[188] | 1 | /*
|
---|
| 2 | * The Apache Software License, Version 1.1
|
---|
| 3 | *
|
---|
| 4 | * Copyright (c) 1999-2003 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: DTDScanner.hpp,v $
|
---|
| 59 | * Revision 1.10 2004/01/29 11:52:30 cargilld
|
---|
| 60 | * Code cleanup changes to get rid of various compiler diagnostic messages.
|
---|
| 61 | *
|
---|
| 62 | * Revision 1.9 2003/07/10 19:50:12 peiyongz
|
---|
| 63 | * Stateless Grammar: create grammar components with grammarPool's memory Manager
|
---|
| 64 | *
|
---|
| 65 | * Revision 1.8 2003/05/16 21:43:19 knoaman
|
---|
| 66 | * Memory manager implementation: Modify constructors to pass in the memory manager.
|
---|
| 67 | *
|
---|
| 68 | * Revision 1.7 2003/05/15 18:54:51 knoaman
|
---|
| 69 | * Partial implementation of the configurable memory manager.
|
---|
| 70 | *
|
---|
| 71 | * Revision 1.6 2003/02/05 22:07:09 tng
|
---|
| 72 | * [Bug 3111] Problem with LexicalHandler::startDTD() and LexicalHandler::endDTD().
|
---|
| 73 | *
|
---|
| 74 | * Revision 1.5 2002/12/04 02:47:25 knoaman
|
---|
| 75 | * scanner re-organization.
|
---|
| 76 | *
|
---|
| 77 | * Revision 1.4 2002/11/04 14:50:40 tng
|
---|
| 78 | * C++ Namespace Support.
|
---|
| 79 | *
|
---|
| 80 | * Revision 1.3 2002/07/11 18:39:48 knoaman
|
---|
| 81 | * Access entities through the DTDGrammar instead of the scanner.
|
---|
| 82 | *
|
---|
| 83 | * Revision 1.2 2002/05/30 16:17:19 tng
|
---|
| 84 | * Add feature to optionally ignore external DTD.
|
---|
| 85 | *
|
---|
| 86 | * Revision 1.1.1.1 2002/02/01 22:22:44 peiyongz
|
---|
| 87 | * sane_include
|
---|
| 88 | *
|
---|
| 89 | * Revision 1.4 2001/06/21 14:25:56 knoaman
|
---|
| 90 | * Fix for bug 1946
|
---|
| 91 | *
|
---|
| 92 | * Revision 1.3 2001/05/11 13:27:10 tng
|
---|
| 93 | * Copyright update.
|
---|
| 94 | *
|
---|
| 95 | * Revision 1.2 2001/04/19 18:17:22 tng
|
---|
| 96 | * Schema: SchemaValidator update, and use QName in Content Model
|
---|
| 97 | *
|
---|
| 98 | * Revision 1.1 2001/03/21 21:56:20 tng
|
---|
| 99 | * Schema: Add Schema Grammar, Schema Validator, and split the DTDValidator into DTDValidator, DTDScanner, and DTDGrammar.
|
---|
| 100 | *
|
---|
| 101 | */
|
---|
| 102 |
|
---|
| 103 |
|
---|
| 104 |
|
---|
| 105 | #if !defined(DTDSCANNER_HPP)
|
---|
| 106 | #define DTDSCANNER_HPP
|
---|
| 107 |
|
---|
| 108 | #include <xercesc/validators/DTD/DTDGrammar.hpp>
|
---|
| 109 | #include <xercesc/validators/DTD/DTDEntityDecl.hpp>
|
---|
| 110 |
|
---|
| 111 | XERCES_CPP_NAMESPACE_BEGIN
|
---|
| 112 |
|
---|
| 113 | class XMLScanner;
|
---|
| 114 |
|
---|
| 115 | /*
|
---|
| 116 | * Default implementation of an XML DTD scanner.
|
---|
| 117 | */
|
---|
| 118 | class DocTypeHandler;
|
---|
| 119 |
|
---|
| 120 | class VALIDATORS_EXPORT DTDScanner : public XMemory
|
---|
| 121 | {
|
---|
| 122 | public:
|
---|
| 123 | // -----------------------------------------------------------------------
|
---|
| 124 | // Class specific types
|
---|
| 125 | //
|
---|
| 126 | // EntityExpRes
|
---|
| 127 | // Returned from scanEntityRef() to indicate how the expanded text
|
---|
| 128 | // was treated.
