Re: [g-a-devel] generateKeyboardEvent Patch for Java Access Bridge

[Matt Wringe <mwringe gmail com>]

On Fri, 18 Mar 2005 13:06:00 +0000, Bill Haneman <Bill Haneman sun com> wrote:
> Instead, your java application should just use the normal CORBA Java
> stubs to call registry.generateKeyboardEvent on the shared
> at-spi-registry object, it shouldn't package a Java reimplementation of
> the registry. 

Oops, your right. I guess seeing that the method was left blank made
me think that someone  hadn't gotten around to doing it. I don't
recall that there was any documentation in there explaining that it
shouldn't be used?
 
> You should not use X keycodes or X keysyms in a Java application; you
> should use the Java-equivalent codes, or use the 'string' version of the
> key event APIs.

Yes, thats my problem I can't find any java equivalent code to return
keycodes or keysyms.

For keysyms, I found java code in java-gnome (see previous message)
that would return them, and I also found that the awt KeyEvent can
also return them (although only for the capitalized character). But
this is really not a proper way of doing it.

For keycodes, wont any java code that return them have to communicate
with X? Its my understanding that keycodes change depending on what
keyboard is plugged in (I could be wrong). So, essentially the java
program will have to return the X keycodes. Right?

You are right that I could just use just be using strings instead of
keysyms or keycodes, if I was just trying to mimic a single keypress.
But to generate more complex keyboard events (ie ctrl alt R), you
would need to use keycodes (cause they are the only ones with key
press and release capabilities).

If you have any idea what the proper way of getting keycodes or
keysyms in java is, could you please let me. Any help on this would be
greatly appreciated.
 
> You haven't indicated what special characters you've found in
> UnknownImpl.java. I assume they're the ones in the Copyright statement,
> in which case I'd just replace them with the nearest Latin equivalents
> and leave them be.

Yes those were the characters I was refering to.


Thanks,

Matt Wringe