I have a 2xN dataframe of chat messages, and I am trying to find the cleanest way to merge consecutive messages that originate from the same speaker. Here is a sample of the
