aboutsummaryrefslogtreecommitdiffstats
path: root/src/libs/cplusplus/pp-engine.cpp
blob: 74ca1ceb5165c9bd732282a6e33079b4b3ef6f32 (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
1827
1828
1829
1830
1831
1832
1833
1834
1835
1836
1837
1838
1839
1840
1841
1842
1843
1844
1845
1846
1847
1848
1849
1850
1851
1852
1853
1854
1855
1856
1857
1858
1859
1860
1861
1862
1863
1864
1865
1866
1867
1868
1869
1870
1871
1872
1873
1874
1875
1876
1877
1878
1879
1880
1881
1882
1883
1884
1885
1886
1887
1888
1889
1890
1891
1892
1893
1894
1895
1896
1897
1898
1899
1900
1901
1902
1903
1904
1905
1906
1907
1908
1909
1910
1911
1912
1913
1914
1915
1916
1917
1918
1919
1920
1921
1922
1923
1924
1925
1926
1927
1928
1929
1930
1931
1932
1933
1934
1935
1936
1937
1938
1939
1940
1941
1942
1943
1944
1945
1946
1947
1948
1949
1950
1951
1952
1953
1954
1955
1956
1957
1958
1959
1960
1961
1962
1963
1964
1965
1966
1967
1968
1969
1970
1971
1972
1973
1974
1975
1976
1977
1978
1979
1980
1981
1982
1983
1984
1985
1986
1987
1988
1989
1990
1991
1992
1993
1994
1995
1996
1997
1998
1999
2000
2001
2002
2003
2004
2005
2006
2007
2008
2009
2010
2011
2012
2013
2014
2015
2016
2017
2018
2019
2020
2021
2022
2023
2024
2025
2026
2027
2028
2029
2030
2031
2032
2033
2034
2035
2036
2037
2038
2039
2040
2041
2042
2043
2044
2045
2046
2047
2048
2049
2050
2051
2052
2053
2054
2055
2056
2057
2058
2059
2060
2061
2062
2063
2064
2065
2066
2067
2068
2069
2070
2071
2072
2073
2074
2075
2076
2077
2078
2079
2080
2081
2082
2083
2084
2085
2086
2087
2088
2089
2090
2091
2092
2093
2094
2095
2096
2097
2098
2099
2100
2101
2102
2103
2104
2105
2106
2107
2108
2109
2110
2111
2112
2113
2114
2115
2116
2117
2118
2119
2120
2121
2122
2123
2124
2125
2126
2127
2128
2129
2130
2131
2132
2133
2134
2135
2136
2137
2138
2139
2140
2141
2142
2143
2144
2145
2146
2147
2148
2149
2150
2151
2152
2153
2154
2155
2156
2157
2158
2159
2160
2161
2162
2163
2164
2165
2166
2167
2168
2169
2170
2171
2172
2173
2174
2175
2176
2177
2178
2179
2180
2181
2182
2183
2184
2185
2186
2187
2188
2189
2190
2191
2192
2193
// Copyright (C) 2016 The Qt Company Ltd.
// SPDX-License-Identifier: LicenseRef-Qt-Commercial OR GPL-3.0-only WITH Qt-GPL-exception-1.0

/*
  Copyright 2005 Roberto Raggi <roberto@kdevelop.org>

  Permission to use, copy, modify, distribute, and sell this software and its
  documentation for any purpose is hereby granted without fee, provided that
  the above copyright notice appear in all copies and that both that
  copyright notice and this permission notice appear in supporting
  documentation.

  The above copyright notice and this permission notice shall be included in
  all copies or substantial portions of the Software.

  THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
  IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
  FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL THE
  KDEVELOP TEAM BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN
  AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN
  CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE SOFTWARE.
*/

#include "pp.h"
#include "pp-cctype.h"

#include <cplusplus/Control.h>
#include <cplusplus/Lexer.h>
#include <cplusplus/Token.h>
#include <cplusplus/Literals.h>
#include <cplusplus/cppassert.h>

#include <utils/filepath.h>
#include <utils/scopedswap.h>

#include <QDebug>
#include <QList>
#include <QDate>
#include <QLoggingCategory>
#include <QTime>
#include <QPair>
#include <QScopeGuard>

#include <cctype>
#include <deque>
#include <list>
#include <algorithm>

// FIXME: This is used for errors that should appear in the editor.
static Q_LOGGING_CATEGORY(lexerLog, "qtc.cpp.lexer", QtWarningMsg)

using namespace Utils;

namespace {
enum {
    MAX_FUNCTION_LIKE_ARGUMENTS_COUNT = 100,
    MAX_INCLUDE_DEPTH = 200,
    MAX_TOKEN_EXPANSION_COUNT = 5000,
    MAX_TOKEN_BUFFER_DEPTH = 16000 // for when macros are using some kind of right-folding, this is the list of "delayed" buffers waiting to be expanded after the current one.
};
}

namespace {
static bool same(const char *a, const char *b, int size)
{
    return strncmp(a, b, size) == 0;
}

static bool isQtReservedWord(const char *name, int size)
{
    if (size < 3)
        return false;

    const char c = name[0];
    if (c == 'Q') {
        if (name[1] == '_') {
            name += 2;
            size -= 2;
            switch (size) {
            case 1:
                return name[0] == 'D' || name[0] == 'Q';
            case 4:
                return same(name, "SLOT", size) || same(name, "EMIT", size);
            case 5:
                return same(name, "SLOTS", size) || same(name, "ENUMS", size)
                        || same(name, "FLAGS", size);
            case 6:
                return same(name, "SIGNAL", size);
            case 7:
                return same(name, "SIGNALS", size) || same(name, "FOREACH", size);
            case 8:
                return same(name, "PROPERTY", size);
            case 9:
                return same(name, "INVOKABLE", size);
            case 10:
                return same(name, "INTERFACES", size);
            case 16:
                return same(name, "PRIVATE_PROPERTY", size);
            }
        }
        return false;
    }

    if (c == 'S')
        return (size == 6 && same(name, "SIGNAL", size)) || (size == 4 && same(name, "SLOT", size));

    if (c == 's')
        return (size == 7 && same(name, "signals", size)) || (size == 5 && same(name, "slots", size));

    if (c == 'f')
        return size == 7 && same(name, "foreach", size);

    if (c == 'e')
        return size == 4 && same(name, "emit", size);

    return false;
}

} // anonymous namespace

namespace CPlusPlus {

namespace Internal {
/// Buffers tokens for the Preprocessor::lex() to read next. Do not use  this
/// class directly, but use Preprocessor::State::pushTokenBuffer .
///
/// New tokens are added when undoing look-ahead, or after expanding a macro.
/// When macro expansion happened, the macro is passed in, and blocked until
/// all tokens generated by it (and by subsequent expansion of those generated
/// tokens) are read from the buffer. See Preprocessor::lex() for details on
/// exactly when the buffer (and subsequently a blocking macro) is removed.
struct TokenBuffer
{
    std::deque<PPToken> tokens;
    std::vector<QByteArray> blockedMacroNames;
    const Macro *macro;
    TokenBuffer *next;

    TokenBuffer(const PPToken *start, const PPToken *end, const Macro *macro, TokenBuffer *next)
        : tokens(start, end), macro(macro), next(next)
    {}

    bool isBlocked(const Macro *macro) const {
        if (!macro)
            return false;

        for (const TokenBuffer *it = this; it; it = it->next) {
            if (it->macro && (it->macro == macro || it->macro->name() == macro->name()))
                return true;
        }
        for (const QByteArray &blockedMacroName : blockedMacroNames) {
            if (macro->name() == blockedMacroName)
                return true;
        }
        return false;
    }
};

struct Value
{
    enum Kind {
        Kind_Long,
        Kind_ULong
    };

    Kind kind;

    union {
        long l;
        unsigned long ul;
    };


    Value()
        : kind(Kind_Long), l(0)
    { }

    inline bool is_ulong () const
    { return kind == Kind_ULong; }

    inline void set_ulong (unsigned long v)
    {
        ul = v;
        kind = Kind_ULong;
    }

    inline void set_long (long v)
    {
        l = v;
        kind = Kind_Long;
    }

    inline bool is_zero () const
    { return l == 0; }

    template<typename T> static int cmpImpl(T v1, T v2)
    {
        if (v1 < v2)
            return -1;
        if (v1 > v2)
            return 1;
        return 0;
    }
    Value cmp(const Value &other) const
    {
        Value v = *this;
        if (v.is_ulong() || other.is_ulong())
            v.set_long(cmpImpl(v.ul, other.ul));
        else
            v.set_long(cmpImpl(v.l, other.l));
        return v;
    }

#define PP_DEFINE_BIN_OP(name, op) \
    inline Value operator op(const Value &other) const \
    { \
        Value v = *this; \
        if (v.is_ulong () || other.is_ulong ()) \
            v.set_ulong (v.ul op other.ul); \
        else \
            v.set_long (v.l op other.l); \
        return v; \
    }

    PP_DEFINE_BIN_OP(op_add, +)
    PP_DEFINE_BIN_OP(op_sub, -)
    PP_DEFINE_BIN_OP(op_mult, *)
    PP_DEFINE_BIN_OP(op_div, /)
    PP_DEFINE_BIN_OP(op_mod, %)
    PP_DEFINE_BIN_OP(op_lhs, <<)
    PP_DEFINE_BIN_OP(op_rhs, >>)
    PP_DEFINE_BIN_OP(op_lt, <)
    PP_DEFINE_BIN_OP(op_gt, >)
    PP_DEFINE_BIN_OP(op_le, <=)
    PP_DEFINE_BIN_OP(op_ge, >=)
    PP_DEFINE_BIN_OP(op_eq, ==)
    PP_DEFINE_BIN_OP(op_ne, !=)
    PP_DEFINE_BIN_OP(op_bit_and, &)
    PP_DEFINE_BIN_OP(op_bit_or, |)
    PP_DEFINE_BIN_OP(op_bit_xor, ^)
    PP_DEFINE_BIN_OP(op_and, &&)
    PP_DEFINE_BIN_OP(op_or, ||)

