10 cryptrand "crypto/rand"
25 "golang.org/x/text/unicode/norm"
27 "github.com/mjl-/bstore"
29 "github.com/mjl-/mox/config"
30 "github.com/mjl-/mox/message"
31 "github.com/mjl-/mox/metrics"
32 "github.com/mjl-/mox/mlog"
33 "github.com/mjl-/mox/mox-"
34 "github.com/mjl-/mox/store"
37type importListener struct {
39 Events chan importEvent
40 Register chan bool // Whether register is successful.
43type importEvent struct {
45 SSEMsg []byte // Full SSE message, including event: ... and data: ... \n\n
46 Event any // nil, importCount, importProblem, importDone, importAborted
47 Cancel func() // For cancelling the context causing abort of the import. Set in first, import-registering, event.
50type importAbortRequest struct {
55var importers = struct {
56 Register chan *importListener
57 Unregister chan *importListener
58 Events chan importEvent
59 Abort chan importAbortRequest
62 make(chan *importListener, 1),
63 make(chan *importListener, 1),
64 make(chan importEvent),
65 make(chan importAbortRequest),
69// ImportManage should be run as a goroutine, it manages imports of mboxes/maildirs, propagating progress over SSE connections.
71 log := mlog.New("httpimport", nil)
73 if x := recover(); x != nil {
74 log.Error("import manage panic", slog.Any("err", x))
76 metrics.PanicInc(metrics.Importmanage)
81 MailboxCounts map[string]int
85 Listeners map[*importListener]struct{}
89 imports := map[string]state{} // Token to state.
92 case l := <-importers.Register:
93 // If we have state, send it so the client is up to date.
94 s, ok := imports[l.Token]
99 s.Listeners[l] = struct{}{}
101 sendEvent := func(kind string, v any) {
102 buf, err := json.Marshal(v)
104 log.Errorx("marshal event", err, slog.String("kind", kind), slog.Any("event", v))
107 ssemsg := fmt.Sprintf("event: %s\ndata: %s\n\n", kind, buf)
110 case l.Events <- importEvent{kind, []byte(ssemsg), nil, nil}:
112 log.Debug("dropped initial import event to slow consumer")
116 for m, c := range s.MailboxCounts {
117 sendEvent("count", importCount{m, c})
119 for _, p := range s.Problems {
120 sendEvent("problem", importProblem{p})
123 sendEvent("done", importDone{})
124 } else if s.Aborted != nil {
125 sendEvent("aborted", importAborted{})
128 case l := <-importers.Unregister:
129 delete(imports[l.Token].Listeners, l)
131 case e := <-importers.Events:
132 s, ok := imports[e.Token]
135 MailboxCounts: map[string]int{},
136 Listeners: map[*importListener]struct{}{},
141 for l := range s.Listeners {
145 log.Debug("dropped import event to slow consumer")
149 s := imports[e.Token]
150 switch x := e.Event.(type) {
152 s.MailboxCounts[x.Mailbox] = x.Count
154 s.Problems = append(s.Problems, x.Message)
165 case a := <-importers.Abort:
166 s, ok := imports[a.Token]
168 a.Response <- errors.New("import not found")
172 a.Response <- errors.New("import already finished")
178 case <-importers.Stop:
182 // Cleanup old state.
183 for t, s := range imports {
184 if len(s.Listeners) > 0 {
187 if s.Done != nil && time.Since(*s.Done) > time.Minute || s.Aborted != nil && time.Since(*s.Aborted) > time.Minute {
194type importCount struct {
198type importProblem struct {
201type importDone struct{}
202type importAborted struct{}
203type importStep struct {
207// importStart prepare the import and launches the goroutine to actually import.
208// importStart is responsible for closing f and removing f.
209func importStart(log mlog.Log, accName string, f *os.File, skipMailboxPrefix string) (string, bool, error) {
212 store.CloseRemoveTempFile(log, f, "upload for import")
217 cryptrand.Read(buf[:])
218 token := fmt.Sprintf("%x", buf)
220 if _, err := f.Seek(0, 0); err != nil {
221 return "", false, fmt.Errorf("seek to start of file: %v", err)
224 // Recognize file format.
