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
use std::{
    collections::HashSet,
    env,
    ffi::{OsStr, OsString},
    io::{self, IsTerminal, Write},
    path::{Path, PathBuf},
    sync::Arc,
};

use {
    clap,
    grep::{
        cli,
        matcher::LineTerminator,
        printer::{
            default_color_specs, ColorSpecs, HyperlinkConfig,
            HyperlinkEnvironment, HyperlinkFormat, JSONBuilder, PathPrinter,
            PathPrinterBuilder, Standard, StandardBuilder, Stats, Summary,
            SummaryBuilder, SummaryKind, JSON,
        },
        regex::{
            RegexMatcher as RustRegexMatcher,
            RegexMatcherBuilder as RustRegexMatcherBuilder,
        },
        searcher::{
            BinaryDetection, Encoding, MmapChoice, Searcher, SearcherBuilder,
        },
    },
    ignore::{
        overrides::{Override, OverrideBuilder},
        types::{FileTypeDef, Types, TypesBuilder},
        {Walk, WalkBuilder, WalkParallel},
    },
    termcolor::{BufferWriter, ColorChoice, WriteColor},
};

#[cfg(feature = "pcre2")]
use grep::pcre2::{
    RegexMatcher as PCRE2RegexMatcher,
    RegexMatcherBuilder as PCRE2RegexMatcherBuilder,
};

use crate::{
    app, config,
    logger::Logger,
    messages::{set_ignore_messages, set_messages},
    search::{PatternMatcher, Printer, SearchWorker, SearchWorkerBuilder},
    subject::{Subject, SubjectBuilder},
    Result,
};

/// The command that ripgrep should execute based on the command line
/// configuration.
#[derive(Clone, Copy, Debug, Eq, PartialEq)]
pub enum Command {
    /// Search using exactly one thread.
    Search,
    /// Search using possibly many threads.
    SearchParallel,
    /// The command line parameters suggest that a search should occur, but
    /// ripgrep knows that a match can never be found (e.g., no given patterns
    /// or --max-count=0).
    SearchNever,
    /// Show the files that would be searched, but don't actually search them,
    /// and use exactly one thread.
    Files,
    /// Show the files that would be searched, but don't actually search them,
    /// and perform directory traversal using possibly many threads.
    FilesParallel,
    /// List all file type definitions configured, including the default file
    /// types and any additional file types added to the command line.
    Types,
    /// Print the version of PCRE2 in use.
    PCRE2Version,
}

impl Command {
    /// Returns true if and only if this command requires executing a search.
    fn is_search(&self) -> bool {
        use self::Command::*;

        match *self {
            Search | SearchParallel => true,
            SearchNever | Files | FilesParallel | Types | PCRE2Version => {
                false
            }
        }
    }
}

/// The primary configuration object used throughout ripgrep. It provides a
/// high-level convenient interface to the provided command line arguments.
///
/// An `Args` object is cheap to clone and can be used from multiple threads
/// simultaneously.
#[derive(Clone, Debug)]
pub struct Args(Arc<ArgsImp>);

#[derive(Clone, Debug)]
struct ArgsImp {
    /// Mid-to-low level routines for extracting CLI arguments.
    matches: ArgMatches,
    /// The command we want to execute.
    command: Command,
    /// The number of threads to use. This is based in part on available
    /// threads, in part on the number of threads requested and in part on the
    /// command we're running.
    threads: usize,
    /// A matcher built from the patterns.
    ///
    /// It's important that this is only built once, since building this goes
    /// through regex compilation and various types of analyses. That is, if
    /// you need many of these (one per thread, for example), it is better to
    /// build it once and then clone it.
    matcher: PatternMatcher,
    /// The paths provided at the command line. This is guaranteed to be
    /// non-empty. (If no paths are provided, then a default path is created.)
    paths: Vec<PathBuf>,
    /// Returns true if and only if `paths` had to be populated with a single
    /// default path.
    using_default_path: bool,
}

impl Args {
    /// Parse the command line arguments for this process.
    ///
    /// If a CLI usage error occurred, then exit the process and print a usage
    /// or error message. Similarly, if the user requested the version of
    /// ripgrep, then print the version and exit.
    ///
    /// Also, initialize a global logger.
    pub fn parse() -> Result<Args> {
        // We parse the args given on CLI. This does not include args from
        // the config. We use the CLI args as an initial configuration while
        // trying to parse config files. If a config file exists and has
        // arguments, then we re-parse argv, otherwise we just use the matches
        // we have here.
        let early_matches = ArgMatches::new(clap_matches(env::args_os())?);
        set_messages(!early_matches.is_present("no-messages"));
        set_ignore_messages(!early_matches.is_present("no-ignore-messages"));

        if let Err(err) = Logger::init() {
            return Err(format!("failed to initialize logger: {}", err).into());
        }
        if early_matches.is_present("trace") {
            log::set_max_level(log::LevelFilter::Trace);
        } else if early_matches.is_present("debug") {
            log::set_max_level(log::LevelFilter::Debug);
        } else {
            log::set_max_level(log::LevelFilter::Warn);
        }

        let matches = early_matches.reconfigure()?;
        // The logging level may have changed if we brought in additional
        // arguments from a configuration file, so recheck it and set the log
        // level as appropriate.
        if matches.is_present("trace") {
            log::set_max_level(log::LevelFilter::Trace);
        } else if matches.is_present("debug") {
            log::set_max_level(log::LevelFilter::Debug);
        } else {
            log::set_max_level(log::LevelFilter::Warn);
        }
        set_messages(!matches.is_present("no-messages"));
        set_ignore_messages(!matches.is_present("no-ignore-messages"));
        matches.to_args()
    }

    /// Return direct access to command line arguments.
    fn matches(&self) -> &ArgMatches {
        &self.0.matches
    }

    /// Return the matcher builder from the patterns.
    fn matcher(&self) -> &PatternMatcher {
        &self.0.matcher
    }

    /// Return the paths found in the command line arguments. This is
    /// guaranteed to be non-empty. In the case where no explicit arguments are
    /// provided, a single default path is provided automatically.
    fn paths(&self) -> &[PathBuf] {
        &self.0.paths
    }

    /// Returns true if and only if `paths` had to be populated with a default
    /// path, which occurs only when no paths were given as command line
    /// arguments.
    pub fn using_default_path(&self) -> bool {
        self.0.using_default_path
    }

    /// Return the printer that should be used for formatting the output of
    /// search results.
    ///
    /// The returned printer will write results to the given writer.
    fn printer<W: WriteColor>(&self, wtr: W) -> Result<Printer<W>> {
        match self.matches().output_kind() {
            OutputKind::Standard => {
                let separator_search = self.command() == Command::Search;
                self.matches()
                    .printer_standard(self.paths(), wtr, separator_search)
                    .map(Printer::Standard)
            }
            OutputKind::Summary => self
                .matches()
                .printer_summary(self.paths(), wtr)
                .map(Printer::Summary),
            OutputKind::JSON => {
                self.matches().printer_json(wtr).map(Printer::JSON)
            }
        }
    }
}

/// High level public routines for building data structures used by ripgrep
/// from command line arguments.
impl Args {
    /// Create a new buffer writer for multi-threaded printing with color
    /// support.
    pub fn buffer_writer(&self) -> Result<BufferWriter> {
        let mut wtr = BufferWriter::stdout(self.matches().color_choice());
        wtr.separator(self.matches().file_separator()?);
        Ok(wtr)
    }

    /// Return the high-level command that ripgrep should run.
    pub fn command(&self) -> Command {
        self.0.command
    }

    /// Builder a path printer that can be used for printing just file paths,
    /// with optional color support.
    ///
    /// The printer will print paths to the given writer.
    pub fn path_printer<W: WriteColor>(
        &self,
        wtr: W,
    ) -> Result<PathPrinter<W>> {
        let mut builder = PathPrinterBuilder::new();
        builder
            .color_specs(self.matches().color_specs()?)
            .hyperlink(self.matches().hyperlink_config()?)
            .separator(self.matches().path_separator()?)
            .terminator(self.matches().path_terminator().unwrap_or(b'\n'));
        Ok(builder.build(wtr))
    }

    /// Returns true if and only if ripgrep should be "quiet."
    pub fn quiet(&self) -> bool {
        self.matches().is_present("quiet")
    }

    /// Returns true if and only if the search should quit after finding the
    /// first match.
    pub fn quit_after_match(&self) -> Result<bool> {
        Ok(self.matches().is_present("quiet") && self.stats()?.is_none())
    }

    /// Build a worker for executing searches.
    ///
    /// Search results are written to the given writer.
    pub fn search_worker<W: WriteColor>(
        &self,
        wtr: W,
    ) -> Result<SearchWorker<W>> {
        let matches = self.matches();
        let matcher = self.matcher().clone();
        let printer = self.printer(wtr)?;
        let searcher = matches.searcher(self.paths())?;
        let mut builder = SearchWorkerBuilder::new();
        builder
            .json_stats(matches.is_present("json"))
            .preprocessor(matches.preprocessor())?
            .preprocessor_globs(matches.preprocessor_globs()?)
            .search_zip(matches.is_present("search-zip"))
            .binary_detection_implicit(matches.binary_detection_implicit())
            .binary_detection_explicit(matches.binary_detection_explicit());
        Ok(builder.build(matcher, searcher, printer))
    }

    /// Returns a zero value for tracking statistics if and only if it has been
    /// requested.
    ///
    /// When this returns a `Stats` value, then it is guaranteed that the
    /// search worker will be configured to track statistics as well.
    pub fn stats(&self) -> Result<Option<Stats>> {
        Ok(if self.command().is_search() && self.matches().stats() {
            Some(Stats::new())
        } else {
            None
        })
    }

    /// Return a builder for constructing subjects. A subject represents a
    /// single unit of something to search. Typically, this corresponds to a
    /// file or a stream such as stdin.
    pub fn subject_builder(&self) -> SubjectBuilder {
        let mut builder = SubjectBuilder::new();
        builder.strip_dot_prefix(self.using_default_path());
        builder
    }

    /// Execute the given function with a writer to stdout that enables color
    /// support based on the command line configuration.
    pub fn stdout(&self) -> cli::StandardStream {
        let color = self.matches().color_choice();
        if self.matches().is_present("line-buffered") {
            cli::stdout_buffered_line(color)
        } else if self.matches().is_present("block-buffered") {
            cli::stdout_buffered_block(color)
        } else {
            cli::stdout(color)
        }
    }

    /// Return the type definitions compiled into ripgrep.
    ///
    /// If there was a problem reading and parsing the type definitions, then
    /// this returns an error.
    pub fn type_defs(&self) -> Result<Vec<FileTypeDef>> {
        Ok(self.matches().types()?.definitions().to_vec())
    }

    /// Return a walker that never uses additional threads.
    pub fn walker(&self) -> Result<Walk> {
        Ok(self
            .matches()
            .walker_builder(self.paths(), self.0.threads)?
            .build())
    }

    /// Returns true if and only if `stat`-related sorting is required
    pub fn needs_stat_sort(&self) -> bool {
        return self.matches().sort_by().map_or(
            false,
            |sort_by| match sort_by.kind {
                SortByKind::LastModified
                | SortByKind::Created
                | SortByKind::LastAccessed => sort_by.check().is_ok(),
                _ => false,
            },
        );
    }

    /// Sort subjects if a sorter is specified, but only if the sort requires
    /// stat calls. Non-stat related sorts are handled during file traversal
    ///
    /// This function assumes that it is known that a stat-related sort is
    /// required, and does not check for it again.
    ///
    /// It is important that that precondition is fulfilled, since this function
    /// consumes the subjects iterator, and is therefore a blocking function.
    pub fn sort_by_stat<I>(&self, subjects: I) -> Vec<Subject>
    where
        I: Iterator<Item = Subject>,
    {
        let sorter = match self.matches().sort_by() {
            Ok(v) => v,
            Err(_) => return subjects.collect(),
        };
        use SortByKind::*;
        let mut keyed = match sorter.kind {
            LastModified => load_timestamps(subjects, |m| m.modified()),
            LastAccessed => load_timestamps(subjects, |m| m.accessed()),
            Created => load_timestamps(subjects, |m| m.created()),
            _ => return subjects.collect(),
        };
        keyed.sort_by(|a, b| sort_by_option(&a.0, &b.0, sorter.reverse));
        keyed.into_iter().map(|v| v.1).collect()
    }

    /// Return a parallel walker that may use additional threads.
    pub fn walker_parallel(&self) -> Result<WalkParallel> {
        Ok(self
            .matches()
            .walker_builder(self.paths(), self.0.threads)?
            .build_parallel())
    }
}

/// `ArgMatches` wraps `clap::ArgMatches` and provides semantic meaning to
/// the parsed arguments.
#[derive(Clone, Debug)]
struct ArgMatches(clap::ArgMatches<'static>);

/// The output format. Generally, this corresponds to the printer that ripgrep
/// uses to show search results.
#[derive(Clone, Copy, Debug, Eq, PartialEq)]
enum OutputKind {
    /// Classic grep-like or ack-like format.
    Standard,
    /// Show matching files and possibly the number of matches in each file.
    Summary,
    /// Emit match information in the JSON Lines format.
    JSON,
}

/// The sort criteria, if present.
#[derive(Clone, Copy, Debug, Eq, PartialEq)]
struct SortBy {
    /// Whether to reverse the sort criteria (i.e., descending order).
    reverse: bool,
    /// The actual sorting criteria.
    kind: SortByKind,
}

#[derive(Clone, Copy, Debug, Eq, PartialEq)]
enum SortByKind {
    /// No sorting at all.
    None,
    /// Sort by path.
    Path,
    /// Sort by last modified time.
    LastModified,
    /// Sort by last accessed time.
    LastAccessed,
    /// Sort by creation time.
    Created,
}

impl SortBy {
    fn asc(kind: SortByKind) -> SortBy {
        SortBy { reverse: false, kind }
    }

    fn desc(kind: SortByKind) -> SortBy {
        SortBy { reverse: true, kind }
    }

    fn none() -> SortBy {
        SortBy::asc(SortByKind::None)
    }

    /// Try to check that the sorting criteria selected is actually supported.
    /// If it isn't, then an error is returned.
    fn check(&self) -> Result<()> {
        match self.kind {
            SortByKind::None | SortByKind::Path => {}
            SortByKind::LastModified => {
                env::current_exe()?.metadata()?.modified()?;
            }
            SortByKind::LastAccessed => {
                env::current_exe()?.metadata()?.accessed()?;
            }
            SortByKind::Created => {
                env::current_exe()?.metadata()?.created()?;
            }
        }
        Ok(())
    }

    /// Load sorters only if they are applicable at the walk stage.
    ///
    /// In particular, sorts that involve `stat` calls are not loaded because
    /// the walk inherently assumes that parent directories are aware of all its
    /// decendent properties, but `stat` does not work that way.
    fn configure_builder_sort(self, builder: &mut WalkBuilder) {
        use SortByKind::*;
        match self.kind {
            Path if self.reverse => {
                builder.sort_by_file_name(|a, b| a.cmp(b).reverse());
            }
            Path => {
                builder.sort_by_file_name(|a, b| a.cmp(b));
            }
            // these use `stat` calls and will be sorted in Args::sort_by_stat()
            LastModified | LastAccessed | Created | None => {}
        };
    }
}

impl SortByKind {
    fn new(kind: &str) -> SortByKind {
        match kind {
            "none" => SortByKind::None,
            "path" => SortByKind::Path,
            "modified" => SortByKind::LastModified,
            "accessed" => SortByKind::LastAccessed,
            "created" => SortByKind::Created,
            _ => SortByKind::None,
        }
    }
}

/// Encoding mode the searcher will use.
#[derive(Clone, Debug)]
enum EncodingMode {
    /// Use an explicit encoding forcefully, but let BOM sniffing override it.
    Some(Encoding),
    /// Use only BOM sniffing to auto-detect an encoding.
    Auto,
    /// Use no explicit encoding and disable all BOM sniffing. This will
    /// always result in searching the raw bytes, regardless of their
    /// true encoding.
    Disabled,
}

impl ArgMatches {
    /// Create an ArgMatches from clap's parse result.
    fn new(clap_matches: clap::ArgMatches<'static>) -> ArgMatches {
        ArgMatches(clap_matches)
    }

    /// Run clap and return the matches using a config file if present. If clap
    /// determines a problem with the user provided arguments (or if --help or
    /// --version are given), then an error/usage/version will be printed and
    /// the process will exit.
    ///
    /// If there are no additional arguments from the environment (e.g., a
    /// config file), then the given matches are returned as is.
    fn reconfigure(self) -> Result<ArgMatches> {
        // If the end user says no config, then respect it.
        if self.is_present("no-config") {
            log::debug!(
                "not reading config files because --no-config is present"
            );
            return Ok(self);
        }
        // If the user wants ripgrep to use a config file, then parse args
        // from that first.
        let mut args = config::args();
        if args.is_empty() {
            return Ok(self);
        }
        let mut cliargs = env::args_os();
        if let Some(bin) = cliargs.next() {
            args.insert(0, bin);
        }
        args.extend(cliargs);
        log::debug!("final argv: {:?}", args);
        Ok(ArgMatches(clap_matches(args)?))
    }

    /// Convert the result of parsing CLI arguments into ripgrep's higher level
    /// configuration structure.
    fn to_args(self) -> Result<Args> {
        // We compute these once since they could be large.
        let patterns = self.patterns()?;
        let matcher = self.matcher(&patterns)?;
        let mut paths = self.paths();
        let using_default_path = if paths.is_empty() {
            paths.push(self.path_default());
            true
        } else {
            false
        };
        // Now figure out the number of threads we'll use and which
        // command will run.
        let is_one_search = self.is_one_search(&paths);
        let threads = if is_one_search { 1 } else { self.threads()? };
        if threads == 1 {
            log::debug!("running in single threaded mode");
        } else {
            log::debug!("running with {threads} threads for parallelism");
        }
        let command = if self.is_present("pcre2-version") {
            Command::PCRE2Version
        } else if self.is_present("type-list") {
            Command::Types
        } else if self.is_present("files") {
            if threads == 1 {
                Command::Files
            } else {
                Command::FilesParallel
            }
        } else if self.can_never_match(&patterns) {
            Command::SearchNever
        } else if threads == 1 {
            Command::Search
        } else {
            Command::SearchParallel
        };
        Ok(Args(Arc::new(ArgsImp {
            matches: self,
            command,
            threads,
            matcher,
            paths,
            using_default_path,
        })))
    }
}

/// High level routines for converting command line arguments into various
/// data structures used by ripgrep.
///
/// Methods are sorted alphabetically.
impl ArgMatches {
    /// Return the matcher that should be used for searching.
    ///
    /// If there was a problem building the matcher (e.g., a syntax error),
    /// then this returns an error.
    fn matcher(&self, patterns: &[String]) -> Result<PatternMatcher> {
        if self.is_present("pcre2") {
            self.matcher_engine("pcre2", patterns)
        } else if self.is_present("auto-hybrid-regex") {
            self.matcher_engine("auto", patterns)
        } else {
            let engine = self.value_of_lossy("engine").unwrap();
            self.matcher_engine(&engine, patterns)
        }
    }

    /// Return the matcher that should be used for searching using engine
    /// as the engine for the patterns.
    ///
    /// If there was a problem building the matcher (e.g., a syntax error),
    /// then this returns an error.
    fn matcher_engine(
        &self,
        engine: &str,
        patterns: &[String],
    ) -> Result<PatternMatcher> {
        match engine {
            "default" => {
                let matcher = match self.matcher_rust(patterns) {
                    Ok(matcher) => matcher,
                    Err(err) => {
                        return Err(From::from(suggest(err.to_string())));
                    }
                };
                Ok(PatternMatcher::RustRegex(matcher))
            }
            #[cfg(feature = "pcre2")]
            "pcre2" => {
                let matcher = self.matcher_pcre2(patterns)?;
                Ok(PatternMatcher::PCRE2(matcher))
            }
            #[cfg(not(feature = "pcre2"))]
            "pcre2" => Err(From::from(
                "PCRE2 is not available in this build of ripgrep",
            )),
            "auto" => {
                let rust_err = match self.matcher_rust(patterns) {
                    Ok(matcher) => {
                        return Ok(PatternMatcher::RustRegex(matcher));
                    }
                    Err(err) => err,
                };
                log::debug!(
                    "error building Rust regex in hybrid mode:\n{}",
                    rust_err,
                );

                let pcre_err = match self.matcher_engine("pcre2", patterns) {
                    Ok(matcher) => return Ok(matcher),
                    Err(err) => err,
                };
                Err(From::from(format!(
                    "regex could not be compiled with either the default \
                     regex engine or with PCRE2.\n\n\
                     default regex engine error:\n{}\n{}\n{}\n\n\
                     PCRE2 regex engine error:\n{}",
                    "~".repeat(79),
                    rust_err,
                    "~".repeat(79),
                    pcre_err,
                )))
            }
            _ => Err(From::from(format!(
                "unrecognized regex engine '{}'",
                engine
            ))),
        }
    }

    /// Build a matcher using Rust's regex engine.
    ///
    /// If there was a problem building the matcher (such as a regex syntax
    /// error), then an error is returned.
    fn matcher_rust(&self, patterns: &[String]) -> Result<RustRegexMatcher> {
        let mut builder = RustRegexMatcherBuilder::new();
        builder
            .case_smart(self.case_smart())
            .case_insensitive(self.case_insensitive())
            .multi_line(true)
            .unicode(self.unicode())
            .octal(false)
            .fixed_strings(self.is_present("fixed-strings"))
            .whole_line(self.is_present("line-regexp"))
            .word(self.is_present("word-regexp"));
        if self.is_present("multiline") {
            builder.dot_matches_new_line(self.is_present("multiline-dotall"));
            if self.is_present("crlf") {
                builder.crlf(true).line_terminator(None);
            }
        } else {
            builder.line_terminator(Some(b'\n')).dot_matches_new_line(false);
            if self.is_present("crlf") {
                builder.crlf(true);
            }
            // We don't need to set this in multiline mode since mulitline
            // matchers don't use optimizations related to line terminators.
            // Moreover, a mulitline regex used with --null-data should
            // be allowed to match NUL bytes explicitly, which this would
            // otherwise forbid.
            if self.is_present("null-data") {
                builder.line_terminator(Some(b'\x00'));
            }
        }
        if let Some(limit) = self.regex_size_limit()? {
            builder.size_limit(limit);
        }
        if let Some(limit) = self.dfa_size_limit()? {
            builder.dfa_size_limit(limit);
        }
        match builder.build_many(patterns) {
            Ok(m) => Ok(m),
            Err(err) => Err(From::from(suggest_multiline(err.to_string()))),
        }
    }

    /// Build a matcher using PCRE2.
    ///
    /// If there was a problem building the matcher (such as a regex syntax
    /// error), then an error is returned.
    #[cfg(feature = "pcre2")]
    fn matcher_pcre2(&self, patterns: &[String]) -> Result<PCRE2RegexMatcher> {
        let mut builder = PCRE2RegexMatcherBuilder::new();
        builder
            .case_smart(self.case_smart())
            .caseless(self.case_insensitive())
            .multi_line(true)
            .fixed_strings(self.is_present("fixed-strings"))
            .whole_line(self.is_present("line-regexp"))
            .word(self.is_present("word-regexp"));
        // For whatever reason, the JIT craps out during regex compilation with
        // a "no more memory" error on 32 bit systems. So don't use it there.
        if cfg!(target_pointer_width = "64") {
            builder
                .jit_if_available(true)
                // The PCRE2 docs say that 32KB is the default, and that 1MB
                // should be big enough for anything. But let's crank it to
                // 10MB.
                .max_jit_stack_size(Some(10 * (1 << 20)));
        }
        if self.unicode() {
            builder.utf(true).ucp(true);
        }
        if self.is_present("multiline") {
            builder.dotall(self.is_present("multiline-dotall"));
        }
        if self.is_present("crlf") {
            builder.crlf(true);
        }
        Ok(builder.build_many(patterns)?)
    }

    /// Build a JSON printer that writes results to the given writer.
    fn printer_json<W: io::Write>(&self, wtr: W) -> Result<JSON<W>> {
        let mut builder = JSONBuilder::new();
        builder
            .pretty(false)
            .max_matches(self.max_count()?)
            .always_begin_end(false);
        Ok(builder.build(wtr))
    }

