I have a program that talks to a device via a serial interface. Structurally it looks like this:
Program A -> module B -> Serial I want to add a protocol layer around the serial port without modifying any of the modules above and I want to be able to use BOTH cases of the program whenever convenient, so at first it seemed like a simple case of sub-classing the Serial class and then (this is the problem) somehow use either the original serial class definition when I wanted the original program functionality or the new class when I wanted the extra layer present. Module B has import serial so my first thought was to put the new module in the current directory and call it "serial.py". This would allow me to run the program with the new Serial class and when I didn't want that extra layer, all I had to do was rename the module to something else and let Module B pick up serial from the site-package area. But (obviously) this doesn't work because the new class's first few statements are: import serial Class Serial(serial.Serial): So when Module B does the import (and gets the new class definition), I get an error message (as the new definition is loaded): AttributeError: 'module' object has no attribute 'Serial' on the Class declaration statement. Is there anyway to have Module B use the new class without modifying the content of Module B? The only other alternative I can think of is to create a copy of the original program, rename the new class file and have two complete copies of the program but that is not ideal :-) Thanks Peter -- http://mail.python.org/mailman/listinfo/python-list
