1 // Copyright (c) 2001-2009 Nokia Corporation and/or its subsidiary(-ies). |
|
2 // All rights reserved. |
|
3 // This component and the accompanying materials are made available |
|
4 // under the terms of "Eclipse Public License v1.0" |
|
5 // which accompanies this distribution, and is available |
|
6 // at the URL "http://www.eclipse.org/legal/epl-v10.html". |
|
7 // |
|
8 // Initial Contributors: |
|
9 // Nokia Corporation - initial contribution. |
|
10 // |
|
11 // Contributors: |
|
12 // |
|
13 // Description: |
|
14 // @file |
|
15 // This file contains the definition of the generic CMDXMLParser class |
|
16 // which is responsible for creating a DOM structure |
|
17 // from a given XML file. |
|
18 // |
|
19 // |
|
20 |
|
21 #include <f32file.h> |
|
22 #include <utf.h> |
|
23 #include <bautils.h> |
|
24 |
|
25 #include <gmxmlconstants.h> |
|
26 #include <gmxmlcomposer.h> |
|
27 #include <gmxmlentityconverter.h> |
|
28 #include <gmxmlnode.h> |
|
29 #include <gmxmldocument.h> |
|
30 #include <gmxmlelement.h> |
|
31 #include <gmxmlparser.h> |
|
32 #include <gmxmlprocessinginstruction.h> |
|
33 #include <gmxmlcomment.h> |
|
34 #include <gmxmlcdatasection.h> |
|
35 #include <gmxmltext.h> |
|
36 |
|
37 #include "GMXMLFileDataSource.h" |
|
38 #ifdef SYMBIAN_ENABLE_SPLIT_HEADERS |
|
39 #include "gmxmldummydtd.h" |
|
40 #endif |
|
41 |
|
42 const TInt KGMXMLDefaultTextBufferSize = 1024; |
|
43 const TInt KUTF8EdgeBufferLen = 6; |
|
44 |
|
45 // |
|
46 // Global functions // |
|
47 // |
|
48 |
|
49 |
|
50 //================================================================================== |
|
51 |
|
52 // |
|
53 //CMDXParser // |
|
54 // |
|
55 //================================================================================== |
|
56 |
|
57 void CMDXMLParser::Panic(TPanicCode aReason) const |
|
58 { |
|
59 _LIT(KClassName, "CMDXMLParser"); |
|
60 User::Panic(KClassName, aReason); |
|
61 } |
|
62 |
|
63 EXPORT_C CMDXMLParser* CMDXMLParser::NewL(MMDXMLParserObserver* aParserObserver) |
|
64 // |
|
65 // Two phase static factory function constructor |
|
66 // @return Created CMDXMLParser |
|
67 // @leave can Leave due to OOM |
|
68 // |
|
69 { |
|
70 CMDXMLParser* self = NewLC(aParserObserver); |
|
71 CleanupStack::Pop(); |
|
72 return self; |
|
73 } |
|
74 |
|
75 EXPORT_C CMDXMLParser* CMDXMLParser::NewL(MMDXMLParserObserver* aParserObserver, MXMLDtd* aDtdRepresentation) |
|
76 // |
|
77 // Two phase static factory function constructor |
|
78 // @return Created CMDXMLParser |
|
79 // @param aDtdRepresentation specid DTD represention class to be used for validation |
|
80 // @leave can Leave due to OOM |
|
81 // |
|
82 { |
|
83 CMDXMLParser* self = NewLC(aParserObserver, aDtdRepresentation); |
|
84 CleanupStack::Pop(); |
|
85 return self; |
|
86 } |
|
87 |
|
88 |
|
89 //================================================================================== |
|
90 |
|
91 EXPORT_C CMDXMLParser* CMDXMLParser::NewLC(MMDXMLParserObserver* aParserObserver) |
|
92 // |
|
93 // Two phase static factory function constructor |
|
94 // @return Created CMDXMLParser |
|
95 // @leave can Leave due to OOM |
|
96 // |
|
97 { |
|
98 CMDXMLParser* self = new (ELeave) CMDXMLParser(aParserObserver); |
|
99 CleanupStack::PushL(self); |
|
100 // This overload of NewLC doesn't take a MXMLDtd*, but we need to provide one to |
|
101 // ConstructL where ownership is taken if we do have one. Just pass NULL. |
|
102 self->ConstructL(NULL); |
|
103 return self; |
|
104 } |
|
105 |
|
106 EXPORT_C CMDXMLParser* CMDXMLParser::NewLC(MMDXMLParserObserver* aParserObserver, MXMLDtd* aDtdRepresentation) |
|
107 // |
|
108 // Two phase static factory function constructor |
|
109 // @return Created CMDXMLParser |
|
110 // @param aDtdRepresentation specid DTD represention class to be used for validation |
|
111 // @leave can Leave due to OOM |
|
112 // |
|
113 { |
|
114 CMDXMLParser* self = new (ELeave) CMDXMLParser(aParserObserver); |
|
115 CleanupStack::PushL(self); |
|
116 self->ConstructL(aDtdRepresentation); |
|
117 return self; |
|
118 } |
|
119 |
|
120 |
|
121 //================================================================================== |
|
122 |
|
123 void CMDXMLParser::ConstructL(MXMLDtd* aDtdRepresentation) |
|
124 // |
|
125 // Second stage constructor |
|
126 // @param aDtdRepresentation The DTD to be used for validation |
|
127 // @leave can Leave due to OOM |
|
128 // |
|
129 { |
|
130 CMDXMLEntityConverter* entityConverter = new(ELeave) CMDXMLEntityConverter(); |
|
131 SetEntityConverter(entityConverter); |
|
132 |
|
133 // This doesn't leave, but if CMDXMLParser::NewLC() Leaves after taking ownership |
|
134 // of this we'll get a double deletion as the caller will have pushed |
|
135 // aDtdRepresentation onto the CleanupStack. As such we can only take ownership |
|
136 // once we are sure we aren't going to leave. |
|
137 iDtdRepresentation = aDtdRepresentation; |
|
138 } |
|
139 |
|
140 CMDXMLParser::CMDXMLParser(MMDXMLParserObserver* aParserObserver) |
|
141 : CActive(EPriorityNormal) |
|
142 // |
|
143 // Constructor |
|
144 // |
|
145 { |
|
146 iParserObserver = aParserObserver; |
|
147 iStoreInvalid = ETrue; |
|
148 CActiveScheduler::Add(this); |
|
149 } |
|
150 |
|
151 //================================================================================== |
|
152 |
|
153 EXPORT_C CMDXMLParser::~CMDXMLParser() |
|
154 { |
|
155 Cancel(); |
|
156 delete iBomBuffer; |
|
157 if( iFileSource == NULL ) |
|
158 { |
|
159 // iFileSource has not been allocated yet, so the file path or file |
|
160 // handle are still owned |
|
161 if( iFileToParse!=NULL ) |
|
162 delete iFileToParse; |
|
163 else |
|
164 iFileHandleToParse.Close(); |
|
165 } |
|
166 else |
|
167 { |
|
168 delete iFileSource; |
|
169 } |
|
170 delete iUTF8EdgeBuffer; |
|
171 delete iXMLDoc; |
|
172 delete iEntityConverter; |
|
173 delete iElementTag; |
|
174 delete iDtdRepresentation; |
|
175 delete iText; |
|
176 } |
|
177 |
|
178 //================================================================================== |
|
179 //Defect fix for INC036136- Enable the use of custom entity converters in GMXML |
|
180 EXPORT_C void CMDXMLParser::SetEntityConverter(CMDXMLEntityConverter* aEntityConverter) |
|
181 /* |
|
182 * Sets the entity converter to be used |
|
183 * and take ownership of the passed entity converter |
|
184 * @param aEntityConverter The entity converter to be used |
|
185 */ |
|
186 { |
|
187 delete iEntityConverter; |
|
188 iEntityConverter = aEntityConverter; |
|
189 } |
|
190 //End Defect fix for INC036136 |
|
191 //================================================================================== |
|
192 |
|
193 EXPORT_C void CMDXMLParser::SetStoreInvalid(TBool aStoreInvalid) |
|
194 { |
|
195 iStoreInvalid = aStoreInvalid; |
|
196 } |
|
197 |
|
198 //================================================================================== |
|
199 // Defect fix for INC105134 - GmXML consumes whitespace characters |
|
200 //================================================================================== |
|
201 EXPORT_C void CMDXMLParser::SetWhiteSpaceHandlingMode(TBool aPreserve) |
|
202 { |
|
203 iPreserve = aPreserve; |
|
204 } |
|
205 |
|
206 // End Defect fix for INC105134 |
|
207 //================================================================================== |
|
208 EXPORT_C CMDXMLDocument* CMDXMLParser::DetachXMLDoc() |
|
209 // |
|
210 // @return CMDXMLDocument* to the created DOM, should be called after the |
|
211 // conclusion of the parser process. Note that internal variable pointing to |
|
212 // the document is set to NULL so this function can only be called once per file |
|
213 // parse. Client application must take ownership of document for cleanup purposes. |
|
214 // |
|
215 { |
|
216 CMDXMLDocument* returnDoc = iXMLDoc; |
|
217 iXMLDoc = NULL; |
|
218 return returnDoc; |
|
219 } |
|
220 |
|
221 //================================================================================== |
|
222 |
|
223 CMDXMLEntityConverter* CMDXMLParser::EntityConverter() |
|
224 // |
|
225 // @return the CMDXMLEntityConverter for use in converting built in entity |
|
226 // and character entities back to their original format |
|
227 // |
|
228 { |
|
229 return iEntityConverter; |
|
230 } |
|
231 |
|
232 //================================================================================== |
|
233 |
|
234 EXPORT_C TInt CMDXMLParser::ParseFile(RFs aRFs, const TDesC& aFileToParse) |
|
235 // |
|
236 // ParseFile opens a file ready for parsing |
|
237 // @param aRFs a resource file session used for file I/O |
|
