An SMTP client and server library for Zig implementing RFC 5321.
35 kB
722 lines
1<!--
2SPDX-FileCopyrightText: © 2026 Jeffrey C. Ollie <jeff@ocjtech.us>
3SPDX-License-Identifier: MIT
4-->
5
6# zig-smtp
7
8An SMTP client and server library for Zig (RFC 5321).
9
10Both the client and the server run over plain `std.Io.Reader`/`std.Io.Writer`
11pairs, so they are transport-agnostic: wrap a TCP stream for real use, or
12fixed in-memory buffers in tests. Requires Zig 0.16.
13
14## Where this lives
15
16The canonical repository is on Forgejo, with mirrors on Tangled and Radicle:
17
18- <https://git.jcollie.dev/jeff/zig-smtp> — issues and pull requests
19- <https://tangled.org/jcollie.dev/zig-smtp>
20
21```sh
22git clone https://git.jcollie.dev/jeff/zig-smtp.git
23```
24
25On [Radicle](https://radicle.xyz/), the peer-to-peer forge, the repository is
26`rad:z3ZKHgoDKEue8FT7sV6fHZdtjxRx1`, which is the only name it has there — a
27Radicle repository is found by its ID and nothing else — so seeding or cloning
28it goes:
29
30```sh
31rad clone rad:z3ZKHgoDKEue8FT7sV6fHZdtjxRx1
32```
33
34Cloning also seeds the repository, which helps keep it available on the
35network.
36
37The API documentation is generated from the doc comments and published at
38<https://jeff.jcollie.page/zig-smtp/>; `zig build docs` builds it locally and
39`zig build docs-serve` serves it for reading.
40
41## Client
42
43```zig
44const zig-smtp = @import("smtp");
45
46var reply_buf: [1024]u8 = undefined;
47var client: zig-smtp.Client = .init(&stream_reader.interface, &stream_writer.interface, &reply_buf);
48
49_ = try client.greet(); // read the 220 greeting
50_ = try client.hello("my-host.example.com"); // EHLO (HELO fallback), returns extensions
51try client.sendMail("me@example.com", &.{"you@example.net"}, message);
52try client.quit();
53```
54
55Line endings in the message are normalized to CRLF and leading dots are
56stuffed automatically. On `error.UnexpectedReply`, `client.last_reply` holds
57the server's actual code and text. `mailFrom`/`rcptTo`/`sendMessage` are also
58available individually.
59
60Addresses and the EHLO domain are checked before they are written: a value
61containing CR, LF or NUL is rejected with `error.UnsafeArgument` rather than
62sent, since it would otherwise end the command line early and let the rest of
63it be read as further SMTP commands. The check is `protocol.isSafeArgument`,
64and it is framing only — it does not claim the address is a well-formed
65mailbox.
66
67Message bodies can also be streamed instead of passed as a slice — from any
68reader via `sendMessageReader(&reader)`, or push-style via `data()`, which
69returns a writer that dot-stuffs and normalizes line endings as content
70flows through it:
71
72```zig
73var data_writer = try client.data();
74try data_writer.interface.print("Subject: report {d}\r\n\r\n", .{id});
75// ... stream as much as needed ...
76try data_writer.end(); // terminates the message, reads the verdict
77```
78
79`envelope` sends MAIL FROM and every RCPT TO at once and reads all their
80replies, which against a server advertising PIPELINING
81([RFC 2920](https://datatracker.ietf.org/doc/html/rfc2920)) turns an envelope
82of *n* recipients from *n*+1 round trips into one. `hello` sets
83`client.pipelining` from the EHLO response and `envelope` falls back to
84waiting for each reply when it is false, so the result is the same either
85way:
86
87```zig
88var codes: [3]u16 = undefined;
89const accepted = try client.envelope(from, recipients, &codes, .{});
90// codes[i] is the RCPT reply code for recipients[i].
91```
92
93A refused recipient is not an error — with several of them the caller is the
94one who can say whether what remains is worth sending — so compare `accepted`
95against `recipients.len`. `sendMail` makes that decision the strict way: if
96any recipient was refused it sends RSET and returns `error.UnexpectedReply`
97without delivering to the others.
98
99DATA is deliberately left out of the group, though RFC 2920 allows it as the
100last command of one. Once a server has answered DATA with 354 the transaction
101is committed, and the only ways out are to send the message or to send an
102empty one to whichever recipients were accepted; stopping the group before
103DATA keeps that choice with the caller, and costs one round trip out of the
104*n*+1 saved.
