summaryrefslogtreecommitdiff
path: root/parser/htmlparser/nsParser.cpp
blob: d1e521750aa906227abd327b6d4903c6e39c332b (plain)
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
417
418
419
420
421
422
423
424
425
426
427
428
429
430
431
432
433
434
435
436
437
438
439
440
441
442
443
444
445
446
447
448
449
450
451
452
453
454
455
456
457
458
459
460
461
462
463
464
465
466
467
468
469
470
471
472
473
474
475
476
477
478
479
480
481
482
483
484
485
486
487
488
489
490
491
492
493
494
495
496
497
498
499
500
501
502
503
504
505
506
507
508
509
510
511
512
513
514
515
516
517
518
519
520
521
522
523
524
525
526
527
528
529
530
531
532
533
534
535
536
537
538
539
540
541
542
543
544
545
546
547
548
549
550
551
552
553
554
555
556
557
558
559
560
561
562
563
564
565
566
567
568
569
570
571
572
573
574
575
576
577
578
579
580
581
582
583
584
585
586
587
588
589
590
591
592
593
594
595
596
597
598
599
600
601
602
603
604
605
606
607
608
609
610
611
612
613
614
615
616
617
618
619
620
621
622
623
624
625
626
627
628
629
630
631
632
633
634
635
636
637
638
639
640
641
642
643
644
645
646
647
648
649
650
651
652
653
654
655
656
657
658
659
660
661
662
663
664
665
666
667
668
669
670
671
672
673
674
675
676
677
678
679
680
681
682
683
684
685
686
687
688
689
690
691
692
693
694
695
696
697
698
699
700
701
702
703
704
705
706
707
708
709
710
711
712
713
714
715
716
717
718
719
720
721
722
723
724
725
726
727
728
729
730
731
732
733
734
735
736
737
738
739
740
741
742
743
744
745
746
747
748
749
750
751
752
753
754
755
756
757
758
759
760
761
762
763
764
765
766
767
768
769
770
771
772
773
774
775
776
777
778
779
780
781
782
783
784
785
786
787
788
789
790
791
792
793
794
795
796
797
798
799
800
801
802
803
804
805
806
807
808
809
810
811
812
813
814
815
816
817
818
819
820
821
822
823
824
825
826
827
828
829
830
831
832
833
834
835
836
837
838
839
840
841
842
843
844
845
846
847
848
849
850
851
852
853
854
855
856
857
858
859
860
861
862
863
864
865
866
867
868
869
870
871
872
873
874
875
876
877
878
879
880
881
882
883
884
885
886
887
888
889
890
891
892
893
894
895
896
897
898
899
900
901
902
903
904
905
906
907
908
909
910
911
912
913
914
915
916
917
918
919
920
921
922
923
924
925
926
927
928
929
930
931
932
933
934
935
936
937
938
939
940
941
942
943
944
945
946
947
948
949
950
951
952
953
954
955
956
957
958
959
960
961
962
963
964
965
966
967
968
969
970
971
972
973
974
975
976
977
978
979
980
981
982
983
984
985
986
987
988
989
990
991
992
993
994
995
996
997
998
999
1000
1001
1002
1003
1004
1005
1006
1007
1008
1009
1010
1011
1012
1013
1014
1015
1016
1017
1018
1019
1020
1021
1022
1023
1024
1025
1026
1027
1028
1029
1030
1031
1032
1033
1034
1035
1036
1037
1038
1039
1040
1041
1042
1043
1044
1045
1046
1047
1048
1049
1050
1051
1052
1053
1054
1055
1056
1057
1058
1059
1060
1061
1062
1063
1064
1065
1066
1067
1068
1069
1070
1071
1072
1073
1074
1075
1076
1077
1078
1079
1080
1081
1082
1083
1084
1085
1086
1087
1088
1089
1090
1091
1092
1093
1094
1095
1096
1097
1098
1099
1100
1101
1102
1103
1104
1105
1106
1107
1108
1109
1110
1111
1112
1113
1114
1115
1116
1117
1118
1119
1120
1121
1122
1123
1124
1125
1126
1127
1128
1129
1130
1131
1132
1133
1134
1135
1136
1137
1138
1139
1140
1141
1142
1143
1144
1145
1146
1147
1148
1149
1150
1151
1152
1153
1154
1155
1156
1157
1158
1159
1160
1161
1162
1163
1164
1165
1166
1167
1168
1169
1170
1171
1172
1173
1174
1175
1176
1177
1178
1179
1180
1181
1182
1183
1184
1185
1186
1187
1188
1189
1190
1191
1192
1193
1194
1195
1196
1197
1198
1199
1200
1201
1202
1203
1204
1205
1206
1207
1208
1209
1210
1211
1212
1213
1214
1215
1216
1217
1218
1219
1220
1221
1222
1223
1224
1225
1226
1227
1228
1229
1230
1231
1232
1233
1234
1235
1236
1237
1238
1239
1240
1241
1242
1243
1244
1245
1246
1247
1248
1249
1250
1251
1252
1253
1254
1255
1256
1257
1258
1259
1260
1261
1262
1263
1264
1265
1266
1267
1268
1269
1270
1271
1272
1273
1274
1275
1276
1277
1278
1279
1280
1281
1282
1283
1284
1285
1286
1287
1288
1289
1290
1291
1292
1293
1294
1295
1296
1297
1298
1299
1300
1301
1302
1303
1304
1305
1306
1307
1308
1309
1310
1311
1312
1313
1314
1315
1316
1317
1318
1319
1320
1321
1322
1323
1324
1325
1326
1327
1328
1329
1330
1331
1332
1333
1334
1335
1336
1337
1338
1339
1340
1341
1342
1343
1344
1345
1346
1347
1348
1349
1350
1351
1352
1353
1354
1355
1356
1357
1358
1359
1360
1361
1362
1363
1364
1365
1366
1367
1368
1369
1370
1371
1372
1373
1374
1375
1376
1377
1378
1379
1380
1381
1382
1383
1384
1385
1386
1387
1388
1389
1390
1391
1392
1393
1394
1395
1396
1397
1398
1399
1400
1401
1402
1403
1404
1405
1406
1407
1408
1409
1410
1411
1412
1413
1414
1415
1416
1417
1418
1419
1420
1421
1422
1423
1424
1425
1426
1427
1428
1429
1430
1431
1432
1433
1434
1435
1436
1437
1438
1439
1440
1441
1442
1443
1444
1445
1446
1447
1448
1449
1450
1451
1452
1453
1454
1455
1456
1457
1458
1459
1460
1461
1462
1463
1464
1465
1466
1467
1468
1469
1470
1471
1472
1473
1474
1475
1476
1477
1478
1479
1480
1481
1482
1483
1484
1485
1486
1487
1488
1489
1490
1491
1492
1493
1494
1495
1496
1497
1498
1499
1500
1501
1502
1503
1504
1505
1506
1507
1508
1509
1510
1511
1512
1513
1514
1515
1516
1517
1518
1519
1520
1521
1522
1523
1524
1525
1526
1527
1528
1529
1530
1531
1532
1533
1534
1535
1536
1537
1538
1539
1540
1541
1542
1543
1544
1545
1546
1547
1548
1549
1550
1551
1552
1553
1554
1555
1556
1557
1558
1559
1560
1561
1562
1563
1564
1565
1566
1567
1568
1569
1570
1571
1572
1573
1574
1575
1576
1577
1578
1579
1580
1581
1582
1583
1584
1585
1586
1587
1588
1589
1590
1591
1592
1593
1594
1595
1596
1597
1598
1599
1600
/* -*- Mode: C++; tab-width: 2; indent-tabs-mode: nil; c-basic-offset: 2 -*- */
/* vim: set sw=2 ts=2 et tw=79: */
/* This Source Code Form is subject to the terms of the Mozilla Public
 * License, v. 2.0. If a copy of the MPL was not distributed with this
 * file, You can obtain one at http://mozilla.org/MPL/2.0/. */