|
---|
| 129 | //
|
---|
| 130 | // IDTypes
|
---|
| 131 | // Type of the ID
|
---|
| 132 | // -----------------------------------------------------------------------
|
---|
| 133 | enum EntityExpRes
|
---|
| 134 | {
|
---|
| 135 | EntityExp_Failed
|
---|
| 136 | , EntityExp_Pushed
|
---|
| 137 | , EntityExp_Returned
|
---|
| 138 | };
|
---|
| 139 |
|
---|
| 140 | enum IDTypes
|
---|
| 141 | {
|
---|
| 142 | IDType_Public
|
---|
| 143 | , IDType_External
|
---|
| 144 | , IDType_Either
|
---|
| 145 | };
|
---|
| 146 |
|
---|
| 147 |
|
---|
| 148 |
|
---|
| 149 | // -----------------------------------------------------------------------
|
---|
| 150 | // Constructors and Destructor
|
---|
| 151 | // -----------------------------------------------------------------------
|
---|
| 152 | DTDScanner
|
---|
| 153 | (
|
---|
| 154 | DTDGrammar* dtdGrammar
|
---|
| 155 | , DocTypeHandler* const docTypeHandler
|
---|
| 156 | , MemoryManager* const grammarPoolMemoryManager
|
---|
| 157 | , MemoryManager* const manager = XMLPlatformUtils::fgMemoryManager
|
---|
| 158 | );
|
---|
| 159 | virtual ~DTDScanner();
|
---|
| 160 |
|
---|
| 161 | // -----------------------------------------------------------------------
|
---|
| 162 | // Getter methods
|
---|
| 163 | // -----------------------------------------------------------------------
|
---|
| 164 | DocTypeHandler* getDocTypeHandler();
|
---|
| 165 | const DocTypeHandler* getDocTypeHandler() const;
|
---|
| 166 |
|
---|
| 167 | // -----------------------------------------------------------------------
|
---|
| 168 | // Setter methods
|
---|
| 169 | //
|
---|
| 170 | // setScannerInfo() is called by the scanner to tell the DTDScanner
|
---|
| 171 | // about the stuff it needs to have access to.
|
---|
| 172 | // -----------------------------------------------------------------------
|
---|
| 173 | void setScannerInfo
|
---|
| 174 | (
|
---|
| 175 | XMLScanner* const owningScanner
|
---|
| 176 | , ReaderMgr* const readerMgr
|
---|
| 177 | , XMLBufferMgr* const bufMgr
|
---|
| 178 | );
|
---|
| 179 |
|
---|
| 180 | void setDocTypeHandler
|
---|
| 181 | (
|
---|
| 182 | DocTypeHandler* const handlerToSet
|
---|
| 183 | );
|
---|
| 184 |
|
---|
| 185 | void scanExtSubsetDecl(const bool inIncludeSect, const bool isDTD);
|
---|
| 186 | bool scanInternalSubset();
|
---|
| 187 | bool scanId
|
---|
| 188 | (
|
---|
| 189 | XMLBuffer& pubIdToFill
|
---|
| 190 | , XMLBuffer& sysIdToFill
|
---|
| 191 | , const IDTypes whatKind
|
---|
| 192 | );
|
---|
| 193 |
|
---|
| 194 | private:
|
---|
| 195 | // -----------------------------------------------------------------------
|
---|
| 196 | // Unimplemented constructors and operators
|
---|
| 197 | // -----------------------------------------------------------------------
|
---|
| 198 | DTDScanner(const DTDScanner &);
|
---|
| 199 | DTDScanner& operator = (const DTDScanner&);
|
---|
| 200 |
|
---|
| 201 | // -----------------------------------------------------------------------
|
---|
| 202 | // Private DTD scanning methods. These are all in XMLValidator2.cpp
|
---|
| 203 | // -----------------------------------------------------------------------
|
---|
| 204 | bool checkForPERef
|
---|
| 205 | (
|
---|
| 206 | const bool inLiteral
|
---|
| 207 | , const bool inMarkup
|
---|
| 208 | );
|
---|
| 209 | bool expandPERef
|
---|
| 210 | (
|
---|
| 211 | const bool scanExternal
|
---|
| 212 | , const bool inLiteral
