From 205bbf61a2dc66095cd6bea09abe0b822b7abf4e Mon Sep 17 00:00:00 2001
From: Ethan Dalool <ethan@voussoir.net>
Date: Tue, 7 Jan 2020 18:23:44 -0800
Subject: [PATCH] Add remove_header_br.

---
 cleanerupper/plugin.py | 10 ++++++++++
 1 file changed, 10 insertions(+)

diff --git a/cleanerupper/plugin.py b/cleanerupper/plugin.py
index 2be0373..692c065 100644
--- a/cleanerupper/plugin.py
+++ b/cleanerupper/plugin.py
@@ -64,6 +64,16 @@ def bring_punctuation_into_italics(html):
             html = re.sub('</{tag}>\\{punct}'.format(**locals()), '{punct}</{tag}>'.format(**locals()), html)
     return html
 
+@html_cleaner
+def remove_header_br(html):
+    html = re.sub(r'<h1>([^\n]+?)\s*<br/>\s*([^\n]+?)</h1>', r'<h1>\1 \2</h1>', html)
+    html = re.sub(r'<h2>([^\n]+?)\s*<br/>\s*([^\n]+?)</h2>', r'<h2>\1 \2</h2>', html)
+    html = re.sub(r'<h3>([^\n]+?)\s*<br/>\s*([^\n]+?)</h3>', r'<h3>\1 \2</h3>', html)
+    html = re.sub(r'<h4>([^\n]+?)\s*<br/>\s*([^\n]+?)</h4>', r'<h4>\1 \2</h4>', html)
+    html = re.sub(r'<h5>([^\n]+?)\s*<br/>\s*([^\n]+?)</h5>', r'<h5>\1 \2</h5>', html)
+    html = re.sub(r'<h6>([^\n]+?)\s*<br/>\s*([^\n]+?)</h6>', r'<h6>\1 \2</h6>', html)
+    return html
+
 @html_cleaner
 def remove_misc_strings(html):
     html = html.replace('epub:type="pagebreak"', '')