#include "nsIAtom.h"
#include "nsParser.h"
#include "nsString.h"
#include "nsCRT.h"
#include "nsScanner.h"
#include "plstr.h"
#include "nsIStringStream.h"
#include "nsIChannel.h"
#include "nsICachingChannel.h"
#include "nsIInputStream.h"
#include "CNavDTD.h"
#include "prenv.h"
#include "prlock.h"
#include "prcvar.h"
#include "nsParserCIID.h"
#include "nsReadableUtils.h"
#include "nsCOMPtr.h"
#include "nsExpatDriver.h"
#include "nsIServiceManager.h"
#include "nsICategoryManager.h"
#include "nsISupportsPrimitives.h"
#include "nsIFragmentContentSink.h"
#include "nsStreamUtils.h"
#include "nsHTMLTokenizer.h"
#include "nsDataHashtable.h"
#include "nsXPCOMCIDInternal.h"
#include "nsMimeTypes.h"
#include "mozilla/CondVar.h"
#include "mozilla/Mutex.h"
#include "nsParserConstants.h"
#include "nsCharsetSource.h"
#include "nsContentUtils.h"
#include "nsThreadUtils.h"
#include "nsIHTMLContentSink.h"

#include "mozilla/dom/EncodingUtils.h"
#include "mozilla/dom/ScriptLoader.h"
#include "mozilla/BinarySearch.h"

using namespace mozilla;
using mozilla::dom::EncodingUtils;

#define NS_PARSER_FLAG_PARSER_ENABLED         0x00000002
#define NS_PARSER_FLAG_OBSERVERS_ENABLED      0x00000004
#define NS_PARSER_FLAG_PENDING_CONTINUE_EVENT 0x00000008
#define NS_PARSER_FLAG_FLUSH_TOKENS           0x00000020
#define NS_PARSER_FLAG_CAN_TOKENIZE           0x00000040

//-------------- Begin ParseContinue Event Definition ------------------------
/*
The parser can be explicitly interrupted by passing a return value of
NS_ERROR_HTMLPARSER_INTERRUPTED from BuildModel on the DTD. This will cause
the parser to stop processing and allow the application to return to the event
loop. The data which was left at the time of interruption will be processed
the next time OnDataAvailable is called. If the parser has received its final
chunk of data then OnDataAvailable will no longer be called by the networking
module, so the parser will schedule a nsParserContinueEvent which will call
the parser to process the remaining data after returning to the event loop.
If the parser is interrupted while processing the remaining data it will
schedule another ParseContinueEvent. The processing of data followed by
scheduling of the continue events will proceed until either:

  1) All of the remaining data can be processed without interrupting
  2) The parser has been cancelled.


This capability is currently used in CNavDTD and nsHTMLContentSink. The
nsHTMLContentSink is notified by CNavDTD when a chunk of tokens is going to be
processed and when each token is processed. The nsHTML content sink records
the time when the chunk has started processing and will return
NS_ERROR_HTMLPARSER_INTERRUPTED if the token processing time has exceeded a
threshold called max tokenizing processing time. This allows the content sink
to limit how much data is processed in a single chunk which in turn gates how
much time is spent away from the event loop. Processing smaller chunks of data
also reduces the time spent in subsequent reflows.

This capability is most apparent when loading large documents. If the maximum
token processing time is set small enough the application will remain
responsive during document load.

A side-effect of this capability is that document load is not complete when
the last chunk of data is passed to OnDataAvailable since  the parser may have
been interrupted when the last chunk of data arrived. The document is complete
when all of the document has been tokenized and there aren't any pending
nsParserContinueEvents. This can cause problems if the application assumes
that it can monitor the load requests to determine when the document load has
been completed. This is what happens in Mozilla. The document is considered
completely loaded when all of the load requests have been satisfied. To delay
the document load until all of the parsing has been completed the
nsHTMLContentSink adds a dummy parser load request which is not removed until
the nsHTMLContentSink's DidBuildModel is called. The CNavDTD will not call
DidBuildModel until the final chunk of data has been passed to the parser
through the OnDataAvailable and there aren't any pending
nsParserContineEvents.

Currently the parser is ignores requests to be interrupted during the
processing of script.  This is because a document.write followed by JavaScript
calls to manipulate the DOM may fail if the parser was interrupted during the
document.write.

For more details @see bugzilla bug 76722
*/


class nsParserContinueEvent : public Runnable
{
public:
  RefPtr<nsParser> mParser;

  explicit nsParserContinueEvent(nsParser* aParser)
    : mParser(aParser)
  {}

  NS_IMETHOD Run() override
  {
    mParser->HandleParserContinueEvent(this);
    return NS_OK;
  }
};

//-------------- End ParseContinue Event Definition ------------------------

/**
 *  default constructor
 */
nsParser::nsParser()
{
  Initialize(true);
}

nsParser::~nsParser()
{
  Cleanup();
}

void
nsParser::Initialize(bool aConstructor)
{
  if (aConstructor) {
    // Raw pointer
    mParserContext = 0;
  }
  else {
    // nsCOMPtrs
    mObserver = nullptr;
    mUnusedInput.Truncate();
  }

  mContinueEvent = nullptr;
  mCharsetSource = kCharsetUninitialized;
  mCharset.AssignLiteral("ISO-8859-1");
  mInternalState = NS_OK;
  mStreamStatus = NS_OK;
  mCommand = eViewNormal;
  mFlags = NS_PARSER_FLAG_OBSERVERS_ENABLED |
           NS_PARSER_FLAG_PARSER_ENABLED |
           NS_PARSER_FLAG_CAN_TOKENIZE;

  mProcessingNetworkData = false;
  mIsAboutBlank = false;
}

void
nsParser::Cleanup()
{
#ifdef DEBUG
  if (mParserContext && mParserContext->mPrevContext) {
    NS_WARNING("Extra parser contexts still on the parser stack");
  }
#endif

  while (mParserContext) {
    CParserContext *pc = mParserContext->mPrevContext;
    delete mParserContext;
    mParserContext = pc;
  }

  // It should not be possible for this flag to be set when we are getting
  // destroyed since this flag implies a pending nsParserContinueEvent, which
  // has an owning reference to |this|.
  NS_ASSERTION(!(mFlags & NS_PARSER_FLAG_PENDING_CONTINUE_EVENT), "bad");
}

NS_IMPL_CYCLE_COLLECTION_CLASS(nsParser)

NS_IMPL_CYCLE_COLLECTION_UNLINK_BEGIN(nsParser)
  NS_IMPL_CYCLE_COLLECTION_UNLINK(mDTD)
  NS_IMPL_CYCLE_COLLECTION_UNLINK(mSink)
  NS_IMPL_CYCLE_COLLECTION_UNLINK(mObserver)
NS_IMPL_CYCLE_COLLECTION_UNLINK_END

NS_IMPL_CYCLE_COLLECTION_TRAVERSE_BEGIN(nsParser)
  NS_IMPL_CYCLE_COLLECTION_TRAVERSE(mDTD)
  NS_IMPL_CYCLE_COLLECTION_TRAVERSE(mSink)
  NS_IMPL_CYCLE_COLLECTION_TRAVERSE(mObserver)
  CParserContext *pc = tmp->mParserContext;
  while (pc) {
    cb.NoteXPCOMChild(pc->mTokenizer);
    pc = pc->mPrevContext;
  }
NS_IMPL_CYCLE_COLLECTION_TRAVERSE_END

