Go Single-character Names

Go loves short names. Even the authoritative page Effective Go uses this example:

f, err := os.Open(name)
if err != nil {
    return err
}
d, err := f.Stat()
if err != nil {
    f.Close()
    return err
}
codeUsing(f, d)

Short names are idiomatic Go, but overusing them creates cryptic, difficult-to-read code that is not idiomatic because, above all, Go is optimized for reading, especially by future engineers. This page describes four cases when single-character names—the shortest names possible—are idiomatic and acceptable.

But first: do not avoid an easy alternative:

f := "/tmp/foo"        // not ideal
tmpFile := "/tmp/foo"  // better: easy alternative

Judicious use of single-character names in Go is idiomatic, but easy alternatives are far better.

Loops

for i := 0; i < 100; i++ {
}

for k, v := range someMap {
}

for _, n := range someList {
}

Loop variables have near universal agreement for being single characters. In fact, not using i, j, and similar in for loops might even be considered bad style because this style is ubiquitous across languages.

However, you will find counter arguments that even loop vars should have short, meaningful names. I agree and advise the same: do not avoid an easy alternative. For example, suppose you have a map keyed on hostname with corresponding IP address:

for hostname, ip := range hostnames {
}

In this case, hostname and ip are clearly superior to k and v—an easy alternative—and they’re still very short.

Blocks


// Do something important with the file. I don't know what
// because I'm a little teapot and this example isn't real.
f, err := os.Open(name)
if err != nil {
    return err
}
d, err := f.Stat()
if err != nil {
    f.Close()
    return err
}
codeUsing(f, d)

That example from Effective Go is a single code block. (To emphasize that it’s a block, I added blank lines and a block code comment.) The variables f and d are declared and used only within the block, which makes it easy for humans (without the help of an IDE) to follow and understand them. This would be a bad, non-idiomatic example if either variable was used outside the block (later in function, for example).

Although acceptable, it’s better to avoid single-character variables in blocks unless the block is short (tens of lines) and well contained. If in doubt, err on the side of caution and use short descriptive variable names.

Receivers

type T struct {}

func (t *T) foo() {
}

Receiver variables are often single characters, as in t for the method receiver on type T. If a single character is too cryptic, then use two or three characters.

Using short descriptive names for receivers is less common in Go; for example:

func (worker *Worker) Slog() {  // less common
func (w *Worker) Slog() {       // more common

You see the more common form, w *Worker, even in Effective Go examples. Two- and three-character receivers are acceptable and sometimes better, like src instead of s.

Never use self or this for Go receiver names.

Arguments

// EscapeString escapes special characters ...
func EscapeString(s string) string {}

// Render renders the parse tree n to the given writer.
func Render(w io.Writer, n *Node) error {}

// Parse returns the parse tree for the HTML from the given Reader.
func Parse(r io.Reader) (*Node, error) {}

Single character arguments are practically the norm in the Go standard library. The three example above are from net/html.

There are three requirements that makes these single-character argument names acceptable and idiomatic:

  1. Great code comment on the function (which is why I included them in the examples)
  2. Not too many arguments
  3. Relatively limited, block-scoped usage

Render exemplifies the first and second; let’s see if it exemplifies the third:

func Render(w io.Writer, n *Node) error {
	if x, ok := w.(writer); ok {
		return render(x, n)
	}
	buf := bufio.NewWriter(w)
	if err := render(buf, n); err != nil {
		return err
	}
	return buf.Flush()
}

Yes: that function is one small block that makes it easy to follow arguments w and n.