From 38e08a7bd39b868ac496f0e0dcbf1bacb7f231b8 Mon Sep 17 00:00:00 2001
From: ldolse <ldolse@yahoo.com>
Date: Mon, 22 Nov 2010 19:19:47 +0800
Subject: [PATCH] had to include heading tags with the previous change, because
 lxml/beautifulsoup were equally unhappy when heading tags contained paragraph
 tags

---
 src/calibre/ebooks/mobi/reader.py | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)

diff --git a/src/calibre/ebooks/mobi/reader.py b/src/calibre/ebooks/mobi/reader.py
index 972b049643..355fc87c0c 100644
--- a/src/calibre/ebooks/mobi/reader.py
+++ b/src/calibre/ebooks/mobi/reader.py
@@ -477,8 +477,8 @@ class MobiReader(object):
         self.processed_html = self.processed_html.replace('<mbp: ', '<mbp:')
         self.processed_html = re.sub(r'<?xml[^>]*>', '', self.processed_html)
         # Rearrange various style tags & paragraph tags so that lxml and BeautifulSoup don't get upset
-        self.processed_html = re.sub(r'(?i)(?P<styletags>(<(i|b|u|em|small|big|strong|tt)>\s*){1,})(?P<para><p[^>]*>)', '\g<para>'+'\g<styletags>', self.processed_html)
-        self.processed_html = re.sub(r'(?i)(?P<para></p[^>]*>)\s*(?P<styletags>(</(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<styletags>(<(h\d+|i|b|u|em|small|big|strong|tt)>\s*){1,})(?P<para><p[^>]*>)', '\g<para>'+'\g<styletags>', self.processed_html)
+        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)
 
     def remove_random_bytes(self, html):
         return re.sub('\x14|\x15|\x19|\x1c|\x1d|\xef|\x12|\x13|\xec|\x08',