusual circumstances, widgets do not hardwire the mapping of user events into widget behavior by using the event manager. Instead, they provide a default mapping of events into behavior that you can override.
The translation manager provides an interface to specify and manage the mapping of X event sequences into widget-supplied functionality, for example, calling procedure Abc when the y key is pressed.
The translation manager uses two kinds of tables to perform translations:
The action tables, which are in the widget class structure, specify the mapping of externally available procedure name strings to the corresponding procedure implemented by the widget class.
A translation table, which is in the widget class structure, specifies the mapping of event sequences to procedure name strings.
You can override the translation table in the class structure for a specific widget instance by supplying a different translation table for the widget instance. The resources XtNtranslations and XtNbaseTranslations are used to modify the class default translation table; see the section called “Translation Table Management”.
All widget class records contain an action table,
an array of
XtActionsRec
entries.
In addition,
an application can register its own action tables with the translation manager
so that the translation tables it provides to widget instances can access
application functionality directly.
The translation action procedure pointer is of type
(*XtActionProc)
.
typedef void (*XtActionProc)(Widget w, XEvent *event, String *params, Cardinal *num_params);
typedef void (*XtActionProc)(Widget w, XEvent *event, String *params, Cardinal *num_params);
w | Specifies the widget that caused the action to be called. |
event | Specifies the event that caused the action to be called. If the action is called after a sequence of events, then the last event in the sequence is used. |
params | Specifies a pointer to the list of strings that were specified in the translation table as arguments to the action, or NULL. |
num_params | Specifies the number of entries in params. |
typedef struct _XtActionsRec {
String string;
XtActionProc proc;
} XtActionsRec, *XtActionList;
The string field is the name used in translation tables to access the procedure. The proc field is a pointer to a procedure that implements the functionality.
When the action list is specified as the
CoreClassPart
actions field, the string pointed to by string must be
permanently allocated prior to or during the execution of the class
initialization procedure and must not be subsequently deallocated.
Action procedures should not assume that the widget in which they are invoked is realized; an accelerator specification can cause an action procedure to be called for a widget that does not yet have a window. Widget writers should also note which of a widget's callback lists are invoked from action procedures and warn clients not to assume the widget is realized in those callbacks.
For example, a Pushbutton widget has procedures to take the following actions:
Set the button to indicate it is activated.
Unset the button back to its normal mode.
Highlight the button borders.
Unhighlight the button borders.
Notify any callbacks that the button has been activated.
The action table for the Pushbutton widget class makes these functions available to translation tables written for Pushbutton or any subclass. The string entry is the name used in translation tables. The procedure entry (usually spelled identically to the string) is the name of the C procedure that implements that function:
XtActionsRec actionTable[] = {
{"Set", Set},
{"Unset", Unset},
{"Highlight", Highlight},
{"Unhighlight", Unhighlight}
{"Notify", Notify},
};
The Intrinsics reserve all action names and parameters starting with the characters “Xt” for future standard enhancements. Users, applications, and widgets should not declare action names or pass parameters starting with these characters except to invoke specified built-in Intrinsics functions.
The actions and num_actions fields of
CoreClassPart
specify the actions implemented by a widget class. These are
automatically registered with the Intrinsics when the class is initialized
and must be allocated in writable storage prior to Core class_part
initialization, and never deallocated. To save memory and optimize
access, the Intrinsics may overwrite the storage in order to compile the
list into an internal representation.
To declare an action table within an application
and register it with the translation manager, use
XtAppAddActions
.
app_context | Specifies the application context. |
actions | Specifies the action table to register. |
num_actions | Specifies the number of entries in this action table. |
If more than one action is registered with the same name,
the most recently registered action is used.
If duplicate actions exist in an action table,
the first is used.
The Intrinsics register an action table containing
XtMenuPopup
and
XtMenuPopdown
as part of
XtCreateApplicationContext
.