238 // @param aFileToParse the file name to parse |
|
239 // @return KErrNone if all OK or file read error code |
|
240 // |
|
241 { |
|
242 //Find out whether the file exists. If not dont start the active object |
|
243 if(!BaflUtils::FileExists(aRFs,aFileToParse)) |
|
244 { |
|
245 return KErrNotFound; |
|
246 } |
|
247 else |
|
248 { |
|
249 //Check whether the file is locked by any other process |
|
250 RFile tempFile; |
|
251 TInt err=tempFile.Open(aRFs, aFileToParse, EFileRead | EFileShareReadersOnly); |
|
252 if(err!=KErrNone) |
|
253 { |
|
254 return err; |
|
255 } |
|
256 tempFile.Close(); |
|
257 } |
|
258 Cancel(); |
|
259 iSuspiciousCharacter = KErrNotFound; |
|
260 iError = KErrNone; |
|
261 iSeverity = EXMLNone; |
|
262 iDocTypeSet = EFalse; |
|
263 iVersionSet = EFalse; |
|
264 |
|
265 /* We need to open our file in a leave-safe place as it involves |
|
266 a heap alloc, and so we'll set up the AO to do that when it runs next. |
|
267 */ |
|
268 delete iFileToParse; |
|
269 iFileToParse = aFileToParse.Alloc(); |
|
270 if(iFileToParse == NULL) |
|
271 { |
|
272 return KErrNoMemory; |
|
273 } |
|
274 |
|
275 iRFs = aRFs; |
|
276 iState = KInitFromFile; |
|
277 |
|
278 SetActive(); |
|
279 TRequestStatus* s=&iStatus; |
|
280 User::RequestComplete(s, KErrNone); |
|
281 |
|
282 return KErrNone; |
|
283 } |
|
284 |
|
285 /** |
|
286 Parses a specified XML file into a DOM object tree. |
|
287 |
|
288 Parses a specified XML file into a DOM object tree using an open file handle. The |
|
289 parser takes ownership of the open file handle and will close handle when completed. |
|
290 |
|
291 @param aFileHandleToParse An open file handle for the file to parse. Ownership of the |
|
292 file handle is passed. |
|
293 @return KErrNone if successful. |
|
294 */ |
|
295 EXPORT_C TInt CMDXMLParser::ParseFile(RFile& aFileHandleToParse) |
|
296 { |
|
297 iFileHandleToParse = aFileHandleToParse; |
|
298 |
|
299 Cancel(); |
|
300 iSuspiciousCharacter = KErrNotFound; |
|
301 iError = KErrNone; |
|
302 iSeverity = EXMLNone; |
|
303 iDocTypeSet = EFalse; |
|
304 iVersionSet = EFalse; |
|
305 |
|
306 iState = KInitFromFile; |
|
307 |
|
308 iStatus = KRequestPending; |
|
309 SetActive(); |
|
310 TRequestStatus* s=&iStatus; |
|
311 User::RequestComplete(s, KErrNone); |
|
312 |
|
313 return KErrNone; |
|
314 } |
|
315 //================================================================================== |
|
316 |
|
317 EXPORT_C void CMDXMLParser::ParseSourceL(MMDXMLParserDataProvider *aSource) |
|
318 { |
|
319 iSuspiciousCharacter = KErrNotFound; |
|
320 iError = KErrNone; |
|
321 iSeverity = EXMLNone; |
|
322 iDocTypeSet = EFalse; |
|
323 iVersionSet = EFalse; |
|
324 |
|
325 iDataSource = aSource; |
|
326 PrepareForReuseL(); |
|
327 GetMoreData(); |
|
328 } |
|
329 |
|
330 //================================================================================== |
|
331 |
|
332 void CMDXMLParser::PrepareForReuseL() |
|
333 { |
|
334 Cancel(); |
|
335 |
|
336 iError = KErrNone; |
|
337 iSeverity = EXMLNone; |
|
338 iDocTypeSet = EFalse; |
|
339 iVersionSet = EFalse; |
|
340 iBytesPerChar = 0; |
|
341 iNextChar = 0; |
|
342 iInputBytesRemaining = 0; |
|
343 iOpened = EFalse; |
|
344 iClosed = EFalse; |
|
345 iNewElement = NULL; |
|
346 iParentElement = NULL; |
|
347 delete iUTF8EdgeBuffer; |
|
348 iUTF8EdgeBuffer = NULL; |
|
349 delete iBomBuffer; |
|
350 iBomBuffer = NULL; |
|
351 |
|
352 CreateDocumentL(); |
|
353 iState = KDetermineCharset; |
|
354 } |
|
355 |
|
356 //================================================================================== |
|
357 TBool CMDXMLParser::DetectFileType() |
|
358 // |
|
359 // Detects the type of a data source - can be Unicode, UTF-8 or ASCII (because ASCII is |
|
360 // a subset of UTF-8). |
|
361 // If the file is empty it is assumed to be Utf-8. |
|
362 { |
|
363 TBuf8<3> bom; |
|
364 |
|
365 // Read the first 3 bytes of the file. These contain any BOM present. If it turns out |
|
366 // there's not a bom we leave the pointer untouched so we can parse these bytes |
|
367 // as usual. |
|
368 if(iInputBufferPtr.Length() < 3) |
|
369 return EFalse; |
|
370 else |
|
371 bom.Copy(iInputBufferPtr.Left(3)); |
|
372 |
|
373 TInt hichar = (CEditableText::EByteOrderMark & 0xFF00)>>8; |
|
374 TInt lochar = CEditableText::EByteOrderMark & 0xFF; |
|
375 |
|
376 if((bom[0] == 0xEF) && (bom[1] == 0xBB) && (bom[2] == 0xBF)) |
|
377 { |
|
378 // Utf-8 with a bom. We don't want to parse the bom, so add 3 bytes offset to the read pos. |
|
379 iBytesPerChar = 1; |
|
380 iNextChar = 3; |
|
381 } |
|
382 else |
|
383 { |
|
384 if((bom[0] == lochar) && (bom[1] == hichar)) |
|
385 { |
|
386 // Little Endian Unicode. Move the read position on 2 bytes to ignore the bom. |
|
387 iBytesPerChar = 2; |
|
388 // would normally skip first 2 characters |
|
389 iNextChar = 2; |
|
390 } |
|
391 else if((bom[0] == hichar) && (bom[1] == lochar)) |
|
392 { |
|
393 // We have a bom, but it indicates endianess opposite to that of the platform. We |
|
394 // don't currently support this so set an error. |
|
395 SetError(KErrNotSupported, EXMLFatal); |
|
396 } |
|
397 else |
|
398 { |
|
399 // Default to Utf-8 |
|
400 iBytesPerChar = 1; |
|
401 } |
|
402 } |
|
403 return ETrue; |
|
404 } |
|
405 |
|
406 //================================================================================== |
|
407 |
|
408 EXPORT_C void CMDXMLParser::DoCancel() |
|
409 // |
|
410 // DoCancel function inherited from CActive base class |
|
411 // |
|
412 { |
|
413 if (iDataSource) |
|
414 { |
|
415 iDataSource->Disconnect(); |
|
416 iDataSource = NULL; |
|
417 } |
|
418 } |
|
419 |
|
420 //================================================================================== |
|
421 |
|
422 EXPORT_C TInt CMDXMLParser::RunError(TInt aError) |
|
423 // |
|
424 // RunError function inherited from CActive base class - intercepts any Leave from |
|
425 // the RunL() function, sets an appropriate errorcode and calls ParseFileCompleteL |
|
426 // |
|
427 { |
|
428 if(iDataSource) |
|
429 { |
|
430 iDataSource->Disconnect(); |
|
431 iDataSource = NULL; |
|
432 } |
|
433 |
|
434 iSeverity = EXMLFatal; |
|
435 iError = aError; |
|
436 if (iFileToParse) |
|
437 { |
|
438 delete iFileToParse; |
|
439 iFileToParse = NULL; |
|
440 } |
|
441 else if( iFileSource==NULL ) |
|
442 { |
|
443 // iFileSource is not set so the ownership of the file handle has not been passed |
|
444 iFileHandleToParse.Close(); |
|
445 } |
|
446 |
|
447 __ASSERT_DEBUG(iParserObserver != NULL, Panic(ENullMemVarParserObserver)); |
|
448 TRAPD(err, iParserObserver->ParseFileCompleteL()); |
|
449 return err; |
|
450 } |
|
451 |
|
452 //================================================================================== |
|
453 |
|
454 TBool CMDXMLParser::DoParseLoopL() |
|
455 // |
|
456 // RunL function inherited from CActive base class - carries out the actual parsing. |
|
457 // @leave can Leave due to OOM |
|
458 // |
|
459 { |
|
460 TBuf<1> singleChar; |
|
461 TGetCharReturn getCharReturn = KError; |
|
462 TInt error; |
|
463 |
|
464 while(iSeverity != EXMLFatal && (getCharReturn = GetChar(singleChar), getCharReturn == KCharReturned) ) |
|
465 { |
|
466 |
|
467 #if 0 // THIS NEEDS TO BE REMOVED - IT'S A VERY HANDY DEBUGGING TOOL WHEN YOU'RE MANUALLY DOING |
|
468 // BYTE ORDERING ON MISALIGNED STREAMS THOUGH... |
|
469 { |
|
470 #define DES_AS_8_BIT(str) (TPtrC8((TText8*)((str).Ptr()), (str).Size())) |
|
471 |
|
472 RFs aRFs; |
|
473 RFile aFile; |
|
474 _LIT(KFileName, "c:\\documents\\SMIL_Test_Files\\echoOutput.txt"); |
|
475 TPtrC file; |
|
476 TInt err; |
|
477 |
|
478 file.Set(KFileName); |
|
479 |
|
480 aRFs.Connect(); |
|
481 |
|
482 err = aFile.Open(aRFs, file, EFileWrite); |
|
483 if(err != KErrNone) |
|
484 err = aFile.Create(aRFs, file, EFileWrite); |
|
485 |
|
486 err = 0; |
|
487 aFile.Seek(ESeekEnd, err); |
|
488 aFile.Write(DES_AS_8_BIT(singleChar)); |
|
489 aFile.Close(); |
|
490 aRFs.Close(); |
|
491 } |
|
492 #endif |
|
493 if((!iOpened) && singleChar != KXMLStartTag) |
|
494 { |
|
495 HandleTextL(singleChar); |
|
496 } |
|
497 |
|
498 if((iOpened) || (singleChar == KXMLStartTag)) |
|
499 { |
|
500 if(singleChar == KXMLStartTag) |
|
501 { |
|
502 if(iOpened) |
|
503 { |
|
504 if((iSuspiciousCharacter == KErrNotFound)) |
|
505 { |
|
506 iSuspiciousCharacter = iElementTag->Length(); |
|
507 } |
|
508 } |
|
509 else |
|
510 { |
|
511 AddTextL(iParentElement); |
|
512 iOpened = ETrue; |
|
513 } |
|
514 } |
|
515 else if(singleChar == KXMLEndTag) |
|
516 { |
|
517 if(iSuspiciousCharacter != KErrNotFound) |
|
518 { |
|
519 TPtrC suspiciousSection = iElementTag->Mid(iSuspiciousCharacter); |
