1/* 2 * Copyright (C) 2005 Frerich Raabe <raabe@kde.org> 3 * Copyright (C) 2006, 2009 Apple Inc. All rights reserved. 4 * Copyright (C) 2007 Alexey Proskuryakov <ap@webkit.org> 5 * 6 * Redistribution and use in source and binary forms, with or without 7 * modification, are permitted provided that the following conditions 8 * are met: 9 * 10 * 1. Redistributions of source code must retain the above copyright 11 * notice, this list of conditions and the following disclaimer. 12 * 2. Redistributions in binary form must reproduce the above copyright 13 * notice, this list of conditions and the following disclaimer in the 14 * documentation and/or other materials provided with the distribution. 15 * 16 * THIS SOFTWARE IS PROVIDED BY THE AUTHOR ``AS IS'' AND ANY EXPRESS OR 17 * IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE IMPLIED WARRANTIES 18 * OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE ARE DISCLAIMED. 19 * IN NO EVENT SHALL THE AUTHOR BE LIABLE FOR ANY DIRECT, INDIRECT, 20 * INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT 21 * NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, 22 * DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY 23 * THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT 24 * (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF 25 * THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE. 26 */ 27 28#include "config.h" 29#include "XPathStep.h" 30 31#include "Attr.h" 32#include "Document.h" 33#include "Element.h" 34#include "NodeTraversal.h" 35#include "XMLNSNames.h" 36#include "XPathParser.h" 37#include "XPathUtil.h" 38 39namespace WebCore { 40namespace XPath { 41 42Step::Step(Axis axis, const NodeTest& nodeTest, const Vector<Predicate*>& predicates) 43 : m_axis(axis) 44 , m_nodeTest(nodeTest) 45 , m_predicates(predicates) 46{ 47} 48 49Step::~Step() 50{ 51 deleteAllValues(m_predicates); 52 deleteAllValues(m_nodeTest.mergedPredicates()); 53} 54 55void Step::optimize() 56{ 57 // Evaluate predicates as part of node test if possible to avoid building unnecessary NodeSets. 58 // E.g., there is no need to build a set of all "foo" nodes to evaluate "foo[@bar]", we can check the predicate while enumerating. 59 // This optimization can be applied to predicates that are not context node list sensitive, or to first predicate that is only context position sensitive, e.g. foo[position() mod 2 = 0]. 60 Vector<Predicate*> remainingPredicates; 61 for (size_t i = 0; i < m_predicates.size(); ++i) { 62 Predicate* predicate = m_predicates[i]; 63 if ((!predicate->isContextPositionSensitive() || m_nodeTest.mergedPredicates().isEmpty()) && !predicate->isContextSizeSensitive() && remainingPredicates.isEmpty()) { 64 m_nodeTest.mergedPredicates().append(predicate); 65 } else 66 remainingPredicates.append(predicate); 67 } 68 swap(remainingPredicates, m_predicates); 69} 70 71void optimizeStepPair(Step* first, Step* second, bool& dropSecondStep) 72{ 73 dropSecondStep = false; 74 75 if (first->m_axis == Step::DescendantOrSelfAxis 76 && first->m_nodeTest.kind() == Step::NodeTest::AnyNodeTest 77 && !first->m_predicates.size() 78 && !first->m_nodeTest.mergedPredicates().size()) { 79 80 ASSERT(first->m_nodeTest.data().isEmpty()); 81 ASSERT(first->m_nodeTest.namespaceURI().isEmpty()); 82 83 // Optimize the common case of "//" AKA /descendant-or-self::node()/child::NodeTest to /descendant::NodeTest. 