summaryrefslogtreecommitdiffstats
path: root/src/textobject.c
blob: d77ec961abae767edda7ff19b3bfd1b591f5cfcf (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
/* vi:set ts=8 sts=4 sw=4 noet:
 *
 * VIM - Vi IMproved	by Bram Moolenaar
 *
 * Do ":help uganda"  in Vim to read copying and usage conditions.
 * Do ":help credits" in Vim to see a list of people who contributed.
 * See README.txt for an overview of the Vim source code.
 */

/*
 * textobject.c: functions for text objects
 */
#include "vim.h"

static int cls(void);
static int skip_chars(int, int);

/*
 * Find the start of the next sentence, searching in the direction specified
 * by the "dir" argument.  The cursor is positioned on the start of the next
 * sentence when found.  If the next sentence is found, return OK.  Return FAIL
 * otherwise.  See ":h sentence" for the precise definition of a "sentence"
 * text object.
 */
    int
findsent(int dir, long count)
{
    pos_T	pos, tpos;
    pos_T	prev_pos;
    int		c;
    int		(*func)(pos_T *);
    int		startlnum;
    int		noskip = FALSE;	    // do not skip blanks
    int		cpo_J;
    int		found_dot;

    pos = curwin->w_cursor;
    if (dir == FORWARD)
	func = incl;
    else
	func = decl;

    while (count--)
    {
	prev_pos = pos;

	/*
	 * if on an empty line, skip up to a non-empty line
	 */
	if (gchar_pos(&pos) == NUL)
	{
	    do
		if ((*func)(&pos) == -1)
		    break;
	    while (gchar_pos(&pos) == NUL);
	    if (dir == FORWARD)
		goto found;
	}
	/*
	 * if on the start of a paragraph or a section and searching forward,
	 * go to the next line
	 */
	else if (dir == FORWARD && pos.col == 0 &&
						startPS(pos.lnum, NUL, FALSE))
	{
	    if (pos.lnum == curbuf->b_ml.ml_line_count)
		return FAIL;
	    ++pos.lnum;
	    goto found;
	}
	else if (dir == BACKWARD)
	    decl(&pos);

	// go back to the previous non-white non-punctuation character
	found_dot = FALSE;
	while (c = gchar_pos(&pos), VIM_ISWHITE(c)
				|| vim_strchr((char_u *)".!?)]\"'", c) != NULL)
	{
	    tpos = pos;
	    if (decl(&tpos) == -1 || (LINEEMPTY(tpos.lnum) && dir == FORWARD))
		break;

	    if (found_dot)
		break;
	    if (vim_strchr((char_u *) ".!?", c) != NULL)
		found_dot = TRUE;

	    if (vim_strchr((char_u *) ")]\"'", c) != NULL
		&& vim_strchr((char_u *) ".!?)]\"'", gchar_pos(&tpos)) == NULL)
		break;

	    decl(&pos);
	}

	// remember the line where the search started
	startlnum = pos.lnum;
	cpo_J = vim_strchr(p_cpo, CPO_ENDOFSENT) != NULL;

	for (;;)		// find end of sentence
	{
	    c = gchar_pos(&pos);
	    if (c == NUL || (pos.col == 0 && startPS(pos.lnum, NUL, FALSE)))
	    {
		if (dir == BACKWARD && pos.lnum != startlnum)
		    ++pos.lnum;
		break;
	    }
	    if (c == '.' || c == '!' || c == '?')
	    {
		tpos = pos;
		do
		    if ((c = inc(&tpos)) == -1)
			break;
		while (vim_strchr((char_u *)")]\"'", c = gchar_pos(&tpos))
			!= NULL);
		if (c == -1  || (!cpo_J && (c == ' ' || c == '\t')) || c == NUL
		    || (cpo_J && (c == ' ' && inc(&tpos) >= 0
			      && gchar_pos(&tpos) == ' ')))
		{
		    pos = tpos;
		    if (gchar_pos(&pos) == NUL) // skip NUL at EOL
			inc(&pos);
		    break;
		}
	    }
	    if ((*func)(&pos) == -1)
	    {
		if (count)
		    return FAIL;
		noskip = TRUE;
		break;
	    }
	}
found:
	    // skip white space
	while (!noskip && ((c = gchar_pos(&pos)) == ' ' || c == '\t'))
	    if (incl(&pos) == -1)
		break;

	if (EQUAL_POS(prev_pos, pos))
	{
	    // didn't actually move, advance one character and try again
	    if ((*func)(&pos) == -1)
	    {
		if (count)
		    return FAIL;
		break;
	    }
	    ++count;
	}
    }

    setpcmark();
    curwin->w_cursor = pos;
    return OK;
}

/*
 * Find the next paragraph or section in direction 'dir'.
 * Paragraphs are currently supposed to be separated by empty lines.
 * If 'what' is NUL we go to the next paragraph.
 * If 'what' is '{' or '}' we go to the next section.
 * If 'both' is TRUE also stop at '}'.
 * Return TRUE if the next paragraph or section was found.
 */
    int
findpar(
    int		*pincl,	    // Return: TRUE if last char is to be included
    int		dir,
    long	count,
    int		what,
    int		both)
{
    linenr_T	curr;
    int		did_skip;   // TRUE after separating lines have been skipped
    int		first;	    // TRUE on first line
    int		posix = (vim_strchr(p_cpo, CPO_PARA) != NULL);
#ifdef FEAT_FOLDING
    linenr_T	fold_first;	// first line of a closed fold
    linenr_T	fold_last;	// last line of a closed fold
    int		fold_skipped;	// TRUE if a closed fold was skipped this
				// iteration
#endif

    curr = curwin->w_cursor.lnum;

    while (count--)
    {
	did_skip = FALSE;
	for (first = TRUE; ; first = FALSE)
	{
	    if (*ml_get(curr) != NUL)
		did_skip = TRUE;

#ifdef FEAT_FOLDING
	    // skip folded lines
	    fold_skipped = FALSE;
	    if (first && hasFolding(curr, &fold_first, &fold_last))
	    {
		curr = ((dir > 0) ? fold_last : fold_first) + dir;
		fold_skipped = TRUE;
	    }
#endif

	    // POSIX has its own ideas of what a paragraph boundary is and it
	    // doesn't match historical Vi: It also stops at a "{" in the
	    // first column and at an empty line.
	    if (!first && did_skip && (startPS(curr, what, both)
			   || (posix && what == NUL && *ml_get(curr) == '{')))
		break;

#ifdef FEAT_FOLDING
	    if (fold_skipped)
		curr -= dir;
#endif
	    if ((curr += dir) < 1 || curr > curbuf->b_ml.ml_line_count)
	    {
		if (count)
		    return FALSE;
		curr -= dir;
		break;
	    }
	}
    }
    setpcmark();
    if (both && *ml_get(curr) == '}')	// include line with '}'
	++curr;
    curwin->w_cursor.lnum = curr;
    if (curr == curbuf->b_ml.ml_line_count && what != '}')
    {
	char_u *line = ml_get(curr);

