[188] | 1 | /*
|
---|
| 2 | * The Apache Software License, Version 1.1
|
---|
| 3 | *
|
---|
| 4 | * Copyright (c) 2002 The Apache Software Foundation. All rights
|
---|
| 5 | * reserved.
|
---|
| 6 | *
|
---|
| 7 | * Redistribution and use in source and binary forms, with or without
|
---|
| 8 | * modification, are permitted provided that the following conditions
|
---|
| 9 | * are met:
|
---|
| 10 | *
|
---|
| 11 | * 1. Redistributions of source code must retain the above copyright
|
---|
| 12 | * notice, this list of conditions and the following disclaimer.
|
---|
| 13 | *
|
---|
| 14 | * 2. Redistributions in binary form must reproduce the above copyright
|
---|
| 15 | * notice, this list of conditions and the following disclaimer in
|
---|
| 16 | * the documentation and/or other materials provided with the
|
---|
| 17 | * distribution.
|
---|
| 18 | *
|
---|
| 19 | * 3. The end-user documentation included with the redistribution,
|
---|
| 20 | * if any, must include the following acknowledgment:
|
---|
| 21 | * "This product includes software developed by the
|
---|
| 22 | * Apache Software Foundation (http://www.apache.org/)."
|
---|
| 23 | * Alternately, this acknowledgment may appear in the software itself,
|
---|
| 24 | * if and wherever such third-party acknowledgments normally appear.
|
---|
| 25 | *
|
---|
| 26 | * 4. The names "Xerces" and "Apache Software Foundation" must
|
---|
| 27 | * not be used to endorse or promote products derived from this
|
---|
| 28 | * software without prior written permission. For written
|
---|
| 29 | * permission, please contact apache\@apache.org.
|
---|
| 30 | *
|
---|
| 31 | * 5. Products derived from this software may not be called "Apache",
|
---|
| 32 | * nor may "Apache" appear in their name, without prior written
|
---|
| 33 | * permission of the Apache Software Foundation.
|
---|
| 34 | *
|
---|
| 35 | * THIS SOFTWARE IS PROVIDED ``AS IS'' AND ANY EXPRESSED OR IMPLIED
|
---|
| 36 | * WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE IMPLIED WARRANTIES
|
---|
| 37 | * OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE ARE
|
---|
| 38 | * DISCLAIMED. IN NO EVENT SHALL THE APACHE SOFTWARE FOUNDATION OR
|
---|
| 39 | * ITS CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL,
|
---|
| 40 | * SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT
|
---|
| 41 | * LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF
|
---|
| 42 | * USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND
|
---|
| 43 | * ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY,
|
---|
| 44 | * OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT
|
---|
| 45 | * OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
|
---|
| 46 | * SUCH DAMAGE.
|
---|
| 47 | * ====================================================================
|
---|
| 48 | *
|
---|
| 49 | * This software consists of voluntary contributions made by many
|
---|
| 50 | * individuals on behalf of the Apache Software Foundation, and was
|
---|
| 51 | * originally based on software copyright (c) 1999, International
|
---|
| 52 | * Business Machines, Inc., http://www.ibm.com . For more information
|
---|
| 53 | * on the Apache Software Foundation, please see
|
---|
| 54 | * <http://www.apache.org/>.
|
---|
| 55 | */
|
---|
| 56 |
|
---|
| 57 | /*
|
---|
| 58 | * $Log: IGXMLScanner.hpp,v $
|
---|
| 59 | * Revision 1.18 2004/01/29 11:46:30 cargilld
|
---|
| 60 | * Code cleanup changes to get rid of various compiler diagnostic messages.
|
---|
| 61 | *
|
---|
| 62 | * Revision 1.17 2003/11/28 19:54:31 knoaman
|
---|
| 63 | * PSVIElement update
|
---|
| 64 | *
|
---|
| 65 | * Revision 1.16 2003/11/28 05:13:29 neilg
|
---|
| 66 | * Fix state-ful duplicate attribute detection when the integrated
|
---|
| 67 | * scanner is in use and namespaces are off. Also, implement
|
---|
| 68 | * change to PSVIHandler interface to remove prefix passing.
