aboutsummaryrefslogtreecommitdiffhomepage
path: root/src/ext_depends/D-YAML/source/dyaml/scanner.d
blob: 77c3e38bef05f61bf132eb5fd441973c7de30d4a (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
1601
1602
1603
1604
1605
1606
1607
1608
1609
1610
1611
1612
1613
1614
1615
1616
1617
1618
1619
1620
1621
1622
1623
1624
1625
1626
1627
1628
1629
1630
1631
1632
1633
1634
1635
1636
1637
1638
1639
1640
1641
1642
1643
1644
1645
1646
1647
1648
1649
1650
1651
1652
1653
1654
1655
1656
1657
1658
1659
1660
1661
1662
1663
1664
1665
1666
1667
1668
1669
1670
1671
1672
1673
1674
1675
1676
1677
1678
1679
1680
1681
1682
1683
1684
1685
1686
1687
1688
1689
1690
1691
1692
1693
1694
1695
1696
1697
1698
1699
1700
1701
1702
1703
1704
1705
1706
1707
1708
1709
1710
1711
1712
1713
1714
1715
1716
1717
1718
1719
1720
1721
1722
1723
1724
1725
1726
1727
1728
1729
1730
1731
1732
1733
1734
1735
1736
1737
1738
1739
1740
1741
1742
1743
1744
1745
1746
1747
1748
1749
1750
1751
1752
1753
1754
1755
1756
1757
1758
1759
1760
1761
1762
1763
1764
1765
1766
1767
1768
1769
1770
1771
1772
1773
1774
1775
1776
1777
1778
1779
1780
1781
1782
1783
1784
1785
1786
1787
1788
1789
1790
1791
1792
1793
1794
1795
1796
1797
1798
1799
1800
1801
1802
1803
1804
1805
1806
1807
1808
1809
1810
1811
1812
1813
1814
1815
1816
1817
1818
1819
1820
1821
1822
1823
1824
1825
1826
//          Copyright Ferdinand Majerech 2011-2014.
// Distributed under the Boost Software License, Version 1.0.
//    (See accompanying file LICENSE_1_0.txt or copy at
//          http://www.boost.org/LICENSE_1_0.txt)

/// YAML scanner.
/// Code based on PyYAML: http://www.pyyaml.org
module dyaml.scanner;


import core.stdc.string;

import std.algorithm;
import std.array;
import std.conv;
import std.ascii : isAlphaNum, isDigit, isHexDigit;
import std.exception;
import std.string;
import std.typecons;
import std.traits : Unqual;
import std.utf;

import dyaml.escapes;
import dyaml.exception;
import dyaml.queue;
import dyaml.reader;
import dyaml.style;
import dyaml.token;

package:
/// Scanner produces tokens of the following types:
/// STREAM-START
/// STREAM-END
/// DIRECTIVE(name, value)
/// DOCUMENT-START
/// DOCUMENT-END
/// BLOCK-SEQUENCE-START
/// BLOCK-MAPPING-START
/// BLOCK-END
/// FLOW-SEQUENCE-START
/// FLOW-MAPPING-START
/// FLOW-SEQUENCE-END
/// FLOW-MAPPING-END
/// BLOCK-ENTRY
/// FLOW-ENTRY
/// KEY
/// VALUE
/// ALIAS(value)
/// ANCHOR(value)
/// TAG(value)
/// SCALAR(value, plain, style)

alias isBreak = among!('\0', '\n', '\r', '\u0085', '\u2028', '\u2029');

alias isBreakOrSpace = among!(' ', '\0', '\n', '\r', '\u0085', '\u2028', '\u2029');

alias isWhiteSpace = among!(' ', '\t', '\0', '\n', '\r', '\u0085', '\u2028', '\u2029');

alias isNonLinebreakWhitespace = among!(' ', '\t');

alias isNonScalarStartCharacter = among!('-', '?', ':', ',', '[', ']', '{', '}',
    '#', '&', '*', '!', '|', '>', '\'', '"', '%', '@', '`', ' ', '\t', '\0', '\n',
    '\r', '\u0085', '\u2028', '\u2029');

alias isURIChar = among!('-', ';', '/', '?', ':', '@', '&', '=', '+', '$', ',',
    '_', '.', '!', '~', '*', '\'', '(', ')', '[', ']', '%');

alias isNSChar = among!(' ', '\n', '\r', '\u0085', '\u2028', '\u2029');

alias isBChar = among!('\n', '\r', '\u0085', '\u2028', '\u2029');

alias isFlowScalarBreakSpace = among!(' ', '\t', '\0', '\n', '\r', '\u0085', '\u2028', '\u2029', '\'', '"', '\\');

alias isNSAnchorName = c => !c.isWhiteSpace && !c.among!('[', ']', '{', '}', ',', '\uFEFF');

/// Marked exception thrown at scanner errors.
///
/// See_Also: MarkedYAMLException
class ScannerException : MarkedYAMLException
{
    mixin MarkedExceptionCtors;
}

/// Generates tokens from data provided by a Reader.
struct Scanner
{
    private:
        /// A simple key is a key that is not denoted by the '?' indicator.
        /// For example:
        ///   ---
        ///   block simple key: value
        ///   ? not a simple key:
        ///   : { flow simple key: value }
        /// We emit the KEY token before all keys, so when we find a potential simple
        /// key, we try to locate the corresponding ':' indicator. Simple keys should be
        /// limited to a single line and 1024 characters.
        ///
        /// 16 bytes on 64-bit.
        static struct SimpleKey
        {
            /// Character index in reader where the key starts.
            uint charIndex = uint.max;
            /// Index of the key token from start (first token scanned being 0).
            uint tokenIndex;
            /// Line the key starts at.
            uint line;
            /// Column the key starts at.
            ushort column;
            /// Is this required to be a simple key?
            bool required;
            /// Is this struct "null" (invalid)?.
            bool isNull;
        }

        /// Block chomping types.
        enum Chomping
        {
            /// Strip all trailing line breaks. '-' indicator.
            strip,
            /// Line break of the last line is preserved, others discarded. Default.
            clip,
            /// All trailing line breaks are preserved. '+' indicator.
            keep
        }

        /// Reader used to read from a file/stream.
        Reader reader_;
        /// Are we done scanning?
        bool done_;

        /// Level of nesting in flow context. If 0, we're in block context.
        uint flowLevel_;
        /// Current indentation level.
        int indent_ = -1;
        /// Past indentation levels. Used as a stack.
        Appender!(int[]) indents_;

        /// Processed tokens not yet emitted. Used as a queue.
        Queue!Token tokens_;

        /// Number of tokens emitted through the getToken method.
        uint tokensTaken_;

        /// Can a simple key start at the current position? A simple key may start:
        /// - at the beginning of the line, not counting indentation spaces
        ///       (in block context),
        /// - after '{', '[', ',' (in the flow context),
        /// - after '?', ':', '-' (in the block context).
        /// In the block context, this flag also signifies if a block collection
        /// may start at the current position.
        bool allowSimpleKey_ = true;

        /// Possible simple keys indexed by flow levels.
        SimpleKey[] possibleSimpleKeys_;

    public:
        /// Construct a Scanner using specified Reader.
        this(Reader reader) @safe nothrow
        {
            // Return the next token, but do not delete it from the queue
            reader_   = reader;
            fetchStreamStart();
        }

        /// Advance to the next token
        void popFront() @safe
        {
            ++tokensTaken_;
            tokens_.pop();
        }

        /// Return the current token
        const(Token) front() @safe
        {
            enforce(!empty, "No token left to peek");
            return tokens_.peek();
        }

        /// Return whether there are any more tokens left.
        bool empty() @safe
        {
            while (needMoreTokens())
            {
                fetchToken();
            }
            return tokens_.empty;
        }