	// Put the cursor on the last character in the last line and make the
	// motion inclusive.
	if ((curwin->w_cursor.col = (colnr_T)STRLEN(line)) != 0)
	{
	    --curwin->w_cursor.col;
	    curwin->w_cursor.col -=
			     (*mb_head_off)(line, line + curwin->w_cursor.col);
	    *pincl = TRUE;
	}
    }
    else
	curwin->w_cursor.col = 0;
    return TRUE;
}

/*
 * check if the string 's' is a nroff macro that is in option 'opt'
 */
    static int
inmacro(char_u *opt, char_u *s)
{
    char_u	*macro;

    for (macro = opt; macro[0]; ++macro)
    {
	// Accept two characters in the option being equal to two characters
	// in the line.  A space in the option matches with a space in the
	// line or the line having ended.
	if (       (macro[0] == s[0]
		    || (macro[0] == ' '
			&& (s[0] == NUL || s[0] == ' ')))
		&& (macro[1] == s[1]
		    || ((macro[1] == NUL || macro[1] == ' ')
			&& (s[0] == NUL || s[1] == NUL || s[1] == ' '))))
	    break;
	++macro;
	if (macro[0] == NUL)
	    break;
    }
    return (macro[0] != NUL);
}

/*
 * startPS: return TRUE if line 'lnum' is the start of a section or paragraph.
 * If 'para' is '{' or '}' only check for sections.
 * If 'both' is TRUE also stop at '}'
 */
    int
startPS(linenr_T lnum, int para, int both)
{
    char_u	*s;

    s = ml_get(lnum);
    if (*s == para || *s == '\f' || (both && *s == '}'))
	return TRUE;
    if (*s == '.' && (inmacro(p_sections, s + 1) ||
					   (!para && inmacro(p_para, s + 1))))
	return TRUE;
    return FALSE;
}

/*
 * The following routines do the word searches performed by the 'w', 'W',
 * 'b', 'B', 'e', and 'E' commands.
 */

/*
 * To perform these searches, characters are placed into one of three
 * classes, and transitions between classes determine word boundaries.
 *
 * The classes are:
 *
 * 0 - white space
 * 1 - punctuation
 * 2 or higher - keyword characters (letters, digits and underscore)
 */

static int	cls_bigword;	// TRUE for "W", "B" or "E"

/*
 * cls() - returns the class of character at curwin->w_cursor
 *
 * If a 'W', 'B', or 'E' motion is being done (cls_bigword == TRUE), chars
 * from class 2 and higher are reported as class 1 since only white space
 * boundaries are of interest.
 */
    static int
cls(void)
{
    int	    c;

    c = gchar_cursor();
    if (c == ' ' || c == '\t' || c == NUL)
	return 0;
    if (enc_dbcs != 0 && c > 0xFF)
    {
	// If cls_bigword, report multi-byte chars as class 1.
	if (enc_dbcs == DBCS_KOR && cls_bigword)
	    return 1;

	// process code leading/trailing bytes
	return dbcs_class(((unsigned)c >> 8), (c & 0xFF));
    }
    if (enc_utf8)
    {
	c = utf_class(c);
	if (c != 0 && cls_bigword)
	    return 1;
	return c;
    }

    // If cls_bigword is TRUE, report all non-blanks as class 1.
    if (cls_bigword)
	return 1;

    if (vim_iswordc(c))
	return 2;
    return 1;
}


/*
 * fwd_word(count, type, eol) - move forward one word
 *
 * Returns FAIL if the cursor was already at the end of the file.
 * If eol is TRUE, last word stops at end of line (for operators).
 */
    int
fwd_word(
    long	count,
    int		bigword,    // "W", "E" or "B"
    int		eol)
{
    int		sclass;	    // starting class
    int		i;
    int		last_line;

    curwin->w_cursor.coladd = 0;
    cls_bigword = bigword;
    while (--count >= 0)
    {
#ifdef FEAT_FOLDING
	// When inside a range of folded lines, move to the last char of the
	// last line.
	if (hasFolding(curwin->w_cursor.lnum, NULL, &curwin->w_cursor.lnum))
	    coladvance((colnr_T)MAXCOL);
#endif
	sclass = cls();

	/*
	 * We always move at least one character, unless on the last
	 * character in the buffer.
	 */
	last_line = (curwin->w_cursor.lnum == curbuf->b_ml.ml_line_count);
	i = inc_cursor();
	if (i == -1 || (i >= 1 && last_line)) // started at last char in file
	    return FAIL;
	if (i >= 1 && eol && count == 0)      // started at last char in line
	    return OK;

	/*
	 * Go one char past end of current word (if any)
	 */
	if (sclass != 0)
	    while (cls() == sclass)
	    {
		i = inc_cursor();
		if (i == -1 || (i >= 1 && eol && count == 0))
		    return OK;
	    }

	/*
	 * go to next non-white
	 */
	while (cls() == 0)
	{
	    /*
	     * We'll stop if we land on a blank line
	     */
	    if (curwin->w_cursor.col == 0 && *ml_get_curline() == NUL)
		break;

	    i = inc_cursor();
	    if (i == -1 || (i >= 1 && eol && count == 0))
		return OK;
	}
    }
    return OK;
}

/*
 * bck_word() - move backward 'count' words
 *
 * If stop is TRUE and we are already on the start of a word, move one less.
 *
 * Returns FAIL if top of the file was reached.
 */
    int
bck_word(long count, int bigword, int stop)
{
    int		sclass;	    // starting class

    curwin->w_cursor.coladd = 0;
    cls_bigword = bigword;
    while (--count >= 0)
    {
#ifdef FEAT_FOLDING
	// When inside a range of folded lines, move to the first char of the
	// first line.
	if (hasFolding(curwin->w_cursor.lnum, &curwin->w_cursor.lnum, NULL))
	    curwin->w_cursor.col = 0;
#endif
	sclass = cls();
	if (dec_cursor() == -1)		// started at start of file
	    return FAIL;

	if (!stop || sclass == cls() || sclass == 0)
	{
	    /*
	     * Skip white space before the word.
	     * Stop on an empty line.
	     */
	    while (cls() == 0)
	    {
		if (curwin->w_cursor.col == 0
				      && LINEEMPTY(curwin->w_cursor.lnum))
		    goto finished;
		if (dec_cursor() == -1) // hit start of file, stop here
		    return OK;
	    }

	    /*
	     * Move backward to start of this word.
	     */
	    if (skip_chars(cls(), BACKWARD))
		return OK;
	}

	inc_cursor();			// overshot - forward one
finished:
	stop = FALSE;
    }
    return OK;
}

