From d558617cd769d5f01693adb929ea044e91d55b02 Mon Sep 17 00:00:00 2001
From: Igor Dvorkin <idvorkin@gmail.com>
Date: Sun, 15 Nov 2020 09:04:22 -0800
Subject: [PATCH 1/3] Add support for headings that include nested block
 elements

---
 markdownify/__init__.py   | 64 +++++++++++++++++++++++++++------------
 tests/test_conversions.py |  8 +++++
 2 files changed, 52 insertions(+), 20 deletions(-)

diff --git a/markdownify/__init__.py b/markdownify/__init__.py
index 2d5daf1..0a376a7 100644
--- a/markdownify/__init__.py
+++ b/markdownify/__init__.py
@@ -61,22 +61,28 @@ class MarkdownConverter(object):
 
     def convert(self, html):
         soup = BeautifulSoup(html, 'html.parser')
-        return self.process_tag(soup, children_only=True)
+        return self.process_tag(soup, convert_as_inline=False, children_only=True)
 
-    def process_tag(self, node, children_only=False):
+    def process_tag(self, node, convert_as_inline, children_only=False):
         text = ''
+        # markdown headings can't include block elements (elements w/newlines)
+        isHeading = node.name.startswith('h')
+        convert_children_as_inline = convert_as_inline
+
+        if not children_only and isHeading:
+            convert_children_as_inline = True
 
         # Convert the children first
         for el in node.children:
             if isinstance(el, NavigableString):
                 text += self.process_text(six.text_type(el))
             else:
-                text += self.process_tag(el)
+                text += self.process_tag(el, convert_children_as_inline)
 
         if not children_only:
             convert_fn = getattr(self, 'convert_%s' % node.name, None)
             if convert_fn and self.should_convert_tag(node.name):
-                text = convert_fn(node, text)
+                text = convert_fn(node, text, convert_as_inline)
 
         return text
 
@@ -89,8 +95,8 @@ class MarkdownConverter(object):
         if m:
             n = int(m.group(1))
 
-            def convert_tag(el, text):
-                return self.convert_hn(n, el, text)
+            def convert_tag(el, text, convert_as_inline):
+                return self.convert_hn(n, el, text, convert_as_inline)
 
             convert_tag.__name__ = 'convert_h%s' % n
             setattr(self, convert_tag.__name__, convert_tag)
@@ -116,10 +122,12 @@ class MarkdownConverter(object):
         text = (text or '').rstrip()
         return '%s\n%s\n\n' % (text, pad_char * len(text)) if text else ''
 
-    def convert_a(self, el, text):
+    def convert_a(self, el, text, convert_as_inline):
         prefix, suffix, text = chomp(text)
         if not text:
             return ''
+        if convert_as_inline:
+            return text
         href = el.get('href')
         title = el.get('title')
         if self.options['autolinks'] and text == href and not title:
@@ -128,22 +136,32 @@ class MarkdownConverter(object):
         title_part = ' "%s"' % title.replace('"', r'\"') if title else ''
         return '%s[%s](%s%s)%s' % (prefix, text, href, title_part, suffix) if href else text
 
-    def convert_b(self, el, text):
-        return self.convert_strong(el, text)
+    def convert_b(self, el, text, convert_as_inline):
+        return self.convert_strong(el, text, convert_as_inline)
+
+    def convert_blockquote(self, el, text, convert_as_inline):
+
+        if convert_as_inline:
+            return text
 
-    def convert_blockquote(self, el, text):
         return '\n' + line_beginning_re.sub('> ', text) if text else ''
 
-    def convert_br(self, el, text):
+    def convert_br(self, el, text, convert_as_inline):
+        if convert_as_inline:
+            return ""
+
         return '  \n'
 
-    def convert_em(self, el, text):
+    def convert_em(self, el, text, convert_as_inline):
         prefix, suffix, text = chomp(text)
         if not text:
             return ''
         return '%s*%s*%s' % (prefix, text, suffix)
 
-    def convert_hn(self, n, el, text):
+    def convert_hn(self, n, el, text, convert_as_inline):
+        if convert_as_inline:
+            return text
+
         style = self.options['heading_style']
         text = text.rstrip()
         if style == UNDERLINED and n <= 2:
@@ -154,10 +172,14 @@ class MarkdownConverter(object):
             return '%s %s %s\n\n' % (hashes, text, hashes)
         return '%s %s\n\n' % (hashes, text)
 
-    def convert_i(self, el, text):
-        return self.convert_em(el, text)
+    def convert_i(self, el, text, convert_as_inline):
+        return self.convert_em(el, text, convert_as_inline)
+
+    def convert_list(self, el, text, convert_as_inline):
+
+        # Converting a list to inline is undefined.
+        # Ignoring convert_to_inline for list.
 
