[358] | 1 | /*
|
---|
| 2 | * Copyright 1999-2001,2004 The Apache Software Foundation.
|
---|
| 3 | *
|
---|
| 4 | * Licensed under the Apache License, Version 2.0 (the "License");
|
---|
| 5 | * you may not use this file except in compliance with the License.
|
---|
| 6 | * You may obtain a copy of the License at
|
---|
| 7 | *
|
---|
| 8 | * http://www.apache.org/licenses/LICENSE-2.0
|
---|
| 9 | *
|
---|
| 10 | * Unless required by applicable law or agreed to in writing, software
|
---|
| 11 | * distributed under the License is distributed on an "AS IS" BASIS,
|
---|
| 12 | * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
|
---|
| 13 | * See the License for the specific language governing permissions and
|
---|
| 14 | * limitations under the License.
|
---|
| 15 | */
|
---|
| 16 |
|
---|
| 17 |
|
---|
| 18 | /*
|
---|
| 19 | * $Log: MemParse.cpp,v $
|
---|
| 20 | * Revision 1.18 2004/09/08 13:55:32 peiyongz
|
---|
| 21 | * Apache License Version 2.0
|
---|
| 22 | *
|
---|
| 23 | * Revision 1.17 2004/09/02 14:59:28 cargilld
|
---|
| 24 | * Add OutOfMemoryException block to samples.
|
---|
| 25 | *
|
---|
| 26 | * Revision 1.16 2003/09/12 18:14:19 neilg
|
---|
| 27 | * enable MemParse to work on OS400; thanks to Jay Hansen.
|
---|
| 28 | *
|
---|
| 29 | * Revision 1.15 2003/08/07 21:21:38 neilg
|
---|
| 30 | * fix segmentation faults that may arise when the parser throws exceptions during document parsing. In general, XMLPlatformUtils::Terminate() should not be called from within a catch statement.
|
---|
| 31 | *
|
---|
| 32 | * Revision 1.14 2003/05/30 09:36:35 gareth
|
---|
| 33 | * Use new macros for iostream.h and std:: issues.
|
---|
| 34 | *
|
---|
| 35 | * Revision 1.13 2002/02/01 22:37:14 peiyongz
|
---|
| 36 | * sane_include
|
---|
| 37 | *
|
---|
| 38 | * Revision 1.12 2001/10/25 15:18:33 tng
|
---|
| 39 | * delete the parser before XMLPlatformUtils::Terminate.
|
---|
| 40 | *
|
---|
| 41 | * Revision 1.11 2001/10/19 18:56:08 tng
|
---|
| 42 | * Pulled the hardcoded "encoding" out of the document itself and made it a #define
|
---|
| 43 | * to make it easier to support other encodings. Patch from David McCreedy.
|
---|
| 44 | * And other modification for consistent help display and return code across samples.
|
---|
| 45 | *
|
---|
| 46 | * Revision 1.10 2001/08/01 19:11:01 tng
|
---|
| 47 | * Add full schema constraint checking flag to the samples and the parser.
|
---|
| 48 | *
|
---|
| 49 | * Revision 1.9 2001/05/11 13:24:55 tng
|
---|
| 50 | * Copyright update.
|
---|
| 51 | *
|
---|
| 52 | * Revision 1.8 2001/05/03 15:59:40 tng
|
---|
| 53 | * Schema: samples update with schema
|
---|
| 54 | *
|
---|
| 55 | * Revision 1.7 2000/09/11 18:43:48 aruna1
|
---|
| 56 | * OS390 related updates
|
---|
| 57 | *
|
---|
| 58 | * Revision 1.6 2000/03/02 19:53:42 roddey
|
---|
| 59 | * This checkin includes many changes done while waiting for the
|
---|
| 60 | * 1.1.0 code to be finished. I can't list them all here, but a list is
|
---|
| 61 | * available elsewhere.