#undef PP_DEFINE_BIN_OP
};

} // namespace Internal

using namespace CPlusPlus;
using namespace CPlusPlus::Internal;

namespace {

inline bool isContinuationToken(const PPToken &tk)
{
    return tk.isNot(T_EOF_SYMBOL) && (! tk.newline() || tk.joined());
}

Macro *macroDefinition(const ByteArrayRef &name,
                       unsigned bytesOffset,
                       unsigned utf16charsOffset,
                       unsigned line,
                       CPlusPlus::Environment *env,
                       Client *client)
{
    Macro *m = env->resolve(name);
    if (client) {
        if (m)
            client->passedMacroDefinitionCheck(bytesOffset, utf16charsOffset, line, *m);
        else
            client->failedMacroDefinitionCheck(bytesOffset, utf16charsOffset, name);
    }
    return m;
}

class RangeLexer
{
    const Token *first;
    const Token *last;
    Token trivial;

public:
    inline RangeLexer(const Token *first, const Token *last)
        : first(first), last(last)
    {
        // WARN: `last' must be a valid iterator.
        trivial.byteOffset = last->byteOffset;
        trivial.utf16charOffset = last->utf16charOffset;
    }

    inline explicit operator bool() const
    { return first != last; }

    inline bool isValid() const
    { return first != last; }

    inline int size() const
    { return std::distance(first, last); }

    inline const Token *dot() const
    { return first; }

    inline const Token &operator*() const
    {
        if (first != last)
            return *first;

        return trivial;
    }

    inline const Token *operator->() const
    {
        if (first != last)
            return first;

        return &trivial;
    }

    inline RangeLexer &operator++()
    {
        ++first;
        return *this;
    }
};

class ExpressionEvaluator
{
    ExpressionEvaluator(const ExpressionEvaluator &other);
    void operator = (const ExpressionEvaluator &other);

public:
    ExpressionEvaluator(Client *client, CPlusPlus::Environment *env)
        : client(client), env(env), _lex(nullptr)
    { }

    Value operator()(const Token *firstToken, const Token *lastToken,
                     const QByteArray &source)
    {
        this->source = source;
        const Value previousValue = switchValue(Value());
        RangeLexer tmp(firstToken, lastToken);
        RangeLexer *previousLex = _lex;
        _lex = &tmp;
        process_expression();
        _lex = previousLex;
        return switchValue(previousValue);
    }

protected:
    Value switchValue(const Value &value)
    {
        Value previousValue = _value;
        _value = value;
        return previousValue;
    }

    bool isTokenDefined() const
    {
        if ((*_lex)->isNot(T_IDENTIFIER))
            return false;
        const ByteArrayRef spell = tokenSpell();
        if (spell.size() != 7)
            return false;
        return spell == "defined";
    }

    const char *tokenPosition() const
    {
        return source.constData() + (*_lex)->byteOffset;
    }

    int tokenLength() const
    {
        return (*_lex)->f.bytes;
    }

    ByteArrayRef tokenSpell() const
    {
        return ByteArrayRef(tokenPosition(), tokenLength());
    }

    inline void process_expression()
    { process_constant_expression(); }

    void process_primary()
    {
        if ((*_lex)->is(T_NUMERIC_LITERAL)) {
            const char *spell = tokenPosition();
            int len = tokenLength();
            while (len) {
                const char ch = spell[len - 1];

                if (! (ch == 'u' || ch == 'U' || ch == 'l' || ch == 'L'))
                    break;
                --len;
            }

            const char *end = spell + len;
            char *vend = const_cast<char *>(end);
            _value.set_long(strtol(spell, &vend, 0));
            // TODO: if (vend != end) error(NaN)
            // TODO: binary literals
            // TODO: float literals
            ++(*_lex);
        } else if (isTokenDefined()) {
            ++(*_lex);
            if ((*_lex)->is(T_IDENTIFIER)) {
                _value.set_long(macroDefinition(tokenSpell(),
                                                (*_lex)->byteOffset,
                                                (*_lex)->utf16charOffset,
                                                (*_lex)->lineno, env, client)
                                != nullptr);
                ++(*_lex);
            } else if ((*_lex)->is(T_LPAREN)) {
                ++(*_lex);
                if ((*_lex)->is(T_IDENTIFIER)) {
                    _value.set_long(macroDefinition(tokenSpell(),
                                                    (*_lex)->byteOffset,
                                                    (*_lex)->utf16charOffset,
                                                    (*_lex)->lineno,
                                                    env, client)
                                    != nullptr);
                    ++(*_lex);
                    if ((*_lex)->is(T_RPAREN))
                        ++(*_lex);
                }
            }
        } else if ((*_lex)->is(T_IDENTIFIER)) {
            _value.set_long(0);
            ++(*_lex);
        } else if ((*_lex)->is(T_MINUS)) {
            ++(*_lex);
            process_primary();
            _value.set_long(- _value.l);
        } else if ((*_lex)->is(T_PLUS)) {
            ++(*_lex);
            process_primary();
        } else if ((*_lex)->is(T_TILDE)) {
            ++(*_lex);
            process_primary();
            _value.set_long(~ _value.l);
        } else if ((*_lex)->is(T_EXCLAIM)) {
            ++(*_lex);
            process_primary();
            _value.set_long(_value.is_zero());
        } else if ((*_lex)->is(T_LPAREN)) {
            ++(*_lex);
            process_expression();
            if ((*_lex)->is(T_RPAREN))
                ++(*_lex);
        }
    }

    Value process_expression_with_operator_precedence(const Value &lhs, int minPrecedence)
    {
        Value result = lhs;

        while (precedence((*_lex)->kind()) >= minPrecedence) {
            const int oper = (*_lex)->kind();
            const int operPrecedence = precedence(oper);
            ++(*_lex);
            process_primary();
            Value rhs = _value;

            for (int LA_token_kind = (*_lex)->kind(), LA_precedence = precedence(LA_token_kind);
                    LA_precedence > operPrecedence && isBinaryOperator(LA_token_kind);
                    LA_token_kind = (*_lex)->kind(), LA_precedence = precedence(LA_token_kind)) {
                rhs = process_expression_with_operator_precedence(rhs, LA_precedence);
            }

            result = evaluate_expression(oper, result, rhs);
        }

        return result;
    }

    void process_constant_expression()
    {
        process_primary();
        _value = process_expression_with_operator_precedence(_value, precedence(T_PIPE_PIPE));

        if ((*_lex)->is(T_QUESTION)) {
            const Value cond = _value;
            ++(*_lex);
            process_constant_expression();
            Value left = _value, right;
            if ((*_lex)->is(T_COLON)) {
                ++(*_lex);
                process_constant_expression();
                right = _value;
            }
            _value = ! cond.is_zero() ? left : right;
        }
    }

private:
    inline int precedence(int tokenKind) const
    {
        switch (tokenKind) {
        case T_PIPE_PIPE:           return 0;
        case T_AMPER_AMPER:         return 1;
        case T_PIPE:                return 2;
        case T_CARET:               return 3;
        case T_AMPER:               return 4;
        case T_EQUAL_EQUAL:
        case T_EXCLAIM_EQUAL:       return 5;
        case T_GREATER:
        case T_LESS:
        case T_LESS_EQUAL:
        case T_GREATER_EQUAL:       return 6;
        case T_LESS_EQUAL_GREATER:  return 7;
        case T_LESS_LESS:
        case T_GREATER_GREATER:     return 8;
        case T_PLUS:
        case T_MINUS:               return 9;
        case T_STAR:
        case T_SLASH:
        case T_PERCENT:             return 10;

        default:
            return -1;
        }
    }

    static inline bool isBinaryOperator(int tokenKind)
    {
        switch (tokenKind) {
        case T_PIPE_PIPE:
        case T_AMPER_AMPER:
        case T_PIPE:
        case T_CARET:
        case T_AMPER:
        case T_EQUAL_EQUAL:
        case T_EXCLAIM_EQUAL:
        case T_GREATER:
        case T_LESS:
        case T_LESS_EQUAL:
        case T_LESS_EQUAL_GREATER:
        case T_GREATER_EQUAL:
        case T_LESS_LESS:
        case T_GREATER_GREATER:
        case T_PLUS:
        case T_MINUS:
        case T_STAR:
        case T_SLASH:
        case T_PERCENT:
            return true;

        default:
            return false;
        }
    }

    static inline Value evaluate_expression(int tokenKind, const Value &lhs, const Value &rhs)
    {
        switch (tokenKind) {
        case T_PIPE_PIPE:            return lhs || rhs;
        case T_AMPER_AMPER:          return lhs && rhs;
        case T_PIPE:                 return lhs | rhs;
        case T_CARET:                return lhs ^ rhs;
        case T_AMPER:                return lhs & rhs;
        case T_EQUAL_EQUAL:          return lhs == rhs;
        case T_EXCLAIM_EQUAL:        return lhs != rhs;
        case T_GREATER:              return lhs > rhs;
        case T_LESS:                 return lhs < rhs;
        case T_LESS_EQUAL:           return lhs <= rhs;
        case T_LESS_EQUAL_GREATER:   return lhs.cmp(rhs);
        case T_GREATER_EQUAL:        return lhs >= rhs;
        case T_LESS_LESS:            return lhs << rhs;
        case T_GREATER_GREATER:      return lhs >> rhs;
        case T_PLUS:                 return lhs + rhs;
        case T_MINUS:                return lhs - rhs;
        case T_STAR:                 return lhs * rhs;
        case T_SLASH:                return rhs.is_zero() ? Value() : lhs / rhs;
        case T_PERCENT:              return rhs.is_zero() ? Value() : lhs % rhs;

        default:
            return Value();
        }
    }

private:
    Client *client;
    Environment *env;
    QByteArray source;
    RangeLexer *_lex;
    Value _value;
};