        /// Set file name.
        void name(string name) @safe pure nothrow @nogc
        {
            reader_.name = name;
        }

    private:
        /// Most scanning error messages have the same format; so build them with this
        /// function.
        string expected(T)(string expected, T found)
        {
            return text("expected ", expected, ", but found ", found);
        }

        /// Determine whether or not we need to fetch more tokens before peeking/getting a token.
        bool needMoreTokens() @safe pure
        {
            if(done_)         { return false; }
            if(tokens_.empty) { return true; }

            /// The current token may be a potential simple key, so we need to look further.
            stalePossibleSimpleKeys();
            return nextPossibleSimpleKey() == tokensTaken_;
        }

        /// Fetch at token, adding it to tokens_.
        void fetchToken() @safe
        {
            // Eat whitespaces and comments until we reach the next token.
            scanToNextToken();

            // Remove obsolete possible simple keys.
            stalePossibleSimpleKeys();

            // Compare current indentation and column. It may add some tokens
            // and decrease the current indentation level.
            unwindIndent(reader_.column);

            // Get the next character.
            const dchar c = reader_.peekByte();

            // Fetch the token.
            if(c == '\0')            { return fetchStreamEnd();     }
            if(checkDirective())     { return fetchDirective();     }
            if(checkDocumentStart()) { return fetchDocumentStart(); }
            if(checkDocumentEnd())   { return fetchDocumentEnd();   }
            // Order of the following checks is NOT significant.
            switch(c)
            {
                case '[':  return fetchFlowSequenceStart();
                case '{':  return fetchFlowMappingStart();
                case ']':  return fetchFlowSequenceEnd();
                case '}':  return fetchFlowMappingEnd();
                case ',':  return fetchFlowEntry();
                case '!':  return fetchTag();
                case '\'': return fetchSingle();
                case '\"': return fetchDouble();
                case '*':  return fetchAlias();
                case '&':  return fetchAnchor();
                case '?':  if(checkKey())        { return fetchKey();        } goto default;
                case ':':  if(checkValue())      { return fetchValue();      } goto default;
                case '-':  if(checkBlockEntry()) { return fetchBlockEntry(); } goto default;
                case '|':  if(flowLevel_ == 0)   { return fetchLiteral();    } break;
                case '>':  if(flowLevel_ == 0)   { return fetchFolded();     } break;
                default:   if(checkPlain())      { return fetchPlain();      }
            }

            throw new ScannerException("While scanning for the next token, found character " ~
                                       "\'%s\', index %s that cannot start any token"
                                       .format(c, to!int(c)), reader_.mark);
        }


        /// Return the token number of the nearest possible simple key.
        uint nextPossibleSimpleKey() @safe pure nothrow @nogc
        {
            uint minTokenNumber = uint.max;
            foreach(k, ref simpleKey; possibleSimpleKeys_)
            {
                if(simpleKey.isNull) { continue; }
                minTokenNumber = min(minTokenNumber, simpleKey.tokenIndex);
            }
            return minTokenNumber;
        }

        /// Remove entries that are no longer possible simple keys.
        ///
        /// According to the YAML specification, simple keys
        /// - should be limited to a single line,
        /// - should be no longer than 1024 characters.
        /// Disabling this will allow simple keys of any length and
        /// height (may cause problems if indentation is broken though).
        void stalePossibleSimpleKeys() @safe pure
        {
            foreach(level, ref key; possibleSimpleKeys_)
            {
                if(key.isNull) { continue; }
                if(key.line != reader_.line || reader_.charIndex - key.charIndex > 1024)
                {
                    enforce(!key.required,
                            new ScannerException("While scanning a simple key",
                                                 Mark(reader_.name, key.line, key.column),
                                                 "could not find expected ':'", reader_.mark));
                    key.isNull = true;
                }
            }
        }

        /// Check if the next token starts a possible simple key and if so, save its position.
        ///
        /// This function is called for ALIAS, ANCHOR, TAG, SCALAR(flow), '[', and '{'.
        void savePossibleSimpleKey() @safe pure
        {
            // Check if a simple key is required at the current position.
            const required = (flowLevel_ == 0 && indent_ == reader_.column);
            assert(allowSimpleKey_ || !required, "A simple key is required only if it is " ~
                   "the first token in the current line. Therefore it is always allowed.");

            if(!allowSimpleKey_) { return; }

            // The next token might be a simple key, so save its number and position.
            removePossibleSimpleKey();
            const tokenCount = tokensTaken_ + cast(uint)tokens_.length;

            const line   = reader_.line;
            const column = reader_.column;
            const key    = SimpleKey(cast(uint)reader_.charIndex, tokenCount, line,
                                     cast(ushort)min(column, ushort.max), required);

            if(possibleSimpleKeys_.length <= flowLevel_)
            {
                const oldLength = possibleSimpleKeys_.length;
                possibleSimpleKeys_.length = flowLevel_ + 1;
                //No need to initialize the last element, it's already done in the next line.
                possibleSimpleKeys_[oldLength .. flowLevel_] = SimpleKey.init;
            }
            possibleSimpleKeys_[flowLevel_] = key;
        }

        /// Remove the saved possible key position at the current flow level.
        void removePossibleSimpleKey() @safe pure
        {
            if(possibleSimpleKeys_.length <= flowLevel_) { return; }

            if(!possibleSimpleKeys_[flowLevel_].isNull)
            {
                const key = possibleSimpleKeys_[flowLevel_];
                enforce(!key.required,
                        new ScannerException("While scanning a simple key",
                                             Mark(reader_.name, key.line, key.column),
                                             "could not find expected ':'", reader_.mark));
                possibleSimpleKeys_[flowLevel_].isNull = true;
            }
        }

        /// Decrease indentation, removing entries in indents_.
        ///
        /// Params:  column = Current column in the file/stream.
        void unwindIndent(const int column) @safe
        {
            if(flowLevel_ > 0)
            {
                // In flow context, tokens should respect indentation.
                // The condition should be `indent >= column` according to the spec.
                // But this condition will prohibit intuitively correct
                // constructions such as
                // key : {
                // }

                // In the flow context, indentation is ignored. We make the scanner less
                // restrictive than what the specification requires.
                // if(pedantic_ && flowLevel_ > 0 && indent_ > column)
                // {
                //     throw new ScannerException("Invalid intendation or unclosed '[' or '{'",
                //                                reader_.mark)
                // }
                return;
            }

            // In block context, we may need to issue the BLOCK-END tokens.
            while(indent_ > column)
            {
                indent_ = indents_.data.back;
                assert(indents_.data.length);
                indents_.shrinkTo(indents_.data.length - 1);
                tokens_.push(blockEndToken(reader_.mark, reader_.mark));
            }
        }

        /// Increase indentation if needed.
        ///
        /// Params:  column = Current column in the file/stream.
        ///
        /// Returns: true if the indentation was increased, false otherwise.
        bool addIndent(int column) @safe
        {
            if(indent_ >= column){return false;}
            indents_ ~= indent_;
            indent_ = column;
            return true;
        }


        /// Add STREAM-START token.
        void fetchStreamStart() @safe nothrow
        {
            tokens_.push(streamStartToken(reader_.mark, reader_.mark, reader_.encoding));
        }

        ///Add STREAM-END token.
        void fetchStreamEnd() @safe
        {
            //Set intendation to -1 .
            unwindIndent(-1);
            removePossibleSimpleKey();
            allowSimpleKey_ = false;
            possibleSimpleKeys_.destroy;

            tokens_.push(streamEndToken(reader_.mark, reader_.mark));
            done_ = true;
        }

