summaryrefslogtreecommitdiffstats
path: root/admin/halog/halog.c
blob: 884a606e83dae97c8cb60cb22f49b30f28285414 (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
/*
 * haproxy log statistics reporter
 *
 * Copyright 2000-2012 Willy Tarreau <w@1wt.eu>
 *
 * This program is free software; you can redistribute it and/or
 * modify it under the terms of the GNU General Public License
 * as published by the Free Software Foundation; either version
 * 2 of the License, or (at your option) any later version.
 *
 */

#include <errno.h>
#include <fcntl.h>
#include <stdio.h>
#include <stdlib.h>
#include <syslog.h>
#include <string.h>
#include <unistd.h>
#include <ctype.h>
#include <time.h>

#include <haproxy/compiler.h>

#include <import/eb32tree.h>
#include <import/eb64tree.h>
#include <import/ebistree.h>
#include <import/ebsttree.h>

#define SOURCE_FIELD 5
#define ACCEPT_FIELD 6
#define SERVER_FIELD 8
#define TIME_FIELD 9
#define STATUS_FIELD 10
#define BYTES_SENT_FIELD 11
#define TERM_CODES_FIELD 14
#define CONN_FIELD 15
#define QUEUE_LEN_FIELD 16
#define METH_FIELD 17
#define URL_FIELD 18
#define MAXLINE 16384
#define QBITS 4

#define SEP(c) ((unsigned char)(c) <= ' ')
#define SKIP_CHAR(p,c) do { while (1) { int __c = (unsigned char)*p++; if (__c == c) break; if (__c <= ' ') { p--; break; } } } while (0)

/* [0] = err/date, [1] = req, [2] = conn, [3] = resp, [4] = data */
static struct eb_root timers[5] = {
	EB_ROOT_UNIQUE, EB_ROOT_UNIQUE, EB_ROOT_UNIQUE,
	EB_ROOT_UNIQUE, EB_ROOT_UNIQUE,
};

struct timer {
	struct eb32_node node;
	unsigned int count;
};

struct srv_st {
	unsigned int st_cnt[6]; /* 0xx to 5xx */
	unsigned int nb_ct, nb_rt, nb_ok;
	unsigned long long cum_ct, cum_rt;
	struct ebmb_node node;
	/* don't put anything else here, the server name will be there */
};

struct url_stat {
	union {
		struct ebpt_node url;
		struct eb64_node val;
	} node;
	char *url;
	unsigned long long total_time;    /* sum(all reqs' times) */
	unsigned long long total_time_ok; /* sum(all OK reqs' times) */
	unsigned long long total_bytes_sent; /* sum(all bytes sent) */
	unsigned int nb_err, nb_req;
};

#define FILT_COUNT_ONLY		0x01
#define FILT_INVERT		0x02
#define FILT_QUIET		0x04
#define FILT_ERRORS_ONLY	0x08
#define FILT_ACC_DELAY		0x10
#define FILT_ACC_COUNT		0x20
#define FILT_GRAPH_TIMERS	0x40
#define FILT_PERCENTILE		0x80
#define FILT_TIME_RESP         0x100

#define FILT_INVERT_ERRORS     0x200
#define FILT_INVERT_TIME_RESP  0x400

#define FILT_COUNT_STATUS      0x800
#define FILT_COUNT_SRV_STATUS 0x1000
#define FILT_COUNT_TERM_CODES 0x2000

#define FILT_COUNT_URL_ONLY  0x004000
#define FILT_COUNT_URL_COUNT 0x008000
#define FILT_COUNT_URL_ERR   0x010000
#define FILT_COUNT_URL_TTOT  0x020000
#define FILT_COUNT_URL_TAVG  0x040000
#define FILT_COUNT_URL_TTOTO 0x080000
#define FILT_COUNT_URL_TAVGO 0x100000

#define FILT_HTTP_ONLY       0x200000
#define FILT_TERM_CODE_NAME  0x400000
#define FILT_INVERT_TERM_CODE_NAME 0x800000

#define FILT_HTTP_STATUS           0x1000000
#define FILT_INVERT_HTTP_STATUS    0x2000000
#define FILT_QUEUE_ONLY            0x4000000
#define FILT_QUEUE_SRV_ONLY        0x8000000

#define FILT_COUNT_URL_BAVG       0x10000000
#define FILT_COUNT_URL_BTOT       0x20000000

#define FILT_COUNT_URL_ANY   (FILT_COUNT_URL_ONLY|FILT_COUNT_URL_COUNT|FILT_COUNT_URL_ERR| \
			      FILT_COUNT_URL_TTOT|FILT_COUNT_URL_TAVG|FILT_COUNT_URL_TTOTO|FILT_COUNT_URL_TAVGO| \
			      FILT_COUNT_URL_BAVG|FILT_COUNT_URL_BTOT)

#define FILT_COUNT_COOK_CODES 0x40000000
#define FILT_COUNT_IP_COUNT   0x80000000

#define FILT2_TIMESTAMP         0x01
#define FILT2_PRESERVE_QUERY    0x02
#define FILT2_EXTRACT_CAPTURE   0x04

unsigned int filter = 0;
unsigned int filter2 = 0;
unsigned int filter_invert = 0;
const char *line;
int linenum = 0;
int parse_err = 0;
int lines_out = 0;
int lines_max = -1;

const char *fgets2(FILE *stream);

void filter_count_url(const char *accept_field, const char *time_field, struct timer **tptr);
void filter_count_ip(const char *source_field, const char *accept_field, const char *time_field, struct timer **tptr);
void filter_count_srv_status(const char *accept_field, const char *time_field, struct timer **tptr);
void filter_count_cook_codes(const char *accept_field, const char *time_field, struct timer **tptr);
void filter_count_term_codes(const char *accept_field, const char *time_field, struct timer **tptr);
void filter_count_status(const char *accept_field, const char *time_field, struct timer **tptr);
void filter_graphs(const char *accept_field, const char *time_field, struct timer **tptr);
void filter_output_line(const char *accept_field, const char *time_field, struct timer **tptr);
void filter_extract_capture(const char *accept_field, const char *time_field, unsigned int, unsigned int);
void filter_accept_holes(const char *accept_field, const char *time_field, struct timer **tptr);

void usage(FILE *output, const char *msg)
{
	fprintf(output,
		"%s"
		"Usage:\n"
		"  halog [-h|--help] for long help\n"
		"  halog [input_filters]* [modifiers]* [output_format] < log\n"
		"    inp = [-e|-E] [-H] [-Q|-QS] [-rt|-RT <time>] [-ad <delay>] [-ac <count>]\n"
		"          [-hs|-HS [min][:[max]]] [-tcn|-TCN <termcode>] [-time [min][:[max]]]\n"
		"    mod = [-q] [-v] [-m <lines>] [-s <skipflds>] [-query]\n"
		"    out = {-c|-u|-uc|-ue|-ua|-ut|-uao|-uto|-uba|-ubt|-hdr <block>:<field>|\n"
		"           -cc|-gt|-pct|-st|-tc|-srv|-ic}\n"
		"\n",
		msg ? msg : ""
		);
}