NS_IMPL_CYCLE_COLLECTING_ADDREF(nsParser)
NS_IMPL_CYCLE_COLLECTING_RELEASE(nsParser)
NS_INTERFACE_MAP_BEGIN_CYCLE_COLLECTION(nsParser)
  NS_INTERFACE_MAP_ENTRY(nsIStreamListener)
  NS_INTERFACE_MAP_ENTRY(nsIParser)
  NS_INTERFACE_MAP_ENTRY(nsIRequestObserver)
  NS_INTERFACE_MAP_ENTRY(nsISupportsWeakReference)
  NS_INTERFACE_MAP_ENTRY_AMBIGUOUS(nsISupports, nsIParser)
NS_INTERFACE_MAP_END

// The parser continue event is posted only if
// all of the data to parse has been passed to ::OnDataAvailable
// and the parser has been interrupted by the content sink
// because the processing of tokens took too long.

nsresult
nsParser::PostContinueEvent()
{
  if (!(mFlags & NS_PARSER_FLAG_PENDING_CONTINUE_EVENT)) {
    // If this flag isn't set, then there shouldn't be a live continue event!
    NS_ASSERTION(!mContinueEvent, "bad");

    // This creates a reference cycle between this and the event that is
    // broken when the event fires.
    nsCOMPtr<nsIRunnable> event = new nsParserContinueEvent(this);
    if (NS_FAILED(NS_DispatchToCurrentThread(event))) {
        NS_WARNING("failed to dispatch parser continuation event");
    } else {
        mFlags |= NS_PARSER_FLAG_PENDING_CONTINUE_EVENT;
        mContinueEvent = event;
    }
  }
  return NS_OK;
}

NS_IMETHODIMP_(void)
nsParser::GetCommand(nsCString& aCommand)
{
  aCommand = mCommandStr;
}

/**
 *  Call this method once you've created a parser, and want to instruct it
 *  about the command which caused the parser to be constructed. For example,
 *  this allows us to select a DTD which can do, say, view-source.
 *
 *  @param   aCommand the command string to set
 */
NS_IMETHODIMP_(void)
nsParser::SetCommand(const char* aCommand)
{
  mCommandStr.Assign(aCommand);
  if (mCommandStr.EqualsLiteral("view-source")) {
    mCommand = eViewSource;
  } else if (mCommandStr.EqualsLiteral("view-fragment")) {
    mCommand = eViewFragment;
  } else {
    mCommand = eViewNormal;
  }
}

/**
 *  Call this method once you've created a parser, and want to instruct it
 *  about the command which caused the parser to be constructed. For example,
 *  this allows us to select a DTD which can do, say, view-source.
 *
 *  @param   aParserCommand the command to set
 */
NS_IMETHODIMP_(void)
nsParser::SetCommand(eParserCommands aParserCommand)
{
  mCommand = aParserCommand;
}

/**
 *  Call this method once you've created a parser, and want to instruct it
 *  about what charset to load
 *
 *  @param   aCharset- the charset of a document
 *  @param   aCharsetSource- the source of the charset
 */
NS_IMETHODIMP_(void)
nsParser::SetDocumentCharset(const nsACString& aCharset, int32_t aCharsetSource)
{
  mCharset = aCharset;
  mCharsetSource = aCharsetSource;
  if (mParserContext && mParserContext->mScanner) {
     mParserContext->mScanner->SetDocumentCharset(aCharset, aCharsetSource);
  }
}

void
nsParser::SetSinkCharset(nsACString& aCharset)
{
  if (mSink) {
    mSink->SetDocumentCharset(aCharset);
  }
}

/**
 *  This method gets called in order to set the content
 *  sink for this parser to dump nodes to.
 *
 *  @param   nsIContentSink interface for node receiver
 */
NS_IMETHODIMP_(void)
nsParser::SetContentSink(nsIContentSink* aSink)
{
  NS_PRECONDITION(aSink, "sink cannot be null!");
  mSink = aSink;

  if (mSink) {
    mSink->SetParser(this);
    nsCOMPtr<nsIHTMLContentSink> htmlSink = do_QueryInterface(mSink);
    if (htmlSink) {
      mIsAboutBlank = true;
    }
  }
}

/**
 * retrieve the sink set into the parser
 * @return  current sink
 */
NS_IMETHODIMP_(nsIContentSink*)
nsParser::GetContentSink()
{
  return mSink;
}

static nsIDTD*
FindSuitableDTD(CParserContext& aParserContext)
{
  // We always find a DTD.
  aParserContext.mAutoDetectStatus = ePrimaryDetect;

  // Quick check for view source.
  MOZ_ASSERT(aParserContext.mParserCommand != eViewSource,
             "The old parser is not supposed to be used for View Source "
             "anymore.");

  // Now see if we're parsing HTML (which, as far as we're concerned, simply
  // means "not XML").
  if (aParserContext.mDocType != eXML) {
    return new CNavDTD();
  }

  // If we're here, then we'd better be parsing XML.
  NS_ASSERTION(aParserContext.mDocType == eXML, "What are you trying to send me, here?");
  return new nsExpatDriver();
}

NS_IMETHODIMP
nsParser::CancelParsingEvents()
{
  if (mFlags & NS_PARSER_FLAG_PENDING_CONTINUE_EVENT) {
    NS_ASSERTION(mContinueEvent, "mContinueEvent is null");
    // Revoke the pending continue parsing event
    mContinueEvent = nullptr;
    mFlags &= ~NS_PARSER_FLAG_PENDING_CONTINUE_EVENT;
  }
  return NS_OK;
}

////////////////////////////////////////////////////////////////////////

/**
 * Evalutes EXPR1 and EXPR2 exactly once each, in that order.  Stores the value
 * of EXPR2 in RV is EXPR2 fails, otherwise RV contains the result of EXPR1
 * (which could be success or failure).
 *
 * To understand the motivation for this construct, consider these example
 * methods:
 *
 *   nsresult nsSomething::DoThatThing(nsIWhatever* obj) {
 *     nsresult rv = NS_OK;
 *     ...
 *     return obj->DoThatThing();
 *     NS_ENSURE_SUCCESS(rv, rv);
 *     ...
 *     return rv;
 *   }
 *
 *   void nsCaller::MakeThingsHappen() {
 *     return mSomething->DoThatThing(mWhatever);
 *   }
 *
 * Suppose, for whatever reason*, we want to shift responsibility for calling
 * mWhatever->DoThatThing() from nsSomething::DoThatThing up to
 * nsCaller::MakeThingsHappen.  We might rewrite the two methods as follows:
 *
 *   nsresult nsSomething::DoThatThing() {
 *     nsresult rv = NS_OK;
 *     ...
 *     ...
 *     return rv;
 *   }
 *
 *   void nsCaller::MakeThingsHappen() {
 *     nsresult rv;
 *     PREFER_LATTER_ERROR_CODE(mSomething->DoThatThing(),
 *                              mWhatever->DoThatThing(),
 *                              rv);
 *     return rv;
 *   }
 *
 * *Possible reasons include: nsCaller doesn't want to give mSomething access
 * to mWhatever, nsCaller wants to guarantee that mWhatever->DoThatThing() will
 * be called regardless of how nsSomething::DoThatThing behaves, &c.
 */
#define PREFER_LATTER_ERROR_CODE(EXPR1, EXPR2, RV) {                          \
  nsresult RV##__temp = EXPR1;                                                \
  RV = EXPR2;                                                                 \
  if (NS_FAILED(RV)) {                                                        \
    RV = RV##__temp;                                                          \
  }                                                                           \
}

