node_modules: update (#314)

Co-authored-by: dawidd6 <9713907+dawidd6@users.noreply.github.com>
This commit is contained in:
Dawid Dziurla
2026-09-07 07:05:39 +02:00
committed by GitHub
parent 40eba5cee7
commit d35571df19
23 changed files with 938 additions and 176 deletions
+3 -3
View File
@@ -94,9 +94,9 @@
}
},
"node_modules/nodemailer": {
"version": "9.0.3",
"resolved": "https://registry.npmjs.org/nodemailer/-/nodemailer-9.0.3.tgz",
"integrity": "sha512-n+YP+NKwR5zRWa60k3GiQ6Q3B4KXCoAw40dAKeCtYn020iNN74aWK2liXIC3ZEATeGql7we3tE3t8QwhY0eskw==",
"version": "9.1.1",
"resolved": "https://registry.npmjs.org/nodemailer/-/nodemailer-9.1.1.tgz",
"integrity": "sha512-izw9mVKFix6YSnC9eLgV6g1opl9DUlRio9ZNcq+Wu9Ujn2UwF+8Nl0B8nz22kEC+CTZCvinkxwJ0DeFbb6NwcQ==",
"license": "MIT-0",
"engines": {
"node": ">=6.0.0"
+61
View File
@@ -1,5 +1,66 @@
# CHANGELOG
## [9.1.1](https://github.com/nodemailer/nodemailer/compare/v9.1.0...v9.1.1) (2026-09-01)
### Bug Fixes
* **mailer:** apply the message access policy in resolveContent ([dc48ed3](https://github.com/nodemailer/nodemailer/commit/dc48ed395c4d6c79ee5c95eb6eff17bafe391474))
* **mailer:** keep message data from reopening the access sandbox ([ab7ef34](https://github.com/nodemailer/nodemailer/commit/ab7ef348b9a97b1fd70e7bfbeb56d4ea4a07946b))
* **mime-node:** inherit the access policy from the tree a node hangs in ([262d550](https://github.com/nodemailer/nodemailer/commit/262d550b1e121e3ff4ef6675d6771a5b2b4ddcec))
## [9.1.0](https://github.com/nodemailer/nodemailer/compare/v9.0.6...v9.1.0) (2026-08-31)
### Features
* **mailer:** cap recipients per message with maxRecipients ([7279ac8](https://github.com/nodemailer/nodemailer/commit/7279ac8dee4f66c032981e6e51e3e7210ad0dcbf))
### Bug Fixes
* **addressparser:** handle address lists in linear time ([9116da9](https://github.com/nodemailer/nodemailer/commit/9116da9528c6524cefaed75185602a7e85d20434))
* **addressparser:** terminate the domain at an RFC 5322 comment ([902b63e](https://github.com/nodemailer/nodemailer/commit/902b63e935435c30f4025901c0902dce64cd8880))
* **mime-node:** apply UTS-46 mapping when encoding a domain ([259c32d](https://github.com/nodemailer/nodemailer/commit/259c32d7d266301e3377a212776c3fff993c0148))
* **mime-node:** dedupe envelope recipients in linear time ([7cc38af](https://github.com/nodemailer/nodemailer/commit/7cc38af418ffa6fc7e86085195ca5ca681694b3e))
* **mime-node:** flatten parsed addresses without concat.apply ([83b8c48](https://github.com/nodemailer/nodemailer/commit/83b8c48cbdb8b3116f2e1ba84af755b2c5661c0f))
* **mime-node:** keep the recipient dedupe linear across address headers ([34da642](https://github.com/nodemailer/nodemailer/commit/34da64282dcdc9b0581c721a27ab2fa226673150))
* **mime-node:** keep URL delimiters away from the domain mapper ([b212ac4](https://github.com/nodemailer/nodemailer/commit/b212ac4e27bce8182478044fcb8d1642ccdad46e))
## [9.0.6](https://github.com/nodemailer/nodemailer/compare/v9.0.5...v9.0.6) (2026-08-27)
### Bug Fixes
* **addressparser:** recover the addr-spec from an angle-addr holding whitespace ([e989a22](https://github.com/nodemailer/nodemailer/commit/e989a22ca4f5161929bf37be8fb07de635016fa7))
* harden copies of user supplied keys and URL fetching ([2f667f4](https://github.com/nodemailer/nodemailer/commit/2f667f4272cb2d7cb479b2e3903ab10600fc0eae))
## [9.0.5](https://github.com/nodemailer/nodemailer/compare/v9.0.4...v9.0.5) (2026-08-07)
### Bug Fixes
* **ci:** retrigger the workflows dropped during the Actions outage ([85d16c1](https://github.com/nodemailer/nodemailer/commit/85d16c103ec69e237c7e55c0e3103f439c135ce3))
* **mailer:** escape specials in List-* header comments ([#1842](https://github.com/nodemailer/nodemailer/issues/1842)) ([75913bb](https://github.com/nodemailer/nodemailer/commit/75913bba032623046dd7fa037b8b880e388d8357))
* **mime-funcs:** star the continuation key of a restarted parameter line ([36bcf1a](https://github.com/nodemailer/nodemailer/commit/36bcf1a21a92b5a283c780e35aacd3080292d98d))
* **mime-node:** keep control chars out of header values and msg-id headers ([15cf6d1](https://github.com/nodemailer/nodemailer/commit/15cf6d1c15cdf60f551618fd54fb727ea7aac94c))
* **mime:** encode DEL in header parameters and List-* comments ([cf69430](https://github.com/nodemailer/nodemailer/commit/cf69430ffac1d246bfbab564daf321f31ed9e1dd))
* **mime:** keep control chars out of the remaining header positions ([5ed9d26](https://github.com/nodemailer/nodemailer/commit/5ed9d26f85eecb48f4b3ae74ad33ed2abf002040))
* **mime:** normalize an address parsed out of a string as well ([63685f7](https://github.com/nodemailer/nodemailer/commit/63685f7dd4aefa75cc72a36f983f32f61cd6733e))
* **mime:** normalize an address so header and envelope agree ([a9343b4](https://github.com/nodemailer/nodemailer/commit/a9343b47e42b9ccb27911ad8e73d4119c6170c85))
* **mime:** stop a header key callback and the dkim tags from injecting ([b7d772e](https://github.com/nodemailer/nodemailer/commit/b7d772ea4ec12ee82e65a9b919af882bf0f125d9))
## [9.0.4](https://github.com/nodemailer/nodemailer/compare/v9.0.3...v9.0.4) (2026-08-04)
### Bug Fixes
* **mime-funcs:** do not let an unpaired surrogate consume the next character ([9797f7f](https://github.com/nodemailer/nodemailer/commit/9797f7f57d47b1e27d8ae050550ba93600dbf9f4))
* **mime-funcs:** keep any surrogate pair intact when chunking base64 mime words ([#1838](https://github.com/nodemailer/nodemailer/issues/1838)) ([5bd3a65](https://github.com/nodemailer/nodemailer/commit/5bd3a657be2d12df1ac2838a1673cb4885e99016))
* **mime-funcs:** percent encode unpaired surrogates in header parameter values ([78f4aa2](https://github.com/nodemailer/nodemailer/commit/78f4aa253d0ebb9c5ba91301adacef5fb6cf5493))
* **mime-node:** escape backslash and quote in the Content-Type name parameter ([#1837](https://github.com/nodemailer/nodemailer/issues/1837)) ([adcfc4f](https://github.com/nodemailer/nodemailer/commit/adcfc4f46445edc8a14136b2bf3f775943425c13))
* **mime:** encode HT/CR/LF in header parameter values instead of quoting them ([#1840](https://github.com/nodemailer/nodemailer/issues/1840)) ([5bc9cab](https://github.com/nodemailer/nodemailer/commit/5bc9cabddcb8d18d16244701ae9facc8fba942a3))
## [9.0.3](https://github.com/nodemailer/nodemailer/compare/v9.0.2...v9.0.3) (2026-06-30)
+3 -6
View File
@@ -8,10 +8,7 @@ Send emails from Node.js easy as cake! 🍰✉️
See [nodemailer.com](https://nodemailer.com/) for documentation and terms.
> [!TIP]
> Check out **[EmailEngine](https://emailengine.app/?utm_source=github-nodemailer&utm_campaign=nodemailer&utm_medium=readme-link)** a self-hosted email gateway that allows making **REST requests against IMAP and SMTP servers**. EmailEngine also sends webhooks whenever something changes on the registered accounts.\
> \
> Using the email accounts registered with EmailEngine, you can receive and [send emails](https://emailengine.app/sending-emails?utm_source=github-nodemailer&utm_campaign=nodemailer&utm_medium=readme-link). EmailEngine supports OAuth2, delayed sends, opens and clicks tracking, bounce detection, etc. All on top of regular email accounts without an external MTA service.
> Nodemailer is developed by the team behind **[EmailEngine](https://emailengine.app/?utm_source=nodemailer-readme&utm_medium=readme&utm_campaign=oss-docs)**, a self-hosted email API that turns any Gmail, Microsoft 365, or IMAP account into a REST endpoint, with managed OAuth2, webhooks for incoming mail, and built-in [sending](https://emailengine.app/sending-emails?utm_source=nodemailer-readme&utm_medium=readme&utm_campaign=oss-docs). If you would rather call an HTTP API than maintain IMAP and SMTP connections yourself, that is what it is for.
## Having an issue?
@@ -25,7 +22,7 @@ You are using an older Node.js version than v6.0. Upgrade Node.js to get support
#### I'm having issues with Gmail
Gmail either works well, or it does not work at all. It is probably easier to switch to an alternative service instead of fixing issues with Gmail. If Gmail does not work for you, then don't use it. Read more about it [here](https://nodemailer.com/usage/using-gmail/).