void die(const char *msg)
{
	usage(stderr, msg);
	exit(1);
}

void help()
{
	usage(stdout, NULL);
	printf(
	       "Input filters - several filters may be combined\n"
	       " -H                      only match lines containing HTTP logs (ignore TCP)\n"
	       " -E                      only match lines without any error (no 5xx status)\n"
	       " -e                      only match lines with errors (status 5xx or negative)\n"
	       " -rt|-RT <time>          only match response times larger|smaller than <time>\n"
	       " -Q|-QS                  only match queued requests (any queue|server queue)\n"
	       " -tcn|-TCN <code>        only match requests with/without termination code <code>\n"
	       " -hs|-HS <[min][:][max]> only match requests with HTTP status codes within/not\n"
	       "                         within min..max. Any of them may be omitted. Exact\n"
	       "                         code is checked for if no ':' is specified.\n"
	       " -time <[min][:max]>     only match requests recorded between timestamps.\n"
	       "                         Any of them may be omitted.\n"
	       "Modifiers\n"
	       " -v                      invert the input filtering condition\n"
	       " -q                      don't report errors/warnings\n"
	       " -m <lines>              limit output to the first <lines> lines\n"
	       " -s <skip_n_fields>      skip n fields from the beginning of a line (default %d)\n"
	       "                         you can also use -n to start from earlier then field %d\n"
	       " -query                  preserve the query string for per-URL (-u*) statistics\n"
	       "\n"
	       "Output format - only one may be used at a time\n"
	       " -c    only report the number of lines that would have been printed\n"
	       " -pct  output connect and response times percentiles\n"
	       " -st   output number of requests per HTTP status code\n"
	       " -cc   output number of requests per cookie code (2 chars)\n"
	       " -tc   output number of requests per termination code (2 chars)\n"
	       " -srv  output statistics per server (time, requests, errors)\n"
	       " -ic   output statistics per ip count (time, requests, errors)\n"
	       " -u*   output statistics per URL (time, requests, errors)\n"
	       "       Additional characters indicate the output sorting key :\n"
	       "       -u : by URL, -uc : request count, -ue : error count\n"
	       "       -ua : average response time, -ut : average total time\n"
	       "       -uao, -uto: average times computed on valid ('OK') requests\n"
	       "       -uba, -ubt: average bytes returned, total bytes returned\n"
	       " -hdr  output captured header at the given <block>:<field>\n"
	       " -ac <count> -ad <delay>:\n"
	       "       Report periods corresponding to a grouped accept of <count> requests at\n"
	       "       the same millisecond after a delay of at least <ad> milliseconds with no\n"
	       "       incoming accept (used to spot network outages). Output format contains:\n"
	       "          <accept_date> <date_ms> <delta_ms from previous one> <nb_entries>\n",
	       (int)SOURCE_FIELD, (int)SOURCE_FIELD
	       );
	exit(0);
}


/* return pointer to first char not part of current field starting at <p>. */

#if defined(__i386__)
/* this one is always faster on 32-bits */
static inline const char *field_stop(const char *p)
{
	asm(
	    /* Look for spaces */
	    "4:                  \n\t"
	    "inc   %0            \n\t"
	    "cmpb  $0x20, -1(%0) \n\t"
	    "ja    4b            \n\t"
	    "jz    3f            \n\t"

	    /* we only get there for control chars 0..31. Leave if we find '\0' */
	    "cmpb  $0x0, -1(%0)  \n\t"
	    "jnz   4b            \n\t"

	    /* return %0-1 = position of the last char we checked */
	    "3:                  \n\t"
	    "dec   %0            \n\t"
	    : "=r" (p)
	    : "0" (p)
	    );
	return p;
}
#else
const char *field_stop(const char *p)
{
	unsigned char c;

	while (1) {
		c = *(p++);
		if (c > ' ')
			continue;
		if (c == ' ' || c == 0)
			break;
	}
	return p - 1;
}
#endif

/* return non-zero if the argument contains at least one zero byte. See principle above. */
static inline __attribute__((unused)) unsigned long long has_zero64(unsigned long long x)
{
	unsigned long long y;

	y = x - 0x0101010101010101ULL; /* generate a carry */
	y &= ~x;                       /* clear the bits that were already set */
	return y & 0x8080808080808080ULL;
}

/* return field <field> (starting from 1) in string <p>. Only consider
 * contiguous spaces (or tabs) as one delimiter. May return pointer to
 * last char if field is not found. Equivalent to awk '{print $field}'.
 */
const char *field_start(const char *p, int field)
{
#ifndef PREFER_ASM
	unsigned char c;
	while (1) {
		/* skip spaces */
		while (1) {
			c = *(p++);
			if (!c) /* end of line */
				return p-1;
			if (c == ' ')
				continue;
			/* other char => new field */
			break;
		}

		/* start of field */
		field--;
		if (!field)
			return p-1;

		/* skip this field */
		while (1) {
#if defined(HA_UNALIGNED_LE64)
			unsigned long long l = *(unsigned long long *)p;
			if (!has_zero64(l)) {
				l ^= 0x2020202020202020;
				l = has_zero64(l);
				if (!l) {
					p += 8;
					continue;
				}
				/* there is at least one space, find it and
				 * skip it now. The lowest byte in <l> with
				 * a 0x80 is the right one, but checking for
				 * it remains slower than testing each byte,
				 * probably due to the numerous short fields.
				 */
				while (*(p++) != ' ')
					;
				break;
			}
#endif
			c = *(p++);
			if (c == '\0')
				return p - 1;
			if (c == ' ')
				break;
		}
	}
#else
	/* This version works optimally on i386 and x86_64 but the code above
	 * shows similar performance. However, depending on the version of GCC
	 * used, inlining rules change and it may have difficulties to make
	 * efficient use of this code at other locations and could result in
	 * worse performance (eg: gcc 4.4). You may want to experience.
	 */
	asm(
	    /* skip spaces */
	    "1:                  \n\t"
	    "inc   %0            \n\t"
	    "cmpb  $0x20, -1(%0) \n\t"
	    "ja    2f            \n\t"
	    "jz    1b            \n\t"

	    /* we only get there for control chars 0..31. Leave if we find '\0' */
	    "cmpb  $0x0, -1(%0)  \n\t"
	    "jz    3f            \n\t"

	    /* start of field at [%0-1]. Check if we need to skip more fields */
	    "2:                  \n\t"
	    "dec   %1            \n\t"
	    "jz    3f            \n\t"

	    /* Look for spaces */
	    "4:                  \n\t"
	    "inc   %0            \n\t"
	    "cmpb  $0x20, -1(%0) \n\t"
	    "jz    1b            \n\t"
	    "ja    4b            \n\t"

	    /* we only get there for control chars 0..31. Leave if we find '\0' */
	    "cmpb  $0x0, -1(%0)  \n\t"
	    "jnz   4b            \n\t"

	    /* return %0-1 = position of the last char we checked */
	    "3:                  \n\t"
	    "dec   %0            \n\t"
	    : "=r" (p)
	    : "r" (field), "0" (p)
	    );
	return p;
#endif
}

/* keep only the <bits> higher bits of <i> */
static inline unsigned int quantify_u32(unsigned int i, int bits)
{
	int high;

	if (!bits)
		return 0;

	if (i)
		high = fls_auto(i);    // 1 to 32
	else
		high = 0;

	if (high <= bits)
		return i;

	return i & ~((1 << (high - bits)) - 1);
}

/* keep only the <bits> higher bits of the absolute value of <i>, as well as
 * its sign. */
static inline int quantify(int i, int bits)
{
	if (i >= 0)
		return quantify_u32(i, bits);
	else
		return -quantify_u32(-i, bits);
}

/* Insert timer value <v> into tree <r>. A pre-allocated node must be passed
 * in <alloc>. It may be NULL, in which case the function will allocate it
 * itself. It will be reset to NULL once consumed. The caller is responsible
 * for freeing the node once not used anymore. The node where the value was
 * inserted is returned.
 */
struct timer *insert_timer(struct eb_root *r, struct timer **alloc, int v)
{
	struct timer *t = *alloc;
	struct eb32_node *n;

	if (!t) {
		t = calloc(1, sizeof(*t));
		if (unlikely(!t)) {
			fprintf(stderr, "%s: not enough memory\n", __FUNCTION__);
			exit(1);
		}
	}
	t->node.key = quantify(v, QBITS);         // keep only the higher QBITS bits

	n = eb32i_insert(r, &t->node);
	if (n == &t->node)
		t = NULL;  /* node inserted, will malloc next time */

