Intermediate ~15 min

Runes

Count the characters in "héllo". Five — obviously. Now ask Go:

len("héllo") // 6?!

Both answers are right; they’re answering different questions. len counts bytes, and é needs two bytes in UTF-8. The string holds six bytes that render as five characters. Everything in this exercise follows from that split.

Bytes are storage, runes are characters

A Go string is a read-only sequence of bytes. A rune is a Unicode code point — what you mean by “character” — and it’s an alias for int32:

'é'  // a rune literal (single quotes!): the number 233 with manners
"é"  // a string of length 2: the bytes 0xC3 0xA9

Single quotes make runes, double quotes make strings. Indexing a string yields a byte (s[1] is a uint8), which is why byte-indexing into non-English text returns beautiful nonsense. The string doesn’t know where its characters are — UTF-8 is a variable-width encoding, and only decoding reveals the boundaries.

Counting what humans count

The fix reaches for the standard library’s decoder:

import "unicode/utf8"

func CountRunes(s string) int {
    return utf8.RuneCountInString(s) // "héllo" -> 5
}

utf8.RuneCountInString walks the bytes and counts decoded characters. Its sibling utf8.DecodeRuneInString decodes one step at a time for manual iteration — but manual iteration has a better vehicle. Ranging over a string decodes as it goes:

n := 0
for range s { // index and rune available; both skippable
    n++
}

That’s the same range from the slice lesson, now doing double duty: over slices it yields indexes and elements, over strings it yields byte offsets and runes. The keyword didn’t change; the collection taught it a new trick.

So when is len(s) correct? When you want bytes — buffer sizes, wire lengths, capacity checks. When you want characters — display widths, input limits, “hello” counts — decode. Mismatching the two is an evergreen source of sliced-in-half accented characters; the strings section ahead will show the safe slicing patterns.

Hint

len counts bytes, and "é" needs two. Reach for the unicode/utf8 package’s rune counter — or range over the string, which decodes as it goes.

Task

Fix CountRunes in calc.go so multibyte characters count once.

Check

go test ./exercises/02_calculator/13_runes/ -v

Source: exercises/02_calculator/13_runes/README.md · Inspired by Mainmatter’s 100 Exercises to Learn Rust (CC BY-NC 4.0) — all Go prose here is original, free for non-commercial use.