From 995a3d09ffb3a405891894693abddefeffa749a3 Mon Sep 17 00:00:00 2001
From: Kovid Goyal <kovid@kovidgoyal.net>
Date: Wed, 16 Dec 2009 13:39:05 -0700
Subject: [PATCH] MOBI Input: Ignore \x02 bytes in data stream for cp1252
 encoded files

---
 src/calibre/ebooks/mobi/reader.py | 4 +++-
 1 file changed, 3 insertions(+), 1 deletion(-)

diff --git a/src/calibre/ebooks/mobi/reader.py b/src/calibre/ebooks/mobi/reader.py
index 7d676a230c..20d31d48b4 100644
--- a/src/calibre/ebooks/mobi/reader.py
+++ b/src/calibre/ebooks/mobi/reader.py
@@ -701,7 +701,9 @@ class MobiReader(object):
         if self.book_header.ancient and '<html' not in self.mobi_html[:300].lower():
             self.mobi_html = self.mobi_html.replace('\r ', '\n\n ')
         self.mobi_html = self.mobi_html.replace('\0', '')
-        self.mobi_html = self.mobi_html.replace('\x1e', '') # record separator
+        if self.book_header.codec == 'cp1252':
+            self.mobi_html = self.mobi_html.replace('\x1e', '') # record separator
+            self.mobi_html = self.mobi_html.replace('\x02', '') # start of text
         return processed_records