| 1 | /**************************************************************************** |
| 2 | ** |
| 3 | ** Copyright (C) 2016 The Qt Company Ltd. |
| 4 | ** Contact: https://www.qt.io/licensing/ |
| 5 | ** |
| 6 | ** This file is part of the QtXmlPatterns module of the Qt Toolkit. |
| 7 | ** |
| 8 | ** $QT_BEGIN_LICENSE:LGPL$ |
| 9 | ** Commercial License Usage |
| 10 | ** Licensees holding valid commercial Qt licenses may use this file in |
| 11 | ** accordance with the commercial license agreement provided with the |
| 12 | ** Software or, alternatively, in accordance with the terms contained in |
| 13 | ** a written agreement between you and The Qt Company. For licensing terms |
| 14 | ** and conditions see https://www.qt.io/terms-conditions. For further |
| 15 | ** information use the contact form at https://www.qt.io/contact-us. |
| 16 | ** |
| 17 | ** GNU Lesser General Public License Usage |
| 18 | ** Alternatively, this file may be used under the terms of the GNU Lesser |
| 19 | ** General Public License version 3 as published by the Free Software |
| 20 | ** Foundation and appearing in the file LICENSE.LGPL3 included in the |
| 21 | ** packaging of this file. Please review the following information to |
| 22 | ** ensure the GNU Lesser General Public License version 3 requirements |
| 23 | ** will be met: https://www.gnu.org/licenses/lgpl-3.0.html. |
| 24 | ** |
| 25 | ** GNU General Public License Usage |
| 26 | ** Alternatively, this file may be used under the terms of the GNU |
| 27 | ** General Public License version 2.0 or (at your option) the GNU General |
| 28 | ** Public license version 3 or any later version approved by the KDE Free |
| 29 | ** Qt Foundation. The licenses are as published by the Free Software |
| 30 | ** Foundation and appearing in the file LICENSE.GPL2 and LICENSE.GPL3 |
| 31 | ** included in the packaging of this file. Please review the following |
| 32 | ** information to ensure the GNU General Public License requirements will |
| 33 | ** be met: https://www.gnu.org/licenses/gpl-2.0.html and |
| 34 | ** https://www.gnu.org/licenses/gpl-3.0.html. |
| 35 | ** |
| 36 | ** $QT_END_LICENSE$ |
| 37 | ** |
| 38 | ****************************************************************************/ |
| 39 | |
| 40 | // |
| 41 | // W A R N I N G |
| 42 | // ------------- |
| 43 | // |
| 44 | // This file is not part of the Qt API. It exists purely as an |
| 45 | // implementation detail. This header file may change from version to |
| 46 | // version without notice, or even be removed. |
| 47 | // |
| 48 | // We mean it. |
| 49 | // |
| 50 | |
| 51 | /** |
| 52 | * @file |
| 53 | * @short This file is included by qacceltreebuilder_p.h. |
| 54 | * If you need includes in this file, put them in qacceltreebuilder_p.h, outside of the namespace. |
| 55 | */ |
| 56 | |
| 57 | template <bool FromDocument> |
| 58 | AccelTreeBuilder<FromDocument>::AccelTreeBuilder(const QUrl &docURI, |
| 59 | const QUrl &baseURI, |
| 60 | const NamePool::Ptr &np, |
| 61 | ReportContext *const context, |
| 62 | Features features) : m_preNumber(-1) |
| 63 | , m_isPreviousAtomic(false) |
| 64 | , m_hasCharacters(false) |
| 65 | , m_isCharactersCompressed(false) |
| 66 | , m_namePool(np) |
| 67 | , m_document(new AccelTree(docURI, baseURI)) |
| 68 | , m_skippedDocumentNodes(0) |
| 69 | , m_documentURI(docURI) |
| 70 | , m_context(context) |
| 71 | , m_features(features) |
| 72 | { |
| 73 | Q_ASSERT(m_namePool); |
| 74 | |
| 75 | /* TODO Perhaps we can merge m_ancestors and m_size |
| 76 | * into one, and store a struct for the two instead? */ |
