canvas_link_migrator 1.0.15 → 1.0.17

checksums.yaml

@@ -1,7 +1,7 @@
 ---
 SHA256:
-  metadata.gz: 36b5ebb39673be4a7105bc459809c1faa333899c9d709d665cabd00f4242401d
-  data.tar.gz: 6702833e94aa9a7fd3bb3265199c2b7caf269e30fd0c82b5c15028d810d44c14
+  metadata.gz: '08e4bb1b132305d9655e4f426f2685e6c524dff6f2e316ee4c32c4898e330e98'
+  data.tar.gz: b000b72459dd32746f8aa31d662df81e180aaff36fd461a88cd6c4d17bfed661
 SHA512:
-  metadata.gz: e8a5d21a3c0a74e5805163d724e42c64a0db4a6d2b18f85f637e401f6ca09089aab287e8afb88b22150cd7eadf90e51a1ba3e78d9bb7892968bbf57055554bfe
-  data.tar.gz: 1fd447f0548f8c84a2fdea0a6a791679eea31c6f4dede5502d4191dcf96b994d4072b303a6ba713151bc03aaddefe6ad003e556ffd4ce39880f8ce792a216522
+  metadata.gz: e731f7af1980e3a1b731b138ba551b921776834f4fe237a005ecf737ddd270ad286d508482c3dd3050b0414c758f2b707d2084fa97c1714a697660e2f1e5dbff
+  data.tar.gz: 9a2100127f1aaf994360e18d6e0370cd67b60239caf9e1fab451bddfc1ae05a885a78abee63cf9d5f0c952b5add8789bf054f97667e0173617cdb7863009e3e7

data/lib/canvas_link_migrator/imported_html_converter.rb

@@ -32,6 +32,18 @@ module CanvasLinkMigrator
     delegate :convert, to: :link_parser
     delegate :resolver_links!, to: :link_resolver
 
+    def convert_single_link(single_link, link_type: false)
+      url = single_link.dup
+      LinkParser::REFERENCE_KEYWORDS.each { |ref| url.gsub!("%24#{ref}%24", "$#{ref}$") }
+      # create the link map for a single link (parse)
+      link_parsing_result = link_parser.parse_single_url(url, link_type)
+      link_parser.handle_parsed_url(url, link_parsing_result, nil, nil, link_type, nil, nil)
+      # resolve_link! on the single element link map
+      link_resolver.resolve_link!(link_parsing_result, link_type)
+      # return the new value
+      link_parsing_result[:new_value]
+    end
+
     def convert_exported_html(input_html)
       new_html = link_parser.convert(input_html, "type", "lookup_id", "field")
       replace!(new_html)

data/lib/canvas_link_migrator/link_parser.rb

@@ -81,7 +81,7 @@ module CanvasLinkMigrator
 
     def convert(html, item_type, mig_id, field, remove_outer_nodes_if_one_child: nil)
       mig_id = mig_id.to_s
-      doc = Nokogiri::HTML5.fragment(html || "")
+      doc = Nokogiri::HTML5.fragment(html || "", max_tree_depth: 10_000)
 
       # Replace source tags with iframes
       doc.search("source[data-media-type],source[data-media-id]").each do |source|
@@ -144,50 +144,63 @@ module CanvasLinkMigrator
       end
 
