From b7b3ec03937be685219f075283d73295f52b5507 Mon Sep 17 00:00:00 2001
From: Erik Stein <erik@classlibrary.net>
Date: Mon, 8 Jan 2018 17:19:08 +0100
Subject: [PATCH] Added html_to_markdown function and template filter.

---
 setup.py                                  |  1 +
 shared/markup/markdown_utils.py           |  6 ++++++
 shared/markup/templatetags/markup_tags.py | 13 +++++++++++++
 3 files changed, 20 insertions(+)

diff --git a/setup.py b/setup.py
index c111c9e..6e94ad9 100644
--- a/setup.py
+++ b/setup.py
@@ -37,6 +37,7 @@ setup(
         # 'django<2', commented out to make `pip install -U` easier
         'beautifulsoup4',
         'lxml',
+        'html2text',
     ],
     classifiers=[
         'Development Status :: 4 - Beta',
diff --git a/shared/markup/markdown_utils.py b/shared/markup/markdown_utils.py
index e0a9156..a0c8aa1 100644
--- a/shared/markup/markdown_utils.py
+++ b/shared/markup/markdown_utils.py
@@ -5,6 +5,8 @@ from __future__ import unicode_literals
 import markdown as markdown_module
 from django.utils.html import strip_tags
 from django.utils.safestring import mark_safe
+
+import html2text
 from shared.utils.text import html_entities_to_unicode
 
 
@@ -80,3 +82,7 @@ def markdown_to_text(text, **kwargs):
     """
     html = markdown_to_html(text, **kwargs)
     return strip_tags(html_entities_to_unicode(html))
+
+
+def html_to_markdown(html):
+    return html2text.html2text(html)
diff --git a/shared/markup/templatetags/markup_tags.py b/shared/markup/templatetags/markup_tags.py
index 71bcf2a..465db02 100644
--- a/shared/markup/templatetags/markup_tags.py
+++ b/shared/markup/templatetags/markup_tags.py
@@ -55,3 +55,16 @@ urlfinder2 = re.compile('\s(http:\/\/\S+)')
 def urlify_markdown(value):
     value = urlfinder.sub(r'<\1>', value)
     return urlfinder2.sub(r' <\1>', value)
+
+
+@register.filter(needs_autoescape=False)
+@stringfilter
+def html_to_markdown(html, autoescape=None):
+    """
+    Converts a HTML string to markdown, using the html2text tool.
+    """
+    if autoescape:
+        esc = conditional_escape
+    else:
+        esc = lambda x: x
+    return markdown_utils.html_to_markdown(esc(text))