[chiragbitoni]

Purpose / Goal

Under the W3C XML 1.0 Recommendation (§2.4 Character Data and Markup):

[14] CharData ::= [^<&]* - ([^<&]* ']]>' [^<&]*)

"The right angle bracket (>) MUST be escaped using '>' or a character reference when it appears in the string ']]>' in content, when that string is not marking the end of a CDATA section."

Previously, XMLValidator.validate() and XMLParser with validation enabled allowed raw ]]> in element text content (such as <a>]]></a> or <a><![CDATA[]]>]]></a>) without reporting an error.

This PR adds a validation check in src/validator.js while reading tag text content to ensure that any literal ]]> sequence outside of a CDATA section is rejected with an InvalidChar error, while valid escaped controls (<a>]]&gt;</a>), CDATA blocks (<a><![CDATA[]]></a>), and attributes (<a attr=']]>'></a>) continue to be accepted.

Fixes #901.

Type

• Bug Fix
• Refactoring / Technology upgrade
• New Feature

[yxshee]

Checked the branch locally (validator spec: 66 specs, 0 failures) and compared these inputs with libxml2 (xmllint --noout), which agrees on all of them: <r>]]>x</r>, <r><![CDATA[x]]>]]></r> and <r>x]]> are rejected, <r>]]&gt;</r>, <r>] ]></r>, and ]]> inside a comment, a PI or an attribute value stay valid. The check sits in the character-data branch, so CDATA sections, comments and PIs are unaffected.

One small ordering note: for text after the root element, e.g. <r></r>]]>, the new branch runs before the reachedRoot check, so the error is InvalidChar ("char ']]>' is not expected.") rather than InvalidXml ("Extra text at the end"), which is what <r></r>x or <r></r>]] report (libxml2 reports all of them as extra content at the end). Moving the ]]> check below the reachedRoot branch would keep that message consistent; either way the document is rejected, so this is only a nit.