        /// Add DIRECTIVE token.
        void fetchDirective() @safe
        {
            // Set intendation to -1 .
            unwindIndent(-1);
            // Reset simple keys.
            removePossibleSimpleKey();
            allowSimpleKey_ = false;

            auto directive = scanDirective();
            tokens_.push(directive);
        }

        /// Add DOCUMENT-START or DOCUMENT-END token.
        void fetchDocumentIndicator(TokenID id)()
            if(id == TokenID.documentStart || id == TokenID.documentEnd)
        {
            // Set indentation to -1 .
            unwindIndent(-1);
            // Reset simple keys. Note that there can't be a block collection after '---'.
            removePossibleSimpleKey();
            allowSimpleKey_ = false;

            Mark startMark = reader_.mark;
            reader_.forward(3);
            tokens_.push(simpleToken!id(startMark, reader_.mark));
        }

        /// Aliases to add DOCUMENT-START or DOCUMENT-END token.
        alias fetchDocumentStart = fetchDocumentIndicator!(TokenID.documentStart);
        alias fetchDocumentEnd = fetchDocumentIndicator!(TokenID.documentEnd);

        /// Add FLOW-SEQUENCE-START or FLOW-MAPPING-START token.
        void fetchFlowCollectionStart(TokenID id)() @safe
        {
            // '[' and '{' may start a simple key.
            savePossibleSimpleKey();
            // Simple keys are allowed after '[' and '{'.
            allowSimpleKey_ = true;
            ++flowLevel_;

            Mark startMark = reader_.mark;
            reader_.forward();
            tokens_.push(simpleToken!id(startMark, reader_.mark));
        }

        /// Aliases to add FLOW-SEQUENCE-START or FLOW-MAPPING-START token.
        alias fetchFlowSequenceStart = fetchFlowCollectionStart!(TokenID.flowSequenceStart);
        alias fetchFlowMappingStart = fetchFlowCollectionStart!(TokenID.flowMappingStart);

        /// Add FLOW-SEQUENCE-START or FLOW-MAPPING-START token.
        void fetchFlowCollectionEnd(TokenID id)()
        {
            // Reset possible simple key on the current level.
            removePossibleSimpleKey();
            // No simple keys after ']' and '}'.
            allowSimpleKey_ = false;
            --flowLevel_;

            Mark startMark = reader_.mark;
            reader_.forward();
            tokens_.push(simpleToken!id(startMark, reader_.mark));
        }

        /// Aliases to add FLOW-SEQUENCE-START or FLOW-MAPPING-START token/
        alias fetchFlowSequenceEnd = fetchFlowCollectionEnd!(TokenID.flowSequenceEnd);
        alias fetchFlowMappingEnd = fetchFlowCollectionEnd!(TokenID.flowMappingEnd);

        /// Add FLOW-ENTRY token;
        void fetchFlowEntry() @safe
        {
            // Reset possible simple key on the current level.
            removePossibleSimpleKey();
            // Simple keys are allowed after ','.
            allowSimpleKey_ = true;

            Mark startMark = reader_.mark;
            reader_.forward();
            tokens_.push(flowEntryToken(startMark, reader_.mark));
        }

        /// Additional checks used in block context in fetchBlockEntry and fetchKey.
        ///
        /// Params:  type = String representing the token type we might need to add.
        ///          id   = Token type we might need to add.
        void blockChecks(string type, TokenID id)()
        {
            enum context = type ~ " keys are not allowed here";
            // Are we allowed to start a key (not neccesarily a simple one)?
            enforce(allowSimpleKey_, new ScannerException(context, reader_.mark));

            if(addIndent(reader_.column))
            {
                tokens_.push(simpleToken!id(reader_.mark, reader_.mark));
            }
        }

        /// Add BLOCK-ENTRY token. Might add BLOCK-SEQUENCE-START in the process.
        void fetchBlockEntry() @safe
        {
            if(flowLevel_ == 0) { blockChecks!("Sequence", TokenID.blockSequenceStart)(); }

            // It's an error for the block entry to occur in the flow context,
            // but we let the parser detect this.

            // Reset possible simple key on the current level.
            removePossibleSimpleKey();
            // Simple keys are allowed after '-'.
            allowSimpleKey_ = true;

            Mark startMark = reader_.mark;
            reader_.forward();
            tokens_.push(blockEntryToken(startMark, reader_.mark));
        }

        /// Add KEY token. Might add BLOCK-MAPPING-START in the process.
        void fetchKey() @safe
        {
            if(flowLevel_ == 0) { blockChecks!("Mapping", TokenID.blockMappingStart)(); }

            // Reset possible simple key on the current level.
            removePossibleSimpleKey();
            // Simple keys are allowed after '?' in the block context.
            allowSimpleKey_ = (flowLevel_ == 0);

            Mark startMark = reader_.mark;
            reader_.forward();
            tokens_.push(keyToken(startMark, reader_.mark));
        }

        /// Add VALUE token. Might add KEY and/or BLOCK-MAPPING-START in the process.
        void fetchValue() @safe
        {
            //Do we determine a simple key?
            if(possibleSimpleKeys_.length > flowLevel_ &&
               !possibleSimpleKeys_[flowLevel_].isNull)
            {
                const key = possibleSimpleKeys_[flowLevel_];
                possibleSimpleKeys_[flowLevel_].isNull = true;
                Mark keyMark = Mark(reader_.name, key.line, key.column);
                const idx = key.tokenIndex - tokensTaken_;

                assert(idx >= 0);

                // Add KEY.
                // Manually inserting since tokens are immutable (need linked list).
                tokens_.insert(keyToken(keyMark, keyMark), idx);

                // If this key starts a new block mapping, we need to add BLOCK-MAPPING-START.
                if(flowLevel_ == 0 && addIndent(key.column))
                {
                    tokens_.insert(blockMappingStartToken(keyMark, keyMark), idx);
                }

                // There cannot be two simple keys in a row.
                allowSimpleKey_ = false;
            }
            // Part of a complex key
            else
            {
                // We can start a complex value if and only if we can start a simple key.
                enforce(flowLevel_ > 0 || allowSimpleKey_,
                        new ScannerException("Mapping values are not allowed here", reader_.mark));

                // If this value starts a new block mapping, we need to add
                // BLOCK-MAPPING-START. It'll be detected as an error later by the parser.
                if(flowLevel_ == 0 && addIndent(reader_.column))
                {
                    tokens_.push(blockMappingStartToken(reader_.mark, reader_.mark));
                }

                // Reset possible simple key on the current level.
                removePossibleSimpleKey();
                // Simple keys are allowed after ':' in the block context.
                allowSimpleKey_ = (flowLevel_ == 0);
            }

            // Add VALUE.
            Mark startMark = reader_.mark;
            reader_.forward();
            tokens_.push(valueToken(startMark, reader_.mark));
        }

        /// Add ALIAS or ANCHOR token.
        void fetchAnchor_(TokenID id)() @safe
            if(id == TokenID.alias_ || id == TokenID.anchor)
        {
            // ALIAS/ANCHOR could be a simple key.
            savePossibleSimpleKey();
            // No simple keys after ALIAS/ANCHOR.
            allowSimpleKey_ = false;

            auto anchor = scanAnchor(id);
            tokens_.push(anchor);
        }

        /// Aliases to add ALIAS or ANCHOR token.
        alias fetchAlias = fetchAnchor_!(TokenID.alias_);
        alias fetchAnchor = fetchAnchor_!(TokenID.anchor);

        /// Add TAG token.
        void fetchTag() @safe
        {
            //TAG could start a simple key.
            savePossibleSimpleKey();
            //No simple keys after TAG.
            allowSimpleKey_ = false;

            tokens_.push(scanTag());
        }

        /// Add block SCALAR token.
        void fetchBlockScalar(ScalarStyle style)() @safe
            if(style == ScalarStyle.literal || style == ScalarStyle.folded)
        {
            // Reset possible simple key on the current level.
            removePossibleSimpleKey();
            // A simple key may follow a block scalar.
            allowSimpleKey_ = true;

            auto blockScalar = scanBlockScalar(style);
            tokens_.push(blockScalar);
        }