       result = parse_url(url, node, attr)
+      handle_parsed_url(url, result, node, attr, item_type, mig_id, field)
+    end
+
+    def handle_parsed_url(url, result, node, attr, item_type, mig_id, field)
       if result[:resolved]
         # resolved, just replace and carry on
-        new_url = result[:new_url] || url
-        unless CanvasLinkMigrator.relative_url?(new_url)
-          # perform configured substitutions
-          if (processed_url = @migration_query_service.process_domain_substitutions(new_url))
-            new_url = processed_url
-          end
-          # relative-ize absolute links outside the course but inside our domain
-          # (analogous to what is done in Api#process_incoming_html_content)
-          begin
-            uri = URI.parse(new_url)
-            account_hosts = @migration_query_service.context_hosts.map { |h| h.split(":").first }
-            if account_hosts.include?(uri.host)
-              uri.scheme = uri.host = uri.port = nil
-              new_url = uri.to_s
-            end
-          rescue URI::InvalidURIError, URI::InvalidComponentError
-            nil
-          end
-        end
-        node[attr] = new_url
+        handle_resolved_link(url, result, node, attr)
       else
-        result.delete(:resolved)
-        if result[:link_type] == :media_object
-          # because we may actually change the media comment node itself
-          # (rather than just replacing a value), we're going to
-          # replace the entire node with a placeholder
-          result[:old_value] = node.to_xml
-          result[:placeholder] = placeholder(result[:old_value])
-          placeholder_node = Nokogiri::HTML5.fragment(result[:placeholder])
+        handle_unresolved_link(url, result, node, attr, item_type, mig_id, field)
+      end
+    end
 
-          node.replace(placeholder_node)
-        else
-          result[:old_value] = node[attr]
-          result[:placeholder] = placeholder(result[:old_value])
-          # replace the inner html of an anchor tag if it matches the href
-          if node.name == "a" && attr == "href" && node["href"] == node.inner_html.delete("\n").strip
-            node.inner_html = result[:placeholder]
+    def handle_unresolved_link(url, result, node, attr, item_type, mig_id, field)
+      result.delete(:resolved)
+      if result[:link_type] == :media_object
+        # because we may actually change the media comment node itself
+        # (rather than just replacing a value), we're going to
+        # replace the entire node with a placeholder
+        result[:old_value] = node ? node.to_xml : result[:rel_path]
+
+        result[:placeholder] = placeholder(result[:old_value])
+        placeholder_node = Nokogiri::HTML5.fragment(result[:placeholder])
+
+        node.replace(placeholder_node) if node
+      else
+        result[:old_value] = node ? node[attr] : url
+        result[:placeholder] = placeholder(result[:old_value])
+        # replace the inner html of an anchor tag if it matches the href
+        if node && node.name == "a" && attr == "href" && node["href"] == node.inner_html.delete("\n").strip
+          node.inner_html = result[:placeholder]
+        end
+        node[attr] = result[:placeholder] if node
+      end
+      add_unresolved_link(result, item_type, mig_id, field)
+    end
+
+    def handle_resolved_link(url, result, node, attr)
+      new_url = result[:new_url] || url
+      unless CanvasLinkMigrator.relative_url?(new_url)
+        # perform configured substitutions
+        if (processed_url = @migration_query_service.process_domain_substitutions(new_url))
+          new_url = processed_url
+        end
+        # relative-ize absolute links outside the course but inside our domain
+        # (analogous to what is done in Api#process_incoming_html_content)
+        begin
+          uri = URI.parse(new_url)
+          account_hosts = @migration_query_service.context_hosts.map { |h| h.split(":").first }
+          if account_hosts.include?(uri.host)
+            uri.scheme = uri.host = uri.port = nil
+            new_url = uri.to_s
           end
-          node[attr] = result[:placeholder]
+        rescue URI::InvalidURIError, URI::InvalidComponentError
+          nil
         end
-        add_unresolved_link(result, item_type, mig_id, field)
       end
+      node[attr] = new_url
     end
 
     def unresolved(type, data = {})
@@ -202,8 +215,11 @@ module CanvasLinkMigrator
       "?type=#{type}&embedded=true"
     end
 
+    def parse_single_url(url, link_type = nil)
+      parse_url(url, nil, nil, link_type)
+    end
     # returns a hash with resolution status and data to hold onto if unresolved
-    def parse_url(url, node, attr)
+    def parse_url(url, node, attr, link_type = nil)
       if url =~ /wiki_page_migration_id=(.*)/
         unresolved(:wiki_page, migration_id: $1)
       elsif url =~ /discussion_topic_migration_id=(.*)/
@@ -211,13 +227,13 @@ module CanvasLinkMigrator
       elsif url =~ %r{\$CANVAS_COURSE_REFERENCE\$/modules/items/([^?]*)(\?.*)?}
         unresolved(:module_item, migration_id: $1, query: $2)
       elsif url =~ %r{\$CANVAS_COURSE_REFERENCE\$/file_ref/([^/?#]+)(.*)}
-        in_media_iframe = (attr == "src" && ["iframe", "source"].include?(node.name) && (node["data-media-id"] || node["data-media-type"]))
-        rest = in_media_iframe ? media_params(node["data-media-type"]) : $2
+        in_media_iframe = node && (attr == "src" && %w[iframe source].include?(node.name) && (node["data-media-id"] || node["data-media-type"]))
+        rest = (in_media_iframe && node) ? media_params(node["data-media-type"]) : $2
         unresolved(:file_ref,
                    migration_id: $1,
                    rest: rest,
                    in_media_iframe: in_media_iframe,
-                   target_blank: node['target'] == "_blank" && node.name == "a" && attr == "href"
+                   target_blank: node && node['target'] == "_blank" && node.name == "a" && attr == "href"
         )
       elsif url =~ %r{(?:\$CANVAS_OBJECT_REFERENCE\$|\$WIKI_REFERENCE\$)/([^/]*)/([^?]*)(\?.*)?}
         if KNOWN_REFERENCE_TYPES.include?($1)
@@ -231,16 +247,21 @@ module CanvasLinkMigrator
         resolved("#{@migration_query_service.context_path}/#{$1}")
       elsif url =~ %r{\$IMS(?:-|_)CC(?:-|_)FILEBASE\$/(.*)}
         rel_path = URI::DEFAULT_PARSER.unescape($1)
-        if (attr == "href" && node["class"]&.include?("instructure_inline_media_comment")) ||
-           (attr == "src" && ["iframe", "source"].include?(node.name) && (node["data-media-id"] || node["data-media-type"]))
+
+        if (attr == "href" && node && node["class"]&.include?("instructure_inline_media_comment")) ||
+           (attr == "src" && node && %w[iframe source].include?(node.name) && (node["data-media-id"] || node["data-media-type"])) ||
+          link_type == :media_object
           unresolved(:media_object, rel_path: rel_path)
         else
           unresolved(:file, rel_path: rel_path)
         end
-      elsif (attr == "src" && ["iframe", "source"].include?(node.name) && (node["data-media-id"] || node["data-media-type"]))
+      elsif (attr == "src" && node && %w[iframe source].include?(node.name) && (node["data-media-id"] || node["data-media-type"])) ||
+        link_type == :media_object
         # media_objects_iframe course copy reference without an attachment id, change to media_attachments_iframe
-        unresolved(:media_object, rel_path: node["src"])
-      elsif @migration_query_service.supports_embedded_images && attr == "src" && (info_match = url.match(%r{\Adata:(?<mime_type>[-\w]+/[-\w+.]+)?;base64,(?<image>.*)}m))
+        rel_path =  node ? node["src"] : url
+        unresolved(:media_object, rel_path: rel_path)
+      elsif (@migration_query_service.supports_embedded_images && attr == "src" && (info_match = url.match(%r{\Adata:(?<mime_type>[-\w]+/[-\w+.]+)?;base64,(?<image>.*)}m))) ||
+        link_type == :image
         result = @migration_query_service.link_embedded_image(info_match)
         if result[:resolved]
           resolved(result[:url])
@@ -249,7 +270,7 @@ module CanvasLinkMigrator
         end
       elsif # rubocop:disable Lint/DuplicateBranch
             # Equation image, leave it alone
-            (attr == "src" && node["class"] && node["class"].include?("equation_image")) || # rubocop:disable Layout/ConditionPosition
+            (attr == "src" && node && node["class"] && node["class"].include?("equation_image")) || # rubocop:disable Layout/ConditionPosition
             # The file is in the context of an AQ, leave the link alone
             url =~ %r{\A/assessment_questions/\d+/files/\d+} ||
             # This points to a specific file already, leave it alone

data/lib/canvas_link_migrator/link_resolver.rb

@@ -54,7 +54,8 @@ module CanvasLinkMigrator
     end
 
     # finds the :new_value to use to replace the placeholder
-    def resolve_link!(link)
+    def resolve_link!(link, preset_type = false)
+      link[:link_type] = preset_type if preset_type
       case link[:link_type]
       when :wiki_page
         if (linked_wiki_url = @migration_id_converter.convert_wiki_page_migration_id_to_slug(link[:migration_id]))
@@ -103,7 +104,7 @@ module CanvasLinkMigrator
         # this part is a little trickier
         # tl;dr we've replaced the entire node with the placeholder
         # see LinkParser for details
-        rel_path = link[:rel_path]
+        rel_path = link[:rel_path] || link[:old_value]
         node = Nokogiri::HTML5.fragment(link[:old_value]).children.first
         new_url = resolve_media_data(node, rel_path)
         new_url ||= resolve_relative_file_url(rel_path)
@@ -112,14 +113,14 @@ module CanvasLinkMigrator
           new_url = rel_path.include?("#{context_path}/file_contents") ? rel_path : missing_relative_file_url(rel_path)
           link[:missing_url] = new_url
         end
-        if ["iframe", "source"].include?(node.name)
+        if %w[iframe source].include?(node.name)
           node["src"] = new_url
         else
           node["href"] = new_url
         end
-        link[:new_value] = node.to_s
+        link[:new_value] = node.name === 'text' ? new_url : node.to_s
       when :file
-        rel_path = link[:rel_path]
+        rel_path = link[:rel_path] || link[:old_value]
         new_url = resolve_relative_file_url(rel_path)
         # leave user urls alone
         new_url ||= rel_path if is_relative_user_url(rel_path)
@@ -249,10 +250,13 @@ module CanvasLinkMigrator
       elsif rel_path&.match(/\/media_attachments_iframe\/\d+/)
         # media attachment from another course or something
         rel_path
-      elsif (file_id, uuid = @migration_id_converter.convert_attachment_media_id(node["data-media-id"]))
+      elsif node["data-media-id"] && (file_id, uuid = @migration_id_converter.convert_attachment_media_id(node["data-media-id"]))
         file_id ? media_attachment_iframe_url(file_id, uuid, node["data-media-type"]) : nil
-      elsif (file_id, uuid = @migration_id_converter.convert_attachment_media_id(rel_path.match(/media_objects(?:_iframe)?\/([^?.]+)/)&.[](1)))
+      elsif (identifier = rel_path.match(/media_objects(?:_iframe)?\/([^?.]+)/)) && (file_id, uuid = @migration_id_converter.convert_attachment_media_id(identifier.[](1)))
         file_id ? media_attachment_iframe_url(file_id, uuid, node["data-media-type"]) : nil
+      elsif node.name == "text" && rel_path[/^[^?]+/] && (mig_id = rel_path[/^[^?]+/].match(/[^\/]+$/)[0])
+        file = @migration_id_converter.lookup_attachment_by_migration_id(mig_id)
+        media_attachment_iframe_url(file["id"], file["uuid"])
       else
         node.delete("class")
         node.delete("id")

data/lib/canvas_link_migrator/version.rb

@@ -1,3 +1,3 @@
 module CanvasLinkMigrator
-  VERSION = "1.0.15"
+  VERSION = "1.0.17"
 end