The translation manager uses a simple algorithm to resolve the name of a procedure specified in a translation table into the actual procedure specified in an action table. When the widget is realized, the translation manager performs a search for the name in the following tables, in order:
The widget's class and all superclass action tables, in subclass-to-superclass order.
The parent's class and all superclass action tables, in subclass-to-superclass order, then on up the ancestor tree.
The action tables registered with
XtAppAddActions
andXtAddActions
from the most recently added table to the oldest table.
As soon as it finds a name, the translation manager stops the search. If it cannot find a name, the translation manager generates a warning message.
An application can specify a procedure that will be called just before
every action routine is dispatched by the translation manager. To do
so, the application supplies a procedure pointer of type
(*XtActionHookProc)
.
typedef void (*XtActionHookProc)(Widget w, XtPointer client_data, String action_name, XEvent* event, String* params, Cardinal* num_params);
typedef void (*XtActionHookProc)(Widget w, XtPointer client_data, String action_name, XEvent* event, String* params, Cardinal* num_params);
w | Specifies the widget whose action is about to be dispatched. |
client_data |
Specifies the application-specific closure that was passed to
|
action_name | Specifies the name of the action to be dispatched. |
event | Specifies the event argument that will be passed to the action routine. |
params | Specifies the action parameters that will be passed to the action routine. |
num_params | Specifies the number of entries in params. |
Action hooks should not modify any of the data pointed to by the arguments other than the client_data argument.
To add an action hook, use
XtAppAddActionHook
.
app | Specifies the application context. |
proc | Specifies the action hook procedure. |
client_data | Specifies application-specific data to be passed to the action hook. |
XtAppAddActionHook
adds the specified procedure to the front of a list
maintained in the application context. In the future, when an action
routine is about to be invoked for any widget in this application
context, either through the translation manager or via
XtCallActionProc
,
the action hook procedures will be called in reverse
order of registration just prior to invoking the action routine.
Action hook procedures are removed automatically and the
XtActionHookId is
destroyed when the application context in which
they were added is destroyed.
To remove an action hook procedure without destroying the application
context, use
XtRemoveActionHook
.
void XtRemoveActionHook(XtActionHookId id);
void XtRemoveActionHook(XtActionHookId id);
id |
Specifies the action hook id returned by
|
XtRemoveActionHook
removes the specified action hook procedure from
the list in which it was registered.
All widget instance records contain a translation table, which is a resource with a default value specified elsewhere in the class record. A translation table specifies what action procedures are invoked for an event or a sequence of events. A translation table is a string containing a list of translations from an event sequence into one or more action procedure calls. The translations are separated from one another by newline characters (ASCII LF). The complete syntax of translation tables is specified in Appendix B.
As an example, the default behavior of Pushbutton is
Highlight on enter window.
Unhighlight on exit window.
Invert on left button down.
Call callbacks and reinvert on left button up.
The following illustrates Pushbutton's default translation table:
static String defaultTranslations =
"<EnterWindow>: Highlight()\\n\\
<LeaveWindow>: Unhighlight()\\n\\
<Btn1Down>: Set()\\n\\
<Btn1Up>: Notify() Unset()";
The tm_table field of the
CoreClassPart
should be filled in at class initialization time with
the string containing the class's default translations.
If a class wants to inherit its superclass's translations,
it can store the special value
XtInheritTranslations
into tm_table.
In Core's class part initialization procedure,
the Intrinsics compile this translation table into an efficient internal form.
Then, at widget creation time,
this default translation table is
combined with the XtNtranslations
and XtNbaseTranslations resources; see
the section called “Translation Table Management”.
The resource conversion mechanism automatically compiles
string translation tables that are specified in the resource database.
If a client uses translation tables that are not retrieved via a
resource conversion,
it must compile them itself using
XtParseTranslationTable
.
The Intrinsics use the compiled form of the translation table to register the necessary events with the event manager. Widgets need do nothing other than specify the action and translation tables for events to be processed by the translation manager.