|
520 if( CheckForStartCData(suspiciousSection) == 0 ) |
|
521 { |
|
522 TInt endCDataLen = TPtrC(KXMLEndCDataSection).Length(); |
|
523 // The suspicious character begins a CDataSection. Check if |
|
524 // this End Tag is closing it. |
|
525 if( suspiciousSection.Right(endCDataLen - 1) |
|
526 == TPtrC(KXMLEndCDataSection).Left(endCDataLen - 1) ) |
|
527 { |
|
528 // Any dodgy characters began the CDataSection or were in it |
|
529 iSuspiciousCharacter = KErrNotFound; |
|
530 } |
|
531 } |
|
532 else if( suspiciousSection.Find(KXMLStartComment) == 0 ) |
|
533 { |
|
534 // The suspicious character begins a comment. Check if |
|
535 // this End Tag is closing it. |
|
536 TInt endCommentLen = TPtrC(KXMLEndComment).Length(); |
|
537 if( suspiciousSection.Right(endCommentLen - 1) |
|
538 == TPtrC(KXMLEndComment).Left(endCommentLen - 1) ) |
|
539 { |
|
540 // Any dodgy characters began the comment or were in it |
|
541 iSuspiciousCharacter = KErrNotFound; |
|
542 } |
|
543 } |
|
544 else if((CheckForStartCData(*iElementTag) == 0) || (iElementTag->Find(KXMLStartComment) == 0)) |
|
545 { |
|
546 // this tag is a CDataSection or comment, we're allowed < |
|
547 iSuspiciousCharacter = KErrNotFound; |
|
548 iClosed = ETrue; |
|
549 } |
|
550 else |
|
551 { |
|
552 // The < was spurious, set an error and close the tag as normal |
|
553 SetError(KErrXMLIllegalCharacter, EXMLWorkable); |
|
554 iClosed = ETrue; |
|
555 iSuspiciousCharacter = KErrNotFound; |
|
556 } |
|
557 } |
|
558 else |
|
559 { |
|
560 iClosed = ETrue; |
|
561 } |
|
562 } |
|
563 |
|
564 // ensure descriptor doesn't overflow end and panics |
|
565 if(iElementTag->Length() == iElementTag->Des().MaxLength()) |
|
566 { |
|
567 iElementTag = iElementTag->ReAllocL(iElementTag->Length() + KNominalTagLength); |
|
568 } |
|
569 |
|
570 iElementTag->Des().Append(singleChar); |
|
571 |
|
572 // if tag is complete and needs adding to the DOM? |
|
573 if(iClosed) |
|
574 { |
|
575 if( !CommentL(iParentElement) |
|
576 && !CDataSectionL(iParentElement) |
|
577 && !VersionIDL() |
|
578 && !DocTypeL() |
|
579 && !ProcessingInstructionL(iParentElement) ) |
|
580 { |
|
581 // is this a regular closing tag |
|
582 if |
|
583 (iElementTag->Left(2) == KXMLStartEndTag) |
|
584 { |
|
585 error = ParseElementEndTag(*iParentElement, iElementTag->Des()); |
|
586 if(error == KErrNone) |
|
587 { |
|
588 if(iParentElement->ParentNode() == NULL) |
|
589 { |
|
590 SetError(KErrXMLBadNesting, EXMLIndeterminate); |
|
591 } |
|
592 else |
|
593 { |
|
594 iParentElement = (CMDXMLElement*) iParentElement->ParentNode(); |
|
595 } |
|
596 } |
|
597 else if(error == KErrNotFound) |
|
598 { |
|
599 CMDXMLElement* tempElement = (CMDXMLElement*) iParentElement->ParentNode(); |
|
600 TInt searchResult = KErrNotFound; |
|
601 |
|
602 while(tempElement != NULL && |
|
603 searchResult == KErrNotFound && |
|
604 tempElement->NodeName() != KXMLDocumentElementNodeName) |
|
605 { |
|
606 searchResult = ParseElementEndTag(*tempElement,iElementTag->Des()); |
|
607 if(searchResult == KErrNone) |
|
608 { |
|
609 iParentElement = tempElement; |
|
610 SetError(KErrXMLBadNesting, EXMLIndeterminate); |
|
611 } |
|
612 else |
|
613 { |
|
614 tempElement = (CMDXMLElement*) tempElement->ParentNode(); |
|
615 } |
|
616 } |
|
617 if(searchResult != KErrNone) |
|
618 { |
|
619 SetError(KErrXMLBadNesting, EXMLIndeterminate); |
|
620 } |
|
621 } |
|
622 else |
|
623 { |
|
624 SetError(error, EXMLIndeterminate); |
|
625 } |
|
626 } |
|
627 // if a new element start tag or start/end tag |
|
628 else |
|
629 { |
|
630 // NOTE ParseStartTagL destroys iElementTag |
|
631 // so following check must be done first |
|
632 |
|
633 // if not single tag with close |
|
634 if(!(iElementTag->Right(2) == KXMLEndStartTag)) |
|
635 { |
|
636 iNewElement = ParseStartTagL(); |
|
637 CleanupStack::PushL(iNewElement); |
|
638 error = iParentElement->AppendChild(iNewElement); |
|
639 |
|
640 if(error == KErrNone) |
|
641 { |
|
642 CleanupStack::Pop(); // iNewElement |
|
643 iParentElement = (CMDXMLElement*) iParentElement->LastChild(); |
|
644 } |
|
645 else |
|
646 { |
|
647 SetError(error, EXMLWorkable); |
|
648 CleanupStack::PopAndDestroy(iNewElement); // iNewElement |
|
649 } |
|
650 } |
|
651 else |
|
652 { |
|
653 iNewElement = ParseStartTagL(); |
|
654 CleanupStack::PushL(iNewElement); |
|
655 error = iParentElement->AppendChild(iNewElement); |
|
656 |
|
657 if(error == KErrNone) |
|
658 { |
|
659 CleanupStack::Pop(iNewElement); // iNewElement |
|
660 } |
|
661 else |
|
662 { |
|
663 SetError(error, EXMLWorkable); |
|
664 CleanupStack::PopAndDestroy(iNewElement); // iNewElement |
|
665 } |
|
666 } |
|
667 } |
|
668 iEndOfTag = ETrue; |
|
669 } |
|
670 |
|
671 if(iEndOfTag) |
|
672 { |
|
673 iEndOfTag = EFalse; |
|
674 iOpened = iClosed = EFalse; |
|
675 iElementTag->Des().Zero(); |
|
676 |
|
677 // reduce size of ElementTag if increased beyond normal limits on last pass |
|
678 if(iElementTag->Des().MaxLength() > KNominalTagLength) |
|
679 { |
|
680 iElementTag = iElementTag->ReAllocL(KNominalTagLength); |
|
681 } |
|
682 } |
|
683 } |
|
684 } |
|
685 } // END OF WHILE LOOP |
|
686 |
|
687 if(getCharReturn == KError) |
|
688 { |
|
689 return EFalse; |
|
690 } |
|
691 else |
|
692 { |
|
693 // GetChar returned KWEaitForChar |
|
694 // GetChar handles pushing the state and requesting more data for us, so we just go active. |
|
695 return ETrue; |
|
696 } |
|
697 } |
|
698 |
|
699 |
|
700 |
|
701 |
|
702 //================================================================================== |
|
703 |
|
704 void CMDXMLParser::RunL() |
|
705 { |
|
706 TRequestStatus* s=&iStatus; |
|
707 TInt err = s->Int(); |
|
708 |
|
709 switch(iState) |
|
710 { |
|
711 case KInitFromFile: |
|
712 delete iFileSource; |
|
713 iFileSource = NULL; |
|
714 if( iFileToParse == NULL ) |
|
715 { |
|
716 // iFileToParse is not set, file was passed by open file handle. |
|
717 iFileSource = CMDXMLParserFileDataSource::NewL(iFileHandleToParse); |
|
718 } |
|
719 else |
|
720 { |
|
721 iFileSource = CMDXMLParserFileDataSource::NewL(iRFs, *iFileToParse); |
|
722 |
|
723 delete iFileToParse; |
|
724 iFileToParse = NULL; |
|
725 } |
|
726 |
|
727 ParseSource(iFileSource); // will go active itself |
|
728 break; |
|
729 |
|
730 case KDetermineCharset: |
|
731 if(!iBytesPerChar) |
|
732 { |
|
733 if(DetectFileType()) |
|
734 { |
|
735 iState = KParseData; |
|
736 SetActive(); |
|
737 User::RequestComplete(s, KErrNone); |
|
738 } |
|
739 else |
|
740 { |
|
741 if(!iBomBuffer) |
|
742 { |
|
743 iBomBuffer = HBufC8::NewL(KUTF8EdgeBufferLen); |
|
744 } |
|
745 |
|
746 TPtr8 bomDes(iBomBuffer->Des()); |
|
747 TInt newLength = bomDes.Length() + iCurrentInputBufferLen; |
|
748 iBomBuffer->ReAlloc(newLength); |
|
749 |
|
750 bomDes.Append(iInputBufferPtr); |
|
751 if(iBomBuffer->Length() >=3) |
|
752 { |
|
753 iInputBufferPtr.Set(bomDes); |
|
754 iCurrentInputBufferLen = newLength; |
|
755 |
|
756 SetActive(); |
|
757 User::RequestComplete(s, KErrNone); |
|
758 } |
|
759 else |
|
760 { |
|
761 GetMoreData(); |
|
762 } |
|
763 } |
|
764 } |
|
765 else |
|
766 { |
|
767 iState = KParseData; |
|
768 SetActive(); |
|
769 User::RequestComplete(s, KErrNone); |
|
770 } |
|
771 break; |
|
772 |
|
773 case KWaitingForData: |
|
774 switch(err) |
|
775 { |
|
776 case MMDXMLParserDataProvider::KMoreData: |
|
777 // We got more data this time, make sure all the parameters are correct. |
|
778 iCurrentInputBufferLen = iInputBufferPtr.Length(); |
|
779 |
|
780 iState = iPreviousState; |
|
781 SetActive(); |
|
782 User::RequestComplete(s, KErrNone); |
|
783 break; |
|
784 |
|
785 case MMDXMLParserDataProvider::KDataStreamEnd: |
|
786 iState = KFinished; |
|
787 SetActive(); |
|
788 User::RequestComplete(s, KErrNone); |
|
789 break; |
|
790 |
|
791 default: |
|
792 case MMDXMLParserDataProvider::KDataStreamError: |
|
793 User::Leave(KErrCorrupt); |
|
794 break; |
|
795 } |
|
796 |
|
797 break; |
|
798 |
|
799 case KParseData: |
|
800 if(!iParentElement) // initialise the parsing |
|
801 { |
|
802 iOpened = EFalse; |
|
803 iClosed = EFalse; |
|
804 // If we're going through the tag and find a < we don't know whether |
|
805 // it will be valid (eg it starts a CDataSection) or whether it's an |
|
806 // illegal character. Store its index so that when we're a bit further |
|
807 // along in the string we can check whether it was allowed and if not |
|
808 // set an Illegal Character error. |
|
809 delete iNewElement; |
|
810 iNewElement = NULL; |
|
811 |
|
812 __ASSERT_DEBUG(iXMLDoc != NULL, Panic(ENullMemVarXMLDoc)); |
|
813 iParentElement = iXMLDoc->DocumentElement(); |
