Escape HTML entities in doc comments (#10252)

* Escape HTML entities in doc comments

* Put a HTML entity out of backticks
because then at least typical Markdown parsers agree with Crystal's parser.

src/html.cr

@@ -81,9 +81,9 @@ module HTML
   }
 
   # Returns a string where named and numeric character references
-  # (e.g. >, >, &x3e;) in *string* are replaced with the corresponding
+  # (e.g. >, >, >) in *string* are replaced with the corresponding
   # unicode characters. This method decodes all HTML5 entities including those
-  # without a trailing semicolon (such as `&copy`).
+  # without a trailing semicolon (such as "&copy").
   #
   # ```
   # require "html"