	*alloc = t;
	return container_of(n, struct timer, node);
}

/* Insert value value <v> into tree <r>. A pre-allocated node must be passed
 * in <alloc>. It may be NULL, in which case the function will allocate it
 * itself. It will be reset to NULL once consumed. The caller is responsible
 * for freeing the node once not used anymore. The node where the value was
 * inserted is returned.
 */
struct timer *insert_value(struct eb_root *r, struct timer **alloc, int v)
{
	struct timer *t = *alloc;
	struct eb32_node *n;

	if (!t) {
		t = calloc(1, sizeof(*t));
		if (unlikely(!t)) {
			fprintf(stderr, "%s: not enough memory\n", __FUNCTION__);
			exit(1);
		}
	}
	t->node.key = v;

	n = eb32i_insert(r, &t->node);
	if (n == &t->node)
		t = NULL;  /* node inserted, will malloc next time */

	*alloc = t;
	return container_of(n, struct timer, node);
}

int str2ic(const char *s)
{
	int i = 0;
	int j, k;

	if (*s != '-') {
		/* positive number */
		while (1) {
			j = (*s++) - '0';
			k = i * 10;
			if ((unsigned)j > 9)
				break;
			i = k + j;
		}
	} else {
		/* negative number */
		s++;
		while (1) {
			j = (*s++) - '0';
			k = i * 10;
			if ((unsigned)j > 9)
				break;
			i = k - j;
		}
	}

	return i;
}


/* Convert "[04/Dec/2008:09:49:40.555]" to an integer equivalent to the time of
 * the day in milliseconds. It returns -1 for all unparsable values. The parser
 * looks ugly but gcc emits far better code that way.
 */
int convert_date(const char *field)
{
	unsigned int h, m, s, ms;
	unsigned char c;
	const char *e;

	h = m = s = ms = 0;
	e = field;

	/* skip the date */
	while (1) {
		c = *(e++);
		if (c == ':')
			break;
		if (!c)
			goto out_err;
	}

	/* hour + ':' */
	while (1) {
		c = *(e++) - '0';
		if (c > 9)
			break;
		h = h * 10 + c;
	}
	if (c == (unsigned char)(0 - '0'))
		goto out_err;

	/* minute + ':' */
	while (1) {
		c = *(e++) - '0';
		if (c > 9)
			break;
		m = m * 10 + c;
	}
	if (c == (unsigned char)(0 - '0'))
		goto out_err;

	/* second + '.' or ']' */
	while (1) {
		c = *(e++) - '0';
		if (c > 9)
			break;
		s = s * 10 + c;
	}
	if (c == (unsigned char)(0 - '0'))
		goto out_err;

	/* if there's a '.', we have milliseconds */
	if (c == (unsigned char)('.' - '0')) {
		/* millisecond second + ']' */
		while (1) {
			c = *(e++) - '0';
			if (c > 9)
				break;
			ms = ms * 10 + c;
		}
		if (c == (unsigned char)(0 - '0'))
			goto out_err;
	}
	return (((h * 60) + m) * 60 + s) * 1000 + ms;
 out_err:
	return -1;
}

/* Convert "[04/Dec/2008:09:49:40.555]" to an unix timestamp.
 * It returns -1 for all unparsable values. The parser
 * looks ugly but gcc emits far better code that way.
 */
int convert_date_to_timestamp(const char *field)
{
	unsigned int d, mo, y, h, m, s;
	unsigned char c;
	const char *e;
	time_t rawtime;
	static struct tm * timeinfo;
	static int last_res;

	d = mo = y = h = m = s = 0;
	e = field;

	e++; // remove '['

	/* day + '/' */
	while (1) {
		c = *(e++) - '0';
		if (c > 9)
			break;
		d = d * 10 + c;
		if (c == (unsigned char)(0 - '0'))
			goto out_err;
	}

	/* month + '/' */
	c = *(e++);
	if (c =='F') {
		mo = 2;
		e = e+3;
	} else if (c =='S') {
		mo = 9;
		e = e+3;
	} else if (c =='O') {
		mo = 10;
		e = e+3;
	} else if (c =='N') {
		mo = 11;
		e = e+3;
	} else if (c == 'D') {
		mo = 12;
		e = e+3;
	} else if (c == 'A') {
		c = *(e++);
		if (c == 'p') {
			mo = 4;
			e = e+2;
		} else if (c == 'u') {
			mo = 8;
			e = e+2;
		} else
			goto out_err;
	} else if (c == 'J') {
		c = *(e++);
		if (c == 'a') {
			mo = 1;
			e = e+2;
		} else if (c == 'u') {
			c = *(e++);
			if (c == 'n') {
				mo = 6;
				e = e+1;
			} else if (c == 'l') {
				mo = 7;
				e++;
			}
		} else
			goto out_err;
	} else if (c == 'M') {
		e++;
		c = *(e++);
		if (c == 'r') {
			mo = 3;
			e = e+1;
		} else if (c == 'y') {
			mo = 5;
			e = e+1;
		} else
			goto out_err;
	} else
		goto out_err;

	/* year + ':' */
	while (1) {
		c = *(e++) - '0';
		if (c > 9)
			break;
		y = y * 10 + c;
		if (c == (unsigned char)(0 - '0'))
			goto out_err;
	}

	/* hour + ':' */
	while (1) {
		c = *(e++) - '0';
		if (c > 9)
			break;
		h = h * 10 + c;
	}
	if (c == (unsigned char)(0 - '0'))
		goto out_err;

	/* minute + ':' */
	while (1) {
		c = *(e++) - '0';
		if (c > 9)
			break;
		m = m * 10 + c;
	}
	if (c == (unsigned char)(0 - '0'))
		goto out_err;

	/* second + '.' or ']' */
	while (1) {
		c = *(e++) - '0';
		if (c > 9)
			break;
		s = s * 10 + c;
	}

	if (likely(timeinfo)) {
		if ((unsigned)timeinfo->tm_min == m &&
		    (unsigned)timeinfo->tm_hour == h &&
		    (unsigned)timeinfo->tm_mday == d &&
		    (unsigned)timeinfo->tm_mon == mo - 1 &&
		    (unsigned)timeinfo->tm_year == y - 1900)
			return last_res + s;
	}
	else {
		time(&rawtime);
		timeinfo = localtime(&rawtime);
	}

	timeinfo->tm_sec = 0;
	timeinfo->tm_min = m;
	timeinfo->tm_hour = h;
	timeinfo->tm_mday = d;
	timeinfo->tm_mon = mo - 1;
	timeinfo->tm_year = y - 1900;
	last_res = mktime(timeinfo);

	return last_res + s;
 out_err:
	return -1;
}

void truncated_line(int linenum, const char *line)
{
	if (!(filter & FILT_QUIET))
		fprintf(stderr, "Truncated line %d: %s\n", linenum, line);
}

int main(int argc, char **argv)
{
	const char *b, *p, *time_field, *accept_field, *source_field;
	const char *filter_term_code_name = NULL;
	const char *output_file = NULL;
	int f, last;
	struct timer *t = NULL;
	struct eb32_node *n;
	struct url_stat *ustat = NULL;
	int val, test;
	unsigned int uval;
	unsigned int filter_acc_delay = 0, filter_acc_count = 0;
	int filter_time_resp = 0;
	int filt_http_status_low = 0, filt_http_status_high = 0;
	unsigned int filt2_timestamp_low = 0, filt2_timestamp_high = 0;
	unsigned int filt2_capture_block = 0, filt2_capture_field = 0;
	int skip_fields = 1;

	void (*line_filter)(const char *accept_field, const char *time_field, struct timer **tptr) = NULL;

	argc--; argv++;
	while (argc > 0) {
		if (*argv[0] != '-')
			break;