An event sequence is a comma-separated list of X event descriptions that describes a specific sequence of X events to map to a set of program actions. Each X event description consists of three parts: The X event type, a prefix consisting of the X modifier bits, and an event-specific suffix.
Various abbreviations are supported to make translation tables easier to read. The events must match incoming events in left-to-right order to trigger the action sequence.
Action sequences specify what program or widget actions to take in response to incoming X events. An action sequence consists of space-separated action procedure call specifications. Each action procedure call consists of the name of an action procedure and a parenthesized list of zero or more comma-separated string parameters to pass to that procedure. The actions are invoked in left-to-right order as specified in the action sequence.
Translation table entries may specify actions that are taken when two
or more identical events occur consecutively within a short time
interval, called the multi-click time. The multi-click time value may
be specified as an application resource with name “multiClickTime” and
class “MultiClickTime” and may also be modified dynamically by the
application. The multi-click time is unique for each Display value and
is retrieved from the resource database by
XtDisplayInitialize
.
If no value is specified, the initial value is 200 milliseconds.
To set the multi-click time dynamically, use
XtSetMultiClickTime
.
void XtSetMultiClickTime(Display *display, int time);
void XtSetMultiClickTime(Display *display, int time);
display | Specifies the display connection. |
time | Specifies the multi-click time in milliseconds. |
XtSetMultiClickTime
sets the time interval used by the translation
manager to determine when multiple events are interpreted as a
repeated event. When a repeat count is specified in a translation
entry, the interval between the timestamps in each pair of repeated
events (e.g., between two
ButtonPress
events) must be less than the
multi-click time in order for the translation actions to be taken.
To read the multi-click time, use
XtGetMultiClickTime
.
int XtGetMultiClickTime(Display *display);
int XtGetMultiClickTime(Display *display);
display | Specifies the display connection. |
XtGetMultiClickTime
returns the time in milliseconds that the
translation manager uses to determine if multiple events are to be
interpreted as a repeated event for purposes of matching a translation
entry containing a repeat count.
Sometimes an application needs to merge its own translations with a widget's translations. For example, a window manager provides functions to move a window. The window manager wishes to bind this operation to a specific pointer button in the title bar without the possibility of user override and bind it to other buttons that may be overridden by the user.
To accomplish this, the window manager should first create the title bar and then should merge the two translation tables into the title bar's translations. One translation table contains the translations that the window manager wants only if the user has not specified a translation for a particular event or event sequence (i.e., those that may be overridden). The other translation table contains the translations that the window manager wants regardless of what the user has specified.
Three Intrinsics functions support this merging:
XtParseTranslationTable | Compiles a translation table. |
XtAugmentTranslations | Merges a compiled translation table into a widget's compiled translation table, ignoring any new translations that conflict with existing translations. |
XtOverrideTranslations | Merges a compiled translation table into a widget's compiled translation table, replacing any existing translations that conflict with new translations. |
To compile a translation table, use
XtParseTranslationTable
.
table | Specifies the translation table to compile. |
The
XtParseTranslationTable
function compiles the translation table, provided in the format given
in Appendix B, into an opaque internal representation
of type
XtTranslations
.
Note that if an empty translation table is required for any purpose,
one can be obtained by calling
XtParseTranslationTable
and passing an empty string.
To merge additional translations into an existing translation table, use
XtAugmentTranslations
.
void XtAugmentTranslations(Widget w, XtTranslations translations);
void XtAugmentTranslations(Widget w, XtTranslations translations);
w | Specifies the widget into which the new translations are to be merged. Must be of class Core or any subclass thereof. |
translations | Specifies the compiled translation table to merge in. |
The
XtAugmentTranslations
function merges the new translations into the existing widget
translations, ignoring any
#replace
,
#augment
,
or
#override
directive that may have been specified
in the translation string. The translation table specified by
translations is not altered by this process.
