matthewwithanm · be-student · Sep 6, 2026
diff --git a/README.rst b/README.rst
@@ -154,7 +154,8 @@ strip_document
   Controls whether leading and/or trailing separation newlines are removed from
   the final converted document. Supported values are ``LSTRIP`` (leading),
   ``RSTRIP`` (trailing), ``STRIP`` (both), and ``None`` (neither). Newlines
-  within the document are unaffected.
+  within the document are unaffected. This also applies when ``convert_soup()``
+  receives an individual BeautifulSoup ``Tag`` rather than a complete soup.
   Defaults to ``STRIP``.
 
 strip_pre

diff --git a/markdownify/__init__.py b/markdownify/__init__.py
@@ -223,7 +223,14 @@ def convert(self, html):
         return self.convert_soup(soup)
 
     def convert_soup(self, soup):
-        return self.process_tag(soup, parent_tags=set())
+        text = self.process_tag(soup, parent_tags=set())
+        # A supplied Tag is also a complete conversion input. BeautifulSoup's
+        # document node already runs this finalizer during process_tag().
+        if soup.name != '[document]':
+            convert_fn = self.get_conv_fn_cached('[document]')
+            if convert_fn is not None:
+                text = convert_fn(soup, text, parent_tags={'[document]'})
+        return text
 
     def process_element(self, node, parent_tags=None):
         if isinstance(node, NavigableString):

diff --git a/tests/test_custom_converter.py b/tests/test_custom_converter.py
@@ -42,3 +42,49 @@ def test_soup():
     html = '<b>test</b>'
     soup = BeautifulSoup(html, 'html.parser')
     assert MarkdownConverter().convert_soup(soup) == '**test**'
+
+
+def test_tag_document_stripping():
+    from markdownify import LSTRIP, RSTRIP, STRIP
+
+    html = '<html><body><div><p>hello</p><p>world</p></div></body></html>'
+    for mode, expected in [(LSTRIP, 'hello\n\nworld\n\n'),
+                           (RSTRIP, '\n\nhello\n\nworld'),
+                           (STRIP, 'hello\n\nworld'),
+                           (None, '\n\nhello\n\nworld\n\n')]:
+        for name in ['html', 'body', 'div']:
+            soup = BeautifulSoup(html, 'html.parser')
+            tag = soup.find(name)
+            before = str(soup)
+            assert MarkdownConverter(strip_document=mode).convert_soup(tag) == expected
+            assert str(soup) == before
+            assert tag.parent is not None
+
+
+def test_tag_document_invalid_strip_mode():
+    import pytest
+
+    soup = BeautifulSoup('<p>hello</p>', 'html.parser')
+    with pytest.raises(ValueError, match='Invalid value for strip_document'):
+        MarkdownConverter(strip_document='invalid').convert_soup(soup.p)
+
+
+def test_document_converter_called_once():
+    class CountingConverter(MarkdownConverter):
+        calls = 0
+
+        def convert__document_(self, el, text, parent_tags):
+            self.calls += 1
+            return super().convert__document_(el, text, parent_tags)
+
+    soup = BeautifulSoup('<div><p>hello</p></div>', 'html.parser')
+    for root in [soup, soup.div, soup.p]:
+        converter = CountingConverter()
+        assert converter.convert_soup(root) == 'hello'
+        assert converter.calls == 1
+
+
+def test_tag_document_converter_exclusion():
+    soup = BeautifulSoup('<p>hello</p>', 'html.parser')
+    for root in [soup, soup.p]:
+        assert MarkdownConverter(strip=['[document]']).convert_soup(root) == '\n\nhello\n\n'