| 77 | m_ancestors.reserve(asize: DefaultNodeStackSize); |
| 78 | m_ancestors.push(t: -1); |
| 79 | |
| 80 | m_size.reserve(asize: DefaultNodeStackSize); |
| 81 | m_size.push(t: 0); |
| 82 | } |
| 83 | |
| 84 | template <bool FromDocument> |
| 85 | void AccelTreeBuilder<FromDocument>::startStructure() |
| 86 | { |
| 87 | if(m_hasCharacters) |
| 88 | { |
| 89 | /* We create a node even if m_characters is empty. |
| 90 | * Remember that `text {""}' creates one text node |
| 91 | * with string value "". */ |
| 92 | |
| 93 | ++m_preNumber; |
| 94 | m_document->basicData.append(t: AccelTree::BasicNodeData(currentDepth(), |
| 95 | currentParent(), |
| 96 | QXmlNodeModelIndex::Text, |
| 97 | m_isCharactersCompressed ? AccelTree::IsCompressed : 0)); |
| 98 | m_document->data.insert(akey: m_preNumber, avalue: m_characters); |
| 99 | ++m_size.top(); |
| 100 | |
| 101 | m_characters.clear(); /* We don't want it added twice. */ |
| 102 | m_hasCharacters = false; |
| 103 | |
| 104 | if(m_isCharactersCompressed) |
| 105 | m_isCharactersCompressed = false; |
| 106 | } |
| 107 | } |
| 108 | |
| 109 | template <bool FromDocument> |
| 110 | void AccelTreeBuilder<FromDocument>::item(const Item &it) |
| 111 | { |
| 112 | Q_ASSERT(it); |
| 113 | |
| 114 | if(it.isAtomicValue()) |
| 115 | { |
| 116 | if(m_isPreviousAtomic) |
| 117 | { |
| 118 | m_characters += QLatin1Char(' '); |
| 119 | m_characters += it.stringValue(); |
| 120 | } |
| 121 | else |
| 122 | { |
| 123 | m_isPreviousAtomic = true; |
| 124 | const QString sv(it.stringValue()); |
| 125 | |
| 126 | if(!sv.isEmpty()) |
| 127 | { |
| 128 | m_characters += sv; |
| 129 | m_hasCharacters = true; |
| 130 | } |
| 131 | } |
| 132 | } |
| 133 | else |
| 134 | sendAsNode(outputItem: it); |
| 135 | } |
| 136 | |
| 137 | template <bool FromDocument> |
| 138 | void AccelTreeBuilder<FromDocument>::startElement(const QXmlName &name) |
| 139 | { |
| 140 | startElement(name, 1, 1); |
| 141 | } |
| 142 | |
| 143 | template <bool FromDocument> |
| 144 | void AccelTreeBuilder<FromDocument>::startElement(const QXmlName &name, qint64 line, qint64 column) |
| 145 | { |
| 146 | startStructure(); |
| 147 | |
| 148 | AccelTree::BasicNodeData data(currentDepth(), currentParent(), QXmlNodeModelIndex::Element, -1, name); |
| 149 | m_document->basicData.append(t: data); |
| 150 | if (m_features & SourceLocationsFeature) |
| 151 | m_document->sourcePositions.insert(akey: m_document->maximumPreNumber(), avalue: qMakePair(x: line, y: column)); |
| 152 | |
| 153 | ++m_preNumber; |
| 154 | m_ancestors.push(t: m_preNumber); |
| 155 | |
| 156 | ++m_size.top(); |
| 157 | m_size.push(t: 0); |
| 158 | |
| 159 | /* With node constructors, we can receive names for which we have no namespace |
| 160 | * constructors, such as in the query '<xs:space/>'. Since the 'xs' prefix has no |
| 161 | * NamespaceConstructor in this case, we synthesize the namespace. |
| 162 | * |
| 163 | * In case we're constructing from an XML document we avoid the call because |
| 164 | * although it's redundant, it's on extra virtual call for each element. */ |
| 165 | if(!FromDocument) |
| 166 | namespaceBinding(nb: QXmlName(name.namespaceURI(), 0, name.prefix())); |
| 167 | |
| 168 | m_isPreviousAtomic = false; |
| 169 | } |
| 170 | |
| 171 | template <bool FromDocument> |
| 172 | void AccelTreeBuilder<FromDocument>::endElement() |
| 173 | { |
| 174 | startStructure(); |
| 175 | const AccelTree::PreNumber index = m_ancestors.pop(); |
| 176 | AccelTree::BasicNodeData &data = m_document->basicData[index]; |
| 177 | |
| 178 | /* Sub trees needs to be included in upper trees, so we add the count of this element |