105
106`mail` and `rcpt` are the parameterized forms of `mailFrom` and `rcptTo`,
107carrying the ESMTP parameters the server advertised — today SMTPUTF8 and the
108DSN set of [RFC 3461](https://datatracker.ietf.org/doc/html/rfc3461):
109
110```zig
111try client.mail("me@example.com", .{ .ret = .hdrs, .envid = "batch 7" });
112try client.rcpt("bob@example.net", .{
113 .notify = .{ .on = .{ .failure = true, .delay = true } },
114 .orcpt = .{ .addr_type = "rfc822", .address = "team@example.net" },
115});
116```
117
118`ENVID` and the `ORCPT` address are xtext-encoded on the way out, so any
119bytes are safe to pass; the length limits RFC 3461 puts on the encoded form
120(100 and 500 characters) are checked and surface as
121`error.ArgumentTooLong`. Check `extensions.dsn` first — a conforming server
122answers an unrecognized parameter with 555.
123
124Setting `client.mode = .lmtp` before `hello` speaks LMTP: `LHLO` goes out in
125place of `EHLO`, and the end of a message brings back one verdict per
126accepted recipient, in the order the RCPT commands were issued. `endResults`
127is how to read them:
128
129```zig
130var data_writer = try client.data();
131try data_writer.interface.writeAll(message);
132var verdicts = try data_writer.endResults();
133while (try verdicts.next()) |reply| {
134 // verdicts.index counts the recipients as they are answered.
135 std.log.info("{s}: {d} {s}", .{ recipients[verdicts.index - 1], reply.code, reply.text });
136}
137```
138
139Every verdict must be read before the session is used again, or the next
140command is answered by a leftover reply. The simpler `end` reads them all
141and reports `error.RecipientRejected` if any was a refusal — without saying
142which, because the replies share one buffer and reading the next overwrites
143the previous.
144
145When the server advertises CHUNKING (`extensions.chunking`), `bdat` and
146`sendMessageChunked` transmit the message with length-framed BDAT chunks
147instead of DATA — verbatim, with no dot-stuffing, so text content must
148already use CRLF line endings.
149
150That framing is also what makes binary content possible.
151`mail(from, .{ .body = .binary_mime })` declares it
152([RFC 3030](https://datatracker.ietf.org/doc/html/rfc3030), needs
153`extensions.binary_mime`), after which the message may hold any octets at
154all — NULs, bare CR, a line that is nothing but a dot — and `data` refuses
155to open a DATA phase for it with `error.BinaryRequiresChunking`, which is
156the 503 the server would have sent, made one round trip earlier. RFC 3030
157is absolute that binary must not be sent to a server that did not advertise
158it, so check the capability first.
159
160### Authentication
161
162The mechanisms themselves live in
163[zig-sasl](https://git.jcollie.dev/jeff/zig-sasl), re-exported here as
164`zig-smtp.sasl`, because nothing about PLAIN or CRAM-MD5 or XOAUTH2 is specific
165to SMTP — POP3 and IMAP want the same ones, and one implementation of each is
166better than three. What is specific to SMTP is `authenticate`: the `AUTH`
167command, the 334 challenges, the `*` that cancels, and the 235 that ends it.
168
169`hello` reports the server's advertised mechanism names in `extensions.auth`,
170exactly as it sent them, for `sasl.Client.selectFromList`:
171
172```zig
173var plain: zig-smtp.sasl.Plain = .init("user", "password");
174var cram: zig-smtp.sasl.CramMd5 = .init("user", "password");
175
176const extensions = try client.hello("my-host.example.com");
177const mechanism = zig-smtp.sasl.Client.selectFromList(
178 &.{ plain.client(), cram.client() }, // in order of preference
179 extensions.auth,
180 client.security == .encrypted,
181) orelse return error.NoSupportedMechanism;
182try client.authenticate(mechanism);
183```
184
185A 535 rejection surfaces as `error.AuthenticationFailed` with the reply in
186`last_reply`.
187
188PLAIN, LOGIN and the OAuth mechanisms put a credential on the wire that an
189eavesdropper could reuse — base64 is not encryption, and a bearer token is
190worth more than a password because it authorizes elsewhere too. The client
191refuses those unless `client.security` is `.encrypted`, returning
192`error.InsecureTransport` before anything is sent, and `selectFromList`
193skips them for the same reason: on a plaintext session the preference order
194above falls through PLAIN to CRAM-MD5, which sends a proof rather than the
195secret.
196
197The library is handed a reader and a writer and cannot see what is underneath
198them, so it assumes the worst: `setTransport` records the answer for a
199STARTTLS upgrade, and a session speaking TLS from the first byte sets
200`client.security = .encrypted` itself. For a connection protected by
201something the library cannot see — a unix socket, an SSH tunnel, a loopback
202test — `client.allow_cleartext_auth = true` permits them without claiming the
203transport is encrypted.
204
205One error is worth knowing about even if it never fires for PLAIN:
206`error.ServerNotAuthenticated` means the server reported success while the
207mechanism had not finished proving what it set out to prove. For a one-way
208mechanism that cannot happen. For SCRAM (via
209[zig-scram](https://git.jcollie.dev/jeff/zig-scram)'s `scram-sasl` module) it
210means the server never produced its own signature — which is what something
211in the middle, holding no verifier, would do.
