From 4288c67c07ed2659db890ea63fdd91ff610e34a9 Mon Sep 17 00:00:00 2001
From: amertak <kadlec.david@gmail.com>
Date: Tue, 23 Mar 2021 19:24:35 +0100
Subject: [PATCH 1/3] fix: Support other encoding than UTF-8

---
 index.js | 28 +++++++++++++++++++++++++++-
 1 file changed, 27 insertions(+), 1 deletion(-)

diff --git a/index.js b/index.js
index 5f0f5fe..8efbcee 100644
--- a/index.js
+++ b/index.js
@@ -2,6 +2,8 @@
 
 var AWS = require('aws-sdk');
 
+/*global TextDecoder */
+
 console.log("AWS Lambda SES Forwarder // @arithmetric // Version 5.0.0");
 
 // Configure the S3 bucket and key prefix for stored raw emails, and the
@@ -195,7 +197,28 @@ exports.fetchMessage = function(data) {
           return reject(
             new Error("Error: Failed to load message body from S3."));
         }
-        data.emailData = result.Body.toString();
+
+        const naivelyReadBody = result.Body.toString();
+        const encodingMatches = naivelyReadBody.match(/text\/plain;\s*charset=([^;]+);/);
+        if (encodingMatches) {
+          try {
+            const encoding = encodingMatches[1]
+            const textDecoder = new TextDecoder(encoding);
+
+            data.emailData = textDecoder.decode(result.Body)
+          } catch (err) {
+            data.log({
+              level: "error",
+              message: "Could not decode body with desired encoding, fallback for naively read body.",
+              error: err,
+              stack: err.stack
+            });
+            data.emailData = naivelyReadBody;
+          }
+        } else {
+          data.emailData = naivelyReadBody;
+        }
+
         return resolve(data);
       });
     });
@@ -274,6 +297,9 @@ exports.processMessage = function(data) {
   // Remove Message-ID header.
   header = header.replace(/^message-id:[\t ]?(.*)\r?\n/mgi, '');
 
+  // JS only supports UTF-8 kinda, we also converted the text of the email into UTF8 in fetchMessage function
+  header = header.replace(/text\/plain;\s*charset=([^;]+);/gi, 'text/plain; charset=UTF-8;')
+
   // Remove all DKIM-Signature headers to prevent triggering an
   // "InvalidParameterValue: Duplicate header 'DKIM-Signature'" error.
   // These signatures will likely be invalid anyways, since the From

From 75abd79bf131ee5bd527d726bf78005677fb5c86 Mon Sep 17 00:00:00 2001
From: amertak <kadlec.david@gmail.com>
Date: Thu, 25 Mar 2021 17:23:02 +0100
Subject: [PATCH 2/3] chore: handle body/header text/any

---
 index.js | 3 ++-
 1 file changed, 2 insertions(+), 1 deletion(-)

diff --git a/index.js b/index.js
index 8efbcee..bc59b73 100644
--- a/index.js
+++ b/index.js
@@ -298,7 +298,8 @@ exports.processMessage = function(data) {
   header = header.replace(/^message-id:[\t ]?(.*)\r?\n/mgi, '');
 
   // JS only supports UTF-8 kinda, we also converted the text of the email into UTF8 in fetchMessage function
-  header = header.replace(/text\/plain;\s*charset=([^;]+);/gi, 'text/plain; charset=UTF-8;')
+  header = header.replace(/content-type:\s*text\/([^;]+);\s*charset=([^;\s]+)/gi, 'Content-Type: text/$1; charset=UTF-8')
+  body = body.replace(/content-type:\s*text\/([^;]+);\s*charset=([^;\s]+)/gi, 'Content-Type: text/$1; charset=UTF-8')
 
   // Remove all DKIM-Signature headers to prevent triggering an
   // "InvalidParameterValue: Duplicate header 'DKIM-Signature'" error.

From d0e34404fed062f0c403ec587ce9804c4a242330 Mon Sep 17 00:00:00 2001
From: amertak <kadlec.david@gmail.com>
Date: Thu, 25 Mar 2021 17:30:41 +0100
Subject: [PATCH 3/3] refactor: reuse regex and replacement text

---
 index.js | 7 +++++--
 1 file changed, 5 insertions(+), 2 deletions(-)

diff --git a/index.js b/index.js
index bc59b73..e169dee 100644
--- a/index.js
+++ b/index.js
@@ -297,9 +297,12 @@ exports.processMessage = function(data) {
   // Remove Message-ID header.
   header = header.replace(/^message-id:[\t ]?(.*)\r?\n/mgi, '');
 
+  const charsetRegex = /content-type:\s*text\/([^;]+);\s*charset=([^;\s]+)/gi
+  const replacementText = 'Content-Type: text/$1; charset=UTF-8'
+
   // JS only supports UTF-8 kinda, we also converted the text of the email into UTF8 in fetchMessage function
-  header = header.replace(/content-type:\s*text\/([^;]+);\s*charset=([^;\s]+)/gi, 'Content-Type: text/$1; charset=UTF-8')
-  body = body.replace(/content-type:\s*text\/([^;]+);\s*charset=([^;\s]+)/gi, 'Content-Type: text/$1; charset=UTF-8')
+  header = header.replace(charsetRegex, replacementText)
+  body = body.replace(charsetRegex, replacementText)
 
   // Remove all DKIM-Signature headers to prevent triggering an
   // "InvalidParameterValue: Duplicate header 'DKIM-Signature'" error.