/**
 * This gets called just prior to the model actually
 * being constructed. It's important to make this the
 * last thing that happens right before parsing, so we
 * can delay until the last moment the resolution of
 * which DTD to use (unless of course we're assigned one).
 */
nsresult
nsParser::WillBuildModel(nsString& aFilename)
{
  if (!mParserContext)
    return kInvalidParserContext;

  if (eUnknownDetect != mParserContext->mAutoDetectStatus)
    return NS_OK;

  if (eDTDMode_unknown == mParserContext->mDTDMode ||
      eDTDMode_autodetect == mParserContext->mDTDMode) {
    if (mIsAboutBlank) {
      mParserContext->mDTDMode = eDTDMode_quirks;
      mParserContext->mDocType = eHTML_Quirks;
    } else {
      mParserContext->mDTDMode = eDTDMode_full_standards;
      mParserContext->mDocType = eXML;
    }
  } // else XML fragment with nested parser context

  NS_ASSERTION(!mDTD || !mParserContext->mPrevContext,
               "Clobbering DTD for non-root parser context!");
  mDTD = FindSuitableDTD(*mParserContext);
  NS_ENSURE_TRUE(mDTD, NS_ERROR_OUT_OF_MEMORY);

  nsITokenizer* tokenizer;
  nsresult rv = mParserContext->GetTokenizer(mDTD, mSink, tokenizer);
  NS_ENSURE_SUCCESS(rv, rv);

  rv = mDTD->WillBuildModel(*mParserContext, tokenizer, mSink);
  nsresult sinkResult = mSink->WillBuildModel(mDTD->GetMode());
  // nsIDTD::WillBuildModel used to be responsible for calling
  // nsIContentSink::WillBuildModel, but that obligation isn't expressible
  // in the nsIDTD interface itself, so it's sounder and simpler to give that
  // responsibility back to the parser. The former behavior of the DTD was to
  // NS_ENSURE_SUCCESS the sink WillBuildModel call, so if the sink returns
  // failure we should use sinkResult instead of rv, to preserve the old error
  // handling behavior of the DTD:
  return NS_FAILED(sinkResult) ? sinkResult : rv;
}

/**
 * This gets called when the parser is done with its input.
 * Note that the parser may have been called recursively, so we
 * have to check for a prev. context before closing out the DTD/sink.
 */
nsresult
nsParser::DidBuildModel(nsresult anErrorCode)
{
  nsresult result = anErrorCode;

  if (IsComplete()) {
    if (mParserContext && !mParserContext->mPrevContext) {
      // Let sink know if we're about to end load because we've been terminated.
      // In that case we don't want it to run deferred scripts.
      bool terminated = mInternalState == NS_ERROR_HTMLPARSER_STOPPARSING;
      if (mDTD && mSink) {
        nsresult dtdResult =  mDTD->DidBuildModel(anErrorCode),
                sinkResult = mSink->DidBuildModel(terminated);
        // nsIDTD::DidBuildModel used to be responsible for calling
        // nsIContentSink::DidBuildModel, but that obligation isn't expressible
        // in the nsIDTD interface itself, so it's sounder and simpler to give
        // that responsibility back to the parser. The former behavior of the
        // DTD was to NS_ENSURE_SUCCESS the sink DidBuildModel call, so if the
        // sink returns failure we should use sinkResult instead of dtdResult,
        // to preserve the old error handling behavior of the DTD:
        result = NS_FAILED(sinkResult) ? sinkResult : dtdResult;
      }

      //Ref. to bug 61462.
      mParserContext->mRequest = nullptr;
    }
  }

  return result;
}

/**
 * This method adds a new parser context to the list,
 * pushing the current one to the next position.
 *
 * @param   ptr to new context
 */
void
nsParser::PushContext(CParserContext& aContext)
{
  NS_ASSERTION(aContext.mPrevContext == mParserContext,
               "Trying to push a context whose previous context differs from "
               "the current parser context.");
  mParserContext = &aContext;
}

/**
 * This method pops the topmost context off the stack,
 * returning it to the user. The next context  (if any)
 * becomes the current context.
 * @update	gess7/22/98
 * @return  prev. context
 */
CParserContext*
nsParser::PopContext()
{
  CParserContext* oldContext = mParserContext;
  if (oldContext) {
    mParserContext = oldContext->mPrevContext;
    if (mParserContext) {
      // If the old context was blocked, propagate the blocked state
      // back to the new one. Also, propagate the stream listener state
      // but don't override onStop state to guarantee the call to DidBuildModel().
      if (mParserContext->mStreamListenerState != eOnStop) {
        mParserContext->mStreamListenerState = oldContext->mStreamListenerState;
      }
    }
  }
  return oldContext;
}

/**
 *  Call this when you want control whether or not the parser will parse
 *  and tokenize input (TRUE), or whether it just caches input to be
 *  parsed later (FALSE).
 *
 *  @param   aState determines whether we parse/tokenize or just cache.
 *  @return  current state
 */
void
nsParser::SetUnusedInput(nsString& aBuffer)
{
  mUnusedInput = aBuffer;
}

/**
 *  Call this when you want to *force* the parser to terminate the
 *  parsing process altogether. This is binary -- so once you terminate
 *  you can't resume without restarting altogether.
 */
NS_IMETHODIMP
nsParser::Terminate(void)
{
  // We should only call DidBuildModel once, so don't do anything if this is
  // the second time that Terminate has been called.
  if (mInternalState == NS_ERROR_HTMLPARSER_STOPPARSING) {
    return NS_OK;
  }

  nsresult result = NS_OK;
  // XXX - [ until we figure out a way to break parser-sink circularity ]
  // Hack - Hold a reference until we are completely done...
  nsCOMPtr<nsIParser> kungFuDeathGrip(this);
  mInternalState = result = NS_ERROR_HTMLPARSER_STOPPARSING;

  // CancelParsingEvents must be called to avoid leaking the nsParser object
  // @see bug 108049
  // If NS_PARSER_FLAG_PENDING_CONTINUE_EVENT is set then CancelParsingEvents
  // will reset it so DidBuildModel will call DidBuildModel on the DTD. Note:
  // The IsComplete() call inside of DidBuildModel looks at the pendingContinueEvents flag.
  CancelParsingEvents();

  // If we got interrupted in the middle of a document.write, then we might
  // have more than one parser context on our parsercontext stack. This has
  // the effect of making DidBuildModel a no-op, meaning that we never call
  // our sink's DidBuildModel and break the reference cycle, causing a leak.
  // Since we're getting terminated, we manually clean up our context stack.
  while (mParserContext && mParserContext->mPrevContext) {
    CParserContext *prev = mParserContext->mPrevContext;
    delete mParserContext;
    mParserContext = prev;
  }

  if (mDTD) {
    mDTD->Terminate();
    DidBuildModel(result);
  } else if (mSink) {
    // We have no parser context or no DTD yet (so we got terminated before we
    // got any data).  Manually break the reference cycle with the sink.
    result = mSink->DidBuildModel(true);
    NS_ENSURE_SUCCESS(result, result);
  }

  return NS_OK;
}

NS_IMETHODIMP
nsParser::ContinueInterruptedParsing()
{
  // If there are scripts executing, then the content sink is jumping the gun
  // (probably due to a synchronous XMLHttpRequest) and will re-enable us
  // later, see bug 460706.
  if (!IsOkToProcessNetworkData()) {
    return NS_OK;
  }

  // If the stream has already finished, there's a good chance
  // that we might start closing things down when the parser
  // is reenabled. To make sure that we're not deleted across
  // the reenabling process, hold a reference to ourselves.
  nsresult result=NS_OK;
  nsCOMPtr<nsIParser> kungFuDeathGrip(this);
  nsCOMPtr<nsIContentSink> sinkDeathGrip(mSink);

