Mail::Body object returned from Mail::Message#body does not inherit message's charset

[sylph01]

It seems like the returned Mail::Body instance from Mail::Message#body always uses the default charset US-ASCII regardless of the Content-Type header in the original message.

Is this behavior intended? If not, I am ready to write up a patch for this. (I might need some help on testing)

[clinejj]

I also experienced this issue with ActionMailbox on Rails (see rails/rails#43038)

Steps to reproduce

Given the email contents below, I would expect the plain-text portion of the email to be decoded using UTF-8 because of the header, but instead Mail uses ASCII-US. We are using the default configuration for ActionMailbox, which according to the guide should set UTF-8.

Email Contents (sent from Gmail)

MIME-Version: 1.0
Date: Tue, 17 Aug 2021 11:32:25 -0400
References: <redacted>
In-Reply-To: <redacted>
Message-ID: <redacted>
Subject: Re: Subject
From: <redacted>
To: <redacted>
Content-Type: multipart/alternative; boundary="00000000000056fcd505c9c30680"

--00000000000056fcd505c9c30680
Content-Type: text/plain; charset="UTF-8"
Content-Transfer-Encoding: quoted-printable

Teamwork =F0=9F=99=8C=F0=9F=8F=BB

-John


On Tue, Aug 17, 2021 at 11:29 AM Redacted <redacted> wrote:

> -- Reply ABOVE THIS LINE --
>

--00000000000056fcd505c9c30680
Content-Type: text/html; charset="UTF-8"
Content-Transfer-Encoding: quoted-printable

<div dir=3D"ltr">Teamwork =F0=9F=99=8C=F0=9F=8F=BB</div>

--00000000000056fcd505c9c30680--

Expected behavior

Mail should report the charset as UTF-8 and decode as such, given the header for the first plain-text body part. Note the charset in the first block here:

irb(main):025:0> email.parts.first
=> #<Mail::Part:52620, Multipart: false, Headers: <Content-Type: text/plain; charset="UTF-8">, <Content-Transfer-Encoding: quoted-printable>>
irb(main):026:0> email.parts.first.body
=>
#<Mail::Body:0x000055b221450370
 @ascii_only=true,
 @boundary=nil,
 @charset="US-ASCII",
 @encoding="quoted-printable",
 @epilogue=nil,
 @part_sort_order=["text/plain", "text/enriched", "text/html", "multipart/alternative"],
 @parts=[],
 @preamble=nil,
 @raw_source=
  "Teamwork =F0=9F=99=8C=F0=9F=8F=BB\r\n\r\n-John\r\n\r\n\r\nOn Tue, Aug 17, 2021 at 11:29 AM Redacted <\r\nredacted> wrote:\r\n\r\n> -- Reply ABOVE THIS LINE --\r\n>">
irb(main):027:0> email.parts.first.headers
=> {}
irb(main):028:0> email.parts.first.header
=>
#<Mail::Header:0x000055b22121b168
 @charset="UTF-8",
 @fields=
  [#<Mail::Field 0x19b40 @name="Content-Type" @unparsed_value="text/plain; charset=\"UTF-8\"" @charset="UTF-8" @field_order_id=23 @field=#<Mail::ContentTypeField:0x000055b221219ca0 @charset="UTF-8", @main_type="text", @sub_type=nil, @parameters=nil, @name="Content-Type", @length=nil, @element=#<Mail::ContentTypeElement:0x000055b221219a20 @main_type="text", @sub_type="plain", @parameters=[{"charset"=>"UTF-8"}]>, @value="text/plain; charset=\"UTF-8\"">>,
   #<Mail::Field 0x19b68 @name="Content-Transfer-Encoding" @unparsed_value="quoted-printable" @charset="UTF-8" @field_order_id=24 @field=#<Mail::ContentTransferEncodingField:0x000055b2211f1188 @charset="UTF-8", @name="Content-Transfer-Encoding", @length=nil, @element=#<Mail::ContentTransferEncodingElement:0x000055b2211f1048 @encoding="quoted-printable">, @value="quoted-printable">>],
 @raw_source="Content-Type: text/plain; charset=\"UTF-8\"\r\nContent-Transfer-Encoding: quoted-printable">

Actual behavior

Mail uses US-ASCII for the body and UTF-8 for the header.

System configuration

Rails version:
6.1.4
Ruby version:
3.0.2

[diesl]

The same error happened to me this week. It would be nice if this could be fixed.

[kwent]

Same here !

[lylo]

Is there a known workaround?

[jens]

Is there a known workaround?

What I do is just call the decoded method directly on the mail or part object.