Gmail either works well, or it does not work at all. It is probably easier to switch to an alternative service instead of fixing issues with Gmail. If Gmail does not work for you, then don't use it. Read more about it [here](https://nodemailer.com/usage/using-gmail/). If the blocker is OAuth2 setup rather than Gmail itself, [EmailEngine](https://emailengine.app/?utm_source=nodemailer-readme&utm_medium=readme&utm_campaign=oss-docs&utm_content=faq-gmail) handles the OAuth2 flow and token refresh for you.
#### I get ETIMEDOUT errors
@@ -55,7 +52,7 @@ let configOptions = {
#### I have issues with DNS / hosts file
Node.js uses [c-ares](https://nodejs.org/en/docs/meta/topics/dependencies/#c-ares) to resolve domain names, not the DNS library provided by the system, so if you have some custom DNS routing set up, it might be ignored. Nodemailer runs [dns.resolve4()](https://nodejs.org/dist/latest-v16.x/docs/api/dns.html#dnsresolve4hostname-options-callback) and [dns.resolve6()](https://nodejs.org/dist/latest-v16.x/docs/api/dns.html#dnsresolve6hostname-options-callback) to resolve hostname into an IP address. If both calls fail, then Nodemailer will fall back to [dns.lookup()](https://nodejs.org/dist/latest-v16.x/docs/api/dns.html#dnslookuphostname-options-callback). If this does not work for you, you can hard code the IP address into the configuration like shown below. In that case, Nodemailer would not perform any DNS lookups.
Node.js uses [c-ares](https://github.com/c-ares/c-ares) to resolve domain names, not the DNS library provided by the system, so if you have some custom DNS routing set up, it might be ignored. Nodemailer runs [dns.resolve4()](https://nodejs.org/dist/latest-v16.x/docs/api/dns.html#dnsresolve4hostname-options-callback) and [dns.resolve6()](https://nodejs.org/dist/latest-v16.x/docs/api/dns.html#dnsresolve6hostname-options-callback) to resolve hostname into an IP address. If both calls fail, then Nodemailer will fall back to [dns.lookup()](https://nodejs.org/dist/latest-v16.x/docs/api/dns.html#dnslookuphostname-options-callback). If this does not work for you, you can hard code the IP address into the configuration like shown below. In that case, Nodemailer would not perform any DNS lookups.
```js
let configOptions = {
+2 -2
View File
@@ -11,8 +11,8 @@ supported way to receive security updates.
| Version | Supported |
| ------- | ------------------ |
| 8.x | :white_check_mark: |
| < 8.0 | :x: |
| 9.x | :white_check_mark: |
| < 9.0 | :x: |
If you are on an older major, please upgrade. See the migration notes at
<https://nodemailer.com/> before updating.
+166 -8
View File
@@ -1,5 +1,134 @@
'use strict';
/**
* Restores the quoting of a local part that was read out of a quoted string.
*
* RFC 5321 allows '@' inside a quoted local part, so handing '"user@evil.com"@good.com'
* on as the bare 'user@evil.com@good.com' leaves it to the consumer which '@' splits the
* domain off. Getting that wrong is a misrouting vector, so the quotes go back on. The
* same holds for the other specials: a ',' or a ';' that loses its quotes reads as a
* recipient separator once the consumer puts the address back into a header.
*
* This module has no dependencies so that it can ship on its own, which is why the two
* grammar tests below are spelled out here instead of shared with lib/mime-node. Keeping
* only what is ambiguous quoted is deliberate, mime-node applies the stricter RFC 5321
* dot-atom rule on top of this when it emits an address.
*
* @param {String} address Address with an unquoted local part
* @return {String} Address with the local part as a quoted-string
*/
function _quoteLocalPart(address) {
const lastAt = address.lastIndexOf('@');
if (lastAt < 0) {
// no domain to split off, nothing can be misrouted
return address;
}
const user = address.substr(0, lastAt);
if (/^[^\s"(),:;<>@[\\\]]+$/.test(user) || /^"(?:[^"\\]|\\[\s\S])*"$/.test(user)) {
// a local part that carries no special reads the same with or without the quotes,
// and one that is already a complete quoted-string needs nothing either
return address;
}
return '"' + user.replace(/["\\]/g, '\\$&') + '"@' + address.substr(lastAt + 1);
}
/**
* Reached for every parsed address, so it is built once rather than per call.
*/
const HAS_WHITESPACE = /\s/;
/**
* An addr-spec that carries its whitespace legally, inside a quoted local part. The
* optional tail is the malformed shape: a real mailbox with wreckage trailing it.
*/
const QUOTED_LOCAL_ADDR = /^("(?:[^"\\]|\\[\s\S])*"@\S+)(?:\s+([\s\S]+))?$/;
/**
* One run holding a single '@' and no whitespace, the shape an addr-spec has to have.
*/
const ADDR_SPEC = /^[^@\s]+@[^@\s]+$/;
/**
* The looser reading applied once the strict one finds nothing, which tolerates the
* further '@' that a domain should not have but malformed headers carry anyway.
*/
const LOOSE_ADDR_SPEC = /^[^@\s]+@\S+$/;
/**
* Recovers the addr-spec from an angle-addr that came back holding unquoted whitespace.
*
* A malformed header can put more than a mailbox between the angle brackets, most often
* because the generator wrote the recipient twice: '<user@example.com user@example.com>'
* or '<example.com user@example.com>'. Whitespace is not addr-spec, so the whole run can
* never be a mailbox anyone could deliver to, and passing it on as the address loses the
* recipient that is sitting right there in the header.
*
* The run that still reads as an addr-spec is kept and whatever is left over becomes
* display text rather than being dropped. Candidates are read strictly first and then
* under the looser grammar, the same two tiers the unquoted-text branch below applies to
* the same problem, so that '<a@b@c.com junk>' and a bare 'a@b@c.com junk' agree on the
* recipient. When several runs qualify the first wins, which is what that branch's looser
* tier does within a token.
*
* A quoted local part is left alone: RFC 5321 allows whitespace inside it, so
* '<"user name"@example.com>' is well formed and means exactly what it says.
*
* @param {Object} data Collected address parts, mutated in place
*/
function _recoverAddrSpec(data) {
if (!HAS_WHITESPACE.test(data.address)) {
return;
}
let address;
let rest;
const quoted = data.address.match(QUOTED_LOCAL_ADDR);
if (quoted) {
if (!quoted[2]) {
// the whitespace sits inside the quoted local part, this is a well formed mailbox
return;
}
// a real mailbox with wreckage trailing it, so peel the addr-spec off whole rather
// than splitting into the quotes
address = quoted[1];
rest = [quoted[2]];
} else {
if (data.address.indexOf('"') >= 0) {
// Splitting on whitespace loses track of where the quoted string starts and ends,
// and this module does not take addresses out of quoted strings: the run picked out
// of '<junk "user@evil.com b"@good.com>' would be an address from the domain the
// quotes were hiding. Every well formed shape was already handled above, so what is
// left is wreckage either way and the original is the honest answer
return;
}
const parts = data.address.split(/\s+/);
let addrIndex = parts.findIndex(part => ADDR_SPEC.test(part));
if (addrIndex < 0) {
addrIndex = parts.findIndex(part => LOOSE_ADDR_SPEC.test(part));
}
if (addrIndex < 0) {
// nothing in there reads as an address, there is no better answer than the original
return;
}
address = parts.splice(addrIndex, 1)[0];
rest = parts;
}
data.address = address;
data.text = [data.text]
.concat(rest)
.filter(part => part)
.join(' ');
}
/**
* Converts tokens for a single address into an address object
*
@@ -55,7 +184,18 @@ function _handleAddress(tokens, depth) {
token.value = token.value.replace(/^[^<]*<\s*/, '');
}
if (prevToken && prevToken.noBreak && data[state].length) {
// A comment is folding whitespace. It may sit inside an addr-spec, on either side
// of the '@', but it cannot join two atoms into one: gluing across it would read
// 'user@example.com(x)evil.com' as the single domain 'example.comevil.com' and
// deliver to a domain the sender never named.
const parts = data[state];
const joins =
prevToken &&
prevToken.noBreak &&
parts.length &&
(prevToken.value !== ')' || parts[parts.length - 1].slice(-1) === '@' || token.value.charAt(0) === '@');
if (joins) {
data[state][data[state].length - 1] += token.value;
if (state === 'text' && insideQuotes) {
data.textWasQuoted[data.textWasQuoted.length - 1] = true;
@@ -103,7 +243,7 @@ function _handleAddress(tokens, depth) {
// Security: Do not extract email addresses from quoted strings.
// RFC 5321 allows @ inside quoted local-parts like "user@domain"@example.com.
// Extracting emails from quoted text leads to misrouting vulnerabilities.
if (!data.textWasQuoted[i] && /^[^@\s]+@[^@\s]+$/.test(data.text[i])) {
if (!data.textWasQuoted[i] && ADDR_SPEC.test(data.text[i])) {
data.address = data.text.splice(i, 1);
data.textWasQuoted.splice(i, 1);
break;
@@ -145,10 +285,16 @@ function _handleAddress(tokens, depth) {
data.text = data.text.concat(data.address.splice(1));
}
// An address is only taken from unquoted text, so anything left in the text at this
// point that still has to serve as the address carries its quoting in this flag
const addressFromQuotedText = !data.address.length && data.textWasQuoted.some(wasQuoted => wasQuoted);
// Join values with spaces
data.text = data.text.join(' ');
data.address = data.address.join(' ');
_recoverAddrSpec(data);
const address = {
address: data.address || data.text || '',
name: data.text || data.address || ''
@@ -162,6 +308,10 @@ function _handleAddress(tokens, depth) {
}
}
if (addressFromQuotedText && address.address) {
address.address = _quoteLocalPart(address.address);
}
addresses.push(address);
}
@@ -360,8 +510,10 @@ function addressparser(str, options) {
addresses.forEach(addr => {
const handled = _handleAddress(addr, depth);
if (handled.length) {
parsedAddresses = parsedAddresses.concat(handled);
// Appended in place. Rebuilding the accumulator with concat() would copy every
// entry collected so far on each address, making a flat list cost O(n^2).
for (let i = 0; i < handled.length; i++) {
parsedAddresses.push(handled[i]);
}
});
@@ -369,14 +521,20 @@ function addressparser(str, options) {
// "Joe Foo, PhD <joe@example.com>" is split on the comma into
// [{name:"Joe Foo", address:""}, {name:"PhD", address:"joe@example.com"}].
