1package message

2

3import (

4 "errors"

5 "fmt"

6 "strings"

7

8 "github.com/mjl-/mox/smtp"

9)

10

11var errBadMessageID = errors.New("not a message-id")

12

13// MessageIDCanonical parses the Message-ID, returning a canonical value that is

14// lower-cased, without <>, and no unneeded quoting. For matching in threading,

15// with References/In-Reply-To. If the message-id is invalid (e.g. no <>), an error

16// is returned. If the message-id could not be parsed as address (localpart "@"

17// domain), the raw value and the bool return parameter true is returned. It is

18// quite common that message-id's don't adhere to the localpart @ domain

19// syntax.

20func MessageIDCanonical(s string) (string, bool, error) {

21 // ../rfc/5322:1383

22

23 s = strings.TrimSpace(s)

24 if !strings.HasPrefix(s, "<") {

25 return "", false, fmt.Errorf("%w: missing <", errBadMessageID)

26 }

27 s = s[1:]

28 // Seen in practice: Message-ID: <valid@valid.example> (added by postmaster@some.example)

29 // Doesn't seem valid, but we allow it.

30 s, rem, have := strings.Cut(s, ">")

31 if !have || (rem != "" && (Pedantic || !strings.HasPrefix(rem, " "))) {

32 return "", false, fmt.Errorf("%w: missing >", errBadMessageID)

33 }

34 // We canonicalize the Message-ID: lower-case, no unneeded quoting.

35 s = strings.ToLower(s)

36 if s == "" {

37 return "", false, fmt.Errorf("%w: empty message-id", errBadMessageID)

38 }

39 addr, err := smtp.ParseAddress(s)

40 if err != nil {

41 // Common reasons for not being an address:

42 // 1. underscore in hostname.

43 // 2. ip literal instead of domain.

44 // 3. two @'s, perhaps intended as time-separator

45 // 4. no @'s, so no domain/host

46 return s, true, nil

47 }

48 // We preserve the unicode-ness of domain.

49 t := strings.Split(s, "@")

50 s = addr.Localpart.String() + "@" + t[len(t)-1]

51 return s, false, nil

52}

53