} // end of anonymous namespace

Preprocessor::State::State()
    : m_lexer(nullptr)
    , m_skipping(MAX_LEVEL)
    , m_trueTest(MAX_LEVEL)
    , m_ifLevel(0)
    , m_tokenBufferDepth(0)
    , m_tokenBuffer(nullptr)
    , m_inPreprocessorDirective(false)
    , m_markExpandedTokens(true)
    , m_noLines(false)
    , m_inCondition(false)
    , m_bytesOffsetRef(0)
    , m_utf16charsOffsetRef(0)
    , m_result(nullptr)
    , m_lineRef(1)
    , m_currentExpansion(nullptr)
    , m_includeGuardState(IncludeGuardState_BeforeIfndef)
{
    m_skipping[m_ifLevel] = false;
    m_trueTest[m_ifLevel] = false;

    m_expansionResult.reserve(256);
    setExpansionStatus(NotExpanding);
}

#define COMPRESS_TOKEN_BUFFER
void Preprocessor::State::pushTokenBuffer(const PPToken *start, const PPToken *end, const Macro *macro)
{
    if (m_tokenBufferDepth <= MAX_TOKEN_BUFFER_DEPTH) {
#ifdef COMPRESS_TOKEN_BUFFER
        if (macro || !m_tokenBuffer) {
            // If there is a new blocking macro (or no token buffer yet), create
            // one.
            m_tokenBuffer = new TokenBuffer(start, end, macro, m_tokenBuffer);
            ++m_tokenBufferDepth;
        } else {
            // No new blocking macro is passed in, so tokens can be prepended to
            // the existing buffer.
            m_tokenBuffer->tokens.insert(m_tokenBuffer->tokens.begin(), start, end);
        }
#else
        m_tokenBuffer = new TokenBuffer(start, end, macro, m_tokenBuffer);
        ++m_tokenBufferDepth;
#endif
    }
}

void Preprocessor::State::popTokenBuffer()
{
    TokenBuffer *r = m_tokenBuffer;
    m_tokenBuffer = m_tokenBuffer->next;
    delete r;

    if (m_tokenBufferDepth)
        --m_tokenBufferDepth;
}

#ifdef DEBUG_INCLUDE_GUARD_TRACKING
QString Preprocessor::State::guardStateToString(int guardState)
{
    switch (guardState) {
    case IncludeGuardState_NoGuard: return QLatin1String("NoGuard");
    case IncludeGuardState_BeforeIfndef: return QLatin1String("BeforeIfndef");
    case IncludeGuardState_AfterIfndef: return QLatin1String("AfterIfndef");
    case IncludeGuardState_AfterDefine: return QLatin1String("AfterDefine");
    case IncludeGuardState_AfterEndif: return QLatin1String("AfterEndif");
    default: return QLatin1String("UNKNOWN");
    }
}
#endif // DEBUG_INCLUDE_GUARD_TRACKING

/**
 * @brief Update the include-guard tracking state.
 *
 * Include guards are the #ifdef/#define/#endif sequence typically found in
 * header files to prevent repeated definition of the contents of that header
 * file. So, for a file to have an include guard, it must look like this:
 * \code
 * #ifndef SOME_ID
 * ... all declarations/definitions/etc. go here ...
 * #endif
 * \endcode
 *
 * SOME_ID is an identifier, and is also the include guard. The only tokens
 * allowed before the #ifndef and after the #endif are comments (in any form)
 * or #line directives. The only other requirement is that a #define SOME_ID
 * occurs inside the #ifndef block, but not nested inside other
 * #if/#ifdef/#ifndef blocks.
 *
 * This function tracks the state, and is called from \c updateIncludeGuardState
 * which handles the most common no-op cases.
 *
 * @param hint indicates what kind of token is encountered in the input
 * @param idToken the identifier token that ought to be in the input
 *        after a #ifndef or a #define .
 */
void Preprocessor::State::updateIncludeGuardState_helper(IncludeGuardStateHint hint, PPToken *idToken)
{
#ifdef DEBUG_INCLUDE_GUARD_TRACKING
    int oldIncludeGuardState = m_includeGuardState;
    QByteArray oldIncludeGuardMacroName = m_includeGuardMacroName;
#endif // DEBUG_INCLUDE_GUARD_TRACKING

    switch (m_includeGuardState) {
    case IncludeGuardState_NoGuard:
        break;
    case IncludeGuardState_BeforeIfndef:
        if (hint == IncludeGuardStateHint_Ifndef
                && idToken && idToken->is(T_IDENTIFIER)) {
            m_includeGuardMacroName = idToken->asByteArrayRef().toByteArray();
            m_includeGuardState = IncludeGuardState_AfterIfndef;
        } else {
            m_includeGuardState = IncludeGuardState_NoGuard;
        }
        break;
    case IncludeGuardState_AfterIfndef:
        if (hint == IncludeGuardStateHint_Define
                && idToken && idToken->is(T_IDENTIFIER)
                && idToken->asByteArrayRef() == m_includeGuardMacroName)
            m_includeGuardState = IncludeGuardState_AfterDefine;
        break;
    case IncludeGuardState_AfterDefine:
        if (hint == IncludeGuardStateHint_Endif)
            m_includeGuardState = IncludeGuardState_AfterEndif;
        break;
    case IncludeGuardState_AfterEndif:
        m_includeGuardState = IncludeGuardState_NoGuard;
        m_includeGuardMacroName.clear();
        break;
    }

#ifdef DEBUG_INCLUDE_GUARD_TRACKING
    qDebug() << "***" << guardStateToString(oldIncludeGuardState)
             << "->" << guardStateToString(m_includeGuardState)
             << "hint:" << hint
             << "guard:" << oldIncludeGuardMacroName << "->" << m_includeGuardMacroName;
#endif // DEBUG_INCLUDE_GUARD_TRACKING
}

const FilePath &Preprocessor::configurationFileName()
{
    const static FilePath configurationFile = FilePath::fromPathPart(u"<configuration>");
    return configurationFile;
}

Preprocessor::Preprocessor(Client *client, Environment *env)
    : m_client(client)
    , m_env(env)
    , m_expandFunctionlikeMacros(true)
    , m_keepComments(false)
{
    m_scratchBuffer.reserve(256);
}

QByteArray Preprocessor::run(const Utils::FilePath &filePath,
                             const QByteArray &source,
                             bool noLines,
                             bool markGeneratedTokens)
{
    return run(filePath.toString(), source, noLines, markGeneratedTokens);
}

QByteArray Preprocessor::run(const QString &fileName,
                             const QByteArray &source,
                             bool noLines,
                             bool markGeneratedTokens)
{
    m_scratchBuffer.clear();

    QByteArray preprocessed, includeGuardMacroName;
    preprocessed.reserve(source.size() * 2); // multiply by 2 because we insert #gen lines.
    preprocess(fileName, source, &preprocessed, &includeGuardMacroName, noLines,
               markGeneratedTokens, false);
    if (m_client && !includeGuardMacroName.isEmpty())
        m_client->markAsIncludeGuard(includeGuardMacroName);
    return preprocessed;
}

void Preprocessor::setCancelChecker(const Preprocessor::CancelChecker &cancelChecker)
{
    m_cancelChecker = cancelChecker;
}

bool Preprocessor::expandFunctionlikeMacros() const
{
    return m_expandFunctionlikeMacros;
}

void Preprocessor::setExpandFunctionlikeMacros(bool expandMacros)
{
    m_expandFunctionlikeMacros = expandMacros;
}

bool Preprocessor::keepComments() const
{
    return m_keepComments;
}

void Preprocessor::setKeepComments(bool keepComments)
{
    m_keepComments = keepComments;
}

void Preprocessor::generateOutputLineMarker(unsigned lineno)
{
    maybeStartOutputLine();
    QByteArray &marker = currentOutputBuffer();
    marker.append("# ");
    marker.append(QByteArray::number(lineno));
    marker.append(" \"");
    marker.append(m_env->currentFileUtf8);
    marker.append("\"\n");
}

void Preprocessor::handleDefined(PPToken *tk)
{
    ScopedBoolSwap s(m_state.m_inPreprocessorDirective, true);
    unsigned lineno = tk->lineno;
    lex(tk); // consume "defined" token
    bool lparenSeen = tk->is(T_LPAREN);
    if (lparenSeen)
        lex(tk); // consume "(" token
    if (tk->isNot(T_IDENTIFIER))
        //### TODO: generate error message
        return;
    PPToken idToken = *tk;
    do {
        lex(tk);
        if (tk->isNot(T_POUND_POUND))
            break;
        lex(tk);
        if (tk->is(T_IDENTIFIER))
            idToken = generateConcatenated(idToken, *tk);
        else
            break;
    } while (isContinuationToken(*tk));


    if (lparenSeen && tk->is(T_RPAREN))
        lex(tk);

    pushToken(tk);

    QByteArray result(1, '0');
    const ByteArrayRef macroName = idToken.asByteArrayRef();
    if (macroDefinition(macroName,
                        idToken.byteOffset + m_state.m_bytesOffsetRef,
                        idToken.utf16charOffset + m_state.m_utf16charsOffsetRef,
                        idToken.lineno, m_env, m_client)) {
        result[0] = '1';
    }
    *tk = generateToken(T_NUMERIC_LITERAL, result.constData(), result.size(), lineno, false);
}

void Preprocessor::pushToken(Preprocessor::PPToken *tk)
{
    const PPToken currentTokenBuffer[] = {*tk};
    m_state.pushTokenBuffer(currentTokenBuffer, currentTokenBuffer + 1, nullptr);
}

void Preprocessor::lex(PPToken *tk)
{
again:
    if (m_state.m_tokenBuffer) {
        // There is a token buffer, so read from there.
        if (m_state.m_tokenBuffer->tokens.empty()) {
            // The token buffer is empty, so pop it, and start over.
            m_state.popTokenBuffer();
            goto again;
        }
        *tk = m_state.m_tokenBuffer->tokens.front();
        m_state.m_tokenBuffer->tokens.pop_front();
        // The token buffer might now be empty. We leave it in, because the
        // token we just read might expand into new tokens, or might be a call
        // to the macro that generated this token. In either case, the macro
        // that generated the token still needs to be blocked (!), which is
        // recorded in the token buffer. Removing the blocked macro and the
        // empty token buffer happens the next time that this function is called.
    } else {
        // No token buffer, so have the lexer scan the next token.
        tk->setSource(m_state.m_source);
        m_state.m_lexer->scan(tk);
    }

