From b56891cdaecd71deeb159ffbb5ca03e3827ed441 Mon Sep 17 00:00:00 2001
From: Kovid Goyal <kovid@kovidgoyal.net>
Date: Tue, 7 Feb 2012 14:10:18 +0530
Subject: [PATCH] MOBI Input: Handle files that have spurious closing </body>
 and/or </html> tags in their markup. Fixes #925833 (prc file fails to read or
 convert)

---
 src/calibre/ebooks/mobi/reader.py | 11 +++++++++++
 1 file changed, 11 insertions(+)

diff --git a/src/calibre/ebooks/mobi/reader.py b/src/calibre/ebooks/mobi/reader.py
index 347e4fda3a..7e69fc89d0 100644
--- a/src/calibre/ebooks/mobi/reader.py
+++ b/src/calibre/ebooks/mobi/reader.py
@@ -516,6 +516,17 @@ class MobiReader(object):
         self.processed_html = re.sub(r'(?i)(?P<para></p[^>]*>)\s*(?P<styletags>(</(h\d+|i|b|u|em|small|big|strong|tt)>\s*){1,})', '\g<styletags>'+'\g<para>', self.processed_html)
         self.processed_html = re.sub(r'(?i)(?P<blockquote>(</(blockquote|div)[^>]*>\s*){1,})(?P<para></p[^>]*>)', '\g<para>'+'\g<blockquote>', self.processed_html)
         self.processed_html = re.sub(r'(?i)(?P<para><p[^>]*>)\s*(?P<blockquote>(<(blockquote|div)[^>]*>\s*){1,})', '\g<blockquote>'+'\g<para>', self.processed_html)
+        bods = htmls = 0
+        for x in re.finditer(ur'</body>|</html>', self.processed_html):
+            if x == '</body>': bods +=1
+            else: htmls += 1
+            if bods > 1 and htmls > 1:
+                break
+        if bods > 1:
+            self.processed_html = self.processed_html.replace('</body>', '')
+        if htmls > 1:
+            self.processed_html = self.processed_html.replace('</html>', '')
+
 
 
     def remove_random_bytes(self, html):