226 magicZip := []byte{0x50, 0x4b, 0x03, 0x04}
227 magicGzip := []byte{0x1f, 0x8b}
228 magic := make([]byte, 4)
229 if _, err := f.ReadAt(magic, 0); err != nil {
230 return "", true, fmt.Errorf("detecting file format: %v", err)
232 if bytes.Equal(magic, magicZip) {
234 } else if !bytes.Equal(magic[:2], magicGzip) {
235 return "", true, fmt.Errorf("file is not a zip or gzip file")
243 return "", false, fmt.Errorf("stat temporary import zip file: %v", err)
245 zr, err = zip.NewReader(f, fi.Size())
247 return "", true, fmt.Errorf("opening zip file: %v", err)
250 gzr, err := gzip.NewReader(f)
252 return "", true, fmt.Errorf("gunzip: %v", err)
254 tr = tar.NewReader(gzr)
257 acc, err := store.OpenAccount(log, accName, false)
259 return "", false, fmt.Errorf("open acount: %v", err)
261 acc.Lock() // Not using WithWLock because importMessage is responsible for unlocking.
263 tx, err := acc.DB.Begin(context.Background(), true)
267 log.Check(xerr, "closing account")
268 return "", false, fmt.Errorf("start transaction: %v", err)
271 // Ensure token is registered before returning, with context that can be canceled.
272 ctx, cancel := context.WithCancel(mox.Shutdown)
273 importers.Events <- importEvent{token, []byte(": keepalive\n\n"), nil, cancel}
275 log.Info("starting import")
276 go importMessages(ctx, log.WithCid(mox.Cid()), token, acc, tx, zr, tr, f, skipMailboxPrefix)
277 f = nil // importMessages is now responsible for closing and removing.
279 return token, false, nil
282// importMessages imports the messages from zip/tgz file f.
283// importMessages is responsible for unlocking and closing acc, and closing tx and f.
284func importMessages(ctx context.Context, log mlog.Log, token string, acc *store.Account, tx *bstore.Tx, zr *zip.Reader, tr *tar.Reader, f *os.File, skipMailboxPrefix string) {
285 // If a fatal processing error occurs, we panic with this type.
286 type importError struct{ Err error }
288 // During import we collect all changes and broadcast them at the end, when successful.
289 var changes []store.Change
291 // ID's of delivered messages. If we have to rollback, we have to remove this files.
294 sendEvent := func(kind string, v any) {
295 buf, err := json.Marshal(v)
297 log.Errorx("marshal event", err, slog.String("kind", kind), slog.Any("event", v))
300 ssemsg := fmt.Sprintf("event: %s\ndata: %s\n\n", kind, buf)
301 importers.Events <- importEvent{token, []byte(ssemsg), v, nil}
304 canceled := func() bool {
307 sendEvent("aborted", importAborted{})
314 problemf := func(format string, args ...any) {
315 msg := fmt.Sprintf(format, args...)
316 sendEvent("problem", importProblem{Message: msg})
320 store.CloseRemoveTempFile(log, f, "uploaded messages")
322 for _, id := range newIDs {
323 p := acc.MessagePath(id)
325 log.Check(err, "closing message file after import error", slog.String("path", p))
329 log.Check(err, "rolling back transaction")
334 log.Check(err, "closing account")
341 if err, ok := x.(importError); ok {
342 log.Errorx("import error", err.Err)
343 problemf("%s (aborting)", err.Err)
344 sendEvent("aborted", importAborted{})
346 log.Error("import panic", slog.Any("err", x))
348 metrics.PanicInc(metrics.Importmessages)
352 ximportcheckf := func(err error, format string, args ...any) {
354 panic(importError{fmt.Errorf("%s: %s", fmt.Sprintf(format, args...), err)})
358 err := acc.ThreadingWait(log)
359 ximportcheckf(err, "waiting for account thread upgrade")
361 conf, _ := acc.Conf()
363 jf, _, err := acc.OpenJunkFilter(ctx, log)
364 if err != nil && !errors.Is(err, store.ErrNoJunkFilter) {
365 ximportcheckf(err, "open junk filter")
369 err := jf.CloseDiscard()
370 log.Check(err, "closing junk filter")
374 // Mailboxes we imported, and message counts.
375 mailboxNames := map[string]*store.Mailbox{}
376 mailboxIDs := map[int64]*store.Mailbox{}
377 mailboxKeywordCounts := map[int64]int{}
378 messages := map[string]int{}
380 maxSize := acc.QuotaMessageSize()
381 du := store.DiskUsage{ID: 1}
383 ximportcheckf(err, "get disk usage")
386 // For maildirs, we are likely to get a possible dovecot-keywords file after having
387 // imported the messages. Once we see the keywords, we use them. But before that
388 // time we remember which messages miss keywords. Once the keywords become
389 // available, we'll fix up the flags for the unknown messages
390 mailboxKeywords := map[string]map[rune]string{} // Mailbox to 'a'-'z' to flag name.
