Hacker News Reader

Comment by rbonvall | Hacker News Reader

I think automatic string concatenation and singleton tuples were not introduced according to some high-level goal. They are just historical baggage. Automatic string concatenation comes from C, and the singleton tuple syntax probably just seemed like a good idea at first.

In hindsight, singleton tuples are not common or useful enough to deserve their own syntax. If the way to create them was something like this:

    t = tuple.single("hello")

we'd thing it's ugly or inconsistent, but definitely not confusing or bug-prone.

View on HN

One place where singleton tuples used to be common is with the old "%"-formatting, specifically in the case where there is a single argument and its value might be a tuple:

    x = (1,2,3)
    #print("the value of x is %s" % x)   # breaks if x is a tuple
    print("the value of x is %s" % (x,)) # works even if x is a tuple

There is a readable way to create singleton tuples, without the sneaky trailing comma or a new function like tuple.single:

    tuple(["hello"])

The square brackets can be slightly annoying. I recall writing the following function to omit them:

    def tup(*args):
        return tuple(args)

This basically lets you use the usual tuple syntax, just prefixed with the word "tup". The advantages are that you don't need a trailing comma for singleton tuples, and it's more obvious that a tuple is being created (it can be difficult to distinguish between tuple literals and parentheses used for grouping in a complex expression).

I am reminded of a somewhat similar issue with empty set literals: {1,2} is a set, {1} is a set, but {} is a dict. The way to create empty sets is using set().

Reply on news.ycombinator.com