mail-server/admin/messages.go

package main

import (
	"bytes"
	"encoding/base64"
	"errors"
	"fmt"
	"io"
	"io/fs"
	"mime"
	"mime/multipart"
	"mime/quotedprintable"
	"net/mail"
	"os"
	"path/filepath"
	"regexp"
	"sort"
	"strings"
	"time"
)

// errBadMailbox marks a mailbox identifier that fails shape validation; it is
// mapped to HTTP 400, never to a filesystem lookup.
var errBadMailbox = errors.New("invalid mailbox")

// mailboxRe accepts only the characters that can appear in a real mailbox
// address: lowercase letters, digits, dot, underscore, dash and one "@".
// Anything else (slashes, "..", uppercase, control chars) is rejected before
// it can reach the filesystem.
var mailboxRe = regexp.MustCompile(`^[a-z0-9._-]+@[a-z0-9.-]+$`)

// splitMailbox validates the full address shape and returns local/domain.
func splitMailbox(mailbox string) (local, domain string, err error) {
	if !mailboxRe.MatchString(mailbox) {
		return "", "", fmt.Errorf("%w: %q", errBadMailbox, mailbox)
	}
	local, domain, _ = strings.Cut(mailbox, "@")
	return local, domain, nil
}

// messageSummary is one row of the unified inbox list.
type messageSummary struct {
	Mailbox   string    `json:"mailbox"`
	Filename  string    `json:"filename"`
	MessageID string    `json:"message_id,omitempty"`
	From      string    `json:"from,omitempty"`
	To        string    `json:"to,omitempty"`
	Subject   string    `json:"subject,omitempty"`
	Date      time.Time `json:"date"`
	Size      int64     `json:"size"`
	Path      string    `json:"path"` // maildir-relative path hint
	when      time.Time // sort key (Date header or file mtime), not serialized
}

// messagesResponse is the JSON payload of GET /api/messages.
type messagesResponse struct {
	Updated  time.Time        `json:"updated"`
	Count    int              `json:"count"`
	Messages []messageSummary `json:"messages"`
}

// listMessages returns the unified inbox: every message file in cur/ and new/
// of every mailbox under root (data/mail-data/<domain>/<localpart>), newest
// first. mailboxFilter matches the full address or the bare localpart; query
// is a case-insensitive substring match on subject and from.
func listMessages(root, mailboxFilter, query string, limit int) ([]messageSummary, error) {
	filter := strings.ToLower(strings.TrimSpace(query))
	var out []messageSummary
	err := walkMailboxes(root, func(local, domain string) error {
		mailbox := local + "@" + domain
		if mailboxFilter != "" && mailbox != mailboxFilter && local != mailboxFilter {
			return nil
		}
		for _, sub := range []string{"cur", "new"} {
			dir := filepath.Join(root, domain, local, sub)
			entries, err := os.ReadDir(dir)
			if err != nil {
				if errors.Is(err, fs.ErrNotExist) {
					continue // no cur/ or new/ yet
				}
				return err
			}
			for _, e := range entries {
				if e.IsDir() {
					continue
				}
				info, err := e.Info()
				if err != nil {
					return err
				}
				if !info.Mode().IsRegular() {
					continue
				}
				path := filepath.Join(dir, e.Name())
				s, err := summarizeMessage(root, path, mailbox, e.Name(), sub, info)
				if err != nil {
					continue // unparseable file (dovecot index, lock): skip
				}
				if filter != "" &&
					!strings.Contains(strings.ToLower(s.Subject), filter) &&
					!strings.Contains(strings.ToLower(s.From), filter) {
					continue
				}
				out = append(out, s)
			}
		}
		return nil
	})
	if err != nil {
		return nil, err
	}
	sort.Slice(out, func(i, j int) bool { return out[i].when.After(out[j].when) })
	if limit > 0 && len(out) > limit {
		out = out[:limit]
	}
	return out, nil
}

// walkMailboxes visits every <domain>/<localpart> maildir under root.
func walkMailboxes(root string, fn func(local, domain string) error) error {
	domains, err := os.ReadDir(root)
	if err != nil {
		return err
	}
	for _, d := range domains {
		if !d.IsDir() {
			continue
		}
		users, err := os.ReadDir(filepath.Join(root, d.Name()))
		if err != nil {
			continue
		}
		for _, u := range users {
			if u.IsDir() {
				if err := fn(u.Name(), d.Name()); err != nil {
					return err
				}
			}
		}
	}
	return nil
}