|
814 |
|
815 delete iElementTag; |
|
816 iElementTag = NULL; |
|
817 iElementTag = HBufC::NewL(KNominalTagLength); |
|
818 } |
|
819 |
|
820 if(!iBytesPerChar) |
|
821 { |
|
822 // assume ascii/UTF8 |
|
823 iBytesPerChar = 1; |
|
824 } |
|
825 |
|
826 if(!DoParseLoopL()) |
|
827 iState = KFinished; |
|
828 |
|
829 break; |
|
830 |
|
831 case KFinished: |
|
832 // Check for any errors that we can pick up now, like missing doctype, or incomplete content |
|
833 CheckForErrors(); |
|
834 |
|
835 // we want to leave this instance in a safe state where it can be restarted again with a single call to ParseSource. |
|
836 // cannot delete this element, as it belongs to the document... |
|
837 iParentElement = NULL; |
|
838 iState = KDetermineCharset; |
|
839 |
|
840 iDataSource->Disconnect(); |
|
841 iDataSource=NULL; |
|
842 __ASSERT_DEBUG(iParserObserver != NULL, Panic(ENullMemVarParserObserver)); |
|
843 iParserObserver->ParseFileCompleteL(); |
|
844 break; |
|
845 |
|
846 default: |
|
847 User::Leave(KErrUnknown); |
|
848 break; |
|
849 } |
|
850 } |
|
851 |
|
852 void CMDXMLParser::CheckForErrors() |
|
853 { |
|
854 if(iError == KErrNone) |
|
855 { |
|
856 if(iParentElement) |
|
857 { |
|
858 // if iParentElement is not pointing to dummy root node, there has been a problem |
|
859 if( (iParentElement == NULL) || (iParentElement->NodeName() != KXMLDocumentElementNodeName) ) |
|
860 { |
|
861 SetError(KErrXMLIncomplete, EXMLWorkable); |
|
862 } |
|
863 else if(!iParentElement->CheckChildren()) |
|
864 { |
|
865 SetError(KErrXMLInvalidChild, EXMLWorkable); |
|
866 } |
|
867 else if(iParentElement->FirstChild() != NULL) |
|
868 { |
|
869 // multiple real (not dummy) root elements |
|
870 TInt count = 0; |
|
871 CMDXMLNode* iterator = iParentElement->FirstChild(); |
|
872 do |
|
873 { |
|
874 if(iterator->NodeType() == CMDXMLNode::EElementNode) |
|
875 { |
|
876 count++; |
|
877 } |
|
878 iterator = iterator->NextSibling(); |
|
879 } |
|
880 while(iterator != NULL); |
|
881 |
|
882 if(count != 1) |
|
883 { |
|
884 SetError(KErrXMLMultipleRootElements, EXMLWorkable); |
|
885 } |
|
886 } |
|
887 } |
|
888 } |
|
889 if(iError == KErrNone && !iDocTypeSet) |
|
890 { |
|
891 SetError(KErrXMLMissingDocTypeTag, EXMLWorkable); |
|
892 } |
|
893 |
|
894 if(iError == KErrNone && !iVersionSet) |
|
895 { |
|
896 SetError(KErrXMLMissingVersionTag, EXMLWorkable); |
|
897 } |
|
898 } |
|
899 |
|
900 |
|
901 void CMDXMLParser::CreateDocumentL() |
|
902 // |
|
903 // Creates a generic or DTD-specific document object |
|
904 // @leave can Leave due to OOM |
|
905 // |
|
906 { |
|
907 delete iXMLDoc; |
|
908 iXMLDoc = NULL; |
|
909 |
|
910 if (iDtdRepresentation != NULL) |
|
911 iXMLDoc = CMDXMLDocument::NewL(*iDtdRepresentation); |
|
912 else |
|
913 iXMLDoc = CMDXMLDocument::NewL(); |
|
914 } |
|
915 |
|
916 |
|
917 //================================================================================== |
|
918 |
|
919 TBool CMDXMLParser::DocTypeL() |
|
920 // |
|
921 // @return Returns true if the current tag is a doctype tag and sets the |
|
922 // Document DocType member accordingly on the first pass of this function. |
|
923 // |
|
924 { |
|
925 TBool returnValue = EFalse; |
|
926 TInt tagIdLen = TPtrC(KXMLDocumentTypes).Length(); |
|
927 |
|
928 __ASSERT_DEBUG(iElementTag != NULL, Panic(ENullMemVarElementTag)); |
|
929 |
|
930 if(iElementTag->Length() > tagIdLen |
|
931 && iElementTag->Left(tagIdLen) == KXMLDocumentTypes) |
|
932 { |
|
933 if(iDocTypeSet) |
|
934 { |
|
935 SetError(KErrXMLDuplicateDocTypeTags, EXMLWorkable); |
|
936 } |
|
937 else |
|
938 { |
|
939 iXMLDoc->SetDocTypeTagL(iElementTag->Des()); |
|
940 iDocTypeSet = ETrue; |
|
941 } |
|
942 returnValue = ETrue; |
|
943 iEndOfTag = ETrue; |
|
944 } |
|
945 |
|
946 return returnValue; |
|
947 } |
|
948 |
|
949 //================================================================================== |
|
950 |
|
951 TBool CMDXMLParser::ProcessingInstructionL(CMDXMLElement* aParentElement) |
|
952 // |
|
953 // creates a new processing instruction if necessary and adds to document |
|
954 // @return Returns true if the current tag is a processing instruction |
|
955 // |
|
956 { |
|
957 TBool returnValue = EFalse; |
|
958 TInt startPILen = TPtrC(KXMLStartProcessingInstruction).Length(); |
|
959 TInt endPILen = TPtrC(KXMLEndProcessingInstruction).Length(); |
|
960 |
|
961 __ASSERT_DEBUG(iElementTag != NULL, Panic(ENullMemVarElementTag)); |
|
962 |
|
963 if((iElementTag->Left(startPILen) == KXMLStartProcessingInstruction) && |
|
964 (iElementTag->Right(endPILen) == KXMLEndProcessingInstruction)) |
|
965 { |
|
966 if(aParentElement != NULL) |
|
967 { |
|
968 CMDXMLProcessingInstruction* inst = CMDXMLProcessingInstruction::NewLC(iXMLDoc); |
|
969 |
|
970 TPtrC instStr = iElementTag->Des().Mid(startPILen, |
|
971 iElementTag->Length() - (startPILen + endPILen)); |
|
972 inst->SetDataL(instStr); |
|
973 |
|
974 __ASSERT_DEBUG(aParentElement != NULL, Panic(ENullParameterParentElement)); |
|
975 TInt error = aParentElement->AppendChild(inst); |
|
976 CleanupStack::Pop(inst); |
|
977 |
|
978 if(error != KErrNone) |
|
979 { |
|
980 SetError(error, EXMLWorkable); |
|
981 } |
|
982 } |
|
983 |
|
984 returnValue = ETrue; |
|
985 iEndOfTag = ETrue; |
|
986 } |
|
987 |
|
988 return returnValue; |
|
989 } |
|
990 |
|
991 |
|
992 //================================================================================== |
|
993 |
|
994 TBool CMDXMLParser::CDataSectionL(CMDXMLElement* aParentElement) |
|
995 // |
|
996 // creates a new CDataSection if necessary and adds to document |
|
997 // @return Returns true if the current tag is a CDataSection |
|
998 // |
|
999 { |
|
1000 TBool returnValue = EFalse; |
|
1001 TInt instLen = TPtrC(KXMLStartCDataSection).Length(); |
|
1002 TInt endCDataLen = TPtrC(KXMLEndCDataSection).Length(); |
|
1003 |
|
1004 __ASSERT_DEBUG(iElementTag != NULL, Panic(ENullMemVarElementTag)); |
|
1005 |
|
1006 if (iElementTag->Left(instLen) == KXMLStartCDataSection) |
|
1007 { |
|
1008 returnValue = ETrue; |
|
1009 if ((iElementTag->Right(endCDataLen) == KXMLEndCDataSection) && (aParentElement != NULL)) |
|
1010 { |
|
1011 CMDXMLCDATASection* inst = CMDXMLCDATASection::NewLC(iXMLDoc); |
|
1012 |
|
1013 TPtrC instStr = iElementTag->Des().Mid(instLen, |
|
1014 iElementTag->Length() - (instLen + endCDataLen)); |
|
1015 inst->SetDataL(instStr); |
|
1016 |
|
1017 __ASSERT_DEBUG(aParentElement != NULL, Panic(ENullParameterParentElement)); |
|
1018 TInt error = aParentElement->AppendChild(inst); |
|
1019 CleanupStack::Pop(); // inst |
|
1020 |
|
1021 if(error != KErrNone) |
|
1022 { |
|
1023 SetError(error, EXMLWorkable); |
|
1024 } |
|
1025 } |
|
1026 iEndOfTag = ETrue; |
|
1027 } |
|
1028 |
|
1029 return returnValue; |
|
1030 } |
|
1031 |
|
1032 |
|
1033 //================================================================================== |
|
1034 |
|
1035 TBool CMDXMLParser::VersionIDL() |
|
1036 // |
|
1037 // @return returns true if the current tag is a version id tag and sets the |
|
1038 // Document Version member accordingly on the first pass of this function. |
|
1039 // |
|
1040 { |
|
1041 TBool returnValue = EFalse; |
|
1042 TInt tagIdLen = TPtrC(KXMLVersion).Length(); |
|
1043 |
|
1044 __ASSERT_DEBUG(iElementTag != NULL, Panic(ENullMemVarElementTag)); |
|
1045 |
|
1046 if(iElementTag->Length() > tagIdLen |
|
1047 && iElementTag->Left(tagIdLen) == KXMLVersion) |
|
1048 { |
|
1049 if(iVersionSet) |
|
1050 { |
|
1051 SetError(KErrXMLDuplicateVersionTags, EXMLWorkable); |
|
1052 } |
|
1053 else |
|
1054 { |
|
1055 iXMLDoc->SetVersionTagL(iElementTag->Des()); |
|
1056 iVersionSet = ETrue; |
|
1057 } |
|
1058 returnValue = ETrue; |
|
1059 iEndOfTag = ETrue; |
|
1060 } |
|
1061 |
|
1062 return returnValue; |
|
1063 } |
|
1064 |
|
1065 //================================================================================== |
|
1066 |
|
1067 TBool CMDXMLParser::CommentL(CMDXMLElement* aParentElement) |
|
1068 // |
|
1069 // creates a new comment if necessary and adds to document |
|
1070 // @return returns true if the current tag is a comment tag |
|
1071 // |
|
1072 { |
|
1073 TBool returnValue = EFalse; |
|
1074 TInt commentLen = TPtrC(KXMLStartComment).Length(); |
|
1075 TInt endCommentLen = TPtrC(KXMLEndComment).Length(); |
|
1076 |
|
1077 __ASSERT_DEBUG(iElementTag != NULL, Panic(ENullMemVarElementTag)); |
|
1078 |
|
1079 if (iElementTag->Left(commentLen) == KXMLStartComment) |
|
1080 { |
|
1081 returnValue = ETrue; |
|
1082 if ((aParentElement != NULL) && (iElementTag->Right(endCommentLen) == KXMLEndComment)) |
|
1083 { |
|
