add webmail
it was far down on the roadmap, but implemented earlier, because it's
interesting, and to help prepare for a jmap implementation. for jmap we need to
implement more client-like functionality than with just imap. internal data
structures need to change. jmap has lots of other requirements, so it's already
a big project. by implementing a webmail now, some of the required data
structure changes become clear and can be made now, so the later jmap
implementation can do things similarly to the webmail code. the webmail
frontend and webmail are written together, making their interface/api much
smaller and simpler than jmap.
one of the internal changes is that we now keep track of per-mailbox
total/unread/unseen/deleted message counts and mailbox sizes. keeping this
data consistent after any change to the stored messages (through the code base)
is tricky, so mox now has a consistency check that verifies the counts are
correct, which runs only during tests, each time an internal account reference
is closed. we have a few more internal "changes" that are propagated for the
webmail frontend (that imap doesn't have a way to propagate on a connection),
like changes to the special-use flags on mailboxes, and used keywords in a
mailbox. more changes that will be required have revealed themselves while
implementing the webmail, and will be implemented next.
the webmail user interface is modeled after the mail clients i use or have
used: thunderbird, macos mail, mutt; and webmails i normally only use for
testing: gmail, proton, yahoo, outlook. a somewhat technical user is assumed,
but still the goal is to make this webmail client easy to use for everyone. the
user interface looks like most other mail clients: a list of mailboxes, a
search bar, a message list view, and message details. there is a top/bottom and
a left/right layout for the list/message view, default is automatic based on
screen size. the panes can be resized by the user. buttons for actions are just
text, not icons. clicking a button briefly shows the shortcut for the action in
the bottom right, helping with learning to operate quickly. any text that is
underdotted has a title attribute that causes more information to be displayed,
e.g. what a button does or a field is about. to highlight potential phishing
attempts, any text (anywhere in the webclient) that switches unicode "blocks"
(a rough approximation to (language) scripts) within a word is underlined
orange. multiple messages can be selected with familiar ui interaction:
clicking while holding control and/or shift keys. keyboard navigation works
with arrows/page up/down and home/end keys, and also with a few basic vi-like
keys for list/message navigation. we prefer showing the text instead of
html (with inlined images only) version of a message. html messages are shown
in an iframe served from an endpoint with CSP headers to prevent dangerous
resources (scripts, external images) from being loaded. the html is also
sanitized, with javascript removed. a user can choose to load external
resources (e.g. images for tracking purposes).
the frontend is just (strict) typescript, no external frameworks. all
incoming/outgoing data is typechecked, both the api request parameters and
response types, and the data coming in over SSE. the types and checking code
are generated with sherpats, which uses the api definitions generated by
sherpadoc based on the Go code. so types from the backend are automatically
propagated to the frontend. since there is no framework to automatically
propagate properties and rerender components, changes coming in over the SSE
connection are propagated explicitly with regular function calls. the ui is
separated into "views", each with a "root" dom element that is added to the
visible document. these views have additional functions for getting changes
propagated, often resulting in the view updating its (internal) ui state (dom).
we keep the frontend compilation simple, it's just a few typescript files that
get compiled (combined and types stripped) into a single js file, no additional
runtime code needed or complicated build processes used. the webmail is served
is served from a compressed, cachable html file that includes style and the
javascript, currently just over 225kb uncompressed, under 60kb compressed (not
minified, including comments). we include the generated js files in the
repository, to keep Go's easily buildable self-contained binaries.
authentication is basic http, as with the account and admin pages. most data
comes in over one long-term SSE connection to the backend. api requests signal
which mailbox/search/messages are requested over the SSE connection. fetching
individual messages, and making changes, are done through api calls. the
operations are similar to imap, so some code has been moved from package
imapserver to package store. the future jmap implementation will benefit from
these changes too. more functionality will probably be moved to the store
package in the future.
the quickstart enables webmail on the internal listener by default (for new
installs). users can enable it on the public listener if they want to. mox
localserve enables it too. to enable webmail on existing installs, add settings
like the following to the listeners in mox.conf, similar to AccountHTTP(S):
WebmailHTTP:
Enabled: true
WebmailHTTPS:
Enabled: true
special thanks to liesbeth, gerben, andrii for early user feedback.
there is plenty still to do, see the list at the top of webmail/webmail.ts.
feedback welcome as always.
2023-08-07 22:57:03 +03:00
|
|
|
package webaccount
|
2023-02-16 11:57:27 +03:00
|
|
|
|
|
|
|
import (
|
|
|
|
"archive/tar"
|
|
|
|
"archive/zip"
|
|
|
|
"bufio"
|
|
|
|
"bytes"
|
|
|
|
"compress/gzip"
|
|
|
|
"context"
|
|
|
|
cryptrand "crypto/rand"
|
|
|
|
"encoding/json"
|
|
|
|
"errors"
|
|
|
|
"fmt"
|
|
|
|
"io"
|
2024-02-08 16:49:01 +03:00
|
|
|
"log/slog"
|
2023-02-16 11:57:27 +03:00
|
|
|
"os"
|
|
|
|
"path"
|
|
|
|
"runtime/debug"
|
add webmail
it was far down on the roadmap, but implemented earlier, because it's
interesting, and to help prepare for a jmap implementation. for jmap we need to
implement more client-like functionality than with just imap. internal data
structures need to change. jmap has lots of other requirements, so it's already
a big project. by implementing a webmail now, some of the required data
structure changes become clear and can be made now, so the later jmap
implementation can do things similarly to the webmail code. the webmail
frontend and webmail are written together, making their interface/api much
smaller and simpler than jmap.
one of the internal changes is that we now keep track of per-mailbox
total/unread/unseen/deleted message counts and mailbox sizes. keeping this
data consistent after any change to the stored messages (through the code base)
is tricky, so mox now has a consistency check that verifies the counts are
correct, which runs only during tests, each time an internal account reference
is closed. we have a few more internal "changes" that are propagated for the
webmail frontend (that imap doesn't have a way to propagate on a connection),
like changes to the special-use flags on mailboxes, and used keywords in a
mailbox. more changes that will be required have revealed themselves while
implementing the webmail, and will be implemented next.
the webmail user interface is modeled after the mail clients i use or have
used: thunderbird, macos mail, mutt; and webmails i normally only use for
testing: gmail, proton, yahoo, outlook. a somewhat technical user is assumed,
but still the goal is to make this webmail client easy to use for everyone. the
user interface looks like most other mail clients: a list of mailboxes, a
search bar, a message list view, and message details. there is a top/bottom and
a left/right layout for the list/message view, default is automatic based on
screen size. the panes can be resized by the user. buttons for actions are just
text, not icons. clicking a button briefly shows the shortcut for the action in
the bottom right, helping with learning to operate quickly. any text that is
underdotted has a title attribute that causes more information to be displayed,
e.g. what a button does or a field is about. to highlight potential phishing
attempts, any text (anywhere in the webclient) that switches unicode "blocks"
(a rough approximation to (language) scripts) within a word is underlined
orange. multiple messages can be selected with familiar ui interaction:
clicking while holding control and/or shift keys. keyboard navigation works
with arrows/page up/down and home/end keys, and also with a few basic vi-like
keys for list/message navigation. we prefer showing the text instead of
html (with inlined images only) version of a message. html messages are shown
in an iframe served from an endpoint with CSP headers to prevent dangerous
resources (scripts, external images) from being loaded. the html is also
sanitized, with javascript removed. a user can choose to load external
resources (e.g. images for tracking purposes).
the frontend is just (strict) typescript, no external frameworks. all
incoming/outgoing data is typechecked, both the api request parameters and
response types, and the data coming in over SSE. the types and checking code
are generated with sherpats, which uses the api definitions generated by
sherpadoc based on the Go code. so types from the backend are automatically
propagated to the frontend. since there is no framework to automatically
propagate properties and rerender components, changes coming in over the SSE
connection are propagated explicitly with regular function calls. the ui is
separated into "views", each with a "root" dom element that is added to the
visible document. these views have additional functions for getting changes
propagated, often resulting in the view updating its (internal) ui state (dom).
we keep the frontend compilation simple, it's just a few typescript files that
get compiled (combined and types stripped) into a single js file, no additional
runtime code needed or complicated build processes used. the webmail is served
is served from a compressed, cachable html file that includes style and the
javascript, currently just over 225kb uncompressed, under 60kb compressed (not
minified, including comments). we include the generated js files in the
repository, to keep Go's easily buildable self-contained binaries.
authentication is basic http, as with the account and admin pages. most data
comes in over one long-term SSE connection to the backend. api requests signal
which mailbox/search/messages are requested over the SSE connection. fetching
individual messages, and making changes, are done through api calls. the
operations are similar to imap, so some code has been moved from package
imapserver to package store. the future jmap implementation will benefit from
these changes too. more functionality will probably be moved to the store
package in the future.
the quickstart enables webmail on the internal listener by default (for new
installs). users can enable it on the public listener if they want to. mox
localserve enables it too. to enable webmail on existing installs, add settings
like the following to the listeners in mox.conf, similar to AccountHTTP(S):
WebmailHTTP:
Enabled: true
WebmailHTTPS:
Enabled: true
special thanks to liesbeth, gerben, andrii for early user feedback.
there is plenty still to do, see the list at the top of webmail/webmail.ts.
feedback welcome as always.
2023-08-07 22:57:03 +03:00
|
|
|
"sort"
|
2023-02-16 11:57:27 +03:00
|
|
|
"strconv"
|
|
|
|
"strings"
|
|
|
|
"time"
|
|
|
|
|
2023-06-24 01:24:43 +03:00
|
|
|
"golang.org/x/exp/maps"
|
2024-02-08 16:49:01 +03:00
|
|
|
|
2023-02-16 11:57:27 +03:00
|
|
|
"golang.org/x/text/unicode/norm"
|
|
|
|
|
|
|
|
"github.com/mjl-/bstore"
|
|
|
|
|
|
|
|
"github.com/mjl-/mox/message"
|
2023-09-11 12:43:49 +03:00
|
|
|
"github.com/mjl-/mox/metrics"
|
2023-02-16 11:57:27 +03:00
|
|
|
"github.com/mjl-/mox/mlog"
|
|
|
|
"github.com/mjl-/mox/mox-"
|
|
|
|
"github.com/mjl-/mox/store"
|
|
|
|
)
|
|
|
|
|
|
|
|
type importListener struct {
|
|
|
|
Token string
|
|
|
|
Events chan importEvent
|
|
|
|
Register chan bool // Whether register is successful.
|
|
|
|
}
|
|
|
|
|
|
|
|
type importEvent struct {
|
|
|
|
Token string
|
|
|
|
SSEMsg []byte // Full SSE message, including event: ... and data: ... \n\n
|
|
|
|
Event any // nil, importCount, importProblem, importDone, importAborted
|
|
|
|
Cancel func() // For cancelling the context causing abort of the import. Set in first, import-registering, event.
|
|
|
|
}
|
|
|
|
|
|
|
|
type importAbortRequest struct {
|
|
|
|
Token string
|
|
|
|
Response chan error
|
|
|
|
}
|
|
|
|
|
|
|
|
var importers = struct {
|
|
|
|
Register chan *importListener
|
|
|
|
Unregister chan *importListener
|
|
|
|
Events chan importEvent
|
|
|
|
Abort chan importAbortRequest
|
2024-06-10 19:18:20 +03:00
|
|
|
Stop chan struct{}
|
2023-02-16 11:57:27 +03:00
|
|
|
}{
|
|
|
|
make(chan *importListener, 1),
|
|
|
|
make(chan *importListener, 1),
|
|
|
|
make(chan importEvent),
|
|
|
|
make(chan importAbortRequest),
|
2024-06-10 19:18:20 +03:00
|
|
|
make(chan struct{}),
|
2023-02-16 11:57:27 +03:00
|
|
|
}
|
|
|
|
|
add webmail
it was far down on the roadmap, but implemented earlier, because it's
interesting, and to help prepare for a jmap implementation. for jmap we need to
implement more client-like functionality than with just imap. internal data
structures need to change. jmap has lots of other requirements, so it's already
a big project. by implementing a webmail now, some of the required data
structure changes become clear and can be made now, so the later jmap
implementation can do things similarly to the webmail code. the webmail
frontend and webmail are written together, making their interface/api much
smaller and simpler than jmap.
one of the internal changes is that we now keep track of per-mailbox
total/unread/unseen/deleted message counts and mailbox sizes. keeping this
data consistent after any change to the stored messages (through the code base)
is tricky, so mox now has a consistency check that verifies the counts are
correct, which runs only during tests, each time an internal account reference
is closed. we have a few more internal "changes" that are propagated for the
webmail frontend (that imap doesn't have a way to propagate on a connection),
like changes to the special-use flags on mailboxes, and used keywords in a
mailbox. more changes that will be required have revealed themselves while
implementing the webmail, and will be implemented next.
the webmail user interface is modeled after the mail clients i use or have
used: thunderbird, macos mail, mutt; and webmails i normally only use for
testing: gmail, proton, yahoo, outlook. a somewhat technical user is assumed,
but still the goal is to make this webmail client easy to use for everyone. the
user interface looks like most other mail clients: a list of mailboxes, a
search bar, a message list view, and message details. there is a top/bottom and
a left/right layout for the list/message view, default is automatic based on
screen size. the panes can be resized by the user. buttons for actions are just
text, not icons. clicking a button briefly shows the shortcut for the action in
the bottom right, helping with learning to operate quickly. any text that is
underdotted has a title attribute that causes more information to be displayed,
e.g. what a button does or a field is about. to highlight potential phishing
attempts, any text (anywhere in the webclient) that switches unicode "blocks"
(a rough approximation to (language) scripts) within a word is underlined
orange. multiple messages can be selected with familiar ui interaction:
clicking while holding control and/or shift keys. keyboard navigation works
with arrows/page up/down and home/end keys, and also with a few basic vi-like
keys for list/message navigation. we prefer showing the text instead of
html (with inlined images only) version of a message. html messages are shown
in an iframe served from an endpoint with CSP headers to prevent dangerous
resources (scripts, external images) from being loaded. the html is also
sanitized, with javascript removed. a user can choose to load external
resources (e.g. images for tracking purposes).
the frontend is just (strict) typescript, no external frameworks. all
incoming/outgoing data is typechecked, both the api request parameters and
response types, and the data coming in over SSE. the types and checking code
are generated with sherpats, which uses the api definitions generated by
sherpadoc based on the Go code. so types from the backend are automatically
propagated to the frontend. since there is no framework to automatically
propagate properties and rerender components, changes coming in over the SSE
connection are propagated explicitly with regular function calls. the ui is
separated into "views", each with a "root" dom element that is added to the
visible document. these views have additional functions for getting changes
propagated, often resulting in the view updating its (internal) ui state (dom).
we keep the frontend compilation simple, it's just a few typescript files that
get compiled (combined and types stripped) into a single js file, no additional
runtime code needed or complicated build processes used. the webmail is served
is served from a compressed, cachable html file that includes style and the
javascript, currently just over 225kb uncompressed, under 60kb compressed (not
minified, including comments). we include the generated js files in the
repository, to keep Go's easily buildable self-contained binaries.
authentication is basic http, as with the account and admin pages. most data
comes in over one long-term SSE connection to the backend. api requests signal
which mailbox/search/messages are requested over the SSE connection. fetching
individual messages, and making changes, are done through api calls. the
operations are similar to imap, so some code has been moved from package
imapserver to package store. the future jmap implementation will benefit from
these changes too. more functionality will probably be moved to the store
package in the future.
the quickstart enables webmail on the internal listener by default (for new
installs). users can enable it on the public listener if they want to. mox
localserve enables it too. to enable webmail on existing installs, add settings
like the following to the listeners in mox.conf, similar to AccountHTTP(S):
WebmailHTTP:
Enabled: true
WebmailHTTPS:
Enabled: true
special thanks to liesbeth, gerben, andrii for early user feedback.
there is plenty still to do, see the list at the top of webmail/webmail.ts.
feedback welcome as always.