|
---|
| 62 | *
|
---|
| 63 | * Revision 1.5 2000/02/11 02:37:01 abagchi
|
---|
| 64 | * Removed StrX::transcode
|
---|
| 65 | *
|
---|
| 66 | * Revision 1.4 2000/02/06 07:47:19 rahulj
|
---|
| 67 | * Year 2K copyright swat.
|
---|
| 68 | *
|
---|
| 69 | * Revision 1.3 2000/01/12 00:27:00 roddey
|
---|
| 70 | * Updates to work with the new URL and input source scheme.
|
---|
| 71 | *
|
---|
| 72 | * Revision 1.2 1999/11/20 01:09:55 rahulj
|
---|
| 73 | * Fixed usage message.
|
---|
| 74 | *
|
---|
| 75 | * Revision 1.1.1.1 1999/11/09 01:09:49 twl
|
---|
| 76 | * Initial checkin
|
---|
| 77 | *
|
---|
| 78 | * Revision 1.7 1999/11/08 20:43:36 rahul
|
---|
| 79 | * Swat for adding in Product name and CVS comment log variable.
|
---|
| 80 | *
|
---|
| 81 | */
|
---|
| 82 |
|
---|
| 83 |
|
---|
| 84 | /**
|
---|
| 85 | * This sample program illustrates how one can use a memory buffer as the
|
---|
| 86 | * input to parser. The memory buffer contains raw bytes representing XML
|
---|
| 87 | * statements.
|
---|
| 88 | *
|
---|
| 89 | * Look at the API documentation for 'MemBufInputSource' for more information
|
---|
| 90 | * on parameters to the constructor.
|
---|
| 91 | *
|
---|
| 92 | */
|
---|
| 93 |
|
---|
| 94 |
|
---|
| 95 | // ---------------------------------------------------------------------------
|
---|
| 96 | // Includes
|
---|
| 97 | // ---------------------------------------------------------------------------
|
---|
| 98 | #include <xercesc/parsers/SAXParser.hpp>
|
---|
| 99 | #include <xercesc/framework/MemBufInputSource.hpp>
|
---|
| 100 | #include "MemParse.hpp"
|
---|
| 101 | #include <xercesc/util/OutOfMemoryException.hpp>
|
---|
| 102 |
|
---|
| 103 | // ---------------------------------------------------------------------------
|
---|
| 104 | // Local const data
|
---|
| 105 | //
|
---|
| 106 | // gXMLInMemBuf
|
---|
| 107 | // Defines the memory buffer contents here which parsed by the XML
|
---|
| 108 | // parser. This is the cheap way to do it, instead of reading it from
|
---|
| 109 | // somewhere. For this demo, its fine.
|
---|
| 110 | //
|
---|
| 111 | // NOTE: If your encoding is not ascii you will need to change
|
---|
| 112 | // the MEMPARSE_ENCODING #define
|
---|
| 113 | //
|
---|
| 114 | // gMemBufId
|
---|
| 115 | // A simple name to give as the system id for the memory buffer. This
|
---|
| 116 | // just for indentification purposes in case of errors. Its not a real
|
---|
| 117 | // system id (and the parser knows that.)