1084 CMDXMLComment* comment = CMDXMLComment::NewLC(iXMLDoc); |
|
1085 |
|
1086 TPtrC commentStr = iElementTag->Des().Mid(commentLen, |
|
1087 iElementTag->Length() - (commentLen + endCommentLen)); |
|
1088 comment->SetDataL(commentStr); |
|
1089 |
|
1090 __ASSERT_DEBUG(aParentElement != NULL, Panic(ENullParameterParentElement)); |
|
1091 TInt error = aParentElement->AppendChild(comment); |
|
1092 CleanupStack::Pop(); // comment |
|
1093 |
|
1094 if(error != KErrNone) |
|
1095 { |
|
1096 SetError(error, EXMLWorkable); |
|
1097 } |
|
1098 iEndOfTag = ETrue; |
|
1099 } |
|
1100 } |
|
1101 |
|
1102 return returnValue; |
|
1103 } |
|
1104 |
|
1105 //================================================================================== |
|
1106 |
|
1107 EXPORT_C void CMDXMLParser::SetSourceCharacterWidth(TMDXMLParserInputCharWidth aWidth) |
|
1108 { |
|
1109 iBytesPerChar = aWidth; |
|
1110 } |
|
1111 |
|
1112 //================================================================================== |
|
1113 |
|
1114 void CMDXMLParser::GetMoreData() |
|
1115 { |
|
1116 |
|
1117 // Prepare these variables. |
|
1118 iNextChar = 0; // reading from the start of the buffer |
|
1119 iCurrentInputBufferLen = 0; // we have no characters in the buffer |
|
1120 iUnicodeConversion.Zero(); // we have no characters in our UTF8->Unicode Conversion buffer. |
|
1121 iUnicodeConversionLen = 0; |
|
1122 iUnicodeReadPos = 0; |
|
1123 |
|
1124 // Request more data from the data provider |
|
1125 __ASSERT_DEBUG(iDataSource != NULL, Panic(ENullMemVarDataSource)); |
|
1126 iDataSource->GetData(iInputBufferPtr, iStatus); |
|
1127 SetActive(); |
|
1128 iPreviousState = iState; |
|
1129 iState = KWaitingForData; |
|
1130 } |
|
1131 |
|
1132 //================================================================================== |
|
1133 |
|
1134 CMDXMLParser::TGetCharReturn CMDXMLParser::GetDoubleByteChar(TDes& aChar) |
|
1135 // when inputing we have a pointer to an 8 bit buffer (iInputBufferPtr), for unicode |
|
1136 // input we point a 16 bit descriptor (tempUnicodeInput) at the 8 bit buffer to |
|
1137 // enable us to read the 2 x 8 bit chars as a single 16 bit char. |
|
1138 // However it isn't always this simple as the data provider interface makes no guarantees |
|
1139 // on the alignment of this data. It's perfectly possible for it to end up with a unicode |
|
1140 // character where the high byte comes from the previous buffer and the low byte comes from |
|
1141 // the current one. This will put the rest of the current buffer out of line, and also all |
|
1142 // subsequent buffers unless an odd length buffer is provided. Hopfully this won't happen often. |
|
1143 { |
|
1144 aChar.Zero(); |
|
1145 |
|
1146 if(iUnicodeInputMisaligned) |
|
1147 { |
|
1148 TUint16 tempOut; |
|
1149 TUint8* tempRead; |
|
1150 tempRead = (TUint8*)(iInputBufferPtr.Ptr()) + iNextChar; |
|
1151 |
|
1152 if(iCurrentInputBufferLen - iNextChar >=1) |
|
1153 { |
|
1154 // if we saved a byte last time, lets use that first |
|
1155 if(iSpareChar.Length()) |
|
1156 { |
|
1157 tempOut = iSpareChar[0]; |
|
1158 iSpareChar.Zero(); |
|
1159 } |
|
1160 else if(iCurrentInputBufferLen - iNextChar >=2) |
|
1161 { |
|
1162 // we didn't save a byte, so we read from the stream. |
|
1163 tempOut = *tempRead; |
|
1164 tempRead++; |
|
1165 iNextChar++; |
|
1166 } |
|
1167 else |
|
1168 { |
|
1169 // our input stream must have been an odd length - this might cause alignment problems, |
|
1170 // so we need to start reading bytewise for a while |
|
1171 iUnicodeInputMisaligned = ETrue; |
|
1172 |
|
1173 TUint8* tempRead = (TUint8*)(iInputBufferPtr.Ptr()) + iNextChar; |
|
1174 TUint16 tempVal = (TUint16)(*tempRead << 8); |
|
1175 iSpareChar.Copy(&tempVal,1); |
|
1176 |
|
1177 GetMoreData(); |
|
1178 return KWaitForChar; |
|
1179 |
|
1180 } |
|
1181 |
|
1182 // second byte (high byte) of our output comes from the input stream in all cases. |
|
1183 tempOut |= ((*tempRead & 0xFF) << 8); |
|
1184 iNextChar++; |
|
1185 |
|
1186 TPtrC16 readDes(&tempOut, 1); |
|
1187 aChar = readDes.Left(1); |
|
1188 } |
|
1189 } |
|
1190 else if(iCurrentInputBufferLen - iNextChar >= 2) |
|
1191 { |
|
1192 // we may be in a position where we don't know we're going to lose a byte |
|
1193 // so we'll test for that ahead of time, and then handle that in the normal way |
|
1194 |
|
1195 // if we execute this, it means that we have two bytes available to read. |
|
1196 const TUint16* word = reinterpret_cast<const TUint16*>((iInputBufferPtr.Ptr() + iNextChar)); |
|
1197 TPtrC16 tempUnicodeInput(word, 2); |
|
1198 aChar = tempUnicodeInput.Left(1); |
|
1199 iNextChar+=2; |
|
1200 return KCharReturned; |
|
1201 } |
|
1202 |
|
1203 TInt bytesRemaining = iCurrentInputBufferLen - iNextChar; |
|
1204 |
|
1205 switch(bytesRemaining) |
|
1206 { |
|
1207 case 1: |
|
1208 { |
|
1209 // our input stream must have been an odd length - this might cause alignment problems, |
|
1210 // so we need to start reading bytewise for a while |
|
1211 |
|
1212 TUint8* tempRead = (TUint8*)(iInputBufferPtr.Ptr()) + iNextChar; |
|
1213 TUint16 tempVal = *tempRead; |
|
1214 iSpareChar.Copy(&tempVal,1); |
|
1215 iUnicodeInputMisaligned = ETrue; |
|
1216 iNextChar++; |
|
1217 if(!aChar.Length()) |
|
1218 { |
|
1219 GetMoreData(); |
|
1220 return KWaitForChar; |
|
1221 } |
|
1222 } |
|
1223 break; |
|
1224 case 0: |
|
1225 { |
|
1226 // we're at the end of this block, and it's turned out to be re-aligned. |
|
1227 // we can read subsequent blocks in 16-bit chunks. |
|
1228 iUnicodeInputMisaligned = EFalse; |
|
1229 } |
|
1230 break; |
|
1231 } |
|
1232 |
|
1233 return KCharReturned; |
|
1234 } |
|
1235 |
|
1236 //================================================================================== |
|
1237 |
|
1238 CMDXMLParser::TGetCharReturn CMDXMLParser::GetSingleByteChar(TDes& aChar) |
|
1239 { |
|
1240 // We have UTF8/ASCII Source, and we must need to convert some more if we got here. |
|
1241 iUnicodeConversion.Zero(); |
|
1242 iUnicodeConversionLen = 0; |
|
1243 iUnicodeReadPos = 0; |
|
1244 |
|
1245 // if we are not operating out of the edge buffer yet, work on the real one |
|
1246 if(!iUTF8EdgeBuffer) |
|
1247 { |
|
1248 // This is an 8 bit encoding, probably UTF-8, but could be ASCII. Because |
|
1249 // ASCII is valid UTF-8 we can just convert what we have to Unicode. |
|
1250 // We're going to convert a number of characters at a time here. |
|
1251 TInt inputBytesRemaining = iCurrentInputBufferLen - iNextChar; |
|
1252 TInt convResult; |
|
1253 TPtrC8 tempPtr( (TUint8*)(iInputBufferPtr.Ptr()) + iNextChar, inputBytesRemaining ); |
|
1254 |
|
1255 convResult = CnvUtfConverter::ConvertToUnicodeFromUtf8(iUnicodeConversion, tempPtr); |
|
1256 if((convResult >= 0) || (convResult == KErrCorrupt)) |
|
1257 { |
|
1258 // Sometimes the UTF8 decoder might return corrupt if it only gets a single character |
|
1259 // in this case we ignore the error and report that we have converted 0 characters |
|
1260 if (convResult == KErrCorrupt) |
|
1261 convResult = tempPtr.Length(); |
|
1262 |
|
1263 // This is the number of bytes converted. |
|
1264 // Keep an eye out in case there is no change in the character consumed count. |
|
1265 TInt bytesConverted = inputBytesRemaining - convResult; |
|
1266 |
|
1267 // We consumed characters. Make our input buffer read position correct. |
|
1268 iNextChar += bytesConverted; |
|
1269 |
|
1270 // Make our intermediate buffering correct and return the first character out of our buffer |
|
1271 // subsequent calls will just return characters from this buffer. |
|
1272 iUnicodeConversionLen = iUnicodeConversion.Length(); |
|
1273 aChar = iUnicodeConversion.Left(1); |
|
1274 iUnicodeReadPos = 1; |
|
1275 |
|
1276 if(convResult && convResult < KUTF8EdgeBufferLen) |
|
1277 { |
|
1278 TUint8* multiByteCheck = (TUint8*)(iInputBufferPtr.Ptr()) + iNextChar; |
|
1279 |
|
1280 // There is a possibility that we've got an edge case here |
|
1281 //check if our left over character is in fact UTF8. |
|
1282 if((0x80 & *multiByteCheck) != 0) |
|
1283 { |
|
1284 // Shift 'convResult' characters off into the edge buffer. |
|
1285 delete iUTF8EdgeBuffer; |
|
1286 iUTF8EdgeBuffer = HBufC8::New(KUTF8EdgeBufferLen); |
|
1287 *iUTF8EdgeBuffer = iInputBufferPtr.Right(convResult); |
|
1288 |
|
1289 TUint8 bitMask = 0x80; |
|
1290 TInt byteCount = 0; |
|
1291 while(bitMask && (bitMask & (iUTF8EdgeBuffer->Des()[0])) != 0) |
|
1292 { |
|
1293 bitMask >>= 1; |
|
1294 byteCount++; |
|
1295 }; |
|
1296 |
|
1297 if(!bitMask) |
|
1298 { |
|
1299 // the utf8 stream appears to be corrupt. |
|
1300 SetError(KError, EXMLFatal); |
|