/*
 * end_word() - move to the end of the word
 *
 * There is an apparent bug in the 'e' motion of the real vi. At least on the
 * System V Release 3 version for the 80386. Unlike 'b' and 'w', the 'e'
 * motion crosses blank lines. When the real vi crosses a blank line in an
 * 'e' motion, the cursor is placed on the FIRST character of the next
 * non-blank line. The 'E' command, however, works correctly. Since this
 * appears to be a bug, I have not duplicated it here.
 *
 * Returns FAIL if end of the file was reached.
 *
 * If stop is TRUE and we are already on the end of a word, move one less.
 * If empty is TRUE stop on an empty line.
 */
    int
end_word(
    long	count,
    int		bigword,
    int		stop,
    int		empty)
{
    int		sclass;	    // starting class

    curwin->w_cursor.coladd = 0;
    cls_bigword = bigword;
    while (--count >= 0)
    {
#ifdef FEAT_FOLDING
	// When inside a range of folded lines, move to the last char of the
	// last line.
	if (hasFolding(curwin->w_cursor.lnum, NULL, &curwin->w_cursor.lnum))
	    coladvance((colnr_T)MAXCOL);
#endif
	sclass = cls();
	if (inc_cursor() == -1)
	    return FAIL;

	/*
	 * If we're in the middle of a word, we just have to move to the end
	 * of it.
	 */
	if (cls() == sclass && sclass != 0)
	{
	    /*
	     * Move forward to end of the current word
	     */
	    if (skip_chars(sclass, FORWARD))
		return FAIL;
	}
	else if (!stop || sclass == 0)
	{
	    /*
	     * We were at the end of a word. Go to the end of the next word.
	     * First skip white space, if 'empty' is TRUE, stop at empty line.
	     */
	    while (cls() == 0)
	    {
		if (empty && curwin->w_cursor.col == 0
					  && LINEEMPTY(curwin->w_cursor.lnum))
		    goto finished;
		if (inc_cursor() == -1)	    // hit end of file, stop here
		    return FAIL;
	    }

	    /*
	     * Move forward to the end of this word.
	     */
	    if (skip_chars(cls(), FORWARD))
		return FAIL;
	}
	dec_cursor();			// overshot - one char backward
finished:
	stop = FALSE;			// we move only one word less
    }
    return OK;
}

/*
 * Move back to the end of the word.
 *
 * Returns FAIL if start of the file was reached.
 */
    int
bckend_word(
    long	count,
    int		bigword,    // TRUE for "B"
    int		eol)	    // TRUE: stop at end of line.
{
    int		sclass;	    // starting class
    int		i;

    curwin->w_cursor.coladd = 0;
    cls_bigword = bigword;
    while (--count >= 0)
    {
	sclass = cls();
	if ((i = dec_cursor()) == -1)
	    return FAIL;
	if (eol && i == 1)
	    return OK;

	/*
	 * Move backward to before the start of this word.
	 */
	if (sclass != 0)
	{
	    while (cls() == sclass)
		if ((i = dec_cursor()) == -1 || (eol && i == 1))
		    return OK;
	}

	/*
	 * Move backward to end of the previous word
	 */
	while (cls() == 0)
	{
	    if (curwin->w_cursor.col == 0 && LINEEMPTY(curwin->w_cursor.lnum))
		break;
	    if ((i = dec_cursor()) == -1 || (eol && i == 1))
		return OK;
	}
    }
    return OK;
}

/*
 * Skip a row of characters of the same class.
 * Return TRUE when end-of-file reached, FALSE otherwise.
 */
    static int
skip_chars(int cclass, int dir)
{
    while (cls() == cclass)
	if ((dir == FORWARD ? inc_cursor() : dec_cursor()) == -1)
	    return TRUE;
    return FALSE;
}

/*
 * Go back to the start of the word or the start of white space
 */
    static void
back_in_line(void)
{
    int		sclass;		    // starting class

    sclass = cls();
    for (;;)
    {
	if (curwin->w_cursor.col == 0)	    // stop at start of line
	    break;
	dec_cursor();
	if (cls() != sclass)		    // stop at start of word
	{
	    inc_cursor();
	    break;
	}
    }
}

    static void
find_first_blank(pos_T *posp)
{
    int	    c;

    while (decl(posp) != -1)
    {
	c = gchar_pos(posp);
	if (!VIM_ISWHITE(c))
	{
	    incl(posp);
	    break;
	}
    }
}

/*
 * Skip count/2 sentences and count/2 separating white spaces.
 */
    static void
findsent_forward(
    long    count,
    int	    at_start_sent)	// cursor is at start of sentence
{
    while (count--)
    {
	findsent(FORWARD, 1L);
	if (at_start_sent)
	    find_first_blank(&curwin->w_cursor);
	if (count == 0 || at_start_sent)
	    decl(&curwin->w_cursor);
	at_start_sent = !at_start_sent;
    }
}

/*
 * Find word under cursor, cursor at end.
 * Used while an operator is pending, and in Visual mode.
 */
    int
current_word(
    oparg_T	*oap,
    long	count,
    int		include,	// TRUE: include word and white space
    int		bigword)	// FALSE == word, TRUE == WORD
{
    pos_T	start_pos;
    pos_T	pos;
    int		inclusive = TRUE;
    int		include_white = FALSE;

    cls_bigword = bigword;
    CLEAR_POS(&start_pos);

    // Correct cursor when 'selection' is exclusive
    if (VIsual_active && *p_sel == 'e' && LT_POS(VIsual, curwin->w_cursor))
	dec_cursor();

    /*
     * When Visual mode is not active, or when the VIsual area is only one
     * character, select the word and/or white space under the cursor.
     */
    if (!VIsual_active || EQUAL_POS(curwin->w_cursor, VIsual))
    {
	/*
	 * Go to start of current word or white space.
	 */
	back_in_line();
	start_pos = curwin->w_cursor;

	/*
	 * If the start is on white space, and white space should be included
	 * ("	word"), or start is not on white space, and white space should
	 * not be included ("word"), find end of word.
	 */
	if ((cls() == 0) == include)
	{
	    if (end_word(1L, bigword, TRUE, TRUE) == FAIL)
		return FAIL;
	}
	else
	{
	    /*
	     * If the start is not on white space, and white space should be
	     * included ("word	 "), or start is on white space and white
	     * space should not be included ("	 "), find start of word.
	     * If we end up in the first column of the next line (single char
	     * word) back up to end of the line.
	     */
	    fwd_word(1L, bigword, TRUE);
	    if (curwin->w_cursor.col == 0)
		decl(&curwin->w_cursor);
	    else
		oneleft();

	    if (include)
		include_white = TRUE;
	}

	if (VIsual_active)
	{
	    // should do something when inclusive == FALSE !
	    VIsual = start_pos;
	    redraw_curbuf_later(UPD_INVERTED);	// update the inversion
	}
	else
	{
	    oap->start = start_pos;
	    oap->motion_type = MCHAR;
	}
	--count;
    }

