Go Tidbit: Ellipsize Strings Without Breaking Unicode

February 6, 2023 #100DaysToOffload #Go #Tidbit

Strings in Go support Unicode. If you come from a programming language like C, you may think of strings as an array of (byte-sized) characters.

In Go, you can convert a string to a byte slice, and access/manipulate each byte. But if you want to truncate or ellipsize the string to a specific length, you have to think of a string like a slice of runes.

Take this string for example: "বাংলাদেশ".

If you want to ellipsize it to a length of n characters, you may run a code like this:

 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
12
13
14
15
16
17
18
19
20
21
22
func ellipsis(s string, n int) string {
	if len(s) > n {
		return s[:n]+"…"
	}
	return s
}

func main() {
	var s = "বাংলাদেশ"
	for n := 1; n <= 8; n++ {
		fmt.Println(ellipsis(s, n))
	}
	// Output:
	// �…
	// ��…
	// ব…
	// ব�…
	// ব��…
	// বা…
	// বা�…
	// বা��…
}

Huh! Something needs to be fixed. What is the sane and desired output is:

1
2
3
4
5
6
7
8
ব…
বা…
বাং…
বাংল…
বাংলা…
বাংলাদ…
বাংলাদে…
বাংলাদেশ

To correctly truncate a string with Unicode characters, convert the string to a slice of runes first:

 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
func ellipsis(s string, n int) string {
	r := []rune(s) // Convert s to a slice of runes and truncate it instead.
	if len(r) > n {
		return string(r[:n]) + "…"
	}
	return s
}

func main() {
	var s = "বাংলাদেশ"
	for n := 1; n <= 8; n++ {
		fmt.Println(ellipsis(s, n))
	}
	// Output:
	// ব…
	// বা…
	// বাং…
	// বাংল…
	// বাংলা…
	// বাংলাদ…
	// বাংলাদে…
	// বাংলাদেশ
}

This post is 15th of my #100DaysToOffload challenge. Want to get involved? Find out more at 100daystooffload.com.

Tags ·

RSS

comments powered by Disqus