|
---|
| 213 | , const bool inMarkup
|
---|
| 214 | , const bool throwEndOfExt = false
|
---|
| 215 | );
|
---|
| 216 | bool getQuotedString(XMLBuffer& toFill);
|
---|
| 217 | XMLAttDef* scanAttDef(DTDElementDecl& elemDecl, XMLBuffer& bufToUse);
|
---|
| 218 | bool scanAttValue
|
---|
| 219 | (
|
---|
| 220 | const XMLCh* const attrName
|
---|
| 221 | , XMLBuffer& toFill
|
---|
| 222 | , const XMLAttDef::AttTypes type
|
---|
| 223 | );
|
---|
| 224 | void scanAttListDecl();
|
---|
| 225 | ContentSpecNode* scanChildren
|
---|
| 226 | (
|
---|
| 227 | const DTDElementDecl& elemDecl
|
---|
| 228 | , XMLBuffer& bufToUse
|
---|
| 229 | );
|
---|
| 230 | bool scanCharRef(XMLCh& toFill, XMLCh& second);
|
---|
| 231 | void scanComment();
|
---|
| 232 | bool scanContentSpec(DTDElementDecl& toFill);
|
---|
| 233 | void scanDefaultDecl(DTDAttDef& toFill);
|
---|
| 234 | void scanElementDecl();
|
---|
| 235 | void scanEntityDecl();
|
---|
| 236 | bool scanEntityDef();
|
---|
| 237 | bool scanEntityLiteral(XMLBuffer& toFill);
|
---|
| 238 | bool scanEntityDef(DTDEntityDecl& decl, const bool isPEDecl);
|
---|
| 239 | EntityExpRes scanEntityRef(XMLCh& firstCh, XMLCh& secondCh, bool& escaped);
|
---|
| 240 | bool scanEnumeration
|
---|
| 241 | (
|
---|
| 242 | const DTDAttDef& attDef
|
---|
| 243 | , XMLBuffer& toFill
|
---|
| 244 | , const bool notation
|
---|
| 245 | );
|
---|
| 246 | bool scanEq();
|
---|
| 247 | void scanIgnoredSection();
|
---|
| 248 | void scanMarkupDecl(const bool parseTextDecl);
|
---|
| 249 | bool scanMixed(DTDElementDecl& toFill);
|
---|
| 250 | void scanNotationDecl();
|
---|
| 251 | void scanPI();
|
---|
| 252 | bool scanPublicLiteral(XMLBuffer& toFill);
|
---|
| 253 | bool scanSystemLiteral(XMLBuffer& toFill);
|
---|
| 254 | void scanTextDecl();
|
---|
| 255 | bool isReadingExternalEntity();
|
---|
| 256 |
|
---|
| 257 |
|
---|
| 258 | // -----------------------------------------------------------------------
|
---|
| 259 | // Private data members
|
---|
| 260 | //
|
---|
| 261 | // fDocTypeHandler
|
---|
| 262 | // This holds the optional doc type handler that can be installed
|
---|
| 263 | // and used to call back for all markup events. It is DTD specific.
|
---|
| 264 | //
|
---|
| 265 | // fDumAttDef
|
---|
| 266 | // fDumElemDecl
|
---|
| 267 | // fDumEntityDecl
|
---|
| 268 | // These are dummy objects into which mark decls are parsed when
|
---|
| 269 | // they are just overrides of previously declared markup decls. In
|
---|
| 270 | // such situations, the first one wins but we need to have somewhere
|
---|
| 271 | // to parse them into. So these are lazily created and used as needed
|
---|
| 272 | // when such markup decls are seen.
|
---|
| 273 | //
|
---|
| 274 | // fInternalSubset
|
---|
| 275 | // This is used to track whether we are in the internal subset or not,
|
---|
| 276 | // in which case we are in the external subset.
|
---|
| 277 | //
|
---|
| 278 | // fNextAttrId
|
---|
| 279 | // Since att defs are per-element, we don't have a validator wide
|
---|
| 280 | // attribute def pool. So we use a simpler data structure in each
|
---|
| 281 | // element decl to store its att defs, and we use this simple counter
|
---|
| 282 | // to apply a unique id to each new attribute.
|
---|
| 283 | //
|
---|
| 284 | // fDTDGrammar
|
---|
| 285 | // The DTD information we scanned like element decl, attribute decl
|
---|
| 286 | // are stored in this Grammar.
