Conversation

Replying to
This is an interesting request. Currently the "tokenization" of terms revolves around white space, lower casing and pulling out work fragments. Some emojis in the higher area of unicode should be searchable but searching for things like :) would be problematic. What type of
1