| 179 | * to our parent. */ |
| 180 | m_size[m_size.count() - 2] += m_size.top(); |
| 181 | |
| 182 | data.setSize(m_size.pop()); |
| 183 | m_isPreviousAtomic = false; |
| 184 | } |
| 185 | |
| 186 | template <bool FromDocument> |
| 187 | void AccelTreeBuilder<FromDocument>::attribute(const QXmlName &name, const QStringRef &value) |
| 188 | { |
| 189 | /* Attributes adds a namespace binding, so lets synthesize one. |
| 190 | * |
| 191 | * We optimize by checking whether we have a namespace for which a binding would |
| 192 | * be generated. Happens relatively rarely. */ |
| 193 | if(name.hasPrefix()) |
| 194 | namespaceBinding(nb: QXmlName(name.namespaceURI(), 0, name.prefix())); |
| 195 | |
| 196 | m_document->basicData.append(t: AccelTree::BasicNodeData(currentDepth(), currentParent(), QXmlNodeModelIndex::Attribute, 0, name)); |
| 197 | ++m_preNumber; |
| 198 | ++m_size.top(); |
| 199 | |
| 200 | m_isPreviousAtomic = false; |
| 201 | |
| 202 | if(name.namespaceURI() == StandardNamespaces::xml && name.localName() == StandardLocalNames::id) |
| 203 | { |
| 204 | const QString normalized(value.toString().simplified()); |
| 205 | |
| 206 | if(QXmlUtils::isNCName(ncName: normalized)) |
| 207 | { |
| 208 | const QXmlName::LocalNameCode id = m_namePool->allocateLocalName(ln: normalized); |
| 209 | |
| 210 | const int oldSize = m_document->m_IDs.count(); |
| 211 | m_document->m_IDs.insert(id, currentParent()); |
| 212 | /* We don't run the value through m_attributeCompress here, because |
| 213 | * the likelyhood of it deing identical to another attribute is |
| 214 | * very small. */ |
| 215 | m_document->data.insert(akey: m_preNumber, avalue: normalized); |
| 216 | |
| 217 | /** |
| 218 | * In the case that we're called for doc-available(), m_context is |
| 219 | * null, and we need to flag somehow that we failed to load this |
| 220 | * document. |
| 221 | */ |
| 222 | if(oldSize == m_document->m_IDs.count() && m_context) // TODO |
| 223 | { |
| 224 | Q_ASSERT(m_context); |
| 225 | m_context->error(QtXmlPatterns::tr(sourceText: "An %1-attribute with value %2 has already been declared." ) |
| 226 | .arg(args: formatKeyword(keyword: "xml:id" ), |
| 227 | args: formatData(data: normalized)), |
| 228 | FromDocument ? ReportContext::FODC0002 : ReportContext::XQDY0091, |
| 229 | this); |
| 230 | } |
| 231 | } |
| 232 | else if(m_context) // TODO |
| 233 | { |
| 234 | Q_ASSERT(m_context); |
| 235 | |
| 236 | /* If we're building from an XML Document(e.g, we're fed from QXmlStreamReader, we raise FODC0002, |
| 237 | * otherwise XQDY0091. */ |
| 238 | m_context->error(QtXmlPatterns::tr(sourceText: "An %1-attribute must have a " |
| 239 | "valid %2 as value, which %3 isn't." ).arg(args: formatKeyword(keyword: "xml:id" ), |
| 240 | args: formatType(np: m_namePool, type: BuiltinTypes::xsNCName), |
| 241 | args: formatData(data: value.toString())), |
| 242 | FromDocument ? ReportContext::FODC0002 : ReportContext::XQDY0091, |
| 243 | this); |
| 244 | } |
| 245 | } |
| 246 | else |
| 247 | m_document->data.insert(akey: m_preNumber, avalue: *m_attributeCompress.insert(value: value.toString())); |
| 248 | } |
| 249 | |
| 250 | template <bool FromDocument> |
| 251 | void AccelTreeBuilder<FromDocument>::characters(const QStringRef &ch) |
| 252 | { |
| 253 | |
| 254 | /* If a text node constructor appears by itself, a node needs to |
| 255 | * be created. Therefore, we set m_hasCharacters |
| 256 | * if we're the only node. |
| 257 | * However, if the text node appears as a child of a document or element |
