support early closing of cdata sections

HtmlUnit · Dec 23, 2023 · 49a31c0 · 49a31c0
1 parent b3feeca
commit 49a31c0
Show file tree

Hide file tree

Showing 3 changed files with 11 additions and 6 deletions.
diff --git a/src/main/java/org/htmlunit/cyberneko/HTMLScanner.java b/src/main/java/org/htmlunit/cyberneko/HTMLScanner.java
@@ -2271,9 +2271,7 @@ protected void scanCDATA() throws IOException {
                 fStringBuffer.append("[CDATA[");
             }
             final boolean eof = scanCDataContent(fStringBuffer);
-            if (!fCDATASections_) {
-                fStringBuffer.append("]]");
-            }
+
             if (fDocumentHandler != null && fElementCount >= fElementDepth) {
                 fEndLineNumber = fCurrentEntity.getLineNumber();
                 fEndColumnNumber = fCurrentEntity.getColumnNumber();
@@ -2453,7 +2451,7 @@ protected boolean scanCDataContent(final XMLString xmlString) throws IOException
                     if (fReportErrors_) {
                         fErrorReporter.reportError("HTML1007", null);
                     }
-                    return true;
+                    break;
                 }
 
                 if (c == ']') {
@@ -2489,6 +2487,10 @@ protected boolean scanCDataContent(final XMLString xmlString) throws IOException
                     }
                     break;
                 }
+                else if (c == '>') {
+                    // don't add the ]] to the buffer
+                    return false;
+                }
                 else if (c == '\n' || c == '\r') {
                     fCurrentEntity.rewind();
                     final int newlines = skipNewlines();
@@ -2500,6 +2502,9 @@ else if (c == '\n' || c == '\r') {
                 appendChar(xmlString, c, null);
             }
 
+            if (!fCDATASections_) {
+                fStringBuffer.append("]]");
+            }
             return c == -1;
         }
 

diff --git a/src/test/resources/org/htmlunit/cyberneko/testfiles/cdata/test-cdata.html b/src/test/resources/org/htmlunit/cyberneko/testfiles/cdata/test-cdata.html
@@ -1,2 +1,2 @@
-<p>P1<![CDATA[<h1>Header</h1>]]>
+<p>P1<![CDATA[Header < &gt;]]>
 <p>P2
diff --git a/src/test/resources/org/htmlunit/cyberneko/testfiles/cdata/test-cdata.html.canonical b/src/test/resources/org/htmlunit/cyberneko/testfiles/cdata/test-cdata.html.canonical
@@ -4,7 +4,7 @@
 (BODY
 (p
 "P1
-#[CDATA[<h1>Header</h1>]]
+#[CDATA[Header < &gt;]]
 "\n
 )p
 (p
-Original file line number
+Diff line change
@@ Expand Up / @@ -4,7 +4,7 @@ @@
     (BODY
     (p
     "P1
-    #[CDATA[<h1>Header</h1>]]
+    #[CDATA[Header < &gt;]]
     "\n
     )p
     (p
@@ Expand Down @@