|
---|
| 69 | *
|
---|
| 70 | * Revision 1.15 2003/11/27 22:52:37 knoaman
|
---|
| 71 | * PSVIElement implementation
|
---|
| 72 | *
|
---|
| 73 | * Revision 1.14 2003/11/27 06:10:31 neilg
|
---|
| 74 | * PSVIAttribute implementation
|
---|
| 75 | *
|
---|
| 76 | * Revision 1.13 2003/11/26 16:20:00 knoaman
|
---|
| 77 | * Store XSModel.
|
---|
| 78 | *
|
---|
| 79 | * Revision 1.12 2003/11/24 05:09:38 neilg
|
---|
| 80 | * implement new, statless, method for detecting duplicate attributes
|
---|
| 81 | *
|
---|
| 82 | * Revision 1.11 2003/10/22 20:22:30 knoaman
|
---|
| 83 | * Prepare for annotation support.
|
---|
| 84 | *
|
---|
| 85 | * Revision 1.10 2003/09/22 19:51:41 neilg
|
---|
| 86 | * scanners should maintain their own pools of undeclared elements, rather than requiring grammars to do this. This makes grammar objects stateless with regard to validation.
|
---|
| 87 | *
|
---|
| 88 | * Revision 1.9 2003/08/14 02:56:41 knoaman
|
---|
| 89 | * Code refactoring to improve performance of validation.
|
---|
| 90 | *
|
---|
| 91 | * Revision 1.8 2003/07/10 19:47:23 peiyongz
|
---|
| 92 | * Stateless Grammar: Initialize scanner with grammarResolver,
|
---|
| 93 | * creating grammar through grammarPool
|
---|
| 94 | *
|
---|
| 95 | * Revision 1.7 2003/05/22 02:10:51 knoaman
|
---|
| 96 | * Default the memory manager.
|
---|
| 97 | *
|
---|
| 98 | * Revision 1.6 2003/05/15 18:26:29 knoaman
|
---|
| 99 | * Partial implementation of the configurable memory manager.
|
---|
| 100 | *
|
---|
| 101 | * Revision 1.5 2003/03/07 18:08:58 tng
|
---|
| 102 | * Return a reference instead of void for operator=
|
---|
| 103 | *
|
---|
| 104 | * Revision 1.4 2003/01/29 19:59:35 gareth
|
---|
| 105 | * we now detect when elements and attributes are validated/ the result of the validation and store that information.
|
---|
| 106 | *
|
---|
| 107 | * Revision 1.3 2003/01/15 15:49:49 knoaman
|
---|
| 108 | * Change constant declaration name to match its value.
|
---|
| 109 | *
|
---|
| 110 | * Revision 1.2 2003/01/13 16:30:18 knoaman
|
---|
| 111 | * [Bug 14469] Validator doesn't enforce xsd:key.
|
---|
| 112 | *
|
---|
| 113 | * Revision 1.1 2002/12/04 02:05:25 knoaman
|
---|
| 114 | * Initial checkin.
|
---|
| 115 | *
|
---|
| 116 | */
|
---|
| 117 |
|
---|
| 118 |
|
---|
| 119 | #if !defined(IGXMLSCANNER_HPP)
|
---|
| 120 | #define IGXMLSCANNER_HPP
|
---|
| 121 |
|
---|
| 122 | #include <xercesc/internal/XMLScanner.hpp>
|
---|
| 123 | #include <xercesc/util/KVStringPair.hpp>
|
---|
| 124 | #include <xercesc/util/NameIdPool.hpp>
|
---|
| 125 | #include <xercesc/util/RefHash3KeysIdPool.hpp>
|
---|
| 126 | #include <xercesc/validators/common/Grammar.hpp>
|
---|
| 127 | #include <xercesc/validators/schema/SchemaElementDecl.hpp>
|
---|
| 128 |
|
---|
| 129 | XERCES_CPP_NAMESPACE_BEGIN
|
---|
| 130 |
|
---|
| 131 | class DTDElementDecl;
|
---|
| 132 | class DTDGrammar;
|
---|
| 133 | class DTDValidator;
|
---|
| 134 | class SchemaValidator;
|
---|
| 135 | class ValueStoreCache;
|
---|
| 136 | class XPathMatcherStack;
|
---|
| 137 | class FieldActivator;
|
---|
| 138 | class IdentityConstraint;
|
---|
| 139 | class ContentLeafNameTypeVector;
|
---|
| 140 | class SchemaAttDef;
|
---|
| 141 | class XMLContentModel;
|
---|
| 142 | class XSModel;
|
---|
| 143 | class PSVIAttributeList;
|
---|
| 144 | class PSVIElement;
|
---|
| 145 |
|
---|
| 146 | // This is an integrated scanner class, which does DTD/XML Schema grammar
|
---|
| 147 | // processing.