391 mailboxMissingKeywordMessages := map[string]map[int64]string{} // Mailbox to message id to string consisting of the unrecognized flags.
393 // Previous mailbox an event was sent for. We send an event for new mailboxes, when
394 // another 100 messages were added, when adding a message to another mailbox, and
395 // finally at the end as a closing statement.
396 var prevMailbox string
398 var modseq store.ModSeq // Assigned on first message, used for all messages.
400 trainMessage := func(m *store.Message, p message.Part, pos string) {
401 words, err := jf.ParseMessage(p)
403 problemf("parsing message %s for updating junk filter: %v (continuing)", pos, err)
406 err = jf.Train(ctx, !m.Junk, words)
408 problemf("training junk filter for message %s: %v (continuing)", pos, err)
411 m.TrainedJunk = &m.Junk
414 openTrainMessage := func(m *store.Message) {
415 path := acc.MessagePath(m.ID)
416 f, err := os.Open(path)
418 problemf("opening message again for training junk filter: %v (continuing)", err)
423 log.Check(err, "closing file after training junkfilter")
425 p, err := m.LoadPart(f)
427 problemf("loading parsed message again for training junk filter: %v (continuing)", err)
430 trainMessage(m, p, fmt.Sprintf("message id %d", m.ID))
433 xensureMailbox := func(name string) *store.Mailbox {
434 // Ensure name is normalized.
435 name = norm.NFC.String(name)
436 name, _, err := config.CheckMailboxName(name, true)
437 ximportcheckf(err, "checking mailbox name")
439 if mb, ok := mailboxNames[name]; ok {
444 var mb *store.Mailbox
445 var parent store.Mailbox
446 for i, e := range strings.Split(name, "/") {
452 if _, ok := mailboxNames[p]; ok {
456 mb, err = acc.MailboxFind(tx, p)
457 ximportcheckf(err, "looking up mailbox %s to import to (aborting)", p)
459 uidvalidity, err := acc.NextUIDValidity(tx)
460 ximportcheckf(err, "finding next uid validity")
464 modseq, err = acc.NextModSeq(tx)
465 ximportcheckf(err, "assigning next modseq")
473 UIDValidity: uidvalidity,
476 // Do not assign special-use flags. This existing account probably already has such mailboxes.
479 ximportcheckf(err, "inserting mailbox in database")
482 if tx.Get(&store.Subscription{Name: p}) != nil {
483 err := tx.Insert(&store.Subscription{Name: p})
484 ximportcheckf(err, "subscribing to imported mailbox")
486 changes = append(changes, store.ChangeAddMailbox{Mailbox: *mb, Flags: []string{`\Subscribed`}})
488 if prevMailbox != "" && mb.Name != prevMailbox {
489 sendEvent("count", importCount{prevMailbox, messages[prevMailbox]})
491 mailboxKeywordCounts[mb.ID] = len(mb.Keywords)
492 mailboxNames[mb.Name] = mb
493 mailboxIDs[mb.ID] = mb
494 sendEvent("count", importCount{mb.Name, 0})
495 prevMailbox = mb.Name
500 xdeliver := func(mb *store.Mailbox, m *store.Message, f *os.File, pos string) {
501 defer store.CloseRemoveTempFile(log, f, "message file for import")
503 m.MailboxOrigID = mb.ID
506 if maxSize > 0 && du.MessageSize+addSize > maxSize {
507 ximportcheckf(fmt.Errorf("account over maximum total size %d", maxSize), "checking quota")
512 modseq, err = acc.NextModSeq(tx)
513 ximportcheckf(err, "assigning next modseq")
518 // Parse message and store parsed information for later fast retrieval.
519 p, err := message.EnsurePart(log.Logger, false, f, m.Size)
521 problemf("parsing message %s: %s (continuing)", pos, err)
523 m.ParsedBuf, err = json.Marshal(p)
524 ximportcheckf(err, "marshal parsed message structure")
526 // Set fields needed for future threading. By doing it now, MessageAdd won't
527 // have to parse the Part again.