// Recombine: a name-only entry followed by an entry with both name and address.
for (let i = parsedAddresses.length - 2; i >= 0; i--) {
// Walked back to front so that a run of fragments folds into one entry in a single
// pass. Splicing each fragment out of the list instead would cost O(n^2).
const mergedAddresses = [];
for (let i = parsedAddresses.length - 1; i >= 0; i--) {
const current = parsedAddresses[i];
const next = parsedAddresses[i + 1];
if (current.address === '' && current.name && !current.group && next.address && next.name) {
const next = mergedAddresses.length ? mergedAddresses[mergedAddresses.length - 1] : null;
if (next && current.address === '' && current.name && !current.group && next.address && next.name) {
next.name = current.name + ', ' + next.name;
parsedAddresses.splice(i, 1);
} else {
mergedAddresses.push(current);
}
}
mergedAddresses.reverse();
parsedAddresses = mergedAddresses;
if (options.flatten) {
const flatAddresses = [];
+6 -1
View File
@@ -10,6 +10,7 @@ const { PassThrough } = require('stream');
const fs = require('fs');
const path = require('path');
const crypto = require('crypto');
const { copyOwnKeys } = require('../shared/objects');
const DKIM_ALGO = 'sha256';
const MAX_MESSAGE_SIZE = 2 * 1024 * 1024; // buffer messages larger than this to disk
@@ -225,7 +226,11 @@ class DKIM {
let options = this.options;
if (extraOptions && Object.keys(extraOptions).length) {
options = Object.assign({}, extraOptions, this.options);
// extraOptions is mail.data._dkim, caller supplied message data. An own
// "__proto__" key there would let every option this signer reads and the
// transport did not set, such as skipFields, answer from the caller
options = copyOwnKeys({}, extraOptions);
copyOwnKeys(options, this.options);
}
const signer = new DKIMSigner(options, this.keys, inputStream, output);
+8 -3
View File
@@ -50,15 +50,20 @@ module.exports = (headers, hashAlgo, bodyHash, options) => {
module.exports.relaxedHeaders = relaxedHeaders;
function generateDKIMHeader(domainName, keySelector, fieldNames, hashAlgo, bodyHash) {
// the caller supplied tag values are interpolated straight into the tag list, and none of
// them has any way to carry a control char, DEL, or one of the delimiters that would close
// the value and open a tag of its own
const cleanTagValue = value => (value || '').toString().replace(/[\x00-\x1f\x7f;=]/g, '');
const dkim = [
'v=1',
'a=rsa-' + hashAlgo,
'c=relaxed/relaxed',
'd=' + punycode.toASCII(domainName),
'd=' + punycode.toASCII(cleanTagValue(domainName)),
'q=dns/txt',
's=' + keySelector,
's=' + cleanTagValue(keySelector),
'bh=' + bodyHash,
'h=' + fieldNames
'h=' + cleanTagValue(fieldNames)
].join('; ');
return mimeFuncs.foldLines('DKIM-Signature: ' + dkim, 76) + ';\r\n b=';
+1
View File
@@ -38,6 +38,7 @@ const ERROR_CODES = {
// Resource errors
EMAXLIMIT: 'Pool resource limit reached (max messages per connection)',
EMAXRECIPIENTS: 'Recipient count exceeds maxRecipients',
// Transport-specific errors
ESENDMAIL: 'Sendmail command error',
+115 -5
View File
@@ -9,9 +9,67 @@ const Cookies = require('./cookies');
const packageData = require('../../package.json');
const net = require('net');
const errors = require('../errors');
const { isProtoKey } = require('../shared/objects');
const MAX_REDIRECTS = 5;
// Only genuine TLS settings are taken from options.tls. That object reaches us straight
// from a user supplied attachment (content.tls), so keys like host, port, path, socketPath
// or lookup would otherwise repoint the request at a destination that never went through
// the URL checks below.
//
// The source of truth is the tls.connect() option list in the Node docs. A key missing
// here is dropped silently, so extend this list rather than working around it.
const TLS_OPTION_KEYS = [
'ALPNProtocols',
'ca',
'cert',
'checkServerIdentity',
'ciphers',
'crl',
'dhparam',
'ecdhCurve',
'honorCipherOrder',
'key',
'maxVersion',
'minVersion',
'passphrase',
'pfx',
'rejectUnauthorized',
'secureContext',
'secureOptions',
'secureProtocol',
'servername',
'sessionIdContext',
'sigalgs'
];
/**
* Resolves a URL only if it is one this module is willing to request.
*
* urllib.parse throws for a host that contains forbidden bytes, and it is called for
* every URL that reaches nmfetch, including ones that arrive from a message attachment
* or from a redirect Location header. An uncaught throw here takes the process down,
* so a URL that does not parse is reported the same way as one with a scheme we refuse.
*
* @param {String} url URL to parse
* @returns {Object|Boolean} Parsed URL, or false if it is not a usable http(s) URL
*/
function parseFetchUrl(url) {
let parsed;
try {
parsed = urllib.parse(url);
} catch (_err) {
return false;
}
if (parsed.protocol !== 'http:' && parsed.protocol !== 'https:') {
return false;
}
return parsed;
}
module.exports = function (url, options) {
return nmfetch(url, options);
};
@@ -26,6 +84,30 @@ function nmfetch(url, options) {
options.redirects = options.redirects || 0;
options.maxRedirects = isNaN(options.maxRedirects) ? MAX_REDIRECTS : options.maxRedirects;
const fetchRes = options.fetchRes;
const parsed = parseFetchUrl(url);
if (!parsed) {
// Only http(s) URLs can be fetched. Any other scheme (file:, gopher:, a
// protocol-relative redirect target etc.) would otherwise be silently served over
// plain HTTP, possibly against an unintended host. Bail out before the cookie jar
// is touched so a refused URL can not seed it, and release a caller supplied body:
// this is the one exit that runs before the error handler below is attached to it,
// so an error on that stream would have nowhere to go and the fd or socket behind
// it would never be released.
if (options.body && typeof options.body.destroy === 'function') {
options.body.on('error', () => false);
options.body.destroy();
}
setImmediate(() => {
const err = new Error('Unsupported protocol for URL ' + url);
err.code = errors.EFETCH;
err.sourceUrl = url;
fetchRes.emit('error', err);
});
return fetchRes;
}
if (options.cookie) {
[].concat(options.cookie || []).forEach(cookie => {
options.cookies.set(cookie, url);
@@ -33,8 +115,6 @@ function nmfetch(url, options) {
options.cookie = false;
}
const fetchRes = options.fetchRes;
const parsed = urllib.parse(url);
let method = (options.method || '').toString().trim().toUpperCase() || 'GET';
let finished = false;
let cookies;
@@ -48,6 +128,10 @@ function nmfetch(url, options) {
};
Object.keys(options.headers || {}).forEach(key => {
// options.headers is the caller's httpHeaders, straight off an attachment
if (isProtoKey(key.toLowerCase().trim())) {
return;
}
headers[key.toLowerCase().trim()] = options.headers[key];
});
@@ -131,7 +215,12 @@ function nmfetch(url, options) {
};
if (options.tls) {
Object.assign(reqOptions, options.tls);
// see TLS_OPTION_KEYS
Object.keys(options.tls).forEach(key => {
if (TLS_OPTION_KEYS.includes(key)) {
reqOptions[key] = options.tls[key];
}
});
}
if (
@@ -216,8 +305,29 @@ function nmfetch(url, options) {
options.method = 'GET';
options.body = false;
const redirectUrl = urllib.resolve(url, res.headers.location);
const redirectParsed = urllib.parse(redirectUrl);
let redirectUrl;
try {
redirectUrl = urllib.resolve(url, res.headers.location);
} catch (_err) {
// the legacy resolver throws on a Location the WHATWG parser also refused,
// so fall through to the check below with what the server actually sent
redirectUrl = res.headers.location;
}
const redirectParsed = parseFetchUrl(redirectUrl);
if (!redirectParsed) {
// Refuse the redirect target here rather than leaving it to the recursive
// call: that call gets its own `finished` flag and no handle on this
// request, so this one would stay open and could emit a second error on
// the shared fetchRes once it times out. Callers listen with req.once().
finished = true;
const err = new Error('Unsupported protocol for URL ' + redirectUrl);
err.code = errors.EFETCH;
err.sourceUrl = redirectUrl;
fetchRes.emit('error', err);
req.abort();
return;
}
// Do not forward credentials when the redirect leaves the original
// security context: a different host, or a downgrade from https to
+1 -1
View File
@@ -33,7 +33,7 @@ class JSONTransport {
// Sendmail strips this header line by itself
mail.message.keepBcc = true;
const envelope = mail.data.envelope || mail.message.getEnvelope();
const envelope = mail.message.getEnvelope();
const messageId = mail.message.messageId();
const recipients = [].concat(envelope.to || []);
+5 -3
View File
@@ -4,7 +4,7 @@
const MimeNode = require('../mime-node');
const mimeFuncs = require('../mime-funcs');
const { parseDataURI } = require('../shared');
const { parseDataURI, copyOwnKeys } = require('../shared');
/**
* Creates the object for composing a MimeNode instance out from the mail options
@@ -205,7 +205,9 @@ class MailComposer {
typeof this.mail.icalEvent === 'object' &&
(this.mail.icalEvent.content || this.mail.icalEvent.path || this.mail.icalEvent.href || this.mail.icalEvent.raw)
) {
icalEvent = Object.assign({}, this.mail.icalEvent);
// an own "__proto__" key would make the copy inherit path/href from caller
// data, and the mapping below then replaces the content the caller did set
icalEvent = copyOwnKeys({}, this.mail.icalEvent);
} else {
icalEvent = {
content: this.mail.icalEvent
@@ -600,7 +602,7 @@ class MailComposer {
}
// Return empty content for excessively long data URLs
return Object.assign({}, element, {
return Object.assign(copyOwnKeys({}, element), {
path: false,
href: false,
content: Buffer.alloc(0),
+27
View File
@@ -15,6 +15,13 @@ const net = require('net');
const dns = require('dns');
const crypto = require('crypto');
/**
* Recipients allowed on one message unless the caller sets its own maxRecipients. A backstop
* against a runaway or hostile recipient list rather than a delivery policy: RFC 5321 only
* asks a server to accept 100, so a real send is bounded far below this.
*/
const DEFAULT_MAX_RECIPIENTS = 100000;
/**
* Creates an object for exposing the Mail API
*
@@ -191,6 +198,26 @@ class Mail extends EventEmitter {
mail.setPriorityHeaders();
mail.setListHeaders();
const maxRecipients = mail.data.maxRecipients === undefined ? DEFAULT_MAX_RECIPIENTS : mail.data.maxRecipients;
const recipientCount = mail.message.getEnvelope().to.length;
if (maxRecipients && recipientCount > maxRecipients) {
const err = new Error(
`Message has ${recipientCount} recipients, which is over the ${maxRecipients} allowed by maxRecipients`
);
err.code = errors.EMAXRECIPIENTS;
this.logger.error(
{
err,
tnx: 'transport',
action: 'send'
},
'Send Error: %s',
err.message
);
return callback(err);
}
this._processPlugins('stream', mail, err => {
if (err) {
this.logger.error(
+67 -36
View File
@@ -4,6 +4,11 @@ const shared = require('../shared');
const MimeNode = require('../mime-node');
const mimeFuncs = require('../mime-funcs');
// Only an own key counts as already set. `key in obj` also matches every member of
// Object.prototype, which silently drops a transporter default legitimately named
// toString or constructor.
const hasOwn = (obj, key) => Object.prototype.hasOwnProperty.call(obj, key);
class MailMessage {
constructor(mailer, data) {
this.mailer = mailer;
@@ -14,34 +19,53 @@ class MailMessage {
const options = mailer.options || {};
const defaults = mailer._defaults || {};
Object.assign(this.data, data);
shared.copyOwnKeys(this.data, data);
this.data.headers = this.data.headers || {};
// apply defaults
Object.keys(defaults).forEach(key => {
if (!(key in this.data)) {
this.data[key] = defaults[key];
} else if (key === 'headers') {
// Apply defaults. `_defaults` is caller supplied too, it is the second argument of
// createTransport, so it needs the same treatment as `data` above
shared.copyOwnKeys(this.data, defaults, key => hasOwn(this.data, key));
// headers is a special case. Allow setting individual default headers
Object.keys(defaults.headers).forEach(key => {
if (!(key in this.data.headers)) {
this.data.headers[key] = defaults.headers[key];
}
});
}
});
shared.copyOwnKeys(this.data.headers, defaults.headers, key => hasOwn(this.data.headers, key));
// force specific keys from transporter options
['disableFileAccess', 'disableUrlAccess', 'normalizeHeaderKey'].forEach(key => {
['disableFileAccess', 'disableUrlAccess', 'normalizeHeaderKey', 'maxRecipients'].forEach(key => {
if (key in options) {
this.data[key] = options[key];
}
});
// The access flags are a sandbox rather than a message field, so `defaults` counts as
// transporter configuration for them. For a transporter plugin it is the only channel
// there is, createTransport leaves `options` undefined for one, and the defaults copy
// above yields to anything the message already set, which let message data switch the
// sandbox back off. Closing is one way here, same as in resolveContent below: either
// side may switch a flag on, neither can switch off what the other closed.
['disableFileAccess', 'disableUrlAccess'].forEach(key => {
if (!(key in options) && hasOwn(defaults, key)) {
this.data[key] = this.data[key] || defaults[key];
}
});
}
resolveContent(...args) {
return shared.resolveContent(...args);
resolveContent(data, key, options, callback) {
// Most plugins call this with the legacy (data, key, callback) signature, which carries
// no access policy. The policy belongs to the message, so apply it here. Explicit
// options may only tighten it, never reopen what the transporter closed.
if (!callback && typeof options === 'function') {
callback = options;
options = false;
}
options = options || {};
const policy = {
disableFileAccess: this.data.disableFileAccess || options.disableFileAccess,
disableUrlAccess: this.data.disableUrlAccess || options.disableUrlAccess
};
return shared.resolveContent(data, key, policy, callback);
}
resolveAll(callback) {
@@ -123,11 +147,12 @@ class MailMessage {
content: value
};
if (args[0][args[1]] && typeof args[0][args[1]] === 'object' && !Buffer.isBuffer(args[0][args[1]])) {
Object.keys(args[0][args[1]]).forEach(key => {
if (!(key in node) && !['content', 'path', 'href', 'raw'].includes(key)) {
node[key] = args[0][args[1]][key];
}
});
// The keys are the caller's, so copying them takes the same "__proto__"
// rule as the constructor. `key in node` stays as the already-set test
// here, unlike for the defaults: it also skips the Object.prototype
// member names, and letting message data land a `toString` string on a
// node only buys a TypeError the first time something stringifies it.
shared.copyOwnKeys(node, args[0][args[1]], key => key in node || ['content', 'path', 'href', 'raw'].includes(key));
}
args[0][args[1]] = node;
@@ -140,7 +165,7 @@ class MailMessage {
}
normalize(callback) {
const envelope = this.data.envelope || this.message.getEnvelope();
const envelope = this.message.getEnvelope();
const messageId = this.message.messageId();
this.resolveAll((err, data) => {
@@ -186,6 +211,9 @@ class MailMessage {
data.normalizedHeaders = {};
Object.keys(data.headers || {}).forEach(key => {
if (shared.isProtoKey(key)) {
return;
}
let value = [].concat(data.headers[key] || []).shift();
value = (value && value.value) || value;
if (value) {
@@ -271,15 +299,16 @@ class MailMessage {
}
if (value && value.url) {
if (key.toLowerCase().trim() === 'id') {
// List-ID: "comment" <domain>
// strip CR/LF so a comment can't inject extra header lines
// strip CR/LF so a comment can't inject extra header lines. DEL is neither
// qtext nor ctext, so it can not be carried literally by either construct
// and has to become an encoded word like any other non-plaintext value
let comment = (value.comment || '').toString().replace(/\r?\n|\r/g, ' ');
if (mimeFuncs.isPlainText(comment)) {
comment = '"' + comment + '"';
} else {
comment = mimeFuncs.encodeWord(comment);
}
const needsEncoding = !mimeFuncs.isPlainText(comment) || /\x7f/.test(comment);
if (key.toLowerCase().trim() === 'id') {
// List-ID: "comment" <domain>, where an unescaped quote or a trailing
// backslash in the comment would swallow the <domain> behind it
comment = needsEncoding ? mimeFuncs.encodeWord(comment) : mimeFuncs.quoteString(comment);
// List-ID expects a bare domain-like identifier, so strip the
// scheme prefix that _formatListUrl adds or passes through
@@ -289,11 +318,11 @@ class MailMessage {
}
// List-*: <http://domain> (comment)
// strip CR/LF so a comment can't inject extra header lines
let comment = (value.comment || '').toString().replace(/\r?\n|\r/g, ' ');
if (!mimeFuncs.isPlainText(comment)) {
comment = mimeFuncs.encodeWord(comment);
}
// the ctext specials go out as quoted-pairs, otherwise a ")" closes the
// comment early and leaves the rest as junk, an unpaired "(" opens a
// nested comment that never closes, and a trailing backslash escapes
// the closing ")" so the comment swallows whatever follows it
comment = needsEncoding ? mimeFuncs.encodeWord(comment) : comment.replace(/[()\\]/g, '\\$&');
return this._formatListUrl(value.url) + (value.comment ? ' (' + comment + ')' : '');
}
@@ -307,7 +336,9 @@ class MailMessage {
}
_formatListUrl(url) {
url = url.replace(/[\s<]+|[\s>]+/g, '');
// a url has no way to carry a control char or DEL, and the angle brackets around it
// are not a quoting construct, so anything left here lands in the header raw
url = url.replace(/[\x00-\x08\x0b\x0c\x0e-\x1f\x7f]/g, '').replace(/[\s<]+|[\s>]+/g, '');
if (/^(https?|mailto|ftp):/.test(url)) {
return '<' + url + '>';
}
+80 -19
View File
@@ -5,19 +5,39 @@
const base64 = require('../base64');
const qp = require('../qp');
const mimeTypes = require('./mime-types');
const { isProtoKey } = require('../shared/objects');
module.exports = {
/**
* Checks if a value is plaintext string (uses only printable 7bit chars)
*
* When isParam is set the value is destined for a header parameter, so HT, CR and LF
* are not plaintext either: a header parameter has no way to carry them. HT is a valid
* fold point, so folding and unfolding a header would rewrite it as a space, and CR/LF
* cannot appear in a header value at all. DEL is neither a token character nor qtext,
* so it can not be carried bare or quoted. Such values have to go through the rfc2231
* parameter continuation encoding instead, the same way a quote already does.