    // Adjust token's line number in order to take into account the environment reference.
    tk->lineno += m_state.m_lineRef - 1;

reclassify:
    if (! m_state.m_inPreprocessorDirective) {
        if (tk->newline() && tk->is(T_POUND)) {
            handlePreprocessorDirective(tk);
            goto reclassify;
        } else if (tk->newline() && skipping()) {
            ScopedBoolSwap s(m_state.m_inPreprocessorDirective, true);
            do {
                lex(tk);
            } while (isContinuationToken(*tk));
            goto reclassify;
        } else if (tk->is(T_IDENTIFIER) && !isQtReservedWord(tk->tokenStart(), tk->bytes())) {
            m_state.updateIncludeGuardState(State::IncludeGuardStateHint_OtherToken);
            if (m_state.m_inCondition && tk->asByteArrayRef() == "defined") {
                handleDefined(tk);
            } else {
                synchronizeOutputLines(*tk);
                if (handleIdentifier(tk))
                    goto again;
            }
        } else if (tk->isNot(T_COMMENT) && tk->isNot(T_EOF_SYMBOL)) {
            m_state.updateIncludeGuardState(State::IncludeGuardStateHint_OtherToken);
        }
    }
}

void Preprocessor::skipPreprocesorDirective(PPToken *tk)
{
    ScopedBoolSwap s(m_state.m_inPreprocessorDirective, true);

    while (isContinuationToken(*tk)) {
        scanComment(tk);
        lex(tk);
    }
}

bool Preprocessor::handleIdentifier(PPToken *tk)
{
    ScopedBoolSwap s(m_state.m_inPreprocessorDirective, !tk->f.expanded);

    static const QByteArray ppLine("__LINE__");
    static const QByteArray ppFile("__FILE__");
    static const QByteArray ppDate("__DATE__");
    static const QByteArray ppTime("__TIME__");

    ByteArrayRef macroNameRef = tk->asByteArrayRef();

    if (macroNameRef.size() == 8
            && macroNameRef[0] == '_'
            && macroNameRef[1] == '_') {
        PPToken newTk;
        if (macroNameRef == ppLine) {
            QByteArray txt = QByteArray::number(tk->lineno);
            newTk = generateToken(T_STRING_LITERAL, txt.constData(), txt.size(), tk->lineno, false);
        } else if (macroNameRef == ppFile) {
            QByteArray txt;
            txt.append('"');
            txt.append(m_env->currentFileUtf8);
            txt.append('"');
            newTk = generateToken(T_STRING_LITERAL, txt.constData(), txt.size(), tk->lineno, false);
        } else if (macroNameRef == ppDate) {
            QByteArray txt;
            txt.append('"');
            txt.append(QDate::currentDate().toString().toUtf8());
            txt.append('"');
            newTk = generateToken(T_STRING_LITERAL, txt.constData(), txt.size(), tk->lineno, false);
        } else if (macroNameRef == ppTime) {
            QByteArray txt;
            txt.append('"');
            txt.append(QTime::currentTime().toString().toUtf8());
            txt.append('"');
            newTk = generateToken(T_STRING_LITERAL, txt.constData(), txt.size(), tk->lineno, false);
        }

        if (newTk.hasSource()) {
            newTk.f.newline = tk->newline();
            newTk.f.whitespace = tk->whitespace();
            *tk = newTk;
            return false;
        }
    }

    Macro *macro = m_env->resolve(macroNameRef);
    if (!macro
        || (tk->expanded() && m_state.m_tokenBuffer && m_state.m_tokenBuffer->isBlocked(macro))) {
        return false;
    }
//    qDebug() << "expanding" << macro->name() << "on line" << tk->lineno;

    // Keep track the of the macro identifier token.
    PPToken idTk = *tk;

    // Expanded tokens which are not generated ones preserve the original line number from
    // their corresponding argument in macro substitution. For expanded tokens which are
    // generated, this information must be taken from somewhere else. What we do is to keep
    // a "reference" line initialize set to the line where expansion happens.
    unsigned baseLine = idTk.lineno - m_state.m_lineRef + 1;

    QVector<PPToken> body = macro->definitionTokens();

    // Within nested expansion we might reach a previously added marker token. In this case,
    // we need to move it from its current possition to outside the nesting.
    PPToken oldMarkerTk;

    if (macro->isFunctionLike()) {
        if (!expandFunctionlikeMacros()
                // Still expand if this originally started with an object-like macro.
                && m_state.m_expansionStatus != Expanding) {
            if (m_client) {
                m_client->notifyMacroReference(m_state.m_bytesOffsetRef + idTk.byteOffset,
                                               m_state.m_utf16charsOffsetRef + idTk.utf16charOffset,
                                               idTk.lineno,
                                               *macro);
            }
            return false;
        }

        // Collect individual tokens that form the macro arguments.
        QVector<QVector<PPToken> > allArgTks;
        bool hasArgs = collectActualArguments(tk, &allArgTks, macro->name());

        // Check whether collecting arguments failed due to a previously added marker
        // that goot nested in a sequence of expansions. If so, store it and try again.
        if (!hasArgs
                && !tk->hasSource()
                && m_state.m_markExpandedTokens
                && (m_state.m_expansionStatus == Expanding
                    || m_state.m_expansionStatus == ReadyForExpansion)) {
            oldMarkerTk = *tk;
            hasArgs = collectActualArguments(tk, &allArgTks, macro->name());
        }

        // Check for matching parameter/argument count.
        bool hasMatchingArgs = false;
        if (hasArgs) {
            const int expectedArgCount = macro->formals().size();
            if (macro->isVariadic() && allArgTks.size() == expectedArgCount - 1)
                allArgTks.push_back(QVector<PPToken>());
            const int actualArgCount = allArgTks.size();
            if (expectedArgCount == actualArgCount
                    || (macro->isVariadic() && actualArgCount > expectedArgCount - 1)
                    // Handle '#define foo()' when invoked as 'foo()'
                    || (expectedArgCount == 0
                        && actualArgCount == 1
                        && allArgTks.at(0).isEmpty())) {
                hasMatchingArgs = true;
            }
        }

        if (!hasArgs || !hasMatchingArgs) {
            //### TODO: error message
            pushToken(tk);
            // If a previous marker was found, make sure to put it back.
            if (oldMarkerTk.bytes())
                pushToken(&oldMarkerTk);
            *tk = idTk;
            return false;
        }

        if (m_client && !idTk.generated()) {
            // Bundle each token sequence into a macro argument "reference" for notification.
            // Even empty ones, which are not necessarily important on its own, but for the matter
            // of couting their number - such as in foo(,)
            QVector<MacroArgumentReference> argRefs;
            for (int i = 0; i < allArgTks.size(); ++i) {
                const QVector<PPToken> &argTks = allArgTks.at(i);
                if (argTks.isEmpty()) {
                    argRefs.push_back(MacroArgumentReference());
                } else {

                    argRefs.push_back(MacroArgumentReference(
                                  m_state.m_bytesOffsetRef + argTks.first().bytesBegin(),
                                  argTks.last().bytesBegin() + argTks.last().bytes()
                                    - argTks.first().bytesBegin(),
                                  m_state.m_utf16charsOffsetRef + argTks.first().utf16charsBegin(),
                                  argTks.last().utf16charsBegin() + argTks.last().utf16chars()
                                    - argTks.first().utf16charsBegin()));
                }
            }

            m_client->startExpandingMacro(m_state.m_bytesOffsetRef + idTk.byteOffset,
                                          m_state.m_utf16charsOffsetRef + idTk.utf16charOffset,
                                          idTk.lineno,
                                          *macro,
                                          argRefs);
        }

        if (allArgTks.size() > MAX_FUNCTION_LIKE_ARGUMENTS_COUNT)
            return false;

        if (!handleFunctionLikeMacro(macro, body, allArgTks, baseLine)) {
            if (m_client && !idTk.expanded())
                m_client->stopExpandingMacro(idTk.byteOffset, *macro);
            return false;
        }
    } else if (m_client && !idTk.generated()) {
        m_client->startExpandingMacro(m_state.m_bytesOffsetRef + idTk.byteOffset,
                                      m_state.m_utf16charsOffsetRef + idTk.utf16charOffset,
                                      idTk.lineno, *macro);
    }

    if (body.isEmpty()) {
        if (m_state.m_markExpandedTokens
                && (m_state.m_expansionStatus == NotExpanding
                    || m_state.m_expansionStatus == JustFinishedExpansion)) {
            // This is not the most beautiful approach but it's quite reasonable. What we do here
            // is to create a fake identifier token which is only composed by whitespaces. It's
            // also not marked as expanded so it it can be treated as a regular token.
            const QByteArray content(int(idTk.bytes() + computeDistance(idTk)), ' ');
            PPToken fakeIdentifier = generateToken(T_IDENTIFIER,
                                                   content.constData(), content.length(),
                                                   idTk.lineno, false, false);
            fakeIdentifier.f.whitespace = true;
            fakeIdentifier.f.expanded = false;
            fakeIdentifier.f.generated = false;
            body.push_back(fakeIdentifier);
        }
    } else {
        // The first body token replaces the macro invocation so its whitespace and
        // newline info is replicated.
        PPToken &bodyTk = body[0];
        bodyTk.f.whitespace = idTk.whitespace();
        bodyTk.f.newline = idTk.newline();