        /// Aliases to add literal or folded block scalar.
        alias fetchLiteral = fetchBlockScalar!(ScalarStyle.literal);
        alias fetchFolded = fetchBlockScalar!(ScalarStyle.folded);

        /// Add quoted flow SCALAR token.
        void fetchFlowScalar(ScalarStyle quotes)()
        {
            // A flow scalar could be a simple key.
            savePossibleSimpleKey();
            // No simple keys after flow scalars.
            allowSimpleKey_ = false;

            // Scan and add SCALAR.
            auto scalar = scanFlowScalar(quotes);
            tokens_.push(scalar);
        }

        /// Aliases to add single or double quoted block scalar.
        alias fetchSingle = fetchFlowScalar!(ScalarStyle.singleQuoted);
        alias fetchDouble = fetchFlowScalar!(ScalarStyle.doubleQuoted);

        /// Add plain SCALAR token.
        void fetchPlain() @safe
        {
            // A plain scalar could be a simple key
            savePossibleSimpleKey();
            // No simple keys after plain scalars. But note that scanPlain() will
            // change this flag if the scan is finished at the beginning of the line.
            allowSimpleKey_ = false;
            auto plain = scanPlain();

            // Scan and add SCALAR. May change allowSimpleKey_
            tokens_.push(plain);
        }

    pure:

        ///Check if the next token is DIRECTIVE:        ^ '%' ...
        bool checkDirective() @safe
        {
            return reader_.peekByte() == '%' && reader_.column == 0;
        }

        /// Check if the next token is DOCUMENT-START:   ^ '---' (' '|'\n')
        bool checkDocumentStart() @safe
        {
            // Check one char first, then all 3, to prevent reading outside the buffer.
            return reader_.column     == 0     &&
                   reader_.peekByte() == '-'   &&
                   reader_.prefix(3)  == "---" &&
                   reader_.peek(3).isWhiteSpace;
        }

        /// Check if the next token is DOCUMENT-END:     ^ '...' (' '|'\n')
        bool checkDocumentEnd() @safe
        {
            // Check one char first, then all 3, to prevent reading outside the buffer.
            return reader_.column     == 0     &&
                   reader_.peekByte() == '.'   &&
                   reader_.prefix(3)  == "..." &&
                   reader_.peek(3).isWhiteSpace;
        }

        /// Check if the next token is BLOCK-ENTRY:      '-' (' '|'\n')
        bool checkBlockEntry() @safe
        {
            return !!reader_.peek(1).isWhiteSpace;
        }

        /// Check if the next token is KEY(flow context):    '?'
        ///
        /// or KEY(block context):   '?' (' '|'\n')
        bool checkKey() @safe
        {
            return (flowLevel_ > 0 || reader_.peek(1).isWhiteSpace);
        }

        /// Check if the next token is VALUE(flow context):  ':'
        ///
        /// or VALUE(block context): ':' (' '|'\n')
        bool checkValue() @safe
        {
            return flowLevel_ > 0 || reader_.peek(1).isWhiteSpace;
        }

        /// Check if the next token is a plain scalar.
        ///
        /// A plain scalar may start with any non-space character except:
        ///   '-', '?', ':', ',', '[', ']', '{', '}',
        ///   '#', '&', '*', '!', '|', '>', '\'', '\"',
        ///   '%', '@', '`'.
        ///
        /// It may also start with
        ///   '-', '?', ':'
        /// if it is followed by a non-space character.
        ///
        /// Note that we limit the last rule to the block context (except the
        /// '-' character) because we want the flow context to be space
        /// independent.
        bool checkPlain() @safe
        {
            const c = reader_.peek();
            if(!c.isNonScalarStartCharacter)
            {
                return true;
            }
            return !reader_.peek(1).isWhiteSpace &&
                   (c == '-' || (flowLevel_ == 0 && (c == '?' || c == ':')));
        }

        /// Move to the next non-space character.
        void findNextNonSpace() @safe
        {
            while(reader_.peekByte() == ' ') { reader_.forward(); }
        }

        /// Scan a string of alphanumeric or "-_" characters.
        ///
        /// Assumes that the caller is building a slice in Reader, and puts the scanned
        /// characters into that slice.
        void scanAlphaNumericToSlice(string name)(const Mark startMark)
        {
            size_t length;
            dchar c = reader_.peek();
            while(c.isAlphaNum || c.among!('-', '_')) { c = reader_.peek(++length); }

            enforce(length > 0, new ScannerException("While scanning " ~ name,
                startMark, expected("alphanumeric, '-' or '_'", c), reader_.mark));

            reader_.sliceBuilder.write(reader_.get(length));
        }

        /// Scan a string.
        ///
        /// Assumes that the caller is building a slice in Reader, and puts the scanned
        /// characters into that slice.
        void scanAnchorAliasToSlice(const Mark startMark) @safe
        {
            size_t length;
            dchar c = reader_.peek();
            while (c.isNSAnchorName)
            {
                c = reader_.peek(++length);
            }

            enforce(length > 0, new ScannerException("While scanning an anchor or alias",
                startMark, expected("a printable character besides '[', ']', '{', '}' and ','", c), reader_.mark));

            reader_.sliceBuilder.write(reader_.get(length));
        }

        /// Scan and throw away all characters until next line break.
        void scanToNextBreak() @safe
        {
            while(!reader_.peek().isBreak) { reader_.forward(); }
        }

        /// Scan all characters until next line break.
        ///
        /// Assumes that the caller is building a slice in Reader, and puts the scanned
        /// characters into that slice.
        void scanToNextBreakToSlice() @safe
        {
            uint length;
            while(!reader_.peek(length).isBreak)
            {
                ++length;
            }
            reader_.sliceBuilder.write(reader_.get(length));
        }


        /// Move to next token in the file/stream.
        ///
        /// We ignore spaces, line breaks and comments.
        /// If we find a line break in the block context, we set
        /// allowSimpleKey` on.
        ///
        /// We do not yet support BOM inside the stream as the
        /// specification requires. Any such mark will be considered as a part
        /// of the document.
        void scanToNextToken() @safe
        {
            // TODO(PyYAML): We need to make tab handling rules more sane. A good rule is:
            //   Tabs cannot precede tokens
            //   BLOCK-SEQUENCE-START, BLOCK-MAPPING-START, BLOCK-END,
            //   KEY(block), VALUE(block), BLOCK-ENTRY
            // So the checking code is
            //   if <TAB>:
            //       allowSimpleKey_ = false
            // We also need to add the check for `allowSimpleKey_ == true` to
            // `unwindIndent` before issuing BLOCK-END.
            // Scanners for block, flow, and plain scalars need to be modified.

            for(;;)
            {
                //All whitespace in flow context is ignored, even whitespace
                // not allowed in other contexts
                if (flowLevel_ > 0)
                {
                    while(reader_.peekByte().isNonLinebreakWhitespace) { reader_.forward(); }
                }
                else
                {
                    findNextNonSpace();
                }
                if(reader_.peekByte() == '#') { scanToNextBreak(); }
                if(scanLineBreak() != '\0')
                {
                    if(flowLevel_ == 0) { allowSimpleKey_ = true; }
                }
                else
                {
                    break;
                }
            }
        }

        /// Scan directive token.
        Token scanDirective() @safe
        {
            Mark startMark = reader_.mark;
            // Skip the '%'.
            reader_.forward();

            // Scan directive name
            reader_.sliceBuilder.begin();
            scanDirectiveNameToSlice(startMark);
            const name = reader_.sliceBuilder.finish();

            reader_.sliceBuilder.begin();

            // Index where tag handle ends and suffix starts in a tag directive value.
            uint tagHandleEnd = uint.max;
            if(name == "YAML")     { scanYAMLDirectiveValueToSlice(startMark); }
            else if(name == "TAG") { tagHandleEnd = scanTagDirectiveValueToSlice(startMark); }
            char[] value = reader_.sliceBuilder.finish();