212
213### TLS
214
215`zig-smtp.Tls` wraps [ianic/tls.zig](https://github.com/ianic/tls.zig) and
216verifies against the system trust store by default (a caller-managed CA
217bundle and an insecure mode are also available). The stream reader/writer
218handed to it need buffers of at least `zig-smtp.Tls.min_buffer_len` bytes, and
219`init` must run at the value's final address (the connection holds interior
220pointers). The standard library's TLS client is deliberately not used: it
221requires the optional TLS 1.3 middlebox-compatibility ChangeCipherSpec
222record, which servers like Exim disable.
223
224Implicit TLS (port 465) — handshake first, then speak SMTP:
225
226```zig
227var tls: zig-smtp.Tls = undefined;
228try tls.init(io, gpa, &stream_reader.interface, &stream_writer.interface, .{
229 .host = "smtp.example.com",
230});
231defer tls.deinit(gpa);
232var client: zig-smtp.Client = .init(tls.reader(), tls.writer(), &reply_buf);
233client.security = .encrypted; // the transport is TLS; `init` cannot tell
234// ... greet, hello, sendMail ...
235try client.quit();
236try tls.end(); // close_notify, before closing the socket
237```
238
239STARTTLS (port 587) — upgrade mid-session, then EHLO again:
240
241```zig
242_ = try client.greet();
243_ = try client.hello("my-host.example.com"); // check .starttls in the result
244try client.starttls();
245var tls: zig-smtp.Tls = undefined;
246try tls.init(io, gpa, &stream_reader.interface, &stream_writer.interface, .{
247 .host = "smtp.example.com",
248});
249client.setTransport(tls.reader(), tls.writer(), .encrypted);
250_ = try client.hello("my-host.example.com"); // server state was reset
251```
252
253## Server
254
255```zig
256var session: zig-smtp.Server = .init(&stream_reader.interface, &stream_writer.interface, .{
257 .context = &my_state,
258 .vtable = &.{
259 .authenticate = onAuth, // optional; enables AUTH PLAIN and LOGIN
260 .rcptTo = onRcptTo, // optional; accept/reject each Recipient
261 .message = onMessage, // required; receives envelope + message data
262 },
263}, .{ .hostname = "mx.example.com" });
264try session.run(gpa);
265```
266
267`Options.auth_mechanisms` is what the session offers for AUTH (RFC 4954),
268advertised by name in the EHLO response and drawn from
269[zig-sasl](https://git.jcollie.dev/jeff/zig-sasl) — so a server can offer
270CRAM-MD5 or EXTERNAL, which it could not when the mechanisms were built in:
271
272```zig
273const check: zsmtp.sasl.Server.PasswordCheck = .{ .context = &app, .verify = verify };
274var plain: zsmtp.sasl.PlainServer = .init(check);
275var login: zsmtp.sasl.LoginServer = .init(check);
276// ... .auth_mechanisms = &.{ plain.server(), login.server() }
277```
278
279**The mechanisms hold per-exchange state, so each session needs its own.**
280Sharing a set between two connections would have them overwrite each other's
281challenges; `Server.init` is per-connection anyway, so building them beside
282it is the natural place.
283
284Where the credential comes from is the mechanism's business, which is why
285there is no longer one callback for it. PLAIN and LOGIN share a
286`PasswordCheck` — asked whether a password is right and told nothing, so an
287application may store a hash — while CRAM-MD5 needs a `PasswordLookup`,
288because it has to compute the same HMAC the client did and therefore needs
289the password itself. That is the argument against offering CRAM-MD5 at all,
290and it is now visible in the types rather than buried.
291
292Whatever the mechanism reports as the authenticated identity reaches every
293`Envelope` as `authenticated_as`, which is what a handler deciding whether to
294relay wants — the envelope sender is whatever the client chose to write.
295
296Setting `Options.require_auth` rejects MAIL with 530 until the client has
297authenticated.
298
299Instead of `message` (which collects the whole body in memory, bounded by
300`max_message_size`), a handler can set `messageReader` to stream it: the
301callback receives an `Io.Reader` yielding the unstuffed message content,
302and anything left unread is drained by the session.
303
304`run` serves one connection until QUIT or disconnect, enforcing command
305sequencing, recipient and message-size limits, and un-stuffing message data.
306Messages may also arrive via BDAT chunks (CHUNKING is advertised); both
307the collecting and streaming handler paths receive the reassembled content.
308MAIL parameters are validated: `SIZE=` (RFC 1870) is rejected early with 552
309when it exceeds `max_message_size`, `BODY=7BIT`/`BODY=8BITMIME` (RFC 6152)
310are accepted, and unrecognized parameters get 555; the declared size and
311body type reach the handler via `Envelope`. Listening, accepting, and
312concurrency are up to the caller.
