From 5a5e63499ef1e64b919401d6ea12dc7f16ab1d0b Mon Sep 17 00:00:00 2001
From: Kovid Goyal <kovid@kovidgoyal.net>
Date: Sat, 19 Jul 2008 20:23:46 -0700
Subject: [PATCH] Fix bug in html2lrf that was causing problems with paragraghs
 inside <a> tags.

---
 src/calibre/ebooks/lrf/html/convert_from.py | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)

diff --git a/src/calibre/ebooks/lrf/html/convert_from.py b/src/calibre/ebooks/lrf/html/convert_from.py
index e602a61156..e1014d69dd 100644
--- a/src/calibre/ebooks/lrf/html/convert_from.py
+++ b/src/calibre/ebooks/lrf/html/convert_from.py
@@ -92,8 +92,8 @@ class HTMLConverter(object, LoggingInterface):
                          # sometimes there are unterminated comments
                         (re.compile(r"<\s*style.*?>(.*?)<\/\s*style\s*>", re.DOTALL|re.IGNORECASE),
                          lambda match: match.group().replace('<!--', '').replace('-->', '')),
-                         # remove <p> tags from within <a> tags
-                        (re.compile(r'%(open)s(.*?)%(close)s'%tag_regex('a'), re.DOTALL|re.IGNORECASE),
+                         # remove <p> tags from within <a href> tags
+                        (re.compile(r'<\s*a\s+[^<>]*href\s*=[^<>]*>(.*?)<\s*/\s*a\s*>', re.DOTALL|re.IGNORECASE),
                          lambda match: re.compile(r'%(open)s|%(close)s'%tag_regex('p'), re.IGNORECASE).sub('', match.group())),
                         
                         # Replace common line break patterns with line breaks