XtAugmentTranslations
logically appends the string representation of the new translations to
the string representation of the widget's current translations and reparses
the result with no warning messages about duplicate left-hand sides, then
stores the result back into the widget instance; i.e.,
if the new translations contain an event or event sequence that
already exists in the widget's translations,
the new translation is ignored.
To overwrite existing translations with new translations, use
XtOverrideTranslations
.
w | Specifies the widget into which the new translations are to be merged. Must be of class Core or any subclass thereof. |
translations | Specifies the compiled translation table to merge in. |
The
XtOverrideTranslations
function merges the new translations into the existing widget
translations, ignoring any
#replace
,
#augment
,
or
#override
directive that may have been
specified in the translation string. The translation table
specified by translations is not altered by this process.
XtOverrideTranslations
logically appends the string representation of the widget's current
translations to the string representation of the new translations and
reparses the result with no warning messages about duplicate left-hand
sides, then stores the result back into the widget instance; i.e.,
if the new translations contain an event or event sequence that
already exists in the widget's translations,
the new translation overrides the widget's translation.
To replace a widget's translations completely, use
XtSetValues
on the XtNtranslations resource and specify a compiled translation table
as the value.
To make it possible for users to easily modify translation tables in their resource files, the string-to-translation-table resource type converter allows the string to specify whether the table should replace, augment, or override any existing translation table in the widget. To specify this, a pound sign (#) is given as the first character of the table followed by one of the keywords “replace”, “augment”, or “override” to indicate whether to replace, augment, or override the existing table. The replace or merge operation is performed during the Core instance initialization. Each merge operation produces a new translation resource value; if the original tables were shared by other widgets, they are unaffected. If no directive is specified, “#replace” is assumed.
At instance initialization the XtNtranslations resource is first fetched. Then, if it was not specified or did not contain “#replace”, the resource database is searched for the resource XtNbaseTranslations. If XtNbaseTranslations is found, it is merged into the widget class translation table. Then the widget translations field is merged into the result or into the class translation table if XtNbaseTranslations was not found. This final table is then stored into the widget translations field. If the XtNtranslations resource specified “#replace”, no merge is done. If neither XtNbaseTranslations or XtNtranslations are specified, the class translation table is copied into the widget instance.
To completely remove existing translations, use
XtUninstallTranslations
.
void XtUninstallTranslations(Widget w);
void XtUninstallTranslations(Widget w);
w | Specifies the widget from which the translations are to be removed. Must be of class Core or any subclass thereof. |
The
XtUninstallTranslations
function causes the entire translation table for the widget to be removed.
It is often desirable to be able to bind events in one widget to actions in another. In particular, it is often useful to be able to invoke menu actions from the keyboard. The Intrinsics provide a facility, called accelerators, that lets you accomplish this. An accelerator table is a translation table that is bound with its actions in the context of a particular widget, the source widget. The accelerator table can then be installed on one or more destination widgets. When an event sequence in the destination widget would cause an accelerator action to be taken, and if the source widget is sensitive, the actions are executed as though triggered by the same event sequence in the accelerator source widget. The event is passed to the action procedure without modification. The action procedures used within accelerators must not assume that the source widget is realized nor that any fields of the event are in reference to the source widget's window if the widget is realized.
Each widget instance contains that widget's exported accelerator table
as a resource.
Each class of widget exports a method that takes a
displayable string representation of the accelerators
so that widgets can display their current accelerators.
The representation is the accelerator table in canonical
translation table form (see Appendix B).
The display_accelerator procedure pointer is of type
(*XtStringProc)
.
w | Specifies the source widget that supplied the accelerators. |
string | Specifies the string representation of the accelerators for this widget. |
Accelerators can be specified in resource files,
and the string representation is the same as for a translation table.
However,
the interpretation of the
#augment
and
#override
directives applies to
what will happen when the accelerator is installed;
that is, whether or not the accelerator translations will override the
translations in the destination widget.
The default is
#augment
,
which means that the accelerator translations have lower priority
than the destination translations.