-    def convert_list(self, el, text):
         nested = False
         while el:
             if el.name == 'li':
@@ -172,7 +194,7 @@ class MarkdownConverter(object):
     convert_ul = convert_list
     convert_ol = convert_list
 
-    def convert_li(self, el, text):
+    def convert_li(self, el, text, convert_as_inline):
         parent = el.parent
         if parent is not None and parent.name == 'ol':
             if parent.get("start"):
@@ -190,16 +212,18 @@ class MarkdownConverter(object):
             bullet = bullets[depth % len(bullets)]
         return '%s %s\n' % (bullet, text or '')
 
-    def convert_p(self, el, text):
+    def convert_p(self, el, text, convert_as_inline):
+        if convert_as_inline:
+            return text
         return '%s\n\n' % text if text else ''
 
-    def convert_strong(self, el, text):
+    def convert_strong(self, el, text, convert_as_inline):
         prefix, suffix, text = chomp(text)
         if not text:
             return ''
         return '%s**%s**%s' % (prefix, text, suffix)
 
-    def convert_img(self, el, text):
+    def convert_img(self, el, text, convert_as_inline):
         alt = el.attrs.get('alt', None) or ''
         src = el.attrs.get('src', None) or ''
         title = el.attrs.get('title', None) or ''
diff --git a/tests/test_conversions.py b/tests/test_conversions.py
index 12ce36b..65dbfd2 100644
--- a/tests/test_conversions.py
+++ b/tests/test_conversions.py
@@ -107,6 +107,14 @@ def test_hn():
     assert md('<h6>Hello</h6>') == '###### Hello\n\n'
 
 
+def test_hn_nested_tag():
+    assert md('<h3>A <b>Bold</b> C </h3>') == '### A **Bold** C\n\n'
+    assert md('<h3>A <p>P</p> C </h3>') == '### A P C\n\n'
+    assert md('<h1>A <p>P</p> C </h1>', heading_style=ATX_CLOSED) == '# A P C #\n\n'
+    assert md('<h1>A <p>P</p> C </h1>', heading_style=ATX) == '# A P C\n\n'
+    assert md('<h3>A <blockquote>BQ</blockquote> C </h3>') == '### A BQ C\n\n'
+
+
 def test_atx_headings():
     assert md('<h1>Hello</h1>', heading_style=ATX) == '# Hello\n\n'
     assert md('<h2>Hello</h2>', heading_style=ATX) == '## Hello\n\n'

From 7780f82c302483a5537175f435d271d66cfc4d84 Mon Sep 17 00:00:00 2001
From: Igor Dvorkin <idvorkin@gmail.com>
Date: Fri, 11 Dec 2020 16:54:14 -0800
Subject: [PATCH 2/3] Using a regexp to determine if a tag is a heading.

---
 markdownify/__init__.py   | 3 ++-
 tests/test_conversions.py | 8 ++++++++
 2 files changed, 10 insertions(+), 1 deletion(-)

diff --git a/markdownify/__init__.py b/markdownify/__init__.py
index 0a376a7..cb12d43 100644
--- a/markdownify/__init__.py
+++ b/markdownify/__init__.py
@@ -6,6 +6,7 @@ import six
 convert_heading_re = re.compile(r'convert_h(\d+)')
 line_beginning_re = re.compile(r'^', re.MULTILINE)
 whitespace_re = re.compile(r'[\r\n\s\t ]+')
+html_heading_re = re.compile(r'h[1-6]')
 
 
 # Heading styles
@@ -66,7 +67,7 @@ class MarkdownConverter(object):
     def process_tag(self, node, convert_as_inline, children_only=False):
         text = ''
         # markdown headings can't include block elements (elements w/newlines)
-        isHeading = node.name.startswith('h')
+        isHeading = html_heading_re.match(node.name) is not None
         convert_children_as_inline = convert_as_inline
 
         if not children_only and isHeading:
diff --git a/tests/test_conversions.py b/tests/test_conversions.py
index 65dbfd2..ab1ce05 100644
--- a/tests/test_conversions.py
+++ b/tests/test_conversions.py
@@ -115,6 +115,14 @@ def test_hn_nested_tag():
     assert md('<h3>A <blockquote>BQ</blockquote> C </h3>') == '### A BQ C\n\n'
 
