From bfe6164a01e19e1b8e032bcfb6cdb278c1acb188 Mon Sep 17 00:00:00 2001
From: malcolm smith <malsmith@treehousesystems.com>
Date: Tue, 21 Apr 2015 22:10:25 -0400
Subject: [PATCH] support funky Microsoft Word XHTML unicode escapes

---
 html2text/__init__.py | 5 +++--
 html2text/utils.py    | 9 +++++++++
 2 files changed, 12 insertions(+), 2 deletions(-)

diff --git a/html2text/__init__.py b/html2text/__init__.py
index 1f3e803fb7..61bd602973 100644
--- a/html2text/__init__.py
+++ b/html2text/__init__.py
@@ -14,6 +14,7 @@
 from html2text import config
 
 from html2text.utils import (
+    wide_unichr,
     name2cp,
     unifiable_n,
     google_text_emphasis,
@@ -703,7 +704,7 @@ def charref(self, name):
             return unifiable_n[c]
         else:
             try:
-                return unichr(c)
+                return wide_unichr(c)
             except NameError:  # Python3
                 return chr(c)
 
@@ -720,7 +721,7 @@ def entityref(self, c):
                     return config.UNIFIABLE[c]
                 else:
                     try:
-                        return unichr(name2cp(c))
+                        return wide_unichr(name2cp(c))
                     except NameError:  # Python3
                         return chr(name2cp(c))
 
diff --git a/html2text/utils.py b/html2text/utils.py
index 0946930f4d..bc9fb48f45 100644
--- a/html2text/utils.py
+++ b/html2text/utils.py
@@ -1,8 +1,17 @@
 import sys
 from html2text import config
+import struct 
 
 from html2text.compat import htmlentitydefs
 
+# Based on http://stackoverflow.com/questions/7105874/valueerror-unichr-arg-not-in-range0x10000-narrow-python-build-please-hel
+def wide_unichr(i):
+    try:
+        return unichr(i)
+    except ValueError:
+        return struct.pack('i', i).decode('utf-32')
+
+
 
 def name2cp(k):
     if k == 'apos':