|
---|
| 287 | //
|
---|
| 288 | // fBufMgr
|
---|
| 289 | // This is the buffer manager of the scanner. This is provided as a
|
---|
| 290 | // convenience so that the DTDScanner doesn't have to create its own
|
---|
| 291 | // buffer manager during the parse process.
|
---|
| 292 | //
|
---|
| 293 | // fReaderMgr
|
---|
| 294 | // This is a pointer to the reader manager that is being used by the scanner.
|
---|
| 295 | //
|
---|
| 296 | // fScanner
|
---|
| 297 | // The pointer to the scanner to which this DTDScanner belongs
|
---|
| 298 | //
|
---|
| 299 | // fPEntityDeclPool
|
---|
| 300 | // This is a pool of EntityDecl objects, which contains all of the
|
---|
| 301 | // parameter entities that are declared in the DTD subsets.
|
---|
| 302 | //
|
---|
| 303 | // fEmptyNamespaceId
|
---|
| 304 | // The uri for all DTD decls
|
---|
| 305 | //
|
---|
| 306 | // fDocTypeReaderId
|
---|
| 307 | // The original reader in the fReaderMgr - to be compared against the
|
---|
| 308 | // current reader to decide whether we are processing an external/internal
|
---|
| 309 | // declaration
|
---|
| 310 | // -----------------------------------------------------------------------
|
---|
| 311 | MemoryManager* fMemoryManager;
|
---|
| 312 | MemoryManager* fGrammarPoolMemoryManager;
|
---|
| 313 | DocTypeHandler* fDocTypeHandler;
|
---|
| 314 | DTDAttDef* fDumAttDef;
|
---|
| 315 | DTDElementDecl* fDumElemDecl;
|
---|
| 316 | DTDEntityDecl* fDumEntityDecl;
|
---|
| 317 | bool fInternalSubset;
|
---|
| 318 | unsigned int fNextAttrId;
|
---|
| 319 | DTDGrammar* fDTDGrammar;
|
---|
| 320 | XMLBufferMgr* fBufMgr;
|
---|
| 321 | ReaderMgr* fReaderMgr;
|
---|
| 322 | XMLScanner* fScanner;
|
---|
| 323 | NameIdPool<DTDEntityDecl>* fPEntityDeclPool;
|
---|
| 324 | unsigned int fEmptyNamespaceId;
|
---|
| 325 | unsigned int fDocTypeReaderId;
|
---|
| 326 | };
|
---|
| 327 |
|
---|
| 328 |
|
---|
| 329 | // ---------------------------------------------------------------------------
|
---|
| 330 | // DTDScanner: Getter methods
|
---|
| 331 | // ---------------------------------------------------------------------------
|
---|
| 332 | inline DocTypeHandler* DTDScanner::getDocTypeHandler()
|
---|
| 333 | {
|
---|
| 334 | return fDocTypeHandler;
|
---|
| 335 | }
|
---|
| 336 |
|
---|
| 337 | inline const DocTypeHandler* DTDScanner::getDocTypeHandler() const
|
---|
| 338 | {
|
---|
| 339 | return fDocTypeHandler;
|
---|
| 340 | }
|
---|
| 341 |
|
---|
| 342 |
|
---|
| 343 | // ---------------------------------------------------------------------------
|
---|
| 344 | // DTDScanner: Setter methods
|
---|
| 345 | // ---------------------------------------------------------------------------
|
---|
| 346 | inline void DTDScanner::setDocTypeHandler(DocTypeHandler* const handlerToSet)
|
---|
| 347 | {
|
---|
| 348 | fDocTypeHandler = handlerToSet;
|
---|
| 349 | }
|
---|
| 350 |
|
---|
| 351 | // -----------------------------------------------------------------------
|
---|
| 352 | // Helper methods
|
---|
| 353 | // -----------------------------------------------------------------------
|
---|
| 354 | inline bool DTDScanner::isReadingExternalEntity() {
|
---|
| 355 | return (fDocTypeReaderId != fReaderMgr->getCurrentReaderNum());
|
---|
| 356 | }
|
---|
| 357 |
|
---|
| 358 | XERCES_CPP_NAMESPACE_END
|
---|
| 359 |
|
---|
| 360 | #endif
|
---|