Szathmáry László honlapja @ DEIK | Nim2 / The ord() function

The ord() function (from std/system, link) is similar to Python.

The signature of the function:

func ord[T: Ordinal | enum](x: T): int {.magic: "Ord", ....}

The input parameter x is not necessarily a char! It can be an enum or an ordinal type. A char is also ordinal.

With ASCII

There's no problem with ASCII characters:

echo ord('A') # 65

echo 'a'.ord # 97

With Unicode

ord() with Unicode characters
print(ord("→")) # 8594	import std/unicode #echo ord('→') # ERROR # a char must be 1 byte long #echo ord("→") # ERROR # ord() cannot be called with a string echo ord("→".runeAt(0)) # 8594

I want to write about Unicode and UTF-8 encoding later.

Is short: a Rune in Nim is an int32. Any Unicode character can fit in an int32. The function .runeAt(0) takes the first Unicode character of the string. Since this arrow (→) cannot be stored in a char (since it's larger than 1 byte), the trick is to store it in a string and then extract the first Unicode character from it (though this string contains only 1 Unicode character). Again: more on this later.

What about Python? In Python, when you declare a char:

c = 'x'

then it's actually a 1-long string :) You can verify its type. Python does the heavy-lifting when you work with Unicode texts.