Gets the latest committed text from the text editing component and removes it from the component's text body. This is used for the "Undo Commit" feature in some input methods, where the committed text reverts to its previous composed state. The composed text will be sent to the component using an InputMethodEvent.

Generally, this feature should only be supported immediately after the text was committed, not after the user performed other operations on the text. When the feature is not supported, return null.

The input method may provide a list of attributes that it is interested in. In that case, information about other attributes that the implementor may have need not be made accessible through the iterator. If the list is null, all available attribute information should be made accessible.

Creates a top-level Swing JFrame for use by the input method. The intended behavior of this window is:

• it floats above all document windows and dialogs
• it and all components that it contains do not receive the focus
• it has lightweight decorations, such as a reduced drag region without title

However, the actual behavior with respect to these three items is platform dependent.

The title may or may not be displayed, depending on the actual type of window created.

If attachToInputContext is true, the new window will share the input context that corresponds to this input method context, so that events for components in the window are automatically dispatched to the input method. Also, when the window is opened using setVisible(true), the input context will prevent deactivate and activate calls to the input method that might otherwise be caused.

Input methods must call Window.dispose on the returned input method window when it is no longer needed.

Creates a top-level window for use by the input method. The intended behavior of this window is:

• it floats above all document windows and dialogs
• it and all components that it contains do not receive the focus
• it has lightweight decorations, such as a reduced drag region without title

However, the actual behavior with respect to these three items is platform dependent.

The title may or may not be displayed, depending on the actual type of window created.

If attachToInputContext is true, the new window will share the input context that corresponds to this input method context, so that events for components in the window are automatically dispatched to the input method. Also, when the window is opened using setVisible(true), the input context will prevent deactivate and activate calls to the input method that might otherwise be caused.

Input methods must call Window.dispose on the returned input method window when it is no longer needed.

Creates an input method event from the arguments given and dispatches it to the client component. For arguments, see java.awt.event.InputMethodEvent#InputMethodEvent .

Enables or disables notification of the current client window's location and state for the specified input method. When notification is enabled, the input method's notifyClientWindowChange method is called as described in that method's specification. Notification is automatically disabled when the input method is disposed.

Gets an iterator providing access to the entire text and attributes contained in the text editing component except for uncommitted text. Uncommitted (composed) text should be ignored for index calculations and should not be made accessible through the iterator.

The input method may provide a list of attributes that it is interested in. In that case, information about other attributes that the implementor may have need not be made accessible through the iterator. If the list is null, all available attribute information should be made accessible.

Gets the length of the entire text contained in the text editing component except for uncommitted (composed) text.

Gets the offset of the insert position in the committed text contained in the text editing component. This is the offset at which characters entered through an input method are inserted. This information is used by an input method, for example, to examine the text surrounding the insert position.

Gets the offset within the composed text for the specified absolute x and y coordinates on the screen. This information is used, for example to handle mouse clicks and the mouse cursor. The offset is relative to the composed text, so offset 0 indicates the beginning of the composed text.

Return null if the location is outside the area occupied by the composed text.

Gets the currently selected text from the text editing component. This may be used for a variety of purposes. One of them is the "Reconvert" feature in some input methods. In this case, the input method will typically send an input method event to replace the selected text with composed text. Depending on the input method's capabilities, this may be the original composed text for the selected text, the latest composed text entered anywhere in the text, or a version of the text that's converted back from the selected text.

The input method may provide a list of attributes that it is interested in. In that case, information about other attributes that the implementor may have need not be made accessible through the iterator. If the list is null, all available attribute information should be made accessible.

Gets the location of a specified offset in the current composed text, or of the selection in committed text. This information is, for example, used to position the candidate window near the composed text, or a composition window near the location where committed text will be inserted.

If the component has composed text (because the most recent InputMethodEvent sent to it contained composed text), then the offset is relative to the composed text - offset 0 indicates the first character in the composed text. The location returned should be for this character.

If the component doesn't have composed text, the offset should be ignored, and the location returned should reflect the beginning (in line direction) of the highlight in the last line containing selected text. For example, for horizontal left-to-right text (such as English), the location to the left of the left-most character on the last line containing selected text is returned. For vertical top-to-bottom text, with lines proceding from right to left, the location to the top of the left-most line containing selected text is returned.

The location is represented as a 0-thickness caret, that is, it has 0 width if the text is drawn horizontally, and 0 height if the text is drawn vertically. Other text orientations need to be mapped to horizontal or vertical orientation. The rectangle uses absolute screen coordinates.