split.go

package netargv

import "errors"

// ErrUnclosedQuote is returned when a quoted span in a message header is never closed.
var ErrUnclosedQuote = errors.New("unclosed quote in message header")

// splitWords splits s into shell-style tokens.
//
// Rules:
//   - Tokens are separated by ASCII whitespace.
//   - A double-quoted span ("...") is a single token; the quotes are stripped.
//   - A single-quoted span ('...') is a single token; the quotes are stripped.
//   - No escape sequences, variable expansion, or glob expansion.
//   - Unclosed quotes are an error.
//   - A backslash outside quotes is not special; line continuation is resolved
//     by the caller before this function is invoked.
func splitWords(s string) ([]string, error) {
	var tokens []string

	i := 0
	for i < len(s) {
		// Skip leading whitespace.
		for i < len(s) && isSpace(s[i]) {
			i++
		}
		if i >= len(s) {
			break
		}

		// Fast path: token contains no quotes — track start index and slice s directly.
		// Only allocate a []byte buffer when a quoted span forces concatenation.
		tokenStart := i
		hasQuote := false

		// Scan forward to detect whether this token contains any quotes before
		// the first whitespace boundary.
		for j := i; j < len(s) && !isSpace(s[j]); j++ {
			if s[j] == '"' || s[j] == '\'' {
				hasQuote = true
				break
			}
		}

		if !hasQuote {
			// No quote in this token: find the end and slice — zero alloc.
			for i < len(s) && !isSpace(s[i]) {
				i++
			}
			tokens = append(tokens, s[tokenStart:i])
			continue
		}

		// Quoted token: build into a []byte buffer so we can strip quote chars
		// and merge bare + quoted spans into one token.
		var cur []byte
		for i < len(s) && !isSpace(s[i]) {
			ch := s[i]
			if ch == '"' || ch == '\'' {
				quote := ch
				i++ // skip opening quote
				for {
					if i >= len(s) {
						return nil, ErrUnclosedQuote
					}
					if s[i] == quote {
						i++ // skip closing quote
						break
					}
					cur = append(cur, s[i])
					i++
				}
				continue
			}
			cur = append(cur, ch)
			i++
		}
		tokens = append(tokens, string(cur))
	}

	return tokens, nil
}

func isSpace(c byte) bool {
	return c == ' ' || c == '\t' || c == '\r' || c == '\n'
}