84 if (second->m_axis == Step::ChildAxis && second->predicatesAreContextListInsensitive()) { 85 first->m_axis = Step::DescendantAxis; 86 first->m_nodeTest = Step::NodeTest(second->m_nodeTest.kind(), second->m_nodeTest.data(), second->m_nodeTest.namespaceURI()); 87 swap(second->m_nodeTest.mergedPredicates(), first->m_nodeTest.mergedPredicates()); 88 swap(second->m_predicates, first->m_predicates); 89 first->optimize(); 90 dropSecondStep = true; 91 } 92 } 93} 94 95bool Step::predicatesAreContextListInsensitive() const 96{ 97 for (size_t i = 0; i < m_predicates.size(); ++i) { 98 Predicate* predicate = m_predicates[i]; 99 if (predicate->isContextPositionSensitive() || predicate->isContextSizeSensitive()) 100 return false; 101 } 102 103 for (size_t i = 0; i < m_nodeTest.mergedPredicates().size(); ++i) { 104 Predicate* predicate = m_nodeTest.mergedPredicates()[i]; 105 if (predicate->isContextPositionSensitive() || predicate->isContextSizeSensitive()) 106 return false; 107 } 108 109 return true; 110} 111 112void Step::evaluate(Node* context, NodeSet& nodes) const 113{ 114 EvaluationContext& evaluationContext = Expression::evaluationContext(); 115 evaluationContext.position = 0; 116 117 nodesInAxis(context, nodes); 118 119 // Check predicates that couldn't be merged into node test. 120 for (unsigned i = 0; i < m_predicates.size(); i++) { 121 Predicate* predicate = m_predicates[i]; 122 123 NodeSet newNodes; 124 if (!nodes.isSorted()) 125 newNodes.markSorted(false); 126 127 for (unsigned j = 0; j < nodes.size(); j++) { 128 Node* node = nodes[j]; 129 130 evaluationContext.node = node; 131 evaluationContext.size = nodes.size(); 132 evaluationContext.position = j + 1; 133 if (predicate->evaluate()) 134 newNodes.append(node); 135 } 136 137 nodes.swap(newNodes); 138 } 139} 140 141static inline Node::NodeType primaryNodeType(Step::Axis axis) 142{ 143 switch (axis) { 144 case Step::AttributeAxis: 145 return Node::ATTRIBUTE_NODE; 146 case Step::NamespaceAxis: 147 return Node::XPATH_NAMESPACE_NODE; 148 default: 149 return Node::ELEMENT_NODE; 150 } 151} 152 153// Evaluate NodeTest without considering merged predicates. 154static inline bool nodeMatchesBasicTest(Node* node, Step::Axis axis, const Step::NodeTest& nodeTest) 155{ 156 switch (nodeTest.kind()) { 157 case Step::NodeTest::TextNodeTest: 158 return node->nodeType() == Node::TEXT_NODE || node->nodeType() == Node::CDATA_SECTION_NODE; 159 case Step::NodeTest::CommentNodeTest: 160 return node->nodeType() == Node::COMMENT_NODE; 161 case Step::NodeTest::ProcessingInstructionNodeTest: { 162 const AtomicString& name = nodeTest.data(); 163 return node->nodeType() == Node::PROCESSING_INSTRUCTION_NODE && (name.isEmpty() || node->nodeName() == name); 164 } 165 case Step::NodeTest::AnyNodeTest: 166 return true; 167 case Step::NodeTest::NameTest: { 168 const AtomicString& name = nodeTest.data(); 169 const AtomicString& namespaceURI = nodeTest.namespaceURI(); 170 171 if (axis == Step::AttributeAxis) { 172 ASSERT(node->isAttributeNode()); 173 174 // In XPath land, namespace nodes are not accessible on the attribute axis. 175 if (node->namespaceURI() == XMLNSNames::xmlnsNamespaceURI) 176 return false; 177 178 if (name == starAtom) 179 return namespaceURI.isEmpty() || node->namespaceURI() == namespaceURI; 180 181 return node->localName() == name && node->namespaceURI() == namespaceURI; 182 } 183 184 // Node test on the namespace axis is not implemented yet, the caller has a check for it. 185 ASSERT(axis != Step::NamespaceAxis); 186 187 // For other axes, the principal node type is element. 