*
* @param {String} value String to be tested
* @param {Boolean} [isParam] Set to true if the value is a header parameter value
* @returns {Boolean} true if it is a plaintext string
*/
isPlainText(value, isParam) {
const re = isParam ? /[\x00-\x08\x0b\x0c\x0e-\x1f"\u0080-\uFFFF]/ : /[\x00-\x08\x0b\x0c\x0e-\x1f\u0080-\uFFFF]/;
const re = isParam ? /[\x00-\x1f\x7f"\u0080-\uFFFF]/ : /[\x00-\x08\x0b\x0c\x0e-\x1f\u0080-\uFFFF]/;
return typeof value === 'string' && !re.test(value);
},
/**
* Wraps a value into a quoted-string. Inside one a quote would end the string early
* and a backslash would escape whatever follows it, so both go out as quoted-pairs.
*
* @param {String} value String to be quoted
* @returns {String} The value as a quoted-string, quotes included
*/
quoteString(value) {
return '"' + (value || '').toString().replace(/["\\]/g, '\\$&') + '"';
},
/**
* Checks if a multi line string containes lines longer than the selected value.
*
@@ -80,8 +100,10 @@ module.exports = {
for (let i = 0, len = encodedStr.length; i < len; i++) {
let chr = encodedStr.charAt(i);
if (/[\ud83c\ud83d\ud83e]/.test(chr) && i < len - 1) {
// composite emoji byte, so add the next byte as well
if (/[\ud800-\udbff]/.test(chr) && /[\udc00-\udfff]/.test(encodedStr.charAt(i + 1))) {
// leading surrogate, so add the trailing surrogate as well
// an unpaired one must not swallow the next unit, that would destroy
// a valid pair following it
chr += encodedStr.charAt(++i);
}
@@ -168,10 +190,12 @@ module.exports = {
buildHeaderValue(structured) {
const paramsArray = [];
Object.keys(structured.params || {}).forEach(param => {
Object.keys(structured.params || {}).forEach(key => {
// filename might include unicode characters so it is a special case
// other values probably do not
const value = structured.params[param];
const value = structured.params[key];
// a parameter name is a token too and it is emitted without any quoting around it
const param = key.replace(/[\x00-\x1f\x7f]/g, '');
if (!this.isPlainText(value, true) || value.length >= 75) {
this.buildHeaderParam(param, value, 50).forEach(encodedParam => {
if (!/[\s"\\;:/=(),<>@[\]?]|^[-']|'$/.test(encodedParam.value) || encodedParam.key.substr(-1) === '*') {
@@ -187,7 +211,11 @@ module.exports = {
}
});
return structured.value + (paramsArray.length ? '; ' + paramsArray.join('; ') : '');
// the value ahead of the parameters is a token, it has no way to carry a control
// char or DEL and there is no quoting construct around it to escape one into
const value = typeof structured.value === 'string' ? structured.value.replace(/[\x00-\x1f\x7f]/g, '') : structured.value;
return value + (paramsArray.length ? '; ' + paramsArray.join('; ') : '');
},
/**
@@ -208,7 +236,7 @@ module.exports = {
buildHeaderParam(key, data, maxLength) {
const list = [];
let encodedStr = typeof data === 'string' ? data : (data || '').toString();
let chr, ord;
let chr;
let line;
let startPos = 0;
let i, len;
@@ -245,8 +273,9 @@ module.exports = {
const encodedStrArr = [];
for (i = 0, len = encodedStr.length; i < len; i++) {
chr = encodedStr.charAt(i);
ord = chr.charCodeAt(0);
if (ord >= 0xd800 && ord <= 0xdbff && i < len - 1) {
if (/[\ud800-\udbff]/.test(chr) && /[\udc00-\udfff]/.test(encodedStr.charAt(i + 1))) {
// an unpaired leading surrogate must not consume the next unit, that
// would tear apart a valid pair following it
chr += encodedStr.charAt(i + 1);
encodedStrArr.push(chr);
i++;
@@ -284,8 +313,11 @@ module.exports = {
line,
encoded
});
// the line we start here holds an encoded char, so it has to be
// flagged as one. otherwise it gets an unstarred continuation key
// and a receiver reads the percent escapes as literal text
line = '';
startPos = i - 1;
encoded = true;
} else {
encoded = true;
i = startPos;
@@ -350,6 +382,16 @@ module.exports = {
value: false,
params: {}
};
// Parameter names come from a caller supplied contentType/contentDisposition. A
// "__proto__" name would target the prototype chain of the params object instead of
// an own property of it, and read back as Object.prototype, so it is dropped.
const setParam = (name, value) => {
if (!isProtoKey(name)) {
response.params[name] = value;
}
};
let key = false;
let value = '';
let type = 'value';
@@ -381,7 +423,7 @@ module.exports = {
if (key === false) {
response.value = value.trim();
} else {
response.params[key] = value.trim();
setParam(key, value.trim());
}
type = 'key';
value = '';
@@ -396,10 +438,10 @@ module.exports = {
if (key === false) {
response.value = value.trim();
} else {
response.params[key] = value.trim();
setParam(key, value.trim());
}
} else if (value.trim()) {
response.params[value.trim().toLowerCase()] = '';
setParam(value.trim().toLowerCase(), '');
}
// handle parameter value continuations
@@ -412,6 +454,14 @@ module.exports = {
actualKey = key.substr(0, match.index);
nr = Number(match[2] || match[3]) || 0;
if (isProtoKey(actualKey)) {
// see setParam. Reading it back would yield Object.prototype, which is
// an object, so the initializer below would be skipped and the write
// that follows would throw out of a header build the caller can not catch
delete response.params[key];
return;
}
if (!response.params[actualKey] || typeof response.params[actualKey] !== 'object') {
response.params[actualKey] = {
charset: false,
@@ -538,7 +588,7 @@ module.exports = {
*/
splitMimeEncodedString: (str, maxlen) => {
const lines = [];
let curLine, match, chr, done;
let curLine, fallbackLine, match, chr, done;
// require at least 12 symbols to fit possible 4 octet UTF-8 sequences
maxlen = Math.max(maxlen || 0, 12);
@@ -551,8 +601,14 @@ module.exports = {
curLine = curLine.substr(0, match.index);
}
// Malformed input (a run of stray UTF-8 continuation bytes) has no split point
// that keeps a character sequence whole, so the loop below walks back to an
// empty line looking for one. Keep the widest chunk that at least does not cut
// a "=XX" escape in half, so the part stays a decodable encoded word.
fallbackLine = curLine.length ? curLine : str.substr(0, maxlen);
done = false;
while (!done) {
while (!done && curLine.length) {
done = true;
// check if not middle of a unicode char sequence
if ((match = str.substr(curLine.length).match(/^[=]([0-9A-F]{2})/i))) {
@@ -565,9 +621,11 @@ module.exports = {
}
}
if (curLine.length) {
lines.push(curLine);
if (!curLine.length) {
curLine = fallbackLine;
}
lines.push(curLine);
str = str.substr(curLine.length);
}
@@ -600,8 +658,11 @@ module.exports = {
// might throw if we try to encode invalid sequences, eg. partial emoji
str = encodeURIComponent(str);
} catch (_E) {
// should never run
return str.replace(/[^\x00-\x1F *'()<>@,;:\\"[\]?=\u007F-\uFFFF]+/g, '');
// an unpaired surrogate has no utf-8 representation, so run the value through a
// utf-8 roundtrip to get the same U+FFFD every other encoder here produces and
// retry. the value must never come back unencoded, it goes into a header parameter
// where a bare quote or semicolon would break it out into a parameter of its own
str = encodeURIComponent(Buffer.from(str, 'utf-8').toString('utf-8'));
}
// ensure chars that are not handled by encodeURICompent are converted as well
+287 -59
View File
@@ -7,6 +7,7 @@ const fs = require('fs');
const punycode = require('../punycode');
const { PassThrough } = require('stream');
const shared = require('../shared');
const urlModule = require('url');
const mimeFuncs = require('../mime-funcs');
const qp = require('../qp');
@@ -21,6 +22,59 @@ const LeUnix = require('./le-unix');
const FORMATTED_HEADERS = ['From', 'Sender', 'To', 'Cc', 'Bcc', 'Reply-To', 'Date', 'References'];
// RFC 5321 atext, plus the non-ascii bytes that SMTPUTF8 (RFC 6531) adds to it. A local part
// built from these, with '.' as a separator, is a dot-atom and can be emitted bare
const ATEXT = "[A-Za-z0-9!#$%&'*+\\-/=?^_`{|}~\\x80-\\uFFFF]";
const DOT_ATOM = new RegExp('^' + ATEXT + '+(?:\\.' + ATEXT + '+)*$');
// A complete quoted-string: everything between the outer quotes is either a plain char or
// a quoted-pair. Anchored, so a value that only starts and ends with a quote does not pass
const QUOTED_STRING = /^"(?:[^"\\]|\\[\s\S])*"$/;
// An address that carries no special anywhere can be emitted bare in a header, everything
// else goes into angle brackets so that the header can not be read as more addresses than
// the envelope carries
const PLAIN_ADDRESS = /^[^\s"(),:;<>@[\\\]]+@[^\s"(),:;<>@[\\\]]+$/;
// domainToASCII and domainToUnicode are WHATWG host parsers rather than plain IDNA
// mappers, so they do more than map: they cut the host at '/', '\\', '?' and '#', drop C0
// controls, and percent-decode. Handing them 'evil.example/mail.corp.example' returns the
// deliverable 'evil.example', which would turn a value the bundled codec leaves as
// unroutable garbage into mail for a domain the sender never named. None of these
// characters are legal in a domain, so keep them away from the mapper.
const URL_PARSER_UNSAFE = /[/\\?#%\x00-\x20\x7F]/;
/**
* Encodes a domain the way browsers, the WHATWG URL Standard and DNS facing resolvers do,
* which is with UTS-46 mapping applied before the Punycode step.