		if (strcmp(argv[0], "-ad") == 0) {
			if (argc < 2) die("missing option for -ad\n");
			argc--; argv++;
			filter |= FILT_ACC_DELAY;
			filter_acc_delay = atol(*argv);
		}
		else if (strcmp(argv[0], "-ac") == 0) {
			if (argc < 2) die("missing option for -ac\n");
			argc--; argv++;
			filter |= FILT_ACC_COUNT;
			filter_acc_count = atol(*argv);
		}
		else if (strcmp(argv[0], "-rt") == 0) {
			if (argc < 2) die("missing option for -rt\n");
			argc--; argv++;
			filter |= FILT_TIME_RESP;
			filter_time_resp = atol(*argv);
		}
		else if (strcmp(argv[0], "-RT") == 0) {
			if (argc < 2) die("missing option for -RT\n");
			argc--; argv++;
			filter |= FILT_TIME_RESP | FILT_INVERT_TIME_RESP;
			filter_time_resp = atol(*argv);
		}
		else if (strcmp(argv[0], "-s") == 0) {
			if (argc < 2) die("missing option for -s\n");
			argc--; argv++;
			skip_fields = atol(*argv);
		}
		else if (strcmp(argv[0], "-m") == 0) {
			if (argc < 2) die("missing option for -m\n");
			argc--; argv++;
			lines_max = atol(*argv);
		}
		else if (strcmp(argv[0], "-e") == 0)
			filter |= FILT_ERRORS_ONLY;
		else if (strcmp(argv[0], "-E") == 0)
			filter |= FILT_ERRORS_ONLY | FILT_INVERT_ERRORS;
		else if (strcmp(argv[0], "-H") == 0)
			filter |= FILT_HTTP_ONLY;
		else if (strcmp(argv[0], "-Q") == 0)
			filter |= FILT_QUEUE_ONLY;
		else if (strcmp(argv[0], "-QS") == 0)
			filter |= FILT_QUEUE_SRV_ONLY;
		else if (strcmp(argv[0], "-c") == 0)
			filter |= FILT_COUNT_ONLY;
		else if (strcmp(argv[0], "-q") == 0)
			filter |= FILT_QUIET;
		else if (strcmp(argv[0], "-v") == 0)
			filter_invert = !filter_invert;
		else if (strcmp(argv[0], "-gt") == 0)
			filter |= FILT_GRAPH_TIMERS;
		else if (strcmp(argv[0], "-pct") == 0)
			filter |= FILT_PERCENTILE;
		else if (strcmp(argv[0], "-st") == 0)
			filter |= FILT_COUNT_STATUS;
		else if (strcmp(argv[0], "-srv") == 0)
			filter |= FILT_COUNT_SRV_STATUS;
		else if (strcmp(argv[0], "-cc") == 0)
			filter |= FILT_COUNT_COOK_CODES;
		else if (strcmp(argv[0], "-tc") == 0)
			filter |= FILT_COUNT_TERM_CODES;
		else if (strcmp(argv[0], "-tcn") == 0) {
			if (argc < 2) die("missing option for -tcn\n");
			argc--; argv++;
			filter |= FILT_TERM_CODE_NAME;
			filter_term_code_name = *argv;
		}
		else if (strcmp(argv[0], "-TCN") == 0) {
			if (argc < 2) die("missing option for -TCN\n");
			argc--; argv++;
			filter |= FILT_TERM_CODE_NAME | FILT_INVERT_TERM_CODE_NAME;
			filter_term_code_name = *argv;
		}
		else if (strcmp(argv[0], "-hs") == 0 || strcmp(argv[0], "-HS") == 0) {
			char *sep, *str;

			if (argc < 2) die("missing option for -hs/-HS ([min]:[max])\n");
			filter |= FILT_HTTP_STATUS;
			if (argv[0][1] == 'H')
				filter |= FILT_INVERT_HTTP_STATUS;

			argc--; argv++;
			str = *argv;
			sep = strchr(str, ':');  /* [min]:[max] */
			if (!sep)
				sep = str; /* make max point to min */
			else
				*sep++ = 0;
			filt_http_status_low = *str ? atol(str) : 0;
			filt_http_status_high = *sep ? atol(sep) : 65535;
		}
		else if (strcmp(argv[0], "-time") == 0) {
			char *sep, *str;

			if (argc < 2) die("missing option for -time ([min]:[max])\n");
			filter2 |= FILT2_TIMESTAMP;

			argc--; argv++;
			str = *argv;
			sep = strchr(str, ':');  /* [min]:[max] */
			filt2_timestamp_low = *str ? atol(str) : 0;
			if (!sep)
				filt2_timestamp_high = 0xFFFFFFFF;
			else
				filt2_timestamp_high = atol(++sep);
		}
		else if (strcmp(argv[0], "-u") == 0)
			filter |= FILT_COUNT_URL_ONLY;
		else if (strcmp(argv[0], "-uc") == 0)
			filter |= FILT_COUNT_URL_COUNT;
		else if (strcmp(argv[0], "-ue") == 0)
			filter |= FILT_COUNT_URL_ERR;
		else if (strcmp(argv[0], "-ua") == 0)
			filter |= FILT_COUNT_URL_TAVG;
		else if (strcmp(argv[0], "-ut") == 0)
			filter |= FILT_COUNT_URL_TTOT;
		else if (strcmp(argv[0], "-uao") == 0)
			filter |= FILT_COUNT_URL_TAVGO;
		else if (strcmp(argv[0], "-uto") == 0)
			filter |= FILT_COUNT_URL_TTOTO;
		else if (strcmp(argv[0], "-uba") == 0)
			filter |= FILT_COUNT_URL_BAVG;
		else if (strcmp(argv[0], "-ubt") == 0)
			filter |= FILT_COUNT_URL_BTOT;
		else if (strcmp(argv[0], "-query") == 0)
			filter2 |= FILT2_PRESERVE_QUERY;
		else if (strcmp(argv[0], "-ic") == 0)
			filter |= FILT_COUNT_IP_COUNT;
		else if (strcmp(argv[0], "-hdr") == 0) {
			char *sep, *str;

			if (argc < 2) die("missing option for -hdr (<block>:<field>)\n");
			filter2 |= FILT2_EXTRACT_CAPTURE;

			argc--; argv++;
			str = *argv;
			sep = strchr(str, ':');
			if (!sep)
				die("missing colon in -hdr (<block>:<field>)\n");
			else
				*sep++ = 0;

			filt2_capture_block = *str ? atol(str) : 1;
			filt2_capture_field = *sep ? atol(sep) : 1;

			if (filt2_capture_block < 1 || filt2_capture_field < 1)
				die("block and field must be at least 1 for -hdr (<block>:<field>)\n");
		}
		else if (strcmp(argv[0], "-o") == 0) {
			if (output_file)
				die("Fatal: output file name already specified.\n");
			if (argc < 2)
				die("Fatal: missing output file name.\n");
			output_file = argv[1];
		}
		else if (strcmp(argv[0], "-h") == 0 || strcmp(argv[0], "--help") == 0)
			help();
		argc--;
		argv++;
	}

	if (!filter && !filter2)
		die("No action specified.\n");

	if (filter & FILT_ACC_COUNT && !filter_acc_count)
		filter_acc_count=1;

	if (filter & FILT_ACC_DELAY && !filter_acc_delay)
		filter_acc_delay = 1;


	/* by default, all lines are printed */
	line_filter = filter_output_line;
	if (filter & (FILT_ACC_COUNT|FILT_ACC_DELAY))
		line_filter = filter_accept_holes;
	else if (filter & (FILT_GRAPH_TIMERS|FILT_PERCENTILE))
		line_filter = filter_graphs;
	else if (filter & FILT_COUNT_STATUS)
		line_filter = filter_count_status;
	else if (filter & FILT_COUNT_COOK_CODES)
		line_filter = filter_count_cook_codes;
	else if (filter & FILT_COUNT_TERM_CODES)
		line_filter = filter_count_term_codes;
	else if (filter & FILT_COUNT_SRV_STATUS)
		line_filter = filter_count_srv_status;
	else if (filter & FILT_COUNT_URL_ANY)
		line_filter = filter_count_url;
	else if (filter & FILT_COUNT_ONLY)
		line_filter = NULL;