188 ASSERT(primaryNodeType(axis) == Node::ELEMENT_NODE); 189 if (node->nodeType() != Node::ELEMENT_NODE) 190 return false; 191 192 if (name == starAtom) 193 return namespaceURI.isEmpty() || namespaceURI == node->namespaceURI(); 194 195 if (node->document()->isHTMLDocument()) { 196 if (node->isHTMLElement()) { 197 // Paths without namespaces should match HTML elements in HTML documents despite those having an XHTML namespace. Names are compared case-insensitively. 198 return equalIgnoringCase(toElement(node)->localName(), name) && (namespaceURI.isNull() || namespaceURI == node->namespaceURI()); 199 } 200 // An expression without any prefix shouldn't match no-namespace nodes (because HTML5 says so). 201 return toElement(node)->hasLocalName(name) && namespaceURI == node->namespaceURI() && !namespaceURI.isNull(); 202 } 203 return toElement(node)->hasLocalName(name) && namespaceURI == node->namespaceURI(); 204 } 205 } 206 ASSERT_NOT_REACHED(); 207 return false; 208} 209 210static inline bool nodeMatches(Node* node, Step::Axis axis, const Step::NodeTest& nodeTest) 211{ 212 if (!nodeMatchesBasicTest(node, axis, nodeTest)) 213 return false; 214 215 EvaluationContext& evaluationContext = Expression::evaluationContext(); 216 217 // Only the first merged predicate may depend on position. 218 ++evaluationContext.position; 219 220 const Vector<Predicate*>& mergedPredicates = nodeTest.mergedPredicates(); 221 for (unsigned i = 0; i < mergedPredicates.size(); i++) { 222 Predicate* predicate = mergedPredicates[i]; 223 224 evaluationContext.node = node; 225 // No need to set context size - we only get here when evaluating predicates that do not depend on it. 226 if (!predicate->evaluate()) 227 return false; 228 } 229 230 return true; 231} 232 233// Result nodes are ordered in axis order. Node test (including merged predicates) is applied. 234void Step::nodesInAxis(Node* context, NodeSet& nodes) const 235{ 236 ASSERT(nodes.isEmpty()); 237 switch (m_axis) { 238 case ChildAxis: 239 if (context->isAttributeNode()) // In XPath model, attribute nodes do not have children. 240 return; 241 242 for (Node* n = context->firstChild(); n; n = n->nextSibling()) 243 if (nodeMatches(n, ChildAxis, m_nodeTest)) 244 nodes.append(n); 245 return; 246 case DescendantAxis: 247 if (context->isAttributeNode()) // In XPath model, attribute nodes do not have children. 248 return; 249 250 for (Node* n = context->firstChild(); n; n = NodeTraversal::next(n, context)) 251 if (nodeMatches(n, DescendantAxis, m_nodeTest)) 252 nodes.append(n); 253 return; 254 case ParentAxis: 255 if (context->isAttributeNode()) { 256 Element* n = static_cast<Attr*>(context)->ownerElement(); 257 if (nodeMatches(n, ParentAxis, m_nodeTest)) 258 nodes.append(n); 259 } else { 260 ContainerNode* n = context->parentNode(); 261 if (n && nodeMatches(n, ParentAxis, m_nodeTest)) 262 nodes.append(n); 263 } 264 return; 265 case AncestorAxis: { 266 Node* n = context; 267 if (context->isAttributeNode()) { 268 n = static_cast<Attr*>(context)->ownerElement(); 269 if (nodeMatches(n, AncestorAxis, m_nodeTest)) 270 nodes.append(n); 271 } 272 for (n = n->parentNode(); n; n = n->parentNode()) 273 if (nodeMatches(n, AncestorAxis, m_nodeTest)) 274 nodes.append(n); 275 nodes.markSorted(false); 276 return; 277 } 278 case FollowingSiblingAxis: 279 if (context->nodeType() == Node::ATTRIBUTE_NODE || 280 context->nodeType() == Node::XPATH_NAMESPACE_NODE) 281 return; 282 283 for (Node* n = context->nextSibling(); n; n = n->nextSibling()) 284 if (nodeMatches(n, FollowingSiblingAxis, m_nodeTest)) 285 nodes.append(n); 286 return; 287 case PrecedingSiblingAxis: 288 if (context->nodeType() == Node::ATTRIBUTE_NODE || 289 context->nodeType() == Node::XPATH_NAMESPACE_NODE) 290 return; 291 292 for (Node* n = context->previousSibling(); n; n = n->previousSibling()) 293 if (nodeMatches(n, PrecedingSiblingAxis, m_nodeTest)) 294 nodes.append(n); 295 296 nodes.markSorted(false); 297 return; 298 case FollowingAxis: 299 if (context->isAttributeNode()) { 300 Node* p = static_cast<Attr*>(context)->ownerElement(); 301 while ((p = NodeTraversal::next(p))) 302 if (nodeMatches(p, FollowingAxis, m_nodeTest)) 303 nodes.append(p); 304 } else { 305 for (Node* p = context; !isRootDomNode(p); p = p->parentNode()) { 306 for (Node* n = p->nextSibling(); n; n = n->nextSibling()) { 307 if (nodeMatches(n, FollowingAxis, m_nodeTest)) 308 nodes.append(n); 309 for (Node* c = n->firstChild(); c; c = NodeTraversal::next(c, n)) 310 if (nodeMatches(c, FollowingAxis, m_nodeTest)) 311 nodes.append(c); 312 } 313 } 314 } 315 return; 316 case PrecedingAxis: { 317 if (context->isAttributeNode()) 318 context = static_cast<Attr*>(context)->ownerElement(); 319 320 Node* n = context; 321 while (ContainerNode* parent = n->parentNode()) { 322 for (n = NodeTraversal::previous(n); n != parent; n = NodeTraversal::previous(n)) 323 if (nodeMatches(n, PrecedingAxis, m_nodeTest)) 324 nodes.append(n); 325 n = parent; 326 } 327 nodes.markSorted(false); 328 return; 329 } 330 case AttributeAxis: { 331 if (!context->isElementNode()) 332 return; 333 334 Element* contextElement = toElement(context); 335 336 // Avoid lazily creating attribute nodes for attributes that we do not need anyway. 337 if (m_nodeTest.kind() == NodeTest::NameTest && m_nodeTest.data() != starAtom) { 338 RefPtr<Node> n = contextElement->getAttributeNodeNS(m_nodeTest.namespaceURI(), m_nodeTest.data()); 339 if (n && n->namespaceURI() != XMLNSNames::xmlnsNamespaceURI) { // In XPath land, namespace nodes are not accessible on the attribute axis. 340 if (nodeMatches(n.get(), AttributeAxis, m_nodeTest)) // Still need to check merged predicates. 341 nodes.append(n.release()); 342 } 343 return; 344 } 345 346 if (!contextElement->hasAttributes()) 347 return; 348 349 for (unsigned i = 0; i < contextElement->attributeCount(); ++i) { 350 RefPtr<Attr> attr = contextElement->ensureAttr(contextElement->attributeItem(i)->name()); 351 if (nodeMatches(attr.get(), AttributeAxis, m_nodeTest)) 352 nodes.append(attr.release()); 353 } 354 return; 355 } 356 case NamespaceAxis: 357 // XPath namespace nodes are not implemented yet. 358 return; 359 case SelfAxis: 360 if (nodeMatches(context, SelfAxis, m_nodeTest)) 361 nodes.append(context); 362 return; 363 case DescendantOrSelfAxis: 364 if (nodeMatches(context, DescendantOrSelfAxis, m_nodeTest)) 365 nodes.append(context); 366 if (context->isAttributeNode()) // In XPath model, attribute nodes do not have children. 367 return; 368 369 for (Node* n = context->firstChild(); n; n = NodeTraversal::next(n, context)) 370 if (nodeMatches(n, DescendantOrSelfAxis, m_nodeTest)) 371 nodes.append(n); 372 return; 373 case AncestorOrSelfAxis: { 374 if (nodeMatches(context, AncestorOrSelfAxis, m_nodeTest)) 375 nodes.append(context); 376 Node* n = context; 377 if (context->isAttributeNode()) { 378 n = static_cast<Attr*>(context)->ownerElement(); 379 if (nodeMatches(n, AncestorOrSelfAxis, m_nodeTest)) 380 nodes.append(n); 381 } 382 for (n = n->parentNode(); n; n = n->parentNode()) 383 if (nodeMatches(n, AncestorOrSelfAxis, m_nodeTest)) 384 nodes.append(n); 385 386 nodes.markSorted(false); 387 return; 388 } 389 } 390 ASSERT_NOT_REACHED(); 391} 392 393 394} 395} 396