2023-08-07 22:57:03 +03:00
|
|
|
// ImportManage should be run as a goroutine, it manages imports of mboxes/maildirs, propagating progress over SSE connections.
|
|
|
|
func ImportManage() {
|
2023-12-05 15:35:58 +03:00
|
|
|
log := mlog.New("httpimport", nil)
|
2023-02-16 11:57:27 +03:00
|
|
|
defer func() {
|
|
|
|
if x := recover(); x != nil {
|
2023-12-05 15:35:58 +03:00
|
|
|
log.Error("import manage panic", slog.Any("err", x))
|
2023-02-16 11:57:27 +03:00
|
|
|
debug.PrintStack()
|
2023-09-15 17:47:17 +03:00
|
|
|
metrics.PanicInc(metrics.Importmanage)
|
2023-02-16 11:57:27 +03:00
|
|
|
}
|
|
|
|
}()
|
|
|
|
|
|
|
|
type state struct {
|
|
|
|
MailboxCounts map[string]int
|
|
|
|
Problems []string
|
|
|
|
Done *time.Time
|
|
|
|
Aborted *time.Time
|
|
|
|
Listeners map[*importListener]struct{}
|
|
|
|
Cancel func()
|
|
|
|
}
|
|
|
|
|
|
|
|
imports := map[string]state{} // Token to state.
|
|
|
|
for {
|
|
|
|
select {
|
|
|
|
case l := <-importers.Register:
|
|
|
|
// If we have state, send it so the client is up to date.
|
2024-06-10 19:18:20 +03:00
|
|
|
s, ok := imports[l.Token]
|
|
|
|
l.Register <- ok
|
|
|
|
if !ok {
|
|
|
|
break
|
|
|
|
}
|
|
|
|
s.Listeners[l] = struct{}{}
|
2023-02-16 11:57:27 +03:00
|
|
|
|
2024-06-10 19:18:20 +03:00
|
|
|
sendEvent := func(kind string, v any) {
|
|
|
|
buf, err := json.Marshal(v)
|
|
|
|
if err != nil {
|
|
|
|
log.Errorx("marshal event", err, slog.String("kind", kind), slog.Any("event", v))
|
|
|
|
return
|
2023-02-16 11:57:27 +03:00
|
|
|
}
|
2024-06-10 19:18:20 +03:00
|
|
|
ssemsg := fmt.Sprintf("event: %s\ndata: %s\n\n", kind, buf)
|
|
|
|
|
|
|
|
select {
|
|
|
|
case l.Events <- importEvent{kind, []byte(ssemsg), nil, nil}:
|
|
|
|
default:
|
|
|
|
log.Debug("dropped initial import event to slow consumer")
|
2023-02-16 11:57:27 +03:00
|
|
|
}
|
2024-06-10 19:18:20 +03:00
|
|
|
}
|
|
|
|
|
|
|
|
for m, c := range s.MailboxCounts {
|
|
|
|
sendEvent("count", importCount{m, c})
|
|
|
|
}
|
|
|
|
for _, p := range s.Problems {
|
|
|
|
sendEvent("problem", importProblem{p})
|
|
|
|
}
|
|
|
|
if s.Done != nil {
|
|
|
|
sendEvent("done", importDone{})
|
|
|
|
} else if s.Aborted != nil {
|
|
|
|
sendEvent("aborted", importAborted{})
|
2023-02-16 11:57:27 +03:00
|
|
|
}
|
|
|
|
|
|
|
|
case l := <-importers.Unregister:
|
|
|
|
delete(imports[l.Token].Listeners, l)
|
|
|
|
|
|
|
|
case e := <-importers.Events:
|
|
|
|
s, ok := imports[e.Token]
|
|
|
|
if !ok {
|
2023-02-16 15:24:51 +03:00
|
|
|
s = state{
|
2023-02-16 11:57:27 +03:00
|
|
|
MailboxCounts: map[string]int{},
|
|
|
|
Listeners: map[*importListener]struct{}{},
|
|
|
|
Cancel: e.Cancel,
|
|
|
|
}
|
|
|
|
imports[e.Token] = s
|
|
|
|
}
|
|
|
|
for l := range s.Listeners {
|
|
|
|
select {
|
|
|
|
case l.Events <- e:
|
|
|
|
default:
|
|
|
|
log.Debug("dropped import event to slow consumer")
|
|
|
|
}
|
|
|
|
}
|
|
|
|
if e.Event != nil {
|
|
|
|
s := imports[e.Token]
|
|
|
|
switch x := e.Event.(type) {
|
|
|
|
case importCount:
|
|
|
|
s.MailboxCounts[x.Mailbox] = x.Count
|
|
|
|
case importProblem:
|
|
|
|
s.Problems = append(s.Problems, x.Message)
|
|
|
|
case importDone:
|
|
|
|
now := time.Now()
|
|
|
|
s.Done = &now
|
|
|
|
case importAborted:
|
|
|
|
now := time.Now()
|
|
|
|
s.Aborted = &now
|
|
|
|
}
|
|
|
|
imports[e.Token] = s
|
|
|
|
}
|
|
|
|
|
|
|
|
case a := <-importers.Abort:
|
|
|
|
s, ok := imports[a.Token]
|
|
|
|
if !ok {
|
|
|
|
a.Response <- errors.New("import not found")
|
|
|
|
return
|
|
|
|
}
|
|
|
|
if s.Done != nil {
|
|
|
|
a.Response <- errors.New("import already finished")
|
|
|
|
return
|
|
|
|
}
|
|
|
|
s.Cancel()
|
|
|
|
a.Response <- nil
|
2024-06-10 19:18:20 +03:00
|
|
|
|
|
|
|
case <-importers.Stop:
|
|
|
|
return
|
2023-02-16 11:57:27 +03:00
|
|
|
}
|
|
|
|
|
|
|
|
// Cleanup old state.
|
|
|
|
for t, s := range imports {
|
|
|
|
if len(s.Listeners) > 0 {
|
|
|
|
continue
|
|
|
|
}
|
|
|
|
if s.Done != nil && time.Since(*s.Done) > time.Minute || s.Aborted != nil && time.Since(*s.Aborted) > time.Minute {
|
|
|
|
delete(imports, t)
|
|
|
|
}
|
|
|
|
}
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
|
|
|
type importCount struct {
|
|
|
|
Mailbox string
|
|
|
|
Count int
|
|
|
|
}
|
|
|
|
type importProblem struct {
|
|
|
|
Message string
|
|
|
|
}
|
|
|
|
type importDone struct{}
|
|
|
|
type importAborted struct{}
|
implement message threading in backend and webmail
we match messages to their parents based on the "references" and "in-reply-to"
headers (requiring the same base subject), and in absense of those headers we
also by only base subject (against messages received max 4 weeks ago).
we store a threadid with messages. all messages in a thread have the same
threadid. messages also have a "thread parent ids", which holds all id's of
parent messages up to the thread root. then there is "thread missing link",
which is set when a referenced immediate parent wasn't found (but possibly
earlier ancestors can still be found and will be in thread parent ids".
threads can be muted: newly delivered messages are automatically marked as
read/seen. threads can be marked as collapsed: if set, the webmail collapses
the thread to a single item in the basic threading view (default is to expand
threads). the muted and collapsed fields are copied from their parent on
message delivery.
the threading is implemented in the webmail. the non-threading mode still works
as before. the new default threading mode "unread" automatically expands only
the threads with at least one unread (not seen) meessage. the basic threading
mode "on" expands all threads except when explicitly collapsed (as saved in the
thread collapsed field). new shortcuts for navigation/interaction threads have
been added, e.g. go to previous/next thread root, toggle collapse/expand of
thread (or double click), toggle mute of thread. some previous shortcuts have
changed, see the help for details.
the message threading are added with an explicit account upgrade step,
automatically started when an account is opened. the upgrade is done in the
background because it will take too long for large mailboxes to block account
operations. the upgrade takes two steps: 1. updating all message records in the
database to add a normalized message-id and thread base subject (with "re:",
"fwd:" and several other schemes stripped). 2. going through all messages in
the database again, reading the "references" and "in-reply-to" headers from
disk, and matching against their parents. this second step is also done at the
end of each import of mbox/maildir mailboxes. new deliveries are matched
immediately against other existing messages, currently no attempt is made to
rematch previously delivered messages (which could be useful for related
messages being delivered out of order).
the threading is not yet exposed over imap.
2023-09-13 09:51:50 +03:00
|
|
|
type importStep struct {
|
|
|
|
Title string
|
|
|
|
}
|
2023-02-16 11:57:27 +03:00
|
|
|
|
|
|
|
// importStart prepare the import and launches the goroutine to actually import.
|
make mox compile on windows, without "mox serve" but with working "mox localserve"
getting mox to compile required changing code in only a few places where
package "syscall" was used: for accessing file access times and for umask
handling. an open problem is how to start a process as an unprivileged user on
windows. that's why "mox serve" isn't implemented yet. and just finding a way
to implement it now may not be good enough in the near future: we may want to
starting using a more complete privilege separation approach, with a process
handling sensitive tasks (handling private keys, authentication), where we may
want to pass file descriptors between processes. how would that work on
windows?
anyway, getting mox to compile for windows doesn't mean it works properly on
windows. the largest issue: mox would normally open a file, rename or remove
it, and finally close it. this happens during message delivery. that doesn't
work on windows, the rename/remove would fail because the file is still open.
so this commit swaps many "remove" and "close" calls. renames are a longer
story: message delivery had two ways to deliver: with "consuming" the
(temporary) message file (which would rename it to its final destination), and
without consuming (by hardlinking the file, falling back to copying). the last
delivery to a recipient of a message (and the only one in the common case of a
single recipient) would consume the message, and the earlier recipients would
not. during delivery, the already open message file was used, to parse the
message. we still want to use that open message file, and the caller now stays
responsible for closing it, but we no longer try to rename (consume) the file.
we always hardlink (or copy) during delivery (this works on windows), and the
caller is responsible for closing and removing (in that order) the original
temporary file. this does cost one syscall more. but it makes the delivery code
(responsibilities) a bit simpler.
there is one more obvious issue: the file system path separator. mox already
used the "filepath" package to join paths in many places, but not everywhere.
and it still used strings with slashes for local file access. with this commit,
the code now uses filepath.FromSlash for path strings with slashes, uses
"filepath" in a few more places where it previously didn't. also switches from
"filepath" to regular "path" package when handling mailbox names in a few
places, because those always use forward slashes, regardless of local file
system conventions. windows can handle forward slashes when opening files, so
test code that passes path strings with forward slashes straight to go stdlib
file i/o functions are left unchanged to reduce code churn. the regular
non-test code, or test code that uses path strings in places other than
standard i/o functions, does have the paths converted for consistent paths
(otherwise we would end up with paths with mixed forward/backward slashes in
log messages).
windows cannot dup a listening socket. for "mox localserve", it isn't
important, and we can work around the issue. the current approach for "mox
serve" (forking a process and passing file descriptors of listening sockets on
"privileged" ports) won't work on windows. perhaps it isn't needed on windows,
and any user can listen on "privileged" ports? that would be welcome.
on windows, os.Open cannot open a directory, so we cannot call Sync on it after
message delivery. a cursory internet search indicates that directories cannot
be synced on windows. the story is probably much more nuanced than that, with
long deep technical details/discussions/disagreement/confusion, like on unix.
for "mox localserve" we can get away with making syncdir a no-op.
2023-10-14 11:54:07 +03:00
|
|
|
// importStart is responsible for closing f and removing f.
|
2024-01-05 13:31:05 +03:00
|
|
|
func importStart(log mlog.Log, accName string, f *os.File, skipMailboxPrefix string) (string, bool, error) {
|
2023-02-16 11:57:27 +03:00
|
|
|
defer func() {
|
|
|
|
if f != nil {
|
2023-11-01 20:57:38 +03:00
|
|
|
store.CloseRemoveTempFile(log, f, "upload for import")
|
2023-02-16 11:57:27 +03:00
|
|
|
}
|
|
|
|
}()
|
|
|
|
|
|
|
|
buf := make([]byte, 16)
|
|
|
|
if _, err := cryptrand.Read(buf); err != nil {
|
2024-01-05 13:31:05 +03:00
|
|
|
return "", false, err
|
2023-02-16 11:57:27 +03:00
|
|
|
}
|
|
|
|
token := fmt.Sprintf("%x", buf)
|
|
|
|
|
|
|
|
if _, err := f.Seek(0, 0); err != nil {
|
2024-01-05 13:31:05 +03:00
|
|
|
return "", false, fmt.Errorf("seek to start of file: %v", err)
|
2023-02-16 11:57:27 +03:00
|
|
|
}
|
|
|
|
|
|
|
|
// Recognize file format.
|
|
|
|
var iszip bool
|
|
|
|
magicZip := []byte{0x50, 0x4b, 0x03, 0x04}
|
|
|
|
magicGzip := []byte{0x1f, 0x8b}
|
|
|
|
magic := make([]byte, 4)
|
|
|
|
if _, err := f.ReadAt(magic, 0); err != nil {
|
2024-01-05 13:31:05 +03:00
|
|
|
return "", true, fmt.Errorf("detecting file format: %v", err)
|
2023-02-16 11:57:27 +03:00
|
|
|
}
|
|
|
|
if bytes.Equal(magic, magicZip) {
|
|
|
|
iszip = true
|
|
|
|
} else if !bytes.Equal(magic[:2], magicGzip) {
|
2024-01-05 13:31:05 +03:00
|
|
|
return "", true, fmt.Errorf("file is not a zip or gzip file")
|
2023-02-16 11:57:27 +03:00
|
|
|
}
|
|
|
|
|
|
|
|
var zr *zip.Reader
|
|
|
|
var tr *tar.Reader
|
|
|
|
if iszip {
|
|
|
|
fi, err := f.Stat()
|
|
|
|
if err != nil {
|
2024-01-05 13:31:05 +03:00
|
|
|
return "", false, fmt.Errorf("stat temporary import zip file: %v", err)
|
2023-02-16 11:57:27 +03:00
|
|
|
}
|
|
|
|
zr, err = zip.NewReader(f, fi.Size())
|
|
|
|
if err != nil {
|
2024-01-05 13:31:05 +03:00
|
|
|
return "", true, fmt.Errorf("opening zip file: %v", err)
|
2023-02-16 11:57:27 +03:00
|
|
|
}
|
|
|
|
} else {
|
|
|
|
gzr, err := gzip.NewReader(f)
|
|
|
|
if err != nil {
|
2024-01-05 13:31:05 +03:00
|
|
|
return "", true, fmt.Errorf("gunzip: %v", err)
|
2023-02-16 11:57:27 +03:00
|
|
|
}
|
|
|
|
tr = tar.NewReader(gzr)
|
|
|
|
}
|
|
|
|
|
2023-12-05 15:35:58 +03:00
|
|
|
acc, err := store.OpenAccount(log, accName)
|
2023-02-16 11:57:27 +03:00
|
|
|
if err != nil {
|
2024-01-05 13:31:05 +03:00
|
|
|
return "", false, fmt.Errorf("open acount: %v", err)
|
2023-02-16 11:57:27 +03:00
|
|
|
}
|
|
|
|
acc.Lock() // Not using WithWLock because importMessage is responsible for unlocking.
|
|
|
|
|
2023-05-22 15:40:36 +03:00
|
|
|
tx, err := acc.DB.Begin(context.Background(), true)
|
2023-02-16 11:57:27 +03:00
|
|
|
if err != nil {
|
|
|
|
acc.Unlock()
|
2023-02-16 15:22:00 +03:00
|
|
|
xerr := acc.Close()
|
|
|
|
log.Check(xerr, "closing account")
|
2024-01-05 13:31:05 +03:00
|
|
|
return "", false, fmt.Errorf("start transaction: %v", err)
|
2023-02-16 11:57:27 +03:00
|
|
|
}
|
|
|
|
|
|
|
|
// Ensure token is registered before returning, with context that can be canceled.
|
|
|
|
ctx, cancel := context.WithCancel(mox.Shutdown)
|
|
|
|
importers.Events <- importEvent{token, []byte(": keepalive\n\n"), nil, cancel}
|
|
|
|
|
|
|
|
log.Info("starting import")
|
|
|
|
go importMessages(ctx, log.WithCid(mox.Cid()), token, acc, tx, zr, tr, f, skipMailboxPrefix)
|
make mox compile on windows, without "mox serve" but with working "mox localserve"
getting mox to compile required changing code in only a few places where
package "syscall" was used: for accessing file access times and for umask
handling. an open problem is how to start a process as an unprivileged user on
windows. that's why "mox serve" isn't implemented yet. and just finding a way
to implement it now may not be good enough in the near future: we may want to
starting using a more complete privilege separation approach, with a process
handling sensitive tasks (handling private keys, authentication), where we may
want to pass file descriptors between processes. how would that work on
windows?
anyway, getting mox to compile for windows doesn't mean it works properly on
windows. the largest issue: mox would normally open a file, rename or remove
it, and finally close it. this happens during message delivery. that doesn't
work on windows, the rename/remove would fail because the file is still open.
so this commit swaps many "remove" and "close" calls. renames are a longer
story: message delivery had two ways to deliver: with "consuming" the
(temporary) message file (which would rename it to its final destination), and
without consuming (by hardlinking the file, falling back to copying). the last
delivery to a recipient of a message (and the only one in the common case of a
single recipient) would consume the message, and the earlier recipients would
not. during delivery, the already open message file was used, to parse the
message. we still want to use that open message file, and the caller now stays
responsible for closing it, but we no longer try to rename (consume) the file.
we always hardlink (or copy) during delivery (this works on windows), and the
caller is responsible for closing and removing (in that order) the original
temporary file. this does cost one syscall more. but it makes the delivery code
(responsibilities) a bit simpler.
there is one more obvious issue: the file system path separator. mox already
used the "filepath" package to join paths in many places, but not everywhere.
and it still used strings with slashes for local file access. with this commit,
the code now uses filepath.FromSlash for path strings with slashes, uses
"filepath" in a few more places where it previously didn't. also switches from
"filepath" to regular "path" package when handling mailbox names in a few
places, because those always use forward slashes, regardless of local file
system conventions. windows can handle forward slashes when opening files, so
test code that passes path strings with forward slashes straight to go stdlib
file i/o functions are left unchanged to reduce code churn. the regular
non-test code, or test code that uses path strings in places other than
standard i/o functions, does have the paths converted for consistent paths
(otherwise we would end up with paths with mixed forward/backward slashes in
log messages).
windows cannot dup a listening socket. for "mox localserve", it isn't
important, and we can work around the issue. the current approach for "mox
serve" (forking a process and passing file descriptors of listening sockets on
"privileged" ports) won't work on windows. perhaps it isn't needed on windows,
and any user can listen on "privileged" ports? that would be welcome.
on windows, os.Open cannot open a directory, so we cannot call Sync on it after
message delivery. a cursory internet search indicates that directories cannot
be synced on windows. the story is probably much more nuanced than that, with
long deep technical details/discussions/disagreement/confusion, like on unix.
for "mox localserve" we can get away with making syncdir a no-op.
