hakurei/internal/rosa/azalea/azalea.go

// Package azalea implements a proof-of-concept, domain-specific language for
// Rosa OS software packaging.
package azalea

import (
	"errors"
	"io"
	"strconv"
	"text/scanner"
)

// idents are runes accepted in an identifier.
var idents = [...]bool{
	'0': true, '1': true, '2': true, '3': true, '4': true, '5': true, '6': true,
	'7': true, '8': true, '9': true,

	'A': true, 'B': true, 'C': true, 'D': true, 'E': true, 'F': true, 'G': true,
	'H': true, 'I': true, 'J': true, 'K': true, 'L': true, 'M': true, 'N': true,
	'O': true, 'P': true, 'Q': true, 'R': true, 'S': true, 'T': true, 'U': true,
	'V': true, 'W': true, 'X': true, 'Y': true, 'Z': true,

	'a': true, 'b': true, 'c': true, 'd': true, 'e': true, 'f': true, 'g': true,
	'h': true, 'i': true, 'j': true, 'k': true, 'l': true, 'm': true, 'n': true,
	'o': true, 'p': true, 'q': true, 'r': true, 's': true, 't': true, 'u': true,
	'v': true, 'w': true, 'x': true, 'y': true, 'z': true,

	'-': true, '_': true,
}

// TokenError describes an unexpected token.
type TokenError [2]rune

func (e TokenError) Error() string {
	return "expected " + scanner.TokenString(e[0]) +
		", found " + scanner.TokenString(e[1])
}

// ExprError is an unexpected token encountered while parsing an expression.
type ExprError rune

func (e ExprError) Error() string {
	return "unexpected token " + scanner.TokenString(rune(e))
}

// must1 returns v, or panics if err is not nil.
func must1[T any](v T, err error) T {
	if err != nil {
		panic(err)
	}
	return v
}

// parser retains the current token.
type parser struct {
	s   scanner.Scanner
	tok rune
}

// scan advances the underlying scanner to the next token, storing its result.
func (p *parser) scan() rune { p.tok = p.s.Scan(); return p.tok }

// scanAs advances the scanner for an expected token.
func (p *parser) scanAs(expects rune) {
	e := TokenError{expects, p.scan()}
	if e[0] != e[1] {
		panic(e)
	}
}

// parseInt parses the current token as a base 10 representation of a 64-bit
// signed integer.
func (p *parser) parseInt() int64 {
	return must1(strconv.ParseInt(p.s.TokenText(), 10, 64))
}

// A String represents an identifier or string literal.
type String struct {
	Value string
	Ident bool
}

// A StringSpec describes a statement evaluating down to a string value.
type StringSpec []String

// parseString parses the current token as a string.
func (p *parser) parseString() string {
	return must1(strconv.Unquote(p.s.TokenText()))
}

// appendStringSpec parses from the next token until the end of the StringSpec.
// It always advances past the final token.
func (p *parser) appendStringSpec(
	op bool,
	data StringSpec,
) (v StringSpec, ok bool) {
	ok = true
	v = data
	for {
		if op {
			switch tok := p.scan(); tok {
			case '+':
				break

			default:
				return
			}
		}

		switch tok := p.scan(); tok {
		case scanner.String, scanner.RawString:
			v = append(v, String{p.parseString(), false})

		case scanner.Ident:
			v = append(v, String{p.s.TokenText(), true})

		default:
			ok = op
			return
		}
		op = true
	}
}

// A KV holds a key/value pair.
type KV struct {
	K string
	V any
}

// An Arg represents an argument of [Func].
type Arg struct {
	K []string
	V any
	R bool
}

// Func is a function call or package declaration.
type Func struct {
	// Function or package identifier.
	Ident string
	// Whether this is a package declaration.
	Package bool
	// Key-value arguments.
	Args []Arg
}

