Manually decoding an encoded email subject. 2022-07-12. UTF-8.

Spam email's subject field:
=?GB2312?B?tPq/qreixrExMzUxMDg3NjUyMw==?=

Subject is Encoded-Word -encoded.
(https://en.wikipedia.org/wiki/MIME#Encoded-Word)
Form: =?  charset ? encoding ? encoded text ?=
Splitting into parts:
Charset:	GB2312
Encoding:	B
Encoded text:	tPq/qreixrExMzUxMDg3NjUyMw==
An encoding of 'Q' is similar to quoted-printable,
while an encoding of 'B' is base-64.

Decoding the base64, using the command line
$ base64 --decode | xxd
tPq/qreixrExMzUxMDg3NjUyMw==
^D
00000000: b4fa bfaa b7a2 c6b1 3133 3531 3038 3736  ........13510876
00000010: 3532 33                                  523

GB2312 (https://en.wikipedia.org/wiki/GB_2312), or EUC-CN,
is a variable-character, but at most two-byte, encoding:
a byte less than 0x80 is the actual ASCII character.
The actual Chinese-encoding characters are all two bytes
in the range 0xA1 to 0xFE.
We use the charts at
https://en.wikipedia.org/wiki/GB_2312#Code_charts
to manually decode this.

Character data, hexadecimal:
B4FA BFAA B7A2 C6B1 31 33 35 31 30 38 37 36 35 32 33.
The single bytes that start with a '3' are all ASCII numbers:
they make up the string "13510876523", presumably a QQ or WeChat number.

The other four characters are
B4FA: row 20, column 90: 代 U+4EE3
BFAA: row 31, column 10: 开 U+5F00
B7A2: row 23, column 02: 发 U+53D1
C6B1: row 38, column 17: 票 U+7968
(\u4EE3\u5F00\u53D1\u7968)
代开发票: "dài kāi fāpiào", 'invoice on behalf of'.

The full subject line therefore translates to
"Invoice on behalf of 13510876523".
It's a scam, they send a malware-ridden file that appears
to be the invoice.