*
* The bundled codec is plain RFC 3492 and maps nothing, so it disagrees with every
* conformant parser on any domain holding a mapped or ignored code point. An invisible
* U+00AD in 'compa\u00ADny.com' encoded to 'xn--company-pka.com' where a validator reads
* 'company.com', which let an allow-listed domain be checked and a different one mailed.
*
* Anything the URL parser does not accept as a hostname, an address literal such as
* '[127.0.0.1]' included, comes back empty and falls through to the bundled codec, which
* leaves those as they were supplied.
*
* @param {String} domain Domain to encode, already lowercased by the caller
* @param {Boolean} toUnicode Return the U-label form instead of the A-label form
* @return {String} Encoded domain
*/
function normalizeDomain(domain, toUnicode) {
// domainToASCII and domainToUnicode landed in Node 7, the bundled codec covers Node 6
const mapper = toUnicode ? urlModule.domainToUnicode : urlModule.domainToASCII;
if (typeof mapper === 'function' && !URL_PARSER_UNSAFE.test(domain)) {
const mapped = mapper(domain);
if (mapped) {
return mapped;
}
}
return toUnicode ? punycode.toUnicode(domain) : punycode.toASCII(domain);
}
/**
* Creates a new mime tree node. Assumes 'multipart/*' as the content type
* if it is a branch, anything else counts as leaf. If rootNode is missing from
@@ -190,6 +244,14 @@ class MimeNode {
* @return {Object} Appended node object
*/
appendChild(childNode) {
// Take the node out of the tree it is in first. Leaving it there keeps it in that
// parent's childNodes, so it still streams as part of the old tree while parentNode
// already points at the new one, and anything read off the parent chain answers for
// the wrong tree.
if (childNode.parentNode && childNode.parentNode !== this) {
childNode.remove();
}
if (childNode.rootNode !== this.rootNode) {
childNode.rootNode = this.rootNode;
childNode._nodeId = ++this.rootNode.nodeCounter;
@@ -519,11 +581,10 @@ class MimeNode {
const formattedHeaders = FORMATTED_HEADERS;
if (value && typeof value === 'object' && !formattedHeaders.includes(key)) {
Object.keys(value).forEach(key => {
if (key !== 'value') {
options[key] = value[key];
}
});
// the keys come from a caller supplied header object and `options.prepared`
// below decides whether the value is emitted raw, so an own "__proto__" key
// here would turn an unfolded value into header injection
shared.copyOwnKeys(options, value, optionKey => optionKey === 'value');
value = (value.value || '').toString();
if (!value.trim()) {
return;
@@ -552,6 +613,11 @@ class MimeNode {
case 'Content-Type':
structured = mimeFuncs.parseHeaderValue(value);
// the type token decides multipart and charset below, so clean it before
// those run and not just on the way out, otherwise a control char makes
// the checks miss and the header ends up claiming a type it is not set up for
structured.value = (structured.value || '').toString().replace(/[\x00-\x1f\x7f]/g, '');
this._handleContentType(structured);
if (
@@ -568,11 +634,18 @@ class MimeNode {
// add support for non-compliant clients like QQ webmail
// we can't build the value with buildHeaderValue as the value is non standard and
// would be converted to parameter continuation encoding that we do not want
param = this._encodeWords(this.filename);
// control chars can not be quoted here: HT is a fold point that unfolding would
// turn into a space, CR/LF can not appear in a header at all and DEL is not
// qtext, so force the mime encoded word that a non-ascii filename would get anyway
param = /[\x00-\x1f\x7f]/.test(this.filename)
? mimeFuncs.encodeWord(this.filename, this._getTextEncoding(this.filename), 52)
: this._encodeWords(this.filename);
if (param !== this.filename || /[\s'"\\;:/=(),<>@[\]?]|^-/.test(param)) {
// include value in quotes if needed
param = '"' + param + '"';
// include value in quotes if needed, escaping backslashes and quotes as
// quoted-pairs exactly like buildHeaderValue does for filename=, otherwise
// a trailing backslash would escape the closing quote
param = JSON.stringify(param);
}
value += '; name=' + param;
}
@@ -595,8 +668,12 @@ class MimeNode {
if (typeof this.normalizeHeaderKey === 'function') {
const normalized = this.normalizeHeaderKey(key, value);
if (normalized && typeof normalized === 'string' && normalized.length) {
key = normalized;
// the result replaces the key on the way into the header, so it gets the same
// treatment the key it replaces already had. a line break here would end the
// header and start one of the caller's own
const cleaned = typeof normalized === 'string' ? normalized.replace(/[\x00-\x1f\x7f]/g, '') : '';
if (cleaned) {
key = cleaned;
}
}
@@ -837,26 +914,23 @@ class MimeNode {
if (envelope.from) {
list = [];
this._convertAddresses(this._parseAddresses(envelope.from), list);
this._convertAddresses(this._parseEnvelopeAddresses(envelope.from), list);
list = list.filter(address => address && address.address);
if (list.length && list[0]) {
this._envelope.from = list[0].address;
}
}
const seenRecipients = new Set();
['to', 'cc', 'bcc'].forEach(key => {
if (envelope[key]) {
this._convertAddresses(this._parseAddresses(envelope[key]), this._envelope.to);
this._convertAddresses(this._parseEnvelopeAddresses(envelope[key]), this._envelope.to, seenRecipients);
}
});
this._envelope.to = this._envelope.to.map(to => to.address).filter(address => address);
const standardFields = ['to', 'cc', 'bcc', 'from'];
Object.keys(envelope).forEach(key => {
if (!standardFields.includes(key)) {
this._envelope[key] = envelope[key];
}
});
shared.copyOwnKeys(this._envelope, envelope, key => standardFields.includes(key));
return this;
}
@@ -868,15 +942,17 @@ class MimeNode {
*/
getAddresses() {
const addresses = {};
const seenByKey = new Map();
this._headers.forEach(header => {
const key = header.key.toLowerCase();
if (['from', 'sender', 'reply-to', 'to', 'cc', 'bcc'].includes(key)) {
if (!Array.isArray(addresses[key])) {
addresses[key] = [];
seenByKey.set(key, new Set());
}
this._convertAddresses(this._parseAddresses(header.value), addresses[key]);
this._convertAddresses(this._parseAddresses(header.value), addresses[key], seenByKey.get(key));
}
});
@@ -897,6 +973,12 @@ class MimeNode {
from: false,
to: []
};
// Built once and carried across the headers. Letting _convertAddresses seed it per
// call would cost O(headers x recipients), and a message can carry many address
// headers: `headers: { to: [...] }` emits one To per entry.
const seenRecipients = new Set();
this._headers.forEach(header => {
const list = [];
if (header.key === 'From' || (!envelope.from && ['Reply-To', 'Sender'].includes(header.key))) {
@@ -905,7 +987,7 @@ class MimeNode {
envelope.from = list[0].address;
}
} else if (['To', 'Cc', 'Bcc'].includes(header.key)) {
this._convertAddresses(this._parseAddresses(header.value), envelope.to);
this._convertAddresses(this._parseAddresses(header.value), envelope.to, seenRecipients);
}
});
@@ -952,6 +1034,26 @@ class MimeNode {
/////// PRIVATE METHODS
/**
* Checks an access policy flag for this node and every node above it. The flags are set
* from the options the node was built with, and createChild only ever sees the options
* the caller passed, so a child of a closed tree starts out open. Reading the answer off
* the parent chain keeps it right whatever order the tree was assembled in.
*
* @param {String} flag Either 'disableFileAccess' or 'disableUrlAccess'
* @return {Boolean} true if this node or an ancestor closed that access
*/
_accessDisabled(flag) {
let node = this;
while (node) {
if (node[flag]) {
return true;
}
node = node.parentNode;
}
return false;
}
/**
* Detects and returns handle to a stream related with the content.
*
@@ -982,7 +1084,7 @@ class MimeNode {
}
if (content && typeof content.path === 'string' && !content.href) {
if (this.disableFileAccess) {
if (this._accessDisabled('disableFileAccess')) {
contentStream = new PassThrough();
setImmediate(() => {
const err = new Error('File access rejected for ' + content.path);
@@ -996,7 +1098,7 @@ class MimeNode {
}
if (content && typeof content.href === 'string') {
if (this.disableUrlAccess) {
if (this._accessDisabled('disableUrlAccess')) {
contentStream = new PassThrough();
setImmediate(() => {
const err = new Error('Url access rejected for ' + content.href);
@@ -1005,7 +1107,9 @@ class MimeNode {
});
return contentStream;
}
// fetch URL
// fetch URL. nmfetch refuses any scheme that is not http(s), and it decides
// that on the parsed URL. Testing the raw string here instead would reject
// forms the parser accepts, such as a leading space or a slash-less authority
return nmfetch(content.href, { headers: content.httpHeaders, tls: content.tls });
}
@@ -1030,17 +1134,81 @@ class MimeNode {
* @return {Array} An array of address objects
*/
_parseAddresses(addresses) {
return [].concat.apply(
[],
[].concat(addresses).map(address => {
// Collected into one list as we go. concat.apply spreads the entries into arguments
// and throws a RangeError once a recipient array is long enough to pass the
// argument limit, which a large Bcc list reaches on its own.
const flattened = [];
[].concat(addresses).forEach(address => {
if (address && address.address) {
address.address = this._normalizeAddress(address.address);
address.name = address.name || '';
return [address];
const normalized = this._normalizeAddress(address.address);
if (normalized === address.address && typeof address.name === 'string') {
// there is nothing to rewrite, so there is nothing to keep off the original
flattened.push(address);
return;
}
return addressparser(address);
})
);
// rewriting would land on the object the caller passed in and might
// still hold a reference to, so rewrite a copy of it instead. An own
// "__proto__" key would make the copy inherit from caller data, and
// _convertAddresses reads `group` off it straight into the envelope
const copy = shared.copyOwnKeys({}, address);
copy.address = normalized;
copy.name = address.name || '';
flattened.push(copy);
return;
}
const parsed = this._normalizeParsedAddresses(addressparser(address));
for (let i = 0; i < parsed.length; i++) {
flattened.push(parsed[i]);
}
});
return flattened;
}
/**
* Normalizes the addresses of a freshly parsed address list, groups included.