#ifdef DEBUG
  if (!(mFlags & NS_PARSER_FLAG_PARSER_ENABLED)) {
    NS_WARNING("Don't call ContinueInterruptedParsing on a blocked parser.");
  }
#endif

  bool isFinalChunk = mParserContext &&
                        mParserContext->mStreamListenerState == eOnStop;

  mProcessingNetworkData = true;
  if (sinkDeathGrip) {
    sinkDeathGrip->WillParse();
  }
  result = ResumeParse(true, isFinalChunk); // Ref. bug 57999
  mProcessingNetworkData = false;

  if (result != NS_OK) {
    result=mInternalState;
  }

  return result;
}

/**
 *  Stops parsing temporarily. That's it will prevent the
 *  parser from building up content model.
 */
NS_IMETHODIMP_(void)
nsParser::BlockParser()
{
  mFlags &= ~NS_PARSER_FLAG_PARSER_ENABLED;
}

/**
 *  Open up the parser for tokenization, building up content
 *  model..etc. However, this method does not resume parsing
 *  automatically. It's the callers' responsibility to restart
 *  the parsing engine.
 */
NS_IMETHODIMP_(void)
nsParser::UnblockParser()
{
  if (!(mFlags & NS_PARSER_FLAG_PARSER_ENABLED)) {
    mFlags |= NS_PARSER_FLAG_PARSER_ENABLED;
  } else {
    NS_WARNING("Trying to unblock an unblocked parser.");
  }
}

NS_IMETHODIMP_(void)
nsParser::ContinueInterruptedParsingAsync()
{
  mSink->ContinueInterruptedParsingAsync();
}

/**
 * Call this to query whether the parser is enabled or not.
 */
NS_IMETHODIMP_(bool)
nsParser::IsParserEnabled()
{
  return (mFlags & NS_PARSER_FLAG_PARSER_ENABLED) != 0;
}

/**
 * Call this to query whether the parser thinks it's done with parsing.
 */
NS_IMETHODIMP_(bool)
nsParser::IsComplete()
{
  return !(mFlags & NS_PARSER_FLAG_PENDING_CONTINUE_EVENT);
}


void nsParser::HandleParserContinueEvent(nsParserContinueEvent *ev)
{
  // Ignore any revoked continue events...
  if (mContinueEvent != ev)
    return;

  mFlags &= ~NS_PARSER_FLAG_PENDING_CONTINUE_EVENT;
  mContinueEvent = nullptr;

  NS_ASSERTION(IsOkToProcessNetworkData(),
               "Interrupted in the middle of a script?");
  ContinueInterruptedParsing();
}

bool
nsParser::IsInsertionPointDefined()
{
  return false;
}

void
nsParser::PushDefinedInsertionPoint()
{
}

void
nsParser::PopDefinedInsertionPoint()
{
}

void
nsParser::MarkAsNotScriptCreated(const char* aCommand)
{
}

bool
nsParser::IsScriptCreated()
{
  return false;
}

/**
 *  This is the main controlling routine in the parsing process.
 *  Note that it may get called multiple times for the same scanner,
 *  since this is a pushed based system, and all the tokens may
 *  not have been consumed by the scanner during a given invocation
 *  of this method.
 */
NS_IMETHODIMP
nsParser::Parse(nsIURI* aURL,
                nsIRequestObserver* aListener,
                void* aKey,
                nsDTDMode aMode)
{

  NS_PRECONDITION(aURL, "Error: Null URL given");

  nsresult result=kBadURL;
  mObserver = aListener;

  if (aURL) {
    nsAutoCString spec;
    nsresult rv = aURL->GetSpec(spec);
    if (rv != NS_OK) {
      return rv;
    }
    NS_ConvertUTF8toUTF16 theName(spec);

    nsScanner* theScanner = new nsScanner(theName, false);
    CParserContext* pc = new CParserContext(mParserContext, theScanner, aKey,
                                            mCommand, aListener);
    if (pc && theScanner) {
      pc->mMultipart = true;
      pc->mContextType = CParserContext::eCTURL;
      pc->mDTDMode = aMode;
      PushContext(*pc);

      result = NS_OK;
    } else {
      result = mInternalState = NS_ERROR_HTMLPARSER_BADCONTEXT;
    }
  }
  return result;
}

/**
 * Used by XML fragment parsing below.
 *
 * @param   aSourceBuffer contains a string-full of real content
 */
nsresult
nsParser::Parse(const nsAString& aSourceBuffer,
                void* aKey,
                bool aLastCall)
{
  nsresult result = NS_OK;

  // Don't bother if we're never going to parse this.
  if (mInternalState == NS_ERROR_HTMLPARSER_STOPPARSING) {
    return result;
  }

  if (!aLastCall && aSourceBuffer.IsEmpty()) {
    // Nothing is being passed to the parser so return
    // immediately. mUnusedInput will get processed when
    // some data is actually passed in.
    // But if this is the last call, make sure to finish up
    // stuff correctly.
    return result;
  }

  // Maintain a reference to ourselves so we don't go away
  // till we're completely done.
  nsCOMPtr<nsIParser> kungFuDeathGrip(this);

  if (aLastCall || !aSourceBuffer.IsEmpty() || !mUnusedInput.IsEmpty()) {
    // Note: The following code will always find the parser context associated
    // with the given key, even if that context has been suspended (e.g., for
    // another document.write call). This doesn't appear to be exactly what IE
    // does in the case where this happens, but this makes more sense.
    CParserContext* pc = mParserContext;
    while (pc && pc->mKey != aKey) {
      pc = pc->mPrevContext;
    }

    if (!pc) {
      // Only make a new context if we don't have one, OR if we do, but has a
      // different context key.
      nsScanner* theScanner = new nsScanner(mUnusedInput);
      NS_ENSURE_TRUE(theScanner, NS_ERROR_OUT_OF_MEMORY);

      eAutoDetectResult theStatus = eUnknownDetect;

      if (mParserContext &&
          mParserContext->mMimeType.EqualsLiteral("application/xml")) {
        // Ref. Bug 90379
        NS_ASSERTION(mDTD, "How come the DTD is null?");

        if (mParserContext) {
          theStatus = mParserContext->mAutoDetectStatus;
          // Added this to fix bug 32022.
        }
      }

      pc = new CParserContext(mParserContext, theScanner, aKey, mCommand,
                              0, theStatus, aLastCall);
      NS_ENSURE_TRUE(pc, NS_ERROR_OUT_OF_MEMORY);

      PushContext(*pc);

      pc->mMultipart = !aLastCall; // By default
      if (pc->mPrevContext) {
        pc->mMultipart |= pc->mPrevContext->mMultipart;
      }