        // Expansions are tracked from a "top-level" basis. This means that each expansion
        // section in the output corresponds to a direct use of a macro (either object-like
        // or function-like) in the source code and all its recurring expansions - they are
        // surrounded by two marker tokens, one at the begin and the other at the end.
        // For instance, the following code will generate 3 expansions in total, but the
        // output will aggregate the tokens in only 2 expansion sections.
        //  - The first corresponds to BAR expanding to FOO and then FOO expanding to T o;
        //  - The second corresponds to FOO expanding to T o;
        //
        // #define FOO(T, o) T o;
        // #define BAR(T, o) FOO(T, o)
        // BAR(Test, x) FOO(Test, y)
        if (m_state.m_markExpandedTokens) {
            if (m_state.m_expansionStatus == NotExpanding
                    || m_state.m_expansionStatus == JustFinishedExpansion) {
                PPToken marker;
                marker.f.expanded = true;
                marker.f.bytes = idTk.bytes();
                marker.byteOffset = idTk.byteOffset;
                marker.lineno = idTk.lineno;
                body.prepend(marker);
                body.append(marker);
                m_state.setExpansionStatus(ReadyForExpansion);
            } else if (oldMarkerTk.bytes()
                       && (m_state.m_expansionStatus == ReadyForExpansion
                           || m_state.m_expansionStatus == Expanding)) {
                body.append(oldMarkerTk);
            }
        }
    }

    const PPToken *start = body.constData();
    m_state.pushTokenBuffer(start, start + body.size(), macro);

    if (m_client && !idTk.generated())
        m_client->stopExpandingMacro(idTk.byteOffset, *macro);

    return true;
}

bool Preprocessor::handleFunctionLikeMacro(const Macro *macro,
                                           QVector<PPToken> &body,
                                           const QVector<QVector<PPToken> > &actuals,
                                           unsigned baseLine)
{
    QVector<PPToken> expanded;

    const auto addToken = [&expanded](PPToken &&tok) {
        if (expanded.isEmpty())
            expanded.reserve(50);
        expanded.push_back(std::move(tok));
    };

    const size_t bodySize = body.size();
    for (size_t i = 0; i < bodySize && expanded.size() < MAX_TOKEN_EXPANSION_COUNT;
            ++i) {
        int expandedSize = expanded.size();
        PPToken bodyTk = body.at(int(i));

        if (bodyTk.is(T_IDENTIFIER)) {
            const ByteArrayRef id = bodyTk.asByteArrayRef();
            const QVector<QByteArray> &formals = macro->formals();
            int j = 0;
            for (; j < formals.size() && expanded.size() < MAX_TOKEN_EXPANSION_COUNT; ++j) {
                if (formals[j] == id) {
                    QVector<PPToken> actualsForThisParam = actuals.at(j);
                    unsigned lineno = baseLine;

                    // Collect variadic arguments
                    if (id == "__VA_ARGS__" || (macro->isVariadic() && j + 1 == formals.size())) {
                        for (int k = j + 1; k < actuals.size(); ++k) {
                            actualsForThisParam.append(generateToken(T_COMMA, ",", 1, lineno, true));
                            actualsForThisParam += actuals.at(k);
                        }
                    }

                    const int actualsSize = actualsForThisParam.size();

                    if (i > 0 && body[int(i) - 1].is(T_POUND)) {
                        QByteArray enclosedString;
                        enclosedString.reserve(256);

                        for (int i = 0; i < actualsSize; ++i) {
                            const PPToken &t = actualsForThisParam.at(i);
                            if (i == 0)
                                lineno = t.lineno;
                            else if (t.whitespace())
                                enclosedString.append(' ');
                            enclosedString.append(t.tokenStart(), t.bytes());
                        }
                        enclosedString.replace("\\", "\\\\");
                        enclosedString.replace("\"", "\\\"");

                        addToken(generateToken(T_STRING_LITERAL,
                                               enclosedString.constData(),
                                               enclosedString.size(),
                                               lineno, true));
                    } else {
                        for (int k = 0; k < actualsSize; ++k) {
                            // Mark the actual tokens (which are the replaced version of the
                            // body's one) as expanded. For the first token we replicate the
                            // body's whitespace info.
                            PPToken actual = actualsForThisParam.at(k);
                            actual.f.expanded = true;
                            if (k == 0)
                                actual.f.whitespace = bodyTk.whitespace();
                            if (k == actualsSize - 1)
                                lineno = actual.lineno;
                            addToken(std::move(actual));
                        }
                    }

                    // Get a better (more up-to-date) value for the base line.
                    baseLine = lineno;

                    break;
                }
            }

            if (j == formals.size()) {
                // No formal macro parameter for this identifier in the body.
                bodyTk.f.generated = true;
                bodyTk.lineno = baseLine;
                addToken(std::move(bodyTk));
            }
        } else if (bodyTk.isNot(T_POUND) && bodyTk.isNot(T_POUND_POUND)) {
            bodyTk.f.generated = true;
            bodyTk.lineno = baseLine;
            addToken(std::move(bodyTk));
        }

        if (i > 1 && body[int(i) - 1].is(T_POUND_POUND)) {
            if (expandedSize < 1 || expanded.size() == expandedSize) //### TODO: [cpp.concat] placemarkers
                continue;
            const PPToken &leftTk = expanded[expandedSize - 1];
            const PPToken &rightTk = expanded[expandedSize];
            expanded[expandedSize - 1] = generateConcatenated(leftTk, rightTk);
            expanded.remove(expandedSize);
        }
    }

    // The "new" body.
    body = expanded;
    body.squeeze();

    return true;
}

void Preprocessor::trackExpansionCycles(PPToken *tk)
{
    if (m_state.m_markExpandedTokens) {
        // Identify a macro expansion section. The format is as follows:
        //
        // # expansion begin x,y ~g l:c
        // ...
        // # expansion end
        //
        // The x and y correspond, respectively, to the offset where the macro invocation happens
        // and the macro name's length. Following that there might be an unlimited number of
        // token marks which are directly mapped to each token that appears in the expansion.
        // Something like ~g indicates that the following g tokens are all generated. While
        // something like l:c indicates that the following token is expanded but not generated
        // and is positioned on line l and column c. Example:
        //
        // #define FOO(X) int f(X = 0)  // line 1
        // FOO(int
        //     a);
        //
        // Output would be:
        // # expansion begin 8,3 ~3 2:4 3:4 ~3
        // int f(int a = 0)
        // # expansion end
        // # 3 filename
        //       ;
        if (tk->expanded() && !tk->hasSource()) {
            if (m_state.m_expansionStatus == ReadyForExpansion) {
                m_state.setExpansionStatus(Expanding);
                m_state.m_expansionResult.clear();
                m_state.m_expandedTokensInfo.clear();
            } else if (m_state.m_expansionStatus == Expanding) {
                m_state.setExpansionStatus(JustFinishedExpansion);

                QByteArray &buffer = currentOutputBuffer();
                maybeStartOutputLine();

                // Offset and length of the macro invocation
                char chunk[40];
                qsnprintf(chunk, sizeof(chunk), "# expansion begin %d,%d", tk->byteOffset,
                          tk->bytes());
                buffer.append(chunk);

                // Expanded tokens
                unsigned generatedCount = 0;
                for (int i = 0; i < m_state.m_expandedTokensInfo.size(); ++i) {
                    const QPair<unsigned, unsigned> &p = m_state.m_expandedTokensInfo.at(i);
                    if (p.first) {
                        if (generatedCount) {
                            qsnprintf(chunk, sizeof(chunk), " ~%d", generatedCount);
                            buffer.append(chunk);
                            generatedCount = 0;
                        }
                        qsnprintf(chunk, sizeof(chunk), " %d:%d", p.first, p.second);
                        buffer.append(chunk);
                    } else {
                        ++generatedCount;
                    }
                }
                if (generatedCount) {
                    qsnprintf(chunk, sizeof(chunk), " ~%d", generatedCount);
                    buffer.append(chunk);
                }
                buffer.append('\n');
                buffer.append(m_state.m_expansionResult);
                maybeStartOutputLine();
                buffer.append("# expansion end\n");
            }

            lex(tk);

            if (tk->expanded() && !tk->hasSource())
                trackExpansionCycles(tk);
        }
    }
}

static void adjustForCommentOrStringNewlines(int *currentLine, const PPToken &tk)
{
    if (tk.isComment() || tk.isStringLiteral())
        (*currentLine) += tk.asByteArrayRef().count('\n');
}

void Preprocessor::synchronizeOutputLines(const PPToken &tk, bool forceLine)
{
    if (m_state.m_expansionStatus != NotExpanding
            || (!forceLine && m_env->currentLine == tk.lineno)) {
        adjustForCommentOrStringNewlines(&m_env->currentLine, tk);
        return;
    }

    if (forceLine || m_env->currentLine > tk.lineno || tk.lineno - m_env->currentLine >= 9) {
        if (m_state.m_noLines) {
            if (!m_state.m_markExpandedTokens)
                currentOutputBuffer().append(' ');
        } else {
            generateOutputLineMarker(tk.lineno);
        }
    } else {
        for (int i = m_env->currentLine; i < tk.lineno; ++i)
            currentOutputBuffer().append('\n');
    }

    m_env->currentLine = tk.lineno;
    adjustForCommentOrStringNewlines(&m_env->currentLine, tk);
}

std::size_t Preprocessor::computeDistance(const Preprocessor::PPToken &tk, bool forceTillLine)
{
    // Find previous non-space character or line begin.
    const char *buffer = tk.bufferStart();
    const char *tokenBegin = tk.tokenStart();
    const char *it = tokenBegin - 1;
    for (; it >= buffer; --it) {
        if (*it == '\n'|| (!pp_isspace(*it) && !forceTillLine))
            break;
    }
    ++it;

    return tokenBegin - it;
}


void Preprocessor::enforceSpacing(const Preprocessor::PPToken &tk, bool forceSpacing)
{
    if (tk.whitespace() || forceSpacing) {
        QByteArray &buffer = currentOutputBuffer();
        // For expanded tokens we simply add a whitespace, if necessary - the exact amount of
        // whitespaces is irrelevant within an expansion section. For real tokens we must be
        // more specific and get the information from the original source.
        if (tk.expanded() && !atStartOfOutputLine()) {
            buffer.append(' ');
        } else {
            const std::size_t spacing = computeDistance(tk, forceSpacing);
            const char *tokenBegin = tk.tokenStart();
            const char *it = tokenBegin - spacing;

