GODT-1077 IMAP sync counting

internal/imap/mailbox_message.go

@@ -226,7 +226,7 @@ func (im *imapMailbox) importMessage(m *pmapi.Message, readers []io.Reader, kr *
 	return im.storeMailbox.ImportMessage(m, body, labels)
 }
 
-func (im *imapMailbox) getMessage(storeMessage storeMessageProvider, items []imap.FetchItem) (msg *imap.Message, err error) {
+func (im *imapMailbox) getMessage(storeMessage storeMessageProvider, items []imap.FetchItem, msgBuildCountHistogram *msgBuildCountHistogram) (msg *imap.Message, err error) { //nolint[funlen]
 	im.log.WithField("msgID", storeMessage.ID()).Trace("Getting message")
 
 	seqNum, err := storeMessage.SequenceNumber()
@@ -268,7 +268,12 @@ func (im *imapMailbox) getMessage(storeMessage storeMessageProvider, items []ima
 			// on our part and we need to compute "real" size of decrypted data.
 			if m.Size <= 0 {
 				im.log.WithField("msgID", storeMessage.ID()).Trace("Size unknown - downloading body")
-				if _, _, err = im.getBodyAndStructure(storeMessage); err != nil {
+				// We are sure the size is not a problem right now. Clients
+				// might not first check sizes of all messages so we couldn't
+				// be sure if seeing 1st or 2nd sync is all right or not.
+				// Therefore, it's better to exclude getting size from the
+				// counting and see build count as real message build.
+				if _, _, err = im.getBodyAndStructure(storeMessage, nil); err != nil {
 					return
 				}
 			}
@@ -279,7 +284,7 @@ func (im *imapMailbox) getMessage(storeMessage storeMessageProvider, items []ima
 				return nil, err
 			}
 		default:
-			if err = im.getLiteralForSection(item, msg, storeMessage); err != nil {
+			if err = im.getLiteralForSection(item, msg, storeMessage, msgBuildCountHistogram); err != nil {
 				return
 			}
 		}
@@ -288,14 +293,14 @@ func (im *imapMailbox) getMessage(storeMessage storeMessageProvider, items []ima
 	return msg, err
 }
 
-func (im *imapMailbox) getLiteralForSection(itemSection imap.FetchItem, msg *imap.Message, storeMessage storeMessageProvider) error {
+func (im *imapMailbox) getLiteralForSection(itemSection imap.FetchItem, msg *imap.Message, storeMessage storeMessageProvider, msgBuildCountHistogram *msgBuildCountHistogram) error {
 	section, err := imap.ParseBodySectionName(itemSection)
 	if err != nil { // Ignore error
 		return nil
 	}
 
 	var literal imap.Literal
-	if literal, err = im.getMessageBodySection(storeMessage, section); err != nil {
+	if literal, err = im.getMessageBodySection(storeMessage, section, msgBuildCountHistogram); err != nil {
 		return err
 	}
 
@@ -313,14 +318,19 @@ func (im *imapMailbox) getBodyStructure(storeMessage storeMessageProvider) (bs *
 		im.log.WithError(err).Debug("Fail to retrieve bodystructure from database")
 	}
 	if bs == nil {
-		if bs, _, err = im.getBodyAndStructure(storeMessage); err != nil {
+		// We are sure the body structure is not a problem right now.
+		// Clients might do first fetch body structure so we couldn't
+		// be sure if seeing 1st or 2nd sync is all right or not.
+		// Therefore, it's better to exclude first body structure fetch
+		// from the counting and see build count as real message build.
+		if bs, _, err = im.getBodyAndStructure(storeMessage, nil); err != nil {
 			return
 		}
 	}
 	return
 }
 
-func (im *imapMailbox) getBodyAndStructure(storeMessage storeMessageProvider) (
+func (im *imapMailbox) getBodyAndStructure(storeMessage storeMessageProvider, msgBuildCountHistogram *msgBuildCountHistogram) (
 	structure *message.BodyStructure,
 	bodyReader *bytes.Reader, err error,
 ) {
@@ -352,6 +362,15 @@ func (im *imapMailbox) getBodyAndStructure(storeMessage storeMessageProvider) (
 					WithField("msgID", m.ID).
 					Warn("Cannot update header while building")
 			}
+			if msgBuildCountHistogram != nil {
+				times, err := storeMessage.IncreaseBuildCount()
+				if err != nil {
+					im.log.WithError(err).
+						WithField("msgID", m.ID).
+						Warn("Cannot increase build count")
+				}
+				msgBuildCountHistogram.add(times)
+			}
 			// Drafts can change and we don't want to cache them.
 			if !isMessageInDraftFolder(m) {
 				cache.SaveMail(id, body, structure)
@@ -379,7 +398,7 @@ func isMessageInDraftFolder(m *pmapi.Message) bool {
 
