Szathmáry László honlapja @ DEIK | Nim2 / slicing an ASCII text

Nim has string slicing, and it's very similar to Python's approach. As long as you want to slice an ASCII text, no problem :) If you want to slice a text that has some fancy Unicode characters, then we'll need a slightly different approach.

Let's keep it simple and see how slicing works on an ASCII string.

(1) the diff
s = "Batman" print(s[0:3]) # Bat; first three characters	let s = "Batman" echo s[0..3] # Batm; closed interval echo s[0..<3] # Bat; position 3 is NOT included

Nim uses a range expression (ex.: echo s[0..3]), which defines a closed interval. In Python, the interval is open on the right side. Thus, if you want a Python-like slicing, write s[0..<3], which makes an interval that is closed on the left side and open on the right side, just like in Python.

(2) examples
s = "Batman" print(s[0]) # B print(s[0:3]) # Bat print(s[:3]) # Bat print(s[1:4]) # atm print(s[3:6]) # man print(s[3:200]) # man; right side is truncated print(s[3:]) # man print(s[-1]) # n print(s[:-1]) # Batma print(s[1:-1]) # atma	let s = "Batman" echo s[0] # B echo s[0..<3] # Bat echo s[0..<3] # Bat; both positions must be indicated echo s[1..<4] # atm echo s[3..<6] # man #echo s[3..200] # ERROR: out of range echo s[3 .. s.high] # man echo s[^1] # n echo s[0 ..< s.high] # Batma echo s[1 ..< s.high] # atma echo s.low # 0; index of the first element echo s.high # 5; index of the last element

With slicing, you can also modify some part of a mutable string:

var s = "batman"

s[0..<3] = "BAT"

echo s # BATman

Of course, if your string is immutable, you can still have the same result:

import std/strutils # strip, split, join

let
s = "batman"
t = s[0..<3].toUpper & s[3 .. s.high]

echo t # BATman

Problem with Unicode

Let's see the problem with Unicode characters:

let s = "László" # my name, BTW :)

echo s[0..<2] # we expect the first two characters: Lá

Output:

L�

The root of the problem is the following: strings are stored in UTF-8 format. The character 'á' is stored in 2 bytes. And s[0..<2] doesn't mean "take the first two characters". No! It means "take the first 2 bytes". And that gives 'L' and half of the character 'á'. That's why we get that funny '�' sign.

In Python it works as expected, we'd get "Lá". Python, as usual, tries to help us and does a lot of extra work behind the scenes.

But we'll see it in detail later.