            // Reproduce the content as in the original line.
            for (; it != tokenBegin; ++it)
                buffer.append(pp_isspace(*it) ? *it : ' ');
        }
    }
}

/// invalid pp-tokens are used as markers to force whitespace checks.
void Preprocessor::preprocess(const QString &fileName, const QByteArray &source,
                              QByteArray *result, QByteArray *includeGuardMacroName,
                              bool noLines,
                              bool markGeneratedTokens, bool inCondition,
                              unsigned bytesOffsetRef, unsigned utf16charOffsetRef,
                              unsigned lineRef)
{
    if (source.isEmpty())
        return;

    ScopedSwap<State> savedState(m_state, State());
    m_state.m_currentFileName = fileName;
    m_state.m_source = source;
    m_state.m_lexer = new Lexer(source.constBegin(), source.constEnd());
    m_state.m_lexer->setScanKeywords(false);
    m_state.m_lexer->setScanAngleStringLiteralTokens(false);
    m_state.m_lexer->setPreprocessorMode(true);
    if (m_keepComments)
        m_state.m_lexer->setScanCommentTokens(true);
    m_state.m_result = result;
    m_state.setExpansionStatus(m_state.m_expansionStatus); // Re-set m_currentExpansion
    m_state.m_noLines = noLines;
    m_state.m_markExpandedTokens = markGeneratedTokens;
    m_state.m_inCondition = inCondition;
    m_state.m_bytesOffsetRef = bytesOffsetRef;
    m_state.m_utf16charsOffsetRef = utf16charOffsetRef;
    m_state.m_lineRef = lineRef;

    ScopedSwap<QString> savedFileName(m_env->currentFile, fileName);
    ScopedSwap<QByteArray> savedUtf8FileName(m_env->currentFileUtf8, fileName.toUtf8());
    ScopedSwap<int> savedCurrentLine(m_env->currentLine, 1);

    if (!m_state.m_noLines)
        generateOutputLineMarker(1);

    PPToken tk(m_state.m_source);
    do {
        lex(&tk);

        // Track the start and end of macro expansion cycles.
        trackExpansionCycles(&tk);

        bool macroExpanded = false;
        if (m_state.m_expansionStatus == Expanding) {
            // Collect the line and column from the tokens undergoing expansion. Those will
            // be available in the expansion section for further referencing about their real
            // location.
            unsigned trackedLine = 0;
            unsigned trackedColumn = 0;
            if (tk.expanded() && !tk.generated()) {
                trackedLine = tk.lineno;
                trackedColumn = unsigned(computeDistance(tk, true));
            }
            m_state.m_expandedTokensInfo.push_back({trackedLine, trackedColumn});
        } else if (m_state.m_expansionStatus == JustFinishedExpansion) {
            m_state.setExpansionStatus(NotExpanding);
            macroExpanded = true;
        }

        // Update environment line information.
        synchronizeOutputLines(tk, macroExpanded);

        // Make sure spacing between tokens is handled properly.
        enforceSpacing(tk, macroExpanded);

        // Finally output the token.
        if (!tk.f.trigraph) {
            currentOutputBuffer().append(tk.tokenStart(), tk.bytes());
        } else {
            switch (tk.kind()) {
            case T_LBRACKET:    currentOutputBuffer().append("["); break;
            case T_RBRACKET:    currentOutputBuffer().append("]"); break;
            case T_LBRACE:      currentOutputBuffer().append("{"); break;
            case T_RBRACE:      currentOutputBuffer().append("}"); break;
            case T_POUND:       currentOutputBuffer().append("#"); break;
            case T_POUND_POUND: currentOutputBuffer().append("##"); break;
            case T_CARET:       currentOutputBuffer().append("^"); break;
            case T_CARET_EQUAL: currentOutputBuffer().append("^="); break;
            case T_PIPE:        currentOutputBuffer().append("|"); break;
            case T_PIPE_EQUAL:  currentOutputBuffer().append("|="); break;
            case T_TILDE:       currentOutputBuffer().append("~"); break;
            case T_TILDE_EQUAL: currentOutputBuffer().append("~="); break;
            default: CPP_ASSERT(0, qDebug() << tk.spell()); break;
            }
        }

    } while (tk.isNot(T_EOF_SYMBOL));

    if (includeGuardMacroName) {
        if (m_state.m_includeGuardState == State::IncludeGuardState_AfterDefine
                || m_state.m_includeGuardState == State::IncludeGuardState_AfterEndif)
            *includeGuardMacroName = m_state.m_includeGuardMacroName;
    }
    delete m_state.m_lexer;
    while (m_state.m_tokenBuffer)
        m_state.popTokenBuffer();
}

bool Preprocessor::scanComment(Preprocessor::PPToken *tk)
{
    if (!tk->isComment())
        return false;
    synchronizeOutputLines(*tk);
    enforceSpacing(*tk, true);
    currentOutputBuffer().append(tk->tokenStart(), tk->bytes());
    return true;
}

bool Preprocessor::consumeComments(PPToken *tk)
{
    while (scanComment(tk))
        lex(tk);
    return tk->isNot(T_EOF_SYMBOL);
}

bool Preprocessor::collectActualArguments(PPToken *tk, QVector<QVector<PPToken> > *actuals,
                                          const QByteArray &parentMacroName)
{
    Q_ASSERT(tk);
    Q_ASSERT(actuals);

    QScopeGuard cleanup([this] {
        if (m_state.m_tokenBuffer && !m_state.m_tokenBuffer->blockedMacroNames.empty())
            m_state.m_tokenBuffer->blockedMacroNames.pop_back();
    });

    if (m_state.m_tokenBuffer)
        m_state.m_tokenBuffer->blockedMacroNames.push_back(parentMacroName);
    else
        cleanup.dismiss();

    lex(tk); // consume the identifier

    bool lastCommentIsCpp = false;
    while (scanComment(tk)) {
        /* After C++ comments we need to add a new line
           e.g.
             #define foo(a, b) int a = b
             foo // comment
             (x, 3);
           can result in
                 // commentint
             x = 3;
        */
        lastCommentIsCpp = tk->is(T_CPP_COMMENT) || tk->is(T_CPP_DOXY_COMMENT);
        lex(tk);
    }
    if (lastCommentIsCpp)
        maybeStartOutputLine();

    if (tk->isNot(T_LPAREN))
        //### TODO: error message
        return false;

    QVector<PPToken> tokens;
    lex(tk);
    scanActualArgument(tk, &tokens);

    actuals->append(tokens);

    while (tk->is(T_COMMA)) {
        lex(tk);

        QVector<PPToken> tokens;
        scanActualArgument(tk, &tokens);
        actuals->append(tokens);
    }

    if (!tk->is(T_RPAREN)) {
        return false;
        //###TODO: error message
    }
    return true;
}

void Preprocessor::scanActualArgument(PPToken *tk, QVector<PPToken> *tokens)
{
    Q_ASSERT(tokens);

    int count = 0;

    while (tk->isNot(T_EOF_SYMBOL)) {
        if (tk->is(T_LPAREN)) {
            ++count;
        } else if (tk->is(T_RPAREN)) {
            if (! count)
                break;
            --count;
        } else if (! count && tk->is(T_COMMA)) {
            break;
        }

        if (m_keepComments
                && (tk->is(T_CPP_COMMENT) || tk->is(T_CPP_DOXY_COMMENT))) {
            // Even in keep comments mode, we cannot preserve C++ style comments inside the
            // expansion. We stick with GCC's approach which is to replace them by C style
            // comments (currently clang just gets rid of them) and transform internals */
            // into *|.
            QByteArray text = m_state.m_source.mid(tk->bytesBegin() + 2,
                                                   tk->bytesEnd() - tk->bytesBegin() - 2);
            const QByteArray &comment = "/*" + text.replace("*/", "*|") + "*/";
            tokens->append(generateToken(T_COMMENT,
                                         comment.constData(), comment.size(),
                                         tk->lineno, false));
        } else {
            tokens->append(*tk);
        }

        lex(tk);
    }
}

void Preprocessor::handlePreprocessorDirective(PPToken *tk)
{
    ScopedBoolSwap s(m_state.m_inPreprocessorDirective, true);

    PPToken poundToken = *tk;
    lex(tk); // scan the directive

    if (tk->newline() && ! tk->joined())
        return; // nothing to do.