            Mark endMark = reader_.mark;

            DirectiveType directive;
            if(name == "YAML")     { directive = DirectiveType.yaml; }
            else if(name == "TAG") { directive = DirectiveType.tag; }
            else
            {
                directive = DirectiveType.reserved;
                scanToNextBreak();
            }

            scanDirectiveIgnoredLine(startMark);

            return directiveToken(startMark, endMark, value, directive, tagHandleEnd);
        }

        /// Scan name of a directive token.
        ///
        /// Assumes that the caller is building a slice in Reader, and puts the scanned
        /// characters into that slice.
        void scanDirectiveNameToSlice(const Mark startMark) @safe
        {
            // Scan directive name.
            scanAlphaNumericToSlice!"a directive"(startMark);

            enforce(reader_.peek().among!(' ', '\0', '\n', '\r', '\u0085', '\u2028', '\u2029'),
                new ScannerException("While scanning a directive", startMark,
                    expected("alphanumeric, '-' or '_'", reader_.peek()), reader_.mark));
        }

        /// Scan value of a YAML directive token. Returns major, minor version separated by '.'.
        ///
        /// Assumes that the caller is building a slice in Reader, and puts the scanned
        /// characters into that slice.
        void scanYAMLDirectiveValueToSlice(const Mark startMark) @safe
        {
            findNextNonSpace();

            scanYAMLDirectiveNumberToSlice(startMark);

            enforce(reader_.peekByte() == '.',
                new ScannerException("While scanning a directive", startMark,
                    expected("digit or '.'", reader_.peek()), reader_.mark));
            // Skip the '.'.
            reader_.forward();

            reader_.sliceBuilder.write('.');
            scanYAMLDirectiveNumberToSlice(startMark);

            enforce(reader_.peek().among!(' ', '\0', '\n', '\r', '\u0085', '\u2028', '\u2029'),
                new ScannerException("While scanning a directive", startMark,
                    expected("digit or '.'", reader_.peek()), reader_.mark));
        }

        /// Scan a number from a YAML directive.
        ///
        /// Assumes that the caller is building a slice in Reader, and puts the scanned
        /// characters into that slice.
        void scanYAMLDirectiveNumberToSlice(const Mark startMark) @safe
        {
            enforce(isDigit(reader_.peek()),
                new ScannerException("While scanning a directive", startMark,
                    expected("digit", reader_.peek()), reader_.mark));

            // Already found the first digit in the enforce(), so set length to 1.
            uint length = 1;
            while(reader_.peek(length).isDigit) { ++length; }

            reader_.sliceBuilder.write(reader_.get(length));
        }

        /// Scan value of a tag directive.
        ///
        /// Assumes that the caller is building a slice in Reader, and puts the scanned
        /// characters into that slice.
        ///
        /// Returns: Length of tag handle (which is before tag prefix) in scanned data
        uint scanTagDirectiveValueToSlice(const Mark startMark) @safe
        {
            findNextNonSpace();
            const startLength = reader_.sliceBuilder.length;
            scanTagDirectiveHandleToSlice(startMark);
            const handleLength = cast(uint)(reader_.sliceBuilder.length  - startLength);
            findNextNonSpace();
            scanTagDirectivePrefixToSlice(startMark);

            return handleLength;
        }

        /// Scan handle of a tag directive.
        ///
        /// Assumes that the caller is building a slice in Reader, and puts the scanned
        /// characters into that slice.
        void scanTagDirectiveHandleToSlice(const Mark startMark) @safe
        {
            scanTagHandleToSlice!"directive"(startMark);
            enforce(reader_.peekByte() == ' ',
                new ScannerException("While scanning a directive handle", startMark,
                    expected("' '", reader_.peek()), reader_.mark));
        }

        /// Scan prefix of a tag directive.
        ///
        /// Assumes that the caller is building a slice in Reader, and puts the scanned
        /// characters into that slice.
        void scanTagDirectivePrefixToSlice(const Mark startMark) @safe
        {
            scanTagURIToSlice!"directive"(startMark);
            enforce(reader_.peek().among!(' ', '\0', '\n', '\r', '\u0085', '\u2028', '\u2029'),
                new ScannerException("While scanning a directive prefix", startMark,
                    expected("' '", reader_.peek()), reader_.mark));
        }

        /// Scan (and ignore) ignored line after a directive.
        void scanDirectiveIgnoredLine(const Mark startMark) @safe
        {
            findNextNonSpace();
            if(reader_.peekByte() == '#') { scanToNextBreak(); }
            enforce(reader_.peek().isBreak,
                new ScannerException("While scanning a directive", startMark,
                      expected("comment or a line break", reader_.peek()), reader_.mark));
            scanLineBreak();
        }


        /// Scan an alias or an anchor.
        ///
        /// The specification does not restrict characters for anchors and
        /// aliases. This may lead to problems, for instance, the document:
        ///   [ *alias, value ]
        /// can be interpteted in two ways, as
        ///   [ "value" ]
        /// and
        ///   [ *alias , "value" ]
        /// Therefore we restrict aliases to ASCII alphanumeric characters.
        Token scanAnchor(const TokenID id) @safe
        {
            const startMark = reader_.mark;
            reader_.forward(); // The */& character was only peeked, so we drop it now

            reader_.sliceBuilder.begin();
            scanAnchorAliasToSlice(startMark);
            // On error, value is discarded as we return immediately
            char[] value = reader_.sliceBuilder.finish();

            assert(!reader_.peek().isNSAnchorName, "Anchor/alias name not fully scanned");

            if(id == TokenID.alias_)
            {
                return aliasToken(startMark, reader_.mark, value);
            }
            if(id == TokenID.anchor)
            {
                return anchorToken(startMark, reader_.mark, value);
            }
            assert(false, "This code should never be reached");
        }

        /// Scan a tag token.
        Token scanTag() @safe
        {
            const startMark = reader_.mark;
            dchar c = reader_.peek(1);

            reader_.sliceBuilder.begin();
            scope(failure) { reader_.sliceBuilder.finish(); }
            // Index where tag handle ends and tag suffix starts in the tag value
            // (slice) we will produce.
            uint handleEnd;

            if(c == '<')
            {
                reader_.forward(2);

                handleEnd = 0;
                scanTagURIToSlice!"tag"(startMark);
                enforce(reader_.peekByte() == '>',
                    new ScannerException("While scanning a tag", startMark,
                        expected("'>'", reader_.peek()), reader_.mark));
                reader_.forward();
            }
            else if(c.isWhiteSpace)
            {
                reader_.forward();
                handleEnd = 0;
                reader_.sliceBuilder.write('!');
            }
            else
            {
                uint length = 1;
                bool useHandle;

                while(!c.isBreakOrSpace)
                {
                    if(c == '!')
                    {
                        useHandle = true;
                        break;
                    }
                    ++length;
                    c = reader_.peek(length);
                }

                if(useHandle)
                {
                    scanTagHandleToSlice!"tag"(startMark);
                    handleEnd = cast(uint)reader_.sliceBuilder.length;
                }
                else
                {
                    reader_.forward();
                    reader_.sliceBuilder.write('!');
                    handleEnd = cast(uint)reader_.sliceBuilder.length;
                }

                scanTagURIToSlice!"tag"(startMark);
            }

            enforce(reader_.peek().isBreakOrSpace,
                new ScannerException("While scanning a tag", startMark, expected("' '", reader_.peek()),
                    reader_.mark));

            char[] slice = reader_.sliceBuilder.finish();
            return tagToken(startMark, reader_.mark, slice, handleEnd);
        }

        /// Scan a block scalar token with specified style.
        Token scanBlockScalar(const ScalarStyle style) @safe
        {
            const startMark = reader_.mark;