#if defined(POSIX_FADV_SEQUENTIAL)
	/* around 20% performance improvement is observed on Linux with this
	 * on cold-cache. Surprisingly, WILLNEED is less performant. Don't
	 * use NOREUSE as it flushes the cache and prevents easy data
	 * manipulation on logs!
	 */
	posix_fadvise(0, 0, 0, POSIX_FADV_SEQUENTIAL);
#endif

	if (!line_filter && /* FILT_COUNT_ONLY ( see above), and no input filter (see below) */
	    !(filter & (FILT_HTTP_ONLY|FILT_TIME_RESP|FILT_ERRORS_ONLY|FILT_HTTP_STATUS|FILT_QUEUE_ONLY|FILT_QUEUE_SRV_ONLY|FILT_TERM_CODE_NAME)) &&
		!(filter2 & (FILT2_TIMESTAMP))) {
		/* read the whole file at once first, ignore it if inverted output */
		if (!filter_invert)
			while ((lines_max < 0 || lines_out < lines_max) && fgets2(stdin) != NULL)
				lines_out++;

		goto skip_filters;
	}

	while ((line = fgets2(stdin)) != NULL) {
		linenum++;
		time_field = NULL; accept_field = NULL;
		source_field = NULL;

		test = 1;

		/* for any line we process, we first ensure that there is a field
		 * looking like the accept date field (beginning with a '[').
		 */
		if (filter & FILT_COUNT_IP_COUNT) {
			/* we need the IP first */
			source_field = field_start(line, SOURCE_FIELD + skip_fields);
			accept_field = field_start(source_field, ACCEPT_FIELD - SOURCE_FIELD + 1);
		}
		else
			accept_field = field_start(line, ACCEPT_FIELD + skip_fields);

		if (unlikely(*accept_field != '[')) {
			parse_err++;
			continue;
		}

		/* the day of month field is begin 01 and 31 */
		if (accept_field[1] < '0' || accept_field[1] > '3') {
			parse_err++;
			continue;
		}

		if (filter2 & FILT2_TIMESTAMP) {
			uval = convert_date_to_timestamp(accept_field);
			test &= (uval>=filt2_timestamp_low && uval<=filt2_timestamp_high) ;
		}

		if (filter & FILT_HTTP_ONLY) {
			/* only report lines with at least 4 timers */
			if (!time_field) {
				time_field = field_start(accept_field, TIME_FIELD - ACCEPT_FIELD + 1);
				if (unlikely(!*time_field)) {
					truncated_line(linenum, line);
					continue;
				}
			}

			field_stop(time_field + 1);
			/* we have field TIME_FIELD in [time_field]..[e-1] */
			p = time_field;
			f = 0;
			while (!SEP(*p)) {
				if (++f == 4)
					break;
				SKIP_CHAR(p, '/');
			}
			test &= (f >= 4);
		}

		if (filter & FILT_TIME_RESP) {
			int tps;

			/* only report lines with response times larger than filter_time_resp */
			if (!time_field) {
				time_field = field_start(accept_field, TIME_FIELD - ACCEPT_FIELD + 1);
				if (unlikely(!*time_field)) {
					truncated_line(linenum, line);
					continue;
				}
			}

			field_stop(time_field + 1);
			/* we have field TIME_FIELD in [time_field]..[e-1], let's check only the response time */

			p = time_field;
			f = 0;
			while (!SEP(*p)) {
				tps = str2ic(p);
				if (tps < 0) {
					tps = -1;
				}
				if (++f == 4)
					break;
				SKIP_CHAR(p, '/');
			}

			if (unlikely(f < 4)) {
				parse_err++;
				continue;
			}

			test &= (tps >= filter_time_resp) ^ !!(filter & FILT_INVERT_TIME_RESP);
		}

		if (filter & (FILT_ERRORS_ONLY | FILT_HTTP_STATUS)) {
			/* Check both error codes (-1, 5xx) and status code ranges */
			if (time_field)
				b = field_start(time_field, STATUS_FIELD - TIME_FIELD + 1);
			else
				b = field_start(accept_field, STATUS_FIELD - ACCEPT_FIELD + 1);

			if (unlikely(!*b)) {
				truncated_line(linenum, line);
				continue;
			}

			val = str2ic(b);
			if (filter & FILT_ERRORS_ONLY)
				test &= (val < 0 || (val >= 500 && val <= 599)) ^ !!(filter & FILT_INVERT_ERRORS);

			if (filter & FILT_HTTP_STATUS)
				test &= (val >= filt_http_status_low && val <= filt_http_status_high) ^ !!(filter & FILT_INVERT_HTTP_STATUS);
		}

		if (filter & (FILT_QUEUE_ONLY|FILT_QUEUE_SRV_ONLY)) {
			/* Check if the server's queue is non-nul */
			if (time_field)
				b = field_start(time_field, QUEUE_LEN_FIELD - TIME_FIELD + 1);
			else
				b = field_start(accept_field, QUEUE_LEN_FIELD - ACCEPT_FIELD + 1);

			if (unlikely(!*b)) {
				truncated_line(linenum, line);
				continue;
			}

			if (*b == '0') {
				if (filter & FILT_QUEUE_SRV_ONLY) {
					test = 0;
				}
				else {
					do {
						b++;
						if (*b == '/') {
							b++;
							break;
						}
					} while (*b);
					test &= ((unsigned char)(*b - '1') < 9);
				}
			}
		}

		if (filter & FILT_TERM_CODE_NAME) {
			/* only report corresponding termination code name */
			if (time_field)
				b = field_start(time_field, TERM_CODES_FIELD - TIME_FIELD + 1);
			else
				b = field_start(accept_field, TERM_CODES_FIELD - ACCEPT_FIELD + 1);

			if (unlikely(!*b)) {
				truncated_line(linenum, line);
				continue;
			}

			test &= (b[0] == filter_term_code_name[0] && b[1] == filter_term_code_name[1]) ^ !!(filter & FILT_INVERT_TERM_CODE_NAME);
		}


		test ^= filter_invert;
		if (!test)
			continue;

		/************** here we process inputs *******************/

		if (line_filter) {
			if (filter & FILT_COUNT_IP_COUNT)
				filter_count_ip(source_field, accept_field, time_field, &t);
			else if (filter2 & FILT2_EXTRACT_CAPTURE)
				filter_extract_capture(accept_field, time_field, filt2_capture_block, filt2_capture_field);
			else
				line_filter(accept_field, time_field, &t);
		}
		else
			lines_out++; /* FILT_COUNT_ONLY was used, so we're just counting lines */
		if (lines_max >= 0 && lines_out >= lines_max)
			break;
	}

 skip_filters:
	/*****************************************************
	 * Here we've finished reading all input. Depending on the
	 * filters, we may still have some analysis to run on the
	 * collected data and to output data in a new format.
	 *************************************************** */

	if (t)
		free(t);

	if (filter & FILT_COUNT_ONLY) {
		printf("%d\n", lines_out);
		exit(0);
	}

	if (filter & (FILT_ACC_COUNT|FILT_ACC_DELAY)) {
		/* sort and count all timers. Output will look like this :
		 * <accept_date> <delta_ms from previous one> <nb entries>
		 */
		n = eb32_first(&timers[0]);

		if (n)
			last = n->key;
		while (n) {
			unsigned int d, h, m, s, ms;

			t = container_of(n, struct timer, node);
			h = n->key;
			d = h - last;
			last = h;

			if (d >= filter_acc_delay && t->count >= filter_acc_count) {
				ms = h % 1000; h = h / 1000;
				s = h % 60; h = h / 60;
				m = h % 60; h = h / 60;
				printf("%02u:%02u:%02u.%03u %d %u %u\n", h, m, s, ms, last, d, t->count);
				lines_out++;
				if (lines_max >= 0 && lines_out >= lines_max)
					break;
			}
			n = eb32_next(n);
		}
	}
	else if (filter & FILT_GRAPH_TIMERS) {
		/* sort all timers */
		for (f = 0; f < 5; f++) {
			struct eb32_node *n;