313
314The server holds back the replies that RFC 2920 §3.2 permits — RSET, MAIL
315FROM and RCPT TO — so that a pipelined group is answered in one write, and
316sends everything pending the moment its input is empty. The condition is what
317makes that safe rather than a deadlock: a reply is only ever held while there
318is another command already waiting to be answered.
319
320Setting `Options.protocol = .lmtp` makes the session speak LMTP
321([RFC 2033](https://datatracker.ietf.org/doc/html/rfc2033)) instead: `LHLO`
322greets and `HELO`/`EHLO` are refused with 500, and the end of a message
323draws one reply per accepted recipient rather than one for the message —
324including a second reply for a recipient named twice. The `recipientResult`
325callback supplies each verdict:
326
327```zig
328fn onRecipientResult(ctx: ?*anyopaque, envelope: zig-smtp.Server.Envelope, index: usize) zig-smtp.Server.Decision {
329 return if (mailboxIsFull(envelope.recipients[index].address))
330 .{ .reject = .{ .code = 452, .text = "4.2.2 Mailbox full" } }
331 else
332 .accept;
333}
334```
335
336Without it every recipient is told the same thing, which is correct but
337gains nothing over SMTP. A message the handler rejected outright is reported
338as that rejection for each recipient, since it failed for all of them. LMTP
339is meant for the hop between a queueing MTA and whatever writes to mailboxes;
340RFC 2033 §5 forbids it on TCP port 25 and advises against wide-area use.
341
342BINARYMIME ([RFC 3030](https://datatracker.ietf.org/doc/html/rfc3030)) is
343advertised alongside CHUNKING, which the RFC requires of anything offering
344it. `BODY=BINARYMIME` arrives as `Envelope.body`, DATA for such a message is
345refused with 503, and the content reaches the handler exactly as it was
346sent — the BDAT path copies octets and has no line structure to normalize.
347
348DSN ([RFC 3461](https://datatracker.ietf.org/doc/html/rfc3461)) is
349advertised. `RET=` and `ENVID=` on MAIL arrive as `Envelope.ret` and
350`Envelope.envid`, and `NOTIFY=` and `ORCPT=` on RCPT arrive as
351`Recipient.notify` and `Recipient.orcpt` — at the `rcptTo` callback, which
352receives the whole `Recipient`, and again on the `Envelope` afterwards. The
353xtext values are decoded, the length limits enforced, and a malformed value
354answered with 501. Like everything else handed to a callback, those slices
355live only for the duration of the call; keep what you need by copying it.
356
357To advertise and accept STARTTLS (TLS 1.3, via
358[ianic/tls.zig](https://github.com/ianic/tls.zig)), pass a certificate key
359pair; the stream buffers must then be at least `zig-smtp.tls.input_buffer_len` /
360`zig-smtp.tls.output_buffer_len` bytes, since the handshake runs over them:
361
362```zig
363var auth: zig-smtp.tls.config.CertKeyPair =
364 try .fromFilePath(gpa, io, .cwd(), "cert.pem", "key.pem");
365defer auth.deinit(gpa);
366
367var session: zig-smtp.Server = .init(&stream_reader.interface, &stream_writer.interface, handler, .{
368 .hostname = "mx.example.com",
369 .tls = .{ .io = io, .auth = &auth },
370});
371try session.run(gpa);
372```
373
374On STARTTLS the session answers 220, performs the server handshake, swaps
375its transport to the encrypted connection, and resets state per RFC 3207 (the
376client must EHLO again). With `.mode = .implicit` the handshake instead runs
377before the greeting (SMTPS, port 465 style):
378
379```zig
380var session: zig-smtp.Server = .init(&stream_reader.interface, &stream_writer.interface, handler, .{
381 .hostname = "mx.example.com",
382 .tls = .{ .io = io, .auth = &auth, .mode = .implicit },
383});
384```
385
386## Demo CLI
387
388```sh
389zig build
390
391# Debug server that prints received messages to stdout
392# (with a cert/key pair it advertises and accepts STARTTLS):
393./zig-out/bin/zig-smtp serve 2525
394./zig-out/bin/zig-smtp serve --tls-cert cert.pem --tls-key key.pem 2525
395./zig-out/bin/zig-smtp serve --tls-cert cert.pem --tls-key key.pem --implicit-tls 2465
396
397# Send a message read from stdin:
398printf 'Subject: hi\r\n\r\nhello\r\n' | \
399 ./zig-out/bin/zig-smtp send 127.0.0.1 2525 me@example.com you@example.net
400
401# Same, over implicit TLS or STARTTLS (--insecure skips cert verification):
402zig-smtp send --tls smtp.example.com 465 me@example.com you@example.net
403zig-smtp send --starttls smtp.example.com 587 me@example.com you@example.net
404
405# Send arbitrary binary content (RFC 3030), framed by BDAT rather than DATA:
406./zig-out/bin/zig-smtp send --binarymime 127.0.0.1 2525 me@example.com you@example.net \
