Parse of email with brackets on from header recover an empty field.

[reinaldocoelho]

Describe the bug
Using MimeMessage.Load(stream) on EML with FROM having format "[Matt] [email protected]", recover an blank From field to Message object.

Platform (please complete the following information):

• OS: Windows
• .NET Runtime: DotNet4.6.1, NetCore 2.1
• .NET Framework: .Net Core, .NET 4.6.1
• MimeKit Version: 2.4.1

To Reproduce
Steps to reproduce the behavior:

1. Create an EML with code below.
2. Simple parse the EML with Mimekit lib.
3. Verify "From" field on "Message" object.
4. Expected to see the email but has blank text.

Expected behavior
Expected to see the email on "From" field but has blank text.

Screenshots

Additional context

1. EML used to test Bug:

`
To: [email protected]
Content-type: text/html; charset=UTF-8
Content-Transfer-Encoding: base64
From: [Invalid Sender] [email protected]
Message-Id: [email protected]
Date: Thu, 9 Jan 2020 11:13:09 +0900 (JST)
Subject: =?UTF-8?B?VGVzdCBzdWJqZWN0IHRvIHZhbGlkYXRlIHByb2JsZW0=?=

PCFET0NUWVBFIEhUTUwgUFVCTElDICItLy9XM0MvL0RURCBIVE1MIDQuMDEgVHJhbnNpdGlvbmFsLy9FTiI+CjxodG1sPgo8aGVhZD4KICA8dGl0bGU+TkIxPC90aXRsZT4KPC9oZWFkPgo8Ym9keT4KCTxoMT4gVGl0bGUgdGVzdCA8L2gxPgoJPHA+VGhpcyBpcyBvbmx5IGEgc2ltcGxlIEhUTUwgYm9keSB0ZXN0LjwvcD4KPC9ib2R5Pgo8L2h0bWw+Cg==
`

[jstedfast]

If I am understanding rfc5322 correctly, the syntax is invalid:

   atext           =   ALPHA / DIGIT /    ; Printable US-ASCII
                       "!" / "#" /        ;  characters not including
                       "$" / "%" /        ;  specials.  Used for atoms.
                       "&" / "'" /
                       "*" / "+" /
                       "-" / "/" /
                       "=" / "?" /
                       "^" / "_" /
                       "`" / "{" /
                       "|" / "}" /
                       "~"

   atom            =   [CFWS] 1*atext [CFWS]

   specials        =   "(" / ")" /        ; Special characters that do
                       "<" / ">" /        ;  not appear in atext
                       "[" / "]" /
                       ":" / ";" /
                       "@" / "\" /
                       "," / "." /
                       DQUOTE

   mailbox         =   name-addr / addr-spec

   name-addr       =   [display-name] angle-addr

   display-name    =   phrase

   word            =   atom / quoted-string

   phrase          =   1*word / obs-phrase

   obs-phrase      =   word *(word / "." / CFWS)

We can see from the syntax definitions above that a phrase (the display-name component of a name-addr-style mailbox address) must either be an atom or a quoted-string and an atom clearly does not allow [ or ].

That said, I don't think there would be any negative consequences for allowing square bracket characters in the name in MimeKit's parser...

The only special meaning I can think of that make use of square brackets are domain-literal tokens which would not appear in any areas that could easily be confused with a "phrase atom".

[jstedfast]

Thanks for the bug report!

[reinaldocoelho]

But in this case, the "Name" in brackets may be invalid, but the address is normal.
If the "From" object has "Name" and "Address", the library should not display the "Name" field, but may display the "Address" field as a result.
Because if we look at the EML, the "From" exists and is understandable.
Is it the best behavior for the library during parse to remove all information because of an invalid character?
Wouldn't it be better to return information that is at least valid?

[jstedfast]

That's the goal. Most of the specials characters were handled when they were encountered in the display-name except square brackets which is what you hit. Now they are handled, too.

[reinaldocoelho]

Cool, will you release a version to this?
The last version on Nuget is the same 2.4.1.

Thank you :-)

[jstedfast]

Yea, I'm hoping to have some free time this weekend (or maybe tonight) to make a release.

[thomas-westergard-scandinn-se]

works: var displayName = "Tom Doe [Corp One]";
works: var displayName = "Tom Doe Corp One";
works: var displayName = "Tom Doe Cörp Öne";
does NOT work: var displayName = "Tom Doe [Cörp Öne]";

var mimeMessage = new MimeMessage();

mimeMessage.From.Add(new MailboxAddress(displayName, emailAddress));

No exception when sending.

@jstedfast

[jstedfast]

@thomas-westergard-scandinn-se This bug is about parsing, not sending.