disable early cdata closing for xhtml

HtmlUnit · Nov 4, 2024 · f9af348 · f9af348
1 parent 94162af
commit f9af348
Show file tree

Hide file tree

Showing 2 changed files with 98 additions and 0 deletions.
diff --git a/src/main/java/org/htmlunit/html/parser/neko/HtmlUnitNekoHtmlParser.java b/src/main/java/org/htmlunit/html/parser/neko/HtmlUnitNekoHtmlParser.java
@@ -180,6 +180,7 @@ public void parse(final WebResponse webResponse, final HtmlPage page,
                 domBuilder.setFeature(HTMLScanner.ALLOW_SELFCLOSING_TAGS, true);
                 domBuilder.setFeature(HTMLScanner.SCRIPT_STRIP_CDATA_DELIMS, true);
                 domBuilder.setFeature(HTMLScanner.STYLE_STRIP_CDATA_DELIMS, true);
+                domBuilder.setFeature(HTMLScanner.CDATA_EARLY_CLOSING, false);
             }
         }
         catch (final Exception e) {

diff --git a/src/test/java/org/htmlunit/html/XHtmlPage2Test.java b/src/test/java/org/htmlunit/html/XHtmlPage2Test.java
@@ -19,6 +19,7 @@
 import org.htmlunit.WebDriverTestCase;
 import org.htmlunit.junit.BrowserRunner;
 import org.htmlunit.junit.BrowserRunner.Alerts;
+import org.htmlunit.junit.BrowserRunner.HtmlUnitNYI;
 import org.junit.Test;
 import org.junit.runner.RunWith;
 import org.openqa.selenium.By;
@@ -83,4 +84,100 @@ public void selfClosingTextarea() throws Exception {
 
         assertEquals("", driver.findElement(By.id("myText")).getText());
     }
+
+    /**
+     * Regression test for Bug #1219.
+     *
+     * @throws Exception if an error occurs
+     */
+    @Test
+    @Alerts({"before<![CDATA[inside]]>after",
+             "<div xmlns=\"http://www.w3.org/1999/xhtml\" id=\"tester\">before<![CDATA[inside]]>after</div>",
+             "beforeinsideafter"})
+    @HtmlUnitNYI(CHROME = {"before<!--[CDATA[inside]]-->after",
+                           "<div id=\"tester\">before<!--[CDATA[inside]]-->after</div>",
+                           "beforeinsideafter"},
+            EDGE = {"before<!--[CDATA[inside]]-->after",
+                    "<div id=\"tester\">before<!--[CDATA[inside]]-->after</div>",
+                    "beforeinsideafter"},
+            FF = {"before<!--[CDATA[inside]]-->after",
+                  "<div id=\"tester\">before<!--[CDATA[inside]]-->after</div>",
+                  "beforeinsideafter"},
+            FF_ESR = {"before<!--[CDATA[inside]]-->after",
+                      "<div id=\"tester\">before<!--[CDATA[inside]]-->after</div>",
+                      "beforeinsideafter"})
+    public void cdata() throws Exception {
+        final String html
+            = "<?xml version=\"1.0\" encoding=\"UTF-8\"?>\n"
+            + "<!DOCTYPE html PUBLIC \n"
+            + "  \"-//W3C//DTD XHTML 1.0 Transitional//EN\" \n"
+            + "  \"http://www.w3.org/TR/xhtml1/DTD/xhtml1-transitional.dtd\">\n"
+            + "<html xmlns=\"http://www.w3.org/1999/xhtml\">\n"
+            + "<head>"
+            + "  <script>\n"
+            + LOG_TITLE_FUNCTION
+            + "    function test() {"
+            + "      log(document.getElementById('tester').innerHTML);\n"
+            + "      log(document.getElementById('tester').outerHTML);\n"
+            + "      log(document.getElementById('tester').textContent);\n"
+            + "    }\n"
+            + "  </script>\n"
+            + "</head>"
+            + "<body onload='test()'>\n"
+            + "  <div id='tester'>before<![CDATA[inside]]>after</div>\n"
+            + "</body>\n"
+            + "</html>";
+
+        final WebDriver driver = loadPage2(html, URL_FIRST, "application/xhtml+xml", ISO_8859_1);
+        assertEquals("beforeafter", driver.findElement(By.id("tester")).getText());
+        verifyTitle2(driver, getExpectedAlerts());
+    }
+
+    /**
+     * Regression test for Bug #1219.
+     *
+     * @throws Exception if an error occurs
+     */
+    @Test
+    @Alerts({"before<![CDATA[<div></div>]]>after",
+             "<div xmlns=\"http://www.w3.org/1999/xhtml\" id=\"tester\">before<![CDATA[<div></div>]]>after</div>",
+             "before<div></div>after"})
+    @HtmlUnitNYI(CHROME = {"before<!--[CDATA[<div></div>]]-->after",
+                           "<div id=\"tester\">before<!--[CDATA[<div></div>]]-->after</div>",
+                           "beforeafter"},
+            EDGE = {"before<!--[CDATA[<div></div>]]-->after",
+                    "<div id=\"tester\">before<!--[CDATA[<div></div>]]-->after</div>",
+                    "beforeafter"},
+            FF = {"before<!--[CDATA[<div></div>]]-->after",
+                  "<div id=\"tester\">before<!--[CDATA[<div></div>]]-->after</div>",
+                  "beforeafter"},
+            FF_ESR = {"before<!--[CDATA[<div></div>]]-->after",
+                      "<div id=\"tester\">before<!--[CDATA[<div></div>]]-->after</div>",
+                      "beforeafter"})
+    public void earlyClosingCdata() throws Exception {
+        final String html
+            = "<?xml version=\"1.0\" encoding=\"UTF-8\"?>\n"
+            + "<!DOCTYPE html PUBLIC \n"
+            + "  \"-//W3C//DTD XHTML 1.0 Transitional//EN\" \n"
+            + "  \"http://www.w3.org/TR/xhtml1/DTD/xhtml1-transitional.dtd\">\n"
+            + "<html xmlns=\"http://www.w3.org/1999/xhtml\">\n"
+            + "<head>"
+            + "  <script>\n"
+            + LOG_TITLE_FUNCTION
+            + "    function test() {"
+            + "      log(document.getElementById('tester').innerHTML);\n"
+            + "      log(document.getElementById('tester').outerHTML);\n"
+            + "      log(document.getElementById('tester').textContent);\n"
+            + "    }\n"
+            + "  </script>\n"
+            + "</head>"
+            + "<body onload='test()'>\n"
+            + "  <div id='tester'>before<![CDATA[<div></div>]]>after</div>\n"
+            + "</body>\n"
+            + "</html>";
+
+        final WebDriver driver = loadPage2(html, URL_FIRST, "application/xhtml+xml", ISO_8859_1);
+        assertEquals("beforeafter", driver.findElement(By.id("tester")).getText());
+        verifyTitle2(driver, getExpectedAlerts());
+    }
 }