Handle the XPath / (root) operator correctly for nodes that aren't attached to the...
authorcevans@google.com <cevans@google.com@268f45cc-cd09-0410-ab3c-d52691b4dbfc>
Wed, 15 Aug 2012 01:28:54 +0000 (01:28 +0000)
committercevans@google.com <cevans@google.com@268f45cc-cd09-0410-ab3c-d52691b4dbfc>
Wed, 15 Aug 2012 01:28:54 +0000 (01:28 +0000)
https://bugs.webkit.org/show_bug.cgi?id=36427

Reviewed by Abhishek Arya.

Source/WebCore:

We now behave the same as Firefox 14.
The consensus seems to be that the XPath spec is ambiguous for the case of detached nodes, and that using the fragment root is more intuitive than the document root for the case of detached nodes.
For example, http://www.w3.org/TR/xpath/ section 2 "Location Paths" is only clear for attached nodes: "A / by itself selects the root node of the document containing the context node. If it is followed by a relative location path, then the location path selects the set of nodes that would be selected by the relative location path relative to the root node of the document containing the context node."

Test: fast/xpath/xpath-detached-nodes.html

* xml/XPathPath.cpp:
(WebCore::XPath::LocationPath::evaluate): Jump to the root of the detached subtree instead of the parent document if the node isn't attached to the document.

LayoutTests:

* fast/xpath/xpath-detached-nodes-expected.txt: Added.
* fast/xpath/xpath-detached-nodes.html: Added.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@125631 268f45cc-cd09-0410-ab3c-d52691b4dbfc

LayoutTests/ChangeLog
LayoutTests/fast/xpath/xpath-detached-nodes-expected.txt [new file with mode: 0644]
LayoutTests/fast/xpath/xpath-detached-nodes.html [new file with mode: 0644]
Source/WebCore/ChangeLog
Source/WebCore/xml/XPathPath.cpp

index 4f47dc8..4820387 100644 (file)
@@ -1,3 +1,13 @@
+2012-08-14  Chris Evans  <cevans@google.com>
+
+        Handle the XPath / (root) operator correctly for nodes that aren't attached to the document.
+        https://bugs.webkit.org/show_bug.cgi?id=36427
+
+        Reviewed by Abhishek Arya.
+
+        * fast/xpath/xpath-detached-nodes-expected.txt: Added.
+        * fast/xpath/xpath-detached-nodes.html: Added.
+
 2012-08-14  Alexandru Chiculita  <achicu@adobe.com>
 
         [CSS Shaders][Chromium] Filters area applied twice when CustomFilterOperation is in the list
diff --git a/LayoutTests/fast/xpath/xpath-detached-nodes-expected.txt b/LayoutTests/fast/xpath/xpath-detached-nodes-expected.txt
new file mode 100644 (file)
index 0000000..3f126f0
--- /dev/null
@@ -0,0 +1,12 @@
+This tests XPath expressions on detached document fragments and nodes. 
+See https://bugs.webkit.org/show_bug.cgi?id=36427
+
+PASS result.numberValue is 1
+PASS result.numberValue is 0
+PASS result.numberValue is 1
+PASS result.numberValue is 0
+PASS result.numberValue is NaN
+PASS successfullyParsed is true
+
+TEST COMPLETE
+
diff --git a/LayoutTests/fast/xpath/xpath-detached-nodes.html b/LayoutTests/fast/xpath/xpath-detached-nodes.html
new file mode 100644 (file)
index 0000000..fa75d38
--- /dev/null
@@ -0,0 +1,39 @@
+<!DOCTYPE html>
+<html>
+<head>
+<script src="../js/resources/js-test-pre.js"></script>
+</head>
+<body>
+<p>This tests XPath expressions on detached document fragments and nodes.
+<br/>See https://bugs.webkit.org/show_bug.cgi?id=36427
+<div id="console"></div>
+
+<script>
+  if (window.testRunner)
+    testRunner.dumpAsText();
+
+  frag = document.createDocumentFragment();
+  child = document.createElement('div');
+  frag.appendChild(child);
+  result = document.evaluate("count(/div)",
+                             child, null, XPathResult.NUMBER_TYPE, null);
+  shouldBe("result.numberValue", "1");
+  result = document.evaluate("count(/html)",
+                             child, null, XPathResult.NUMBER_TYPE, null);
+  shouldBe("result.numberValue", "0");
+
+  ele = document.createElement('p');
+  ele.appendChild(document.createElement('h1'));
+  result = document.evaluate("count(/h1)",
+                             ele, null, XPathResult.NUMBER_TYPE, null);
+  shouldBe("result.numberValue", "1");
+  result = document.evaluate("count(/html)",
+                             ele, null, XPathResult.NUMBER_TYPE, null);
+  shouldBe("result.numberValue", "0");
+  result = document.evaluate("/* | *",
+                             ele, null, XPathResult.NUMBER_TYPE, null);
+  shouldBe("result.numberValue", "NaN");
+</script>
+<script src="../js/resources/js-test-post.js"></script>
+</body>
+</html>
index 73c0036..7a2ff67 100644 (file)
@@ -1,3 +1,19 @@
+2012-08-14  Chris Evans  <cevans@google.com>
+
+        Handle the XPath / (root) operator correctly for nodes that aren't attached to the document.
+        https://bugs.webkit.org/show_bug.cgi?id=36427
+
+        Reviewed by Abhishek Arya.
+
+        We now behave the same as Firefox 14.
+        The consensus seems to be that the XPath spec is ambiguous for the case of detached nodes, and that using the fragment root is more intuitive than the document root for the case of detached nodes.
+        For example, http://www.w3.org/TR/xpath/ section 2 "Location Paths" is only clear for attached nodes: "A / by itself selects the root node of the document containing the context node. If it is followed by a relative location path, then the location path selects the set of nodes that would be selected by the relative location path relative to the root node of the document containing the context node."
+
+        Test: fast/xpath/xpath-detached-nodes.html
+
+        * xml/XPathPath.cpp:
+        (WebCore::XPath::LocationPath::evaluate): Jump to the root of the detached subtree instead of the parent document if the node isn't attached to the document.
+
 2012-08-14  Alexandru Chiculita  <achicu@adobe.com>
 
         [CSS Shaders][Chromium] Filters area applied twice when CustomFilterOperation is in the list
index ba1ded3..db88c98 100644 (file)
@@ -93,11 +93,16 @@ Value LocationPath::evaluate() const
 {
     EvaluationContext& evaluationContext = Expression::evaluationContext();
     EvaluationContext backupContext = evaluationContext;
-    // For absolute location paths, the context node is ignored - the
-    // document's root node is used instead.
+    // For absolute location paths, the context node is ignored. The
+    // document's root node is used for attached nodes, otherwise the root
+    // node of the detached subtree is used.
     Node* context = evaluationContext.node.get();
-    if (m_absolute && context->nodeType() != Node::DOCUMENT_NODE) 
-        context = context->ownerDocument();
+    if (m_absolute && context->nodeType() != Node::DOCUMENT_NODE)  {
+        if (context->inDocument())
+            context = context->ownerDocument();
+        else
+            context = context->highestAncestor();
+    }
 
     NodeSet nodes;
     nodes.append(context);