Correctly parse CDATA sections

gettalong · Nov 18, 2024 · e5a9d64 · e5a9d64
1 parent b970b40
commit e5a9d64
Show file tree

Hide file tree

Showing 5 changed files with 34 additions and 5 deletions.
diff --git a/lib/kramdown/converter/kramdown.rb b/lib/kramdown/converter/kramdown.rb
@@ -75,14 +75,17 @@ def convert_blank(_el, _opts)
       ESCAPED_CHAR_RE = /(\$\$|[\\*_`\[\]{"'|])|^ {0,3}(:)/
 
       def convert_text(el, opts)
-        if opts[:raw_text]
+        if opts[:raw_text] || (@stack.last.type == :html_element && @stack.last.options[:content_model] == :raw)
           el.value
         else
-          el.value.gsub(/\A\n/) do
+          result = el.value.gsub(/\A\n/) do
             opts[:prev] && opts[:prev].type == :br ? '' : "\n"
-          end.gsub(/\s+/, ' ').gsub(ESCAPED_CHAR_RE) do
+          end
+          result.gsub!(/\s+/, ' ') unless el.options[:cdata]
+          result.gsub!(ESCAPED_CHAR_RE) do
             $1 || !opts[:prev] || opts[:prev].type == :br ? "\\#{$1 || $2}" : $&
           end
+          result
         end
       end
 

diff --git a/lib/kramdown/parser/html.rb b/lib/kramdown/parser/html.rb
@@ -29,6 +29,7 @@ module Constants
         HTML_DOCTYPE_RE = /<!DOCTYPE.*?>/im
         HTML_COMMENT_RE = /<!--(.*?)-->/m
         HTML_INSTRUCTION_RE = /<\?(.*?)\?>/m
+        HTML_CDATA_RE = /<!\[CDATA\[(.*?)\]\]>/m
         HTML_ATTRIBUTE_RE = /\s*(#{REXML::Parsers::BaseParser::UNAME_STR})(?:\s*=\s*(?:(\p{Word}+)|("|')(.*?)\3))?/m
         HTML_TAG_RE = /<((?>#{REXML::Parsers::BaseParser::UNAME_STR}))\s*((?>\s+#{REXML::Parsers::BaseParser::UNAME_STR}(?:\s*=\s*(?:\p{Word}+|("|').*?\3))?)*)\s*(\/)?>/m
         HTML_TAG_CLOSE_RE = /<\/(#{REXML::Parsers::BaseParser::UNAME_STR})\s*>/m
@@ -136,7 +137,7 @@ def handle_raw_html_tag(name)
           end
         end
 
-        HTML_RAW_START = /(?=<(#{REXML::Parsers::BaseParser::UNAME_STR}|\/|!--|\?))/ # :nodoc:
+        HTML_RAW_START = /(?=<(#{REXML::Parsers::BaseParser::UNAME_STR}|\/|!--|\?|!\[CDATA\[))/ # :nodoc:
 
         # Parse raw HTML from the current source position, storing the found elements in +el+.
         # Parsing continues until one of the following criteria are fulfilled:
@@ -160,6 +161,8 @@ def parse_raw_html(el, &block)
                 @tree.children << Element.new(:xml_comment, result, nil, category: :block, location: line)
               elsif (result = @src.scan(HTML_INSTRUCTION_RE))
                 @tree.children << Element.new(:xml_pi, result, nil, category: :block, location: line)
+              elsif @src.scan(HTML_CDATA_RE)
+                @tree.children << Element.new(:text, @src[1], nil, cdata: true, location: line)
               elsif @src.scan(HTML_TAG_RE)
                 if method(:handle_html_start_tag).arity.abs >= 1
                   handle_html_start_tag(line, &block)

diff --git a/lib/kramdown/parser/kramdown/html.rb b/lib/kramdown/parser/kramdown/html.rb
@@ -14,6 +14,7 @@ module Parser
     class Kramdown
 
       include Kramdown::Parser::Html::Parser
+      include Kramdown::Utils::Html
 
       # Mapping of markdown attribute value to content model. I.e. :raw when "0", :default when "1"
       # (use default content model for the HTML element), :span when "span", :block when block and
@@ -95,13 +96,15 @@ def parse_block_html
       end
       define_parser(:block_html, HTML_BLOCK_START)
 
-      HTML_SPAN_START = /<(#{REXML::Parsers::BaseParser::UNAME_STR}|!--|\/)/
+      HTML_SPAN_START = /<(#{REXML::Parsers::BaseParser::UNAME_STR}|!--|\/|!\[CDATA\[)/
 
       # Parse the HTML at the current position as span-level HTML.
       def parse_span_html
         line = @src.current_line_number
         if (result = @src.scan(HTML_COMMENT_RE))
           @tree.children << Element.new(:xml_comment, result, nil, category: :span, location: line)
+        elsif @src.scan(HTML_CDATA_RE)
+          add_text(escape_html(@src[1]))
         elsif (result = @src.scan(HTML_TAG_CLOSE_RE))
           warning("Found invalidly used HTML closing tag for '#{@src[1]}' on line #{line}")
           add_text(result)

diff --git a/test/testcases/block/09_html/cdata_section.html b/test/testcases/block/09_html/cdata_section.html
@@ -0,0 +1,10 @@
+<figure>
++-----------------------+
+| Use XML, be Happy :-) |
+|_______________________|
+
+</figure>
+
+<p>Inline section &lt;goes&gt; here.</p>
+
+<p>Inline <kbd>test *case* *section* &lt;goes&gt; here</kbd> end.</p>
diff --git a/test/testcases/block/09_html/cdata_section.text b/test/testcases/block/09_html/cdata_section.text
@@ -0,0 +1,10 @@
+<figure><![CDATA[
++-----------------------+
+| Use XML, be Happy :-) |
+|_______________________|
+     ]]>
+</figure>
+
+Inline <![CDATA[section <goes>]]> here.
+
+Inline <kbd>test *case* <![CDATA[*section* <goes>]]> here</kbd> end.