Back in the days of the Commodore PET, I wanted to convert the keyboard
into Dvorak layout. Nowadays, I'd find the keyboard translation table, edit that and blow a new ROM.
Usually keyboards are wired as a matrix with the lines in the cable representing the X and Y scan lines and each switch at the junction of a pair of scan lines. The output scan lines have resistor pullups so they are normally all in the high state. One set of 'input' scan lines (say X) is then sequentially made LOW and the other set of scan 'output' lines is read in and decoded. When a switch is closed, one output line will go low when some specific input line is activated. This scan operation can happen very fast, so the keyboard is repeatedly scanned at 60 times a second or faster. (This matrix arrangement minimizes the number of required connections to the keys.) Then a computer decodes the XY coordinates into a key code. This would not be difficult to do with an arduino type processor: you'd use the output parallel port lines as the keyboard output scan lines, and another parallel input port as the input scan lines. Send the decoded information out the Arduino usb port. We used to give this to our EE students at Ryerson as an exercise in assembly language programming. Peter -- Peter Hiscocks Syscomp Electronic Design Limited, Toronto http://www.syscompdesign.com USB Oscilloscope and Waveform Generator 647-839-0325