|
---|
| 118 | // ---------------------------------------------------------------------------
|
---|
| 119 |
|
---|
| 120 | #ifndef MEMPARSE_ENCODING
|
---|
| 121 | #if defined(OS390)
|
---|
| 122 | #define MEMPARSE_ENCODING "ibm-1047-s390"
|
---|
| 123 | #elif defined(OS400)
|
---|
| 124 | #define MEMPARSE_ENCODING "ibm037"
|
---|
| 125 | #else
|
---|
| 126 | #define MEMPARSE_ENCODING "ascii"
|
---|
| 127 | #endif
|
---|
| 128 | #endif /* ifndef MEMPARSE_ENCODING */
|
---|
| 129 |
|
---|
| 130 | static const char* gXMLInMemBuf =
|
---|
| 131 | "\
|
---|
| 132 | <?xml version='1.0' encoding='" MEMPARSE_ENCODING "'?>\n\
|
---|
| 133 | <!DOCTYPE company [\n\
|
---|
| 134 | <!ELEMENT company (product,category,developedAt)>\n\
|
---|
| 135 | <!ELEMENT product (#PCDATA)>\n\
|
---|
| 136 | <!ELEMENT category (#PCDATA)>\n\
|
---|
| 137 | <!ATTLIST category idea CDATA #IMPLIED>\n\
|
---|
| 138 | <!ELEMENT developedAt (#PCDATA)>\n\
|
---|
| 139 | ]>\n\n\
|
---|
| 140 | <company>\n\
|
---|
| 141 | <product>XML4C</product>\n\
|
---|
| 142 | <category idea='great'>XML Parsing Tools</category>\n\
|
---|
| 143 | <developedAt>\n\
|
---|
| 144 | IBM Center for Java Technology, Silicon Valley, Cupertino, CA\n\
|
---|
| 145 | </developedAt>\n\
|
---|
| 146 | </company>\
|
---|
| 147 | ";
|
---|
| 148 |
|
---|
| 149 | static const char* gMemBufId = "prodInfo";
|
---|
| 150 |
|
---|
| 151 |
|
---|
| 152 |
|
---|
| 153 | // ---------------------------------------------------------------------------
|
---|
| 154 | // Local helper methods
|
---|
| 155 | // ---------------------------------------------------------------------------
|
---|
| 156 | void usage()
|
---|
| 157 | {
|
---|
| 158 | XERCES_STD_QUALIFIER cout << "\nUsage:\n"
|
---|
| 159 | " MemParse [options]\n\n"
|
---|
| 160 | "This program uses the SAX Parser to parse a memory buffer\n"
|
---|
| 161 | "containing XML statements, and reports the number of\n"
|
---|
| 162 | "elements and attributes found.\n\n"
|
---|
| 163 | "Options:\n"
|
---|
| 164 | " -v=xxx Validation scheme [always | never | auto*].\n"
|
---|
| 165 | " -n Enable namespace processing. Defaults to off.\n"
|
---|
| 166 | " -s Enable schema processing. Defaults to off.\n"
|
---|
| 167 | " -f Enable full schema constraint checking. Defaults to off.\n"
|
---|
| 168 | " -? Show this help.\n\n"
|
---|
| 169 | " * = Default if not provided explicitly.\n"
|
---|
| 170 | << XERCES_STD_QUALIFIER endl;
|
---|
| 171 | }
|
---|
| 172 |
|
---|
| 173 |
|
---|
| 174 | // ---------------------------------------------------------------------------
|
---|
| 175 | // Program entry point
|
---|
| 176 | // ---------------------------------------------------------------------------
|
---|
| 177 | int main(int argC, char* argV[])
|
---|
| 178 | {
|
---|
| 179 | // Initialize the XML4C2 system
|
---|
| 180 | try
|
---|
| 181 | {
|
---|
| 182 | XMLPlatformUtils::Initialize();
|
---|
| 183 | }
|
---|
| 184 | catch (const XMLException& toCatch)
|
---|
| 185 | {
|
---|
| 186 | XERCES_STD_QUALIFIER cerr << "Error during initialization! Message:\n"
|
---|
| 187 | << StrX(toCatch.getMessage()) << XERCES_STD_QUALIFIER endl;
|
---|
| 188 | return 1;
|
---|
| 189 | }
|
---|
| 190 |
|
---|
| 191 | SAXParser::ValSchemes valScheme = SAXParser::Val_Auto;
|
---|
| 192 | bool doNamespaces = false;
|
---|
| 193 | bool doSchema = false;
|
---|
| 194 | bool schemaFullChecking = false;
|
---|
| 195 |
|
---|
| 196 | int argInd;
|
---|
| 197 | for (argInd = 1; argInd < argC; argInd++)
|
---|
| 198 | {
|
---|
| 199 | // Break out on first parm not starting with a dash
|
---|
| 200 | if (argV[argInd][0] != '-')
|
---|
| 201 | {
|
---|
| 202 | usage();
|
---|
| 203 | XMLPlatformUtils::Terminate();
|
---|
| 204 | return 1;
|
---|
| 205 | }
|
---|
| 206 |
|
---|
| 207 | // Watch for special case help request
|
---|
| 208 | if (!strcmp(argV[argInd], "-?"))
