Email difference algorithm

I would like to replicate gmails functionality to "magically" not show irrelevant quoted stuff in emails but still showing mostly relevant stuf开发者_Python百科f. Are there any libraries which can help me find the text that is actually new and should be shown? Or do you have any suggestions on how to proceed?

I do know which two messages belong together and which one is the answer to the other but I would love to only show relevant text.