1301 return KError; |
|
1302 } |
|
1303 // we need to find byteCount characters to make up the character currently stored in the edge |
|
1304 // buffer. |
|
1305 iRequiredUTF8Bytes = byteCount - iUTF8EdgeBuffer->Length(); |
|
1306 |
|
1307 // set the variables up so that we return any converted characters, and then begin work |
|
1308 // on the edge buffer (where we've already cached the remaining bytes if any) |
|
1309 // NOTE: We will return all the characters which we preconverted into iUnicodeConversion *before* |
|
1310 // we begin dealing with the edge buffer, because of the structure of this function. |
|
1311 iNextChar = iCurrentInputBufferLen; |
|
1312 |
|
1313 if (bytesConverted == 0) |
|
1314 // If no bytes were converted then there is nothing to return, |
|
1315 // we need to wait for more data and then conbine it with what we have |
|
1316 // just put on the edge buffer. |
|
1317 { |
|
1318 // need more bytes to finish this character. |
|
1319 GetMoreData(); |
|
1320 return KWaitForChar; |
|
1321 } |
|
1322 } |
|
1323 } |
|
1324 } |
|
1325 |
|
1326 else |
|
1327 { |
|
1328 return KError; // something failed in the UTF8 Converter. |
|
1329 } |
|
1330 } |
|
1331 else |
|
1332 { |
|
1333 // We are converting the UTF8 Edge Buffer. We know that in it's current state, it |
|
1334 // can't be converted to Unicode. |
|
1335 // Decide if we have enough characters in our current input stream to convert to an |
|
1336 // output character (or two) yet. |
|
1337 |
|
1338 if(iUTF8EdgeBuffer->Length() >= KUTF8EdgeBufferLen) |
|
1339 { |
|
1340 // our edge buffer reached the maximum length for a UTF8 character |
|
1341 // and we haven't managed to convert a unicode output. |
|
1342 // this means that the input stream is corrupt. |
|
1343 delete iUTF8EdgeBuffer; |
|
1344 iUTF8EdgeBuffer = NULL; |
|
1345 |
|
1346 // Report a fatal error. |
|
1347 SetError(KError, EXMLFatal); |
|
1348 return KError; |
|
1349 } |
|
1350 else |
|
1351 { |
|
1352 TInt convResult; |
|
1353 |
|
1354 // we know how many bytes are required in order to complete the utf8 buffer |
|
1355 TInt bytesAvailable = iCurrentInputBufferLen - iNextChar; |
|
1356 |
|
1357 if(bytesAvailable >= iRequiredUTF8Bytes) |
|
1358 { |
|
1359 // we have enough bytes to complete this character. |
|
1360 // Go ahead and convert then return it. |
|
1361 iUTF8EdgeBuffer->Des().Append(iInputBufferPtr.Mid(iNextChar, iRequiredUTF8Bytes)); |
|
1362 iUnicodeConversion.Zero(); |
|
1363 convResult = CnvUtfConverter::ConvertToUnicodeFromUtf8(iUnicodeConversion, iUTF8EdgeBuffer->Des()); |
|
1364 |
|
1365 // regardless if we managed to convert this buffer or not, we don't need it any more. |
|
1366 delete iUTF8EdgeBuffer; |
|
1367 iUTF8EdgeBuffer = NULL; |
|
1368 |
|
1369 // make sure we report any error in the conversion |
|
1370 if(convResult != 0) |
|
1371 { |
|
1372 // we either incorrectly calculated the required number of bytes or the |
|
1373 // stream is corrupt. Either way we have a fatal error. |
|
1374 SetError(KError, EXMLFatal); |
|
1375 return KError; |
|
1376 } |
|
1377 |
|
1378 // Make our intermediate buffering correct and return the first character out of our buffer |
|
1379 // subsequent calls will just return characters from this buffer. |
|
1380 iUnicodeConversionLen = iUnicodeConversion.Length(); |
|
1381 aChar = iUnicodeConversion.Left(1); |
|
1382 iUnicodeReadPos = 1; |
|
1383 |
|
1384 // set up the main input buffers so that the next char comes from the input stream. |
|
1385 iNextChar += iRequiredUTF8Bytes; |
|
1386 iRequiredUTF8Bytes = 0; |
|
1387 } |
|
1388 else |
|
1389 { |
|
1390 // we haven't got enough bytes to complete this character, store the |
|
1391 // available byte(s) and request more data. |
|
1392 iUTF8EdgeBuffer->Des().Append(iInputBufferPtr.Mid(iNextChar, bytesAvailable)); |
|
1393 |
|
1394 // Move the next character index on for as many bytes as we have just added to the edge buffer |
|
1395 iNextChar += bytesAvailable; |
|
1396 // We can reduce the number of bytes require by the number of bytes added to the edge buffer |
|
1397 iRequiredUTF8Bytes -= bytesAvailable; |
|
1398 |
|
1399 // need more bytes to finish this character. |
|
1400 GetMoreData(); |
|
1401 return KWaitForChar; |
|
1402 } |
|
1403 } |
|
1404 } |
|
1405 |
|
1406 return KCharReturned; |
|
1407 } |
|
1408 |
|
1409 //================================================================================== |
|
1410 |
|
1411 CMDXMLParser::TGetCharReturn CMDXMLParser::GetChar(TDes& aChar) |
|
1412 // |
|
1413 // Fetch one character from the input file |
|
1414 // @param aChar the returned character. |
|
1415 // @return returns true if a character returned or false if the file is finished |
|
1416 // |
|
1417 { |
|
1418 // first test - see if we're providing preconverted characters. |
|
1419 if(iUnicodeConversionLen && iUnicodeReadPos < iUnicodeConversionLen) |
|
1420 { |
|
1421 // return one of the preconverted chars. |
|
1422 aChar = iUnicodeConversion.Mid(iUnicodeReadPos, 1); |
|
1423 iUnicodeReadPos++; |
|
1424 |
|
1425 return KCharReturned; |
|
1426 } |
|
1427 |
|
1428 // Second test - see if we require more data. If we have converted data and we require |
|
1429 // more data, this code is not intelligent enough to request data from the provider |
|
1430 // early, but that's ok. |
|
1431 |
|
1432 // Buffer length held as a member variable for performance reasons |
|
1433 // this function will be accessed thousands of times |
|
1434 if(iCurrentInputBufferLen <= iNextChar) |
|
1435 { |
|
1436 GetMoreData(); |
|
1437 return KWaitForChar; |
|
1438 } |
|
1439 |
|
1440 // return the character, handling any of the buffer shuffling we need to do. |
|
1441 if(iBytesPerChar == 2) |
|
1442 return GetDoubleByteChar(aChar); |
|
1443 else |
|
1444 return GetSingleByteChar(aChar); |
|
1445 } |
|
1446 |
|
1447 //================================================================================== |
|
1448 |
|
1449 CMDXMLElement* CMDXMLParser::ParseStartTagL() |
|
1450 // |
|
1451 // Parse a start of element tag and create an element with attributes set. |
|
1452 // @return Returns a pointer to the created element |
|
1453 // @leave can Leave due to OOM |
|
1454 // |
|
1455 { |
|
1456 __ASSERT_DEBUG(iElementTag != NULL, Panic(ENullMemVarElementTag)); |
|
1457 |
|
1458 // there must be at least two angle brackets and a single character to be meaningful |
|
1459 if(iElementTag->Length() < 3) |
|
1460 return NULL; |
|
1461 |
|
1462 CMDXMLElement* newElement = NULL; |
|
1463 TPtr elementTagPtr = iElementTag->Des(); |
|
1464 |
|
1465 // remove the angle brackets and trim white space |
|
1466 if(iElementTag->Right(TPtrC(KXMLEndStartTag).Length()) == KXMLEndStartTag) |
|
1467 elementTagPtr = iElementTag->Left(iElementTag->Length() - TPtrC(KXMLEndStartTag).Length()); |
|
1468 else |
|
1469 elementTagPtr = iElementTag->Left(iElementTag->Length() - TPtrC(KXMLEndTag).Length()); |
|
1470 |
|
1471 elementTagPtr = iElementTag->Right(iElementTag->Length() - TPtrC(KXMLStartTag).Length()); |
|
1472 elementTagPtr.Trim(); |
|
1473 |
|
1474 // find out where the name ends and the attributes begin |
|
1475 TLex16 element(elementTagPtr); |
|
1476 element.SkipCharacters(); |
|
1477 TInt endOfName = element.Offset(); |
|
1478 |
|
1479 // separate out the name from the attributes |
|
1480 HBufC* elementName = (iElementTag->Left(endOfName)).AllocLC(); |
|
1481 TPtr elementNamePtr = elementName->Des(); |
|
1482 elementNamePtr.TrimRight(); |
|
1483 |
|
1484 TInt error = KErrNone; |
|
1485 |
|
1486 __ASSERT_DEBUG(iXMLDoc != NULL, Panic(ENullMemVarXMLDoc)); |
|
1487 |
|
1488 TBool validElement = iXMLDoc->ValidElementNameL(elementNamePtr); |
|
1489 if(validElement || iStoreInvalid) |
|
1490 { |
|
1491 // remove the actual name from the tag so we only pass on the attributes |
|
1492 elementTagPtr = iElementTag->Right(iElementTag->Length() - endOfName); |
|
1493 elementTagPtr.TrimLeft(); |
|
1494 newElement = CMDXMLElement::NewLC(iXMLDoc->CanElementHaveChildren(elementNamePtr), iXMLDoc, elementNamePtr); |
|
1495 |
|
1496 error = ParseElementAttributesL(*newElement, elementTagPtr); |
|
1497 CleanupStack::Pop(); |
|
1498 } |
|
1499 if(!validElement) |
|
1500 { |
|
1501 error = KErrXMLInvalidElement; |
|
1502 } |
|
1503 |
|
1504 CleanupStack::PopAndDestroy(elementName); // elementName |
|
1505 |
|
1506 if(error != KErrNone) |
|
1507 { |
|
1508 SetError(error, EXMLWorkable); |
|
1509 } |
|
1510 |
|
1511 return newElement; |
|
1512 |
|
1513 } |
|
1514 |
|
1515 |
|
1516 TInt CMDXMLParser::ParseElementAttributesL(CMDXMLElement& aElement, TDes& aTagToParse) |
|
1517 // |
|
1518 // This function is used to parse the attributes. |
