Feature #18986: Raises Exception for Range#last(n) with Float::INFINITY

• Status: Open
• Assignee: -
• Target version: -

[masasakano (Masa Sakano)]

When a (Numeric) argument is given to Range#last(n) for a Range with Float::INFINITY, it runs into an infinite loop (see a small code snippet below).
There is no use case for the feature as far as I can think of, and it is hard to detect.

(6..Float::INFINITY).last(1)  # => infinite loop!!
(6..).last(1)   # cannot get the last element of endless range (RangeError)
(-Float::INFINITY..4).last(1) # can't iterate from Float (TypeError)
(..4).last(1)                 # can't iterate from Float (TypeError)

Let me suggest raising an Exception (maybe RangeError, in the same way as an endless Range) when last(n) is specified for a Range whose end value is Float::INFINITY.
I note (6..Float::INFINITY).to_a has the same issue.