      // Start fix bug 40143
      if (pc->mMultipart) {
        pc->mStreamListenerState = eOnDataAvail;
        if (pc->mScanner) {
          pc->mScanner->SetIncremental(true);
        }
      } else {
        pc->mStreamListenerState = eOnStop;
        if (pc->mScanner) {
          pc->mScanner->SetIncremental(false);
        }
      }
      // end fix for 40143

      pc->mContextType=CParserContext::eCTString;
      pc->SetMimeType(NS_LITERAL_CSTRING("application/xml"));
      pc->mDTDMode = eDTDMode_full_standards;

      mUnusedInput.Truncate();

      pc->mScanner->Append(aSourceBuffer);
      // Do not interrupt document.write() - bug 95487
      result = ResumeParse(false, false, false);
    } else {
      pc->mScanner->Append(aSourceBuffer);
      if (!pc->mPrevContext) {
        // Set stream listener state to eOnStop, on the final context - Fix 68160,
        // to guarantee DidBuildModel() call - Fix 36148
        if (aLastCall) {
          pc->mStreamListenerState = eOnStop;
          pc->mScanner->SetIncremental(false);
        }

        if (pc == mParserContext) {
          // If pc is not mParserContext, then this call to ResumeParse would
          // do the wrong thing and try to continue parsing using
          // mParserContext. We need to wait to actually resume parsing on pc.
          ResumeParse(false, false, false);
        }
      }
    }
  }

  return result;
}

NS_IMETHODIMP
nsParser::ParseFragment(const nsAString& aSourceBuffer,
                        nsTArray<nsString>& aTagStack)
{
  nsresult result = NS_OK;
  nsAutoString  theContext;
  uint32_t theCount = aTagStack.Length();
  uint32_t theIndex = 0;

  // Disable observers for fragments
  mFlags &= ~NS_PARSER_FLAG_OBSERVERS_ENABLED;

  for (theIndex = 0; theIndex < theCount; theIndex++) {
    theContext.Append('<');
    theContext.Append(aTagStack[theCount - theIndex - 1]);
    theContext.Append('>');
  }

  if (theCount == 0) {
    // Ensure that the buffer is not empty. Because none of the DTDs care
    // about leading whitespace, this doesn't change the result.
    theContext.Assign(' ');
  }

  // First, parse the context to build up the DTD's tag stack. Note that we
  // pass false for the aLastCall parameter.
  result = Parse(theContext,
                 (void*)&theContext,
                 false);
  if (NS_FAILED(result)) {
    mFlags |= NS_PARSER_FLAG_OBSERVERS_ENABLED;
    return result;
  }

  if (!mSink) {
    // Parse must have failed in the XML case and so the sink was killed.
    return NS_ERROR_HTMLPARSER_STOPPARSING;
  }

  nsCOMPtr<nsIFragmentContentSink> fragSink = do_QueryInterface(mSink);
  NS_ASSERTION(fragSink, "ParseFragment requires a fragment content sink");

  fragSink->WillBuildContent();
  // Now, parse the actual content. Note that this is the last call
  // for HTML content, but for XML, we will want to build and parse
  // the end tags.  However, if tagStack is empty, it's the last call
  // for XML as well.
  if (theCount == 0) {
    result = Parse(aSourceBuffer,
                   &theContext,
                   true);
    fragSink->DidBuildContent();
  } else {
    // Add an end tag chunk, so expat will read the whole source buffer,
    // and not worry about ']]' etc.
    result = Parse(aSourceBuffer + NS_LITERAL_STRING("</"),
                   &theContext,
                   false);
    fragSink->DidBuildContent();

    if (NS_SUCCEEDED(result)) {
      nsAutoString endContext;
      for (theIndex = 0; theIndex < theCount; theIndex++) {
         // we already added an end tag chunk above
        if (theIndex > 0) {
          endContext.AppendLiteral("</");
        }

        nsString& thisTag = aTagStack[theIndex];
        // was there an xmlns=?
        int32_t endOfTag = thisTag.FindChar(char16_t(' '));
        if (endOfTag == -1) {
          endContext.Append(thisTag);
        } else {
          endContext.Append(Substring(thisTag,0,endOfTag));
        }

        endContext.Append('>');
      }

      result = Parse(endContext,
                     &theContext,
                     true);
    }
  }

  mFlags |= NS_PARSER_FLAG_OBSERVERS_ENABLED;

  return result;
}

/**
 *  This routine is called to cause the parser to continue parsing its
 *  underlying stream.  This call allows the parse process to happen in
 *  chunks, such as when the content is push based, and we need to parse in
 *  pieces.
 *
 *  An interesting change in how the parser gets used has led us to add extra
 *  processing to this method.  The case occurs when the parser is blocked in
 *  one context, and gets a parse(string) call in another context.  In this
 *  case, the parserContexts are linked. No problem.
 *
 *  The problem is that Parse(string) assumes that it can proceed unabated,
 *  but if the parser is already blocked that assumption is false. So we
 *  needed to add a mechanism here to allow the parser to continue to process
 *  (the pop and free) contexts until 1) it get's blocked again; 2) it runs
 *  out of contexts.
 *
 *
 *  @param   allowItertion : set to true if non-script resumption is requested
 *  @param   aIsFinalChunk : tells us when the last chunk of data is provided.
 *  @return  error code -- 0 if ok, non-zero if error.
 */
nsresult
nsParser::ResumeParse(bool allowIteration, bool aIsFinalChunk,
                      bool aCanInterrupt)
{
  nsresult result = NS_OK;

  if ((mFlags & NS_PARSER_FLAG_PARSER_ENABLED) &&
      mInternalState != NS_ERROR_HTMLPARSER_STOPPARSING) {

    result = WillBuildModel(mParserContext->mScanner->GetFilename());
    if (NS_FAILED(result)) {
      mFlags &= ~NS_PARSER_FLAG_CAN_TOKENIZE;
      return result;
    }

    if (mDTD) {
      mSink->WillResume();
      bool theIterationIsOk = true;

      while (result == NS_OK && theIterationIsOk) {
        if (!mUnusedInput.IsEmpty() && mParserContext->mScanner) {
          // -- Ref: Bug# 22485 --
          // Insert the unused input into the source buffer
          // as if it was read from the input stream.
          // Adding UngetReadable() per vidur!!
          mParserContext->mScanner->UngetReadable(mUnusedInput);
          mUnusedInput.Truncate(0);
        }

        // Only allow parsing to be interrupted in the subsequent call to
        // build model.
        nsresult theTokenizerResult = (mFlags & NS_PARSER_FLAG_CAN_TOKENIZE)
                                      ? Tokenize(aIsFinalChunk)
                                      : NS_OK;
        result = BuildModel();

        if (result == NS_ERROR_HTMLPARSER_INTERRUPTED && aIsFinalChunk) {
          PostContinueEvent();
        }

        theIterationIsOk = theTokenizerResult != kEOF &&
                           result != NS_ERROR_HTMLPARSER_INTERRUPTED;

        // Make sure not to stop parsing too early. Therefore, before shutting
        // down the parser, it's important to check whether the input buffer
        // has been scanned to completion (theTokenizerResult should be kEOF).
        // kEOF -> End of buffer.

        // If we're told to block the parser, we disable all further parsing
        // (and cache any data coming in) until the parser is re-enabled.
        if (NS_ERROR_HTMLPARSER_BLOCK == result) {
          mSink->WillInterrupt();
          if (mFlags & NS_PARSER_FLAG_PARSER_ENABLED) {
            // If we were blocked by a recursive invocation, don't re-block.
            BlockParser();
          }
          return NS_OK;
        }
        if (NS_ERROR_HTMLPARSER_STOPPARSING == result) {
          // Note: Parser Terminate() calls DidBuildModel.
          if (mInternalState != NS_ERROR_HTMLPARSER_STOPPARSING) {
            DidBuildModel(mStreamStatus);
            mInternalState = result;
          }

          return NS_OK;
        }
        if ((NS_OK == result && theTokenizerResult == kEOF) ||
             result == NS_ERROR_HTMLPARSER_INTERRUPTED) {
          bool theContextIsStringBased =
            CParserContext::eCTString == mParserContext->mContextType;