|
---|
| 148 | class XMLPARSER_EXPORT IGXMLScanner : public XMLScanner
|
---|
| 149 | {
|
---|
| 150 | public :
|
---|
| 151 | // -----------------------------------------------------------------------
|
---|
| 152 | // Constructors and Destructor
|
---|
| 153 | // -----------------------------------------------------------------------
|
---|
| 154 | IGXMLScanner
|
---|
| 155 | (
|
---|
| 156 | XMLValidator* const valToAdopt
|
---|
| 157 | , GrammarResolver* const grammarResolver
|
---|
| 158 | , MemoryManager* const manager = XMLPlatformUtils::fgMemoryManager
|
---|
| 159 | );
|
---|
| 160 | IGXMLScanner
|
---|
| 161 | (
|
---|
| 162 | XMLDocumentHandler* const docHandler
|
---|
| 163 | , DocTypeHandler* const docTypeHandler
|
---|
| 164 | , XMLEntityHandler* const entityHandler
|
---|
| 165 | , XMLErrorReporter* const errReporter
|
---|
| 166 | , XMLValidator* const valToAdopt
|
---|
| 167 | , GrammarResolver* const grammarResolver
|
---|
| 168 | , MemoryManager* const manager = XMLPlatformUtils::fgMemoryManager
|
---|
| 169 | );
|
---|
| 170 | virtual ~IGXMLScanner();
|
---|
| 171 |
|
---|
| 172 | // -----------------------------------------------------------------------
|
---|
| 173 | // XMLScanner public virtual methods
|
---|
| 174 | // -----------------------------------------------------------------------
|
---|
| 175 | virtual const XMLCh* getName() const;
|
---|
| 176 | virtual NameIdPool<DTDEntityDecl>* getEntityDeclPool();
|
---|
| 177 | virtual const NameIdPool<DTDEntityDecl>* getEntityDeclPool() const;
|
---|
| 178 | virtual unsigned int resolveQName
|
---|
| 179 | (
|
---|
| 180 | const XMLCh* const qName
|
---|
| 181 | , XMLBuffer& prefixBufToFill
|
---|
| 182 | , const short mode
|
---|
| 183 | , int& prefixColonPos
|
---|
| 184 | );
|
---|
| 185 | virtual void scanDocument
|
---|
| 186 | (
|
---|
| 187 | const InputSource& src
|
---|
| 188 | );
|
---|
| 189 | virtual bool scanNext(XMLPScanToken& toFill);
|
---|
| 190 | virtual Grammar* loadGrammar
|
---|
| 191 | (
|
---|
| 192 | const InputSource& src
|
---|
| 193 | , const short grammarType
|
---|
| 194 | , const bool toCache = false
|
---|
| 195 | );
|
---|
| 196 |
|
---|
| 197 | private :
|
---|
| 198 | // -----------------------------------------------------------------------
|
---|
| 199 | // Unimplemented constructors and operators
|
---|
| 200 | // -----------------------------------------------------------------------
|
---|
| 201 | IGXMLScanner();
|
---|
| 202 | IGXMLScanner(const IGXMLScanner&);
|
---|
| 203 | IGXMLScanner& operator=(const IGXMLScanner&);
|
---|
| 204 |
|
---|
| 205 | // -----------------------------------------------------------------------
|
---|
| 206 | // XMLScanner virtual methods
|
---|
| 207 | // -----------------------------------------------------------------------
|
---|
| 208 | virtual void scanCDSection();
|
---|
| 209 | virtual void scanCharData(XMLBuffer& toToUse);
|
---|
| 210 | virtual EntityExpRes scanEntityRef
|
---|
| 211 | (
|
---|
| 212 | const bool inAttVal
|
---|
| 213 | , XMLCh& firstCh
|
---|
| 214 | , XMLCh& secondCh
|
---|
