Skip to content

Instantly share code, notes, and snippets.

@lionel-rowe

lionel-rowe/outlook-mixed-encoding-subject-bug.eml

Last active February 28, 2022 18:51
Show Gist options
  • Star 0 You must be signed in to star a gist
  • Fork 0 You must be signed in to fork a gist
  • Save lionel-rowe/a2df1c39f5549798bc0c0f99ba09acdb to your computer and use it in GitHub Desktop.
Save lionel-rowe/a2df1c39f5549798bc0c0f99ba09acdb to your computer and use it in GitHub Desktop.
Download ZIP
Raw
outlook-mixed-encoding-subject-bug.eml
Date: Thu, 01 Jan 1970 00:00:00 +0000 (UTC)
From: ABC <abc@example.com>
To: XYZ <xyz@example.com>
Subject:
=?UTF-8?Q?=E7=BB=9F=E4=B8=80=E7=A0=81?=
=?UTF-8?Q?=E5=8F=98=E6=88=90=E4=B9=B1=E7=A0=81?=
=?us-ascii?Q?_removing_this_part_fixes_the_mojibake?=
MIME-Version: 1.0
Content-Type: text/plain;
email body
@lionel-rowe
Copy link
Author

lionel-rowe commented Feb 28, 2022 •
edited

Explanation of the corrupted "g; d8" subject line:

Due to assumed us-ascii encoding, Outlook thinks it can safely ignore the most-significant bit of each byte. The first 6 bytes, which should be e7 bb 9f e4 b8 80, are thus interpreted as 67 3b 1f 64 38 00 instead. 00, the null byte, is interpreted as terminating the string, so the rest gets truncated. Meanwhile, 67 3b 1f 64 38 correspond to the characters g ; \x1f d 8, of which \x1f is a control character that gets rendered as a space.

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment