Search: remove non-generic parser code

Closes #10272.

We can't remove the code that generates the fjson files,
since they are being used by the embed API (v2),
that API is mentioned as deprecated,
we should do the actual deprecation by contacting
projects using it.

Author: stsewd

readthedocs/embed/tests/data/mkdocs/latest/index.json

@@ -281,14 +281,8 @@ def test_embed_sphinx(self, storage_mock, section, client):
         assert response.data == expected
         assert response['Cache-tag'] == 'project,project:latest'
 
-    @mock.patch('readthedocs.embed.views.build_media_storage')
-    def test_embed_mkdocs(self, storage_mock, client):
-        json_file = data_path / 'mkdocs/latest/index.json'
-        storage_mock.exists.return_value = True
-        storage_mock.open.side_effect = self._mock_open(
-            json_file.open().read()
-        )
-
+    def test_embed_mkdocs(self, client):
+        """API v2 doesn't support mkdocs."""
         self.version.documentation_type = MKDOCS
         self.version.save()
 
@@ -303,31 +297,7 @@ def test_embed_mkdocs(self, storage_mock, client):
             }
         )
 
-        expected = {
-            'content': mock.ANY,  # too long to compare here
-            'headers': [
-                {'Overview': 'overview'},
-                {'Installation': 'installation'},
-                {'Getting Started': 'getting-started'},
-                {'Adding pages': 'adding-pages'},
-                {'Theming our documentation': 'theming-our-documentation'},
-                {'Changing the Favicon Icon': 'changing-the-favicon-icon'},
-                {'Building the site': 'building-the-site'},
-                {'Other Commands and Options': 'other-commands-and-options'},
-                {'Deploying': 'deploying'},
-                {'Getting help': 'getting-help'},
-            ],
-            'url': 'http://project.readthedocs.io/en/latest/index.html',
-            'meta': {
-                'project': 'project',
-                'version': 'latest',
-                'doc': 'index',
-                'section': 'Installation',
-            },
-        }
-
-        assert response.status_code == status.HTTP_200_OK
-        assert response.data == expected
+        assert response.status_code == status.HTTP_404_NOT_FOUND
 
     def test_no_access(self, client, settings):
         settings.RTD_DEFAULT_FEATURES = {}

readthedocs/embed/tests/test_api.py

@@ -145,6 +145,7 @@ def do_embed(*, project, version, doc=None, path=None, section=None, url=None):
 
     content = None
     headers = None
+    # Embed API v2 supports Sphinx only.
     if version.is_sphinx_type:
         file_content = _get_doc_content(
             project=project,
@@ -160,18 +161,7 @@ def do_embed(*, project, version, doc=None, path=None, section=None, url=None):
             url=url,
         )
     else:
-        # TODO: this should read from the html file itself,
-        # we don't have fjson files for mkdocs.
-        file_content = _get_doc_content(
-            project=project,
-            version=version,
-            doc=doc,
-        )
-        content, headers, section = parse_mkdocs(
-            content=file_content,
-            section=section,
-            url=url,
-        )
+        return None
 
     if content is None:
         return None
@@ -310,47 +300,3 @@ def dump(obj):
 
     ret = [dump(clean_references(obj, url)) for obj in query_result]
     return ret, headers, section
-
-
-def parse_mkdocs(content, section, url):  # pylint: disable=unused-argument
-    """Get the embed content for the section."""
-    ret = []
-    headers = []
-
-    if not content or not content.get('content'):
-        return (None, None, section)
-
-    body = content['content']
-    for element in PQ(body)('h2'):
-        headers.append(recurse_while_none(element))
-
-    if not section and headers:
-        # If no section is sent, return the content of the first one
-        section = list(headers[0].keys())[0].lower()
-
-    if section:
-        body_obj = PQ(body)
-        escaped_section = escape_selector(section)
-        section_list = body_obj(
-            ':header:contains("{title}")'.format(title=str(escaped_section)))
-        for num in range(len(section_list)):
-            header2 = section_list.eq(num)
-            # h2_title = h2.text().strip()
-            # section_id = h2.attr('id')
-            h2_content = ""
-            next_p = header2.next()
-            while next_p:
-                if next_p[0].tag == 'h2':
-                    break
-                h2_html = next_p.outerHtml()
-                if h2_html:
-                    h2_content += "\n%s\n" % h2_html
-                next_p = next_p.next()
-            if h2_content:
-                ret.append(h2_content)
-                # ret.append({
-                #     'id': section_id,
-                #     'title': h2_title,
-                #     'content': h2_content,
-                # })
-    return (ret, headers, section)