 
+def test_hr():
+    assert md('<hr>hr</hr>') == 'hr'
+
+
+def test_head():
+    assert md('<head>head</head>') == 'head'
+
+
 def test_atx_headings():
     assert md('<h1>Hello</h1>', heading_style=ATX) == '# Hello\n\n'
     assert md('<h2>Hello</h2>', heading_style=ATX) == '## Hello\n\n'

From 05ea8dc58ab35e1d3e8bae5b84df77bd4e3dc14d Mon Sep 17 00:00:00 2001
From: Igor Dvorkin <idvorkin@gmail.com>
Date: Sun, 13 Dec 2020 17:39:08 +0000
Subject: [PATCH 3/3] Add many tests and support image tag

---
 markdownify/__init__.py   |  3 +++
 tests/test_conversions.py | 38 ++++++++++++++++++++++++++++++++++----
 2 files changed, 37 insertions(+), 4 deletions(-)

diff --git a/markdownify/__init__.py b/markdownify/__init__.py
index cb12d43..2cd8fc8 100644
--- a/markdownify/__init__.py
+++ b/markdownify/__init__.py
@@ -229,6 +229,9 @@ class MarkdownConverter(object):
         src = el.attrs.get('src', None) or ''
         title = el.attrs.get('title', None) or ''
         title_part = ' "%s"' % title.replace('"', r'\"') if title else ''
+        if convert_as_inline:
+            return alt
+
         return '![%s](%s%s)' % (alt, src, title_part)
 
 
diff --git a/tests/test_conversions.py b/tests/test_conversions.py
index ab1ce05..f5fc1c2 100644
--- a/tests/test_conversions.py
+++ b/tests/test_conversions.py
@@ -107,12 +107,42 @@ def test_hn():
     assert md('<h6>Hello</h6>') == '###### Hello\n\n'
 
 
-def test_hn_nested_tag():
-    assert md('<h3>A <b>Bold</b> C </h3>') == '### A **Bold** C\n\n'
-    assert md('<h3>A <p>P</p> C </h3>') == '### A P C\n\n'
+def test_hn_nested_tag_heading_style():
     assert md('<h1>A <p>P</p> C </h1>', heading_style=ATX_CLOSED) == '# A P C #\n\n'
     assert md('<h1>A <p>P</p> C </h1>', heading_style=ATX) == '# A P C\n\n'
-    assert md('<h3>A <blockquote>BQ</blockquote> C </h3>') == '### A BQ C\n\n'
+
+
+def test_hn_nested_simple_tag():
+    tag_to_markdown = [
+        ("strong", "**strong**"),
+        ("b", "**b**"),
+        ("em", "*em*"),
+        ("i", "*i*"),
+        ("p", "p"),
+        ("a", "a"),
+        ("div", "div"),
+        ("blockquote", "blockquote"),
+    ]
+
+    for tag, markdown in tag_to_markdown:
+        assert md('<h3>A <' + tag + '>' + tag + '</' + tag + '> B</h3>') == '### A ' + markdown + ' B\n\n'
+
+    assert md('<h3>A <br>B</h3>', heading_style=ATX) == '### A B\n\n'
+
+    # Nested lists not supported
+    # assert md('<h3>A <ul><li>li1</i><li>l2</li></ul></h3>', heading_style=ATX) == '### A li1 li2 B\n\n'
+
+
+def test_hn_nested_img():
+    assert md('<img src="/path/to/img.jpg" alt="Alt text" title="Optional title" />') == '![Alt text](/path/to/img.jpg "Optional title")'
+    assert md('<img src="/path/to/img.jpg" alt="Alt text" />') == '![Alt text](/path/to/img.jpg)'
+    image_attributes_to_markdown = [
+        ("", ""),
+        ("alt='Alt Text'", "Alt Text"),
+        ("alt='Alt Text' title='Optional title'", "Alt Text"),
+    ]
+    for image_attributes, markdown in image_attributes_to_markdown:
+        assert md('<h3>A <img src="/path/to/img.jpg " ' + image_attributes + '/> B</h3>') == '### A ' + markdown + ' B\n\n'
 
 
 def test_hr():