[PATCH 34/42] Xi: hook up pointer emulation for emulating touches

[Chase Douglas]

On 12/14/2011 07:02 PM, Peter Hutterer wrote:
> Includes a hack for implicit grab activation, because integrating this
> requires a larger rewrite and I'm not sleeping enough as it is.
> Right now, we deliver the event and check before/after if there is an
> implicit grab on. If one activated, then store the event in the grab and
> switch the listener type to a grab listener.
> 
> Signed-off-by: Peter Hutterer <peter.hutterer at who-t.net>
> ---
>  Xi/exevents.c |  172 ++++++++++++++++++++++++++++++++++++++++++++++++++++++++-
>  1 files changed, 171 insertions(+), 1 deletions(-)
> 
> diff --git a/Xi/exevents.c b/Xi/exevents.c
> index d199493..ee8c828 100644
> --- a/Xi/exevents.c
> +++ b/Xi/exevents.c
> @@ -1188,6 +1188,148 @@ RetrieveTouchDeliveryData(DeviceIntPtr dev, TouchPointInfoPtr ti,
>      return TRUE;
>  }
>  
> +static int
> +DeliverTouchEmulatedEvent(DeviceIntPtr dev, TouchPointInfoPtr ti, InternalEvent *ev,
> +                          TouchListener *listener, ClientPtr client,
> +                          WindowPtr win, GrabPtr grab, XI2Mask *xi2mask)
> +{
> +    InternalEvent motion, button;
> +    InternalEvent *ptrev = &motion;
> +    int nevents;
> +    DeviceIntPtr kbd;
> +
> +    /* We don't deliver pointer events to non-owners */
> +    if (!TouchResourceIsOwner(ti, listener->listener))
> +        return Success;
> +
> +    nevents = TouchConvertToPointerEvent(ev, &motion, &button);
> +    BUG_WARN(nevents == 0);
> +    if (nevents == 0)
> +        return BadValue;
> +
> +    if (nevents > 1)
> +        ptrev = &button;
> +
> +    kbd = GetMaster(dev, KEYBOARD_OR_FLOAT);
> +    event_set_state(dev, kbd, &ptrev->device_event);
> +    ptrev->device_event.corestate = event_get_corestate(dev, kbd);
> +
> +    if (grab)
> +    {
> +        /* this side-steps the usual activation mechansims, but... */
> +        if (ev->any.type == ET_TouchBegin)
> +            ActivatePassiveGrab(dev, grab, ptrev, ev); /* also delivers the event */
> +        else {
> +            int deliveries = 0;
> +            /* 'grab' is the passive grab, but if the grab isn't active,
> +             * don't deliver */
> +            if (!dev->deviceGrab.grab)
> +                return Success;
> +
> +            if (grab->ownerEvents)
> +            {
> +                WindowPtr focus = NullWindow;
> +                WindowPtr win = dev->spriteInfo->sprite->win;
> +                deliveries = DeliverDeviceEvents(win, ptrev, grab, focus, dev);
> +            }
> +
> +            if (!deliveries)
> +                DeliverOneGrabbedEvent(ptrev, dev, grab->grabtype);
> +
> +            if (ev->any.type == ET_TouchEnd &&
> +                !dev->button->buttonsDown &&
> +                dev->deviceGrab.fromPassiveGrab &&
> +                GrabIsPointerGrab(grab))
> +                (*dev->deviceGrab.DeactivateGrab)(dev);
> +        }
> +    } else
> +    {
> +        GrabPtr devgrab = dev->deviceGrab.grab;
> +
> +        DeliverDeviceEvents(win, ptrev, grab, win, dev);
> +        /* FIXME: bad hack
> +         * Implicit passive grab activated in response to this event. Store
> +         * the event.
> +         */
> +        if (!devgrab && dev->deviceGrab.grab && dev->deviceGrab.implicitGrab)
> +        {
> +            TouchListener *listener;
> +
> +            devgrab = dev->deviceGrab.grab;
> +
> +            *dev->deviceGrab.sync.event = ev->device_event;

It took me a bit to realize that an implicit grab means the listener
must be a pointer selection, which would be the last listener because
all the preceding listeners represent grabs. Maybe a comment would be in
order, like:

The listener array has a sequence of grabs and then one event selection.
Implicit grab activation occurs through delivering an event selection.
Thus, we update the last listener in the array.

> +            listener = &ti->listeners[ti->num_listeners - 1];
> +            listener->listener = devgrab->resource;

Otherwise,

Reviewed-by: Chase Douglas <chase.douglas at canonical.com>