| 258 | * node it is discarded if it's empty. |
| 259 | */ |
| 260 | if(m_hasCharacters && m_isCharactersCompressed) |
| 261 | { |
| 262 | m_characters = CompressedWhitespace::decompress(input: m_characters); |
| 263 | m_isCharactersCompressed = false; |
| 264 | } |
| 265 | |
| 266 | m_characters += ch; |
| 267 | |
| 268 | m_isPreviousAtomic = false; |
| 269 | m_hasCharacters = !m_characters.isEmpty() || m_preNumber == -1; /* -1 is our start value. */ |
| 270 | } |
| 271 | |
| 272 | template <bool FromDocument> |
| 273 | void AccelTreeBuilder<FromDocument>::whitespaceOnly(const QStringRef &ch) |
| 274 | { |
| 275 | Q_ASSERT(!ch.isEmpty()); |
| 276 | Q_ASSERT(ch.toString().trimmed().isEmpty()); |
| 277 | |
| 278 | /* This gets problematic due to how QXmlStreamReader works(which |
| 279 | * is the only one we get whitespaceOnly() events from). Namely, text intermingled |
| 280 | * with CDATA gets reported as individual Characters events, and |
| 281 | * QXmlStreamReader::isWhitespace() can return differently for each of those. However, |
| 282 | * it will occur very rarely, so this workaround of 1) mistakenly compressing 2) decompressing 3) |
| 283 | * appending, will happen infrequently. |
| 284 | */ |
| 285 | if(m_hasCharacters) |
| 286 | { |
| 287 | if(m_isCharactersCompressed) |
| 288 | { |
| 289 | m_characters = CompressedWhitespace::decompress(input: m_characters); |
| 290 | m_isCharactersCompressed = false; |
| 291 | } |
| 292 | |
| 293 | m_characters.append(s: ch.toString()); |
| 294 | } |
| 295 | else |
| 296 | { |
| 297 | /* We haven't received a text node previously. */ |
| 298 | m_characters = CompressedWhitespace::compress(input: ch); |
| 299 | m_isCharactersCompressed = true; |
| 300 | m_isPreviousAtomic = false; |
| 301 | m_hasCharacters = true; |
| 302 | } |
| 303 | } |
| 304 | |
| 305 | template <bool FromDocument> |
| 306 | void AccelTreeBuilder<FromDocument>::processingInstruction(const QXmlName &target, |
| 307 | const QString &data) |
| 308 | { |
| 309 | startStructure(); |
| 310 | ++m_preNumber; |
| 311 | m_document->data.insert(akey: m_preNumber, avalue: data); |
| 312 | |
| 313 | m_document->basicData.append(t: AccelTree::BasicNodeData(currentDepth(), |
| 314 | currentParent(), |
| 315 | QXmlNodeModelIndex::ProcessingInstruction, |
| 316 | 0, |
| 317 | target)); |
| 318 | ++m_size.top(); |
| 319 | m_isPreviousAtomic = false; |
| 320 | } |
| 321 | |
| 322 | template <bool FromDocument> |
| 323 | void AccelTreeBuilder<FromDocument>::(const QString &content) |
| 324 | { |
| 325 | startStructure(); |
| 326 | m_document->basicData.append(t: AccelTree::BasicNodeData(currentDepth(), currentParent(), QXmlNodeModelIndex::Comment, 0)); |
| 327 | ++m_preNumber; |
| 328 | m_document->data.insert(akey: m_preNumber, avalue: content); |
| 329 | ++m_size.top(); |
| 330 | } |
| 331 | |
| 332 | template <bool FromDocument> |
| 333 | void AccelTreeBuilder<FromDocument>::namespaceBinding(const QXmlName &nb) |
| 334 | { |
| 335 | /* Note, because attribute() sometimes generate namespaceBinding() calls, this function |
| 336 | * can be called after attributes, in contrast to what the class documentation says. This is ok, |
| 337 | * as long as we're not dealing with public API. */ |
| 338 | |
| 339 | /* If we've received attributes, it means the element's size have changed and m_preNumber have advanced, |
| 340 | * so "reverse back" to the actual element. */ |
| 341 | const AccelTree::PreNumber pn = m_preNumber - m_size.top(); |
| 342 | |
| 343 | QVector<QXmlName> &nss = m_document->namespaces[pn]; |
| 344 | |
| 345 | /* "xml" hasn't been declared for each node, AccelTree::namespaceBindings() adds it, so avoid it |