    /*
     * When count is still > 0, extend with more objects.
     */
    while (count > 0)
    {
	inclusive = TRUE;
	if (VIsual_active && LT_POS(curwin->w_cursor, VIsual))
	{
	    /*
	     * In Visual mode, with cursor at start: move cursor back.
	     */
	    if (decl(&curwin->w_cursor) == -1)
		return FAIL;
	    if (include != (cls() != 0))
	    {
		if (bck_word(1L, bigword, TRUE) == FAIL)
		    return FAIL;
	    }
	    else
	    {
		if (bckend_word(1L, bigword, TRUE) == FAIL)
		    return FAIL;
		(void)incl(&curwin->w_cursor);
	    }
	}
	else
	{
	    /*
	     * Move cursor forward one word and/or white area.
	     */
	    if (incl(&curwin->w_cursor) == -1)
		return FAIL;
	    if (include != (cls() == 0))
	    {
		if (fwd_word(1L, bigword, TRUE) == FAIL && count > 1)
		    return FAIL;
		/*
		 * If end is just past a new-line, we don't want to include
		 * the first character on the line.
		 * Put cursor on last char of white.
		 */
		if (oneleft() == FAIL)
		    inclusive = FALSE;
	    }
	    else
	    {
		if (end_word(1L, bigword, TRUE, TRUE) == FAIL)
		    return FAIL;
	    }
	}
	--count;
    }

    if (include_white && (cls() != 0
		 || (curwin->w_cursor.col == 0 && !inclusive)))
    {
	/*
	 * If we don't include white space at the end, move the start
	 * to include some white space there. This makes "daw" work
	 * better on the last word in a sentence (and "2daw" on last-but-one
	 * word).  Also when "2daw" deletes "word." at the end of the line
	 * (cursor is at start of next line).
	 * But don't delete white space at start of line (indent).
	 */
	pos = curwin->w_cursor;	// save cursor position
	curwin->w_cursor = start_pos;
	if (oneleft() == OK)
	{
	    back_in_line();
	    if (cls() == 0 && curwin->w_cursor.col > 0)
	    {
		if (VIsual_active)
		    VIsual = curwin->w_cursor;
		else
		    oap->start = curwin->w_cursor;
	    }
	}
	curwin->w_cursor = pos;	// put cursor back at end
    }

    if (VIsual_active)
    {
	if (*p_sel == 'e' && inclusive && LTOREQ_POS(VIsual, curwin->w_cursor))
	    inc_cursor();
	if (VIsual_mode == 'V')
	{
	    VIsual_mode = 'v';
	    redraw_cmdline = TRUE;		// show mode later
	}
    }
    else
	oap->inclusive = inclusive;

    return OK;
}

/*
 * Find sentence(s) under the cursor, cursor at end.
 * When Visual active, extend it by one or more sentences.
 */
    int
current_sent(oparg_T *oap, long count, int include)
{
    pos_T	start_pos;
    pos_T	pos;
    int		start_blank;
    int		c;
    int		at_start_sent;
    long	ncount;

    start_pos = curwin->w_cursor;
    pos = start_pos;
    findsent(FORWARD, 1L);	// Find start of next sentence.

    /*
     * When the Visual area is bigger than one character: Extend it.
     */
    if (VIsual_active && !EQUAL_POS(start_pos, VIsual))
    {
extend:
	if (LT_POS(start_pos, VIsual))
	{
	    /*
	     * Cursor at start of Visual area.
	     * Find out where we are:
	     * - in the white space before a sentence
	     * - in a sentence or just after it
	     * - at the start of a sentence
	     */
	    at_start_sent = TRUE;
	    decl(&pos);
	    while (LT_POS(pos, curwin->w_cursor))
	    {
		c = gchar_pos(&pos);
		if (!VIM_ISWHITE(c))
		{
		    at_start_sent = FALSE;
		    break;
		}
		incl(&pos);
	    }
	    if (!at_start_sent)
	    {
		findsent(BACKWARD, 1L);
		if (EQUAL_POS(curwin->w_cursor, start_pos))
		    at_start_sent = TRUE;  // exactly at start of sentence
		else
		    // inside a sentence, go to its end (start of next)
		    findsent(FORWARD, 1L);
	    }
	    if (include)	// "as" gets twice as much as "is"
		count *= 2;
	    while (count--)
	    {
		if (at_start_sent)
		    find_first_blank(&curwin->w_cursor);
		c = gchar_cursor();
		if (!at_start_sent || (!include && !VIM_ISWHITE(c)))
		    findsent(BACKWARD, 1L);
		at_start_sent = !at_start_sent;
	    }
	}
	else
	{
	    /*
	     * Cursor at end of Visual area.
	     * Find out where we are:
	     * - just before a sentence
	     * - just before or in the white space before a sentence
	     * - in a sentence
	     */
	    incl(&pos);
	    at_start_sent = TRUE;
	    // not just before a sentence
	    if (!EQUAL_POS(pos, curwin->w_cursor))
	    {
		at_start_sent = FALSE;
		while (LT_POS(pos, curwin->w_cursor))
		{
		    c = gchar_pos(&pos);
		    if (!VIM_ISWHITE(c))
		    {
			at_start_sent = TRUE;
			break;
		    }
		    incl(&pos);
		}
		if (at_start_sent)	// in the sentence
		    findsent(BACKWARD, 1L);
		else		// in/before white before a sentence
		    curwin->w_cursor = start_pos;
	    }

	    if (include)	// "as" gets twice as much as "is"
		count *= 2;
	    findsent_forward(count, at_start_sent);
	    if (*p_sel == 'e')
		++curwin->w_cursor.col;
	}
	return OK;
    }

    /*
     * If the cursor started on a blank, check if it is just before the start
     * of the next sentence.
     */
    while (c = gchar_pos(&pos), VIM_ISWHITE(c))	// VIM_ISWHITE() is a macro
	incl(&pos);
    if (EQUAL_POS(pos, curwin->w_cursor))
    {
	start_blank = TRUE;
	find_first_blank(&start_pos);	// go back to first blank
    }
    else
    {
	start_blank = FALSE;
	findsent(BACKWARD, 1L);
	start_pos = curwin->w_cursor;
    }
    if (include)
	ncount = count * 2;
    else
    {
	ncount = count;
	if (start_blank)
	    --ncount;
    }
    if (ncount > 0)
	findsent_forward(ncount, TRUE);
    else
	decl(&curwin->w_cursor);

    if (include)
    {
	/*
	 * If the blank in front of the sentence is included, exclude the
	 * blanks at the end of the sentence, go back to the first blank.
	 * If there are no trailing blanks, try to include leading blanks.
	 */
	if (start_blank)
	{
	    find_first_blank(&curwin->w_cursor);
	    c = gchar_pos(&curwin->w_cursor);	// VIM_ISWHITE() is a macro
	    if (VIM_ISWHITE(c))
		decl(&curwin->w_cursor);
	}
	else if (c = gchar_cursor(), !VIM_ISWHITE(c))
	    find_first_blank(&start_pos);
    }

    if (VIsual_active)
    {
	// Avoid getting stuck with "is" on a single space before a sentence.
	if (EQUAL_POS(start_pos, curwin->w_cursor))
	    goto extend;
	if (*p_sel == 'e')
	    ++curwin->w_cursor.col;
	VIsual = start_pos;
	VIsual_mode = 'v';
	redraw_cmdline = TRUE;		// show mode later
	redraw_curbuf_later(UPD_INVERTED);	// update the inversion
    }
    else
    {
	// include a newline after the sentence, if there is one
	if (incl(&curwin->w_cursor) == -1)
	    oap->inclusive = TRUE;
	else
	    oap->inclusive = FALSE;
	oap->start = start_pos;
	oap->motion_type = MCHAR;
    }
    return OK;
}