            // Scan the header.
            reader_.forward();

            const indicators = scanBlockScalarIndicators(startMark);

            const chomping   = indicators[0];
            const increment  = indicators[1];
            scanBlockScalarIgnoredLine(startMark);

            // Determine the indentation level and go to the first non-empty line.
            Mark endMark;
            uint indent = max(1, indent_ + 1);

            reader_.sliceBuilder.begin();
            alias Transaction = SliceBuilder.Transaction;
            // Used to strip the last line breaks written to the slice at the end of the
            // scalar, which may be needed based on chomping.
            Transaction breaksTransaction = Transaction(&reader_.sliceBuilder);
            // Read the first indentation/line breaks before the scalar.
            size_t startLen = reader_.sliceBuilder.length;
            if(increment == int.min)
            {
                auto indentation = scanBlockScalarIndentationToSlice();
                endMark = indentation[1];
                indent  = max(indent, indentation[0]);
            }
            else
            {
                indent += increment - 1;
                endMark = scanBlockScalarBreaksToSlice(indent);
            }

            // int.max means there's no line break (int.max is outside UTF-32).
            dchar lineBreak = cast(dchar)int.max;

            // Scan the inner part of the block scalar.
            while(reader_.column == indent && reader_.peekByte() != '\0')
            {
                breaksTransaction.commit();
                const bool leadingNonSpace = !reader_.peekByte().among!(' ', '\t');
                // This is where the 'interesting' non-whitespace data gets read.
                scanToNextBreakToSlice();
                lineBreak = scanLineBreak();


                // This transaction serves to rollback data read in the
                // scanBlockScalarBreaksToSlice() call.
                breaksTransaction = Transaction(&reader_.sliceBuilder);
                startLen = reader_.sliceBuilder.length;
                // The line breaks should actually be written _after_ the if() block
                // below. We work around that by inserting
                endMark = scanBlockScalarBreaksToSlice(indent);

                // This will not run during the last iteration (see the if() vs the
                // while()), hence breaksTransaction rollback (which happens after this
                // loop) will never roll back data written in this if() block.
                if(reader_.column == indent && reader_.peekByte() != '\0')
                {
                    // Unfortunately, folding rules are ambiguous.

                    // This is the folding according to the specification:
                    if(style == ScalarStyle.folded && lineBreak == '\n' &&
                       leadingNonSpace && !reader_.peekByte().among!(' ', '\t'))
                    {
                        // No breaks were scanned; no need to insert the space in the
                        // middle of slice.
                        if(startLen == reader_.sliceBuilder.length)
                        {
                            reader_.sliceBuilder.write(' ');
                        }
                    }
                    else
                    {
                        // We need to insert in the middle of the slice in case any line
                        // breaks were scanned.
                        reader_.sliceBuilder.insert(lineBreak, startLen);
                    }

                    ////this is Clark Evans's interpretation (also in the spec
                    ////examples):
                    //
                    //if(style == ScalarStyle.folded && lineBreak == '\n')
                    //{
                    //    if(startLen == endLen)
                    //    {
                    //        if(!" \t"d.canFind(reader_.peekByte()))
                    //        {
                    //            reader_.sliceBuilder.write(' ');
                    //        }
                    //        else
                    //        {
                    //            chunks ~= lineBreak;
                    //        }
                    //    }
                    //}
                    //else
                    //{
                    //    reader_.sliceBuilder.insertBack(lineBreak, endLen - startLen);
                    //}
                }
                else
                {
                    break;
                }
            }

            // If chompint is Keep, we keep (commit) the last scanned line breaks
            // (which are at the end of the scalar). Otherwise re remove them (end the
            // transaction).
            if(chomping == Chomping.keep)  { breaksTransaction.commit(); }
            else                           { breaksTransaction.end(); }
            if(chomping != Chomping.strip && lineBreak != int.max)
            {
                // If chomping is Keep, we keep the line break but the first line break
                // that isn't stripped (since chomping isn't Strip in this branch) must
                // be inserted _before_ the other line breaks.
                if(chomping == Chomping.keep)
                {
                    reader_.sliceBuilder.insert(lineBreak, startLen);
                }
                // If chomping is not Keep, breaksTransaction was cancelled so we can
                // directly write the first line break (as it isn't stripped - chomping
                // is not Strip)
                else
                {
                    if (lineBreak != '\0')
                    {
                        reader_.sliceBuilder.write(lineBreak);
                    }
                }
            }

            char[] slice = reader_.sliceBuilder.finish();
            return scalarToken(startMark, endMark, slice, style);
        }

        /// Scan chomping and indentation indicators of a scalar token.
        Tuple!(Chomping, int) scanBlockScalarIndicators(const Mark startMark) @safe
        {
            auto chomping = Chomping.clip;
            int increment = int.min;
            dchar c       = reader_.peek();

            /// Indicators can be in any order.
            if(getChomping(c, chomping))
            {
                getIncrement(c, increment, startMark);
            }
            else
            {
                const gotIncrement = getIncrement(c, increment, startMark);
                if(gotIncrement) { getChomping(c, chomping); }
            }

            enforce(c.among!(' ', '\0', '\n', '\r', '\u0085', '\u2028', '\u2029'),
                new ScannerException("While scanning a block scalar", startMark,
                expected("chomping or indentation indicator", c), reader_.mark));

            return tuple(chomping, increment);
        }

        /// Get chomping indicator, if detected. Return false otherwise.
        ///
        /// Used in scanBlockScalarIndicators.
        ///
        /// Params:
        ///
        /// c        = The character that may be a chomping indicator.
        /// chomping = Write the chomping value here, if detected.
        bool getChomping(ref dchar c, ref Chomping chomping) @safe
        {
            if(!c.among!('+', '-')) { return false; }
            chomping = c == '+' ? Chomping.keep : Chomping.strip;
            reader_.forward();
            c = reader_.peek();
            return true;
        }

        /// Get increment indicator, if detected. Return false otherwise.
        ///
        /// Used in scanBlockScalarIndicators.
        ///
        /// Params:
        ///
        /// c         = The character that may be an increment indicator.
        ///             If an increment indicator is detected, this will be updated to
        ///             the next character in the Reader.
        /// increment = Write the increment value here, if detected.
        /// startMark = Mark for error messages.
        bool getIncrement(ref dchar c, ref int increment, const Mark startMark) @safe
        {
            if(!c.isDigit) { return false; }
            // Convert a digit to integer.
            increment = c - '0';
            assert(increment < 10 && increment >= 0, "Digit has invalid value");

            enforce(increment > 0,
                new ScannerException("While scanning a block scalar", startMark,
                    expected("indentation indicator in range 1-9", "0"), reader_.mark));

            reader_.forward();
            c = reader_.peek();
            return true;
        }

        /// Scan (and ignore) ignored line in a block scalar.
        void scanBlockScalarIgnoredLine(const Mark startMark) @safe
        {
            findNextNonSpace();
            if(reader_.peekByte()== '#') { scanToNextBreak(); }

            enforce(reader_.peek().isBreak,
                new ScannerException("While scanning a block scalar", startMark,
                    expected("comment or line break", reader_.peek()), reader_.mark));

            scanLineBreak();
        }

        /// Scan indentation in a block scalar, returning line breaks, max indent and end mark.
        ///
        /// Assumes that the caller is building a slice in Reader, and puts the scanned
        /// characters into that slice.
        Tuple!(uint, Mark) scanBlockScalarIndentationToSlice() @safe
        {
            uint maxIndent;
            Mark endMark = reader_.mark;

            while(reader_.peek().among!(' ', '\n', '\r', '\u0085', '\u2028', '\u2029'))
            {
                if(reader_.peekByte() != ' ')
                {
                    reader_.sliceBuilder.write(scanLineBreak());
                    endMark = reader_.mark;
                    continue;
                }
                reader_.forward();
                maxIndent = max(reader_.column, maxIndent);
            }