2023-10-14 11:54:07 +03:00
|
|
|
f = nil // importMessages is now responsible for closing and removing.
|
2023-02-16 11:57:27 +03:00
|
|
|
|
2024-01-05 13:31:05 +03:00
|
|
|
return token, false, nil
|
2023-02-16 11:57:27 +03:00
|
|
|
}
|
|
|
|
|
|
|
|
// importMessages imports the messages from zip/tgz file f.
|
|
|
|
// importMessages is responsible for unlocking and closing acc, and closing tx and f.
|
2023-12-05 15:35:58 +03:00
|
|
|
func 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) {
|
2023-02-16 11:57:27 +03:00
|
|
|
// If a fatal processing error occurs, we panic with this type.
|
|
|
|
type importError struct{ Err error }
|
|
|
|
|
|
|
|
// During import we collect all changes and broadcast them at the end, when successful.
|
|
|
|
var changes []store.Change
|
|
|
|
|
|
|
|
// ID's of delivered messages. If we have to rollback, we have to remove this files.
|
|
|
|
var deliveredIDs []int64
|
|
|
|
|
|
|
|
sendEvent := func(kind string, v any) {
|
|
|
|
buf, err := json.Marshal(v)
|
|
|
|
if err != nil {
|
2023-12-05 15:35:58 +03:00
|
|
|
log.Errorx("marshal event", err, slog.String("kind", kind), slog.Any("event", v))
|
2023-02-16 11:57:27 +03:00
|
|
|
return
|
|
|
|
}
|
|
|
|
ssemsg := fmt.Sprintf("event: %s\ndata: %s\n\n", kind, buf)
|
|
|
|
importers.Events <- importEvent{token, []byte(ssemsg), v, nil}
|
|
|
|
}
|
|
|
|
|
|
|
|
canceled := func() bool {
|
|
|
|
select {
|
|
|
|
case <-ctx.Done():
|
|
|
|
sendEvent("aborted", importAborted{})
|
|
|
|
return true
|
|
|
|
default:
|
|
|
|
return false
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
implement message threading in backend and webmail
we match messages to their parents based on the "references" and "in-reply-to"
headers (requiring the same base subject), and in absense of those headers we
also by only base subject (against messages received max 4 weeks ago).
we store a threadid with messages. all messages in a thread have the same
threadid. messages also have a "thread parent ids", which holds all id's of
parent messages up to the thread root. then there is "thread missing link",
which is set when a referenced immediate parent wasn't found (but possibly
earlier ancestors can still be found and will be in thread parent ids".
threads can be muted: newly delivered messages are automatically marked as
read/seen. threads can be marked as collapsed: if set, the webmail collapses
the thread to a single item in the basic threading view (default is to expand
threads). the muted and collapsed fields are copied from their parent on
message delivery.
the threading is implemented in the webmail. the non-threading mode still works
as before. the new default threading mode "unread" automatically expands only
the threads with at least one unread (not seen) meessage. the basic threading
mode "on" expands all threads except when explicitly collapsed (as saved in the
thread collapsed field). new shortcuts for navigation/interaction threads have
been added, e.g. go to previous/next thread root, toggle collapse/expand of
thread (or double click), toggle mute of thread. some previous shortcuts have
changed, see the help for details.
the message threading are added with an explicit account upgrade step,
automatically started when an account is opened. the upgrade is done in the
background because it will take too long for large mailboxes to block account
operations. the upgrade takes two steps: 1. updating all message records in the
database to add a normalized message-id and thread base subject (with "re:",
"fwd:" and several other schemes stripped). 2. going through all messages in
the database again, reading the "references" and "in-reply-to" headers from
disk, and matching against their parents. this second step is also done at the
end of each import of mbox/maildir mailboxes. new deliveries are matched
immediately against other existing messages, currently no attempt is made to
rematch previously delivered messages (which could be useful for related
messages being delivered out of order).
the threading is not yet exposed over imap.
2023-09-13 09:51:50 +03:00
|
|
|
problemf := func(format string, args ...any) {
|
|
|
|
msg := fmt.Sprintf(format, args...)
|
|
|
|
sendEvent("problem", importProblem{Message: msg})
|
|
|
|
}
|
|
|
|
|
2023-02-16 11:57:27 +03:00
|
|
|
defer func() {
|
2023-11-01 20:57:38 +03:00
|
|
|
store.CloseRemoveTempFile(log, f, "uploaded messages")
|
2023-02-16 15:22:00 +03:00
|
|
|
|
2023-02-16 11:57:27 +03:00
|
|
|
for _, id := range deliveredIDs {
|
|
|
|
p := acc.MessagePath(id)
|
2023-02-16 15:22:00 +03:00
|
|
|
err := os.Remove(p)
|
2023-12-05 15:35:58 +03:00
|
|
|
log.Check(err, "closing message file after import error", slog.String("path", p))
|
2023-02-16 11:57:27 +03:00
|
|
|
}
|
|
|
|
if tx != nil {
|
2023-02-16 15:22:00 +03:00
|
|
|
err := tx.Rollback()
|
|
|
|
log.Check(err, "rolling back transaction")
|
2023-02-16 11:57:27 +03:00
|
|
|
}
|
|
|
|
if acc != nil {
|
|
|
|
acc.Unlock()
|
2023-02-16 15:22:00 +03:00
|
|
|
err := acc.Close()
|
|
|
|
log.Check(err, "closing account")
|
2023-02-16 11:57:27 +03:00
|
|
|
}
|
|
|
|
|
|
|
|
x := recover()
|
|
|
|
if x == nil {
|
|
|
|
return
|
|
|
|
}
|
|
|
|
if err, ok := x.(importError); ok {
|
|
|
|
log.Errorx("import error", err.Err)
|
|
|
|
problemf("%s (aborting)", err.Err)
|
|
|
|
sendEvent("aborted", importAborted{})
|
|
|
|
} else {
|
2023-12-05 15:35:58 +03:00
|
|
|
log.Error("import panic", slog.Any("err", x))
|
2023-02-16 11:57:27 +03:00
|
|
|
debug.PrintStack()
|
2023-09-15 17:47:17 +03:00
|
|
|
metrics.PanicInc(metrics.Importmessages)
|
2023-02-16 11:57:27 +03:00
|
|
|
}
|
|
|
|
}()
|
|
|
|
|
implement message threading in backend and webmail
we match messages to their parents based on the "references" and "in-reply-to"
headers (requiring the same base subject), and in absense of those headers we
also by only base subject (against messages received max 4 weeks ago).
we store a threadid with messages. all messages in a thread have the same
threadid. messages also have a "thread parent ids", which holds all id's of
parent messages up to the thread root. then there is "thread missing link",
which is set when a referenced immediate parent wasn't found (but possibly
earlier ancestors can still be found and will be in thread parent ids".
threads can be muted: newly delivered messages are automatically marked as
read/seen. threads can be marked as collapsed: if set, the webmail collapses
the thread to a single item in the basic threading view (default is to expand
threads). the muted and collapsed fields are copied from their parent on
message delivery.
the threading is implemented in the webmail. the non-threading mode still works
as before. the new default threading mode "unread" automatically expands only
the threads with at least one unread (not seen) meessage. the basic threading
mode "on" expands all threads except when explicitly collapsed (as saved in the
thread collapsed field). new shortcuts for navigation/interaction threads have
been added, e.g. go to previous/next thread root, toggle collapse/expand of
thread (or double click), toggle mute of thread. some previous shortcuts have
changed, see the help for details.
the message threading are added with an explicit account upgrade step,
automatically started when an account is opened. the upgrade is done in the
background because it will take too long for large mailboxes to block account
operations. the upgrade takes two steps: 1. updating all message records in the
database to add a normalized message-id and thread base subject (with "re:",
"fwd:" and several other schemes stripped). 2. going through all messages in
the database again, reading the "references" and "in-reply-to" headers from
disk, and matching against their parents. this second step is also done at the
end of each import of mbox/maildir mailboxes. new deliveries are matched
immediately against other existing messages, currently no attempt is made to
rematch previously delivered messages (which could be useful for related
messages being delivered out of order).
the threading is not yet exposed over imap.
2023-09-13 09:51:50 +03:00
|
|
|
ximportcheckf := func(err error, format string, args ...any) {
|
|
|
|
if err != nil {
|
|
|
|
panic(importError{fmt.Errorf("%s: %s", fmt.Sprintf(format, args...), err)})
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
|
|
|
err := acc.ThreadingWait(log)
|
|
|
|
ximportcheckf(err, "waiting for account thread upgrade")
|
|
|
|
|
2023-02-16 11:57:27 +03:00
|
|
|
conf, _ := acc.Conf()
|
|
|
|
|
2023-05-22 15:40:36 +03:00
|
|
|
jf, _, err := acc.OpenJunkFilter(ctx, log)
|
2023-02-16 11:57:27 +03:00
|
|
|
if err != nil && !errors.Is(err, store.ErrNoJunkFilter) {
|
|
|
|
ximportcheckf(err, "open junk filter")
|
|
|
|
}
|
|
|
|
defer func() {
|
|
|
|
if jf != nil {
|
|
|
|
err := jf.CloseDiscard()
|
|
|
|
log.Check(err, "closing junk filter")
|
|
|
|
}
|
|
|
|
}()
|
|
|
|
|
|
|
|
// Mailboxes we imported, and message counts.
|
|
|
|
mailboxes := map[string]store.Mailbox{}
|
|
|
|
messages := map[string]int{}
|
|
|
|
|
2023-12-20 22:54:12 +03:00
|
|
|
maxSize := acc.QuotaMessageSize()
|
|
|
|
du := store.DiskUsage{ID: 1}
|
|
|
|
err = tx.Get(&du)
|
|
|
|
ximportcheckf(err, "get disk usage")
|
|
|
|
var addSize int64
|
|
|
|
|
2023-06-24 01:24:43 +03:00
|
|
|
// For maildirs, we are likely to get a possible dovecot-keywords file after having
|
|
|
|
// imported the messages. Once we see the keywords, we use them. But before that
|
|
|
|
// time we remember which messages miss a keywords. Once the keywords become
|
|
|
|
// available, we'll fix up the flags for the unknown messages
|
2023-02-16 11:57:27 +03:00
|
|
|
mailboxKeywords := map[string]map[rune]string{} // Mailbox to 'a'-'z' to flag name.
|
|
|
|
mailboxMissingKeywordMessages := map[string]map[int64]string{} // Mailbox to message id to string consisting of the unrecognized flags.
|
|
|
|
|
add webmail
it was far down on the roadmap, but implemented earlier, because it's
interesting, and to help prepare for a jmap implementation. for jmap we need to
implement more client-like functionality than with just imap. internal data
structures need to change. jmap has lots of other requirements, so it's already
a big project. by implementing a webmail now, some of the required data
structure changes become clear and can be made now, so the later jmap
implementation can do things similarly to the webmail code. the webmail
frontend and webmail are written together, making their interface/api much
smaller and simpler than jmap.
one of the internal changes is that we now keep track of per-mailbox
total/unread/unseen/deleted message counts and mailbox sizes. keeping this
data consistent after any change to the stored messages (through the code base)
is tricky, so mox now has a consistency check that verifies the counts are
correct, which runs only during tests, each time an internal account reference
is closed. we have a few more internal "changes" that are propagated for the
webmail frontend (that imap doesn't have a way to propagate on a connection),
like changes to the special-use flags on mailboxes, and used keywords in a
mailbox. more changes that will be required have revealed themselves while
implementing the webmail, and will be implemented next.
the webmail user interface is modeled after the mail clients i use or have
used: thunderbird, macos mail, mutt; and webmails i normally only use for
testing: gmail, proton, yahoo, outlook. a somewhat technical user is assumed,
but still the goal is to make this webmail client easy to use for everyone. the
user interface looks like most other mail clients: a list of mailboxes, a
search bar, a message list view, and message details. there is a top/bottom and
a left/right layout for the list/message view, default is automatic based on
screen size. the panes can be resized by the user. buttons for actions are just
text, not icons. clicking a button briefly shows the shortcut for the action in
the bottom right, helping with learning to operate quickly. any text that is
underdotted has a title attribute that causes more information to be displayed,
e.g. what a button does or a field is about. to highlight potential phishing
attempts, any text (anywhere in the webclient) that switches unicode "blocks"
(a rough approximation to (language) scripts) within a word is underlined
orange. multiple messages can be selected with familiar ui interaction:
clicking while holding control and/or shift keys. keyboard navigation works
with arrows/page up/down and home/end keys, and also with a few basic vi-like
keys for list/message navigation. we prefer showing the text instead of
html (with inlined images only) version of a message. html messages are shown
in an iframe served from an endpoint with CSP headers to prevent dangerous
resources (scripts, external images) from being loaded. the html is also
sanitized, with javascript removed. a user can choose to load external
resources (e.g. images for tracking purposes).
the frontend is just (strict) typescript, no external frameworks. all
incoming/outgoing data is typechecked, both the api request parameters and
response types, and the data coming in over SSE. the types and checking code
are generated with sherpats, which uses the api definitions generated by
sherpadoc based on the Go code. so types from the backend are automatically
propagated to the frontend. since there is no framework to automatically
propagate properties and rerender components, changes coming in over the SSE
connection are propagated explicitly with regular function calls. the ui is
separated into "views", each with a "root" dom element that is added to the
visible document. these views have additional functions for getting changes
propagated, often resulting in the view updating its (internal) ui state (dom).
we keep the frontend compilation simple, it's just a few typescript files that
get compiled (combined and types stripped) into a single js file, no additional
runtime code needed or complicated build processes used. the webmail is served
is served from a compressed, cachable html file that includes style and the
javascript, currently just over 225kb uncompressed, under 60kb compressed (not
minified, including comments). we include the generated js files in the
repository, to keep Go's easily buildable self-contained binaries.
authentication is basic http, as with the account and admin pages. most data
comes in over one long-term SSE connection to the backend. api requests signal
which mailbox/search/messages are requested over the SSE connection. fetching
individual messages, and making changes, are done through api calls. the
operations are similar to imap, so some code has been moved from package
imapserver to package store. the future jmap implementation will benefit from
these changes too. more functionality will probably be moved to the store
package in the future.
the quickstart enables webmail on the internal listener by default (for new
installs). users can enable it on the public listener if they want to. mox
localserve enables it too. to enable webmail on existing installs, add settings
like the following to the listeners in mox.conf, similar to AccountHTTP(S):
WebmailHTTP:
Enabled: true
WebmailHTTPS:
Enabled: true
special thanks to liesbeth, gerben, andrii for early user feedback.
there is plenty still to do, see the list at the top of webmail/webmail.ts.
feedback welcome as always.
2023-08-07 22:57:03 +03:00
|
|
|
// We keep the mailboxes we deliver to up to date with count and keywords (non-system flags).
|
|
|
|
destMailboxCounts := map[int64]store.MailboxCounts{}
|
2023-06-24 01:24:43 +03:00
|
|
|
destMailboxKeywords := map[int64]map[string]bool{}
|
|
|
|
|
2023-02-16 11:57:27 +03:00
|
|
|
// Previous mailbox an event was sent for. We send an event for new mailboxes, when
|
|
|
|
// another 100 messages were added, when adding a message to another mailbox, and
|
|
|
|
// finally at the end as a closing statement.
|
|
|
|
var prevMailbox string
|
|
|
|
|
2023-07-24 22:21:05 +03:00
|
|
|
var modseq store.ModSeq // Assigned on first message, used for all messages.
|
|
|
|
|
2023-02-16 11:57:27 +03:00
|
|
|
trainMessage := func(m *store.Message, p message.Part, pos string) {
|
|
|
|
words, err := jf.ParseMessage(p)
|
|
|
|
if err != nil {
|
|
|
|
problemf("parsing message %s for updating junk filter: %v (continuing)", pos, err)
|
|
|
|
return
|
|
|
|
}
|
2023-05-22 15:40:36 +03:00
|
|
|
err = jf.Train(ctx, !m.Junk, words)
|
2023-02-16 11:57:27 +03:00
|
|
|
if err != nil {
|
|
|
|
problemf("training junk filter for message %s: %v (continuing)", pos, err)
|
|
|
|
return
|
|
|
|
}
|
|
|
|
m.TrainedJunk = &m.Junk
|
|
|
|
}
|
|
|
|
|
|
|
|
openTrainMessage := func(m *store.Message) {
|
|
|
|
path := acc.MessagePath(m.ID)
|
|
|
|
f, err := os.Open(path)
|
|
|
|
if err != nil {
|
|
|
|
problemf("opening message again for training junk filter: %v (continuing)", err)
|
|
|
|
return
|
|
|
|
}
|
2023-02-16 15:22:00 +03:00
|
|
|
defer func() {
|
|
|
|
err := f.Close()
|
|
|
|
log.Check(err, "closing file after training junkfilter")
|
|
|
|
}()
|
2023-02-16 11:57:27 +03:00
|
|
|
p, err := m.LoadPart(f)
|
|
|
|
if err != nil {
|
|
|
|
problemf("loading parsed message again for training junk filter: %v (continuing)", err)
|
|
|
|
return
|
|
|
|
}
|
|
|
|
trainMessage(m, p, fmt.Sprintf("message id %d", m.ID))
|
|
|
|
}
|
|
|
|
|
|
|
|
xensureMailbox := func(name string) store.Mailbox {
|
|
|
|
name = norm.NFC.String(name)
|
|
|
|
if strings.ToLower(name) == "inbox" {
|
|
|
|
name = "Inbox"
|
|
|
|
}
|
|
|
|
|
|
|
|
if mb, ok := mailboxes[name]; ok {
|
|
|
|
return mb
|
|
|
|
}
|
|
|
|
|
|
|
|
var p string
|
|
|
|
var mb store.Mailbox
|
|
|
|
for i, e := range strings.Split(name, "/") {
|
|
|
|
if i == 0 {
|
|
|
|
p = e
|
|
|
|
} else {
|
|
|
|
p = path.Join(p, e)
|
|
|
|
}
|
|
|
|
if _, ok := mailboxes[p]; ok {
|
|
|
|
continue
|
|
|
|
}
|
|
|
|
|
|
|
|
q := bstore.QueryTx[store.Mailbox](tx)
|
|
|
|
q.FilterNonzero(store.Mailbox{Name: p})
|
|
|
|
var err error
|
|
|
|
mb, err = q.Get()
|
|
|
|
if err == bstore.ErrAbsent {
|
|
|
|
uidvalidity, err := acc.NextUIDValidity(tx)
|
|
|
|
ximportcheckf(err, "finding next uid validity")
|
|
|
|
mb = store.Mailbox{
|
|
|
|
Name: p,
|
|
|
|
UIDValidity: uidvalidity,
|
|
|
|
UIDNext: 1,
|
add webmail
it was far down on the roadmap, but implemented earlier, because it's
interesting, and to help prepare for a jmap implementation. for jmap we need to
implement more client-like functionality than with just imap. internal data
structures need to change. jmap has lots of other requirements, so it's already
a big project. by implementing a webmail now, some of the required data
structure changes become clear and can be made now, so the later jmap
implementation can do things similarly to the webmail code. the webmail
frontend and webmail are written together, making their interface/api much
smaller and simpler than jmap.