/*
 * Find block under the cursor, cursor at end.
 * "what" and "other" are two matching parenthesis/brace/etc.
 */
    int
current_block(
    oparg_T	*oap,
    long	count,
    int		include,	// TRUE == include white space
    int		what,		// '(', '{', etc.
    int		other)		// ')', '}', etc.
{
    pos_T	old_pos;
    pos_T	*pos = NULL;
    pos_T	start_pos;
    pos_T	*end_pos;
    pos_T	old_start, old_end;
    char_u	*save_cpo;
    int		sol = FALSE;		// '{' at start of line

    old_pos = curwin->w_cursor;
    old_end = curwin->w_cursor;		// remember where we started
    old_start = old_end;

    /*
     * If we start on '(', '{', ')', '}', etc., use the whole block inclusive.
     */
    if (!VIsual_active || EQUAL_POS(VIsual, curwin->w_cursor))
    {
	setpcmark();
	if (what == '{')		// ignore indent
	    while (inindent(1))
		if (inc_cursor() != 0)
		    break;
	if (gchar_cursor() == what)
	    // cursor on '(' or '{', move cursor just after it
	    ++curwin->w_cursor.col;
    }
    else if (LT_POS(VIsual, curwin->w_cursor))
    {
	old_start = VIsual;
	curwin->w_cursor = VIsual;	    // cursor at low end of Visual
    }
    else
	old_end = VIsual;

    /*
     * Search backwards for unclosed '(', '{', etc..
     * Put this position in start_pos.
     * Ignore quotes here.  Keep the "M" flag in 'cpo', as that is what the
     * user wants.
     */
    save_cpo = p_cpo;
    p_cpo = (char_u *)(vim_strchr(p_cpo, CPO_MATCHBSL) != NULL ? "%M" : "%");
    if ((pos = findmatch(NULL, what)) != NULL)
    {
	while (count-- > 0)
	{
	    if ((pos = findmatch(NULL, what)) == NULL)
		break;
	    curwin->w_cursor = *pos;
	    start_pos = *pos;   // the findmatch for end_pos will overwrite *pos
	}
    }
    else
    {
	while (count-- > 0)
	{
	    if ((pos = findmatchlimit(NULL, what, FM_FORWARD, 0)) == NULL)
		break;
	    curwin->w_cursor = *pos;
	    start_pos = *pos;   // the findmatch for end_pos will overwrite *pos
	}
    }
    p_cpo = save_cpo;

    /*
     * Search for matching ')', '}', etc.
     * Put this position in curwin->w_cursor.
     */
    if (pos == NULL || (end_pos = findmatch(NULL, other)) == NULL)
    {
	curwin->w_cursor = old_pos;
	return FAIL;
    }
    curwin->w_cursor = *end_pos;

    /*
     * Try to exclude the '(', '{', ')', '}', etc. when "include" is FALSE.
     * If the ending '}', ')' or ']' is only preceded by indent, skip that
     * indent.  But only if the resulting area is not smaller than what we
     * started with.
     */
    while (!include)
    {
	incl(&start_pos);
	sol = (curwin->w_cursor.col == 0);
	decl(&curwin->w_cursor);
	while (inindent(1))
	{
	    sol = TRUE;
	    if (decl(&curwin->w_cursor) != 0)
		break;
	}

	/*
	 * In Visual mode, when the resulting area is not bigger than what we
	 * started with, extend it to the next block, and then exclude again.
	 * Don't try to expand the area if the area is empty.
	 */
	if (!LT_POS(start_pos, old_start) && !LT_POS(old_end, curwin->w_cursor)
		&& !EQUAL_POS(start_pos, curwin->w_cursor)
		&& VIsual_active)
	{
	    curwin->w_cursor = old_start;
	    decl(&curwin->w_cursor);
	    if ((pos = findmatch(NULL, what)) == NULL)
	    {
		curwin->w_cursor = old_pos;
		return FAIL;
	    }
	    start_pos = *pos;
	    curwin->w_cursor = *pos;
	    if ((end_pos = findmatch(NULL, other)) == NULL)
	    {
		curwin->w_cursor = old_pos;
		return FAIL;
	    }
	    curwin->w_cursor = *end_pos;
	}
	else
	    break;
    }

    if (VIsual_active)
    {
	if (*p_sel == 'e')
	    inc(&curwin->w_cursor);
	if (sol && gchar_cursor() != NUL)
	    inc(&curwin->w_cursor);	// include the line break
	VIsual = start_pos;
	VIsual_mode = 'v';
	redraw_curbuf_later(UPD_INVERTED);	// update the inversion
	showmode();
    }
    else
    {
	oap->start = start_pos;
	oap->motion_type = MCHAR;
	oap->inclusive = FALSE;
	if (sol)
	    incl(&curwin->w_cursor);
	else if (LTOREQ_POS(start_pos, curwin->w_cursor))
	    // Include the character under the cursor.
	    oap->inclusive = TRUE;
	else
	    // End is before the start (no text in between <>, [], etc.): don't
	    // operate on any text.
	    curwin->w_cursor = start_pos;
    }

    return OK;
}

#if defined(FEAT_EVAL) || defined(PROTO)
/*
 * Return TRUE if the cursor is on a "<aaa>" tag.  Ignore "<aaa/>".
 * When "end_tag" is TRUE return TRUE if the cursor is on "</aaa>".
 */
    static int
in_html_tag(
    int		end_tag)
{
    char_u	*line = ml_get_curline();
    char_u	*p;
    int		c;
    int		lc = NUL;
    pos_T	pos;

    if (enc_dbcs)
    {
	char_u	*lp = NULL;

	// We search forward until the cursor, because searching backwards is
	// very slow for DBCS encodings.
	for (p = line; p < line + curwin->w_cursor.col; MB_PTR_ADV(p))
	    if (*p == '>' || *p == '<')
	    {
		lc = *p;
		lp = p;
	    }
	if (*p != '<')	    // check for '<' under cursor
	{
	    if (lc != '<')
		return FALSE;
	    p = lp;
	}
    }
    else
    {
	for (p = line + curwin->w_cursor.col; p > line; )
	{
	    if (*p == '<')	// find '<' under/before cursor
		break;
	    MB_PTR_BACK(line, p);
	    if (*p == '>')	// find '>' before cursor
		break;
	}
	if (*p != '<')
	    return FALSE;
    }

    pos.lnum = curwin->w_cursor.lnum;
    pos.col = (colnr_T)(p - line);

    MB_PTR_ADV(p);
    if (end_tag)
	// check that there is a '/' after the '<'
	return *p == '/';

    // check that there is no '/' after the '<'
    if (*p == '/')
	return FALSE;

    // check that the matching '>' is not preceded by '/'
    for (;;)
    {
	if (inc(&pos) < 0)
	    return FALSE;
	c = *ml_get_pos(&pos);
	if (c == '>')
	    break;
	lc = c;
    }
    return lc != '/';
}