407 < some-binary-file
408
409# Speak LMTP (RFC 2033) instead of SMTP. The server reports one verdict per
410# recipient, and --fail-delivery makes one of them fail to show it:
411./zig-out/bin/zig-smtp serve --lmtp --fail-delivery bad@example.net 2529
412printf 'Subject: hi\r\n\r\nhello\r\n' | \
413 ./zig-out/bin/zig-smtp send --lmtp 127.0.0.1 2529 me@example.com \
414 good@example.net bad@example.net
415
416# Request a delivery status notification (RFC 3461):
417zig-smtp send --ret hdrs --envid 'batch 7' --notify success,failure \
418 --orcpt team@example.net 127.0.0.1 2525 me@example.com you@example.net
419
420# Authenticate. Over a plaintext connection this refuses PLAIN and LOGIN
421# rather than put the password on the wire; --allow-cleartext-auth overrides
422# that for a connection protected by other means:
423zig-smtp send --starttls --user me --password secret smtp.example.com 587 \
424 me@example.com you@example.net
425```
426
427## Status
428
429TLS is supported on both sides via
430[ianic/tls.zig](https://github.com/ianic/tls.zig): the client does implicit
431TLS and STARTTLS via `zig-smtp.Tls`, and the server accepts both STARTTLS and
432implicit TLS (TLS 1.3 only). AUTH covers PLAIN, LOGIN, and CRAM-MD5 on the
433client and PLAIN and LOGIN on the server. Message bodies can be streamed on
434both sides, and the server validates MAIL and RCPT parameters (SIZE=, BODY=,
435and the DSN set RET=, ENVID=, NOTIFY=, ORCPT=). Both sides also speak LMTP,
436where a message ends with one verdict per recipient rather than one for the
437message, and both use PIPELINING, which collapses an envelope into a single
438round trip.
439
440## Known gaps
441
442Measured against the implementations people are likely to be coming from —
443Postfix, Exim and Haraka on the server side, Go's `net/smtp`, Python's
444`smtplib`, lettre and Nodemailer on the client side. Kept here so the list
445is one thing rather than a rediscovery each time.
446
447### Out of scope, not missing
448
449- **Message composition.** No MIME builder, headers, attachments, transfer
450 encodings, `Message-ID` or `Date` generation. zig-smtp carries a message that
451 already exists; building one is RFC 5322's job and belongs in a library of
452 its own.
453- **DSN report generation**
454 ([RFC 3464](https://datatracker.ietf.org/doc/html/rfc3464)). The SMTP half
455 of DSN — RFC 3461's `RET`, `ENVID`, `NOTIFY` and `ORCPT` — is implemented
456 on both sides, but nothing here builds the `multipart/report` message that
457 carries a delivery status back to the sender. That is message composition
458 by another name, so it goes with the library above.
459- **Everything an MTA does around a session.** No queue, no retry schedule,
460 no MX resolution, no routing, no mailbox store. "Server" here means a
461 session handler: listening, accepting and concurrency are the caller's.
462
463### Protocol
464
465- **`AUTH=` on MAIL FROM** ([RFC 4954 §5](https://datatracker.ietf.org/doc/html/rfc4954#section-5)),
466 which a trusted relay uses to forward the identity that originally
467 authenticated. The client-side *mechanisms* are no longer a gap — PLAIN,
468 LOGIN, CRAM-MD5, EXTERNAL, XOAUTH2, OAUTHBEARER and SCRAM all come from
469 zig-sasl, and the server offers whichever of their server halves it is
470 given.
471- **Client certificates** — neither side can present or verify one.
472- **No enhanced status code accessor** — the server emits `x.y.z` on every
473 reply, but `Reply` exposes only `code` and the raw text.
474- `EXPN` is unrecognized rather than unimplemented, so it answers 500 where
475 [RFC 5321 §4.2.4](https://datatracker.ietf.org/doc/html/rfc5321#section-4.2.4)
476 wants 502.
477- Niche and absent: REQUIRETLS, MT-PRIORITY, DELIVERBY, FUTURERELEASE, ETRN.
478
479### Server
480
481- **No `Received:` header.**
482 [RFC 5321 §4.4](https://datatracker.ietf.org/doc/html/rfc5321#section-4.4)
483 requires a receiving server to stamp one.
484- **The handler never sees the connection** — no connect callback, no peer
485 address, no TLS state. Greylisting, DNSBLs, SPF and per-IP policy cannot
486 be built on top, and a `Received:` header cannot be written without it.
487- **No timeouts**, so a client that connects and says nothing holds the
488 session forever;
489 [RFC 5321 §4.5.3.2](https://datatracker.ietf.org/doc/html/rfc5321#section-4.5.3.2)
490 specifies per-command limits.
491- **No abuse limits** beyond `max_recipients`: unlimited failed AUTH
492 attempts, no error-count disconnect, no command budget.