one of the internal changes is that we now keep track of per-mailbox
total/unread/unseen/deleted message counts and mailbox sizes. keeping this
data consistent after any change to the stored messages (through the code base)
is tricky, so mox now has a consistency check that verifies the counts are
correct, which runs only during tests, each time an internal account reference
is closed. we have a few more internal "changes" that are propagated for the
webmail frontend (that imap doesn't have a way to propagate on a connection),
like changes to the special-use flags on mailboxes, and used keywords in a
mailbox. more changes that will be required have revealed themselves while
implementing the webmail, and will be implemented next.
the webmail user interface is modeled after the mail clients i use or have
used: thunderbird, macos mail, mutt; and webmails i normally only use for
testing: gmail, proton, yahoo, outlook. a somewhat technical user is assumed,
but still the goal is to make this webmail client easy to use for everyone. the
user interface looks like most other mail clients: a list of mailboxes, a
search bar, a message list view, and message details. there is a top/bottom and
a left/right layout for the list/message view, default is automatic based on
screen size. the panes can be resized by the user. buttons for actions are just
text, not icons. clicking a button briefly shows the shortcut for the action in
the bottom right, helping with learning to operate quickly. any text that is
underdotted has a title attribute that causes more information to be displayed,
e.g. what a button does or a field is about. to highlight potential phishing
attempts, any text (anywhere in the webclient) that switches unicode "blocks"
(a rough approximation to (language) scripts) within a word is underlined
orange. multiple messages can be selected with familiar ui interaction:
clicking while holding control and/or shift keys. keyboard navigation works
with arrows/page up/down and home/end keys, and also with a few basic vi-like
keys for list/message navigation. we prefer showing the text instead of
html (with inlined images only) version of a message. html messages are shown
in an iframe served from an endpoint with CSP headers to prevent dangerous
resources (scripts, external images) from being loaded. the html is also
sanitized, with javascript removed. a user can choose to load external
resources (e.g. images for tracking purposes).
the frontend is just (strict) typescript, no external frameworks. all
incoming/outgoing data is typechecked, both the api request parameters and
response types, and the data coming in over SSE. the types and checking code
are generated with sherpats, which uses the api definitions generated by
sherpadoc based on the Go code. so types from the backend are automatically
propagated to the frontend. since there is no framework to automatically
propagate properties and rerender components, changes coming in over the SSE
connection are propagated explicitly with regular function calls. the ui is
separated into "views", each with a "root" dom element that is added to the
visible document. these views have additional functions for getting changes
propagated, often resulting in the view updating its (internal) ui state (dom).
we keep the frontend compilation simple, it's just a few typescript files that
get compiled (combined and types stripped) into a single js file, no additional
runtime code needed or complicated build processes used. the webmail is served
is served from a compressed, cachable html file that includes style and the
javascript, currently just over 225kb uncompressed, under 60kb compressed (not
minified, including comments). we include the generated js files in the
repository, to keep Go's easily buildable self-contained binaries.
authentication is basic http, as with the account and admin pages. most data
comes in over one long-term SSE connection to the backend. api requests signal
which mailbox/search/messages are requested over the SSE connection. fetching
individual messages, and making changes, are done through api calls. the
operations are similar to imap, so some code has been moved from package
imapserver to package store. the future jmap implementation will benefit from
these changes too. more functionality will probably be moved to the store
package in the future.
the quickstart enables webmail on the internal listener by default (for new
installs). users can enable it on the public listener if they want to. mox
localserve enables it too. to enable webmail on existing installs, add settings
like the following to the listeners in mox.conf, similar to AccountHTTP(S):
WebmailHTTP:
Enabled: true
WebmailHTTPS:
Enabled: true
special thanks to liesbeth, gerben, andrii for early user feedback.
there is plenty still to do, see the list at the top of webmail/webmail.ts.
feedback welcome as always.
2023-08-07 22:57:03 +03:00
|
|
|
HaveCounts: true,
|
2023-02-16 11:57:27 +03:00
|
|
|
// Do not assign special-use flags. This existing account probably already has such mailboxes.
|
|
|
|
}
|
|
|
|
err = tx.Insert(&mb)
|
|
|
|
ximportcheckf(err, "inserting mailbox in database")
|
|
|
|
|
2023-08-01 11:14:02 +03:00
|
|
|
if tx.Get(&store.Subscription{Name: p}) != nil {
|
|
|
|
err := tx.Insert(&store.Subscription{Name: p})
|
2023-02-17 20:35:11 +03:00
|
|
|
ximportcheckf(err, "subscribing to imported mailbox")
|
|
|
|
}
|
add webmail
it was far down on the roadmap, but implemented earlier, because it's
interesting, and to help prepare for a jmap implementation. for jmap we need to
implement more client-like functionality than with just imap. internal data
structures need to change. jmap has lots of other requirements, so it's already
a big project. by implementing a webmail now, some of the required data
structure changes become clear and can be made now, so the later jmap
implementation can do things similarly to the webmail code. the webmail
frontend and webmail are written together, making their interface/api much
smaller and simpler than jmap.
one of the internal changes is that we now keep track of per-mailbox
total/unread/unseen/deleted message counts and mailbox sizes. keeping this
data consistent after any change to the stored messages (through the code base)
is tricky, so mox now has a consistency check that verifies the counts are
correct, which runs only during tests, each time an internal account reference
is closed. we have a few more internal "changes" that are propagated for the
webmail frontend (that imap doesn't have a way to propagate on a connection),
like changes to the special-use flags on mailboxes, and used keywords in a
mailbox. more changes that will be required have revealed themselves while
implementing the webmail, and will be implemented next.
the webmail user interface is modeled after the mail clients i use or have
used: thunderbird, macos mail, mutt; and webmails i normally only use for
testing: gmail, proton, yahoo, outlook. a somewhat technical user is assumed,
but still the goal is to make this webmail client easy to use for everyone. the
user interface looks like most other mail clients: a list of mailboxes, a
search bar, a message list view, and message details. there is a top/bottom and
a left/right layout for the list/message view, default is automatic based on
screen size. the panes can be resized by the user. buttons for actions are just
text, not icons. clicking a button briefly shows the shortcut for the action in
the bottom right, helping with learning to operate quickly. any text that is
underdotted has a title attribute that causes more information to be displayed,
e.g. what a button does or a field is about. to highlight potential phishing
attempts, any text (anywhere in the webclient) that switches unicode "blocks"
(a rough approximation to (language) scripts) within a word is underlined
orange. multiple messages can be selected with familiar ui interaction:
clicking while holding control and/or shift keys. keyboard navigation works
with arrows/page up/down and home/end keys, and also with a few basic vi-like
keys for list/message navigation. we prefer showing the text instead of
html (with inlined images only) version of a message. html messages are shown
in an iframe served from an endpoint with CSP headers to prevent dangerous
resources (scripts, external images) from being loaded. the html is also
sanitized, with javascript removed. a user can choose to load external
resources (e.g. images for tracking purposes).
the frontend is just (strict) typescript, no external frameworks. all
incoming/outgoing data is typechecked, both the api request parameters and
response types, and the data coming in over SSE. the types and checking code
are generated with sherpats, which uses the api definitions generated by
sherpadoc based on the Go code. so types from the backend are automatically
propagated to the frontend. since there is no framework to automatically
propagate properties and rerender components, changes coming in over the SSE
connection are propagated explicitly with regular function calls. the ui is
separated into "views", each with a "root" dom element that is added to the
visible document. these views have additional functions for getting changes
propagated, often resulting in the view updating its (internal) ui state (dom).
we keep the frontend compilation simple, it's just a few typescript files that
get compiled (combined and types stripped) into a single js file, no additional
runtime code needed or complicated build processes used. the webmail is served
is served from a compressed, cachable html file that includes style and the
javascript, currently just over 225kb uncompressed, under 60kb compressed (not
minified, including comments). we include the generated js files in the
repository, to keep Go's easily buildable self-contained binaries.
authentication is basic http, as with the account and admin pages. most data
comes in over one long-term SSE connection to the backend. api requests signal
which mailbox/search/messages are requested over the SSE connection. fetching
individual messages, and making changes, are done through api calls. the
operations are similar to imap, so some code has been moved from package
imapserver to package store. the future jmap implementation will benefit from
these changes too. more functionality will probably be moved to the store
package in the future.
the quickstart enables webmail on the internal listener by default (for new
installs). users can enable it on the public listener if they want to. mox
localserve enables it too. to enable webmail on existing installs, add settings
like the following to the listeners in mox.conf, similar to AccountHTTP(S):
WebmailHTTP:
Enabled: true
WebmailHTTPS:
Enabled: true
special thanks to liesbeth, gerben, andrii for early user feedback.
there is plenty still to do, see the list at the top of webmail/webmail.ts.
feedback welcome as always.
2023-08-07 22:57:03 +03:00
|
|
|
changes = append(changes, store.ChangeAddMailbox{Mailbox: mb, Flags: []string{`\Subscribed`}})
|
2023-02-16 11:57:27 +03:00
|
|
|
} else if err != nil {
|
|
|
|
ximportcheckf(err, "creating mailbox %s (aborting)", p)
|
|
|
|
}
|
|
|
|
if prevMailbox != "" && mb.Name != prevMailbox {
|
|
|
|
sendEvent("count", importCount{prevMailbox, messages[prevMailbox]})
|
|
|
|
}
|
|
|
|
mailboxes[mb.Name] = mb
|
|
|
|
sendEvent("count", importCount{mb.Name, 0})
|
|
|
|
prevMailbox = mb.Name
|
|
|
|
}
|
|
|
|
return mb
|
|
|
|
}
|
|
|
|
|
|
|
|
xdeliver := func(mb store.Mailbox, m *store.Message, f *os.File, pos string) {
|
|
|
|
defer func() {
|
make mox compile on windows, without "mox serve" but with working "mox localserve"
getting mox to compile required changing code in only a few places where
package "syscall" was used: for accessing file access times and for umask
handling. an open problem is how to start a process as an unprivileged user on
windows. that's why "mox serve" isn't implemented yet. and just finding a way
to implement it now may not be good enough in the near future: we may want to
starting using a more complete privilege separation approach, with a process
handling sensitive tasks (handling private keys, authentication), where we may
want to pass file descriptors between processes. how would that work on
windows?
anyway, getting mox to compile for windows doesn't mean it works properly on
windows. the largest issue: mox would normally open a file, rename or remove
it, and finally close it. this happens during message delivery. that doesn't
work on windows, the rename/remove would fail because the file is still open.
so this commit swaps many "remove" and "close" calls. renames are a longer
story: message delivery had two ways to deliver: with "consuming" the
(temporary) message file (which would rename it to its final destination), and
without consuming (by hardlinking the file, falling back to copying). the last
delivery to a recipient of a message (and the only one in the common case of a
single recipient) would consume the message, and the earlier recipients would
not. during delivery, the already open message file was used, to parse the
message. we still want to use that open message file, and the caller now stays
responsible for closing it, but we no longer try to rename (consume) the file.
we always hardlink (or copy) during delivery (this works on windows), and the
caller is responsible for closing and removing (in that order) the original
temporary file. this does cost one syscall more. but it makes the delivery code
(responsibilities) a bit simpler.
there is one more obvious issue: the file system path separator. mox already
used the "filepath" package to join paths in many places, but not everywhere.
and it still used strings with slashes for local file access. with this commit,
the code now uses filepath.FromSlash for path strings with slashes, uses
"filepath" in a few more places where it previously didn't. also switches from
"filepath" to regular "path" package when handling mailbox names in a few
places, because those always use forward slashes, regardless of local file
system conventions. windows can handle forward slashes when opening files, so
test code that passes path strings with forward slashes straight to go stdlib
file i/o functions are left unchanged to reduce code churn. the regular
non-test code, or test code that uses path strings in places other than
standard i/o functions, does have the paths converted for consistent paths
(otherwise we would end up with paths with mixed forward/backward slashes in
log messages).
windows cannot dup a listening socket. for "mox localserve", it isn't
important, and we can work around the issue. the current approach for "mox
serve" (forking a process and passing file descriptors of listening sockets on
"privileged" ports) won't work on windows. perhaps it isn't needed on windows,
and any user can listen on "privileged" ports? that would be welcome.
on windows, os.Open cannot open a directory, so we cannot call Sync on it after
message delivery. a cursory internet search indicates that directories cannot
be synced on windows. the story is probably much more nuanced than that, with
long deep technical details/discussions/disagreement/confusion, like on unix.
for "mox localserve" we can get away with making syncdir a no-op.
2023-10-14 11:54:07 +03:00
|
|
|
name := f.Name()
|
|
|
|
err = f.Close()
|
|
|
|
log.Check(err, "closing temporary message file for delivery")
|
|
|
|
err := os.Remove(name)
|
|
|
|
log.Check(err, "removing temporary message file for delivery")
|
2023-02-16 11:57:27 +03:00
|
|
|
}()
|
|
|
|
m.MailboxID = mb.ID
|
|
|
|
m.MailboxOrigID = mb.ID
|
|
|
|
|
2023-12-20 22:54:12 +03:00
|
|
|
addSize += m.Size
|
|
|
|
if maxSize > 0 && du.MessageSize+addSize > maxSize {
|
|
|
|
ximportcheckf(fmt.Errorf("account over maximum total size %d", maxSize), "checking quota")
|
|
|
|
}
|
|
|
|
|
2023-07-24 22:21:05 +03:00
|
|
|
if modseq == 0 {
|
|
|
|
var err error
|
|
|
|
modseq, err = acc.NextModSeq(tx)
|
|
|
|
ximportcheckf(err, "assigning next modseq")
|
|
|
|
}
|
|
|
|
m.CreateSeq = modseq
|
|
|
|
m.ModSeq = modseq
|
|
|
|
|
add webmail
it was far down on the roadmap, but implemented earlier, because it's
interesting, and to help prepare for a jmap implementation. for jmap we need to
implement more client-like functionality than with just imap. internal data
structures need to change. jmap has lots of other requirements, so it's already
a big project. by implementing a webmail now, some of the required data
structure changes become clear and can be made now, so the later jmap
implementation can do things similarly to the webmail code. the webmail
frontend and webmail are written together, making their interface/api much
smaller and simpler than jmap.
one of the internal changes is that we now keep track of per-mailbox
total/unread/unseen/deleted message counts and mailbox sizes. keeping this
data consistent after any change to the stored messages (through the code base)
is tricky, so mox now has a consistency check that verifies the counts are
correct, which runs only during tests, each time an internal account reference
is closed. we have a few more internal "changes" that are propagated for the
webmail frontend (that imap doesn't have a way to propagate on a connection),
like changes to the special-use flags on mailboxes, and used keywords in a
mailbox. more changes that will be required have revealed themselves while
implementing the webmail, and will be implemented next.
the webmail user interface is modeled after the mail clients i use or have
used: thunderbird, macos mail, mutt; and webmails i normally only use for
testing: gmail, proton, yahoo, outlook. a somewhat technical user is assumed,
but still the goal is to make this webmail client easy to use for everyone. the
user interface looks like most other mail clients: a list of mailboxes, a
search bar, a message list view, and message details. there is a top/bottom and
a left/right layout for the list/message view, default is automatic based on
screen size. the panes can be resized by the user. buttons for actions are just
text, not icons. clicking a button briefly shows the shortcut for the action in
the bottom right, helping with learning to operate quickly. any text that is
underdotted has a title attribute that causes more information to be displayed,
e.g. what a button does or a field is about. to highlight potential phishing
attempts, any text (anywhere in the webclient) that switches unicode "blocks"
(a rough approximation to (language) scripts) within a word is underlined
orange. multiple messages can be selected with familiar ui interaction:
clicking while holding control and/or shift keys. keyboard navigation works
with arrows/page up/down and home/end keys, and also with a few basic vi-like
keys for list/message navigation. we prefer showing the text instead of
html (with inlined images only) version of a message. html messages are shown
in an iframe served from an endpoint with CSP headers to prevent dangerous
resources (scripts, external images) from being loaded. the html is also
sanitized, with javascript removed. a user can choose to load external
resources (e.g. images for tracking purposes).
the frontend is just (strict) typescript, no external frameworks. all
incoming/outgoing data is typechecked, both the api request parameters and
response types, and the data coming in over SSE. the types and checking code
are generated with sherpats, which uses the api definitions generated by
sherpadoc based on the Go code. so types from the backend are automatically
propagated to the frontend. since there is no framework to automatically
propagate properties and rerender components, changes coming in over the SSE
connection are propagated explicitly with regular function calls. the ui is
separated into "views", each with a "root" dom element that is added to the
visible document. these views have additional functions for getting changes
propagated, often resulting in the view updating its (internal) ui state (dom).
we keep the frontend compilation simple, it's just a few typescript files that
get compiled (combined and types stripped) into a single js file, no additional
runtime code needed or complicated build processes used. the webmail is served
is served from a compressed, cachable html file that includes style and the
javascript, currently just over 225kb uncompressed, under 60kb compressed (not
minified, including comments). we include the generated js files in the
repository, to keep Go's easily buildable self-contained binaries.
authentication is basic http, as with the account and admin pages. most data
comes in over one long-term SSE connection to the backend. api requests signal
which mailbox/search/messages are requested over the SSE connection. fetching
individual messages, and making changes, are done through api calls. the
operations are similar to imap, so some code has been moved from package
imapserver to package store. the future jmap implementation will benefit from
these changes too. more functionality will probably be moved to the store
package in the future.
the quickstart enables webmail on the internal listener by default (for new
installs). users can enable it on the public listener if they want to. mox
localserve enables it too. to enable webmail on existing installs, add settings
like the following to the listeners in mox.conf, similar to AccountHTTP(S):
WebmailHTTP:
Enabled: true
WebmailHTTPS:
Enabled: true
special thanks to liesbeth, gerben, andrii for early user feedback.
there is plenty still to do, see the list at the top of webmail/webmail.ts.
feedback welcome as always.