    static const QByteArray ppDefine("define");
    static const QByteArray ppIf("if");
    static const QByteArray ppIfDef("ifdef");
    static const QByteArray ppIfNDef("ifndef");
    static const QByteArray ppEndIf("endif");
    static const QByteArray ppElse("else");
    static const QByteArray ppUndef("undef");
    static const QByteArray ppElif("elif");
    static const QByteArray ppInclude("include");
    static const QByteArray ppIncludeNext("include_next");
    static const QByteArray ppImport("import");
    //### TODO:
    // line
    // error
    // pragma

    if (tk->is(T_IDENTIFIER)) {
        const ByteArrayRef directive = tk->asByteArrayRef();

        if (!skipping() && directive == ppDefine) {
            handleDefineDirective(tk);
        } else if (directive == ppIfNDef) {
            handleIfDefDirective(true, tk);
        } else if (directive == ppEndIf) {
            handleEndIfDirective(tk, poundToken);
        } else {
            m_state.updateIncludeGuardState(State::IncludeGuardStateHint_OtherToken);

            if (!skipping() && directive == ppUndef)
                handleUndefDirective(tk);
            else if (!skipping() && (directive == ppInclude
                                    || directive == ppImport))
                handleIncludeDirective(tk, false);
            else if (!skipping() && directive == ppIncludeNext)
                handleIncludeDirective(tk, true);
            else if (directive == ppIf)
                handleIfDirective(tk);
            else if (directive == ppIfDef)
                handleIfDefDirective(false, tk);
            else if (directive == ppElse)
                handleElseDirective(tk, poundToken);
            else if (directive == ppElif)
                handleElifDirective(tk, poundToken);
        }
    }

    skipPreprocesorDirective(tk);
}


void Preprocessor::handleIncludeDirective(PPToken *tk, bool includeNext)
{
    if (m_cancelChecker && m_cancelChecker())
        return;

    GuardLocker depthLocker(m_includeDepthGuard);
    if (m_includeDepthGuard.lockCount() > MAX_INCLUDE_DEPTH) {
        qCWarning(lexerLog) << "Maximum include depth exceeded" << m_state.m_currentFileName;
        return;
    }

    m_state.m_lexer->setScanAngleStringLiteralTokens(true);
    lex(tk); // consume "include" token
    m_state.m_lexer->setScanAngleStringLiteralTokens(false);
    const unsigned line = tk->lineno;
    QByteArray included;

    if (tk->is(T_STRING_LITERAL) || tk->is(T_ANGLE_STRING_LITERAL)) {
        included = tk->asByteArrayRef().toByteArray();
        lex(tk); // consume string token
    } else {
        included = expand(tk);
    }
    included = included.trimmed();

    if (included.isEmpty()) {
        //### TODO: error message
        return;
    }

//    qDebug("include [[%s]]", included.constData());
    Client::IncludeType mode;
    if (includeNext)
        mode = Client::IncludeNext;
    else if (included.at(0) == '"')
        mode = Client::IncludeLocal;
    else if (included.at(0) == '<')
        mode = Client::IncludeGlobal;
    else
        return; //### TODO: add error message?

    if (m_client) {
        QString inc = QString::fromUtf8(included.constData() + 1, included.size() - 2);
        m_client->sourceNeeded(line, FilePath::fromString(inc), mode);
    }
}

void Preprocessor::handleDefineDirective(PPToken *tk)
{
    const unsigned defineOffset = tk->byteOffset;
    lex(tk); // consume "define" token

    if (!consumeComments(tk))
        return;

    if (tk->isNot(T_IDENTIFIER))
        return;

    Macro macro;
    macro.setFilePath(FilePath::fromString(m_env->currentFile));
    macro.setLine(tk->lineno);
    QByteArray macroName = tk->asByteArrayRef().toByteArray();
    macro.setName(macroName);
    macro.setBytesOffset(tk->byteOffset);
    macro.setUtf16charOffset(tk->utf16charOffset);

    PPToken idToken(*tk);

    lex(tk);

    if (isContinuationToken(*tk) && tk->is(T_LPAREN) && ! tk->whitespace()) {
        macro.setFunctionLike(true);

        lex(tk); // skip `('
        if (!consumeComments(tk))
            return;

        bool hasIdentifier = false;
        if (isContinuationToken(*tk) && tk->is(T_IDENTIFIER)) {
            hasIdentifier = true;
            macro.addFormal(tk->asByteArrayRef().toByteArray());

            lex(tk);
            if (!consumeComments(tk))
                return;

            while (isContinuationToken(*tk) && tk->is(T_COMMA)) {
                lex(tk);
                if (!consumeComments(tk))
                    return;

                if (isContinuationToken(*tk) && tk->is(T_IDENTIFIER)) {
                    macro.addFormal(tk->asByteArrayRef().toByteArray());
                    lex(tk);
                    if (!consumeComments(tk))
                        return;
                } else {
                    hasIdentifier = false;
                }
            }
        }

        if (tk->is(T_DOT_DOT_DOT)) {
            macro.setVariadic(true);
            if (!hasIdentifier)
                macro.addFormal("__VA_ARGS__");
            lex(tk); // consume elipsis token
            if (!consumeComments(tk))
                return;
        }
        if (isContinuationToken(*tk) && tk->is(T_RPAREN))
            lex(tk); // consume ")" token
    } else {
        if (m_state.m_ifLevel == 1)
            m_state.updateIncludeGuardState(State::IncludeGuardStateHint_Define, &idToken);
    }

    QVector<PPToken> bodyTokens;
    unsigned previousBytesOffset = 0;
    unsigned previousUtf16charsOffset = 0;
    unsigned previousLine = 0;
    Macro *macroReference = nullptr;
    while (isContinuationToken(*tk)) {
        // Macro tokens are always marked as expanded. However, only for object-like macros
        // we mark them as generated too. For function-like macros we postpone it until the
        // formals are identified in the bodies.
        tk->f.expanded = true;
        if (!macro.isFunctionLike())
            tk->f.generated = true;

        // Identifiers must not be eagerly expanded inside defines, but we should still notify
        // in the case they are macros.
        if (tk->is(T_IDENTIFIER) && m_client) {
            macroReference = m_env->resolve(tk->asByteArrayRef());
            if (macroReference) {
                if (!macroReference->isFunctionLike()) {
                    m_client->notifyMacroReference(tk->byteOffset, tk->utf16charOffset,
                                                   tk->lineno, *macroReference);
                    macroReference = nullptr;
                }
            }
        } else if (macroReference) {
            if (m_client && tk->is(T_LPAREN)) {
                m_client->notifyMacroReference(previousBytesOffset, previousUtf16charsOffset,
                                               previousLine, *macroReference);
            }
            macroReference = nullptr;
        }

        previousBytesOffset = tk->byteOffset;
        previousUtf16charsOffset = tk->utf16charOffset;
        previousLine = tk->lineno;

        if (!scanComment(tk))
            bodyTokens.push_back(*tk);

        lex(tk);
    }

    if (isQtReservedWord(macroName.data(), macroName.size())) {
        QByteArray macroId = macro.name();

        if (macro.isFunctionLike()) {
            macroId += '(';
            bool fst = true;
            const QVector<QByteArray> formals = macro.formals();
            for (const QByteArray &formal : formals) {
                if (! fst)
                    macroId += ", ";
                fst = false;
                macroId += formal;
            }
            macroId += ')';
        }

        bodyTokens.clear();
        macro.setDefinition(macroId, bodyTokens);
    } else if (!bodyTokens.isEmpty()) {
        PPToken &firstBodyToken = bodyTokens[0];
        int start = firstBodyToken.byteOffset;
        int len = tk->byteOffset - start;
        QByteArray bodyText = firstBodyToken.source().mid(start, len).trimmed();

        const int bodySize = bodyTokens.size();
        for (int i = 0; i < bodySize; ++i) {
            PPToken &t = bodyTokens[i];
            if (t.hasSource())
                t.squeezeSource();
        }
        macro.setDefinition(bodyText, bodyTokens);
    }

    macro.setLength(tk->byteOffset - defineOffset);
    m_env->bind(macro);

//    qDebug() << "adding macro" << macro.name() << "defined at" << macro.fileName() << ":"<<macro.line();

    if (m_client)
        m_client->macroAdded(macro);
}

QByteArray Preprocessor::expand(PPToken *tk, PPToken *lastConditionToken)
{
    unsigned line = tk->lineno;
    unsigned bytesBegin = tk->bytesBegin();
    const int originalOffset = tk->originalOffset();
    unsigned utf16charsBegin = tk->utf16charsBegin();
    PPToken lastTk;
    while (isContinuationToken(*tk)) {
        lastTk = *tk;
        lex(tk);
    }
    // Gather the exact spelling of the content in the source.
    QByteArray condition(m_state.m_source.mid(originalOffset, lastTk.originalOffset() + lastTk.bytes()
                                              - originalOffset));

//    qDebug("*** Condition before: [%s]", condition.constData());
    QByteArray result;
    result.reserve(256);
    preprocess(m_state.m_currentFileName, condition, &result, nullptr, true, false, true,
               bytesBegin, utf16charsBegin, line);
    result.squeeze();
//    qDebug("*** Condition after: [%s]", result.constData());

    if (lastConditionToken)
        *lastConditionToken = lastTk;

    return result;
}

const PPToken Preprocessor::evalExpression(PPToken *tk, Value &result)
{
    PPToken lastConditionToken;
    const QByteArray expanded = expand(tk, &lastConditionToken);
    Lexer lexer(expanded.constData(), expanded.constData() + expanded.size());
    lexer.setPreprocessorMode(true);
    std::vector<Token> buf;
    Token t;
    do {
        lexer.scan(&t);
        buf.push_back(t);
    } while (t.isNot(T_EOF_SYMBOL));
    ExpressionEvaluator eval(m_client, m_env);
    result = eval(&buf[0], &buf[buf.size() - 1], expanded);
    return lastConditionToken;
}

void Preprocessor::handleIfDirective(PPToken *tk)
{
    lex(tk); // consume "if" token
    Value result;
    const PPToken lastExpressionToken = evalExpression(tk, result);

    if (!checkConditionalNesting())
        return;

    const bool value = !result.is_zero();

    const bool wasSkipping = m_state.m_skipping[m_state.m_ifLevel];
    ++m_state.m_ifLevel;
    m_state.m_trueTest[m_state.m_ifLevel] = value;
    if (wasSkipping) {
        m_state.m_skipping[m_state.m_ifLevel] = wasSkipping;
    } else {
        bool startSkipping = !value;
        m_state.m_skipping[m_state.m_ifLevel] = startSkipping;
        if (startSkipping && m_client)
            startSkippingBlocks(lastExpressionToken);
    }