|
---|
| 209 | {
|
---|
| 210 | usage();
|
---|
| 211 | XMLPlatformUtils::Terminate();
|
---|
| 212 | return 1;
|
---|
| 213 | }
|
---|
| 214 | else if (!strncmp(argV[argInd], "-v=", 3)
|
---|
| 215 | || !strncmp(argV[argInd], "-V=", 3))
|
---|
| 216 | {
|
---|
| 217 | const char* const parm = &argV[argInd][3];
|
---|
| 218 |
|
---|
| 219 | if (!strcmp(parm, "never"))
|
---|
| 220 | valScheme = SAXParser::Val_Never;
|
---|
| 221 | else if (!strcmp(parm, "auto"))
|
---|
| 222 | valScheme = SAXParser::Val_Auto;
|
---|
| 223 | else if (!strcmp(parm, "always"))
|
---|
| 224 | valScheme = SAXParser::Val_Always;
|
---|
| 225 | else
|
---|
| 226 | {
|
---|
| 227 | XERCES_STD_QUALIFIER cerr << "Unknown -v= value: " << parm << XERCES_STD_QUALIFIER endl;
|
---|
| 228 | return 2;
|
---|
| 229 | }
|
---|
| 230 | }
|
---|
| 231 | else if (!strcmp(argV[argInd], "-n")
|
---|
| 232 | || !strcmp(argV[argInd], "-N"))
|
---|
| 233 | {
|
---|
| 234 | doNamespaces = true;
|
---|
| 235 | }
|
---|
| 236 | else if (!strcmp(argV[argInd], "-s")
|
---|
| 237 | || !strcmp(argV[argInd], "-S"))
|
---|
| 238 | {
|
---|
| 239 | doSchema = true;
|
---|
| 240 | }
|
---|
| 241 | else if (!strcmp(argV[argInd], "-f")
|
---|
| 242 | || !strcmp(argV[argInd], "-F"))
|
---|
| 243 | {
|
---|
| 244 | schemaFullChecking = true;
|
---|
| 245 | }
|
---|
| 246 | else
|
---|
| 247 | {
|
---|
| 248 | XERCES_STD_QUALIFIER cerr << "Unknown option '" << argV[argInd]
|
---|
| 249 | << "', ignoring it\n" << XERCES_STD_QUALIFIER endl;
|
---|
| 250 | }
|
---|
| 251 | }
|
---|
| 252 |
|
---|
| 253 | //
|
---|
| 254 | // Create a SAX parser object. Then, according to what we were told on
|
---|
| 255 | // the command line, set it to validate or not.
|
---|
| 256 | //
|
---|
| 257 | SAXParser* parser = new SAXParser;
|
---|
| 258 | parser->setValidationScheme(valScheme);
|
---|
| 259 | parser->setDoNamespaces(doNamespaces);
|
---|
| 260 | parser->setDoSchema(doSchema);
|
---|
| 261 | parser->setValidationSchemaFullChecking(schemaFullChecking);
|
---|
| 262 |
|
---|
| 263 | //
|
---|
| 264 | // Create our SAX handler object and install it on the parser, as the
|
---|
| 265 | // document and error handlers.
|
---|
| 266 | //
|
---|
| 267 | MemParseHandlers handler;
|
---|
| 268 | parser->setDocumentHandler(&handler);
|
---|
| 269 | parser->setErrorHandler(&handler);
|
---|
| 270 |
|
---|
| 271 | //
|
---|
| 272 | // Create MemBufferInputSource from the buffer containing the XML
|
---|
| 273 | // statements.
|
---|
| 274 | //
|
---|
| 275 | // NOTE: We are using strlen() here, since we know that the chars in
|
---|
| 276 | // our hard coded buffer are single byte chars!!! The parameter wants
|
---|
| 277 | // the number of BYTES, not chars, so when you create a memory buffer
|
---|
| 278 | // give it the byte size (which just happens to be the same here.)