528 p.SetReaderAt(store.FileMsgReader(m.MsgPrefix, f))
529 m.PrepareThreading(log, &p)
531 if m.Received.IsZero() {
532 if p.Envelope != nil && !p.Envelope.Date.IsZero() {
533 m.Received = p.Envelope.Date
535 m.Received = time.Now()
539 // We set the flags that Deliver would set now and train ourselves. This prevents
540 // Deliver from training, which would open the junk filter, change it, and write it
541 // back to disk, for each message (slow).
542 m.JunkFlagsForMailbox(*mb, conf)
543 if jf != nil && m.NeedsTraining() {
544 trainMessage(m, p, pos)
547 opts := store.AddOpts{
551 SkipUpdateDiskUsage: true,
552 SkipCheckQuota: true,
555 if err := acc.MessageAdd(log, tx, mb, m, f, opts); err != nil {
556 problemf("delivering message %s: %s (continuing)", pos, err)
559 newIDs = append(newIDs, m.ID)
560 changes = append(changes, m.ChangeAddUID(*mb))
562 if messages[mb.Name]%100 == 0 || prevMailbox != mb.Name {
563 prevMailbox = mb.Name
564 sendEvent("count", importCount{mb.Name, messages[mb.Name]})
568 ximportMbox := func(mailbox, filename string, r io.Reader) {
570 problemf("empty mailbox name for mbox file %s (skipping)", filename)
573 mb := xensureMailbox(mailbox)
575 mr := store.NewMboxReader(log, store.CreateMessageTemp, filename, r)
577 m, mf, pos, err := mr.Next()
580 } else if err != nil {
581 ximportcheckf(err, "next message in mbox file")
584 xdeliver(mb, m, mf, pos)
588 ximportMaildir := func(mailbox, filename string, r io.Reader) {
590 problemf("empty mailbox name for maildir file %s (skipping)", filename)
593 mb := xensureMailbox(mailbox)
595 f, err := store.CreateMessageTemp(log, "import")
596 ximportcheckf(err, "creating temp message")
599 store.CloseRemoveTempFile(log, f, "message to import")
603 // Copy data, changing bare \n into \r\n.
604 br := bufio.NewReader(r)
605 w := bufio.NewWriter(f)
608 line, err := br.ReadBytes('\n')
609 if err != nil && err != io.EOF {
610 ximportcheckf(err, "reading message")
613 if !bytes.HasSuffix(line, []byte("\r\n")) {
614 line = append(line[:len(line)-1], "\r\n"...)
617 n, err := w.Write(line)
618 ximportcheckf(err, "writing message")
626 ximportcheckf(err, "writing message")
628 var received time.Time
629 t := strings.SplitN(path.Base(filename), ".", 2)
630 if v, err := strconv.ParseInt(t[0], 10, 64); err == nil {
631 received = time.Unix(v, 0)
634 // Parse flags. See https://cr.yp.to/proto/maildir.html.
635 var keepFlags strings.Builder
636 var flags store.Flags
637 keywords := map[string]bool{}
638 t = strings.SplitN(path.Base(filename), ":2,", 2)
640 for _, c := range t[1] {
643 // Passed, doesn't map to a common IMAP flag.
645 flags.Answered = true
655 if c >= 'a' && c <= 'z' {
656 dovecotKeywords, ok := mailboxKeywords[mailbox]
658 // No keywords file seen yet, we'll try later if it comes in.
659 keepFlags.WriteString(string(c))
660 } else if kw, ok := dovecotKeywords[c]; ok {
661 flagSet(&flags, keywords, kw)
671 Keywords: slices.Sorted(maps.Keys(keywords)),
674 xdeliver(mb, &m, f, filename)
676 if keepFlags.String() != "" {
677 if _, ok := mailboxMissingKeywordMessages[mailbox]; !ok {
678 mailboxMissingKeywordMessages[mailbox] = map[int64]string{}
680 mailboxMissingKeywordMessages[mailbox][m.ID] = keepFlags.String()
684 importFile := func(name string, r io.Reader) {
687 if strings.HasPrefix(name, skipMailboxPrefix) {
688 name = strings.TrimPrefix(name[len(skipMailboxPrefix):], "/")
691 if before, ok := strings.CutSuffix(name, "/"); ok {
693 dir := path.Dir(name)
694 switch path.Base(dir) {
695 case "new", "cur", "tmp":
696 // Maildir, ensure it exists.
697 mailbox := path.Dir(dir)
698 xensureMailbox(mailbox)
700 // Otherwise, this is just a directory that probably holds mbox files and maildirs.
704 if strings.HasSuffix(path.Base(name), ".mbox") {
705 mailbox := name[:len(name)-len(".mbox")]
706 ximportMbox(mailbox, origName, r)
709 dir := path.Dir(name)
710 dirbase := path.Base(dir)
712 case "new", "cur", "tmp":
713 mailbox := path.Dir(dir)
714 ximportMaildir(mailbox, origName, r)
718 if path.Base(name) != "dovecot-keywords" {
719 problemf("unrecognized file %s (skipping)", origName)
723 // Handle dovecot-keywords.
724 mailbox := path.Dir(name)
725 dovecotKeywords := map[rune]string{}
726 words, err := store.ParseDovecotKeywordsFlags(r, log)
727 log.Check(err, "parsing dovecot keywords for mailbox", slog.String("mailbox", mailbox))
728 for i, kw := range words {
729 dovecotKeywords['a'+rune(i)] = kw
731 mailboxKeywords[mailbox] = dovecotKeywords
733 for id, chars := range mailboxMissingKeywordMessages[mailbox] {
734 var flags, zeroflags store.Flags
735 keywords := map[string]bool{}
736 for _, c := range chars {
737 kw, ok := dovecotKeywords[c]
739 problemf("unspecified dovecot message flag %c for message id %d (continuing)", c, id)
742 flagSet(&flags, keywords, kw)
744 if flags == zeroflags && len(keywords) == 0 {
748 m := store.Message{ID: id}
750 ximportcheckf(err, "get imported message for flag update")
752 mb := mailboxIDs[m.MailboxID]
753 mb.Sub(m.MailboxCounts())
756 m.Flags = m.Flags.Set(flags, flags)
757 m.Keywords = slices.Sorted(maps.Keys(keywords))
759 mb.Add(m.MailboxCounts())
761 mb.Keywords, _ = store.MergeKeywords(mb.Keywords, m.Keywords)
763 // We train before updating, training may set m.TrainedJunk.
764 if jf != nil && m.NeedsTraining() {
768 ximportcheckf(err, "updating message after flag update")
769 changes = append(changes, m.ChangeFlags(oflags, *mb))
771 delete(mailboxMissingKeywordMessages, mailbox)
775 for _, f := range zr.File {
781 problemf("opening file %s in zip: %v", f.Name, err)
784 importFile(f.Name, zf)
786 log.Check(err, "closing file from zip")
796 } else if err != nil {
797 problemf("reading next tar header: %v (aborting)", err)
800 importFile(h.Name, tr)
805 for _, count := range messages {
808 log.Debug("messages imported", slog.Int("total", total))
810 // Send final update for count of last-imported mailbox.
811 if prevMailbox != "" {
812 sendEvent("count", importCount{prevMailbox, messages[prevMailbox]})
817 sendEvent("step", importStep{"matching messages with threads"})
818 err = acc.AssignThreads(ctx, log, tx, newIDs[0], 0, io.Discard)
819 ximportcheckf(err, "assigning messages to threads")
822 // Update mailboxes with counts and keywords.
823 for _, mb := range mailboxIDs {
825 ximportcheckf(err, "updating mailbox count and keywords")
827 changes = append(changes, mb.ChangeCounts())
828 if len(mb.Keywords) != mailboxKeywordCounts[mb.ID] {
829 changes = append(changes, mb.ChangeKeywords())
833 err = acc.AddMessageSize(log, tx, addSize)
834 ximportcheckf(err, "updating disk usage after import")
838 ximportcheckf(err, "commit")
842 if err := jf.Close(); err != nil {
843 problemf("saving changes of training junk filter: %v (continuing)", err)
844 log.Errorx("saving changes of training junk filter", err)
849 store.BroadcastChanges(acc, changes)
852 log.Check(err, "closing account after import")
855 sendEvent("done", importDone{})
858func flagSet(flags *store.Flags, keywords map[string]bool, word string) {
860 case "forwarded", "$forwarded":
861 flags.Forwarded = true
862 case "junk", "$junk":
864 case "notjunk", "$notjunk", "nonjunk", "$nonjunk":
866 case "phishing", "$phishing":
867 flags.Phishing = true
868 case "mdnsent", "$mdnsent":
871 if err := store.CheckKeyword(word); err == nil {
872 keywords[word] = true