}

void Preprocessor::handleElifDirective(PPToken *tk, const PPToken &poundToken)
{
    if (m_state.m_ifLevel == 0) {
        qCWarning(lexerLog) << "#elif without #if";
        handleIfDirective(tk);
    } else {
        lex(tk); // consume "elif" token
        if (m_state.m_skipping[m_state.m_ifLevel - 1]) {
            // we keep on skipping because we are nested in a skipped block
            m_state.m_skipping[m_state.m_ifLevel] = true;
        } else if (m_state.m_trueTest[m_state.m_ifLevel]) {
            if (!m_state.m_skipping[m_state.m_ifLevel]) {
                // start skipping because the preceding then-part was not skipped
                m_state.m_skipping[m_state.m_ifLevel] = true;
                if (m_client)
                    startSkippingBlocks(poundToken);
            }
        } else {
            // preceding then-part was skipped, so calculate if we should start
            // skipping, depending on the condition
            Value result;
            evalExpression(tk, result);

            bool startSkipping = result.is_zero();
            m_state.m_trueTest[m_state.m_ifLevel] = !startSkipping;
            m_state.m_skipping[m_state.m_ifLevel] = startSkipping;
            if (m_client && !startSkipping)
                m_client->stopSkippingBlocks(poundToken.utf16charOffset - 1);
        }
    }
}

void Preprocessor::handleElseDirective(PPToken *tk, const PPToken &poundToken)
{
    lex(tk); // consume "else" token

    if (m_state.m_ifLevel != 0) {
        if (m_state.m_skipping[m_state.m_ifLevel - 1]) {
            // we keep on skipping because we are nested in a skipped block
            m_state.m_skipping[m_state.m_ifLevel] = true;
        } else {
            bool wasSkipping = m_state.m_skipping[m_state.m_ifLevel];
            bool startSkipping = m_state.m_trueTest[m_state.m_ifLevel];
            m_state.m_skipping[m_state.m_ifLevel] = startSkipping;

            if (m_client && wasSkipping && !startSkipping)
                m_client->stopSkippingBlocks(poundToken.utf16charOffset - 1);
            else if (m_client && !wasSkipping && startSkipping)
                startSkippingBlocks(poundToken);
        }
    } else {
        qCWarning(lexerLog) << "#else without #if";
    }
}

void Preprocessor::handleEndIfDirective(PPToken *tk, const PPToken &poundToken)
{
    if (m_state.m_ifLevel == 0) {
        qCWarning(lexerLog) << "#endif without #if";
        if (!tk->generated())
            qCWarning(lexerLog) << "on line" << tk->lineno << "of file"
                              << m_state.m_currentFileName.toUtf8().constData();
    } else {
        bool wasSkipping = m_state.m_skipping[m_state.m_ifLevel];
        m_state.m_skipping[m_state.m_ifLevel] = false;
        m_state.m_trueTest[m_state.m_ifLevel] = false;
        --m_state.m_ifLevel;
        if (m_client && wasSkipping && !m_state.m_skipping[m_state.m_ifLevel])
            m_client->stopSkippingBlocks(poundToken.utf16charOffset - 1);

        if (m_state.m_ifLevel == 0)
            m_state.updateIncludeGuardState(State::IncludeGuardStateHint_Endif);
    }

    lex(tk); // consume "endif" token
}

void Preprocessor::handleIfDefDirective(bool checkUndefined, PPToken *tk)
{
    lex(tk); // consume "ifdef" token
    if (tk->is(T_IDENTIFIER)) {
        if (checkUndefined && m_state.m_ifLevel == 0)
            m_state.updateIncludeGuardState(State::IncludeGuardStateHint_Ifndef, tk);

        bool value = false;
        const ByteArrayRef macroName = tk->asByteArrayRef();
        if (Macro *macro = macroDefinition(macroName, tk->byteOffset, tk->utf16charOffset,
                                           tk->lineno, m_env, m_client)) {
            value = true;

            // the macro is a feature constraint(e.g. QT_NO_XXX)
            if (checkUndefined && macroName.startsWith("QT_NO_")) {
                if (macro->filePath() == configurationFileName()) {
                    // and it' defined in a pro file (e.g. DEFINES += QT_NO_QOBJECT)

                    value = false; // take the branch
                }
            }
        } else if (Environment::isBuiltinMacro(macroName)) {
            value = true;
        }

        if (checkUndefined)
            value = !value;

        const bool wasSkipping = m_state.m_skipping[m_state.m_ifLevel];

        if (checkConditionalNesting()) {
            ++m_state.m_ifLevel;
            m_state.m_trueTest[m_state.m_ifLevel] = value;
            m_state.m_skipping[m_state.m_ifLevel] = wasSkipping ? wasSkipping : !value;

            if (m_client && !wasSkipping && !value)
                startSkippingBlocks(*tk);
        }

        lex(tk); // consume the identifier
    } else {
        qCWarning(lexerLog) << "#ifdef without identifier";
    }
}

void Preprocessor::handleUndefDirective(PPToken *tk)
{
    lex(tk); // consume "undef" token
    if (tk->is(T_IDENTIFIER)) {
        const ByteArrayRef macroName = tk->asByteArrayRef();
        const unsigned bytesOffset = tk->byteOffset + m_state.m_bytesOffsetRef;
        const unsigned utf16charsOffset = tk->utf16charOffset + m_state.m_utf16charsOffsetRef;
        // Track macro use if previously defined
        if (m_client) {
            if (const Macro *existingMacro = m_env->resolve(macroName)) {
                m_client->notifyMacroReference(bytesOffset, utf16charsOffset,
                                               tk->lineno, *existingMacro);
            }
        }
        synchronizeOutputLines(*tk);
        Macro *macro = m_env->remove(macroName);

        if (m_client && macro) {
            macro->setBytesOffset(bytesOffset);
            macro->setUtf16charOffset(utf16charsOffset);
            m_client->macroAdded(*macro);
        }
        lex(tk); // consume macro name
    } else {
        qCWarning(lexerLog) << "#undef without identifier";
    }
}

PPToken Preprocessor::generateToken(enum Kind kind,
                                    const char *content, int length,
                                    unsigned lineno,
                                    bool addQuotes,
                                    bool addToControl)
{
    // When the token is a generated token, the column position cannot be
    // reconstructed, but we also have to prevent it from searching the whole
    // scratch buffer. So inserting a newline before the new token will give
    // an indent width of 0 (zero).
    m_scratchBuffer.append('\n');

    const size_t pos = m_scratchBuffer.size();

    if (kind == T_STRING_LITERAL && addQuotes)
        m_scratchBuffer.append('"');
    m_scratchBuffer.append(content, length);
    if (kind == T_STRING_LITERAL && addQuotes) {
        m_scratchBuffer.append('"');
        length += 2;
    }

    PPToken tk(m_scratchBuffer);
    tk.f.kind = kind;
    if (m_state.m_lexer->control() && addToControl) {
        if (kind == T_STRING_LITERAL)
            tk.string = m_state.m_lexer->control()->stringLiteral(m_scratchBuffer.constData() + pos, length);
        else if (kind == T_IDENTIFIER)
            tk.identifier = m_state.m_lexer->control()->identifier(m_scratchBuffer.constData() + pos, length);
        else if (kind == T_NUMERIC_LITERAL)
            tk.number = m_state.m_lexer->control()->numericLiteral(m_scratchBuffer.constData() + pos, length);
    }
    tk.byteOffset = unsigned(pos);
    tk.f.bytes = length;
    tk.f.generated = true;
    tk.f.expanded = true;
    tk.lineno = lineno;

    return tk;
}

PPToken Preprocessor::generateConcatenated(const PPToken &leftTk, const PPToken &rightTk)
{
    QByteArray newText;
    newText.reserve(leftTk.bytes() + rightTk.bytes());
    newText.append(leftTk.tokenStart(), leftTk.bytes());
    newText.append(rightTk.tokenStart(), rightTk.bytes());
    PPToken result = generateToken(T_IDENTIFIER, newText.constData(), newText.size(), leftTk.lineno, true);
    result.f.whitespace = leftTk.whitespace();
    return result;
}

void Preprocessor::startSkippingBlocks(const Preprocessor::PPToken &tk) const
{
    if (!m_client)
        return;

    unsigned utf16charIter = tk.utf16charsEnd();
    const char *source = tk.source().constData() + tk.bytesEnd();
    const char *sourceEnd = tk.source().constEnd();
    unsigned char yychar = *source;

    do {
        if (yychar == '\n') {
            m_client->startSkippingBlocks(utf16charIter + 1);
            return;
        }
        Lexer::yyinp_utf8(source, yychar, utf16charIter);
    } while (source < sourceEnd);
}

bool Preprocessor::atStartOfOutputLine() const
{
    const QByteArray *buffer = m_state.m_currentExpansion;
    return buffer->isEmpty() || buffer->endsWith('\n');
}

void Preprocessor::maybeStartOutputLine()
{
    QByteArray &buffer = currentOutputBuffer();
    if (buffer.isEmpty())
        return;
    if (!buffer.endsWith('\n'))
        buffer.append('\n');
    // If previous line ends with \ (possibly followed by whitespace), add another \n
    const char *start = buffer.constData();
    const char *ch = start + buffer.length() - 2;
    while (ch > start && (*ch != '\n') && pp_isspace(*ch))
        --ch;
    if (*ch == '\\')
        buffer.append('\n');
}

bool Preprocessor::checkConditionalNesting() const
{
    if (m_state.m_ifLevel >= MAX_LEVEL - 1) {
        qCWarning(lexerLog) << "#if/#ifdef nesting exceeding maximum level" << MAX_LEVEL;
        return false;
    }
    return true;
}


} // namespace CPlusPlus