          if (mParserContext->mStreamListenerState == eOnStop ||
              !mParserContext->mMultipart || theContextIsStringBased) {
            if (!mParserContext->mPrevContext) {
              if (mParserContext->mStreamListenerState == eOnStop) {
                DidBuildModel(mStreamStatus);
                return NS_OK;
              }
            } else {
              CParserContext* theContext = PopContext();
              if (theContext) {
                theIterationIsOk = allowIteration && theContextIsStringBased;
                if (theContext->mCopyUnused) {
                  if (!theContext->mScanner->CopyUnusedData(mUnusedInput)) {
                    mInternalState = NS_ERROR_OUT_OF_MEMORY;
                  }
                }

                delete theContext;
              }

              result = mInternalState;
              aIsFinalChunk = mParserContext &&
                              mParserContext->mStreamListenerState == eOnStop;
              // ...then intentionally fall through to mSink->WillInterrupt()...
            }
          }
        }

        if (theTokenizerResult == kEOF ||
            result == NS_ERROR_HTMLPARSER_INTERRUPTED) {
          result = (result == NS_ERROR_HTMLPARSER_INTERRUPTED) ? NS_OK : result;
          mSink->WillInterrupt();
        }
      }
    } else {
      mInternalState = result = NS_ERROR_HTMLPARSER_UNRESOLVEDDTD;
    }
  }

  return (result == NS_ERROR_HTMLPARSER_INTERRUPTED) ? NS_OK : result;
}

/**
 *  This is where we loop over the tokens created in the
 *  tokenization phase, and try to make sense out of them.
 */
nsresult
nsParser::BuildModel()
{
  nsITokenizer* theTokenizer = nullptr;

  nsresult result = NS_OK;
  if (mParserContext) {
    result = mParserContext->GetTokenizer(mDTD, mSink, theTokenizer);
  }

  if (NS_SUCCEEDED(result)) {
    if (mDTD) {
      result = mDTD->BuildModel(theTokenizer, mSink);
    }
  } else {
    mInternalState = result = NS_ERROR_HTMLPARSER_BADTOKENIZER;
  }
  return result;
}

/*******************************************************************
  These methods are used to talk to the netlib system...
 *******************************************************************/

nsresult
nsParser::OnStartRequest(nsIRequest *request, nsISupports* aContext)
{
  NS_PRECONDITION(eNone == mParserContext->mStreamListenerState,
                  "Parser's nsIStreamListener API was not setup "
                  "correctly in constructor.");
  if (mObserver) {
    mObserver->OnStartRequest(request, aContext);
  }
  mParserContext->mStreamListenerState = eOnStart;
  mParserContext->mAutoDetectStatus = eUnknownDetect;
  mParserContext->mRequest = request;

  NS_ASSERTION(!mParserContext->mPrevContext,
               "Clobbering DTD for non-root parser context!");
  mDTD = nullptr;

  nsresult rv;
  nsAutoCString contentType;
  nsCOMPtr<nsIChannel> channel = do_QueryInterface(request);
  if (channel) {
    rv = channel->GetContentType(contentType);
    if (NS_SUCCEEDED(rv)) {
      mParserContext->SetMimeType(contentType);
    }
  }

  rv = NS_OK;

  return rv;
}

static bool
ExtractCharsetFromXmlDeclaration(const unsigned char* aBytes, int32_t aLen,
                                 nsCString& oCharset)
{
  // This code is rather pointless to have. Might as well reuse expat as
  // seen in nsHtml5StreamParser. -- hsivonen
  oCharset.Truncate();
  if ((aLen >= 5) &&
      ('<' == aBytes[0]) &&
      ('?' == aBytes[1]) &&
      ('x' == aBytes[2]) &&
      ('m' == aBytes[3]) &&
      ('l' == aBytes[4])) {
    int32_t i;
    bool versionFound = false, encodingFound = false;
    for (i = 6; i < aLen && !encodingFound; ++i) {
      // end of XML declaration?
      if ((((char*) aBytes)[i] == '?') &&
          ((i + 1) < aLen) &&
          (((char*) aBytes)[i + 1] == '>')) {
        break;
      }
      // Version is required.
      if (!versionFound) {
        // Want to avoid string comparisons, hence looking for 'n'
        // and only if found check the string leading to it. Not
        // foolproof, but fast.
        // The shortest string allowed before this is  (strlen==13):
        // <?xml version
        if ((((char*) aBytes)[i] == 'n') &&
            (i >= 12) &&
            (0 == PL_strncmp("versio", (char*) (aBytes + i - 6), 6))) {
          // Fast forward through version
          char q = 0;
          for (++i; i < aLen; ++i) {
            char qi = ((char*) aBytes)[i];
            if (qi == '\'' || qi == '"') {
              if (q && q == qi) {
                //  ending quote
                versionFound = true;
                break;
              } else {
                // Starting quote
                q = qi;
              }
            }
          }
        }
      } else {
        // encoding must follow version
        // Want to avoid string comparisons, hence looking for 'g'
        // and only if found check the string leading to it. Not
        // foolproof, but fast.
        // The shortest allowed string before this (strlen==26):
        // <?xml version="1" encoding
        if ((((char*) aBytes)[i] == 'g') && (i >= 25) && (0 == PL_strncmp(
            "encodin", (char*) (aBytes + i - 7), 7))) {
          int32_t encStart = 0;
          char q = 0;
          for (++i; i < aLen; ++i) {
            char qi = ((char*) aBytes)[i];
            if (qi == '\'' || qi == '"') {
              if (q && q == qi) {
                int32_t count = i - encStart;
                // encoding value is invalid if it is UTF-16
                if (count > 0 && PL_strncasecmp("UTF-16",
                    (char*) (aBytes + encStart), count)) {
                  oCharset.Assign((char*) (aBytes + encStart), count);
                }
                encodingFound = true;
                break;
              } else {
                encStart = i + 1;
                q = qi;
              }
            }
          }
        }
      } // if (!versionFound)
    } // for
  }
  return !oCharset.IsEmpty();
}

inline char
GetNextChar(nsACString::const_iterator& aStart,
            nsACString::const_iterator& aEnd)
{
  NS_ASSERTION(aStart != aEnd, "end of buffer");
  return (++aStart != aEnd) ? *aStart : '\0';
}

static nsresult
NoOpParserWriteFunc(nsIInputStream* in,
                void* closure,
                const char* fromRawSegment,
                uint32_t toOffset,
                uint32_t count,
                uint32_t *writeCount)
{
  *writeCount = count;
  return NS_OK;
}

typedef struct {
  bool mNeedCharsetCheck;
  nsParser* mParser;
  nsScanner* mScanner;
  nsIRequest* mRequest;
} ParserWriteStruct;

/*
 * This function is invoked as a result of a call to a stream's
 * ReadSegments() method. It is called for each contiguous buffer
 * of data in the underlying stream or pipe. Using ReadSegments
 * allows us to avoid copying data to read out of the stream.
 */
static nsresult
ParserWriteFunc(nsIInputStream* in,
                void* closure,
                const char* fromRawSegment,
                uint32_t toOffset,
                uint32_t count,
                uint32_t *writeCount)
{
  nsresult result;
  ParserWriteStruct* pws = static_cast<ParserWriteStruct*>(closure);
  const unsigned char* buf =
    reinterpret_cast<const unsigned char*> (fromRawSegment);
  uint32_t theNumRead = count;

  if (!pws) {
    return NS_ERROR_FAILURE;
  }

  if (pws->mNeedCharsetCheck) {
    pws->mNeedCharsetCheck = false;
    int32_t source;
    nsAutoCString preferred;
    nsAutoCString maybePrefer;
    pws->mParser->GetDocumentCharset(preferred, source);