// summarizeMessage reads the headers of one message file. Files that do not
// parse as RFC 5322 (empty, binary index leftovers) still appear in the list
// with filename and size only.
func summarizeMessage(root, path, mailbox, filename, sub string, info fs.FileInfo) (messageSummary, error) {
	s := messageSummary{
		Mailbox:  mailbox,
		Filename: filename,
		Size:     info.Size(),
		when:     info.ModTime(),
	}
	if rel, err := filepath.Rel(root, path); err == nil {
		s.Path = rel
	}
	f, err := os.Open(path)
	if err != nil {
		return s, err
	}
	defer f.Close()
	msg, err := mail.ReadMessage(f)
	if err != nil {
		return s, nil // not a message: keep the row, no headers
	}
	h := msg.Header
	s.MessageID = decodeHeader(h.Get("Message-Id"))
	s.From = headerAddress(h.Get("From"))
	s.To = headerAddress(h.Get("To"))
	s.Subject = decodeHeader(h.Get("Subject"))
	if d, err := mail.ParseDate(h.Get("Date")); err == nil {
		s.Date, s.when = d, d
	}
	return s, nil
}

// headerAddress extracts the first address from an RFC 5322 address header,
// falling back to the raw value when parsing fails.
func headerAddress(v string) string {
	v = strings.TrimSpace(v)
	if v == "" {
		return ""
	}
	if addrs, err := mail.ParseAddressList(v); err == nil && len(addrs) > 0 {
		return addrs[0].String()
	}
	return v
}

// attachmentInfo describes one MIME attachment part.
type attachmentInfo struct {
	Filename string `json:"filename,omitempty"`
	Size     int64  `json:"size"`
	CID      string `json:"cid,omitempty"`
}

// headerMap keeps header values per name (multiple values allowed, e.g.
// Received) with a convenience Getter like net/http.Header.
type headerMap map[string][]string

func (h headerMap) Get(key string) string {
	if v := h[key]; len(v) > 0 {
		return v[0]
	}
	return ""
}

// messageDetail is the JSON payload of GET /api/messages/<mailbox>/<filename>:
// decoded headers, the unfolded text/plain and text/html bodies and a list of
// attachments. Read-only: the underlying message file is never modified.
type messageDetail struct {
	Mailbox     string           `json:"mailbox"`
	Filename    string           `json:"filename"`
	Size        int64            `json:"size"`
	Headers     headerMap        `json:"headers"`
	Text        string           `json:"text"`
	HTML        string           `json:"html"`
	Attachments []attachmentInfo `json:"attachments"`
}

// Subject, From and To expose decoded display values from the header map
// (headers are stored decoded by parseMessage).
func (d messageDetail) Subject() string { return d.Headers.Get("Subject") }
func (d messageDetail) From() string    { return headerAddress(d.Headers.Get("From")) }
func (d messageDetail) To() string      { return headerAddress(d.Headers.Get("To")) }

// readMessage resolves mailbox+filename against the real listing (a request
// filename can never build a path of its own) and parses the message. It
// returns fs.ErrNotExist when the file is not in cur/ or new/ of the mailbox.
func readMessage(root, mailbox, filename string) (messageDetail, error) {
	local, domain, err := splitMailbox(mailbox)
	if err != nil {
		return messageDetail{}, err
	}
	full := ""
	for _, sub := range []string{"cur", "new"} {
		dir := filepath.Join(root, domain, local, sub)
		entries, err := os.ReadDir(dir)
		if err != nil {
			if errors.Is(err, fs.ErrNotExist) {
				continue
			}
			return messageDetail{}, err
		}
		for _, e := range entries {
			if !e.IsDir() && e.Name() == filename {
				full = filepath.Join(dir, e.Name())
				break
			}
		}
	}
	if full == "" {
		return messageDetail{}, os.ErrNotExist
	}
	info, err := os.Stat(full)
	if err != nil {
		return messageDetail{}, err
	}
	data, err := os.ReadFile(full)
	if err != nil {
		return messageDetail{}, err
	}
	d, err := parseMessage(data, mailbox, filename)
	if err != nil {
		return messageDetail{}, err
	}
	d.Size = info.Size()
	return d, nil
}

