def unique$\_$tokens$($tweets$,$ tokenizer$=$nltk$.$TweetTokenizer$().$tokenize$)$:

$"""$Tokenizes the texts of an iterable of Tweet objects.

Accepts an iterable of either tweet dictionary objects

or tweet objects encoded as JSON text.

For backwards compatibility in the Twitter API, the function

should support Tweets with either a $`$full$\_$text$`$ field or

a $`$text$`$ field, defaulting to $`$full$\_$text$`$ if it exists.

Returns the set of unique tokens.

$"""$

pass

# TODO: Implement this function to return the set of

# unique tokens parsed from the tweet texts by the

# NLTK TweetTokenizer