| 215 | , bool& escaped
|
---|
| 216 | );
|
---|
| 217 | virtual void scanDocTypeDecl();
|
---|
| 218 | virtual void scanReset(const InputSource& src);
|
---|
| 219 | virtual void sendCharData(XMLBuffer& toSend);
|
---|
| 220 |
|
---|
| 221 | // -----------------------------------------------------------------------
|
---|
| 222 | // Private helper methods
|
---|
| 223 | // -----------------------------------------------------------------------
|
---|
| 224 | void commonInit();
|
---|
| 225 | void cleanUp();
|
---|
| 226 | InputSource* resolveSystemId(const XMLCh* const sysId); // return owned by caller
|
---|
| 227 |
|
---|
| 228 | // Spaces are not allowed in URI, so %20 is used instead.
|
---|
| 229 | // Convert %20 to spaces before resolving the URI
|
---|
| 230 | void normalizeURI(const XMLCh* const systemURI, XMLBuffer& normalizedURI);
|
---|
| 231 |
|
---|
| 232 | unsigned int buildAttList
|
---|
| 233 | (
|
---|
| 234 | const RefVectorOf<KVStringPair>& providedAttrs
|
---|
| 235 | , const unsigned int attCount
|
---|
| 236 | , XMLElementDecl* elemDecl
|
---|
| 237 | , RefVectorOf<XMLAttr>& toFill
|
---|
| 238 | );
|
---|
| 239 | bool normalizeAttValue
|
---|
| 240 | (
|
---|
| 241 | const XMLAttDef* const attDef
|
---|
| 242 | , const XMLCh* const name
|
---|
| 243 | , const XMLCh* const value
|
---|
| 244 | , XMLBuffer& toFill
|
---|
| 245 | );
|
---|
| 246 | bool normalizeAttRawValue
|
---|
| 247 | (
|
---|
| 248 | const XMLCh* const attrName
|
---|
| 249 | , const XMLCh* const value
|
---|
| 250 | , XMLBuffer& toFill
|
---|
| 251 | );
|
---|
| 252 | unsigned int resolvePrefix
|
---|
| 253 | (
|
---|
| 254 | const XMLCh* const prefix
|
---|
| 255 | , const ElemStack::MapModes mode
|
---|
| 256 | );
|
---|
| 257 | unsigned int resolvePrefix
|
---|
| 258 | (
|
---|
| 259 | const XMLCh* const prefix
|
---|
| 260 | , XMLBuffer& uriBufToFill
|
---|
| 261 | , const ElemStack::MapModes mode
|
---|
| 262 | );
|
---|
| 263 | void updateNSMap
|
---|
| 264 | (
|
---|
| 265 | const XMLCh* const attrName
|
---|
| 266 | , const XMLCh* const attrValue
|
---|
| 267 | );
|
---|
| 268 | void scanRawAttrListforNameSpaces(int attCount);
|
---|
| 269 | void parseSchemaLocation(const XMLCh* const schemaLocationStr);
|
---|
| 270 | void resolveSchemaGrammar(const XMLCh* const loc, const XMLCh* const uri);
|
---|
| 271 | bool switchGrammar(const XMLCh* const newGrammarNameSpace);
|
---|
| 272 | bool laxElementValidation(QName* element, ContentLeafNameTypeVector* cv,
|
---|
| 273 | const XMLContentModel* const cm,
|
---|
| 274 | const unsigned int parentElemDepth);
|
---|
| 275 | bool anyAttributeValidation(SchemaAttDef* attWildCard,
|
---|
| 276 | unsigned int uriId,
|
---|
| 277 | bool& skipThisOne,
|
---|
| 278 | bool& laxThisOne);
|
---|
| 279 | void resizeElemState();
|
---|
| 280 | void processSchemaLocation(XMLCh* const schemaLoc);
|
---|
| 281 |
|
---|
| 282 | // -----------------------------------------------------------------------
|
---|
| 283 | // Private scanning methods
|
---|
| 284 | // -----------------------------------------------------------------------
|
---|