/*
 * Find tag block under the cursor, cursor at end.
 */
    int
current_tagblock(
    oparg_T	*oap,
    long	count_arg,
    int		include)	// TRUE == include white space
{
    long	count = count_arg;
    long	n;
    pos_T	old_pos;
    pos_T	start_pos;
    pos_T	end_pos;
    pos_T	old_start, old_end;
    char_u	*spat, *epat;
    char_u	*p;
    char_u	*cp;
    int		len;
    int		r;
    int		do_include = include;
    int		save_p_ws = p_ws;
    int		retval = FAIL;
    int		is_inclusive = TRUE;

    p_ws = FALSE;

    old_pos = curwin->w_cursor;
    old_end = curwin->w_cursor;		    // remember where we started
    old_start = old_end;
    if (!VIsual_active || *p_sel == 'e')
	decl(&old_end);			    // old_end is inclusive

    /*
     * If we start on "<aaa>" select that block.
     */
    if (!VIsual_active || EQUAL_POS(VIsual, curwin->w_cursor))
    {
	setpcmark();

	// ignore indent
	while (inindent(1))
	    if (inc_cursor() != 0)
		break;

	if (in_html_tag(FALSE))
	{
	    // cursor on start tag, move to its '>'
	    while (*ml_get_cursor() != '>')
		if (inc_cursor() < 0)
		    break;
	}
	else if (in_html_tag(TRUE))
	{
	    // cursor on end tag, move to just before it
	    while (*ml_get_cursor() != '<')
		if (dec_cursor() < 0)
		    break;
	    dec_cursor();
	    old_end = curwin->w_cursor;
	}
    }
    else if (LT_POS(VIsual, curwin->w_cursor))
    {
	old_start = VIsual;
	curwin->w_cursor = VIsual;	    // cursor at low end of Visual
    }
    else
	old_end = VIsual;

again:
    /*
     * Search backwards for unclosed "<aaa>".
     * Put this position in start_pos.
     */
    for (n = 0; n < count; ++n)
    {
	if (do_searchpair((char_u *)"<[^ \t>/!]\\+\\%(\\_s\\_[^>]\\{-}[^/]>\\|$\\|\\_s\\=>\\)",
		    (char_u *)"",
		    (char_u *)"</[^>]*>", BACKWARD, NULL, 0,
						  NULL, (linenr_T)0, 0L) <= 0)
	{
	    curwin->w_cursor = old_pos;
	    goto theend;
	}
    }
    start_pos = curwin->w_cursor;

    /*
     * Search for matching "</aaa>".  First isolate the "aaa".
     */
    inc_cursor();
    p = ml_get_cursor();
    for (cp = p; *cp != NUL && *cp != '>' && !VIM_ISWHITE(*cp); MB_PTR_ADV(cp))
	;
    len = (int)(cp - p);
    if (len == 0)
    {
	curwin->w_cursor = old_pos;
	goto theend;
    }
    spat = alloc(len + 39);
    epat = alloc(len + 9);
    if (spat == NULL || epat == NULL)
    {
	vim_free(spat);
	vim_free(epat);
	curwin->w_cursor = old_pos;
	goto theend;
    }
    sprintf((char *)spat, "<%.*s\\>\\%%(\\_s\\_[^>]\\{-}\\_[^/]>\\|\\_s\\?>\\)\\c", len, p);
    sprintf((char *)epat, "</%.*s>\\c", len, p);

    r = do_searchpair(spat, (char_u *)"", epat, FORWARD, NULL,
						    0, NULL, (linenr_T)0, 0L);

    vim_free(spat);
    vim_free(epat);

    if (r < 1 || LT_POS(curwin->w_cursor, old_end))
    {
	// Can't find other end or it's before the previous end.  Could be a
	// HTML tag that doesn't have a matching end.  Search backwards for
	// another starting tag.
	count = 1;
	curwin->w_cursor = start_pos;
	goto again;
    }

    if (do_include)
    {
	// Include up to the '>'.
	while (*ml_get_cursor() != '>')
	    if (inc_cursor() < 0)
		break;
    }
    else
    {
	char_u *c = ml_get_cursor();

	// Exclude the '<' of the end tag.
	// If the closing tag is on new line, do not decrement cursor, but
	// make operation exclusive, so that the linefeed will be selected
	if (*c == '<' && !VIsual_active && curwin->w_cursor.col == 0)
	    // do not decrement cursor
	    is_inclusive = FALSE;
	else if (*c == '<')
	    dec_cursor();
    }
    end_pos = curwin->w_cursor;

    if (!do_include)
    {
	// Exclude the start tag.
	curwin->w_cursor = start_pos;
	while (inc_cursor() >= 0)
	    if (*ml_get_cursor() == '>')
	    {
		inc_cursor();
		start_pos = curwin->w_cursor;
		break;
	    }
	curwin->w_cursor = end_pos;

	// If we are in Visual mode and now have the same text as before set
	// "do_include" and try again.
	if (VIsual_active && EQUAL_POS(start_pos, old_start)
						&& EQUAL_POS(end_pos, old_end))
	{
	    do_include = TRUE;
	    curwin->w_cursor = old_start;
	    count = count_arg;
	    goto again;
	}
    }

    if (VIsual_active)
    {
	// If the end is before the start there is no text between tags, select
	// the char under the cursor.
	if (LT_POS(end_pos, start_pos))
	    curwin->w_cursor = start_pos;
	else if (*p_sel == 'e')
	    inc_cursor();
	VIsual = start_pos;
	VIsual_mode = 'v';
	redraw_curbuf_later(UPD_INVERTED);	// update the inversion
	showmode();
    }
    else
    {
	oap->start = start_pos;
	oap->motion_type = MCHAR;
	if (LT_POS(end_pos, start_pos))
	{
	    // End is before the start: there is no text between tags; operate
	    // on an empty area.
	    curwin->w_cursor = start_pos;
	    oap->inclusive = FALSE;
	}
	else
	    oap->inclusive = is_inclusive;
    }
    retval = OK;

theend:
    p_ws = save_p_ws;
    return retval;
}
#endif

    int
current_par(
    oparg_T	*oap,
    long	count,
    int		include,	// TRUE == include white space
    int		type)		// 'p' for paragraph, 'S' for section
{
    linenr_T	start_lnum;
    linenr_T	end_lnum;
    int		white_in_front;
    int		dir;
    int		start_is_white;
    int		prev_start_is_white;
    int		retval = OK;
    int		do_white = FALSE;
    int		t;
    int		i;

    if (type == 'S')	    // not implemented yet
	return FAIL;

    start_lnum = curwin->w_cursor.lnum;

