1package message

2

3import (

4 "errors"

5 "fmt"

6 "strings"

7

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

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

10)

11

12var errBadMessageID = errors.New("not a message-id")

13

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

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

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

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

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

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

20// syntax.

21func MessageIDCanonical(s string) (string, bool, error) {

22 // ../rfc/5322:1383

23

24 s = strings.TrimSpace(s)

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

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

27 }

28 s = s[1:]

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

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

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

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

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

34 }

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

36 s = strings.ToLower(s)

37 if s == "" {

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

39 }

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

41 if err != nil {

42 // Common reasons for not being an address:

43 // 1. underscore in hostname.

44 // 2. ip literal instead of domain.

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

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

47 return s, true, nil

48 }

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

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

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

52 return s, false, nil

53}

54