493- **No `require_tls`** to go with `require_auth`.
494- **No PROXY protocol, XCLIENT or XFORWARD**, so the real peer address is
495 lost behind a load balancer.
496- No filter or milter hook, and so no DKIM, SPF, DMARC or ARC.
497- No logging or tracing hooks.
498- `max_message_size` is not enforced in `messageReader` mode.
499
500### Client
501
502- **`sendMail` is all-or-nothing on recipients** — a refused RCPT abandons
503 the transaction, where `smtplib.sendmail` delivers to the rest and reports
504 the refusals. `envelope` gives a caller the per-recipient codes to decide
505 for itself, but no higher-level call does that decision for it.
506- **No `SIZE=` on MAIL**, though the client parses the capability off EHLO:
507 `max_size` is read and never used, so nothing checks that a message fits
508 before transmitting it.
509- No MX resolution or connect helper, no 4xx retry or backoff, no connection
510 reuse helper.
511
512## Standards
513
514- [RFC 5321](https://datatracker.ietf.org/doc/html/rfc5321) — Simple Mail
515 Transfer Protocol: the command/reply protocol, multiline replies,
516 dot-stuffing, reply classes, and ESMTP parameter syntax (client and
517 server).
518- [RFC 1870](https://datatracker.ietf.org/doc/html/rfc1870) — SIZE:
519 advertised and enforced by the server (oversize declarations are rejected
520 with 552 before DATA); parsed from EHLO by the client.
521- [RFC 6152](https://datatracker.ietf.org/doc/html/rfc6152) — 8BITMIME:
522 advertised by the server and `BODY=` validated; parsed by the client.
523- [RFC 3030](https://datatracker.ietf.org/doc/html/rfc3030) — CHUNKING
524 (BDAT) and BINARYMIME: client and server, with length-based framing and no
525 dot-stuffing. `BODY=BINARYMIME` is advertised, accepted and delivered bit
526 for bit, and DATA is refused with 503 for a message that declared it,
527 since binary content cannot be framed by a line holding a single dot.
528- [RFC 3461](https://datatracker.ietf.org/doc/html/rfc3461) — DSN:
529 advertised by the server, which parses and validates `RET=`/`ENVID=` on
530 MAIL and `NOTIFY=`/`ORCPT=` on RCPT and hands them to the handler; the
531 client sends them through `mail`/`rcpt`. Includes the xtext codec of §4.
532 Generating the report message itself (RFC 3464) is out of scope.
533- [RFC 2033](https://datatracker.ietf.org/doc/html/rfc2033) — LMTP: client
534 and server, via `Client.mode` and `Server.Options.protocol`. `LHLO`
535 replaces `EHLO` and the end of a message draws one reply per accepted
536 recipient instead of one for the message, after DATA and after `BDAT
537 LAST` alike.
538- [RFC 2920](https://datatracker.ietf.org/doc/html/rfc2920) — PIPELINING:
539 the client sends a whole envelope as one group through `envelope`, and the
540 server holds back the replies it is allowed to (RSET, MAIL, RCPT) so they
541 leave together, sending everything pending the moment its input runs dry.
542- [RFC 3207](https://datatracker.ietf.org/doc/html/rfc3207) — STARTTLS:
543 client and server, including the mandatory post-handshake state reset.
544- [RFC 8314](https://datatracker.ietf.org/doc/html/rfc8314) — implicit TLS
545 (SMTPS): client (`Tls` before any SMTP traffic) and server
546 (`.mode = .implicit`).
547- [RFC 4954](https://datatracker.ietf.org/doc/html/rfc4954) — AUTH: client
548 and server, including initial responses, empty challenges and `*`
549 cancellation. The client drives any mechanism from
550 [zig-sasl](https://git.jcollie.dev/jeff/zig-sasl), and the server offers
551 whichever of their server halves it is handed — PLAIN
552 ([RFC 4616](https://datatracker.ietf.org/doc/html/rfc4616)), the de-facto
553 [LOGIN](https://datatracker.ietf.org/doc/html/draft-murchison-sasl-login-00),
554 CRAM-MD5 and EXTERNAL among them.
555- [RFC 3463](https://datatracker.ietf.org/doc/html/rfc3463) /
556 [RFC 2034](https://datatracker.ietf.org/doc/html/rfc2034) — enhanced
557 status codes: carried in every server reply and advertised via
558 ENHANCEDSTATUSCODES; detected by the client.
559- [RFC 6531](https://datatracker.ietf.org/doc/html/rfc6531) — SMTPUTF8:
560 client (`mailFromUtf8`) and server (advertised; non-ASCII addresses
561 require the parameter and must be valid UTF-8, rejected with 553 5.6.7
562 per [RFC 6533](https://datatracker.ietf.org/doc/html/rfc6533) otherwise;
563 the flag reaches handlers via `Envelope.smtputf8`).