*
* Everything this method returns carries a normalized address, whether it arrived as an
* object or was parsed out of a header value. Without this the two shapes disagree, and
* a consumer reading the parsed form back is handed the ambiguous
* 'user@evil.com@good.com' that the header and the envelope no longer carry.
*
* @param {Array} parsed An array of address objects, as returned by addressparser
* @return {Array} The same array, with every address normalized
*/
_normalizeParsedAddresses(parsed) {
// addressparser builds these objects, so no caller holds a reference to rewrite around
parsed.forEach(entry => {
if (entry.address) {
entry.address = this._normalizeAddress(entry.address);
} else if (entry.group) {
this._normalizeParsedAddresses(entry.group);
}
});
return parsed;
}
/**
* Parses the addresses of an explicitly set envelope.
*
* An envelope value is an addr-spec and never a display name, so a bare local username
* such as 'root' is the address here. Header parsing has to read the same value as a
* display name, as a value with no '@' in it can not be an addr-spec in a header.
*
* @param {Mixed} addresses Addresses to be parsed
* @return {Array} An array of address objects
*/
_parseEnvelopeAddresses(addresses) {
return this._parseAddresses(addresses).map(entry => {
if (entry.address || entry.group || !entry.name || /[\s@]/.test(entry.name)) {
return entry;
}
return { address: this._normalizeAddress(entry.name), name: '' };
});
}
/**
@@ -1054,6 +1222,9 @@ class MimeNode {
.toString()
// no newlines in keys
.replace(/\r?\n|\r/g, ' ')
// a field name is printable ascii without the colon, so a control char or DEL
// can only be dropped, there is no quoting construct around a field name
.replace(/[\x00-\x1f\x7f]/g, '')
.trim()
.toLowerCase()
// use uppercase words, except MIME
@@ -1114,7 +1285,13 @@ class MimeNode {
case 'Message-ID':
case 'In-Reply-To':
case 'Content-Id':
value = (value || '').toString().replace(/\r?\n|\r/g, ' ');
// a msg-id is structured, so an encoded word inside the angle brackets would
// be read as literal text. drop the characters that can not appear in a header
// at all, but leave HT alone, it separates the ids of a multi id value
value = (value || '')
.toString()
.replace(/\r?\n|\r/g, ' ')
.replace(/[\x00-\x08\x0b\x0c\x0e-\x1f\x7f]/g, '');
if (value.charAt(0) !== '<') {
value = '<' + value;
@@ -1134,6 +1311,7 @@ class MimeNode {
elm = (elm || '')
.toString()
.replace(/\r?\n|\r/g, ' ')
.replace(/[\x00-\x08\x0b\x0c\x0e-\x1f\x7f]/g, '')
.trim();
return elm.replace(/<[^>]*>/g, str => str.replace(/\s/g, '')).split(/\s+/);
})
@@ -1156,7 +1334,7 @@ class MimeNode {
}
value = (value || '').toString().replace(/\r?\n|\r/g, ' ');
return this._encodeWords(value);
return this._encodeHeaderText(value);
case 'Content-Type':
case 'Content-Disposition':
@@ -1165,8 +1343,7 @@ class MimeNode {
default:
value = (value || '').toString().replace(/\r?\n|\r/g, ' ');
// encodeWords only encodes if needed, otherwise the original string is returned
return this._encodeWords(value);
return this._encodeHeaderText(value);
}
}
@@ -1177,26 +1354,44 @@ class MimeNode {
* @param {Array} [uniqueList] An array to be populated with addresses
* @return {String} address string
*/
_convertAddresses(addresses, uniqueList) {
_convertAddresses(addresses, uniqueList, seenAddresses) {
const values = [];
uniqueList = uniqueList || [];
// Membership is checked once per address, so scanning uniqueList itself would make
// a recipient list cost O(n^2). Groups recurse with the same set so that a nested
// group still dedupes against the addresses collected around it, and a caller that
// passes a partly filled list (To, then Cc, then Bcc) keeps deduping across headers.
if (!seenAddresses) {
seenAddresses = new Set();
for (let i = 0; i < uniqueList.length; i++) {
seenAddresses.add(uniqueList[i].address);
}
}
[].concat(addresses || []).forEach(address => {
if (address.address) {
address.address = this._normalizeAddress(address.address);
if (!address.name) {
values.push(address.address.indexOf(' ') >= 0 ? `<${address.address}>` : `${address.address}`);
// an address that carries a special, be it a quoted local part or a domain
// that could not be normalized, is only unambiguous inside angle brackets.
// Without them a ',' or a ';' anywhere in it reads as a recipient separator
// and the header would list more recipients than the envelope carries
values.push(PLAIN_ADDRESS.test(address.address) ? address.address : `<${address.address}>`);
} else {
values.push(`${this._encodeAddressName(address.name)} <${address.address}>`);
}
if (!uniqueList.some(a => a.address === address.address)) {
if (!seenAddresses.has(address.address)) {
seenAddresses.add(address.address);
uniqueList.push(address);
}
} else if (address.group) {
const groupListAddresses = (address.group.length ? this._convertAddresses(address.group, uniqueList) : '').trim();
const groupListAddresses = (
address.group.length ? this._convertAddresses(address.group, uniqueList, seenAddresses) : ''
).trim();
values.push(`${this._encodeAddressName(address.name)}:${groupListAddresses};`);
}
});
@@ -1213,45 +1408,63 @@ class MimeNode {
_normalizeAddress(address) {
address = (address || '')
.toString()
.replace(/[\x00-\x1F<>]+/g, ' ') // remove unallowed characters
.replace(/[\x00-\x1F\x7F<>]+/g, ' ') // remove unallowed characters
.trim();
const lastAt = address.lastIndexOf('@');
if (lastAt < 0) {
// Bare username
if (!address) {
// callers use an empty value to detect a missing address
return address;
}
let user = address.substr(0, lastAt);
const lastAt = address.lastIndexOf('@');
if (lastAt < 0) {
// Bare username, there is no domain to split off
return this._normalizeLocalPart(address);
}
const user = address.substr(0, lastAt);
const domain = address.substr(lastAt + 1);
// Usernames are not touched and are kept as is even if these include unicode.
// Unicode in the local part is kept as is, see _normalizeLocalPart for the rest of it.
// A domain has no quoting construct to fall back on, so whatever is not a valid domain
// is kept as supplied and it is _convertAddresses that keeps such an address unambiguous.
// Domains are punycoded when the local part is ASCII ('safe@jõgeva.ee' -> 'safe@xn--jgeva-dua.ee').
// When the local part contains non-ASCII bytes the address already requires SMTPUTF8,
// so the domain is kept (or decoded back) as UTF-8 for symmetry on both sides of '@'.
let encodedDomain = domain;
// A non-ASCII local part already requires SMTPUTF8, so the domain stays UTF-8 for
// symmetry on both sides of the '@' rather than being encoded to an A-label
const smtputf8 = /[\x80-\uFFFF]/.test(user);
try {
if (/[\x80-\uFFFF]/.test(user)) {
encodedDomain = punycode.toUnicode(domain.toLowerCase());
} else {
encodedDomain = punycode.toASCII(domain.toLowerCase());
}
encodedDomain = normalizeDomain(domain.toLowerCase(), smtputf8);
} catch (_err) {
// keep domain as supplied
}
if (user.indexOf(' ') >= 0) {
if (user.charAt(0) !== '"') {
user = '"' + user;
}
if (user.substr(-1) !== '"') {
user = user + '"';
}
return `${this._normalizeLocalPart(user)}@${encodedDomain}`;
}
return `${user}@${encodedDomain}`;
/**
* Normalizes the local part of an address into a form that can be emitted as is.
*
* A local part is either a dot-atom or a quoted-string, anything else is not a valid
* addr-spec. The quotes of a quoted local part get lost along the way, and a bare
* 'user@evil.com@good.com' leaves it to the receiver which '@' splits the domain off,
* while the split here is always at the last one. So whatever is not already one of
* the two valid forms goes back out as a quoted-string.
*
* @param {String} user Local part of an address
* @return {String} Local part as a dot-atom or as a quoted-string
*/
_normalizeLocalPart(user) {
if (DOT_ATOM.test(user) || QUOTED_STRING.test(user)) {
return user;
}
return mimeFuncs.quoteString(user);
}
/**
@@ -1263,7 +1476,7 @@ class MimeNode {
_encodeAddressName(name) {
if (!/^[\w ]*$/.test(name)) {
if (/^[\x20-\x7e]*$/.test(name)) {
return '"' + name.replace(/([\\"])/g, '\\$1') + '"';
return mimeFuncs.quoteString(name);
} else {
return mimeFuncs.encodeWord(name, this._getTextEncoding(name), 52);
}
@@ -1271,6 +1484,21 @@ class MimeNode {
return name;
}
/**
* Encodes an unstructured header value. Such a value can only carry VCHAR and WSP, so a
* control char or DEL has to be forced into the mime encoded word that a non-ascii value
* would get anyway. HT stays as it is, it is valid folding whitespace here.
