gigaparsec/bytes/regexp.go

package bytes

import (
	"errors"
	"fmt"
	"io"
	"regexp"
	"strings"
	"unicode/utf8"

	"git.codemonkeysoftware.net/b/gigaparsec"
	"git.codemonkeysoftware.net/b/gigaparsec/cursor"
)

// RuneReader is an io.RuneReader backed by a Cursor, for compatibility
// with the regexp package.
type RuneReader struct {
	cursor cursor.Cursor[byte]
	start  uint64
	err    error
}

func NewRuneReader(c cursor.Cursor[byte]) *RuneReader {
	return &RuneReader{cursor: c, start: c.Pos()}
}

func (rr *RuneReader) ReadRune() (r rune, size int, err error) {
	defer func() {
		rr.err = err
	}()
	var b [4]byte
	s := b[:]
	n, next, err := rr.cursor.Read(s)
	if err != nil && !errors.Is(err, io.EOF) {
		rr.cursor = next
		return 0, 0, fmt.Errorf("ReadRune: %w", err)
	}
	if n == 0 {
		return 0, 0, io.EOF
	}
	s = s[:n]
	r, size = utf8.DecodeRune(s)
	rr.cursor = rr.cursor.At(rr.cursor.Pos() + uint64(size))
	return r, size, nil
}

func (rr *RuneReader) Cursor() cursor.Cursor[byte] {
	return rr.cursor
}

func (rr *RuneReader) Error() error {
	return rr.err
}

func (rr *RuneReader) Count() uint64 {
	return rr.cursor.Pos() - rr.start
}

func Regexp(str string) gigaparsec.Parser[byte, string] {
	if !strings.HasPrefix(str, "^") && !strings.HasPrefix(str, `\A`) {
		str = "^" + str
	}
	re := regexp.MustCompile(str)
	expected := fmt.Sprintf("match `%s`", str)
	return func(input gigaparsec.State[byte]) (gigaparsec.Result[byte, string], error) {
		r := NewRuneReader(input.Cursor())
		idx := re.FindReaderIndex(r)
		err := r.Error()
		if err != nil && !errors.Is(err, io.EOF) {
			return gigaparsec.Result[byte, string]{}, fmt.Errorf("Regexp: reader error: %w", err)
		}
		if idx == nil {
			got := make([]byte, r.Count())
			_, _, err = input.Cursor().Read(got)
			if err != nil {
				return gigaparsec.Result[byte, string]{}, fmt.Errorf("Regexp: unexpected error: %w", err)
			}
			return gigaparsec.Fail[byte, string](false, gigaparsec.MakeMessage(input.Pos(), string(got), expected)), nil
		}
		// Alas, this is a little wasteful because a Regexp can only return indices
		// when searching a RuneReader.
		dst := make([]byte, idx[1]-idx[0])
		n, _, err := input.Cursor().Read(dst)
		if err != nil {
			return gigaparsec.Result[byte, string]{}, fmt.Errorf("Regexp: unexpected error: %w", err)
		}
		next := input.At(input.Pos() + n)
		return gigaparsec.Succeed(true, string(dst), next, gigaparsec.MessageOK(input.Pos())), nil
	}
}