| 285 | bool basicAttrValueScan
|
---|
| 286 | (
|
---|
| 287 | const XMLCh* const attrName
|
---|
| 288 | , XMLBuffer& toFill
|
---|
| 289 | );
|
---|
| 290 | unsigned int rawAttrScan
|
---|
| 291 | (
|
---|
| 292 | const XMLCh* const elemName
|
---|
| 293 | , RefVectorOf<KVStringPair>& toFill
|
---|
| 294 | , bool& isEmpty
|
---|
| 295 | );
|
---|
| 296 | bool scanAttValue
|
---|
| 297 | (
|
---|
| 298 | const XMLAttDef* const attDef
|
---|
| 299 | , const XMLCh* const attrName
|
---|
| 300 | , XMLBuffer& toFill
|
---|
| 301 | );
|
---|
| 302 | bool scanContent();
|
---|
| 303 | void scanEndTag(bool& gotData);
|
---|
| 304 | bool scanStartTag(bool& gotData);
|
---|
| 305 | bool scanStartTagNS(bool& gotData);
|
---|
| 306 |
|
---|
| 307 | // -----------------------------------------------------------------------
|
---|
| 308 | // IdentityConstraints Activation methods
|
---|
| 309 | // -----------------------------------------------------------------------
|
---|
| 310 | void activateSelectorFor(IdentityConstraint* const ic, const int initialDepth);
|
---|
| 311 |
|
---|
| 312 | // -----------------------------------------------------------------------
|
---|
| 313 | // Grammar preparsing methods
|
---|
| 314 | // -----------------------------------------------------------------------
|
---|
| 315 | Grammar* loadXMLSchemaGrammar(const InputSource& src, const bool toCache = false);
|
---|
| 316 | Grammar* loadDTDGrammar(const InputSource& src, const bool toCache = false);
|
---|
| 317 |
|
---|
| 318 | // -----------------------------------------------------------------------
|
---|
| 319 | // PSVI handling methods
|
---|
| 320 | // -----------------------------------------------------------------------
|
---|
| 321 | void endElementPSVI(SchemaElementDecl* const elemDecl,
|
---|
| 322 | DatatypeValidator* const memberDV);
|
---|
| 323 | void resetPSVIElemContext();
|
---|
| 324 |
|
---|
| 325 | // -----------------------------------------------------------------------
|
---|
| 326 | // Data members
|
---|
| 327 | //
|
---|
| 328 | // fRawAttrList
|
---|
| 329 | // During the initial scan of the attributes we can only do a raw
|
---|
| 330 | // scan for key/value pairs. So this vector is used to store them
|
---|
| 331 | // until they can be processed (and put into fAttrList.)
|
---|
| 332 | //
|
---|
| 333 | // fDTDValidator
|
---|
| 334 | // The DTD validator instance.
|
---|
| 335 | //
|
---|
| 336 | // fSchemaValidator
|
---|
| 337 | // The Schema validator instance.
|
---|
| 338 | //
|
---|
| 339 | // fSeeXsi
|
---|
| 340 | // This flag indicates a schema has been seen.
|
---|
| 341 | //
|
---|
| 342 | // fElemState
|
---|
| 343 | // fElemStateSize
|
---|
| 344 | // Stores an element next state from DFA content model - used for
|
---|
| 345 | // wildcard validation
|
---|
| 346 | //
|
---|
| 347 | // fMatcherStack
|
---|
| 348 | // Stack of active XPath matchers for identity constraints. All
|
---|
| 349 | // active XPath matchers are notified of startElement, characters
|
---|
| 350 | // and endElement callbacks in order to perform their matches.