// parseMessage splits a raw message into decoded headers, text/html bodies
// and attachment metadata. multipart bodies are unfolded recursively.
func parseMessage(data []byte, mailbox, filename string) (messageDetail, error) {
	msg, err := mail.ReadMessage(bytes.NewReader(data))
	if err != nil {
		return messageDetail{}, err
	}
	d := messageDetail{
		Mailbox:  mailbox,
		Filename: filename,
		Headers:  headerMap{},
	}
	for k, vs := range msg.Header {
		decoded := make([]string, len(vs))
		for i, v := range vs {
			decoded[i] = decodeHeader(v)
		}
		d.Headers[k] = decoded
	}
	mediatype, params, _ := mime.ParseMediaType(msg.Header.Get("Content-Type"))
	if mediatype == "" {
		mediatype = "text/plain"
	}
	if strings.HasPrefix(mediatype, "multipart/") {
		if err := parseMultipart(msg.Body, params["boundary"], &d); err != nil {
			return messageDetail{}, err
		}
		return d, nil
	}
	body, err := io.ReadAll(decodeTransfer(msg.Header.Get("Content-Transfer-Encoding"), msg.Body))
	if err != nil {
		return messageDetail{}, err
	}
	addPart(msg.Header, mediatype, body, &d)
	return d, nil
}

// parseMultipart walks one multipart/* body, recursing into nested
// multipart parts (mixed/alternative/related).
func parseMultipart(r io.Reader, boundary string, d *messageDetail) error {
	mr := multipart.NewReader(r, boundary)
	for {
		part, err := mr.NextPart()
		if err == io.EOF {
			return nil
		}
		if err != nil {
			return err
		}
		mediatype, params, _ := mime.ParseMediaType(part.Header.Get("Content-Type"))
		if mediatype == "" {
			mediatype = "text/plain"
		}
		if strings.HasPrefix(mediatype, "multipart/") {
			if err := parseMultipart(part, params["boundary"], d); err != nil {
				return err
			}
			part.Close()
			continue
		}
		data, err := io.ReadAll(decodeTransfer(part.Header.Get("Content-Transfer-Encoding"), part))
		if err != nil {
			part.Close()
			return err
		}
		addPart(part.Header, mediatype, data, d)
		part.Close()
	}
}

// addPart dispatches one leaf MIME part: text/plain and text/html feed the
// body views, everything else becomes an attachment entry.
func addPart(h interface{ Get(string) string }, mediatype string, data []byte, d *messageDetail) {
	switch mediatype {
	case "text/plain":
		d.Text += string(data)
	case "text/html":
		d.HTML += string(data)
	default:
		a := attachmentInfo{Size: int64(len(data))}
		if cd := h.Get("Content-Disposition"); cd != "" {
			if _, p, err := mime.ParseMediaType(cd); err == nil {
				a.Filename = p["filename"]
			}
		}
		if a.Filename == "" {
			if _, p, err := mime.ParseMediaType(h.Get("Content-Type")); err == nil {
				a.Filename = p["name"]
			}
		}
		if cid := h.Get("Content-Id"); cid != "" {
			a.CID = strings.Trim(cid, "<>")
		}
		d.Attachments = append(d.Attachments, a)
	}
}

// decodeTransfer wraps a part body with the decoder for its
// Content-Transfer-Encoding; base64 and quoted-printable are the only
// encodings that need one, everything else is read verbatim.
func decodeTransfer(enc string, r io.Reader) io.Reader {
	switch strings.ToLower(strings.TrimSpace(enc)) {
	case "base64":
		return base64.NewDecoder(base64.StdEncoding, r)
	case "quoted-printable":
		return quotedprintable.NewReader(r)
	default:
		return r
	}
}

// wordDecoder decodes RFC 2047 encoded words in header values.
var wordDecoder = &mime.WordDecoder{}

func decodeHeader(v string) string {
	if v == "" {
		return ""
	}
	if d, err := wordDecoder.DecodeHeader(v); err == nil {
		return d
	}
	return v
}