cdodge/fix-updates-parsing: seems like studio now uses a different HTML…

cdodge/fix-updates-parsing: seems like studio now uses a different HTML formatting, e.g. sections & articles rather than ol & li

make backwards compatible to read older versions of Studio produced updates

add some tests

lms/djangoapps/api_manager/courses/content.py

@@ -7,10 +7,33 @@ from textwrap import dedent
 
 TEST_COURSE_UPDATES_CONTENT = dedent(
     """
+    <section>
+      <article>
+        <h2>April 18, 2014</h2>
+        This does not have a paragraph tag around it
+      </article>
+      <article>
+        <h2>April 17, 2014</h2>
+        Some text before paragraph tag<p>This is inside paragraph tag</p>Some text after tag
+      </article>
+      <article>
+        <h2>April 16, 2014</h2>
+        Some text before paragraph tag<p>This is inside paragraph tag</p>Some text after tag<p>one more</p>
+      </article>
+      <article>
+        <h2>April 15, 2014</h2>
+        <p>A perfectly</p><p>formatted piece</p><p>of HTML</p>
+      </article>
+    </section>
+    """
+)
+
+TEST_COURSE_UPDATES_CONTENT_LEGACY = dedent(
+    """
     <ol>
       <li>
         <h2>April 18, 2014</h2>
-        This does not have a paragraph tag around it
+        This is some legacy content
       </li>
       <li>
         <h2>April 17, 2014</h2>

lms/djangoapps/api_manager/courses/tests.py

@@ -15,7 +15,7 @@ from django.test.utils import override_settings
 from courseware.tests.modulestore_config import TEST_DATA_MIXED_MODULESTORE
 from xmodule.modulestore.tests.factories import CourseFactory, ItemFactory
 
-from .content import TEST_COURSE_OVERVIEW_CONTENT, TEST_COURSE_UPDATES_CONTENT
+from .content import TEST_COURSE_OVERVIEW_CONTENT, TEST_COURSE_UPDATES_CONTENT, TEST_COURSE_UPDATES_CONTENT_LEGACY
 from .content import TEST_STATIC_TAB1_CONTENT, TEST_STATIC_TAB2_CONTENT
 
 TEST_API_KEY = str(uuid.uuid4())
@@ -560,6 +560,38 @@ class CoursesApiTests(TestCase):
         response = self.do_get(test_uri)
         self.assertEqual(response.status_code, 404)
 
+    def test_courses_updates_legacy(self):
+        #try a bogus course_id to test failure case
+        test_course = CourseFactory.create()
+        ItemFactory.create(
+            category="course_info",
+            parent_location=test_course.location,
+            data=TEST_COURSE_UPDATES_CONTENT_LEGACY,
+            display_name="updates"
+        )
+        test_uri = self.base_courses_uri + '/' + test_course.id + '/updates'
+        response = self.do_get(test_uri)
+        self.assertEqual(response.status_code, 200)
+        self.assertGreater(len(response.data), 0)
+        self.assertEqual(response.data['content'], TEST_COURSE_UPDATES_CONTENT_LEGACY)
+
+        # then try parsed
+        test_uri = self.base_courses_uri + '/' + test_course.id + '/updates?parse=True'
+        response = self.do_get(test_uri)
+        self.assertEqual(response.status_code, 200)
+        self.assertGreater(len(response.data), 0)
+
+        postings = response.data['postings']
+        self.assertEqual(len(postings), 4)
+        self.assertEqual(postings[0]['date'], 'April 18, 2014')
+        self.assertEqual(postings[0]['content'], 'This is some legacy content')
+        self.assertEqual(postings[1]['date'], 'April 17, 2014')
+        self.assertEqual(postings[1]['content'], 'Some text before paragraph tag<p>This is inside paragraph tag</p>Some text after tag')
+        self.assertEqual(postings[2]['date'], 'April 16, 2014')
+        self.assertEqual(postings[2]['content'], 'Some text before paragraph tag<p>This is inside paragraph tag</p>Some text after tag<p>one more</p>')
+        self.assertEqual(postings[3]['date'], 'April 15, 2014')
+        self.assertEqual(postings[3]['content'], '<p>A perfectly</p><p>formatted piece</p><p>of HTML</p>')
+
     def test_static_tab_list_get(self):
         test_uri = '{}/{}/static_tabs'.format(self.base_courses_uri, self.test_course_id)
         response = self.do_get(test_uri)

lms/djangoapps/api_manager/courses/views.py

@@ -224,7 +224,11 @@ def _parse_updates_html(html):
     tree = etree.parse(StringIO(html), parser)
 
     # get all of the individual postings
-    postings = tree.findall('/body/ol/li')
+    postings = tree.findall('/body/section/article')
+
+    # be backwards compatible
+    if not postings:
+        postings = tree.findall('/body/ol/li')
 
     result = []
     for posting in postings: