Use Nokogiri::HTML5::Inference for parsing HTML fragments

Gemfile.lock

@@ -21,6 +21,7 @@ PATH
       deface (~> 1.9)
       html_press (~> 0.8.2)
       nokogiri (~> 1.0)
+      nokogiri-html5-inference (~> 0.2)
       phlex (~> 1.6)
       phlex-rails (>= 0.9, < 2.0)
       syntax_tree (~> 6.0)
@@ -180,6 +181,8 @@ GEM
       racc (~> 1.4)
     nokogiri (1.16.4-x86_64-linux)
       racc (~> 1.4)
+    nokogiri-html5-inference (0.2.0)
+      nokogiri (~> 1.14)
     parallel (1.22.1)
     parser (3.1.2.1)
       ast (~> 2.4.1)

gem/Gemfile

@@ -8,3 +8,6 @@ gem "maxitest", "~> 4.4"
 gem "minitest", "~> 5.0"
 gem "rake", "~> 13.0"
 gem "rubocop", require: false, github: "joeldrapper/rubocop", branch: "rubocop-user-agent"
+# gem "nokogiri-html5-inference", github: "flavorjones/nokogiri-html5-inference", branch: "2-use-template-context-node"
+
+gem "nokogiri-html5-inference", github: "flavorjones/nokogiri-html5-inference", branch: "flavorjones-always-return-nodeset"

gem/Gemfile.lock

@@ -1,3 +1,11 @@
+GIT
+  remote: https://github.com/flavorjones/nokogiri-html5-inference.git
+  revision: 7426a3601d30bf6cb9794d6b5c2e767151a0fa6f
+  branch: flavorjones-always-return-nodeset
+  specs:
+    nokogiri-html5-inference (0.2.0)
+      nokogiri (~> 1.14)
+
 GIT
   remote: https://github.com/joeldrapper/rubocop.git
   revision: 8d3c1a7dd04a0dee701e26ee506b28af0ffe3b98
@@ -21,6 +29,7 @@ PATH
       deface (~> 1.9)
       html_press (~> 0.8.2)
       nokogiri (~> 1.0)
+      nokogiri-html5-inference (~> 0.2)
       phlex (~> 1.6)
       phlex-rails (>= 0.9, < 2.0)
       syntax_tree (~> 6.0)
@@ -178,6 +187,7 @@ PLATFORMS
 DEPENDENCIES
   maxitest (~> 4.4)
   minitest (~> 5.0)
+  nokogiri-html5-inference!
   phlexing!
   rake (~> 13.0)
   rubocop!

gem/lib/phlexing/name_suggestor.rb

@@ -33,7 +33,7 @@ def self.wrap(name)
     def self.extract(document, method)
       return [] unless document
 
-      document.children.map { |element| send(method, element) }.compact
+      document.map { |element| send(method, element) }.compact
     end
 
     def self.extract_id_from_element(element)

gem/lib/phlexing/parser.rb

@@ -1,32 +1,15 @@
 # frozen_string_literal: true
 
 require "nokogiri"
+require "nokogiri/html5/inference"
 
 module Phlexing
   class Parser
     def self.call(source)
       source = ERBTransformer.call(source)
       source = Minifier.call(source)
 
-      # Credit:
-      # https://github.com/spree/deface/blob/6bf18df76715ee3eb3d0cd1b6eda822817ace91c/lib/deface/parser.rb#L105-L111
-      #
-
-      html_tag = /<html(( .*?(?:(?!>)[\s\S])*>)|>)/i
-      head_tag = /<head(( .*?(?:(?!>)[\s\S])*>)|>)/i
-      body_tag = /<body(( .*?(?:(?!>)[\s\S])*>)|>)/i
-
-      if source =~ html_tag
-        Nokogiri::HTML::Document.parse(source)
-      elsif source =~ head_tag && source =~ body_tag
-        Nokogiri::HTML::Document.parse(source).css("html").first
-      elsif source =~ head_tag
-        Nokogiri::HTML::Document.parse(source).css("head").first
-      elsif source =~ body_tag
-        Nokogiri::HTML::Document.parse(source).css("body").first
-      else
-        Nokogiri::HTML::DocumentFragment.parse(source)
-      end
+      Nokogiri::HTML5::Inference.parse(source)
     end
   end
 end

gem/lib/phlexing/template_generator.rb

@@ -231,6 +231,8 @@ def handle_node(node, level = 0)
         handle_document_node(node, level)
       in Nokogiri::XML::Comment
         handle_html_comment_node(node)
+      in Nokogiri::XML::NodeSet
+        node.each { |n| handle_node(n, level + 1) }
       end
     end
   end

gem/phlexing.gemspec

@@ -32,6 +32,7 @@ Gem::Specification.new do |spec|
   spec.add_dependency "deface", "~> 1.9"
   spec.add_dependency "html_press", "~> 0.8.2"
   spec.add_dependency "nokogiri", "~> 1.0"
+  spec.add_dependency "nokogiri-html5-inference", "~> 0.2"
   spec.add_dependency "phlex", "~> 1.6"
   spec.add_dependency "phlex-rails", ">= 0.9", "< 2.0"
   spec.add_dependency "syntax_tree", "~> 6.0"

gem/test/phlexing/converter/tags_test.rb

@@ -8,10 +8,15 @@ class Phlexing::Converter::TagsTest < Minitest::Spec
     assert_phlex_template "span", %(<span></span>)
     assert_phlex_template "p", %(<p></p>)
     assert_phlex_template "template_tag", %(<template></template>)
-    assert_phlex_template "html", %(<html></html>)
-    assert_phlex_template "head", %(<head></head>)
+    assert_phlex_template "head\n\nbody", %(<head></head>)
     assert_phlex_template "header", %(<header></header>)
     assert_phlex_template "body", %(<body></body>)
+    assert_phlex_template <<~PHLEX.strip, %(<html></html>)
+      html do
+        head
+        body
+      end
+    PHLEX
   end
 
   it "basic self closing tag" do
@@ -133,11 +138,9 @@ class Phlexing::Converter::TagsTest < Minitest::Spec
     HTML
 
     expected = <<~PHLEX.strip
-      html do
-        head
+      head
 
-        body
-      end
+      body
     PHLEX
 
     assert_phlex_template expected, html
@@ -150,11 +153,9 @@ class Phlexing::Converter::TagsTest < Minitest::Spec
     HTML
 
     expected = <<~PHLEX.strip
-      html do
-        head
+      head
 
-        body
-      end
+      body
     PHLEX
 
     assert_phlex_template expected, html
@@ -167,6 +168,8 @@ class Phlexing::Converter::TagsTest < Minitest::Spec
 
     expected = <<~PHLEX.strip
       head
+
+      body
     PHLEX
 
     assert_phlex_template expected, html
@@ -179,6 +182,8 @@ class Phlexing::Converter::TagsTest < Minitest::Spec
 
     expected = <<~PHLEX.strip
       head(id: "123")
+
+      body
     PHLEX
 
     assert_phlex_template expected, html

gem/test/phlexing/converter/uppercase_tags_test.rb

@@ -8,9 +8,14 @@ class Phlexing::Converter::UppercaseTagsTest < Minitest::Spec
     assert_phlex_template "span", %(<SPAN></SPAN>)
     assert_phlex_template "p", %(<P></P>)
     assert_phlex_template "template_tag", %(<TEMPLATE></TEMPLATE>)
-    assert_phlex_template "html", %(<HTML></HTML>)
-    assert_phlex_template "head", %(<HEAD></HEAD>)
+    assert_phlex_template "head\n\nbody", %(<HEAD></HEAD>)
     assert_phlex_template "body", %(<BODY></BODY>)
+    assert_phlex_template <<~PHLEX.strip, %(<HTML></HTML>)
+      html do
+        head
+        body
+      end
+    PHLEX
   end
 
   it "standlone uppercase body tag" do
@@ -32,6 +37,8 @@ class Phlexing::Converter::UppercaseTagsTest < Minitest::Spec
 
     expected = <<~PHLEX.strip
       head
+
+      body
     PHLEX
 
     assert_phlex_template expected, html
@@ -43,7 +50,10 @@ class Phlexing::Converter::UppercaseTagsTest < Minitest::Spec
     HTML
 
     expected = <<~PHLEX.strip
-      html
+      html do
+        head
+        body
+      end
     PHLEX
 
     assert_phlex_template expected, html
@@ -56,11 +66,8 @@ class Phlexing::Converter::UppercaseTagsTest < Minitest::Spec
     HTML
 
     expected = <<~PHLEX.strip
-      html do
-        head
-
-        body
-      end
+      head
+      body
     PHLEX
 
     assert_phlex_template expected, html
@@ -76,7 +83,6 @@ class Phlexing::Converter::UppercaseTagsTest < Minitest::Spec
 
     expected = <<~PHLEX.strip
       html do
-        whitespace
         head
         whitespace
         body