From 3b9f6cd4e5173e974f2410397e7bae11e29f1b4d Mon Sep 17 00:00:00 2001
From: =?UTF-8?q?La=CC=81szlo=CC=81=20Ka=CC=81rolyi?= <laszlo@karolyi.hu>
Date: Thu, 14 May 2015 22:51:48 +0200
Subject: [PATCH] Fixing input charset issues by autodetecting

---
 .gitignore                             |  3 +++
 README.md                              |  7 ++++++-
 email-responder.py                     | 10 +++++++---
 email-responder.sh                     |  9 +++++++++
 requirements.txt                       |  1 +
 init-schema.sh => tools/init-schema.sh |  0
 tools/install.sh                       | 10 ++++++++++
 7 files changed, 36 insertions(+), 4 deletions(-)
 mode change 100755 => 100644 email-responder.py
 create mode 100755 email-responder.sh
 create mode 100644 requirements.txt
 rename init-schema.sh => tools/init-schema.sh (100%)
 create mode 100755 tools/install.sh

diff --git a/.gitignore b/.gitignore
index ba74660..18505b2 100644
--- a/.gitignore
+++ b/.gitignore
@@ -55,3 +55,6 @@ docs/_build/
 
 # PyBuilder
 target/
+
+# Ignore the in here installed virtualenv
+virtualenv/
diff --git a/README.md b/README.md
index 38bf399..7f6794c 100644
--- a/README.md
+++ b/README.md
@@ -24,6 +24,9 @@ Used python3 modules:
 - `email` to construct the email.
 - `syslog` for logging the sending of emails into syslog
 
+External modules:
+- `chardet`, to autodetect the input charset (which can vary)
+
 ## Requirements:
 
 - A mailserver.
@@ -35,7 +38,9 @@ Used python3 modules:
 
 - Install `python3` (do I need to say this?)
 - Clone this repository into a random directory, reachable for dovecot2
-- Initialize the sqlite db with running `init-schema.sh`
+- Change to that directory
+- Install virtualenv and external modules by running `tools/install.sh`
+- Initialize the sqlite db with running `tools/init-schema.sh`
 - Copy your replies (mentioned above) into the same repo directory
 - Edit your sieve configuration to pipe your given role into this tool.
 
diff --git a/email-responder.py b/email-responder.py
old mode 100755
new mode 100644
index 3152ec9..9c187c8
--- a/email-responder.py
+++ b/email-responder.py
@@ -2,12 +2,12 @@
 
 import sys
 import os
-import io
 import smtplib
 import sqlite3
 import datetime
 import syslog
 import codecs
+import chardet
 
 from email.mime.text import MIMEText
 from email.mime.multipart import MIMEMultipart
@@ -19,8 +19,12 @@ from email.header import decode_header, make_header
 my_dir = os.path.dirname(os.path.realpath(__file__))
 os.chdir(my_dir)
 
-stdin_utf8 = io.TextIOWrapper(sys.stdin.buffer, encoding='utf-8')
-original_headers = Parser().parsestr(stdin_utf8.read())
+input_bytes = sys.stdin.buffer.read()  # This is bytes
+
+encoding_result = chardet.detect(input_bytes)
+input_decoded = input_bytes.decode(encoding_result['encoding'])
+
+original_headers = Parser().parsestr(input_decoded)
 # with open('email1.txt') as fp:
 #     original_headers = Parser().parsestr(fp.read())
 
diff --git a/email-responder.sh b/email-responder.sh
new file mode 100755
index 0000000..c7d5127
--- /dev/null
+++ b/email-responder.sh
@@ -0,0 +1,9 @@
+#!/usr/bin/env bash
+
+PROJECT_DIR=$(cd $(dirname ${BASH_SOURCE[0]});pwd)
+
+cd $PROJECT_DIR
+
+source virtualenv/bin/activate
+
+exec python3 email-responder.py
diff --git a/requirements.txt b/requirements.txt
new file mode 100644
index 0000000..7e01c11
--- /dev/null
+++ b/requirements.txt
@@ -0,0 +1 @@
+chardet==2.3.0
diff --git a/init-schema.sh b/tools/init-schema.sh
similarity index 100%
rename from init-schema.sh
rename to tools/init-schema.sh
diff --git a/tools/install.sh b/tools/install.sh
new file mode 100755
index 0000000..43ad554
--- /dev/null
+++ b/tools/install.sh
@@ -0,0 +1,10 @@
+#!/usr/bin/env bash
+
+PROJECT_DIR=$(cd $(dirname ${BASH_SOURCE[0]});cd ..;pwd)
+
+cd $PROJECT_DIR
+
+pyvenv virtualenv
+source virtualenv/bin/activate
+
+pip install -r requirements.txt