HN Explorer

2 comments

[ 2.5 ms ] story [ 20.2 ms ] thread
nicolewhite
Neat. This is something that R does out of the box:

  > s = c(1:3, 14, 29, 92:97)
  > s
   [1]  1  2  3 14 29 92 93 94 95 96 97
I think I prefer the 1:3 syntax versus 1-3, especially when dealing with negative numbers.

In R:

  > -6:-3
  [1] -6 -5 -4 -3
In intspan:

  >>> s = intspan('-6--3')
  >>> list(s)
  [-6, -5, -4, -3]
So the syntax is a little funky.

Another thing is that R will allow you to go from big to small:

  > 2:-3
  [1]  2  1  0 -1 -2 -3
Whereas intspan doesn't seem to support this at all:

  >>> s = intspan('2--3')
  >>> list(s)
  []
Very cool though!
jonathaneunice
Intspan's string representation was inspired by network news / UUCP, back in the first days of Usenet / Internet before the real Internet. So it was really only designed for whole numbers, not negatives. The syntax for negatives IS very clunky, mostly because they are rarely exercised/used in most of the use cases.

I like that "generalized range" operator from R. And it's a nice closed interval, not the half-open annoyance of Python's range() function.