*
* @param {String} value Header value to encode
* @returns {String} Mime word encoded string if needed
*/
_encodeHeaderText(value) {
return /[\x00-\x08\x0b\x0c\x0e-\x1f\x7f]/.test(value)
? mimeFuncs.encodeWord(value, this._getTextEncoding(value), 52)
: // encodeWords only encodes if needed, otherwise the original string is returned
this._encodeWords(value);
}
/**
* If needed, mime encodes the name part
*
+5 -2
View File
@@ -62,14 +62,17 @@ class SendmailTransport {
// Sendmail strips this header line by itself
mail.message.keepBcc = true;
const envelope = mail.data.envelope || mail.message.getEnvelope();
const envelope = mail.message.getEnvelope();
const messageId = mail.message.messageId();
let returned;
const hasInvalidAddresses = []
.concat(envelope.from || [])
.concat(envelope.to || [])
.some(addr => /^-/.test(addr));
// a local part is either a dot-atom or a quoted-string, so a leading dash sits at
// offset 0 or, behind the opening quote, at offset 1. Only the first shape is read
// as an option by sendmail, but both are the address this guard keeps out of argv
.some(addr => /^"?-/.test(addr));
if (hasInvalidAddresses) {
const err = new Error('Can not send mail. Invalid envelope addresses.');
err.code = errors.ESENDMAIL;
+4 -3
View File
@@ -66,7 +66,7 @@ class SESTransport extends EventEmitter {
fromHeader = mimeNode._convertAddresses(mimeNode._parseAddresses(fromHeader.value));
}
const envelope = mail.data.envelope || mail.message.getEnvelope();
const envelope = mail.message.getEnvelope();
const messageId = mail.message.messageId();
const recipients = [].concat(envelope.to || []);
@@ -130,7 +130,8 @@ class SESTransport extends EventEmitter {
return callback(err);
}
const sesMessage = Object.assign(
// mail.data.ses is caller supplied message data, so copy its own keys only
const sesMessage = shared.copyOwnKeys(
{
Content: {
Raw: {
@@ -143,7 +144,7 @@ class SESTransport extends EventEmitter {
ToAddresses: envelope.to
}
},
mail.data.ses || {}
mail.data.ses
);
this.getRegion((err, region) => {
+29 -13
View File
@@ -7,10 +7,15 @@ const util = require('util');
const fs = require('fs');
const nmfetch = require('../fetch');
const errors = require('../errors');
const objects = require('./objects');
const dns = require('dns');
const net = require('net');
const os = require('os');
// re-exported for the callers that already depend on this module, see ./objects
const isProtoKey = (module.exports.isProtoKey = objects.isProtoKey);
module.exports.copyOwnKeys = objects.copyOwnKeys;
const DNS_TTL = 5 * 60 * 1000;
const CACHE_CLEANUP_INTERVAL = 30 * 1000; // Minimum 30 seconds between cleanups
const MAX_CACHE_SIZE = 1000; // Maximum number of entries in cache
@@ -355,7 +360,9 @@ module.exports.parseConnectionUrl = str => {
return;
}
if (!(lKey in obj)) {
// `in` already keeps "__proto__" out, but only as a side effect of it being an
// Object.prototype member. Say it, so the protection survives a change to the check
if (!isProtoKey(lKey) && !(lKey in obj)) {
obj[lKey] = value;
}
});
@@ -470,7 +477,7 @@ module.exports.parseDataURI = uri => {
// Ensure there's a key before the '='
const key = entry.substring(0, sepPos).trim();
const value = entry.substring(sepPos + 1).trim();
if (key) {
if (key && !isProtoKey(key)) {
params[key] = value;
}
}
@@ -561,19 +568,24 @@ function resolveContentValue(data, key, options, callback) {
}
callback(null, value);
});
} else if (/^https?:\/\//i.test(content.path || content.href)) {
if (options.disableUrlAccess) {
return setImmediate(() => {
const err = new Error('Url access rejected for ' + (content.path || content.href));
err.code = errors.EURLACCESS;
callback(err);
});
}
return resolveStream(nmfetch(content.path || content.href, { headers: content.httpHeaders, tls: content.tls }), callback);
} else if (/^data:/i.test(content.path || content.href)) {
const parsedDataUri = module.exports.parseDataURI(content.path || content.href);
return callback(null, parsedDataUri && parsedDataUri.data ? parsedDataUri.data : Buffer.alloc(0));
} else if (content.href || /^https?:\/\//i.test(content.path)) {
// An href is always a URL, and so is a path that looks like one. Let nmfetch
// decide whether it is fetchable, it validates the parsed URL. Testing the raw
// string here instead would let a file: href fall through to the "return as is"
// default below and travel on inside the resolved message.
const url = content.href || content.path;
if (options.disableUrlAccess) {
return setImmediate(() => {
const err = new Error('Url access rejected for ' + url);
err.code = errors.EURLACCESS;
callback(err);
});
}
return resolveStream(nmfetch(url, { headers: content.httpHeaders, tls: content.tls }), callback);
} else if (content.path) {
if (options.disableFileAccess) {
return setImmediate(() => {
@@ -603,10 +615,14 @@ module.exports.assign = function (/* target, ... sources */) {
args.forEach(source => {
Object.keys(source || {}).forEach(key => {
if (isProtoKey(key)) {
return;
}
if (['tls', 'auth'].includes(key) && source[key] && typeof source[key] === 'object') {
// tls and auth are special keys that need to be enumerated separately
// other objects are passed as is
target[key] = Object.assign(target[key] || {}, source[key]);
// other objects are passed as is. Enumerating is a copy of user supplied
// keys just like the loop above, so it gets the same treatment
target[key] = module.exports.copyOwnKeys(target[key] || {}, source[key]);
} else {
target[key] = source[key];
}
+42
View File
@@ -0,0 +1,42 @@
'use strict';
// Safe copying of objects whose keys come from the caller.
//
// This lives in its own leaf module, like ./url.js, so that every layer can reach it.
// lib/shared/index.js requires lib/fetch, so lib/fetch can not require lib/shared back,
// and lib/mime-funcs is a leaf that would otherwise pull in dns/net/os/fs for a string
// comparison. lib/shared/index.js re-exports both functions for the callers that already
// depend on it.
/**
* Detects a key that can not be copied onto a plain object with `target[key] = value`.
*
* "__proto__" is the only one: assigning it runs the inherited setter and replaces the
* prototype of the target instead of adding a property to it, so a caller can smuggle
* values past validation that only inspects own keys. JSON.parse produces such a key
* where an object literal can not. "constructor" and "prototype" have no such setter and
* become ordinary own properties, so dropping them would only discard legitimate values.
*
* @param {String} key Key to check
* @returns {Boolean} true if the key must not be copied
*/
module.exports.isProtoKey = key => key === '__proto__';
/**
* Copies own enumerable keys from a source object to a target object. Every copy that
* walks the keys of user supplied data goes through here, see isProtoKey.
*
* @param {Object} target Object to copy the keys to
* @param {Object} source Object to copy the keys from
* @param {Function} [skip] Optional predicate, return true to leave a key out
* @returns {Object} The target object
*/
module.exports.copyOwnKeys = (target, source, skip) => {
Object.keys(source || {}).forEach(key => {
if (module.exports.isProtoKey(key) || (skip && skip(key))) {
return;
}
target[key] = source[key];
});
return target;
};
+1 -1
View File
@@ -42,7 +42,7 @@ class StreamTransport {
// We probably need this in the output
mail.message.keepBcc = true;
const envelope = mail.data.envelope || mail.message.getEnvelope();
const envelope = mail.message.getEnvelope();
const messageId = mail.message.messageId();
const recipients = [].concat(envelope.to || []);
+14
View File
@@ -577,6 +577,20 @@
"port": 587
},
"TurboSMTP": {
"description": "TurboSMTP",
"host": "pro.turbo-smtp.com",
"port": 465,
"secure": true
},
"TurboSMTP-EU": {
"description": "TurboSMTP (EU region)",
"host": "pro.eu.turbo-smtp.com",
"port": 465,
"secure": true
},
"Tutanota": {
"description": "Tutanota (Tuta Mail)",
"domains": ["tutanota.com", "tuta.com", "tutanota.de", "tuta.io"],
+8 -8
View File
@@ -1,6 +1,6 @@
{
"name": "nodemailer",
"version": "9.0.3",
"version": "9.1.1",
"description": "Easy as cake e-mail sending from your Node.js applications",
"main": "lib/nodemailer.js",
"scripts": {
@@ -27,19 +27,19 @@
},
"homepage": "https://nodemailer.com/",
"devDependencies": {
"@aws-sdk/client-sesv2": "3.1068.0",
"@aws-sdk/client-sesv2": "3.1121.0",
"bunyan": "1.8.15",
"c8": "11.0.0",
"eslint": "10.5.0",
"c8": "12.0.0",
"eslint": "10.9.1",
"eslint-config-prettier": "10.1.8",
"globals": "17.6.0",
"globals": "17.11.0",
"libbase64": "1.3.0",
"libmime": "5.3.8",
"libmime": "5.4.2",
"libqp": "2.1.1",
"prettier": "3.8.4",
"prettier": "3.9.6",
"proxy": "1.0.2",
"proxy-test-server": "1.0.0",
"smtp-server": "3.19.0"
"smtp-server": "3.19.4"
},
"engines": {
"node": ">=6.0.0"