// parseExpr scans and parses the current expression. A nil return indicates
// [scanner.EOF].
func (p *parser) parseExpr() (any, bool) {
	switch p.tok {
	case scanner.EOF:
		return nil, false

	case scanner.Int:
		return p.parseInt(), false

	case scanner.String, scanner.RawString:
		v, ok := p.appendStringSpec(true, StringSpec{
			{p.parseString(), false},
		})
		if !ok {
			panic(TokenError{scanner.String, p.tok})
		}
		return v, true

	case scanner.Ident:
		var v Func
		v.Ident = p.s.TokenText()
		if v.Package = v.Ident == "package"; v.Package {
			p.scanAs(scanner.Ident)
			v.Ident = p.s.TokenText()
		}

		p.scan()
		switch p.tok {
		case '{':
			for {
				p.scan()
				switch p.tok {
				case '}':
					return v, false

				case scanner.Ident:
					break

				default:
					panic(TokenError{scanner.Ident, p.tok})
				}

				var next bool
				arg := Arg{K: []string{p.s.TokenText()}}
				delim := true
			arg:
				for {
					p.scan()
					switch p.tok {
					case ',':
						if delim {
							delim = false
							continue
						}
						panic(ExprError(p.tok))

					case scanner.Ident:
						if delim {
							panic(TokenError{',', p.tok})
						}
						delim = true
						arg.K = append(arg.K, p.s.TokenText())

					default:
						break arg
					}
				}
				switch p.tok {
				case '=':
					break

				case '*':
					arg.R = true
					p.scanAs('=')

				default:
					panic(TokenError{'=', p.tok})
				}
				p.scan()
				arg.V, next = p.parseExpr()
				v.Args = append(v.Args, arg)
				if !next {
					p.scanAs(';')
				}
			}

		case ';':
			return StringSpec{{v.Ident, true}}, true

		case '+':
			s, ok := p.appendStringSpec(false, StringSpec{
				{v.Ident, true},
			})
			if !ok {
				panic(TokenError{scanner.String, p.tok})
			}
			return s, p.tok != scanner.EOF

		case scanner.EOF:
			return StringSpec{{v.Ident, true}}, false

		default:
			return StringSpec{{v.Ident, true}}, true
		}

	case '{':
		var v []KV
		for {
			p.scan()
			switch p.tok {
			case '}':
				return v, false

			case scanner.String:
				pair := KV{K: p.parseString()}
				p.scan()
				switch p.tok {
				case ';':
					break

				case ':':
					var next bool
					p.scan()
					pair.V, next = p.parseExpr()
					if !next {
						p.scanAs(';')
					}
					break

				default:
					panic(ExprError(p.tok))
				}
				v = append(v, pair)

			default:
				panic(ExprError(p.tok))
			}
		}

	case '[':
		var (
			v     []any
			e     any
			delim bool
			next  bool
		)
		for {
			if !next {
				p.scan()
			}
			switch p.tok {
			case ',':
				if delim {
					delim = false
					next = false
					continue
				}
				panic(ExprError(','))
			case ']':
				return v, false
			case scanner.EOF:
				panic(ExprError(scanner.EOF))
			default:
				if delim {
					panic(TokenError{',', p.tok})
				}
				delim = true
				break
			}

			e, next = p.parseExpr()
			v = append(v, e)
		}

	default:
		panic(ExprError(p.tok))
	}
}

// ScanError is the error count parsing all expressions.
type ScanError int

func (ScanError) Error() string {
	return "aborting due to scanning errors"
}

// Parse parses expressions from r.
func Parse(r io.Reader) (e []any, err error) {
	var p parser
	p.s.Init(r)

	p.s.Mode = scanner.ScanIdents |
		scanner.ScanInts |
		scanner.ScanStrings |
		scanner.ScanRawStrings |
		scanner.ScanComments |
		scanner.SkipComments
	p.s.IsIdentRune = func(ch rune, i int) bool {
		if i == 0 && ch >= '0' && ch <= '9' {
			return false
		}
		return ch > 0 && ch < rune(len(idents)) && idents[ch]
	}

	defer func() {
		v := recover()
		if v == nil {
			return
		}

		_err, ok := v.(error)
		if !ok {
			panic(v)
		}

		if err == nil {
			err = _err
			return
		}
		err = errors.Join(err, _err)
	}()

	p.scan()
	for {
		expr, next := p.parseExpr()
		if expr == nil {
			break
		}
		e = append(e, expr)
		if !next {
			p.scan()
		}
	}

	if p.s.ErrorCount != 0 {
		err = ScanError(p.s.ErrorCount)
	}
	return
}