|
1519 // @param aElement The element to which the attributes belong |
|
1520 // @param aTagToParse The tag to be parsed |
|
1521 // @return Returns KErrNone if both attribute name & value are valid |
|
1522 // KErrXMLBadAttributeName if attribute name is invalid or KErrXMLBadAttributeValue is invalid |
|
1523 // @leave can Leave due to OOM |
|
1524 // |
|
1525 { |
|
1526 TInt error = KErrNone; |
|
1527 TInt attributeError = KErrNone; |
|
1528 HBufC* attributeName = NULL; |
|
1529 HBufC* attributeValue = NULL; |
|
1530 TBuf<1> attributeDelimiter; // may be " or ' |
|
1531 TInt offset = KErrNone; |
|
1532 |
|
1533 offset = aTagToParse.Find(KEqualSign); |
|
1534 while(offset != KErrNotFound) |
|
1535 { |
|
1536 attributeName = TPtrC(aTagToParse.Left(offset)).AllocLC(); |
|
1537 TPtr attributeNamePtr = attributeName->Des(); |
|
1538 attributeNamePtr.TrimRight(); // remove white space that existed between name and equal sign |
|
1539 |
|
1540 // remove current attribute name and equal sign from string |
|
1541 aTagToParse = aTagToParse.Right(aTagToParse.Length() - (offset + 1)); |
|
1542 aTagToParse.TrimLeft(); // remove white space that existed between equal sign and delimiter |
|
1543 |
|
1544 if(error == KErrNone && aTagToParse.Length() < 2) // a name must be at followed by at least 2 delimiters |
|
1545 { |
|
1546 error = KErrXMLBadAttributeName; |
|
1547 // In this case, there is insufficient tag left to contain any more attributes |
|
1548 } |
|
1549 else if(error == KErrNone && aElement.IsAttributeSpecified(*attributeName)) |
|
1550 { |
|
1551 error = KErrXMLDuplicateAttributeName; |
|
1552 // We need to remove the attribute value from the tag string or it will be |
|
1553 // picked up as part of the next attribute name. |
|
1554 attributeDelimiter = aTagToParse.Left(1); |
|
1555 |
|
1556 // Check in case we've got a missing " at the beginning of the attribute. If |
|
1557 // we do, we need to try a different strategy to find the end of this attribute |
|
1558 if (attributeDelimiter != KQuotation && attributeDelimiter != KApostrophe) |
|
1559 { |
|
1560 offset = LocateNextAttribute(aTagToParse); |
|
1561 } |
|
1562 else |
|
1563 { |
|
1564 // remove start delimiter then search for next one (end delimiter) |
|
1565 aTagToParse = aTagToParse.Right(aTagToParse.Length() - 1); |
|
1566 offset = FindDelimiter(aTagToParse, attributeDelimiter); |
|
1567 } |
|
1568 |
|
1569 if(offset != KErrNotFound) |
|
1570 { |
|
1571 // remove current attribute value and delimiter |
|
1572 aTagToParse = aTagToParse.Right(aTagToParse.Length() - (offset + 1)); |
|
1573 aTagToParse.TrimLeft(); // remove white space that existed between delimiter and next name |
|
1574 } |
|
1575 } |
|
1576 else |
|
1577 { |
|
1578 attributeDelimiter = aTagToParse.Left(1); |
|
1579 |
|
1580 if (attributeDelimiter != KQuotation && attributeDelimiter != KApostrophe) |
|
1581 { |
|
1582 // This attribute doesn't have a valid delimiter. Try and find the beginning of the next |
|
1583 // attribute and just cut this one. |
|
1584 TInt nextAttribute = LocateNextAttribute(aTagToParse); |
|
1585 if(nextAttribute > 0) |
|
1586 { |
|
1587 // Add one to next attribute because the offset includes the whitespace before it |
|
1588 aTagToParse = aTagToParse.Right(aTagToParse.Length() - (nextAttribute + 1)); |
|
1589 } |
|
1590 |
|
1591 if (error == KErrNone) |
|
1592 { |
|
1593 error = KErrXMLBadAttributeValue; |
|
1594 } |
|
1595 } |
|
1596 else |
|
1597 { |
|
1598 // remove start delimiter then search for next one (end delimiter) |
|
1599 aTagToParse = aTagToParse.Right(aTagToParse.Length() - 1); |
|
1600 |
|
1601 offset = FindDelimiter(aTagToParse, attributeDelimiter); |
|
1602 if(offset != KErrNotFound) |
|
1603 { |
|
1604 attributeValue = TPtrC(aTagToParse.Left(offset)).AllocLC(); |
|
1605 TPtr attributeValuePtr = attributeValue->Des(); |
|
1606 attributeValuePtr.TrimRight(); // remove white space that existed between value and delimiter |
|
1607 |
|
1608 // remove current attribute value and delimiter |
|
1609 aTagToParse = aTagToParse.Right(aTagToParse.Length() - (offset + 1)); |
|
1610 aTagToParse.TrimLeft(); // remove white space that existed between delimiter and next name |
|
1611 |
|
1612 // Entity convert this attribute |
|
1613 attributeError = ParseSingleAttributeL(attributeValuePtr); |
|
1614 if( attributeError != KErrNone && error == KErrNone) |
|
1615 { |
|
1616 error = attributeError; |
|
1617 } |
|
1618 |
|
1619 attributeError = aElement.SetAttributeL(*attributeName, *attributeValue, iStoreInvalid); |
|
1620 if( attributeError != KErrNone && error == KErrNone) |
|
1621 { |
|
1622 error = KErrXMLInvalidAttribute; |
|
1623 } |
|
1624 CleanupStack::PopAndDestroy(attributeValue); //attributeValue |
|
1625 } |
|
1626 else if(error == KErrNone) |
|
1627 { |
|
1628 error = KErrXMLBadAttributeValue; |
|
1629 } |
|
1630 } |
|
1631 } |
|
1632 |
|
1633 // next attribute pair |
|
1634 offset = aTagToParse.Find(KEqualSign); |
|
1635 CleanupStack::PopAndDestroy(attributeName); //attributeName |
|
1636 } |
|
1637 |
|
1638 if(error == KErrNone && aTagToParse.Length() != 0) |
|
1639 { |
|
1640 error = KErrXMLBadAttributeValue; |
|
1641 } |
|
1642 |
|
1643 return error; |
|
1644 } |
|
1645 |
|
1646 TInt CMDXMLParser::LocateNextAttribute(const TDesC& aTagToParse) |
|
1647 { |
|
1648 // Find the next attribute by looking for an = then search back for a ' '. |
|
1649 // This is useful when you've hit rubbish parsing the content of a start tag |
|
1650 // and are looking for somewhere sensible to start. |
|
1651 TInt nextAttribute = KErrNotFound; |
|
1652 TInt offset = aTagToParse.Find(KEqualSign); |
|
1653 |
|
1654 // If the = is the first character then there isn't space for a ' ' so |
|
1655 // don't bother looking |
|
1656 if(offset > 0) |
|
1657 { |
|
1658 TPtrC invalidText = aTagToParse.Left(offset); |
|
1659 nextAttribute = invalidText.LocateReverse(' '); |
|
1660 } |
|
1661 |
|
1662 return nextAttribute; |
|
1663 } |
|
1664 |
|
1665 TInt CMDXMLParser::ParseElementEndTag(CMDXMLElement& aElement, const TDesC& aTagToParse) |
|
1666 // |
|
1667 // Parses an end tag. In fact, at this point the end tag must match |
|
1668 // the tag name of the start tag. |
|
1669 // @param aTagToParse Text of the end tag. |
|
1670 // @return Returns KErrNone if the end tag matches the start tag or KErrNotFound if there is a mismatch. |
|
1671 // |
|
1672 { |
|
1673 // The tag should be of the form '</tag>' where tag is the name of this element so we will |
|
1674 // check and strip off the surrounding </ > and then compare the remains with this # |
|
1675 // node name. |
|
1676 TInt retVal = KErrNone; |
|
1677 if( aTagToParse.Length() != (aElement.NodeName().Length()+3)) |
|
1678 { |
|
1679 retVal = KErrNotFound; |
|
1680 } |
|
1681 else |
|
1682 { |
|
1683 TInt startEndTagLen = TPtrC(KXMLStartEndTag).Length(); |
|
1684 TInt endTagLen = TPtrC(KXMLEndTag).Length(); |
|
1685 |
|
1686 if((aTagToParse.Left(startEndTagLen).Compare(KXMLStartEndTag) == 0) && |
|
1687 (aTagToParse.Right(endTagLen).Compare(KXMLEndTag) == 0)) |
|
1688 { |
|
1689 if(aElement.NodeName().Compare(aTagToParse.Mid(2, |
|
1690 aTagToParse.Length() - (startEndTagLen + endTagLen))) != 0) |
|
1691 { |
|
1692 retVal = KErrNotFound; |
|
1693 } |
|
1694 } |
|
1695 } |
|
1696 return retVal; |
|
1697 } |
|
1698 |
|
1699 |
|
1700 //================================================================================== |
|
1701 |
|
1702 EXPORT_C void CMDXMLParser::SetError(const TInt aErrorCode, const TXMLErrorCodeSeverity aSeverity) |
|
1703 // |
|
1704 // Sets iError to new errorcode if more serious than any error so far encountered |
|
1705 // |
|
1706 { |
|
1707 if(iSeverity > aSeverity) |
|
1708 { |
|
1709 iSeverity = aSeverity; |
|
1710 iError = aErrorCode; |
|
1711 } |
|
1712 } |
|
1713 |
|
1714 //================================================================================== |
|
1715 |
|
1716 EXPORT_C TInt CMDXMLParser::Error() const |
|
1717 { |
|
1718 return iError; |
|
1719 } |
|
1720 |
|
1721 //================================================================================== |
|
1722 |
|
1723 EXPORT_C TXMLErrorCodeSeverity CMDXMLParser::ErrorSeverity() const |
|
1724 { |
|
1725 return iSeverity; |
|
1726 } |
|
1727 |
|
1728 //================================================================================== |
|
1729 |
|
1730 void CMDXMLParser::HandleTextL(TDes& aChar) |
|
1731 // |
|
1732 // Called when a character is read in and found to bo outside of an element tag |
|
1733 // |
|
1734 { |
|
1735 // Save the text in a buffer. |
|
1736 // This text will get added as as a child element when the next tag is encounted. |
|
1737 if (iText == NULL) |
|
1738 iText = HBufC::NewL(KGMXMLDefaultTextBufferSize); |
|
1739 |
|
1740 if (iText->Length() == iText->Des().MaxLength()) |
|
1741 // The buffer will overflow if we add another character. |
|
1742 // Need to reallocate. |
|
1743 { |
|
1744 iText = iText->ReAllocL(iText->Des().MaxLength() + KGMXMLDefaultTextBufferSize); |
|
1745 } |
|
1746 |
|
1747 iText->Des().Append(aChar); |
|
1748 } |
|
1749 |
|
1750 TBool CMDXMLParser::EndOfCDataSection() |
|
1751 { |
|
1752 TBool endOfCData = EFalse; |
|
1753 TPtrC cdataEndSection(KXMLEndCDataSection); |
|
1754 TInt instLen = TPtrC(KXMLEndCDataSection).Length()-1; |
|
1755 |
|
1756 __ASSERT_DEBUG(iElementTag != NULL, Panic(ENullMemVarElementTag)); |
|
1757 if(iElementTag->Right(instLen) == cdataEndSection.Left(instLen)) |
|
1758 { |
|
1759 if(iElementTag->Left(instLen) == KXMLStartCDataSection) |
|
1760 endOfCData = ETrue; |
|
1761 } |
|
1762 |
|
1763 return endOfCData; |
|
1764 } |
|
1765 |
|
1766 TInt CMDXMLParser::CheckForStartCData(const TDesC& aTextToCheck) |
|
1767 { |
|
1768 TInt index; |
|
1769 index = aTextToCheck.Find(KXMLStartCDataSection); |
|
1770 return index; |
|
1771 } |
|
1772 |
|
1773 TInt CMDXMLParser::FindDelimiter(TDesC& aDataToSearch, TDesC& aDelimiterToFind) |
|
1774 { |
|
1775 TInt currentOffset = 0; |
|
1776 TInt nextDelimiter = KErrNotFound; |
|
1777 TBool valid = EFalse; |
|
1778 TPtrC unsearchedData(aDataToSearch); |
|
1779 |
|
1780 while (!valid && ((nextDelimiter = unsearchedData.Find(aDelimiterToFind)) != KErrNotFound)) |
|
1781 { |
|
1782 // If this isn't the first time round the loop (When currentOffset == 0) we're moved |
|
1783 // our attention to the character after the delimiter we found, so add one to currentOffset |
|
1784 if(currentOffset != 0) |
|
1785 { |
|
1786 currentOffset += 1; |
|
1787 } |
|
1788 // We have a delimiter, add the position of this to currentOffset |
|
1789 currentOffset += nextDelimiter; |
|
1790 |
|
1791 // Check whether this delimiter is in a CDataSection, it's valid if it isn't |
|
1792 TPtrC delimiterToCheck = aDataToSearch.Left(currentOffset); |
|
1793 valid = !InCDataSection(delimiterToCheck); |
|
1794 |
|
1795 // Move on to the next section of text in case this one wasn't valid |
|
1796 unsearchedData.Set(aDataToSearch.Mid(currentOffset + 1)); |
|
1797 } |
|
1798 |
|
1799 if ((nextDelimiter == KErrNotFound) && (!valid)) |
|
1800 { |
|
1801 return KErrNotFound; |
|
1802 } |
|
1803 else |
|
1804 { |
|
1805 return currentOffset; |
|
1806 } |
|
1807 } |
|
1808 |
|
1809 void CMDXMLParser::AddTextL(CMDXMLElement* aParentElement) |
|
1810 { |
|
1811 if ((aParentElement != NULL) && (iText != NULL)) |
|
1812 // Add any buffered text to the parent element unless it contains only whitespace |
|
1813 { |
|
1814 // Strip off any leading whitespace |
|
1815 TInt stripCounter = 0; |
|
1816 |
|
1817 if (!iPreserve) // GmXML consumes whitespace characters |
|
1818 { |
|
1819 TBool endOfWhitespace = EFalse; |
|
1820 while ((stripCounter < iText->Length()) && (!endOfWhitespace)) |
|
1821 { |
|
1822 // If character is not 0x20 (space) and not between 0x09 and 0x0d |
|
1823 // it isn't whitespace |
|
1824 if( ((*iText)[stripCounter] != 0x20) && |
|
1825 !((*iText)[stripCounter] >= 0x09 && (*iText)[stripCounter] <= 0x0d)) |
|
1826 { |
|
1827 endOfWhitespace = ETrue; |
|
1828 } |
|
1829 else |
|
1830 { |
|
1831 stripCounter++; |
|
1832 } |
|
1833 } |
|
1834 } |
|
1835 |
|
1836 HBufC* strippedText = TPtrC(iText->Right(iText->Length() - stripCounter)).AllocLC(); |
|
1837 |
|
1838 if (strippedText->Length() > 0) |
|
1839 // If there is anything left of the stripped text then entity convert and add it. |
|
1840 { |
|
1841 TPtr toConvert = strippedText->Des(); |
|
1842 TInt error = iEntityConverter->EntityToTextL(toConvert); |
|
1843 if( error != KErrNone ) |
|
1844 { |
|
1845 SetError(error, EXMLIndeterminate); |
|
1846 } |
|
1847 CMDXMLText* textElement = CMDXMLText::NewLC(iXMLDoc); |
|
1848 textElement->SetDataL(*strippedText); |
|
1849 CleanupStack::Pop(textElement); |
|
1850 TInt err = aParentElement->AppendChild(textElement); |
|
1851 if(err != KErrNone) |
|
1852 { |
|
1853 SetError(err, EXMLWorkable); |
|
1854 } |
|
1855 } |
|
1856 |
|
1857 CleanupStack::PopAndDestroy(strippedText); |
|
1858 iText->Des().Zero(); |
|
1859 } |
|
1860 } |
|
1861 |
|
1862 TBool CMDXMLParser::InCDataSection(TDesC& aDataToSearch) |
|
1863 { |
|
1864 TBool inCDataSection = EFalse; |
|
1865 TInt startCData = CheckForStartCData(aDataToSearch); |
|
1866 TInt endCData = 0; |
|
1867 |
|
1868 while ((startCData != KErrNotFound) && !inCDataSection) |
|
1869 { |
|
1870 // We only want to look for the end of the CDataSection in the part of |
|
1871 // aDataToSearch after the start of the CDataSection. We know that the |
|
1872 // first (TPtrC)KXMLStartCDataSection.Length() of the data we're looking |
|
1873 // at won't match because it's the start tag, but it's probably more |
|
1874 // efficient to check the extra few characters than to work out the |
|
1875 // length of the tag so we can ignore them. |
|
1876 startCData += endCData; |
|
1877 TPtrC afterStart = aDataToSearch.Mid(startCData); |
|
1878 endCData = afterStart.Find(KXMLEndCDataSection); |
|
1879 if (endCData == KErrNotFound) |
|
1880 { |
|
1881 // We haven't found a match for the start of the CDataSection so |
|
1882 // we must still be in it -> "<" is valid. |
|
1883 inCDataSection = ETrue; |
|
1884 } |
|
1885 else |
|
1886 { |
|
1887 // We found a match for the start of the CDataSection. Check to |
|
1888 // see if another one has started since then. |
|
1889 endCData += startCData; |
|
1890 TPtrC afterEnd = aDataToSearch.Mid(endCData); |
|
1891 startCData = CheckForStartCData(afterEnd); |
|
1892 } |
|
1893 } |
|
1894 |
|
1895 return inCDataSection; |
|
1896 } |
|
1897 |
|
1898 TInt CMDXMLParser::ParseSingleAttributeL(TDes& aAttributeValue) |
|
1899 { |
|
1900 TInt error = KErrNone; |
|
1901 TInt beginSection = 0; |
|
1902 TInt endSection = aAttributeValue.Find(KXMLStartCDataSection); |
|
1903 |
|
1904 // We've found at least one CDataSection |
|
1905 while(endSection != KErrNotFound) |
|
1906 { |
|
1907 // Entity convert this plain text section |
|
1908 HBufC* textToConvert = TPtrC(aAttributeValue.Mid(beginSection, endSection)).AllocLC(); |
|
1909 TPtr toConvert = textToConvert->Des(); |
|
1910 error = iEntityConverter->EntityToTextL(toConvert); |
|
1911 aAttributeValue.Replace(beginSection, endSection, *textToConvert); |
|
1912 CleanupStack::PopAndDestroy(textToConvert); |
|
1913 |
|
1914 // Move on our markers. We start the new section at the end of the old one. |
|
1915 beginSection += endSection; |
|
1916 // The end of this new section is the end of the CDataSection |
|
1917 endSection = TPtrC(aAttributeValue.Mid(beginSection)).Find(KXMLEndCDataSection); |
|
1918 |
|
1919 if(endSection != KErrNotFound) |
|
1920 { |
|
1921 // Now move on our markers again. Start at the end of the CDataSection, |
|
1922 // plus the length of the end tag, and continue to the beginning of the next one. |
|
1923 beginSection += endSection + TPtrC(KXMLEndCDataSection).Length(); |
|
1924 endSection = TPtrC(aAttributeValue.Mid(beginSection)).Find(KXMLStartCDataSection); |
|
1925 } |
|
1926 else |
|
1927 { |
|
1928 // There's an unterminated CDataSection in our attribute |
|
1929 error = KErrXMLBadAttributeValue; |
|
1930 } |
|
1931 } |
|
1932 |
|
1933 // There are no more CDataSections, entity convert the rest of the string |
|
1934 if(!error) |
|
1935 { |
|
1936 HBufC* textToConvert = TPtrC(aAttributeValue.Mid(beginSection)).AllocLC(); |
|
1937 TPtr toConvert = textToConvert->Des(); |
|
1938 error = iEntityConverter->EntityToTextL(toConvert); |
|
1939 aAttributeValue.Replace(beginSection, (aAttributeValue.Length()-beginSection), *textToConvert); |
|
1940 CleanupStack::PopAndDestroy(textToConvert); |
|
1941 } |
|
1942 |
|
1943 return error; |
|
1944 } |
|
1945 |
|
1946 |
|
1947 EXPORT_C void CMDXMLParser::PlaceholderForRemovedExport1(MMDXMLParserObserver* /*aParserObserver*/) |
|
1948 { |
|
1949 User::Panic(KLDRIMPORT, KLdrImportedOrdinalDoesNotExist); |
|
1950 } |
|
1951 |
|
1952 EXPORT_C void CMDXMLParser::PlaceholderForRemovedExport2(MMDXMLParserObserver* /*aParserObserver*/, MXMLDtd* /*aDtdRepresentation*/) |
|
1953 { |
|
1954 User::Panic(KLDRIMPORT, KLdrImportedOrdinalDoesNotExist); |
|
1955 } |
|
1956 |
|
1957 EXPORT_C void CMDXMLParser::PlaceholderForRemovedExport3() |
|
1958 { |
|
1959 User::Panic(KLDRIMPORT, KLdrImportedOrdinalDoesNotExist); |
|
1960 } |
|
1961 |
|
1962 // End of File |
|