2023-08-07 22:57:03 +03:00
|
|
|
mc := destMailboxCounts[mb.ID]
|
|
|
|
mc.Add(m.MailboxCounts())
|
|
|
|
destMailboxCounts[mb.ID] = mc
|
|
|
|
|
2023-06-24 01:24:43 +03:00
|
|
|
if len(m.Keywords) > 0 {
|
|
|
|
if destMailboxKeywords[mb.ID] == nil {
|
|
|
|
destMailboxKeywords[mb.ID] = map[string]bool{}
|
|
|
|
}
|
|
|
|
for _, k := range m.Keywords {
|
|
|
|
destMailboxKeywords[mb.ID][k] = true
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
2023-02-16 11:57:27 +03:00
|
|
|
// Parse message and store parsed information for later fast retrieval.
|
2023-12-05 15:35:58 +03:00
|
|
|
p, err := message.EnsurePart(log.Logger, false, f, m.Size)
|
2023-02-16 11:57:27 +03:00
|
|
|
if err != nil {
|
|
|
|
problemf("parsing message %s: %s (continuing)", pos, err)
|
|
|
|
}
|
|
|
|
m.ParsedBuf, err = json.Marshal(p)
|
|
|
|
ximportcheckf(err, "marshal parsed message structure")
|
|
|
|
|
implement message threading in backend and webmail
we match messages to their parents based on the "references" and "in-reply-to"
headers (requiring the same base subject), and in absense of those headers we
also by only base subject (against messages received max 4 weeks ago).
we store a threadid with messages. all messages in a thread have the same
threadid. messages also have a "thread parent ids", which holds all id's of
parent messages up to the thread root. then there is "thread missing link",
which is set when a referenced immediate parent wasn't found (but possibly
earlier ancestors can still be found and will be in thread parent ids".
threads can be muted: newly delivered messages are automatically marked as
read/seen. threads can be marked as collapsed: if set, the webmail collapses
the thread to a single item in the basic threading view (default is to expand
threads). the muted and collapsed fields are copied from their parent on
message delivery.
the threading is implemented in the webmail. the non-threading mode still works
as before. the new default threading mode "unread" automatically expands only
the threads with at least one unread (not seen) meessage. the basic threading
mode "on" expands all threads except when explicitly collapsed (as saved in the
thread collapsed field). new shortcuts for navigation/interaction threads have
been added, e.g. go to previous/next thread root, toggle collapse/expand of
thread (or double click), toggle mute of thread. some previous shortcuts have
changed, see the help for details.
the message threading are added with an explicit account upgrade step,
automatically started when an account is opened. the upgrade is done in the
background because it will take too long for large mailboxes to block account
operations. the upgrade takes two steps: 1. updating all message records in the
database to add a normalized message-id and thread base subject (with "re:",
"fwd:" and several other schemes stripped). 2. going through all messages in
the database again, reading the "references" and "in-reply-to" headers from
disk, and matching against their parents. this second step is also done at the
end of each import of mbox/maildir mailboxes. new deliveries are matched
immediately against other existing messages, currently no attempt is made to
rematch previously delivered messages (which could be useful for related
messages being delivered out of order).
the threading is not yet exposed over imap.
2023-09-13 09:51:50 +03:00
|
|
|
// Set fields needed for future threading. By doing it now, DeliverMessage won't
|
|
|
|
// have to parse the Part again.
|
|
|
|
p.SetReaderAt(store.FileMsgReader(m.MsgPrefix, f))
|
|
|
|
m.PrepareThreading(log, &p)
|
|
|
|
|
2023-02-16 11:57:27 +03:00
|
|
|
if m.Received.IsZero() {
|
|
|
|
if p.Envelope != nil && !p.Envelope.Date.IsZero() {
|
|
|
|
m.Received = p.Envelope.Date
|
|
|
|
} else {
|
|
|
|
m.Received = time.Now()
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
|
|
|
// We set the flags that Deliver would set now and train ourselves. This prevents
|
|
|
|
// Deliver from training, which would open the junk filter, change it, and write it
|
|
|
|
// back to disk, for each message (slow).
|
2023-09-21 16:19:11 +03:00
|
|
|
m.JunkFlagsForMailbox(mb, conf)
|
2023-02-16 11:57:27 +03:00
|
|
|
if jf != nil && m.NeedsTraining() {
|
|
|
|
trainMessage(m, p, pos)
|
|
|
|
}
|
|
|
|
|
|
|
|
const sync = false
|
|
|
|
const notrain = true
|
implement message threading in backend and webmail
we match messages to their parents based on the "references" and "in-reply-to"
headers (requiring the same base subject), and in absense of those headers we
also by only base subject (against messages received max 4 weeks ago).
we store a threadid with messages. all messages in a thread have the same
threadid. messages also have a "thread parent ids", which holds all id's of
parent messages up to the thread root. then there is "thread missing link",
which is set when a referenced immediate parent wasn't found (but possibly
earlier ancestors can still be found and will be in thread parent ids".
threads can be muted: newly delivered messages are automatically marked as
read/seen. threads can be marked as collapsed: if set, the webmail collapses
the thread to a single item in the basic threading view (default is to expand
threads). the muted and collapsed fields are copied from their parent on
message delivery.
the threading is implemented in the webmail. the non-threading mode still works
as before. the new default threading mode "unread" automatically expands only
the threads with at least one unread (not seen) meessage. the basic threading
mode "on" expands all threads except when explicitly collapsed (as saved in the
thread collapsed field). new shortcuts for navigation/interaction threads have
been added, e.g. go to previous/next thread root, toggle collapse/expand of
thread (or double click), toggle mute of thread. some previous shortcuts have
changed, see the help for details.
the message threading are added with an explicit account upgrade step,
automatically started when an account is opened. the upgrade is done in the
background because it will take too long for large mailboxes to block account
operations. the upgrade takes two steps: 1. updating all message records in the
database to add a normalized message-id and thread base subject (with "re:",
"fwd:" and several other schemes stripped). 2. going through all messages in
the database again, reading the "references" and "in-reply-to" headers from
disk, and matching against their parents. this second step is also done at the
end of each import of mbox/maildir mailboxes. new deliveries are matched
immediately against other existing messages, currently no attempt is made to
rematch previously delivered messages (which could be useful for related
messages being delivered out of order).
the threading is not yet exposed over imap.
2023-09-13 09:51:50 +03:00
|
|
|
const nothreads = true
|
2023-12-20 22:54:12 +03:00
|
|
|
const updateDiskUsage = false
|
|
|
|
if err := acc.DeliverMessage(log, tx, m, f, sync, notrain, nothreads, updateDiskUsage); err != nil {
|
2023-04-20 15:16:56 +03:00
|
|
|
problemf("delivering message %s: %s (continuing)", pos, err)
|
|
|
|
return
|
|
|
|
}
|
2023-02-16 11:57:27 +03:00
|
|
|
deliveredIDs = append(deliveredIDs, m.ID)
|
add webmail
it was far down on the roadmap, but implemented earlier, because it's
interesting, and to help prepare for a jmap implementation. for jmap we need to
implement more client-like functionality than with just imap. internal data
structures need to change. jmap has lots of other requirements, so it's already
a big project. by implementing a webmail now, some of the required data
structure changes become clear and can be made now, so the later jmap
implementation can do things similarly to the webmail code. the webmail
frontend and webmail are written together, making their interface/api much
smaller and simpler than jmap.
one of the internal changes is that we now keep track of per-mailbox
total/unread/unseen/deleted message counts and mailbox sizes. keeping this
data consistent after any change to the stored messages (through the code base)
is tricky, so mox now has a consistency check that verifies the counts are
correct, which runs only during tests, each time an internal account reference
is closed. we have a few more internal "changes" that are propagated for the
webmail frontend (that imap doesn't have a way to propagate on a connection),
like changes to the special-use flags on mailboxes, and used keywords in a
mailbox. more changes that will be required have revealed themselves while
implementing the webmail, and will be implemented next.
the webmail user interface is modeled after the mail clients i use or have
used: thunderbird, macos mail, mutt; and webmails i normally only use for
testing: gmail, proton, yahoo, outlook. a somewhat technical user is assumed,
but still the goal is to make this webmail client easy to use for everyone. the
user interface looks like most other mail clients: a list of mailboxes, a
search bar, a message list view, and message details. there is a top/bottom and
a left/right layout for the list/message view, default is automatic based on
screen size. the panes can be resized by the user. buttons for actions are just
text, not icons. clicking a button briefly shows the shortcut for the action in
the bottom right, helping with learning to operate quickly. any text that is
underdotted has a title attribute that causes more information to be displayed,
e.g. what a button does or a field is about. to highlight potential phishing
attempts, any text (anywhere in the webclient) that switches unicode "blocks"
(a rough approximation to (language) scripts) within a word is underlined
orange. multiple messages can be selected with familiar ui interaction:
clicking while holding control and/or shift keys. keyboard navigation works
with arrows/page up/down and home/end keys, and also with a few basic vi-like
keys for list/message navigation. we prefer showing the text instead of
html (with inlined images only) version of a message. html messages are shown
in an iframe served from an endpoint with CSP headers to prevent dangerous
resources (scripts, external images) from being loaded. the html is also
sanitized, with javascript removed. a user can choose to load external
resources (e.g. images for tracking purposes).
the frontend is just (strict) typescript, no external frameworks. all
incoming/outgoing data is typechecked, both the api request parameters and
response types, and the data coming in over SSE. the types and checking code
are generated with sherpats, which uses the api definitions generated by
sherpadoc based on the Go code. so types from the backend are automatically
propagated to the frontend. since there is no framework to automatically
propagate properties and rerender components, changes coming in over the SSE
connection are propagated explicitly with regular function calls. the ui is
separated into "views", each with a "root" dom element that is added to the
visible document. these views have additional functions for getting changes
propagated, often resulting in the view updating its (internal) ui state (dom).
we keep the frontend compilation simple, it's just a few typescript files that
get compiled (combined and types stripped) into a single js file, no additional
runtime code needed or complicated build processes used. the webmail is served
is served from a compressed, cachable html file that includes style and the
javascript, currently just over 225kb uncompressed, under 60kb compressed (not
minified, including comments). we include the generated js files in the
repository, to keep Go's easily buildable self-contained binaries.
authentication is basic http, as with the account and admin pages. most data
comes in over one long-term SSE connection to the backend. api requests signal
which mailbox/search/messages are requested over the SSE connection. fetching
individual messages, and making changes, are done through api calls. the
operations are similar to imap, so some code has been moved from package
imapserver to package store. the future jmap implementation will benefit from
these changes too. more functionality will probably be moved to the store
package in the future.
the quickstart enables webmail on the internal listener by default (for new
installs). users can enable it on the public listener if they want to. mox
localserve enables it too. to enable webmail on existing installs, add settings
like the following to the listeners in mox.conf, similar to AccountHTTP(S):
WebmailHTTP:
Enabled: true
WebmailHTTPS:
Enabled: true
special thanks to liesbeth, gerben, andrii for early user feedback.
there is plenty still to do, see the list at the top of webmail/webmail.ts.
feedback welcome as always.
2023-08-07 22:57:03 +03:00
|
|
|
changes = append(changes, m.ChangeAddUID())
|
2023-02-16 11:57:27 +03:00
|
|
|
messages[mb.Name]++
|
|
|
|
if messages[mb.Name]%100 == 0 || prevMailbox != mb.Name {
|
|
|
|
prevMailbox = mb.Name
|
|
|
|
sendEvent("count", importCount{mb.Name, messages[mb.Name]})
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
|
|
|
ximportMbox := func(mailbox, filename string, r io.Reader) {
|
|
|
|
if mailbox == "" {
|
|
|
|
problemf("empty mailbox name for mbox file %s (skipping)", filename)
|
|
|
|
return
|
|
|
|
}
|
|
|
|
mb := xensureMailbox(mailbox)
|
|
|
|
|
2023-12-05 15:35:58 +03:00
|
|
|
mr := store.NewMboxReader(log, store.CreateMessageTemp, filename, r)
|
2023-02-16 11:57:27 +03:00
|
|
|
for {
|
|
|
|
m, mf, pos, err := mr.Next()
|
|
|
|
if err == io.EOF {
|
|
|
|
break
|
|
|
|
} else if err != nil {
|
|
|
|
ximportcheckf(err, "next message in mbox file")
|
|
|
|
}
|
|
|
|
|
|
|
|
xdeliver(mb, m, mf, pos)
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
|
|
|
ximportMaildir := func(mailbox, filename string, r io.Reader) {
|
|
|
|
if mailbox == "" {
|
|
|
|
problemf("empty mailbox name for maildir file %s (skipping)", filename)
|
|
|
|
return
|
|
|
|
}
|
|
|
|
mb := xensureMailbox(mailbox)
|
|
|
|
|
2023-12-05 15:35:58 +03:00
|
|
|
f, err := store.CreateMessageTemp(log, "import")
|
2023-02-16 11:57:27 +03:00
|
|
|
ximportcheckf(err, "creating temp message")
|
|
|
|
defer func() {
|
|
|
|
if f != nil {
|
2023-11-01 20:57:38 +03:00
|
|
|
store.CloseRemoveTempFile(log, f, "message to import")
|
2023-02-16 11:57:27 +03:00
|
|
|
}
|
|
|
|
}()
|
|
|
|
|
|
|
|
// Copy data, changing bare \n into \r\n.
|
|
|
|
br := bufio.NewReader(r)
|
|
|
|
w := bufio.NewWriter(f)
|
|
|
|
var size int64
|
|
|
|
for {
|
|
|
|
line, err := br.ReadBytes('\n')
|
|
|
|
if err != nil && err != io.EOF {
|
|
|
|
ximportcheckf(err, "reading message")
|
|
|
|
}
|
|
|
|
if len(line) > 0 {
|
|
|
|
if !bytes.HasSuffix(line, []byte("\r\n")) {
|
|
|
|
line = append(line[:len(line)-1], "\r\n"...)
|
|
|
|
}
|
|
|
|
|
|
|
|
n, err := w.Write(line)
|
|
|
|
ximportcheckf(err, "writing message")
|
|
|
|
size += int64(n)
|
|
|
|
}
|
|
|
|
if err == io.EOF {
|
|
|
|
break
|
|
|
|
}
|
|
|
|
}
|
|
|
|
err = w.Flush()
|
|
|
|
ximportcheckf(err, "writing message")
|
|
|
|
|
|
|
|
var received time.Time
|
|
|
|
t := strings.SplitN(path.Base(filename), ".", 2)
|
|
|
|
if v, err := strconv.ParseInt(t[0], 10, 64); err == nil {
|
|
|
|
received = time.Unix(v, 0)
|
|
|
|
}
|
|
|
|
|
|
|
|
// Parse flags. See https://cr.yp.to/proto/maildir.html.
|
|
|
|
var keepFlags string
|
2023-06-24 01:24:43 +03:00
|
|
|
var flags store.Flags
|
|
|
|
keywords := map[string]bool{}
|
2023-02-16 11:57:27 +03:00
|
|
|
t = strings.SplitN(path.Base(filename), ":2,", 2)
|
|
|
|
if len(t) == 2 {
|
|
|
|
for _, c := range t[1] {
|
|
|
|
switch c {
|
|
|
|
case 'P':
|
|
|
|
// Passed, doesn't map to a common IMAP flag.
|
|
|
|
case 'R':
|
|
|
|
flags.Answered = true
|
|
|
|
case 'S':
|
|
|
|
flags.Seen = true
|
|
|
|
case 'T':
|
|
|
|
flags.Deleted = true
|
|
|
|
case 'D':
|
|
|
|
flags.Draft = true
|
|
|
|
case 'F':
|
|
|
|
flags.Flagged = true
|
|
|
|
default:
|
|
|
|
if c >= 'a' && c <= 'z' {
|
2023-06-24 01:24:43 +03:00
|
|
|
dovecotKeywords, ok := mailboxKeywords[mailbox]
|
2023-02-16 11:57:27 +03:00
|
|
|
if !ok {
|
|
|
|
// No keywords file seen yet, we'll try later if it comes in.