            return tuple(maxIndent, endMark);
        }

        /// Scan line breaks at lower or specified indentation in a block scalar.
        ///
        /// Assumes that the caller is building a slice in Reader, and puts the scanned
        /// characters into that slice.
        Mark scanBlockScalarBreaksToSlice(const uint indent) @safe
        {
            Mark endMark = reader_.mark;

            for(;;)
            {
                while(reader_.column < indent && reader_.peekByte() == ' ') { reader_.forward(); }
                if(!reader_.peek().among!('\n', '\r', '\u0085', '\u2028', '\u2029'))  { break; }
                reader_.sliceBuilder.write(scanLineBreak());
                endMark = reader_.mark;
            }

            return endMark;
        }

        /// Scan a qouted flow scalar token with specified quotes.
        Token scanFlowScalar(const ScalarStyle quotes) @safe
        {
            const startMark = reader_.mark;
            const quote     = reader_.get();

            reader_.sliceBuilder.begin();

            scanFlowScalarNonSpacesToSlice(quotes, startMark);

            while(reader_.peek() != quote)
            {
                scanFlowScalarSpacesToSlice(startMark);
                scanFlowScalarNonSpacesToSlice(quotes, startMark);
            }
            reader_.forward();

            auto slice = reader_.sliceBuilder.finish();
            return scalarToken(startMark, reader_.mark, slice, quotes);
        }

        /// Scan nonspace characters in a flow scalar.
        ///
        /// Assumes that the caller is building a slice in Reader, and puts the scanned
        /// characters into that slice.
        void scanFlowScalarNonSpacesToSlice(const ScalarStyle quotes, const Mark startMark)
            @safe
        {
            for(;;)
            {
                dchar c = reader_.peek();

                size_t numCodePoints;
                while(!reader_.peek(numCodePoints).isFlowScalarBreakSpace) { ++numCodePoints; }

                if (numCodePoints > 0) { reader_.sliceBuilder.write(reader_.get(numCodePoints)); }

                c = reader_.peek();
                if(quotes == ScalarStyle.singleQuoted && c == '\'' && reader_.peek(1) == '\'')
                {
                    reader_.forward(2);
                    reader_.sliceBuilder.write('\'');
                }
                else if((quotes == ScalarStyle.doubleQuoted && c == '\'') ||
                        (quotes == ScalarStyle.singleQuoted && c.among!('"', '\\')))
                {
                    reader_.forward();
                    reader_.sliceBuilder.write(c);
                }
                else if(quotes == ScalarStyle.doubleQuoted && c == '\\')
                {
                    reader_.forward();
                    c = reader_.peek();
                    if(c.among!(escapes))
                    {
                        reader_.forward();
                        // Escaping has been moved to Parser as it can't be done in
                        // place (in a slice) in case of '\P' and '\L' (very uncommon,
                        // but we don't want to break the spec)
                        char[2] escapeSequence = ['\\', cast(char)c];
                        reader_.sliceBuilder.write(escapeSequence);
                    }
                    else if(c.among!(escapeHexCodeList))
                    {
                        const hexLength = dyaml.escapes.escapeHexLength(c);
                        reader_.forward();

                        foreach(i; 0 .. hexLength) {
                            enforce(reader_.peek(i).isHexDigit,
                                new ScannerException("While scanning a double quoted scalar", startMark,
                                    expected("escape sequence of hexadecimal numbers",
                                        reader_.peek(i)), reader_.mark));
                        }
                        char[] hex = reader_.get(hexLength);

                        enforce((hex.length > 0) && (hex.length <= 8),
                            new ScannerException("While scanning a double quoted scalar", startMark,
                                  "overflow when parsing an escape sequence of " ~
                                  "hexadecimal numbers.", reader_.mark));

                        char[2] escapeStart = ['\\', cast(char) c];
                        reader_.sliceBuilder.write(escapeStart);
                        reader_.sliceBuilder.write(hex);

                    }
                    else if(c.among!('\n', '\r', '\u0085', '\u2028', '\u2029'))
                    {
                        scanLineBreak();
                        scanFlowScalarBreaksToSlice(startMark);
                    }
                    else
                    {
                        throw new ScannerException("While scanning a double quoted scalar", startMark,
                              text("found unsupported escape character ", c),
                              reader_.mark);
                    }
                }
                else { return; }
            }
        }

        /// Scan space characters in a flow scalar.
        ///
        /// Assumes that the caller is building a slice in Reader, and puts the scanned
        /// spaces into that slice.
        void scanFlowScalarSpacesToSlice(const Mark startMark) @safe
        {
            // Increase length as long as we see whitespace.
            size_t length;
            while(reader_.peekByte(length).among!(' ', '\t')) { ++length; }
            auto whitespaces = reader_.prefixBytes(length);

            // Can check the last byte without striding because '\0' is ASCII
            const c = reader_.peek(length);
            enforce(c != '\0',
                new ScannerException("While scanning a quoted scalar", startMark,
                    "found unexpected end of buffer", reader_.mark));

            // Spaces not followed by a line break.
            if(!c.among!('\n', '\r', '\u0085', '\u2028', '\u2029'))
            {
                reader_.forward(length);
                reader_.sliceBuilder.write(whitespaces);
                return;
            }

            // There's a line break after the spaces.
            reader_.forward(length);
            const lineBreak = scanLineBreak();

            if(lineBreak != '\n') { reader_.sliceBuilder.write(lineBreak); }

            // If we have extra line breaks after the first, scan them into the
            // slice.
            const bool extraBreaks = scanFlowScalarBreaksToSlice(startMark);

            // No extra breaks, one normal line break. Replace it with a space.
            if(lineBreak == '\n' && !extraBreaks) { reader_.sliceBuilder.write(' '); }
        }

        /// Scan line breaks in a flow scalar.
        ///
        /// Assumes that the caller is building a slice in Reader, and puts the scanned
        /// line breaks into that slice.
        bool scanFlowScalarBreaksToSlice(const Mark startMark) @safe
        {
            // True if at least one line break was found.
            bool anyBreaks;
            for(;;)
            {
                // Instead of checking indentation, we check for document separators.
                const prefix = reader_.prefix(3);
                enforce(!(prefix == "---" || prefix == "...") ||
                    !reader_.peek(3).isWhiteSpace,
                    new ScannerException("While scanning a quoted scalar", startMark,
                        "found unexpected document separator", reader_.mark));

                // Skip any whitespaces.
                while(reader_.peekByte().among!(' ', '\t')) { reader_.forward(); }

                // Encountered a non-whitespace non-linebreak character, so we're done.
                if(!reader_.peek().among!(' ', '\n', '\r', '\u0085', '\u2028', '\u2029')) { break; }

                const lineBreak = scanLineBreak();
                anyBreaks = true;
                reader_.sliceBuilder.write(lineBreak);
            }
            return anyBreaks;
        }

        /// Scan plain scalar token (no block, no quotes).
        Token scanPlain() @safe
        {
            // We keep track of the allowSimpleKey_ flag here.
            // Indentation rules are loosed for the flow context
            const startMark = reader_.mark;
            Mark endMark = startMark;
            const indent = indent_ + 1;

            // We allow zero indentation for scalars, but then we need to check for
            // document separators at the beginning of the line.
            // if(indent == 0) { indent = 1; }

            reader_.sliceBuilder.begin();

            alias Transaction = SliceBuilder.Transaction;
            Transaction spacesTransaction;
            // Stop at a comment.
            while(reader_.peekByte() != '#')
            {
                // Scan the entire plain scalar.
                size_t length;
                dchar c = reader_.peek(length);
                for(;;)
                {
                    const cNext = reader_.peek(length + 1);
                    if(c.isWhiteSpace ||
                       (flowLevel_ == 0 && c == ':' && cNext.isWhiteSpace) ||
                       (flowLevel_ > 0 && c.among!(',', ':', '?', '[', ']', '{', '}')))
                    {
                        break;
                    }
                    ++length;
                    c = cNext;
                }