The
#replace
directive is ignored for accelerator tables.
To parse an accelerator table, use
XtParseAcceleratorTable
.
source | Specifies the accelerator table to compile. |
The
XtParseAcceleratorTable
function compiles the accelerator table into an opaque internal representation.
The client
should set the XtNaccelerators resource of
each widget that is to be activated by these translations
to the returned value.
To install accelerators from a widget on another widget, use
XtInstallAccelerators
.
void XtInstallAccelerators(Widget destination, Widget source);
void XtInstallAccelerators(Widget destination, Widget source);
destination | Specifies the widget on which the accelerators are to be installed. Must be of class Core or any subclass thereof. |
source | Specifies the widget from which the accelerators are to come. Must be of class Core or any subclass thereof. |
The
XtInstallAccelerators
function installs the accelerators resource value from
source onto destination
by merging the source accelerators into the destination translations.
If the source display_accelerator field is non-NULL,
XtInstallAccelerators
calls it with the source widget and a string representation
of the accelerator table,
which indicates that its accelerators have been installed
and that it should display them appropriately.
The string representation of the accelerator table is its
canonical translation table representation.
As a convenience for installing all accelerators from a widget and all its
descendants onto one destination, use
XtInstallAllAccelerators
.
destination | Specifies the widget on which the accelerators are to be installed. Must be of class Core or any subclass thereof. |
source | Specifies the root widget of the widget tree from which the accelerators are to come. Must be of class Core or any subclass thereof. |
The
XtInstallAllAccelerators
function recursively descends the widget tree rooted at source
and installs the accelerators resource value
of each widget encountered onto destination.
A common use is to call
XtInstallAllAccelerators
and pass the application main window as the source.
The translation manager provides support for automatically translating
KeyCodes in incoming key events into KeySyms.
KeyCode-to-KeySym translator procedure pointers are of type
(*XtKeyProc)
.
typedef void (*XtKeyProc)(Display *display, KeyCode keycode, Modifiers modifiers, Modifiers *modifiers_return, KeySym *keysym_return);
typedef void (*XtKeyProc)(Display *display, KeyCode keycode, Modifiers modifiers, Modifiers *modifiers_return, KeySym *keysym_return);
display | Specifies the display that the KeyCode is from. |
keycode | Specifies the KeyCode to translate. |
modifiers | Specifies the modifiers to the KeyCode. |
modifiers_return | Specifies a location in which to store a mask that indicates the subset of all modifiers that are examined by the key translator for the specified keycode. |
keysym_return | Specifies a location in which to store the resulting KeySym. |
This procedure takes a KeyCode and modifiers and produces a KeySym. For any given key translator function and keyboard encoding, modifiers_return will be a constant per KeyCode that indicates the subset of all modifiers that are examined by the key translator for that KeyCode.
The KeyCode-to-KeySym translator procedure
must be implemented such that multiple calls with the same
display, keycode, and modifiers return the same
result until either a new case converter, an
(*XtCaseProc)
,
is installed or a
MappingNotify
event is received.
The Intrinsics maintain tables internally to map KeyCodes to KeySyms for each open display. Translator procedures and other clients may share a single copy of this table to perform the same mapping.
To return a pointer to the KeySym-to-KeyCode mapping table for a
particular display, use
XtGetKeysymTable
.
KeySym *XtGetKeysymTable(Display *display, KeyCode *min_keycode_return, int *keysyms_per_keycode_return);
KeySym *XtGetKeysymTable(Display *display, KeyCode *min_keycode_return, int *keysyms_per_keycode_return);
display | Specifies the display whose table is required. |
min_keycode_return | Returns the minimum KeyCode valid for the display. |
keysyms_per_keycode_return | Returns the number of KeySyms stored for each KeyCode. |
XtGetKeysymTable
returns a pointer to the Intrinsics' copy of the
server's KeyCode-to-KeySym table. This table must not be modified.