 // This will download message (or read from cache) and pick up the section,
 // extract data (header,body, both) and trim the output if needed.
-func (im *imapMailbox) getMessageBodySection(storeMessage storeMessageProvider, section *imap.BodySectionName) (literal imap.Literal, err error) { // nolint[funlen]
+func (im *imapMailbox) getMessageBodySection(storeMessage storeMessageProvider, section *imap.BodySectionName, msgBuildCountHistogram *msgBuildCountHistogram) (literal imap.Literal, err error) { // nolint[funlen]
 	var (
 		structure  *message.BodyStructure
 		bodyReader *bytes.Reader
@@ -410,7 +429,7 @@ func (im *imapMailbox) getMessageBodySection(storeMessage storeMessageProvider,
 		}
 	} else {
 		// The rest of cases need download and decrypt.
-		structure, bodyReader, err = im.getBodyAndStructure(storeMessage)
+		structure, bodyReader, err = im.getBodyAndStructure(storeMessage, msgBuildCountHistogram)
 		if err != nil {
 			return
 		}

internal/imap/mailbox_messages.go

@@ -482,12 +482,13 @@ func (im *imapMailbox) SearchMessages(isUID bool, criteria *imap.SearchCriteria)
 //
 // Messages must be sent to msgResponse. When the function returns, msgResponse must be closed.
 func (im *imapMailbox) ListMessages(isUID bool, seqSet *imap.SeqSet, items []imap.FetchItem, msgResponse chan<- *imap.Message) error {
+	msgBuildCountHistogram := newMsgBuildCountHistogram()
 	return im.logCommand(func() error {
-		return im.listMessages(isUID, seqSet, items, msgResponse)
-	}, "FETCH", isUID, seqSet, items)
+		return im.listMessages(isUID, seqSet, items, msgResponse, msgBuildCountHistogram)
+	}, "FETCH", isUID, seqSet, items, msgBuildCountHistogram)
 }
 
-func (im *imapMailbox) listMessages(isUID bool, seqSet *imap.SeqSet, items []imap.FetchItem, msgResponse chan<- *imap.Message) (err error) { //nolint[funlen]
+func (im *imapMailbox) listMessages(isUID bool, seqSet *imap.SeqSet, items []imap.FetchItem, msgResponse chan<- *imap.Message, msgBuildCountHistogram *msgBuildCountHistogram) (err error) { //nolint[funlen]
 	defer func() {
 		close(msgResponse)
 		if err != nil {
@@ -544,7 +545,7 @@ func (im *imapMailbox) listMessages(isUID bool, seqSet *imap.SeqSet, items []ima
 			return nil, err
 		}
 
-		msg, err := im.getMessage(storeMessage, items)
+		msg, err := im.getMessage(storeMessage, items, msgBuildCountHistogram)
 		if err != nil {
 			err = fmt.Errorf("list message build: %v", err)
 			l.WithField("metaID", storeMessage.ID()).Error(err)

internal/imap/msg_build_counts.go

@@ -0,0 +1,65 @@
+// Copyright (c) 2021 Proton Technologies AG
+//
+// This file is part of ProtonMail Bridge.
+//
+// ProtonMail Bridge is free software: you can redistribute it and/or modify
+// it under the terms of the GNU General Public License as published by
+// the Free Software Foundation, either version 3 of the License, or
+// (at your option) any later version.
+//
+// ProtonMail Bridge is distributed in the hope that it will be useful,
+// but WITHOUT ANY WARRANTY; without even the implied warranty of
+// MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+// GNU General Public License for more details.
+//
+// You should have received a copy of the GNU General Public License
+// along with ProtonMail Bridge.  If not, see <https://www.gnu.org/licenses/>.
+
+package imap
+
+import (
+	"fmt"
+	"sync"
+)
+
+// msgBuildCountHistogram is used to analyse and log the number of repetitive
+// downloads of requested messages per one fetch. The number of builds per each
+// messageID is stored in persistent database. The msgBuildCountHistogram will
+// take this number for each message in ongoing fetch and create histogram of
+// repeats.
+//
+// Example: During `fetch 1:300` there were
+// - 100 messages were downloaded first time
+// - 100 messages were downloaded second time
+// - 99 messages were downloaded 10th times
+// - 1 messages were downloaded 100th times
+type msgBuildCountHistogram struct {
+	// Key represents how many times message was build.
+	// Value stores how many messages are build X times based on the key.
+	counts map[uint32]uint32
+	lock   sync.Locker
+}
+
+func newMsgBuildCountHistogram() *msgBuildCountHistogram {
+	return &msgBuildCountHistogram{
+		counts: map[uint32]uint32{},
+		lock:   &sync.Mutex{},
+	}
+}
+
+func (c *msgBuildCountHistogram) String() string {
+	res := ""
+	for nRebuild, counts := range c.counts {
+		if res != "" {
+			res += ", "
+		}
+		res += fmt.Sprintf("[%d]:%d", nRebuild, counts)
+	}
+	return res
+}
+
+func (c *msgBuildCountHistogram) add(nRebuild uint32) {
+	c.lock.Lock()
+	defer c.lock.Unlock()
+	c.counts[nRebuild]++
+}

internal/imap/store.go

@@ -103,6 +103,7 @@ type storeMessageProvider interface {
 	SetContentTypeAndHeader(string, mail.Header) error
 	SetBodyStructure(*pkgMsg.BodyStructure) error
 	GetBodyStructure() (*pkgMsg.BodyStructure, error)
+	IncreaseBuildCount() (uint32, error)
 }
 
 type storeUserWrap struct {