    // This code was bogus when I found it. It expects the BOM or the XML
    // declaration to be entirely in the first network buffer. -- hsivonen
    if (nsContentUtils::CheckForBOM(buf, count, maybePrefer)) {
      // The decoder will swallow the BOM. The UTF-16 will re-sniff for
      // endianness. The value of preferred is now either "UTF-8" or "UTF-16".
      preferred.Assign(maybePrefer);
      source = kCharsetFromByteOrderMark;
    } else if (source < kCharsetFromChannel) {
      nsAutoCString declCharset;

      if (ExtractCharsetFromXmlDeclaration(buf, count, declCharset)) {
        if (EncodingUtils::FindEncodingForLabel(declCharset, maybePrefer)) {
          preferred.Assign(maybePrefer);
          source = kCharsetFromMetaTag;
        }
      }
    }

    pws->mParser->SetDocumentCharset(preferred, source);
    pws->mParser->SetSinkCharset(preferred);

  }

  result = pws->mScanner->Append(fromRawSegment, theNumRead);
  if (NS_SUCCEEDED(result)) {
    *writeCount = count;
  }

  return result;
}

nsresult
nsParser::OnDataAvailable(nsIRequest *request, nsISupports* aContext,
                          nsIInputStream *pIStream, uint64_t sourceOffset,
                          uint32_t aLength)
{
  NS_PRECONDITION((eOnStart == mParserContext->mStreamListenerState ||
                   eOnDataAvail == mParserContext->mStreamListenerState),
            "Error: OnStartRequest() must be called before OnDataAvailable()");
  NS_PRECONDITION(NS_InputStreamIsBuffered(pIStream),
                  "Must have a buffered input stream");

  nsresult rv = NS_OK;

  if (mIsAboutBlank) {
    MOZ_ASSERT(false, "Must not get OnDataAvailable for about:blank");
    // ... but if an extension tries to feed us data for about:blank in a
    // release build, silently ignore the data.
    uint32_t totalRead;
    rv = pIStream->ReadSegments(NoOpParserWriteFunc,
                                nullptr,
                                aLength,
                                &totalRead);
    return rv;
  }

  CParserContext *theContext = mParserContext;

  while (theContext && theContext->mRequest != request) {
    theContext = theContext->mPrevContext;
  }

  if (theContext) {
    theContext->mStreamListenerState = eOnDataAvail;

    if (eInvalidDetect == theContext->mAutoDetectStatus) {
      if (theContext->mScanner) {
        nsScannerIterator iter;
        theContext->mScanner->EndReading(iter);
        theContext->mScanner->SetPosition(iter, true);
      }
    }

    uint32_t totalRead;
    ParserWriteStruct pws;
    pws.mNeedCharsetCheck = true;
    pws.mParser = this;
    pws.mScanner = theContext->mScanner;
    pws.mRequest = request;

    rv = pIStream->ReadSegments(ParserWriteFunc, &pws, aLength, &totalRead);
    if (NS_FAILED(rv)) {
      return rv;
    }

    if (IsOkToProcessNetworkData()) {
      nsCOMPtr<nsIParser> kungFuDeathGrip(this);
      nsCOMPtr<nsIContentSink> sinkDeathGrip(mSink);
      mProcessingNetworkData = true;
      if (sinkDeathGrip) {
        sinkDeathGrip->WillParse();
      }
      rv = ResumeParse();
      mProcessingNetworkData = false;
    }
  } else {
    rv = NS_ERROR_UNEXPECTED;
  }

  return rv;
}

/**
 *  This is called by the networking library once the last block of data
 *  has been collected from the net.
 */
nsresult
nsParser::OnStopRequest(nsIRequest *request, nsISupports* aContext,
                        nsresult status)
{
  nsresult rv = NS_OK;

  CParserContext *pc = mParserContext;
  while (pc) {
    if (pc->mRequest == request) {
      pc->mStreamListenerState = eOnStop;
      pc->mScanner->SetIncremental(false);
      break;
    }

    pc = pc->mPrevContext;
  }

  mStreamStatus = status;

  if (IsOkToProcessNetworkData() && NS_SUCCEEDED(rv)) {
    mProcessingNetworkData = true;
    if (mSink) {
      mSink->WillParse();
    }
    rv = ResumeParse(true, true);
    mProcessingNetworkData = false;
  }

  // If the parser isn't enabled, we don't finish parsing till
  // it is reenabled.


  // XXX Should we wait to notify our observers as well if the
  // parser isn't yet enabled?
  if (mObserver) {
    mObserver->OnStopRequest(request, aContext, status);
  }

  return rv;
}


/*******************************************************************
  Here come the tokenization methods...
 *******************************************************************/


/**
 *  Part of the code sandwich, this gets called right before
 *  the tokenization process begins. The main reason for
 *  this call is to allow the delegate to do initialization.
 */
bool
nsParser::WillTokenize(bool aIsFinalChunk)
{
  if (!mParserContext) {
    return true;
  }

  nsITokenizer* theTokenizer;
  nsresult result = mParserContext->GetTokenizer(mDTD, mSink, theTokenizer);
  NS_ENSURE_SUCCESS(result, false);
  return NS_SUCCEEDED(theTokenizer->WillTokenize(aIsFinalChunk));
}


/**
 * This is the primary control routine to consume tokens.
 * It iteratively consumes tokens until an error occurs or
 * you run out of data.
 */
nsresult nsParser::Tokenize(bool aIsFinalChunk)
{
  nsITokenizer* theTokenizer;

  nsresult result = NS_ERROR_NOT_AVAILABLE;
  if (mParserContext) {
    result = mParserContext->GetTokenizer(mDTD, mSink, theTokenizer);
  }

  if (NS_SUCCEEDED(result)) {
    bool flushTokens = false;

    bool killSink = false;

    WillTokenize(aIsFinalChunk);
    while (NS_SUCCEEDED(result)) {
      mParserContext->mScanner->Mark();
      result = theTokenizer->ConsumeToken(*mParserContext->mScanner,
                                          flushTokens);
      if (NS_FAILED(result)) {
        mParserContext->mScanner->RewindToMark();
        if (kEOF == result){
          break;
        }
        if (NS_ERROR_HTMLPARSER_STOPPARSING == result) {
          killSink = true;
          result = Terminate();
          break;
        }
      } else if (flushTokens && (mFlags & NS_PARSER_FLAG_OBSERVERS_ENABLED)) {
        // I added the extra test of NS_PARSER_FLAG_OBSERVERS_ENABLED to fix Bug# 23931.
        // Flush tokens on seeing </SCRIPT> -- Ref: Bug# 22485 --
        // Also remember to update the marked position.
        mFlags |= NS_PARSER_FLAG_FLUSH_TOKENS;
        mParserContext->mScanner->Mark();
        break;
      }
    }

    if (killSink) {
      mSink = nullptr;
    }
  } else {
    result = mInternalState = NS_ERROR_HTMLPARSER_BADTOKENIZER;
  }

  return result;
}

/**
 * Get the channel associated with this parser
 *
 * @param aChannel out param that will contain the result
 * @return NS_OK if successful
 */
NS_IMETHODIMP
nsParser::GetChannel(nsIChannel** aChannel)
{
  nsresult result = NS_ERROR_NOT_AVAILABLE;
  if (mParserContext && mParserContext->mRequest) {
    result = CallQueryInterface(mParserContext->mRequest, aChannel);
  }
  return result;
}

/**
 * Get the DTD associated with this parser
 */
NS_IMETHODIMP
nsParser::GetDTD(nsIDTD** aDTD)
{
  if (mParserContext) {
    NS_IF_ADDREF(*aDTD = mDTD);
  }

  return NS_OK;
}

/**
 * Get this as nsIStreamListener
 */
nsIStreamListener*
nsParser::GetStreamListener()
{
  return this;
}