|
---|
| 351 | //
|
---|
| 352 | // fValueStoreCache
|
---|
| 353 | // Cache of value stores for identity constraint fields.
|
---|
| 354 | //
|
---|
| 355 | // fFieldActivator
|
---|
| 356 | // Activates fields within a certain scope when a selector matches
|
---|
| 357 | // its xpath.
|
---|
| 358 | // fDTDElemNonDeclPool
|
---|
| 359 | // registry of "faulted-in" DTD element decls
|
---|
| 360 | // fSchemaElemNonDeclPool
|
---|
| 361 | // registry for elements without decls in the grammar
|
---|
| 362 | // fElemCount
|
---|
| 363 | // count of the number of start tags seen so far (starts at 1).
|
---|
| 364 | // Used for duplicate attribute detection/processing of required/defaulted attributes
|
---|
| 365 | // fAttDefRegistry
|
---|
| 366 | // mapping from XMLAttDef instances to the count of the last
|
---|
| 367 | // start tag where they were utilized.
|
---|
| 368 | // fUndeclaredAttrRegistry
|
---|
| 369 | // mapping of attr QNames to the count of the last start tag in which they occurred
|
---|
| 370 | // fUndeclaredAttrRegistryNS
|
---|
| 371 | // mapping of namespaceId/localName pairs to the count of the last
|
---|
| 372 | // start tag in which they occurred.
|
---|
| 373 | // fPSVIAttrList
|
---|
| 374 | // PSVI attribute list implementation that needs to be
|
---|
| 375 | // filled when a PSVIHandler is registered
|
---|
| 376 | //
|
---|
| 377 | // -----------------------------------------------------------------------
|
---|
| 378 | bool fSeeXsi;
|
---|
| 379 | Grammar::GrammarType fGrammarType;
|
---|
| 380 | unsigned int fElemStateSize;
|
---|
| 381 | unsigned int* fElemState;
|
---|
| 382 | XMLBuffer fContent;
|
---|
| 383 | RefVectorOf<KVStringPair>* fRawAttrList;
|
---|
| 384 | DTDValidator* fDTDValidator;
|
---|
| 385 | SchemaValidator* fSchemaValidator;
|
---|
| 386 | DTDGrammar* fDTDGrammar;
|
---|
| 387 | XPathMatcherStack* fMatcherStack;
|
---|
| 388 | ValueStoreCache* fValueStoreCache;
|
---|
| 389 | FieldActivator* fFieldActivator;
|
---|
| 390 | ValueVectorOf<XMLCh*>* fLocationPairs;
|
---|
| 391 | NameIdPool<DTDElementDecl>* fDTDElemNonDeclPool;
|
---|
| 392 | RefHash3KeysIdPool<SchemaElementDecl>* fSchemaElemNonDeclPool;
|
---|
| 393 | unsigned int fElemCount;
|
---|
| 394 | RefHashTableOf<unsigned int>* fAttDefRegistry;
|
---|
| 395 | RefHashTableOf<unsigned int>* fUndeclaredAttrRegistry;
|
---|
| 396 | RefHash2KeysTableOf<unsigned int>* fUndeclaredAttrRegistryNS;
|
---|
| 397 | PSVIAttributeList * fPSVIAttrList;
|
---|
| 398 | XSModel* fModel;
|
---|
| 399 | PSVIElement* fPSVIElement;
|
---|
| 400 | ValueStackOf<bool>* fErrorStack;
|
---|
| 401 | PSVIElemContext fPSVIElemContext;
|
---|
| 402 | };
|
---|
| 403 |
|
---|
| 404 | inline const XMLCh* IGXMLScanner::getName() const
|
---|
| 405 | {
|
---|
| 406 | return XMLUni::fgIGXMLScanner;
|
---|
| 407 | }
|
---|
| 408 |
|
---|
| 409 |
|
---|
| 410 | XERCES_CPP_NAMESPACE_END
|
---|
| 411 |
|
---|
| 412 | #endif
|
---|