There are keysyms_per_keycode_return KeySyms associated with each
KeyCode, located in the table with indices starting at index
(test_keycode - min_keycode_return) * keysyms_per_keycode_return
for KeyCode test_keycode. Any entries that have no KeySyms associated
with them contain the value
NoSymbol
.
Clients should not cache the KeySym table but should call
XtGetKeysymTable
each time the value is
needed, as the table may change prior to dispatching each event.
For more information on this table, see Section 12.7 in Xlib — C Language X Interface..
To register a key translator, use
XtSetKeyTranslator
.
void XtSetKeyTranslator(Display *display, XtKeyProc proc);
void XtSetKeyTranslator(Display *display, XtKeyProc proc);
display | Specifies the display from which to translate the events. |
proc | Specifies the procedure to perform key translations. |
The
XtSetKeyTranslator
function sets the specified procedure as the current key translator.
The default translator is
XtTranslateKey
,
an
(*XtKeyProc)
that uses the Shift, Lock, numlock, and group modifiers
with the interpretations defined in X Window System Protocol, Section 5.
It is provided so that new translators can call it to get default
KeyCode-to-KeySym translations and so that the default translator
can be reinstalled.
To invoke the currently registered KeyCode-to-KeySym translator,
use
XtTranslateKeycode
.
void XtTranslateKeycode(Display *display, KeyCode keycode, Modifiers modifiers, Modifiers *modifiers_return, KeySym *keysym_return);
void XtTranslateKeycode(Display *display, KeyCode keycode, Modifiers modifiers, Modifiers *modifiers_return, KeySym *keysym_return);
display | Specifies the display that the KeyCode is from. |
keycode | Specifies the KeyCode to translate. |
modifiers | Specifies the modifiers to the KeyCode. |
modifiers_return | Returns a mask that indicates the modifiers actually used to generate the KeySym. |
keysym_return | Returns the resulting KeySym. |
The
XtTranslateKeycode
function passes the specified arguments
directly to the currently registered KeyCode-to-KeySym translator.
To handle capitalization of nonstandard KeySyms, the Intrinsics allow
clients to register case conversion routines.
Case converter procedure pointers are of type
(*XtCaseProc)
.
typedef void (*XtCaseProc)(Display *display, KeySym keysym, KeySym *lower_return, KeySym *upper_return);
typedef void (*XtCaseProc)(Display *display, KeySym keysym, KeySym *lower_return, KeySym *upper_return);
display | Specifies the display connection for which the conversion is required. |
keysym | Specifies the KeySym to convert. |
lower_return | Specifies a location into which to store the lowercase equivalent for the KeySym. |
upper_return | Specifies a location into which to store the uppercase equivalent for the KeySym. |
If there is no case distinction, this procedure should store the KeySym into both return values.
To register a case converter, use
XtRegisterCaseConverter
.
void XtRegisterCaseConverter(Display *display, XtCaseProc proc, KeySym start, KeySym stop);
void XtRegisterCaseConverter(Display *display, XtCaseProc proc, KeySym start, KeySym stop);
display | Specifies the display from which the key events are to come. |
proc |
Specifies the
|
start | Specifies the first KeySym for which this converter is valid. |
stop | Specifies the last KeySym for which this converter is valid. |
The
XtRegisterCaseConverter
registers the specified case converter.
The start and stop arguments provide the inclusive range of KeySyms
for which this converter is to be called.
The new converter overrides any previous converters for KeySyms in that range.
No interface exists to remove converters;
you need to register an identity converter.
When a new converter is registered,
the Intrinsics refresh the keyboard state if necessary.
The default converter understands case conversion for all
Latin KeySyms defined in X Window System Protocol, Appendix A.
