Prior to this change all <tspan> elements without x, and y attributes would
just inherit the parent elements such attributes and be placed in the same
location. This didn't respect whether these attributes were consumed already
by the parent.

Having multiple x and y elements, or having a different number of x and y
elements was also not handled in line with the rules for SVG.

This allows the LookAndFeel of submenus to query the target component
used for the top-level menu. getTargetComponent() isn't suitable for
this because the target component is set to null for submenus, and this
behaviour can't be changed without potentially breaking code that relies
on the current behaviour.

This fixes a bug, where VoiceOver would interact with the InnerNSView in
the AudioPluginHost. Instead the focus now jumps from the minimise
button directly to the plugin's first accessible element.

On MacOS the rule is: when returning accessibilityChildren an ignored
child must be replaced with its own children. When returning the
accessibilityParent, an ignored element must be replaced with its own
parent.

The accessibilityValue method of the AccessibilityCheckBox and
AccessibilityRadioButton protocols is required to return a
boolean valued NSNumber.

This patch should resolve an issue introduced in
0ab30555fc where arrow keys and other
directional keys (home, end, page up, page down) stopped working as
expected.

With this patch in place,
- IME input in plugins should work correctly, including for languages
  with a selection palette (Japanese) and languages where multiple
  keypresses combine to a single character (Korean).
- Keyboard shortcuts should work (cut, copy, paste)
- Directional keys should work

The default value of this property changed in Sonoma, but we still want
to use the old default.

While improving Android IME support (da38c1ed), text editor destructors
were updated to explicitly pass keyboard focus elsewhere.
As far as I remember, the change was intended to prevent the text input
system from trying to send input events to components while they were
being destroyed, in which case the TextInputTarget and Component bases
may be 'valid', but the data members referenced by the TextInputTarget
implementation may have been destroyed.

The motivation for removing these lines is that giving away focus and
sending a focus event can cause all components to become unfocused. This
is problematic in the case of slider text editors - pressing 'enter'
will cause the TextEditor to be destroyed, but the parent component will
fail to gain focus, so pressing 'tab' will not have any effect.

Running under Gnome, calling ComponentPeer::forceSetBounds immediately
after creating a new peer causes the peer to emit a visibility-changed
event because the window manager sometimes reports that the new window
is in IconicState.

_NET_WM_STATE_HIDDEN seems to more accurately reflect whether the window
is really minimised or off-screen.