			n = eb32_first(&timers[f]);
			while (n) {
				int i;
				double d;
				int val;

				t = container_of(n, struct timer, node);
				last = n->key;
				val = t->count;

				i = (last < 0) ? -last : last;
				i = fls_auto(i) - QBITS;

				if (i > 0)
					d = val / (double)(1 << i);
				else
					d = val;

				if (d > 0.0)
					printf("%d %d %f\n", f, last, d+1.0);

				n = eb32_next(n);
			}
		}
	}
	else if (filter & FILT_PERCENTILE) {
		/* report timers by percentile :
		 *    <percent> <total> <max_req_time> <max_conn_time> <max_resp_time> <max_data_time>
		 * We don't count errs.
		 */
		struct eb32_node *n[5];
		unsigned long cum[5];
		double step;

		if (!lines_out)
			goto empty;

		for (f = 1; f < 5; f++) {
			n[f] = eb32_first(&timers[f]);
			cum[f] = container_of(n[f], struct timer, node)->count;
		}

		for (step = 1; step <= 1000;) {
			unsigned int thres = lines_out * (step / 1000.0);

			printf("%3.1f %u ", step/10.0, thres);
			for (f = 1; f < 5; f++) {
				struct eb32_node *next;
				while (cum[f] < thres) {
					/* need to find other keys */
					next = eb32_next(n[f]);
					if (!next)
						break;
					n[f] = next;
					cum[f] += container_of(next, struct timer, node)->count;
				}

				/* value still within $step % of total */
				printf("%d ", n[f]->key);
			}
			putchar('\n');
			if (step >= 100 && step < 900)
				step += 50;  // jump 5% by 5% between those steps.
			else if (step >= 20 && step < 980)
				step += 10;
			else
				step += 1;
		}
	}
	else if (filter & FILT_COUNT_STATUS) {
		/* output all statuses in the form of <status> <occurrences> */
		n = eb32_first(&timers[0]);
		while (n) {
			t = container_of(n, struct timer, node);
			printf("%d %u\n", n->key, t->count);
			lines_out++;
			if (lines_max >= 0 && lines_out >= lines_max)
				break;
			n = eb32_next(n);
		}
	}
	else if (filter & FILT_COUNT_SRV_STATUS) {
		struct ebmb_node *srv_node;
		struct srv_st *srv;

		printf("#srv_name 1xx 2xx 3xx 4xx 5xx other tot_req req_ok pct_ok avg_ct avg_rt\n");

		srv_node = ebmb_first(&timers[0]);
		while (srv_node) {
			int tot_rq;

			srv = container_of(srv_node, struct srv_st, node);

			tot_rq = 0;
			for (f = 0; f <= 5; f++)
				tot_rq += srv->st_cnt[f];

			printf("%s %u %u %u %u %u %u %d %u %.1f %d %d\n",
			       srv_node->key, srv->st_cnt[1], srv->st_cnt[2],
			       srv->st_cnt[3], srv->st_cnt[4], srv->st_cnt[5], srv->st_cnt[0],
			       tot_rq,
			       srv->nb_ok, (double)srv->nb_ok * 100.0 / (tot_rq?tot_rq:1),
			       (int)(srv->cum_ct / (srv->nb_ct?srv->nb_ct:1)), (int)(srv->cum_rt / (srv->nb_rt?srv->nb_rt:1)));
			srv_node = ebmb_next(srv_node);
			lines_out++;
			if (lines_max >= 0 && lines_out >= lines_max)
				break;
		}
	}
	else if (filter & (FILT_COUNT_TERM_CODES|FILT_COUNT_COOK_CODES)) {
		/* output all statuses in the form of <code> <occurrences> */
		n = eb32_first(&timers[0]);
		while (n) {
			t = container_of(n, struct timer, node);
			printf("%c%c %u\n", (n->key >> 8), (n->key) & 255, t->count);
			lines_out++;
			if (lines_max >= 0 && lines_out >= lines_max)
				break;
			n = eb32_next(n);
		}
	}
	else if (filter & (FILT_COUNT_URL_ANY|FILT_COUNT_IP_COUNT)) {
		struct eb_node *node, *next;

		if (!(filter & FILT_COUNT_URL_ONLY)) {
			/* we have to sort on another criterion. We'll use timers[1] for the
			 * destination tree.
			 */

			timers[1] = EB_ROOT; /* reconfigure to accept duplicates */
			for (node = eb_first(&timers[0]); node; node = next) {
				next = eb_next(node);
				eb_delete(node);

				ustat = container_of(node, struct url_stat, node.url.node);

				if (filter & (FILT_COUNT_URL_COUNT|FILT_COUNT_IP_COUNT))
					ustat->node.val.key = ustat->nb_req;
				else if (filter & FILT_COUNT_URL_ERR)
					ustat->node.val.key = ustat->nb_err;
				else if (filter & FILT_COUNT_URL_TTOT)
					ustat->node.val.key = ustat->total_time;
				else if (filter & FILT_COUNT_URL_TAVG)
					ustat->node.val.key = ustat->nb_req ? ustat->total_time / ustat->nb_req : 0;
				else if (filter & FILT_COUNT_URL_TTOTO)
					ustat->node.val.key = ustat->total_time_ok;
				else if (filter & FILT_COUNT_URL_TAVGO)
					ustat->node.val.key = (ustat->nb_req - ustat->nb_err) ? ustat->total_time_ok / (ustat->nb_req - ustat->nb_err) : 0;
				else if (filter & FILT_COUNT_URL_BAVG)
					ustat->node.val.key = ustat->nb_req ? ustat->total_bytes_sent / ustat->nb_req : 0;
				else if (filter & FILT_COUNT_URL_BTOT)
					ustat->node.val.key = ustat->total_bytes_sent;
				else
					ustat->node.val.key = 0;

				eb64_insert(&timers[1], &ustat->node.val);
			}
			/* switch trees */
			timers[0] = timers[1];
		}

		if (FILT_COUNT_IP_COUNT)
			printf("#req err ttot tavg oktot okavg bavg btot src\n");
		else
			printf("#req err ttot tavg oktot okavg bavg btot url\n");

		/* scan the tree in its reverse sorting order */
		node = eb_last(&timers[0]);
		while (node) {
			ustat = container_of(node, struct url_stat, node.url.node);
			printf("%u %u %llu %llu %llu %llu %llu %llu %s\n",
			       ustat->nb_req,
			       ustat->nb_err,
			       ustat->total_time,
			       ustat->nb_req ? ustat->total_time / ustat->nb_req : 0,
			       ustat->total_time_ok,
			       (ustat->nb_req - ustat->nb_err) ? ustat->total_time_ok / (ustat->nb_req - ustat->nb_err) : 0,
			       ustat->nb_req ? ustat->total_bytes_sent / ustat->nb_req : 0,
			       ustat->total_bytes_sent,
			       ustat->url);

			node = eb_prev(node);
			lines_out++;
			if (lines_max >= 0 && lines_out >= lines_max)
				break;
		}
	}

 empty:
	if (!(filter & FILT_QUIET))
		fprintf(stderr, "%d lines in, %d lines out, %d parsing errors\n",
			linenum, lines_out, parse_err);
	exit(0);
}

void filter_output_line(const char *accept_field, const char *time_field, struct timer **tptr)
{
	puts(line);
	lines_out++;
}

void filter_extract_capture(const char *accept_field, const char *time_field, unsigned int block, unsigned int field)
{
	const char *e, *f;

	if (time_field)
		e = field_start(time_field, METH_FIELD - TIME_FIELD + 1);
	else
		e = field_start(accept_field, METH_FIELD - ACCEPT_FIELD + 1);

	while (block-- > 0) {
		/* Scan until the start of a capture block ('{') until the URL ('"'). */
		while ((*e != '"' && *e != '{') && *e) {
			/* Note: some syslog servers escape quotes ! */
			if (*e == '\\' && e[1] == '"')
				break;

			e = field_start(e, 2);
		}

		if (unlikely(!*e)) {
			truncated_line(linenum, line);
			return;
		}

		/* We reached the URL, no more captures will follow. */
		if (*e != '{') {
			puts("");
			lines_out++;
			return;
		}