564
565TLS itself (TLS 1.3, [RFC 8446](https://datatracker.ietf.org/doc/html/rfc8446))
566is provided by [ianic/tls.zig](https://github.com/ianic/tls.zig).
567
568## References cited
569
570The specifications this implementation was written against, and the outside
571work it borrows from, in the RFC citation format so that a reference here
572matches one anywhere else. The **Standards** section above says what is
573implemented of each; this one says what each document *is*. Every entry is
574also filed in the project bibliography, so a citation can be taken from there
575rather than composed; the RFCs are keyed by their DOIs (`10.17487/RFC5321`
576and so on).
577
578- **[RFC1870]** Klensin, J., Freed, N., and K. Moore, "SMTP Service
579 Extension for Message Size Declaration", RFC 1870, November 1995,
580 <https://www.rfc-editor.org/info/rfc1870>.
581- **[RFC2033]** Myers, J., "Local Mail Transfer Protocol", RFC 2033,
582 October 1996, <https://www.rfc-editor.org/info/rfc2033>.
583- **[RFC2034]** Freed, N., "SMTP Service Extension for Returning Enhanced
584 Error Codes", RFC 2034, October 1996,
585 <https://www.rfc-editor.org/info/rfc2034>.
586- **[RFC2195]** Klensin, J., Catoe, R., and P. Krumviede, "IMAP/POP
587 AUTHorize Extension for Simple Challenge/Response", RFC 2195,
588 September 1997, <https://www.rfc-editor.org/info/rfc2195>.
589- **[RFC2920]** Freed, N., "SMTP Service Extension for Command Pipelining",
590 RFC 2920, September 2000, <https://www.rfc-editor.org/info/rfc2920>.
591- **[RFC3030]** Vaudreuil, G., "SMTP Service Extensions for Transmission of
592 Large and Binary MIME Messages", RFC 3030, December 2000,
593 <https://www.rfc-editor.org/info/rfc3030>.
594- **[RFC3207]** Hoffman, P., "SMTP Service Extension for Secure SMTP over
595 Transport Layer Security", RFC 3207, February 2002,
596 <https://www.rfc-editor.org/info/rfc3207>.
597- **[RFC3461]** Moore, K., "Simple Mail Transfer Protocol (SMTP) Service
598 Extension for Delivery Status Notifications (DSNs)", RFC 3461,
599 January 2003, <https://www.rfc-editor.org/info/rfc3461>.
600- **[RFC3463]** Vaudreuil, G., "Enhanced Mail System Status Codes",
601 RFC 3463, January 2003, <https://www.rfc-editor.org/info/rfc3463>.
602- **[RFC3464]** Moore, K. and G. Vaudreuil, "An Extensible Message Format
603 for Delivery Status Notifications", RFC 3464, January 2003,
604 <https://www.rfc-editor.org/info/rfc3464>. *(Cited as out of scope: the
605 report message itself.)*
606- **[RFC4616]** Zeilenga, K., "The PLAIN Simple Authentication and Security
607 Layer (SASL) Mechanism", RFC 4616, August 2006,
608 <https://www.rfc-editor.org/info/rfc4616>.
609- **[RFC4954]** Siemborski, R. and A. Melnikov, "SMTP Service Extension for
610 Authentication", RFC 4954, July 2007,
611 <https://www.rfc-editor.org/info/rfc4954>.
612- **[RFC5321]** Klensin, J., "Simple Mail Transfer Protocol", RFC 5321,
613 October 2008, <https://www.rfc-editor.org/info/rfc5321>.
614- **[RFC5322]** Resnick, P., Ed., "Internet Message Format", RFC 5322,
615 October 2008, <https://www.rfc-editor.org/info/rfc5322>. *(Cited as out
616 of scope: the format of the message this library carries.)*
617- **[RFC6152]** Klensin, J., Freed, N., Rose, M., and D. Crocker, "SMTP
618 Service Extension for 8-bit MIME Transport", RFC 6152, March 2011,
619 <https://www.rfc-editor.org/info/rfc6152>.
620- **[RFC6531]** Yao, J. and W. Mao, "SMTP Extension for Internationalized
621 Email", RFC 6531, February 2012,
622 <https://www.rfc-editor.org/info/rfc6531>.