|
---|
| 279 | //
|
---|
| 280 | MemBufInputSource* memBufIS = new MemBufInputSource
|
---|
| 281 | (
|
---|
| 282 | (const XMLByte*)gXMLInMemBuf
|
---|
| 283 | , strlen(gXMLInMemBuf)
|
---|
| 284 | , gMemBufId
|
---|
| 285 | , false
|
---|
| 286 | );
|
---|
| 287 |
|
---|
| 288 | //
|
---|
| 289 | // Get the starting time and kick off the parse of the indicated
|
---|
| 290 | // file. Catch any exceptions that might propogate out of it.
|
---|
| 291 | //
|
---|
| 292 | unsigned long duration;
|
---|
| 293 | int errorCount = 0;
|
---|
| 294 | int errorCode = 0;
|
---|
| 295 | try
|
---|
| 296 | {
|
---|
| 297 | const unsigned long startMillis = XMLPlatformUtils::getCurrentMillis();
|
---|
| 298 | parser->parse(*memBufIS);
|
---|
| 299 | const unsigned long endMillis = XMLPlatformUtils::getCurrentMillis();
|
---|
| 300 | duration = endMillis - startMillis;
|
---|
| 301 | errorCount = parser->getErrorCount();
|
---|
| 302 | }
|
---|
| 303 | catch (const OutOfMemoryException&)
|
---|
| 304 | {
|
---|
| 305 | XERCES_STD_QUALIFIER cerr << "OutOfMemoryException" << XERCES_STD_QUALIFIER endl;
|
---|
| 306 | errorCode = 5;
|
---|
| 307 | }
|
---|
| 308 | catch (const XMLException& e)
|
---|
| 309 | {
|
---|
| 310 | XERCES_STD_QUALIFIER cerr << "\nError during parsing memory stream:\n"
|
---|
| 311 | << "Exception message is: \n"
|
---|
| 312 | << StrX(e.getMessage()) << "\n" << XERCES_STD_QUALIFIER endl;
|
---|
| 313 | errorCode = 4;
|
---|
| 314 | }
|
---|
| 315 | if(errorCode) {
|
---|
| 316 | XMLPlatformUtils::Terminate();
|
---|
| 317 | return errorCode;
|
---|
| 318 | }
|
---|
| 319 |
|
---|
| 320 | // Print out the stats that we collected and time taken.
|
---|
| 321 | if (!errorCount) {
|
---|
| 322 | XERCES_STD_QUALIFIER cout << "\nFinished parsing the memory buffer containing the following "
|
---|
| 323 | << "XML statements:\n\n"
|
---|
| 324 | << gXMLInMemBuf
|
---|
| 325 | << "\n\n\n"
|
---|
| 326 | << "Parsing took " << duration << " ms ("
|
---|
| 327 | << handler.getElementCount() << " elements, "
|
---|
| 328 | << handler.getAttrCount() << " attributes, "
|
---|
| 329 | << handler.getSpaceCount() << " spaces, "
|
---|
| 330 | << handler.getCharacterCount() << " characters).\n" << XERCES_STD_QUALIFIER endl;
|
---|
| 331 | }
|
---|
| 332 |
|
---|
| 333 | //
|
---|
| 334 | // Delete the parser itself. Must be done prior to calling Terminate, below.
|
---|
| 335 | //
|
---|
| 336 | delete parser;
|
---|
| 337 |
|
---|
| 338 | delete memBufIS;
|
---|
| 339 |
|
---|
| 340 | // And call the termination method
|
---|
| 341 | XMLPlatformUtils::Terminate();
|
---|
| 342 |
|
---|
| 343 | if (errorCount > 0)
|
---|
| 344 | return 4;
|
---|
| 345 | else
|
---|
| 346 | return 0;
|
---|
| 347 | }
|
---|
| 348 |
|
---|