|
|
|
|
keepFlags += string(c)
|
2023-06-24 01:24:43 +03:00
|
|
|
} else if kw, ok := dovecotKeywords[c]; ok {
|
add webmail
it was far down on the roadmap, but implemented earlier, because it's
interesting, and to help prepare for a jmap implementation. for jmap we need to
implement more client-like functionality than with just imap. internal data
structures need to change. jmap has lots of other requirements, so it's already
a big project. by implementing a webmail now, some of the required data
structure changes become clear and can be made now, so the later jmap
implementation can do things similarly to the webmail code. the webmail
frontend and webmail are written together, making their interface/api much
smaller and simpler than jmap.
one of the internal changes is that we now keep track of per-mailbox
total/unread/unseen/deleted message counts and mailbox sizes. keeping this
data consistent after any change to the stored messages (through the code base)
is tricky, so mox now has a consistency check that verifies the counts are
correct, which runs only during tests, each time an internal account reference
is closed. we have a few more internal "changes" that are propagated for the
webmail frontend (that imap doesn't have a way to propagate on a connection),
like changes to the special-use flags on mailboxes, and used keywords in a
mailbox. more changes that will be required have revealed themselves while
implementing the webmail, and will be implemented next.
the webmail user interface is modeled after the mail clients i use or have
used: thunderbird, macos mail, mutt; and webmails i normally only use for
testing: gmail, proton, yahoo, outlook. a somewhat technical user is assumed,
but still the goal is to make this webmail client easy to use for everyone. the
user interface looks like most other mail clients: a list of mailboxes, a
search bar, a message list view, and message details. there is a top/bottom and
a left/right layout for the list/message view, default is automatic based on
screen size. the panes can be resized by the user. buttons for actions are just
text, not icons. clicking a button briefly shows the shortcut for the action in
the bottom right, helping with learning to operate quickly. any text that is
underdotted has a title attribute that causes more information to be displayed,
e.g. what a button does or a field is about. to highlight potential phishing
attempts, any text (anywhere in the webclient) that switches unicode "blocks"
(a rough approximation to (language) scripts) within a word is underlined
orange. multiple messages can be selected with familiar ui interaction:
clicking while holding control and/or shift keys. keyboard navigation works
with arrows/page up/down and home/end keys, and also with a few basic vi-like
keys for list/message navigation. we prefer showing the text instead of
html (with inlined images only) version of a message. html messages are shown
in an iframe served from an endpoint with CSP headers to prevent dangerous
resources (scripts, external images) from being loaded. the html is also
sanitized, with javascript removed. a user can choose to load external
resources (e.g. images for tracking purposes).
the frontend is just (strict) typescript, no external frameworks. all
incoming/outgoing data is typechecked, both the api request parameters and
response types, and the data coming in over SSE. the types and checking code
are generated with sherpats, which uses the api definitions generated by
sherpadoc based on the Go code. so types from the backend are automatically
propagated to the frontend. since there is no framework to automatically
propagate properties and rerender components, changes coming in over the SSE
connection are propagated explicitly with regular function calls. the ui is
separated into "views", each with a "root" dom element that is added to the
visible document. these views have additional functions for getting changes
propagated, often resulting in the view updating its (internal) ui state (dom).
we keep the frontend compilation simple, it's just a few typescript files that
get compiled (combined and types stripped) into a single js file, no additional
runtime code needed or complicated build processes used. the webmail is served
is served from a compressed, cachable html file that includes style and the
javascript, currently just over 225kb uncompressed, under 60kb compressed (not
minified, including comments). we include the generated js files in the
repository, to keep Go's easily buildable self-contained binaries.
authentication is basic http, as with the account and admin pages. most data
comes in over one long-term SSE connection to the backend. api requests signal
which mailbox/search/messages are requested over the SSE connection. fetching
individual messages, and making changes, are done through api calls. the
operations are similar to imap, so some code has been moved from package
imapserver to package store. the future jmap implementation will benefit from
these changes too. more functionality will probably be moved to the store
package in the future.
the quickstart enables webmail on the internal listener by default (for new
installs). users can enable it on the public listener if they want to. mox
localserve enables it too. to enable webmail on existing installs, add settings
like the following to the listeners in mox.conf, similar to AccountHTTP(S):
WebmailHTTP:
Enabled: true
WebmailHTTPS:
Enabled: true
special thanks to liesbeth, gerben, andrii for early user feedback.
there is plenty still to do, see the list at the top of webmail/webmail.ts.
feedback welcome as always.
2023-08-07 22:57:03 +03:00
|
|
|
flagSet(&flags, keywords, kw)
|
2023-02-16 11:57:27 +03:00
|
|
|
}
|
|
|
|
}
|
|
|
|
}
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
|
|
|
m := store.Message{
|
|
|
|
Received: received,
|
|
|
|
Flags: flags,
|
2023-06-24 01:24:43 +03:00
|
|
|
Keywords: maps.Keys(keywords),
|
2023-02-16 11:57:27 +03:00
|
|
|
Size: size,
|
|
|
|
}
|
|
|
|
xdeliver(mb, &m, f, filename)
|
|
|
|
f = nil
|
|
|
|
if keepFlags != "" {
|
|
|
|
if _, ok := mailboxMissingKeywordMessages[mailbox]; !ok {
|
|
|
|
mailboxMissingKeywordMessages[mailbox] = map[int64]string{}
|
|
|
|
}
|
|
|
|
mailboxMissingKeywordMessages[mailbox][m.ID] = keepFlags
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
|
|
|
importFile := func(name string, r io.Reader) {
|
|
|
|
origName := name
|
|
|
|
|
|
|
|
if strings.HasPrefix(name, skipMailboxPrefix) {
|
|
|
|
name = strings.TrimPrefix(name[len(skipMailboxPrefix):], "/")
|
|
|
|
}
|
|
|
|
|
|
|
|
if strings.HasSuffix(name, "/") {
|
|
|
|
name = strings.TrimSuffix(name, "/")
|
|
|
|
dir := path.Dir(name)
|
|
|
|
switch path.Base(dir) {
|
|
|
|
case "new", "cur", "tmp":
|
|
|
|
// Maildir, ensure it exists.
|
|
|
|
mailbox := path.Dir(dir)
|
|
|
|
xensureMailbox(mailbox)
|
|
|
|
}
|
|
|
|
// Otherwise, this is just a directory that probably holds mbox files and maildirs.
|
|
|
|
return
|
|
|
|
}
|
|
|
|
|
|
|
|
if strings.HasSuffix(path.Base(name), ".mbox") {
|
|
|
|
mailbox := name[:len(name)-len(".mbox")]
|
|
|
|
ximportMbox(mailbox, origName, r)
|
|
|
|
return
|
|
|
|
}
|
|
|
|
dir := path.Dir(name)
|
|
|
|
dirbase := path.Base(dir)
|
|
|
|
switch dirbase {
|
|
|
|
case "new", "cur", "tmp":
|
|
|
|
mailbox := path.Dir(dir)
|
|
|
|
ximportMaildir(mailbox, origName, r)
|
|
|
|
default:
|
|
|
|
if path.Base(name) == "dovecot-keywords" {
|
|
|
|
mailbox := path.Dir(name)
|
2023-06-24 01:24:43 +03:00
|
|
|
dovecotKeywords := map[rune]string{}
|
add webmail
it was far down on the roadmap, but implemented earlier, because it's
interesting, and to help prepare for a jmap implementation. for jmap we need to
implement more client-like functionality than with just imap. internal data
structures need to change. jmap has lots of other requirements, so it's already
a big project. by implementing a webmail now, some of the required data
structure changes become clear and can be made now, so the later jmap
implementation can do things similarly to the webmail code. the webmail
frontend and webmail are written together, making their interface/api much
smaller and simpler than jmap.
one of the internal changes is that we now keep track of per-mailbox
total/unread/unseen/deleted message counts and mailbox sizes. keeping this
data consistent after any change to the stored messages (through the code base)
is tricky, so mox now has a consistency check that verifies the counts are
correct, which runs only during tests, each time an internal account reference
is closed. we have a few more internal "changes" that are propagated for the
webmail frontend (that imap doesn't have a way to propagate on a connection),
like changes to the special-use flags on mailboxes, and used keywords in a
mailbox. more changes that will be required have revealed themselves while
implementing the webmail, and will be implemented next.
the webmail user interface is modeled after the mail clients i use or have
used: thunderbird, macos mail, mutt; and webmails i normally only use for
testing: gmail, proton, yahoo, outlook. a somewhat technical user is assumed,
but still the goal is to make this webmail client easy to use for everyone. the
user interface looks like most other mail clients: a list of mailboxes, a
search bar, a message list view, and message details. there is a top/bottom and
a left/right layout for the list/message view, default is automatic based on
screen size. the panes can be resized by the user. buttons for actions are just
text, not icons. clicking a button briefly shows the shortcut for the action in
the bottom right, helping with learning to operate quickly. any text that is
underdotted has a title attribute that causes more information to be displayed,
e.g. what a button does or a field is about. to highlight potential phishing
attempts, any text (anywhere in the webclient) that switches unicode "blocks"
(a rough approximation to (language) scripts) within a word is underlined
orange. multiple messages can be selected with familiar ui interaction:
clicking while holding control and/or shift keys. keyboard navigation works
with arrows/page up/down and home/end keys, and also with a few basic vi-like
keys for list/message navigation. we prefer showing the text instead of
html (with inlined images only) version of a message. html messages are shown
in an iframe served from an endpoint with CSP headers to prevent dangerous
resources (scripts, external images) from being loaded. the html is also
sanitized, with javascript removed. a user can choose to load external
resources (e.g. images for tracking purposes).
the frontend is just (strict) typescript, no external frameworks. all
incoming/outgoing data is typechecked, both the api request parameters and
response types, and the data coming in over SSE. the types and checking code
are generated with sherpats, which uses the api definitions generated by
sherpadoc based on the Go code. so types from the backend are automatically
propagated to the frontend. since there is no framework to automatically
propagate properties and rerender components, changes coming in over the SSE
connection are propagated explicitly with regular function calls. the ui is
separated into "views", each with a "root" dom element that is added to the
visible document. these views have additional functions for getting changes
propagated, often resulting in the view updating its (internal) ui state (dom).
we keep the frontend compilation simple, it's just a few typescript files that
get compiled (combined and types stripped) into a single js file, no additional
runtime code needed or complicated build processes used. the webmail is served
is served from a compressed, cachable html file that includes style and the
javascript, currently just over 225kb uncompressed, under 60kb compressed (not
minified, including comments). we include the generated js files in the
repository, to keep Go's easily buildable self-contained binaries.
authentication is basic http, as with the account and admin pages. most data
comes in over one long-term SSE connection to the backend. api requests signal
which mailbox/search/messages are requested over the SSE connection. fetching
individual messages, and making changes, are done through api calls. the
operations are similar to imap, so some code has been moved from package
imapserver to package store. the future jmap implementation will benefit from
these changes too. more functionality will probably be moved to the store
package in the future.
the quickstart enables webmail on the internal listener by default (for new
installs). users can enable it on the public listener if they want to. mox
localserve enables it too. to enable webmail on existing installs, add settings
like the following to the listeners in mox.conf, similar to AccountHTTP(S):
WebmailHTTP:
Enabled: true
WebmailHTTPS:
Enabled: true
special thanks to liesbeth, gerben, andrii for early user feedback.
there is plenty still to do, see the list at the top of webmail/webmail.ts.
feedback welcome as always.
2023-08-07 22:57:03 +03:00
|
|
|
words, err := store.ParseDovecotKeywordsFlags(r, log)
|
2023-12-05 15:35:58 +03:00
|
|
|
log.Check(err, "parsing dovecot keywords for mailbox", slog.String("mailbox", mailbox))
|
2023-02-16 11:57:27 +03:00
|
|
|
for i, kw := range words {
|
2023-06-24 01:24:43 +03:00
|
|
|
dovecotKeywords['a'+rune(i)] = kw
|
2023-02-16 11:57:27 +03:00
|
|
|
}
|
2023-06-24 01:24:43 +03:00
|
|
|
mailboxKeywords[mailbox] = dovecotKeywords
|
2023-02-16 11:57:27 +03:00
|
|
|
|
|
|
|
for id, chars := range mailboxMissingKeywordMessages[mailbox] {
|
|
|
|
var flags, zeroflags store.Flags
|
2023-06-24 01:24:43 +03:00
|
|
|
keywords := map[string]bool{}
|
2023-02-16 11:57:27 +03:00
|
|
|
for _, c := range chars {
|
2023-06-24 01:24:43 +03:00
|
|
|
kw, ok := dovecotKeywords[c]
|
2023-02-16 11:57:27 +03:00
|
|
|
if !ok {
|
2023-06-24 01:24:43 +03:00
|
|
|
problemf("unspecified dovecot message flag %c for message id %d (continuing)", c, id)
|
2023-02-16 11:57:27 +03:00
|
|
|
continue
|
|
|
|
}
|
add webmail
it was far down on the roadmap, but implemented earlier, because it's
interesting, and to help prepare for a jmap implementation. for jmap we need to
implement more client-like functionality than with just imap. internal data
structures need to change. jmap has lots of other requirements, so it's already
a big project. by implementing a webmail now, some of the required data
structure changes become clear and can be made now, so the later jmap
implementation can do things similarly to the webmail code. the webmail
frontend and webmail are written together, making their interface/api much
smaller and simpler than jmap.
one of the internal changes is that we now keep track of per-mailbox
total/unread/unseen/deleted message counts and mailbox sizes. keeping this
data consistent after any change to the stored messages (through the code base)
is tricky, so mox now has a consistency check that verifies the counts are
correct, which runs only during tests, each time an internal account reference
is closed. we have a few more internal "changes" that are propagated for the
webmail frontend (that imap doesn't have a way to propagate on a connection),
like changes to the special-use flags on mailboxes, and used keywords in a
mailbox. more changes that will be required have revealed themselves while
implementing the webmail, and will be implemented next.
the webmail user interface is modeled after the mail clients i use or have
used: thunderbird, macos mail, mutt; and webmails i normally only use for
testing: gmail, proton, yahoo, outlook. a somewhat technical user is assumed,
but still the goal is to make this webmail client easy to use for everyone. the
user interface looks like most other mail clients: a list of mailboxes, a
search bar, a message list view, and message details. there is a top/bottom and
a left/right layout for the list/message view, default is automatic based on
screen size. the panes can be resized by the user. buttons for actions are just
text, not icons. clicking a button briefly shows the shortcut for the action in
the bottom right, helping with learning to operate quickly. any text that is
underdotted has a title attribute that causes more information to be displayed,
e.g. what a button does or a field is about. to highlight potential phishing
attempts, any text (anywhere in the webclient) that switches unicode "blocks"
(a rough approximation to (language) scripts) within a word is underlined
orange. multiple messages can be selected with familiar ui interaction:
clicking while holding control and/or shift keys. keyboard navigation works
with arrows/page up/down and home/end keys, and also with a few basic vi-like
keys for list/message navigation. we prefer showing the text instead of
html (with inlined images only) version of a message. html messages are shown
in an iframe served from an endpoint with CSP headers to prevent dangerous
resources (scripts, external images) from being loaded. the html is also
sanitized, with javascript removed. a user can choose to load external
resources (e.g. images for tracking purposes).
the frontend is just (strict) typescript, no external frameworks. all
incoming/outgoing data is typechecked, both the api request parameters and
response types, and the data coming in over SSE. the types and checking code
are generated with sherpats, which uses the api definitions generated by
sherpadoc based on the Go code. so types from the backend are automatically
propagated to the frontend. since there is no framework to automatically
propagate properties and rerender components, changes coming in over the SSE
connection are propagated explicitly with regular function calls. the ui is
separated into "views", each with a "root" dom element that is added to the
visible document. these views have additional functions for getting changes
propagated, often resulting in the view updating its (internal) ui state (dom).
we keep the frontend compilation simple, it's just a few typescript files that
get compiled (combined and types stripped) into a single js file, no additional
runtime code needed or complicated build processes used. the webmail is served
is served from a compressed, cachable html file that includes style and the
javascript, currently just over 225kb uncompressed, under 60kb compressed (not
minified, including comments). we include the generated js files in the
repository, to keep Go's easily buildable self-contained binaries.
authentication is basic http, as with the account and admin pages. most data
comes in over one long-term SSE connection to the backend. api requests signal
which mailbox/search/messages are requested over the SSE connection. fetching
individual messages, and making changes, are done through api calls. the
operations are similar to imap, so some code has been moved from package
imapserver to package store. the future jmap implementation will benefit from
these changes too. more functionality will probably be moved to the store
package in the future.
the quickstart enables webmail on the internal listener by default (for new
installs). users can enable it on the public listener if they want to. mox
localserve enables it too. to enable webmail on existing installs, add settings
like the following to the listeners in mox.conf, similar to AccountHTTP(S):
WebmailHTTP:
Enabled: true
WebmailHTTPS:
Enabled: true
special thanks to liesbeth, gerben, andrii for early user feedback.
there is plenty still to do, see the list at the top of webmail/webmail.ts.
feedback welcome as always.