                // It's not clear what we should do with ':' in the flow context.
                enforce(flowLevel_ == 0 || c != ':' ||
                   reader_.peek(length + 1).isWhiteSpace ||
                   reader_.peek(length + 1).among!(',', '[', ']', '{', '}'),
                    new ScannerException("While scanning a plain scalar", startMark,
                        "found unexpected ':' . Please check " ~
                        "http://pyyaml.org/wiki/YAMLColonInFlowContext for details.",
                        reader_.mark));

                if(length == 0) { break; }

                allowSimpleKey_ = false;

                reader_.sliceBuilder.write(reader_.get(length));

                endMark = reader_.mark;

                spacesTransaction.commit();
                spacesTransaction = Transaction(&reader_.sliceBuilder);

                const startLength = reader_.sliceBuilder.length;
                scanPlainSpacesToSlice();
                if(startLength == reader_.sliceBuilder.length ||
                   (flowLevel_ == 0 && reader_.column < indent))
                {
                    break;
                }
            }

            spacesTransaction.end();
            char[] slice = reader_.sliceBuilder.finish();

            return scalarToken(startMark, endMark, slice, ScalarStyle.plain);
        }

        /// Scan spaces in a plain scalar.
        ///
        /// Assumes that the caller is building a slice in Reader, and puts the spaces
        /// into that slice.
        void scanPlainSpacesToSlice() @safe
        {
            // The specification is really confusing about tabs in plain scalars.
            // We just forbid them completely. Do not use tabs in YAML!

            // Get as many plain spaces as there are.
            size_t length;
            while(reader_.peekByte(length) == ' ') { ++length; }
            char[] whitespaces = reader_.prefixBytes(length);
            reader_.forward(length);

            const dchar c = reader_.peek();
            if(!c.isNSChar)
            {
                // We have spaces, but no newline.
                if(whitespaces.length > 0) { reader_.sliceBuilder.write(whitespaces); }
                return;
            }

            // Newline after the spaces (if any)
            const lineBreak = scanLineBreak();
            allowSimpleKey_ = true;

            static bool end(Reader reader_) @safe pure
            {
                const prefix = reader_.prefix(3);
                return ("---" == prefix || "..." == prefix)
                        && reader_.peek(3).among!(' ', '\t', '\0', '\n', '\r', '\u0085', '\u2028', '\u2029');
            }

            if(end(reader_)) { return; }

            bool extraBreaks;

            alias Transaction = SliceBuilder.Transaction;
            auto transaction = Transaction(&reader_.sliceBuilder);
            if(lineBreak != '\n') { reader_.sliceBuilder.write(lineBreak); }
            while(reader_.peek().isNSChar)
            {
                if(reader_.peekByte() == ' ') { reader_.forward(); }
                else
                {
                    const lBreak = scanLineBreak();
                    extraBreaks  = true;
                    reader_.sliceBuilder.write(lBreak);

                    if(end(reader_)) { return; }
                }
            }
            transaction.commit();

            // No line breaks, only a space.
            if(lineBreak == '\n' && !extraBreaks) { reader_.sliceBuilder.write(' '); }
        }

        /// Scan handle of a tag token.
        ///
        /// Assumes that the caller is building a slice in Reader, and puts the scanned
        /// characters into that slice.
        void scanTagHandleToSlice(string name)(const Mark startMark)
        {
            dchar c = reader_.peek();
            enum contextMsg = "While scanning a " ~ name;
            enforce(c == '!',
                new ScannerException(contextMsg, startMark, expected("'!'", c), reader_.mark));

            uint length = 1;
            c = reader_.peek(length);
            if(c != ' ')
            {
                while(c.isAlphaNum || c.among!('-', '_'))
                {
                    ++length;
                    c = reader_.peek(length);
                }
                enforce(c == '!',
                    new ScannerException(contextMsg, startMark, expected("'!'", c), reader_.mark));
                ++length;
            }

            reader_.sliceBuilder.write(reader_.get(length));
        }

        /// Scan URI in a tag token.
        ///
        /// Assumes that the caller is building a slice in Reader, and puts the scanned
        /// characters into that slice.
        void scanTagURIToSlice(string name)(const Mark startMark)
        {
            // Note: we do not check if URI is well-formed.
            dchar c = reader_.peek();
            const startLen = reader_.sliceBuilder.length;
            {
                uint length;
                while(c.isAlphaNum || c.isURIChar)
                {
                    if(c == '%')
                    {
                        auto chars = reader_.get(length);
                        reader_.sliceBuilder.write(chars);
                        length = 0;
                        scanURIEscapesToSlice!name(startMark);
                    }
                    else { ++length; }
                    c = reader_.peek(length);
                }
                if(length > 0)
                {
                    auto chars = reader_.get(length);
                    reader_.sliceBuilder.write(chars);
                    length = 0;
                }
            }
            // OK if we scanned something, error otherwise.
            enum contextMsg = "While parsing a " ~ name;
            enforce(reader_.sliceBuilder.length > startLen,
                new ScannerException(contextMsg, startMark, expected("URI", c), reader_.mark));
        }

        // Not @nogc yet because std.utf.decode is not @nogc
        /// Scan URI escape sequences.
        ///
        /// Assumes that the caller is building a slice in Reader, and puts the scanned
        /// characters into that slice.
        void scanURIEscapesToSlice(string name)(const Mark startMark)
        {
            import core.exception : UnicodeException;
            // URI escapes encode a UTF-8 string. We store UTF-8 code units here for
            // decoding into UTF-32.
            Appender!string buffer;


            enum contextMsg = "While scanning a " ~ name;
            while(reader_.peekByte() == '%')
            {
                reader_.forward();
                char[2] nextByte = [reader_.peekByte(), reader_.peekByte(1)];

                enforce(nextByte[0].isHexDigit && nextByte[1].isHexDigit,
                    new ScannerException(contextMsg, startMark,
                        expected("URI escape sequence of 2 hexadecimal " ~
                            "numbers", nextByte), reader_.mark));

                buffer ~= nextByte[].to!ubyte(16);

                reader_.forward(2);
            }
            try
            {
                foreach (dchar chr; buffer.data)
                {
                    reader_.sliceBuilder.write(chr);
                }
            }
            catch (UnicodeException)
            {
                throw new ScannerException(contextMsg, startMark,
                        "Invalid UTF-8 data encoded in URI escape sequence",
                        reader_.mark);
            }
        }


        /// Scan a line break, if any.
        ///
        /// Transforms:
        ///   '\r\n'      :   '\n'
        ///   '\r'        :   '\n'
        ///   '\n'        :   '\n'
        ///   '\u0085'    :   '\n'
        ///   '\u2028'    :   '\u2028'
        ///   '\u2029     :   '\u2029'
        ///   no break    :   '\0'
        dchar scanLineBreak() @safe
        {
            // Fast path for ASCII line breaks.
            const b = reader_.peekByte();
            if(b < 0x80)
            {
                if(b == '\n' || b == '\r')
                {
                    if(reader_.prefix(2) == "\r\n") { reader_.forward(2); }
                    else { reader_.forward(); }
                    return '\n';
                }
                return '\0';
            }

            const c = reader_.peek();
            if(c == '\x85')
            {
                reader_.forward();
                return '\n';
            }
            if(c == '\u2028' || c == '\u2029')
            {
                reader_.forward();
                return c;
            }
            return '\0';
        }
}

// Issue 309 - https://github.com/dlang-community/D-YAML/issues/309
@safe unittest
{
    enum str = q"EOS
exp: |
  foobar
EOS".chomp;

    auto r = new Reader(cast(ubyte[])str.dup);
    auto s = Scanner(r);
    auto elems = s.map!"a.value".filter!"a.length > 0".array;
    assert(elems[1] == "foobar");
}