To determine uppercase and lowercase equivalents for a KeySym, use
XtConvertCase
.
void XtConvertCase(Display *display, KeySym keysym, KeySym *lower_return, KeySym *upper_return);
void XtConvertCase(Display *display, KeySym keysym, KeySym *lower_return, KeySym *upper_return);
display | Specifies the display that the KeySym came from. |
keysym | Specifies the KeySym to convert. |
lower_return | Returns the lowercase equivalent of the KeySym. |
upper_return | Returns the uppercase equivalent of the KeySym. |
The
XtConvertCase
function calls the appropriate converter and returns the results.
A user-supplied
(*XtKeyProc)
may need to use this function.
When an action procedure is invoked on a
KeyPress
or
KeyRelease
event, it often has a need to retrieve the KeySym and modifiers
corresponding to the event that caused it to be invoked. In order to
avoid repeating the processing that was just performed by the
Intrinsics to match the translation entry, the KeySym and modifiers
are stored for the duration of the action procedure and are made
available to the client.
To retrieve the KeySym and modifiers that matched the final event
specification in the translation table entry, use
XtGetActionKeysym
.
KeySym XtGetActionKeysym(XEvent *event, Modifiers *modifiers_return);
KeySym XtGetActionKeysym(XEvent *event, Modifiers *modifiers_return);
event | Specifies the event pointer passed to the action procedure by the Intrinsics. |
modifiers_return | Returns the modifiers that caused the match, if non-NULL. |
If
XtGetActionKeysym
is called after an action procedure has been
invoked by the Intrinsics and before that action procedure returns, and
if the event pointer has the same value as the event pointer passed to
that action routine, and if the event is a
KeyPress
or
KeyRelease
event, then
XtGetActionKeysym
returns the KeySym that matched the final
event specification in the translation table and, if modifiers_return
is non-NULL, the modifier state actually used to generate this KeySym;
otherwise, if the event is a
KeyPress
or
KeyRelease
event, then
XtGetActionKeysym
calls
XtTranslateKeycode
and returns the results;
else it returns
NoSymbol
and does not examine modifiers_return.
Note that if an action procedure invoked by the Intrinsics
invokes a subsequent action procedure (and so on) via
XtCallActionProc
,
the nested action procedure may also call
XtGetActionKeysym
to retrieve the Intrinsics' KeySym and modifiers.
To return the list of KeyCodes that map to a particular KeySym in
the keyboard mapping table maintained by the Intrinsics, use
XtKeysymToKeycodeList
.
void XtKeysymToKeycodeList(Display *display, KeySym keysym, KeyCode **keycodes_return, Cardinal *keycount_return);
void XtKeysymToKeycodeList(Display *display, KeySym keysym, KeyCode **keycodes_return, Cardinal *keycount_return);
display | Specifies the display whose table is required. |
keysym | Specifies the KeySym for which to search. |
keycodes_return | Returns a list of KeyCodes that have keysym associated with them, or NULL if keycount_return is 0. |
keycount_return | Returns the number of KeyCodes in the keycode list. |
The
XtKeysymToKeycodeList
procedure returns all the KeyCodes that have keysym
in their entry for the keyboard mapping table associated with display.
For each entry in the
table, the first four KeySyms (groups 1 and 2) are interpreted as
specified by X Window System Protocol, Section 5. If no KeyCodes map to the
specified KeySym, keycount_return is zero and *keycodes_return is NULL.
The caller should free the storage pointed to by keycodes_return using
XtFree
when it is no longer useful. If the caller needs to examine
the KeyCode-to-KeySym table for a particular KeyCode, it should call
XtGetKeysymTable
.
To register button and key grabs for a widget's window according to the
event bindings in the widget's translation table, use
XtRegisterGrabAction
.
void XtRegisterGrabAction(XtActionProc action_proc, Boolean owner_events, unsigned int event_mask, int pointer_mode);
void XtRegisterGrabAction(XtActionProc action_proc, Boolean owner_events, unsigned int event_mask, int pointer_mode);
action_proc | Specifies the action procedure to search for in translation tables. |
owner_events | |
event_mask | |
pointer_mode | |
keyboard_mode |
Specify arguments to
|
XtRegisterGrabAction
adds the specified action_proc to a list known to
the translation manager. When a widget is realized, or when the
translations of a realized widget or the accelerators installed on a
realized widget are modified, its translation table and any installed
accelerators are scanned for action procedures on this list.