		/* e points the the opening brace of the capture block. */

		e++;
	}

	/* We are in the first field of the selected capture block. */

	while (--field > 0) {
		while ((*e != '|' && *e != '}') && *e)
			e++;

		if (unlikely(!*e)) {
			truncated_line(linenum, line);
			return;
		}

		if (*e != '|') {
			puts("");
			lines_out++;
			return;
		}

		/* e points to the pipe. */

		e++;
	}

	f = e;

	while ((*f != '|' && *f != '}') && *f)
		f++;

	if (unlikely(!*f)) {
		truncated_line(linenum, line);
		return;
	}

	fwrite(e, f - e, 1, stdout);
	putchar('\n');
	lines_out++;
}

void filter_accept_holes(const char *accept_field, const char *time_field, struct timer **tptr)
{
	struct timer *t2;
	int val;

	val = convert_date(accept_field);
	if (unlikely(val < 0)) {
		truncated_line(linenum, line);
		return;
	}

	t2 = insert_value(&timers[0], tptr, val);
	t2->count++;
	return;
}

void filter_count_status(const char *accept_field, const char *time_field, struct timer **tptr)
{
	struct timer *t2;
	const char *b;
	int val;

	if (time_field)
		b = field_start(time_field, STATUS_FIELD - TIME_FIELD + 1);
	else
		b = field_start(accept_field, STATUS_FIELD - ACCEPT_FIELD + 1);

	if (unlikely(!*b)) {
		truncated_line(linenum, line);
		return;
	}

	val = str2ic(b);

	t2 = insert_value(&timers[0], tptr, val);
	t2->count++;
}

void filter_count_cook_codes(const char *accept_field, const char *time_field, struct timer **tptr)
{
	struct timer *t2;
	const char *b;
	int val;

	if (time_field)
		b = field_start(time_field, TERM_CODES_FIELD - TIME_FIELD + 1);
	else
		b = field_start(accept_field, TERM_CODES_FIELD - ACCEPT_FIELD + 1);

	if (unlikely(!*b)) {
		truncated_line(linenum, line);
		return;
	}

	val = 256 * b[2] + b[3];

	t2 = insert_value(&timers[0], tptr, val);
	t2->count++;
}

void filter_count_term_codes(const char *accept_field, const char *time_field, struct timer **tptr)
{
	struct timer *t2;
	const char *b;
	int val;

	if (time_field)
		b = field_start(time_field, TERM_CODES_FIELD - TIME_FIELD + 1);
	else
		b = field_start(accept_field, TERM_CODES_FIELD - ACCEPT_FIELD + 1);

	if (unlikely(!*b)) {
		truncated_line(linenum, line);
		return;
	}

	val = 256 * b[0] + b[1];

	t2 = insert_value(&timers[0], tptr, val);
	t2->count++;
}

void filter_count_srv_status(const char *accept_field, const char *time_field, struct timer **tptr)
{
	const char *b, *e, *p;
	int f, err, array[5];
	struct ebmb_node *srv_node;
	struct srv_st *srv;
	int val;

	/* the server field is before the status field, so let's
	 * parse them in the proper order.
	 */
	b = field_start(accept_field, SERVER_FIELD - ACCEPT_FIELD + 1);
	if (unlikely(!*b)) {
		truncated_line(linenum, line);
		return;
	}

	e = field_stop(b + 1);  /* we have the server name in [b]..[e-1] */

	/* the chance that a server name already exists is extremely high,
	 * so let's perform a normal lookup first.
	 */
	srv_node = ebst_lookup_len(&timers[0], b, e - b);
	srv = container_of(srv_node, struct srv_st, node);

	if (!srv_node) {
		/* server not yet in the tree, let's create it */
		srv = (void *)calloc(1, sizeof(struct srv_st) + e - b + 1);
		srv_node = &srv->node;
		memcpy(&srv_node->key, b, e - b);
		srv_node->key[e - b] = '\0';
		ebst_insert(&timers[0], srv_node);
	}

	/* let's collect the connect and response times */
	if (!time_field) {
		time_field = field_start(e, TIME_FIELD - SERVER_FIELD);
		if (unlikely(!*time_field)) {
			truncated_line(linenum, line);
			return;
		}
	}

	e = field_stop(time_field + 1);
	/* we have field TIME_FIELD in [time_field]..[e-1] */

	p = time_field;
	err = 0;
	f = 0;
	while (!SEP(*p)) {
		array[f] = str2ic(p);
		if (array[f] < 0) {
			array[f] = -1;
			err = 1;
		}
		if (++f == 5)
			break;
		SKIP_CHAR(p, '/');
	}

	if (unlikely(f < 5)){
		parse_err++;
		return;
	}

	/* OK we have our timers in array[2,3] */
	if (!err)
		srv->nb_ok++;

	if (array[2] >= 0) {
		srv->cum_ct += array[2];
		srv->nb_ct++;
	}

	if (array[3] >= 0) {
		srv->cum_rt += array[3];
		srv->nb_rt++;
	}

	/* we're interested in the 5 HTTP status classes (1xx ... 5xx), and
	 * the invalid ones which will be reported as 0.
	 */
	b = field_start(e, STATUS_FIELD - TIME_FIELD);
	if (unlikely(!*b)) {
		truncated_line(linenum, line);
		return;
	}

	val = 0;
	if (*b >= '1' && *b <= '5')
		val = *b - '0';

	srv->st_cnt[val]++;
}

void filter_count_url(const char *accept_field, const char *time_field, struct timer **tptr)
{
	struct url_stat *ustat = NULL;
	struct ebpt_node *ebpt_old;
	const char *b, *e;
	int f, err, array[5];
	int val;

	/* let's collect the response time */
	if (!time_field) {
		time_field = field_start(accept_field, TIME_FIELD - ACCEPT_FIELD + 1);  // avg 115 ns per line
		if (unlikely(!*time_field)) {
			truncated_line(linenum, line);
			return;
		}
	}

	/* we have the field TIME_FIELD starting at <time_field>. We'll
	 * parse the 5 timers to detect errors, it takes avg 55 ns per line.
	 */
	e = time_field; err = 0; f = 0;
	while (!SEP(*e)) {
		array[f] = str2ic(e);
		if (array[f] < 0) {
			array[f] = -1;
			err = 1;
		}
		if (++f == 5)
			break;
		SKIP_CHAR(e, '/');
	}
	if (f < 5) {
		parse_err++;
		return;
	}

	/* OK we have our timers in array[3], and err is >0 if at
	 * least one -1 was seen. <e> points to the first char of
	 * the last timer. Let's prepare a new node with that.
	 */
	if (unlikely(!ustat))
		ustat = calloc(1, sizeof(*ustat));

	ustat->nb_err = err;
	ustat->nb_req = 1;

	/* use array[4] = total time in case of error */
	ustat->total_time = (array[3] >= 0) ? array[3] : array[4];
	ustat->total_time_ok = (array[3] >= 0) ? array[3] : 0;

	e = field_start(e, BYTES_SENT_FIELD - TIME_FIELD + 1);
	val = str2ic(e);
	ustat->total_bytes_sent = val;