    /*
     * When visual area is more than one line: extend it.
     */
    if (VIsual_active && start_lnum != VIsual.lnum)
    {
extend:
	if (start_lnum < VIsual.lnum)
	    dir = BACKWARD;
	else
	    dir = FORWARD;
	for (i = count; --i >= 0; )
	{
	    if (start_lnum ==
			   (dir == BACKWARD ? 1 : curbuf->b_ml.ml_line_count))
	    {
		retval = FAIL;
		break;
	    }

	    prev_start_is_white = -1;
	    for (t = 0; t < 2; ++t)
	    {
		start_lnum += dir;
		start_is_white = linewhite(start_lnum);
		if (prev_start_is_white == start_is_white)
		{
		    start_lnum -= dir;
		    break;
		}
		for (;;)
		{
		    if (start_lnum == (dir == BACKWARD
					    ? 1 : curbuf->b_ml.ml_line_count))
			break;
		    if (start_is_white != linewhite(start_lnum + dir)
			    || (!start_is_white
				    && startPS(start_lnum + (dir > 0
							     ? 1 : 0), 0, 0)))
			break;
		    start_lnum += dir;
		}
		if (!include)
		    break;
		if (start_lnum == (dir == BACKWARD
					    ? 1 : curbuf->b_ml.ml_line_count))
		    break;
		prev_start_is_white = start_is_white;
	    }
	}
	curwin->w_cursor.lnum = start_lnum;
	curwin->w_cursor.col = 0;
	return retval;
    }

    /*
     * First move back to the start_lnum of the paragraph or white lines
     */
    white_in_front = linewhite(start_lnum);
    while (start_lnum > 1)
    {
	if (white_in_front)	    // stop at first white line
	{
	    if (!linewhite(start_lnum - 1))
		break;
	}
	else		// stop at first non-white line of start of paragraph
	{
	    if (linewhite(start_lnum - 1) || startPS(start_lnum, 0, 0))
		break;
	}
	--start_lnum;
    }

    /*
     * Move past the end of any white lines.
     */
    end_lnum = start_lnum;
    while (end_lnum <= curbuf->b_ml.ml_line_count && linewhite(end_lnum))
	++end_lnum;

    --end_lnum;
    i = count;
    if (!include && white_in_front)
	--i;
    while (i--)
    {
	if (end_lnum == curbuf->b_ml.ml_line_count)
	    return FAIL;

	if (!include)
	    do_white = linewhite(end_lnum + 1);

	if (include || !do_white)
	{
	    ++end_lnum;
	    /*
	     * skip to end of paragraph
	     */
	    while (end_lnum < curbuf->b_ml.ml_line_count
		    && !linewhite(end_lnum + 1)
		    && !startPS(end_lnum + 1, 0, 0))
		++end_lnum;
	}

	if (i == 0 && white_in_front && include)
	    break;

	/*
	 * skip to end of white lines after paragraph
	 */
	if (include || do_white)
	    while (end_lnum < curbuf->b_ml.ml_line_count
						   && linewhite(end_lnum + 1))
		++end_lnum;
    }

    /*
     * If there are no empty lines at the end, try to find some empty lines at
     * the start (unless that has been done already).
     */
    if (!white_in_front && !linewhite(end_lnum) && include)
	while (start_lnum > 1 && linewhite(start_lnum - 1))
	    --start_lnum;

    if (VIsual_active)
    {
	// Problem: when doing "Vipipip" nothing happens in a single white
	// line, we get stuck there.  Trap this here.
	if (VIsual_mode == 'V' && start_lnum == curwin->w_cursor.lnum)
	    goto extend;
	if (VIsual.lnum != start_lnum)
	{
	    VIsual.lnum = start_lnum;
	    VIsual.col = 0;
	}
	VIsual_mode = 'V';
	redraw_curbuf_later(UPD_INVERTED);	// update the inversion
	showmode();
    }
    else
    {
	oap->start.lnum = start_lnum;
	oap->start.col = 0;
	oap->motion_type = MLINE;
    }
    curwin->w_cursor.lnum = end_lnum;
    curwin->w_cursor.col = 0;

    return OK;
}

/*
 * Search quote char from string line[col].
 * Quote character escaped by one of the characters in "escape" is not counted
 * as a quote.
 * Returns column number of "quotechar" or -1 when not found.
 */
    static int
find_next_quote(
    char_u	*line,
    int		col,
    int		quotechar,
    char_u	*escape)	// escape characters, can be NULL
{
    int		c;

    for (;;)
    {
	c = line[col];
	if (c == NUL)
	    return -1;
	else if (escape != NULL && vim_strchr(escape, c))
	{
	    ++col;
	    if (line[col] == NUL)
		return -1;
	}
	else if (c == quotechar)
	    break;
	if (has_mbyte)
	    col += (*mb_ptr2len)(line + col);
	else
	    ++col;
    }
    return col;
}

/*
 * Search backwards in "line" from column "col_start" to find "quotechar".
 * Quote character escaped by one of the characters in "escape" is not counted
 * as a quote.
 * Return the found column or zero.
 */
    static int
find_prev_quote(
    char_u	*line,
    int		col_start,
    int		quotechar,
    char_u	*escape)	// escape characters, can be NULL
{
    int		n;

    while (col_start > 0)
    {
	--col_start;
	col_start -= (*mb_head_off)(line, line + col_start);
	n = 0;
	if (escape != NULL)
	    while (col_start - n > 0 && vim_strchr(escape,
					     line[col_start - n - 1]) != NULL)
	    ++n;
	if (n & 1)
	    col_start -= n;	// uneven number of escape chars, skip it
	else if (line[col_start] == quotechar)
	    break;
    }
    return col_start;
}

/*
 * Find quote under the cursor, cursor at end.
 * Returns TRUE if found, else FALSE.
 */
    int
current_quote(
    oparg_T	*oap,
    long	count,
    int		include,	// TRUE == include quote char
    int		quotechar)	// Quote character
{
    char_u	*line = ml_get_curline();
    int		col_end;
    int		col_start = curwin->w_cursor.col;
    int		inclusive = FALSE;
    int		vis_empty = TRUE;	// Visual selection <= 1 char
    int		vis_bef_curs = FALSE;	// Visual starts before cursor
    int		did_exclusive_adj = FALSE;  // adjusted pos for 'selection'
    int		inside_quotes = FALSE;	// Looks like "i'" done before
    int		selected_quote = FALSE;	// Has quote inside selection
    int		i;
    int		restore_vis_bef = FALSE; // restore VIsual on abort

