Merge pull request #4 from Easy-D/preserve-regular-blockquotes-claw

Preserve regular blockquotes
tictail · Aug 25, 2015 · 07dade1 · 07dade1
2 parents f2a2229 + 36c0d39
commit 07dade1
Show file tree

Hide file tree

Showing 2 changed files with 22 additions and 3 deletions.
diff --git a/claw/html_quotations.py b/claw/html_quotations.py
@@ -140,9 +140,10 @@ def cut_by_id(html_message):
 
 
 def cut_blockquote(html_message):
-    ''' Cuts blockquote with wrapping elements. '''
-    quote = html_message.find('.//blockquote')
-    if quote is not None:
+    ''' Cuts the last non-nested blockquote with wrapping elements. '''
+    quote = html_message.xpath('(.//blockquote)[not(ancestor::blockquote)][last()]')
+    if quote:
+        quote = quote[0]
         quote.getparent().remove(quote)
         return True
 

diff --git a/tests/html_quotations_test.py b/tests/html_quotations_test.py
@@ -49,6 +49,24 @@ def test_quotation_splitter_outside_blockquote():
         RE_WHITESPACE.sub('', quotations.extract_from_html(msg_body)))
 
 
+def test_regular_blockquote():
+    msg_body = """Reply
+<blockquote>Regular</blockquote>
+
+<div>
+  On 11-Apr-2011, at 6:54 PM, Bob &lt;bob@example.com&gt; wrote:
+</div>
+
+<blockquote>
+  <div>
+    <blockquote>Nested</blockquote>
+  </div>
+</blockquote>
+"""
+    eq_("<html><body><p>Reply</p><blockquote>Regular</blockquote><div></div></body></html>",
+        RE_WHITESPACE.sub('', quotations.extract_from_html(msg_body)))
+
+
 def test_no_blockquote():
     msg_body = """
 <html>