2023-08-07 22:57:03 +03:00
|
|
|
flagSet(&flags, keywords, kw)
|
2023-02-16 11:57:27 +03:00
|
|
|
}
|
2023-06-24 01:24:43 +03:00
|
|
|
if flags == zeroflags && len(keywords) == 0 {
|
2023-02-16 11:57:27 +03:00
|
|
|
continue
|
|
|
|
}
|
2023-06-24 01:24:43 +03:00
|
|
|
|
2023-02-16 11:57:27 +03:00
|
|
|
m := store.Message{ID: id}
|
|
|
|
err := tx.Get(&m)
|
|
|
|
ximportcheckf(err, "get imported message for flag update")
|
2023-06-24 01:24:43 +03:00
|
|
|
|
add webmail
it was far down on the roadmap, but implemented earlier, because it's
interesting, and to help prepare for a jmap implementation. for jmap we need to
implement more client-like functionality than with just imap. internal data
structures need to change. jmap has lots of other requirements, so it's already
a big project. by implementing a webmail now, some of the required data
structure changes become clear and can be made now, so the later jmap
implementation can do things similarly to the webmail code. the webmail
frontend and webmail are written together, making their interface/api much
smaller and simpler than jmap.
one of the internal changes is that we now keep track of per-mailbox
total/unread/unseen/deleted message counts and mailbox sizes. keeping this
data consistent after any change to the stored messages (through the code base)
is tricky, so mox now has a consistency check that verifies the counts are
correct, which runs only during tests, each time an internal account reference
is closed. we have a few more internal "changes" that are propagated for the
webmail frontend (that imap doesn't have a way to propagate on a connection),
like changes to the special-use flags on mailboxes, and used keywords in a
mailbox. more changes that will be required have revealed themselves while
implementing the webmail, and will be implemented next.
the webmail user interface is modeled after the mail clients i use or have
used: thunderbird, macos mail, mutt; and webmails i normally only use for
testing: gmail, proton, yahoo, outlook. a somewhat technical user is assumed,
but still the goal is to make this webmail client easy to use for everyone. the
user interface looks like most other mail clients: a list of mailboxes, a
search bar, a message list view, and message details. there is a top/bottom and
a left/right layout for the list/message view, default is automatic based on
screen size. the panes can be resized by the user. buttons for actions are just
text, not icons. clicking a button briefly shows the shortcut for the action in
the bottom right, helping with learning to operate quickly. any text that is
underdotted has a title attribute that causes more information to be displayed,
e.g. what a button does or a field is about. to highlight potential phishing
attempts, any text (anywhere in the webclient) that switches unicode "blocks"
(a rough approximation to (language) scripts) within a word is underlined
orange. multiple messages can be selected with familiar ui interaction:
clicking while holding control and/or shift keys. keyboard navigation works
with arrows/page up/down and home/end keys, and also with a few basic vi-like
keys for list/message navigation. we prefer showing the text instead of
html (with inlined images only) version of a message. html messages are shown
in an iframe served from an endpoint with CSP headers to prevent dangerous
resources (scripts, external images) from being loaded. the html is also
sanitized, with javascript removed. a user can choose to load external
resources (e.g. images for tracking purposes).
the frontend is just (strict) typescript, no external frameworks. all
incoming/outgoing data is typechecked, both the api request parameters and
response types, and the data coming in over SSE. the types and checking code
are generated with sherpats, which uses the api definitions generated by
sherpadoc based on the Go code. so types from the backend are automatically
propagated to the frontend. since there is no framework to automatically
propagate properties and rerender components, changes coming in over the SSE
connection are propagated explicitly with regular function calls. the ui is
separated into "views", each with a "root" dom element that is added to the
visible document. these views have additional functions for getting changes
propagated, often resulting in the view updating its (internal) ui state (dom).
we keep the frontend compilation simple, it's just a few typescript files that
get compiled (combined and types stripped) into a single js file, no additional
runtime code needed or complicated build processes used. the webmail is served
is served from a compressed, cachable html file that includes style and the
javascript, currently just over 225kb uncompressed, under 60kb compressed (not
minified, including comments). we include the generated js files in the
repository, to keep Go's easily buildable self-contained binaries.
authentication is basic http, as with the account and admin pages. most data
comes in over one long-term SSE connection to the backend. api requests signal
which mailbox/search/messages are requested over the SSE connection. fetching
individual messages, and making changes, are done through api calls. the
operations are similar to imap, so some code has been moved from package
imapserver to package store. the future jmap implementation will benefit from
these changes too. more functionality will probably be moved to the store
package in the future.
the quickstart enables webmail on the internal listener by default (for new
installs). users can enable it on the public listener if they want to. mox
localserve enables it too. to enable webmail on existing installs, add settings
like the following to the listeners in mox.conf, similar to AccountHTTP(S):
WebmailHTTP:
Enabled: true
WebmailHTTPS:
Enabled: true
special thanks to liesbeth, gerben, andrii for early user feedback.
there is plenty still to do, see the list at the top of webmail/webmail.ts.
feedback welcome as always.
2023-08-07 22:57:03 +03:00
|
|
|
mc := destMailboxCounts[m.MailboxID]
|
|
|
|
mc.Sub(m.MailboxCounts())
|
|
|
|
|
|
|
|
oflags := m.Flags
|
2023-02-16 11:57:27 +03:00
|
|
|
m.Flags = m.Flags.Set(flags, flags)
|
2023-06-24 01:24:43 +03:00
|
|
|
m.Keywords = maps.Keys(keywords)
|
add webmail
it was far down on the roadmap, but implemented earlier, because it's
interesting, and to help prepare for a jmap implementation. for jmap we need to
implement more client-like functionality than with just imap. internal data
structures need to change. jmap has lots of other requirements, so it's already
a big project. by implementing a webmail now, some of the required data
structure changes become clear and can be made now, so the later jmap
implementation can do things similarly to the webmail code. the webmail
frontend and webmail are written together, making their interface/api much
smaller and simpler than jmap.
one of the internal changes is that we now keep track of per-mailbox
total/unread/unseen/deleted message counts and mailbox sizes. keeping this
data consistent after any change to the stored messages (through the code base)
is tricky, so mox now has a consistency check that verifies the counts are
correct, which runs only during tests, each time an internal account reference
is closed. we have a few more internal "changes" that are propagated for the
webmail frontend (that imap doesn't have a way to propagate on a connection),
like changes to the special-use flags on mailboxes, and used keywords in a
mailbox. more changes that will be required have revealed themselves while
implementing the webmail, and will be implemented next.
the webmail user interface is modeled after the mail clients i use or have
used: thunderbird, macos mail, mutt; and webmails i normally only use for
testing: gmail, proton, yahoo, outlook. a somewhat technical user is assumed,
but still the goal is to make this webmail client easy to use for everyone. the
user interface looks like most other mail clients: a list of mailboxes, a
search bar, a message list view, and message details. there is a top/bottom and
a left/right layout for the list/message view, default is automatic based on
screen size. the panes can be resized by the user. buttons for actions are just
text, not icons. clicking a button briefly shows the shortcut for the action in
the bottom right, helping with learning to operate quickly. any text that is
underdotted has a title attribute that causes more information to be displayed,
e.g. what a button does or a field is about. to highlight potential phishing
attempts, any text (anywhere in the webclient) that switches unicode "blocks"
(a rough approximation to (language) scripts) within a word is underlined
orange. multiple messages can be selected with familiar ui interaction:
clicking while holding control and/or shift keys. keyboard navigation works
with arrows/page up/down and home/end keys, and also with a few basic vi-like
keys for list/message navigation. we prefer showing the text instead of
html (with inlined images only) version of a message. html messages are shown
in an iframe served from an endpoint with CSP headers to prevent dangerous
resources (scripts, external images) from being loaded. the html is also
sanitized, with javascript removed. a user can choose to load external
resources (e.g. images for tracking purposes).
the frontend is just (strict) typescript, no external frameworks. all
incoming/outgoing data is typechecked, both the api request parameters and
response types, and the data coming in over SSE. the types and checking code
are generated with sherpats, which uses the api definitions generated by
sherpadoc based on the Go code. so types from the backend are automatically
propagated to the frontend. since there is no framework to automatically
propagate properties and rerender components, changes coming in over the SSE
connection are propagated explicitly with regular function calls. the ui is
separated into "views", each with a "root" dom element that is added to the
visible document. these views have additional functions for getting changes
propagated, often resulting in the view updating its (internal) ui state (dom).
we keep the frontend compilation simple, it's just a few typescript files that
get compiled (combined and types stripped) into a single js file, no additional
runtime code needed or complicated build processes used. the webmail is served
is served from a compressed, cachable html file that includes style and the
javascript, currently just over 225kb uncompressed, under 60kb compressed (not
minified, including comments). we include the generated js files in the
repository, to keep Go's easily buildable self-contained binaries.
authentication is basic http, as with the account and admin pages. most data
comes in over one long-term SSE connection to the backend. api requests signal
which mailbox/search/messages are requested over the SSE connection. fetching
individual messages, and making changes, are done through api calls. the
operations are similar to imap, so some code has been moved from package
imapserver to package store. the future jmap implementation will benefit from
these changes too. more functionality will probably be moved to the store
package in the future.
the quickstart enables webmail on the internal listener by default (for new
installs). users can enable it on the public listener if they want to. mox
localserve enables it too. to enable webmail on existing installs, add settings
like the following to the listeners in mox.conf, similar to AccountHTTP(S):
WebmailHTTP:
Enabled: true
WebmailHTTPS:
Enabled: true
special thanks to liesbeth, gerben, andrii for early user feedback.
there is plenty still to do, see the list at the top of webmail/webmail.ts.
feedback welcome as always.
2023-08-07 22:57:03 +03:00
|
|
|
sort.Strings(m.Keywords)
|
|
|
|
|
|
|
|
mc.Add(m.MailboxCounts())
|
|
|
|
destMailboxCounts[m.MailboxID] = mc
|
2023-06-24 01:24:43 +03:00
|
|
|
|
|
|
|
if len(m.Keywords) > 0 {
|
|
|
|
if destMailboxKeywords[m.MailboxID] == nil {
|
|
|
|
destMailboxKeywords[m.MailboxID] = map[string]bool{}
|
|
|
|
}
|
|
|
|
for _, k := range m.Keywords {
|
|
|
|
destMailboxKeywords[m.MailboxID][k] = true
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
2023-02-16 11:57:27 +03:00
|
|
|
// We train before updating, training may set m.TrainedJunk.
|
|
|
|
if jf != nil && m.NeedsTraining() {
|
|
|
|
openTrainMessage(&m)
|
|
|
|
}
|
|
|
|
err = tx.Update(&m)
|
|
|
|
ximportcheckf(err, "updating message after flag update")
|
add webmail
it was far down on the roadmap, but implemented earlier, because it's
interesting, and to help prepare for a jmap implementation. for jmap we need to
implement more client-like functionality than with just imap. internal data
structures need to change. jmap has lots of other requirements, so it's already
a big project. by implementing a webmail now, some of the required data
structure changes become clear and can be made now, so the later jmap
implementation can do things similarly to the webmail code. the webmail
frontend and webmail are written together, making their interface/api much
smaller and simpler than jmap.
one of the internal changes is that we now keep track of per-mailbox
total/unread/unseen/deleted message counts and mailbox sizes. keeping this
data consistent after any change to the stored messages (through the code base)
is tricky, so mox now has a consistency check that verifies the counts are
correct, which runs only during tests, each time an internal account reference
is closed. we have a few more internal "changes" that are propagated for the
webmail frontend (that imap doesn't have a way to propagate on a connection),
like changes to the special-use flags on mailboxes, and used keywords in a
mailbox. more changes that will be required have revealed themselves while
implementing the webmail, and will be implemented next.
the webmail user interface is modeled after the mail clients i use or have
used: thunderbird, macos mail, mutt; and webmails i normally only use for
testing: gmail, proton, yahoo, outlook. a somewhat technical user is assumed,
but still the goal is to make this webmail client easy to use for everyone. the
user interface looks like most other mail clients: a list of mailboxes, a
search bar, a message list view, and message details. there is a top/bottom and
a left/right layout for the list/message view, default is automatic based on
screen size. the panes can be resized by the user. buttons for actions are just
text, not icons. clicking a button briefly shows the shortcut for the action in
the bottom right, helping with learning to operate quickly. any text that is
underdotted has a title attribute that causes more information to be displayed,
e.g. what a button does or a field is about. to highlight potential phishing
attempts, any text (anywhere in the webclient) that switches unicode "blocks"
(a rough approximation to (language) scripts) within a word is underlined
orange. multiple messages can be selected with familiar ui interaction:
clicking while holding control and/or shift keys. keyboard navigation works
with arrows/page up/down and home/end keys, and also with a few basic vi-like
keys for list/message navigation. we prefer showing the text instead of
html (with inlined images only) version of a message. html messages are shown
in an iframe served from an endpoint with CSP headers to prevent dangerous
resources (scripts, external images) from being loaded. the html is also
sanitized, with javascript removed. a user can choose to load external
resources (e.g. images for tracking purposes).
the frontend is just (strict) typescript, no external frameworks. all
incoming/outgoing data is typechecked, both the api request parameters and
response types, and the data coming in over SSE. the types and checking code
are generated with sherpats, which uses the api definitions generated by
sherpadoc based on the Go code. so types from the backend are automatically
propagated to the frontend. since there is no framework to automatically
propagate properties and rerender components, changes coming in over the SSE
connection are propagated explicitly with regular function calls. the ui is
separated into "views", each with a "root" dom element that is added to the
visible document. these views have additional functions for getting changes
propagated, often resulting in the view updating its (internal) ui state (dom).
we keep the frontend compilation simple, it's just a few typescript files that
get compiled (combined and types stripped) into a single js file, no additional
runtime code needed or complicated build processes used. the webmail is served
is served from a compressed, cachable html file that includes style and the
javascript, currently just over 225kb uncompressed, under 60kb compressed (not
minified, including comments). we include the generated js files in the
repository, to keep Go's easily buildable self-contained binaries.
authentication is basic http, as with the account and admin pages. most data
comes in over one long-term SSE connection to the backend. api requests signal
which mailbox/search/messages are requested over the SSE connection. fetching
individual messages, and making changes, are done through api calls. the
operations are similar to imap, so some code has been moved from package
imapserver to package store. the future jmap implementation will benefit from
these changes too. more functionality will probably be moved to the store
package in the future.
the quickstart enables webmail on the internal listener by default (for new
installs). users can enable it on the public listener if they want to. mox
localserve enables it too. to enable webmail on existing installs, add settings
like the following to the listeners in mox.conf, similar to AccountHTTP(S):
WebmailHTTP:
Enabled: true
WebmailHTTPS:
Enabled: true
special thanks to liesbeth, gerben, andrii for early user feedback.
there is plenty still to do, see the list at the top of webmail/webmail.ts.
feedback welcome as always.
2023-08-07 22:57:03 +03:00
|
|
|
changes = append(changes, m.ChangeFlags(oflags))
|
2023-02-16 11:57:27 +03:00
|
|
|
}
|
|
|
|
delete(mailboxMissingKeywordMessages, mailbox)
|
|
|
|
} else {
|
|
|
|
problemf("unrecognized file %s (skipping)", origName)
|
|
|
|
}
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
|
|
|
if zr != nil {
|
|
|
|
for _, f := range zr.File {
|
|
|
|
if canceled() {
|
|
|
|
return
|
|
|
|
}
|
|
|
|
zf, err := f.Open()
|
|
|
|
if err != nil {
|
|
|
|
problemf("opening file %s in zip: %v", f.Name, err)
|
|
|
|
continue
|
|
|
|
}
|
|
|
|
importFile(f.Name, zf)
|
2023-02-16 15:22:00 +03:00
|
|
|
err = zf.Close()
|
|
|
|
log.Check(err, "closing file from zip")
|
2023-02-16 11:57:27 +03:00
|
|
|
}
|
|
|
|
} else {
|
|
|
|
for {
|
|
|
|
if canceled() {
|
|
|
|
return
|
|
|
|
}
|
|
|
|
h, err := tr.Next()
|
|
|
|
if err == io.EOF {
|
|
|
|
break
|
|
|
|
} else if err != nil {
|
|
|
|
problemf("reading next tar header: %v (aborting)", err)
|
|
|
|
return
|
|
|
|
}
|
|
|
|
importFile(h.Name, tr)
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
|
|
|
total := 0
|
|
|
|
for _, count := range messages {
|
|
|
|
total += count
|
|
|
|
}
|
2023-12-05 15:35:58 +03:00
|
|
|
log.Debug("messages imported", slog.Int("total", total))
|
2023-02-16 11:57:27 +03:00
|
|
|
|
|
|
|
// Send final update for count of last-imported mailbox.
|
|
|
|
if prevMailbox != "" {
|
|
|
|
sendEvent("count", importCount{prevMailbox, messages[prevMailbox]})
|
|
|
|
}
|
|
|
|
|
implement message threading in backend and webmail
we match messages to their parents based on the "references" and "in-reply-to"
headers (requiring the same base subject), and in absense of those headers we
also by only base subject (against messages received max 4 weeks ago).
we store a threadid with messages. all messages in a thread have the same
threadid. messages also have a "thread parent ids", which holds all id's of
parent messages up to the thread root. then there is "thread missing link",
which is set when a referenced immediate parent wasn't found (but possibly
earlier ancestors can still be found and will be in thread parent ids".
threads can be muted: newly delivered messages are automatically marked as
read/seen. threads can be marked as collapsed: if set, the webmail collapses
the thread to a single item in the basic threading view (default is to expand
threads). the muted and collapsed fields are copied from their parent on
message delivery.
the threading is implemented in the webmail. the non-threading mode still works
as before. the new default threading mode "unread" automatically expands only
the threads with at least one unread (not seen) meessage. the basic threading
mode "on" expands all threads except when explicitly collapsed (as saved in the
thread collapsed field). new shortcuts for navigation/interaction threads have
been added, e.g. go to previous/next thread root, toggle collapse/expand of
thread (or double click), toggle mute of thread. some previous shortcuts have
changed, see the help for details.
the message threading are added with an explicit account upgrade step,
automatically started when an account is opened. the upgrade is done in the
background because it will take too long for large mailboxes to block account
operations. the upgrade takes two steps: 1. updating all message records in the
database to add a normalized message-id and thread base subject (with "re:",
"fwd:" and several other schemes stripped). 2. going through all messages in
the database again, reading the "references" and "in-reply-to" headers from
disk, and matching against their parents. this second step is also done at the
end of each import of mbox/maildir mailboxes. new deliveries are matched
immediately against other existing messages, currently no attempt is made to
rematch previously delivered messages (which could be useful for related
messages being delivered out of order).
the threading is not yet exposed over imap.