data/spec/canvas_link_migrator/imported_html_converter_spec.rb

@@ -403,13 +403,40 @@ describe CanvasLinkMigrator::ImportedHtmlConverter do
 
     it "converts iframe srcs that point to non-media files" do
       test_string = <<~HTML
-      <p><iframe style="width: 100%; height: 100vh; border: none;" src="$IMS-CC-FILEBASE$/subfolder/test.png?canvas_download=1"></iframe></p>
+        <p><iframe style="width: 100%; height: 100vh; border: none;" src="$IMS-CC-FILEBASE$/subfolder/test.png?canvas_download=1"></iframe></p>
       HTML
       converted_string = <<~HTML
-      <p><iframe style="width: 100%; height: 100vh; border: none;" src="/courses/2/files/7/download?verifier=u7"></iframe></p>
+        <p><iframe style="width: 100%; height: 100vh; border: none;" src="/courses/2/files/7/download?verifier=u7"></iframe></p>
       HTML
       html = @converter.convert_exported_html(test_string)
       expect(html[0]).to eq converted_string
     end
   end
+
+  describe ".convert_single_link" do
+    before(:each) do
+      @path = "/courses/2/"
+      @converter = CanvasLinkMigrator::ImportedHtmlConverter.new(resource_map: JSON.parse(File.read("spec/fixtures/canvas_resource_map.json")))
+    end
+
+    it "converts a wiki reference" do
+      test_string = "%24WIKI_REFERENCE%24/wiki/test-wiki-page?query=blah"
+      link = @converter.convert_single_link(test_string)
+      expect(link).to eq "#{@path}pages/test-wiki-page?query=blah"
+    end
+
+    it "converts a wiki reference with migration id" do
+      test_string = "%24WIKI_REFERENCE%24/pages/A?query=blah"
+      link = @converter.convert_single_link(test_string)
+      expect(link).to eq "#{@path}pages/slug-a?query=blah"
+    end
+
+    it "formats media properly if explicitly set to do so" do
+      test_string = "$IMS-CC-FILEBASE$/subfolder/with a space/yodawg.mov?canvas_=1&canvas_qs_type=video&canvas_qs_amp=&canvas_qs_embedded=true&media_attachment=true"
+      link = @converter.convert_single_link(test_string, link_type: :media_object)
+      expect(link).to eq "/media_attachments_iframe/9?embedded=true&verifier=u9"
+      link = @converter.convert_single_link(test_string)
+      expect(link).to eq "/courses/2/files/9?verifier=u9&type=video&amp=&embedded=true"
+    end
+  end
 end

data/spec/canvas_link_migrator/link_parser_spec.rb

@@ -69,5 +69,19 @@ describe CanvasLinkMigrator::LinkParser do
       doc = Nokogiri::HTML5.fragment(%Q(<a id="media_comment_m-4uoGqVdEqXhpqu2ZMytHSy9XMV73aQ7E" class="instructure_inline_media_comment" data-media_comment_type="video" data-alt=""></a>))
       expect{ parser.convert(doc.to_html, "type", "lookup_id", "field") }.not_to raise_error
     end
+
+    it "handles deeply nested html up to 10.000 levels" do
+      deeply_nested_html = "<div>" * 9999
+      deeply_nested_html += "<a target=\"_blank\"></a>"
+      deeply_nested_html += "</div>" * 9999
+      expect{ parser.convert(deeply_nested_html, "type", "lookup_id", "field") }.not_to raise_error
+    end
+
+    it "raises error when html is beyond 10.000 depth" do
+      deeply_nested_html = "<div>" * 10_000
+      deeply_nested_html += "<a target=\"_blank\"></a>"
+      deeply_nested_html += "</div>" * 10_000
+      expect{ parser.convert(deeply_nested_html, "type", "lookup_id", "field") }.to raise_error("Document tree depth limit exceeded")
+    end
   end
 end

metadata

@@ -1,7 +1,7 @@
 --- !ruby/object:Gem::Specification
 name: canvas_link_migrator
 version: !ruby/object:Gem::Version
-  version: 1.0.15
+  version: 1.0.17
 platform: ruby
 authors:
 - Mysti Lilla
@@ -11,7 +11,7 @@ authors:
 autorequire: 
 bindir: bin
 cert_chain: []
-date: 2024-10-24 00:00:00.000000000 Z
+date: 2025-03-05 00:00:00.000000000 Z
 dependencies:
 - !ruby/object:Gem::Dependency
   name: activesupport