	/* the line may be truncated because of a bad request or anything like this,
	 * without a method. Also, if it does not begin with an quote, let's skip to
	 * the next field because it's a capture. Let's fall back to the "method" itself
	 * if there's nothing else.
	 */
	e = field_start(e, METH_FIELD - BYTES_SENT_FIELD + 1);
	while (*e != '"' && *e) {
		/* Note: some syslog servers escape quotes ! */
		if (*e == '\\' && e[1] == '"')
			break;
		e = field_start(e, 2);
	}

	if (unlikely(!*e)) {
		truncated_line(linenum, line);
		free(ustat);
		return;
	}

	b = field_start(e, URL_FIELD - METH_FIELD + 1); // avg 40 ns per line
	if (!*b)
		b = e;

	/* stop at end of field or first ';' or '?', takes avg 64 ns per line */
	e = b;
	do {
		if (*e == ' '||
		    (!(filter2 & FILT2_PRESERVE_QUERY) && (*e == '?' || *e == ';'))) {
			*(char *)e = 0;
			break;
		}
		e++;
	} while (*e);

	/* now instead of copying the URL for a simple lookup, we'll link
	 * to it from the node we're trying to insert. If it returns a
	 * different value, it was already there. Otherwise we just have
	 * to dynamically realloc an entry using strdup().
	 */
	ustat->node.url.key = (char *)b;
	ebpt_old = ebis_insert(&timers[0], &ustat->node.url);

	if (ebpt_old != &ustat->node.url) {
		struct url_stat *ustat_old;
		/* node was already there, let's update previous one */
		ustat_old = container_of(ebpt_old, struct url_stat, node.url);
		ustat_old->nb_req ++;
		ustat_old->nb_err += ustat->nb_err;
		ustat_old->total_time += ustat->total_time;
		ustat_old->total_time_ok += ustat->total_time_ok;
		ustat_old->total_bytes_sent += ustat->total_bytes_sent;
	} else {
		ustat->url = ustat->node.url.key = strdup(ustat->node.url.key);
		ustat = NULL; /* node was used */
	}
}

void filter_count_ip(const char *source_field, const char *accept_field, const char *time_field, struct timer **tptr)
{
	struct url_stat *ustat = NULL;
	struct ebpt_node *ebpt_old;
	const char *b, *e;
	int f, err, array[5];
	int val;

	/* let's collect the response time */
	if (!time_field) {
		time_field = field_start(accept_field, TIME_FIELD - ACCEPT_FIELD + 1);  // avg 115 ns per line
		if (unlikely(!*time_field)) {
			truncated_line(linenum, line);
			return;
		}
	}

	/* we have the field TIME_FIELD starting at <time_field>. We'll
	 * parse the 5 timers to detect errors, it takes avg 55 ns per line.
	 */
	e = time_field; err = 0; f = 0;
	while (!SEP(*e)) {
		if (f == 0 || f == 4) {
			array[f] = str2ic(e);
			if (array[f] < 0) {
				array[f] = -1;
				err = 1;
			}
		}
		if (++f == 5)
			break;
		SKIP_CHAR(e, '/');
	}
	if (f < 5) {
		parse_err++;
		return;
	}

	/* OK we have our timers in array[0], and err is >0 if at
	 * least one -1 was seen. <e> points to the first char of
	 * the last timer. Let's prepare a new node with that.
	 */
	if (unlikely(!ustat))
		ustat = calloc(1, sizeof(*ustat));

	ustat->nb_err = err;
	ustat->nb_req = 1;

	/* use array[4] = total time in case of error */
	ustat->total_time = (array[0] >= 0) ? array[0] : array[4];
	ustat->total_time_ok = (array[0] >= 0) ? array[0] : 0;

	e = field_start(e, BYTES_SENT_FIELD - TIME_FIELD + 1);
	val = str2ic(e);
	ustat->total_bytes_sent = val;

	/* the source might be IPv4 or IPv6, so we always strip the port by
	 * removing the last colon.
	 */
	b = source_field;
	e = field_stop(b + 1);
	while (e > b && e[-1] != ':')
		e--;
	*(char *)(e - 1) = '\0';

	/* now instead of copying the src for a simple lookup, we'll link
	 * to it from the node we're trying to insert. If it returns a
	 * different value, it was already there. Otherwise we just have
	 * to dynamically realloc an entry using strdup(). We're using the
	 * <url> field of the node to store the source address.
	 */
	ustat->node.url.key = (char *)b;
	ebpt_old = ebis_insert(&timers[0], &ustat->node.url);

	if (ebpt_old != &ustat->node.url) {
		struct url_stat *ustat_old;
		/* node was already there, let's update previous one */
		ustat_old = container_of(ebpt_old, struct url_stat, node.url);
		ustat_old->nb_req ++;
		ustat_old->nb_err += ustat->nb_err;
		ustat_old->total_time += ustat->total_time;
		ustat_old->total_time_ok += ustat->total_time_ok;
		ustat_old->total_bytes_sent += ustat->total_bytes_sent;
	} else {
		ustat->url = ustat->node.url.key = strdup(ustat->node.url.key);
		ustat = NULL; /* node was used */
	}
}

void filter_graphs(const char *accept_field, const char *time_field, struct timer **tptr)
{
	struct timer *t2;
	const char *p;
	int f, err, array[5];

	if (!time_field) {
		time_field = field_start(accept_field, TIME_FIELD - ACCEPT_FIELD + 1);
		if (unlikely(!*time_field)) {
			truncated_line(linenum, line);
			return;
		}
	}

	field_stop(time_field + 1);
	/* we have field TIME_FIELD in [time_field]..[e-1] */

	p = time_field;
	err = 0;
	f = 0;
	while (!SEP(*p)) {
		array[f] = str2ic(p);
		if (array[f] < 0) {
			array[f] = -1;
			err = 1;
		}
		if (++f == 5)
			break;
		SKIP_CHAR(p, '/');
	}

	if (unlikely(f < 5)) {
		parse_err++;
		return;
	}

	/* if we find at least one negative time, we count one error
	 * with a time equal to the total session time. This will
	 * emphasize quantum timing effects associated to known
	 * timeouts. Note that on some buggy machines, it is possible
	 * that the total time is negative, hence the reason to reset
	 * it.
	 */

	if (filter & FILT_GRAPH_TIMERS) {
		if (err) {
			if (array[4] < 0)
				array[4] = -1;
			t2 = insert_timer(&timers[0], tptr, array[4]);  // total time
			t2->count++;
		} else {
			int v;

			t2 = insert_timer(&timers[1], tptr, array[0]); t2->count++;  // req
			t2 = insert_timer(&timers[2], tptr, array[2]); t2->count++;  // conn
			t2 = insert_timer(&timers[3], tptr, array[3]); t2->count++;  // resp

			v = array[4] - array[0] - array[1] - array[2] - array[3]; // data time
			if (v < 0 && !(filter & FILT_QUIET))
				fprintf(stderr, "ERR: %s (%d %d %d %d %d => %d)\n",
					line, array[0], array[1], array[2], array[3], array[4], v);
			t2 = insert_timer(&timers[4], tptr, v); t2->count++;
			lines_out++;
		}
	} else { /* percentile */
		if (err) {
			if (array[4] < 0)
				array[4] = -1;
			t2 = insert_value(&timers[0], tptr, array[4]);  // total time
			t2->count++;
		} else {
			int v;

			t2 = insert_value(&timers[1], tptr, array[0]); t2->count++;  // req
			t2 = insert_value(&timers[2], tptr, array[2]); t2->count++;  // conn
			t2 = insert_value(&timers[3], tptr, array[3]); t2->count++;  // resp

			v = array[4] - array[0] - array[1] - array[2] - array[3]; // data time
			if (v < 0 && !(filter & FILT_QUIET))
				fprintf(stderr, "ERR: %s (%d %d %d %d %d => %d)\n",
					line, array[0], array[1], array[2], array[3], array[4], v);
			t2 = insert_value(&timers[4], tptr, v); t2->count++;
			lines_out++;
		}
	}
}


/*
 * Local variables:
 *  c-indent-level: 8
 *  c-basic-offset: 8
 * End:
 */