    // When 'selection' is "exclusive" move the cursor to where it would be
    // with 'selection' "inclusive", so that the logic is the same for both.
    // The cursor then is moved forward after adjusting the area.
    if (VIsual_active)
    {
	// this only works within one line
	if (VIsual.lnum != curwin->w_cursor.lnum)
	    return FALSE;

	vis_bef_curs = LT_POS(VIsual, curwin->w_cursor);
	vis_empty = EQUAL_POS(VIsual, curwin->w_cursor);
	if (*p_sel == 'e')
	{
	    if (vis_bef_curs)
	    {
		dec_cursor();
		did_exclusive_adj = TRUE;
	    }
	    else if (!vis_empty)
	    {
		dec(&VIsual);
		did_exclusive_adj = TRUE;
	    }
	    vis_empty = EQUAL_POS(VIsual, curwin->w_cursor);
	    if (!vis_bef_curs && !vis_empty)
	    {
		// VIsual needs to be the start of Visual selection.
		pos_T t = curwin->w_cursor;

		curwin->w_cursor = VIsual;
		VIsual = t;
		vis_bef_curs = TRUE;
		restore_vis_bef = TRUE;
	    }
	}
    }

    if (!vis_empty)
    {
	// Check if the existing selection exactly spans the text inside
	// quotes.
	if (vis_bef_curs)
	{
	    inside_quotes = VIsual.col > 0
			&& line[VIsual.col - 1] == quotechar
			&& line[curwin->w_cursor.col] != NUL
			&& line[curwin->w_cursor.col + 1] == quotechar;
	    i = VIsual.col;
	    col_end = curwin->w_cursor.col;
	}
	else
	{
	    inside_quotes = curwin->w_cursor.col > 0
			&& line[curwin->w_cursor.col - 1] == quotechar
			&& line[VIsual.col] != NUL
			&& line[VIsual.col + 1] == quotechar;
	    i = curwin->w_cursor.col;
	    col_end = VIsual.col;
	}

	// Find out if we have a quote in the selection.
	while (i <= col_end)
	{
	    // check for going over the end of the line, which can happen if
	    // the line was changed after the Visual area was selected.
	    if (line[i] == NUL)
		break;
	    if (line[i++] == quotechar)
	    {
		selected_quote = TRUE;
		break;
	    }
	}
    }

    if (!vis_empty && line[col_start] == quotechar)
    {
	// Already selecting something and on a quote character.  Find the
	// next quoted string.
	if (vis_bef_curs)
	{
	    // Assume we are on a closing quote: move to after the next
	    // opening quote.
	    col_start = find_next_quote(line, col_start + 1, quotechar, NULL);
	    if (col_start < 0)
		goto abort_search;
	    col_end = find_next_quote(line, col_start + 1, quotechar,
							      curbuf->b_p_qe);
	    if (col_end < 0)
	    {
		// We were on a starting quote perhaps?
		col_end = col_start;
		col_start = curwin->w_cursor.col;
	    }
	}
	else
	{
	    col_end = find_prev_quote(line, col_start, quotechar, NULL);
	    if (line[col_end] != quotechar)
		goto abort_search;
	    col_start = find_prev_quote(line, col_end, quotechar,
							      curbuf->b_p_qe);
	    if (line[col_start] != quotechar)
	    {
		// We were on an ending quote perhaps?
		col_start = col_end;
		col_end = curwin->w_cursor.col;
	    }
	}
    }
    else

    if (line[col_start] == quotechar || !vis_empty)
    {
	int	first_col = col_start;

	if (!vis_empty)
	{
	    if (vis_bef_curs)
		first_col = find_next_quote(line, col_start, quotechar, NULL);
	    else
		first_col = find_prev_quote(line, col_start, quotechar, NULL);
	}

	// The cursor is on a quote, we don't know if it's the opening or
	// closing quote.  Search from the start of the line to find out.
	// Also do this when there is a Visual area, a' may leave the cursor
	// in between two strings.
	col_start = 0;
	for (;;)
	{
	    // Find open quote character.
	    col_start = find_next_quote(line, col_start, quotechar, NULL);
	    if (col_start < 0 || col_start > first_col)
		goto abort_search;
	    // Find close quote character.
	    col_end = find_next_quote(line, col_start + 1, quotechar,
							      curbuf->b_p_qe);
	    if (col_end < 0)
		goto abort_search;
	    // If is cursor between start and end quote character, it is
	    // target text object.
	    if (col_start <= first_col && first_col <= col_end)
		break;
	    col_start = col_end + 1;
	}
    }
    else
    {
	// Search backward for a starting quote.
	col_start = find_prev_quote(line, col_start, quotechar, curbuf->b_p_qe);
	if (line[col_start] != quotechar)
	{
	    // No quote before the cursor, look after the cursor.
	    col_start = find_next_quote(line, col_start, quotechar, NULL);
	    if (col_start < 0)
		goto abort_search;
	}

	// Find close quote character.
	col_end = find_next_quote(line, col_start + 1, quotechar,
							      curbuf->b_p_qe);
	if (col_end < 0)
	    goto abort_search;
    }

    // When "include" is TRUE, include spaces after closing quote or before
    // the starting quote.
    if (include)
    {
	if (VIM_ISWHITE(line[col_end + 1]))
	    while (VIM_ISWHITE(line[col_end + 1]))
		++col_end;
	else
	    while (col_start > 0 && VIM_ISWHITE(line[col_start - 1]))
		--col_start;
    }

    // Set start position.  After vi" another i" must include the ".
    // For v2i" include the quotes.
    if (!include && count < 2 && (vis_empty || !inside_quotes))
	++col_start;
    curwin->w_cursor.col = col_start;
    if (VIsual_active)
    {
	// Set the start of the Visual area when the Visual area was empty, we
	// were just inside quotes or the Visual area didn't start at a quote
	// and didn't include a quote.
	if (vis_empty
		|| (vis_bef_curs
		    && !selected_quote
		    && (inside_quotes
			|| (line[VIsual.col] != quotechar
			    && (VIsual.col == 0
				|| line[VIsual.col - 1] != quotechar)))))
	{
	    VIsual = curwin->w_cursor;
	    redraw_curbuf_later(UPD_INVERTED);
	}
    }
    else
    {
	oap->start = curwin->w_cursor;
	oap->motion_type = MCHAR;
    }

    // Set end position.
    curwin->w_cursor.col = col_end;
    if ((include || count > 1 // After vi" another i" must include the ".
		|| (!vis_empty && inside_quotes)
	) && inc_cursor() == 2)
	inclusive = TRUE;
    if (VIsual_active)
    {
	if (vis_empty || vis_bef_curs)
	{
	    // decrement cursor when 'selection' is not exclusive
	    if (*p_sel != 'e')
		dec_cursor();
	}
	else
	{
	    // Cursor is at start of Visual area.  Set the end of the Visual
	    // area when it was just inside quotes or it didn't end at a
	    // quote.
	    if (inside_quotes
		    || (!selected_quote
			&& line[VIsual.col] != quotechar
			&& (line[VIsual.col] == NUL
			    || line[VIsual.col + 1] != quotechar)))
	    {
		dec_cursor();
		VIsual = curwin->w_cursor;
	    }
	    curwin->w_cursor.col = col_start;
	}
	if (VIsual_mode == 'V')
	{
	    VIsual_mode = 'v';
	    redraw_cmdline = TRUE;		// show mode later
	}
    }
    else
    {
	// Set inclusive and other oap's flags.
	oap->inclusive = inclusive;
    }

    return OK;

abort_search:
    if (VIsual_active && *p_sel == 'e')
    {
	if (did_exclusive_adj)
	    inc_cursor();
	if (restore_vis_bef)
	{
	    pos_T t = curwin->w_cursor;

	    curwin->w_cursor = VIsual;
	    VIsual = t;
	}
    }
    return FALSE;
}