2023-09-13 09:51:50 +03:00
|
|
|
// Match threads.
|
|
|
|
if len(deliveredIDs) > 0 {
|
|
|
|
sendEvent("step", importStep{"matching messages with threads"})
|
|
|
|
err = acc.AssignThreads(ctx, log, tx, deliveredIDs[0], 0, io.Discard)
|
|
|
|
ximportcheckf(err, "assigning messages to threads")
|
|
|
|
}
|
|
|
|
|
add webmail
it was far down on the roadmap, but implemented earlier, because it's
interesting, and to help prepare for a jmap implementation. for jmap we need to
implement more client-like functionality than with just imap. internal data
structures need to change. jmap has lots of other requirements, so it's already
a big project. by implementing a webmail now, some of the required data
structure changes become clear and can be made now, so the later jmap
implementation can do things similarly to the webmail code. the webmail
frontend and webmail are written together, making their interface/api much
smaller and simpler than jmap.
one of the internal changes is that we now keep track of per-mailbox
total/unread/unseen/deleted message counts and mailbox sizes. keeping this
data consistent after any change to the stored messages (through the code base)
is tricky, so mox now has a consistency check that verifies the counts are
correct, which runs only during tests, each time an internal account reference
is closed. we have a few more internal "changes" that are propagated for the
webmail frontend (that imap doesn't have a way to propagate on a connection),
like changes to the special-use flags on mailboxes, and used keywords in a
mailbox. more changes that will be required have revealed themselves while
implementing the webmail, and will be implemented next.
the webmail user interface is modeled after the mail clients i use or have
used: thunderbird, macos mail, mutt; and webmails i normally only use for
testing: gmail, proton, yahoo, outlook. a somewhat technical user is assumed,
but still the goal is to make this webmail client easy to use for everyone. the
user interface looks like most other mail clients: a list of mailboxes, a
search bar, a message list view, and message details. there is a top/bottom and
a left/right layout for the list/message view, default is automatic based on
screen size. the panes can be resized by the user. buttons for actions are just
text, not icons. clicking a button briefly shows the shortcut for the action in
the bottom right, helping with learning to operate quickly. any text that is
underdotted has a title attribute that causes more information to be displayed,
e.g. what a button does or a field is about. to highlight potential phishing
attempts, any text (anywhere in the webclient) that switches unicode "blocks"
(a rough approximation to (language) scripts) within a word is underlined
orange. multiple messages can be selected with familiar ui interaction:
clicking while holding control and/or shift keys. keyboard navigation works
with arrows/page up/down and home/end keys, and also with a few basic vi-like
keys for list/message navigation. we prefer showing the text instead of
html (with inlined images only) version of a message. html messages are shown
in an iframe served from an endpoint with CSP headers to prevent dangerous
resources (scripts, external images) from being loaded. the html is also
sanitized, with javascript removed. a user can choose to load external
resources (e.g. images for tracking purposes).
the frontend is just (strict) typescript, no external frameworks. all
incoming/outgoing data is typechecked, both the api request parameters and
response types, and the data coming in over SSE. the types and checking code
are generated with sherpats, which uses the api definitions generated by
sherpadoc based on the Go code. so types from the backend are automatically
propagated to the frontend. since there is no framework to automatically
propagate properties and rerender components, changes coming in over the SSE
connection are propagated explicitly with regular function calls. the ui is
separated into "views", each with a "root" dom element that is added to the
visible document. these views have additional functions for getting changes
propagated, often resulting in the view updating its (internal) ui state (dom).
we keep the frontend compilation simple, it's just a few typescript files that
get compiled (combined and types stripped) into a single js file, no additional
runtime code needed or complicated build processes used. the webmail is served
is served from a compressed, cachable html file that includes style and the
javascript, currently just over 225kb uncompressed, under 60kb compressed (not
minified, including comments). we include the generated js files in the
repository, to keep Go's easily buildable self-contained binaries.
authentication is basic http, as with the account and admin pages. most data
comes in over one long-term SSE connection to the backend. api requests signal
which mailbox/search/messages are requested over the SSE connection. fetching
individual messages, and making changes, are done through api calls. the
operations are similar to imap, so some code has been moved from package
imapserver to package store. the future jmap implementation will benefit from
these changes too. more functionality will probably be moved to the store
package in the future.
the quickstart enables webmail on the internal listener by default (for new
installs). users can enable it on the public listener if they want to. mox
localserve enables it too. to enable webmail on existing installs, add settings
like the following to the listeners in mox.conf, similar to AccountHTTP(S):
WebmailHTTP:
Enabled: true
WebmailHTTPS:
Enabled: true
special thanks to liesbeth, gerben, andrii for early user feedback.
there is plenty still to do, see the list at the top of webmail/webmail.ts.
feedback welcome as always.
2023-08-07 22:57:03 +03:00
|
|
|
// Update mailboxes with counts and keywords.
|
|
|
|
for mbID, mc := range destMailboxCounts {
|
2023-06-24 01:24:43 +03:00
|
|
|
mb := store.Mailbox{ID: mbID}
|
|
|
|
err := tx.Get(&mb)
|
add webmail
it was far down on the roadmap, but implemented earlier, because it's
interesting, and to help prepare for a jmap implementation. for jmap we need to
implement more client-like functionality than with just imap. internal data
structures need to change. jmap has lots of other requirements, so it's already
a big project. by implementing a webmail now, some of the required data
structure changes become clear and can be made now, so the later jmap
implementation can do things similarly to the webmail code. the webmail
frontend and webmail are written together, making their interface/api much
smaller and simpler than jmap.
one of the internal changes is that we now keep track of per-mailbox
total/unread/unseen/deleted message counts and mailbox sizes. keeping this
data consistent after any change to the stored messages (through the code base)
is tricky, so mox now has a consistency check that verifies the counts are
correct, which runs only during tests, each time an internal account reference
is closed. we have a few more internal "changes" that are propagated for the
webmail frontend (that imap doesn't have a way to propagate on a connection),
like changes to the special-use flags on mailboxes, and used keywords in a
mailbox. more changes that will be required have revealed themselves while
implementing the webmail, and will be implemented next.
the webmail user interface is modeled after the mail clients i use or have
used: thunderbird, macos mail, mutt; and webmails i normally only use for
testing: gmail, proton, yahoo, outlook. a somewhat technical user is assumed,
but still the goal is to make this webmail client easy to use for everyone. the
user interface looks like most other mail clients: a list of mailboxes, a
search bar, a message list view, and message details. there is a top/bottom and
a left/right layout for the list/message view, default is automatic based on
screen size. the panes can be resized by the user. buttons for actions are just
text, not icons. clicking a button briefly shows the shortcut for the action in
the bottom right, helping with learning to operate quickly. any text that is
underdotted has a title attribute that causes more information to be displayed,
e.g. what a button does or a field is about. to highlight potential phishing
attempts, any text (anywhere in the webclient) that switches unicode "blocks"
(a rough approximation to (language) scripts) within a word is underlined
orange. multiple messages can be selected with familiar ui interaction:
clicking while holding control and/or shift keys. keyboard navigation works
with arrows/page up/down and home/end keys, and also with a few basic vi-like
keys for list/message navigation. we prefer showing the text instead of
html (with inlined images only) version of a message. html messages are shown
in an iframe served from an endpoint with CSP headers to prevent dangerous
resources (scripts, external images) from being loaded. the html is also
sanitized, with javascript removed. a user can choose to load external
resources (e.g. images for tracking purposes).
the frontend is just (strict) typescript, no external frameworks. all
incoming/outgoing data is typechecked, both the api request parameters and
response types, and the data coming in over SSE. the types and checking code
are generated with sherpats, which uses the api definitions generated by
sherpadoc based on the Go code. so types from the backend are automatically
propagated to the frontend. since there is no framework to automatically
propagate properties and rerender components, changes coming in over the SSE
connection are propagated explicitly with regular function calls. the ui is
separated into "views", each with a "root" dom element that is added to the
visible document. these views have additional functions for getting changes
propagated, often resulting in the view updating its (internal) ui state (dom).
we keep the frontend compilation simple, it's just a few typescript files that
get compiled (combined and types stripped) into a single js file, no additional
runtime code needed or complicated build processes used. the webmail is served
is served from a compressed, cachable html file that includes style and the
javascript, currently just over 225kb uncompressed, under 60kb compressed (not
minified, including comments). we include the generated js files in the
repository, to keep Go's easily buildable self-contained binaries.
authentication is basic http, as with the account and admin pages. most data
comes in over one long-term SSE connection to the backend. api requests signal
which mailbox/search/messages are requested over the SSE connection. fetching
individual messages, and making changes, are done through api calls. the
operations are similar to imap, so some code has been moved from package
imapserver to package store. the future jmap implementation will benefit from
these changes too. more functionality will probably be moved to the store
package in the future.
the quickstart enables webmail on the internal listener by default (for new
installs). users can enable it on the public listener if they want to. mox
localserve enables it too. to enable webmail on existing installs, add settings
like the following to the listeners in mox.conf, similar to AccountHTTP(S):
WebmailHTTP:
Enabled: true
WebmailHTTPS:
Enabled: true
special thanks to liesbeth, gerben, andrii for early user feedback.
there is plenty still to do, see the list at the top of webmail/webmail.ts.
feedback welcome as always.
2023-08-07 22:57:03 +03:00
|
|
|
ximportcheckf(err, "loading mailbox for counts and keywords")
|
|
|
|
|
|
|
|
if mb.MailboxCounts != mc {
|
|
|
|
mb.MailboxCounts = mc
|
|
|
|
changes = append(changes, mb.ChangeCounts())
|
|
|
|
}
|
|
|
|
|
|
|
|
keywords := destMailboxKeywords[mb.ID]
|
|
|
|
var mbKwChanged bool
|
|
|
|
mb.Keywords, mbKwChanged = store.MergeKeywords(mb.Keywords, maps.Keys(keywords))
|
|
|
|
|
|
|
|
err = tx.Update(&mb)
|
|
|
|
ximportcheckf(err, "updating mailbox count and keywords")
|
|
|
|
if mbKwChanged {
|
|
|
|
changes = append(changes, mb.ChangeKeywords())
|
2023-06-24 01:24:43 +03:00
|
|
|
}
|
|
|
|
}
|
|
|
|
|
2023-12-20 22:54:12 +03:00
|
|
|
err = acc.AddMessageSize(log, tx, addSize)
|
|
|
|
ximportcheckf(err, "updating disk usage after import")
|
|
|
|
|
2023-02-16 11:57:27 +03:00
|
|
|
err = tx.Commit()
|
|
|
|
tx = nil
|
|
|
|
ximportcheckf(err, "commit")
|
|
|
|
deliveredIDs = nil
|
|
|
|
|
|
|
|
if jf != nil {
|
2023-02-16 15:22:00 +03:00
|
|
|
if err := jf.Close(); err != nil {
|
2023-02-16 11:57:27 +03:00
|
|
|
problemf("saving changes of training junk filter: %v (continuing)", err)
|
|
|
|
log.Errorx("saving changes of training junk filter", err)
|
|
|
|
}
|
|
|
|
jf = nil
|
|
|
|
}
|
|
|
|
|
2023-07-23 16:28:37 +03:00
|
|
|
store.BroadcastChanges(acc, changes)
|
2023-02-16 11:57:27 +03:00
|
|
|
acc.Unlock()
|
2023-02-16 15:22:00 +03:00
|
|
|
err = acc.Close()
|
|
|
|
log.Check(err, "closing account after import")
|
2023-02-16 11:57:27 +03:00
|
|
|
acc = nil
|
|
|
|
|
|
|
|
sendEvent("done", importDone{})
|
|
|
|
}
|
|
|
|
|
2023-06-24 01:24:43 +03:00
|
|
|
func flagSet(flags *store.Flags, keywords map[string]bool, word string) {
|
2023-02-16 11:57:27 +03:00
|
|
|
switch word {
|
|
|
|
case "forwarded", "$forwarded":
|
|
|
|
flags.Forwarded = true
|
|
|
|
case "junk", "$junk":
|
|
|
|
flags.Junk = true
|
|
|
|
case "notjunk", "$notjunk", "nonjunk", "$nonjunk":
|
|
|
|
flags.Notjunk = true
|
|
|
|
case "phishing", "$phishing":
|
|
|
|
flags.Phishing = true
|
|
|
|
case "mdnsent", "$mdnsent":
|
|
|
|
flags.MDNSent = true
|
2023-06-24 01:24:43 +03:00
|
|
|
default:
|
add webmail
it was far down on the roadmap, but implemented earlier, because it's
interesting, and to help prepare for a jmap implementation. for jmap we need to
implement more client-like functionality than with just imap. internal data
structures need to change. jmap has lots of other requirements, so it's already
a big project. by implementing a webmail now, some of the required data
structure changes become clear and can be made now, so the later jmap
implementation can do things similarly to the webmail code. the webmail
frontend and webmail are written together, making their interface/api much
smaller and simpler than jmap.
one of the internal changes is that we now keep track of per-mailbox
total/unread/unseen/deleted message counts and mailbox sizes. keeping this
data consistent after any change to the stored messages (through the code base)
is tricky, so mox now has a consistency check that verifies the counts are
correct, which runs only during tests, each time an internal account reference
is closed. we have a few more internal "changes" that are propagated for the
webmail frontend (that imap doesn't have a way to propagate on a connection),
like changes to the special-use flags on mailboxes, and used keywords in a
mailbox. more changes that will be required have revealed themselves while
implementing the webmail, and will be implemented next.
the webmail user interface is modeled after the mail clients i use or have
used: thunderbird, macos mail, mutt; and webmails i normally only use for
testing: gmail, proton, yahoo, outlook. a somewhat technical user is assumed,
but still the goal is to make this webmail client easy to use for everyone. the
user interface looks like most other mail clients: a list of mailboxes, a
search bar, a message list view, and message details. there is a top/bottom and
a left/right layout for the list/message view, default is automatic based on
screen size. the panes can be resized by the user. buttons for actions are just
text, not icons. clicking a button briefly shows the shortcut for the action in
the bottom right, helping with learning to operate quickly. any text that is
underdotted has a title attribute that causes more information to be displayed,
e.g. what a button does or a field is about. to highlight potential phishing
attempts, any text (anywhere in the webclient) that switches unicode "blocks"
(a rough approximation to (language) scripts) within a word is underlined
orange. multiple messages can be selected with familiar ui interaction:
clicking while holding control and/or shift keys. keyboard navigation works
with arrows/page up/down and home/end keys, and also with a few basic vi-like
keys for list/message navigation. we prefer showing the text instead of
html (with inlined images only) version of a message. html messages are shown
in an iframe served from an endpoint with CSP headers to prevent dangerous
resources (scripts, external images) from being loaded. the html is also
sanitized, with javascript removed. a user can choose to load external
resources (e.g. images for tracking purposes).
the frontend is just (strict) typescript, no external frameworks. all
incoming/outgoing data is typechecked, both the api request parameters and
response types, and the data coming in over SSE. the types and checking code
are generated with sherpats, which uses the api definitions generated by
sherpadoc based on the Go code. so types from the backend are automatically
propagated to the frontend. since there is no framework to automatically
propagate properties and rerender components, changes coming in over the SSE
connection are propagated explicitly with regular function calls. the ui is
separated into "views", each with a "root" dom element that is added to the
visible document. these views have additional functions for getting changes
propagated, often resulting in the view updating its (internal) ui state (dom).
we keep the frontend compilation simple, it's just a few typescript files that
get compiled (combined and types stripped) into a single js file, no additional
runtime code needed or complicated build processes used. the webmail is served
is served from a compressed, cachable html file that includes style and the
javascript, currently just over 225kb uncompressed, under 60kb compressed (not
minified, including comments). we include the generated js files in the
repository, to keep Go's easily buildable self-contained binaries.
authentication is basic http, as with the account and admin pages. most data
comes in over one long-term SSE connection to the backend. api requests signal
which mailbox/search/messages are requested over the SSE connection. fetching
individual messages, and making changes, are done through api calls. the
operations are similar to imap, so some code has been moved from package
imapserver to package store. the future jmap implementation will benefit from
these changes too. more functionality will probably be moved to the store
package in the future.
the quickstart enables webmail on the internal listener by default (for new
installs). users can enable it on the public listener if they want to. mox
localserve enables it too. to enable webmail on existing installs, add settings
like the following to the listeners in mox.conf, similar to AccountHTTP(S):
WebmailHTTP:
Enabled: true
WebmailHTTPS:
Enabled: true
special thanks to liesbeth, gerben, andrii for early user feedback.
there is plenty still to do, see the list at the top of webmail/webmail.ts.
feedback welcome as always.
2023-08-07 22:57:03 +03:00
|
|
|
if err := store.CheckKeyword(word); err == nil {
|
2023-06-24 01:24:43 +03:00
|
|
|
keywords[word] = true
|
|
|
|
}
|
2023-02-16 11:57:27 +03:00
|
|
|
}
|
|
|
|
}
|