623- **[RFC6533]** Hansen, T., Ed., Newman, C., and A. Melnikov,
624 "Internationalized Delivery Status and Disposition Notifications",
625 RFC 6533, February 2012, <https://www.rfc-editor.org/info/rfc6533>.
626- **[RFC7628]** Mills, W., Showalter, T., and H. Tschofenig, "A Set of
627 Simple Authentication and Security Layer (SASL) Mechanisms for OAuth",
628 RFC 7628, August 2015, <https://www.rfc-editor.org/info/rfc7628>.
629 *(Cited as a gap.)*
630- **[RFC7677]** Hansen, T., "SCRAM-SHA-256 and SCRAM-SHA-256-PLUS Simple
631 Authentication and Security Layer (SASL) Mechanisms", RFC 7677,
632 November 2015, <https://www.rfc-editor.org/info/rfc7677>. *(Cited as a
633 gap.)*
634- **[RFC8314]** Moore, K. and C. Newman, "Cleartext Considered Obsolete:
635 Use of Transport Layer Security (TLS) for Email Submission and Access",
636 RFC 8314, January 2018, <https://www.rfc-editor.org/info/rfc8314>.
637- **[RFC8446]** Rescorla, E., "The Transport Layer Security (TLS) Protocol
638 Version 1.3", RFC 8446, August 2018,
639 <https://www.rfc-editor.org/info/rfc8446>.
640- **[SASL-LOGIN]** Murchison, K. and M. Crispin, "The LOGIN SASL
641 Mechanism", Work in Progress, Internet-Draft,
642 draft-murchison-sasl-login-00, August 2003,
643 <https://datatracker.ietf.org/doc/html/draft-murchison-sasl-login-00>.
644 The draft expired and LOGIN was never standardized; it is implemented
645 here because servers still ask for it.
646- **[TLS.ZIG]** Ianic, "tls.zig — TLS 1.2/1.3 implementation in Zig",
647 <https://github.com/ianic/tls.zig>. Provides the TLS on both sides; see
648 the **TLS** section for why the standard library's client is not used.
649- **[ISEMAIL]** Sayers, D., "is_email — an email address validator and its
650 test suite", BSD-3-Clause, <https://github.com/dominicsayers/isemail>.
651 The address corpus the path parser is checked against; see **Tests**.
652- **[EXIM]** The Exim Maintainers, "Exim Internet Mailer",
653 GPL-2.0-or-later, <https://www.exim.org/>. The protocol torture script
654 and the gauntlet unit test's dialogue are adapted from its test suite.
655
656## Tests
657
658```sh
659zig build test
660zig build test --fuzz # run the fuzz tests under the fuzzer (endless)
661```
662
663The fuzz tests cover parser crash-safety (`Command.parse`, `Reply.read`),
664whole-session robustness against arbitrary bytes on both the client and
665server side, and two differential properties: the streaming `DataWriter`
666must produce byte-identical output to the slice-based `writeStuffed` under
667fuzzer-chosen chunk boundaries, and the collecting and streaming server
668DATA paths must yield identical message content.
669
670### Protocol torture testing with exim's test client
671
672Exim's scriptable SMTP test client (`test/src/client.c` in the exim
673source) sends raw protocol lines and asserts reply prefixes. The exim
674source is declared as a *lazy* Zig dependency, fetched only on demand:
675
676```sh
677zig build -Dexim-client # fetches exim, installs zig-out/bin/exim-client
678./zig-out/bin/zig-smtp serve 2525 &
679./zig-out/bin/exim-client 127.0.0.1 2525 < test/protocol-torture.script
680```
681
682### Address corpus testing with the is_email suite
683
684Dominic Sayers' [is_email](https://github.com/dominicsayers/isemail) test
685suite (BSD-3-Clause) is declared as a *lazy* Zig dependency; nothing from
686it is copied into this repository. On demand, the corpus test embeds its
687XML test files, extracts the 125 addresses valid at the RFC 5321 layer,
688and checks that each passes through the path parser byte-for-byte:
689
690```sh
691zig build test -Disemail-corpus # fetches the suite and runs the corpus test
692```
693
694Without the option the corpus test is skipped.
695
696`test/protocol-torture.script` is a 28-reply dialogue distilled from
697exim's own test suite (syntax errors, sequencing violations, parameter
698validation, dot-stuffing); the same dialogue is asserted byte-for-byte
699as a unit test in `Server.zig`.
700
701The library is MIT-licensed; the small amount of test-only material adapted
702from exim's test suite (the torture script and the gauntlet unit test's
703dialogue) is GPL-2.0-or-later, marked with SPDX snippet tags and REUSE.toml
704annotations.
705
706Note: Zig 0.16.0's fuzz *driver* is broken out of the box (its bundled
707test runner fails to compile in fuzz mode, and the coverage server panics
708on a test binary with no fuzz tests); both are fixed on Zig master. Until
709then, fuzzing needs a patched copy of the standard library via
710`zig build --zig-lib-dir <patched-lib> test --fuzz`. The fuzz tests
711themselves also run once per invocation as part of the normal
712`zig build test` suite.
713
714Interoperability against third-party implementations is covered by a NixOS
715VM test (`nix/interop-test.nix`): the zig-smtp client delivers mail to Postfix
716and Exim over plaintext, STARTTLS, and implicit TLS against each, and swaks
717delivers to the zig-smtp server over plaintext and STARTTLS.
718
719```sh
720nix build .#zig-smtp # build the package
721nix build .#checks.x86_64-linux.interop # run the VM interop test
722```