If any are invoked on
ButtonPress
or
KeyPress
events as the only or final event
in a sequence, the Intrinsics will call
XtGrabButton
or
XtGrabKey
for the widget with every button or KeyCode which maps to the
event detail field, passing the specified owner_events, event_mask,
pointer_mode, and keyboard_mode. For
ButtonPress
events, the modifiers
specified in the grab are determined directly from the translation
specification and confine_to and cursor are specified as
None
.
For
KeyPress
events, if the translation table entry specifies colon (:) in
the modifier list, the modifiers are determined by calling the key
translator procedure registered for the display and calling
XtGrabKey
for every combination of standard modifiers which map the KeyCode to
the specified event detail KeySym, and ORing any modifiers specified in
the translation table entry, and event_mask is ignored. If the
translation table entry does not specify colon in the modifier list,
the modifiers specified in the grab are those specified in the
translation table entry only. For both
ButtonPress
and
KeyPress
events, don't-care modifiers are ignored unless the translation entry
explicitly specifies “Any” in the modifiers field.
If the specified action_proc is already registered for the calling process, the new values will replace the previously specified values for any widgets that become realized following the call, but existing grabs are not altered on currently realized widgets.
When translations or installed accelerators are modified for a
realized widget, any previous key or button grabs registered
as a result of the old bindings are released if they do not appear in
the new bindings and are not explicitly grabbed by the client with
XtGrabKey
or
XtGrabButton
.
Normally action procedures are invoked by the Intrinsics when an
event or event sequence arrives for a widget. To
invoke an action procedure directly, without generating
(or synthesizing) events, use
XtCallActionProc
.
void XtCallActionProc(Widget widget, String action, XEvent *event, String *params, Cardinal num_params);
void XtCallActionProc(Widget widget, String action, XEvent *event, String *params, Cardinal num_params);
widget | Specifies the widget in which the action is to be invoked. Must be of class Core or any subclass thereof. |
action | Specifies the name of the action routine. |
event | Specifies the contents of the event passed to the action routine. |
params | Specifies the contents of the params passed to the action routine. |
num_params | Specifies the number of entries in params. |
XtCallActionProc
searches for the named action routine in the same
manner and order as translation tables are bound, as described in
Section 10.1.2, except that application action tables are searched, if
necessary, as of the time of the call to
XtCallActionProc
.
If found,
the action routine is invoked with the specified widget, event pointer,
and parameters. It is the responsibility of the caller to ensure that
the contents of the event, params, and num_params arguments are
appropriate for the specified action routine and, if necessary, that
the specified widget is realized or sensitive. If the named action
routine cannot be found,
XtCallActionProc
generates a warning message and returns.
Occasionally a subclass will require the pointers to one or more of
its superclass's action procedures. This would be needed, for
example, in order to envelop the superclass's action. To retrieve
the list of action procedures registered in the superclass's
actions field, use
XtGetActionList
.
void XtGetActionList(WidgetClass widget_class, XtActionList *actions_return, Cardinal *num_actions_return);
void XtGetActionList(WidgetClass widget_class, XtActionList *actions_return, Cardinal *num_actions_return);
widget_class | Specifies the widget class whose actions are to be returned. |
actions_return | Returns the action list. |
num_actions_return | Returns the number of action procedures declared by the class. |
XtGetActionList
returns the action table defined by the specified
widget class. This table does not include actions defined by the
superclasses. If widget_class is not initialized, or is not
coreWidgetClass
or a subclass thereof, or if the class does not define any actions,
*actions_return will be NULL and *num_actions_return
will be zero.
If *actions_return is non-NULL the client is responsible for freeing
the table using
XtFree
when it is no longer needed.