Path: utzoo!utgpu!news-server.csri.toronto.edu!mailrus!uflorida!haven!uvaarpa!hudson!astsun7.astro.Virginia.EDU!cak3g
From: cak3g@astsun7.astro.Virginia.EDU (Colin Klipsch)
Newsgroups: comp.sys.mac.programmer
Subject: Re: Help!
Keywords: jGNEFilter, GetNextEvent, events
Message-ID: <2735@hudson.acc.virginia.edu>
Date: 20 Mar 90 20:51:12 GMT
References: <45183.25FD20CE@cmhgate.FIDONET.ORG> <25ff4e70.643e@polyslo.CalPoly.EDU> <14463@reed.UUCP>
Sender: news@hudson.acc.virginia.edu
Reply-To: cak3g@astsun7.astro.Virginia.EDU (Colin Klipsch)
Organization: University of Virginia
Lines: 71

In article <14463@reed.UUCP> chaffee@reed.UUCP (Alex Chaffee) writes:
>In article <25ff4e70.643e@polyslo.CalPoly.EDU> rcfische@polyslo.CalPoly.EDU 
><in the menu, even when the menu gets re-drawn.  Then you might want
><to look at hooking into GetNextEvent so that you can intercept mouse
><downs and see if they're on your icon.
>
>A small suggestion - rather than patching GetNextEvent, take advantage of
>the low-memory global jGNEFilter, which is sort of a built-in tail patch on
>GNE.  It's documeted in a tech note (I can't remember the number); I could
>give you source code if you're interested.
>
><Ray Fischer
><rcfische@polyslo.calpoly.edu
>-- 
>Alex Chaffee
>chaffee@reed.UUCP
>Reed College, Portland OR 97202
>____________________

Tech Note #85 talks about jGNEFilter.  It is a low-memory global which
points to code that will be executed upon every event, after the Mac
has done some of its own processing.  I used to use this method of
event filtering for TappyType, and it worked as follows. . .

In your INIT:

- Load your event-filter code, presumably stored as another
  resource, into the system heap.
- Make sure it's locked.  (i.e. set its resLocked bit beforehand)
- Detach it with _DetachResource.
- Take the whatever pointer you find in jGNEfilter and store it
  somewhere safe in your event-filter.
- Store the address of your event-filter in jGNEfilter.

Now every time an event happens, your event-filter (tucked safely
inside the system heap) will be called.  Hopefully.

In your event-filter:

- Upon entry, A1 should point to the event record.  Do whatever
  you want to do with the event.
- After you've done your thing, jump to the old value of jGNEfilter
  that you saved previously.

In effect, you are "splicing" your event-filter into the flow of
event processing.  If everyone were to do the same thing, then
many different event filters could co-exist.

Two Bad Things: Unfortunately, not everyone plays nicely, and thus
it may matter when your INIT runs in relation to others.  Secondly,
jGNEfilter is a low memory global, and thus it is not guaranteed to
be around in the long run.

Patching GetNextEvent, as a method of event filtering, won't work
except as a tail patch, which is a Thing Not To Be Done.

The scheme I use now is head-patching SystemEvent, which also seems to be
called after every event.  SystemEvent is "guaranteed" to be supported,
and you don't have to share a precarious memory global with random
strangers.

----------------------------------------------------------------------
"May the forces of evil become confused on the way to your house."
                                                      -- George Carlin
                            \ /
Bemusedly,                   | DISCLAIMER:
  Colin Klipsch              |   This text is actually a horrendously
  UVA Astronomy Department   |   garbled excerpt from _Mating_Rituals_
  Charlottesville, Virginia  |   of_West_African_Ostriches_, Vol IV,
  cak3g@virginia.edu         |   by Davis & Griffin, 1913, pp. 134-135
____________________________/ \_______________________________________