parse wiki pages in migrations with HTML5

this matches the version used by Canvas proper when it
sanitizes user HTML, to prevent surprises in migrations

test plan:
 - put the HTML attached to the ticket in a page
 - it should survive an export/import or course copy round trip

flag = none
fixes LS-2241

Change-Id: Iae07c73e3b994dddf814ad5ee7dc4c2bd0284559
Reviewed-on: https://gerrit.instructure.com/c/canvas-lms/+/267889
Tested-by: Service Cloud Jenkins <svc.cloudjenkins@instructure.com>
Reviewed-by: Nate Armstrong <narmstrong@instructure.com>
QA-Review: Nate Armstrong <narmstrong@instructure.com>
Product-Review: Jeremy Stanley <jeremy@instructure.com>

lib/canvas/migration/xml_helper.rb

@@ -37,7 +37,7 @@ module XMLHelper
   def get_node_val(node, selector, default=nil)
     node.at_css(selector) ? node.at_css(selector).text : default
   end
-  
+
   # You can't do a css selector that only looks for direct
   # descendants of the current node, so you have to iterate
   # over the children and see if it's there.
@@ -88,13 +88,17 @@ module XMLHelper
     path = path.gsub('\\', '/') if path
     path
   end
-  
+
   def open_file(path)
     # this is cheating; we use HTML because we don't deal properly with namespaces
     # even though it's really XML
     File.exist?(path) ? ::Nokogiri::HTML(File.open(path)) : nil
   end
 
+  def open_file_html5(path)
+    File.exist?(path) ? ::Nokogiri::HTML5(File.open(path)) : nil
+  end
+
   def open_file_xml(path)
     File.exist?(path) ? create_xml_doc(File.open(path)) : nil
   end
@@ -116,4 +120,4 @@ module XMLHelper
     doc
   end
 end
-end
+end

lib/cc/importer/canvas/wiki_converter.rb

@@ -27,7 +27,7 @@ module CC::Importer::Canvas
       wiki_dir = @package_root.item_path(WIKI_FOLDER)
       Dir["#{wiki_dir}/**/**"].each do |path|
         next if File.directory?(path)
-        doc = open_file(path)
+        doc = open_file_html5(path)
         wikis << convert_wiki(doc, path)
       end
 

spec/models/content_migration/course_copy_wiki_spec.rb

@@ -24,7 +24,7 @@ describe ContentMigration do
     include_examples "course copy"
 
     it "should copy wiki page attributes" do
-      page = @copy_from.wiki_pages.create!(:title => "title", :body => "<p>butter</p>",
+      page = @copy_from.wiki_pages.create!(:title => "title", :body => "<address><ul></ul></address>",
                                                 :editing_roles => "teachers", :todo_date => Time.zone.now)
 
       run_course_copy
@@ -33,6 +33,7 @@ describe ContentMigration do
 
       attrs = [:title, :body, :editing_roles, :todo_date]
       expect(page.attributes.slice(*attrs)).to eq page_to.attributes.slice(*attrs)
+      expect(page_to.body.strip).to eq "<address><ul></ul></address>"
     end
 
     it "should reset user on re-import" do