| 346 | * such that we don't get duplicates. */ |
| 347 | if(nb.prefix() == StandardPrefixes::xml) |
| 348 | return; |
| 349 | |
| 350 | /* If we already have the binding, skip it. */ |
| 351 | const int len = nss.count(); |
| 352 | for(int i = 0; i < len; ++i) |
| 353 | { |
| 354 | if(nss.at(i).prefix() == nb.prefix()) |
| 355 | return; |
| 356 | } |
| 357 | |
| 358 | nss.append(t: nb); |
| 359 | } |
| 360 | |
| 361 | template <bool FromDocument> |
| 362 | void AccelTreeBuilder<FromDocument>::startDocument() |
| 363 | { |
| 364 | /* If we have already received nodes, we can't add a document node. */ |
| 365 | if(m_preNumber == -1) /* -1 is our start value. */ |
| 366 | { |
| 367 | m_size.push(t: 0); |
| 368 | m_document->basicData.append(t: AccelTree::BasicNodeData(0, -1, QXmlNodeModelIndex::Document, -1)); |
| 369 | ++m_preNumber; |
| 370 | m_ancestors.push(t: m_preNumber); |
| 371 | } |
| 372 | else |
| 373 | ++m_skippedDocumentNodes; |
| 374 | |
| 375 | m_isPreviousAtomic = false; |
| 376 | } |
| 377 | |
| 378 | template <bool FromDocument> |
| 379 | void AccelTreeBuilder<FromDocument>::endDocument() |
| 380 | { |
| 381 | if(m_skippedDocumentNodes == 0) |
| 382 | { |
| 383 | /* Create text nodes, if we've received any. We do this only if we're the |
| 384 | * top node because if we're getting this event as being a child of an element, |
| 385 | * text nodes or atomic values can appear after us, and which must get |
| 386 | * merged with the previous text. |
| 387 | * |
| 388 | * We call startStructure() before we pop the ancestor, such that the text node becomes |
| 389 | * a child of this document node. */ |
| 390 | startStructure(); |
| 391 | |
| 392 | m_document->basicData.first().setSize(m_size.pop()); |
| 393 | m_ancestors.pop(); |
| 394 | } |
| 395 | else |
| 396 | --m_skippedDocumentNodes; |
| 397 | |
| 398 | m_isPreviousAtomic = false; |
| 399 | } |
| 400 | |
| 401 | template <bool FromDocument> |
| 402 | void AccelTreeBuilder<FromDocument>::atomicValue(const QVariant &value) |
| 403 | { |
| 404 | Q_UNUSED(value); |
| 405 | // TODO |
| 406 | } |
| 407 | |
| 408 | template <bool FromDocument> |
| 409 | QAbstractXmlNodeModel::Ptr AccelTreeBuilder<FromDocument>::builtDocument() |
| 410 | { |
| 411 | /* Create a text node, if we have received text in some way. */ |
| 412 | startStructure(); |
| 413 | m_document->printStats(np: m_namePool); |
| 414 | |
| 415 | return m_document; |
| 416 | } |
| 417 | |
| 418 | template <bool FromDocument> |
| 419 | NodeBuilder::Ptr AccelTreeBuilder<FromDocument>::create(const QUrl &baseURI) const |
| 420 | { |
| 421 | Q_UNUSED(baseURI); |
| 422 | return NodeBuilder::Ptr(new AccelTreeBuilder(QUrl(), baseURI, m_namePool, m_context)); |
| 423 | } |
| 424 | |
| 425 | template <bool FromDocument> |
| 426 | void AccelTreeBuilder<FromDocument>::startOfSequence() |
| 427 | { |
| 428 | } |
| 429 | |
| 430 | template <bool FromDocument> |
| 431 | void AccelTreeBuilder<FromDocument>::endOfSequence() |
| 432 | { |
| 433 | } |
| 434 | |
| 435 | template <bool FromDocument> |
| 436 | const SourceLocationReflection *AccelTreeBuilder<FromDocument>::actualReflection() const |
| 437 | { |
| 438 | return this; |
| 439 | } |
| 440 | |
| 441 | template <bool FromDocument> |
| 442 | QSourceLocation AccelTreeBuilder<FromDocument>::sourceLocation() const |
| 443 | { |
| 444 | if(m_documentURI.isEmpty()) |
| 445 | return QSourceLocation(QUrl(QLatin1String("AnonymousNodeTree" ))); |
| 446 | else |
| 447 | return QSourceLocation(m_documentURI); |
| 448 | } |
| 449 | |
| 450 | |