    /// Build a Standard printer that writes results to the given writer.
    ///
    /// The given paths are used to configure aspects of the printer.
    ///
    /// If `separator_search` is true, then the returned printer will assume
    /// the responsibility of printing a separator between each set of
    /// search results, when appropriate (e.g., when contexts are enabled).
    /// When it's set to false, the caller is responsible for handling
    /// separators.
    ///
    /// In practice, we want the printer to handle it in the single threaded
    /// case but not in the multi-threaded case.
    fn printer_standard<W: WriteColor>(
        &self,
        paths: &[PathBuf],
        wtr: W,
        separator_search: bool,
    ) -> Result<Standard<W>> {
        let mut builder = StandardBuilder::new();
        builder
            .color_specs(self.color_specs()?)
            .hyperlink(self.hyperlink_config()?)
            .stats(self.stats())
            .heading(self.heading())
            .path(self.with_filename(paths))
            .only_matching(self.is_present("only-matching"))
            .per_match(self.is_present("vimgrep"))
            .per_match_one_line(true)
            .replacement(self.replacement())
            .max_columns(self.max_columns()?)
            .max_columns_preview(self.max_columns_preview())
            .max_matches(self.max_count()?)
            .column(self.column())
            .byte_offset(self.is_present("byte-offset"))
            .trim_ascii(self.is_present("trim"))
            .separator_search(None)
            .separator_context(self.context_separator())
            .separator_field_match(self.field_match_separator())
            .separator_field_context(self.field_context_separator())
            .separator_path(self.path_separator()?)
            .path_terminator(self.path_terminator());
        if separator_search {
            builder.separator_search(self.file_separator()?);
        }
        Ok(builder.build(wtr))
    }

    /// Build a Summary printer that writes results to the given writer.
    ///
    /// The given paths are used to configure aspects of the printer.
    ///
    /// This panics if the output format is not `OutputKind::Summary`.
    fn printer_summary<W: WriteColor>(
        &self,
        paths: &[PathBuf],
        wtr: W,
    ) -> Result<Summary<W>> {
        let mut builder = SummaryBuilder::new();
        builder
            .kind(self.summary_kind().expect("summary format"))
            .color_specs(self.color_specs()?)
            .hyperlink(self.hyperlink_config()?)
            .stats(self.stats())
            .path(self.with_filename(paths))
            .max_matches(self.max_count()?)
            .exclude_zero(!self.is_present("include-zero"))
            .separator_field(b":".to_vec())
            .separator_path(self.path_separator()?)
            .path_terminator(self.path_terminator());
        Ok(builder.build(wtr))
    }

    /// Build a searcher from the command line parameters.
    fn searcher(&self, paths: &[PathBuf]) -> Result<Searcher> {
        let (ctx_before, ctx_after) = self.contexts()?;
        let line_term = if self.is_present("crlf") {
            LineTerminator::crlf()
        } else if self.is_present("null-data") {
            LineTerminator::byte(b'\x00')
        } else {
            LineTerminator::byte(b'\n')
        };
        let mut builder = SearcherBuilder::new();
        builder
            .line_terminator(line_term)
            .invert_match(self.is_present("invert-match"))
            .line_number(self.line_number(paths))
            .multi_line(self.is_present("multiline"))
            .before_context(ctx_before)
            .after_context(ctx_after)
            .passthru(self.is_present("passthru"))
            .memory_map(self.mmap_choice(paths))
            .stop_on_nonmatch(self.is_present("stop-on-nonmatch"));
        match self.encoding()? {
            EncodingMode::Some(enc) => {
                builder.encoding(Some(enc));
            }
            EncodingMode::Auto => {} // default for the searcher
            EncodingMode::Disabled => {
                builder.bom_sniffing(false);
            }
        }
        Ok(builder.build())
    }

    /// Return a builder for recursively traversing a directory while
    /// respecting ignore rules.
    ///
    /// If there was a problem parsing the CLI arguments necessary for
    /// constructing the builder, then this returns an error.
    fn walker_builder(
        &self,
        paths: &[PathBuf],
        threads: usize,
    ) -> Result<WalkBuilder> {
        let mut builder = WalkBuilder::new(&paths[0]);
        for path in &paths[1..] {
            builder.add(path);
        }
        if !self.no_ignore_files() {
            for path in self.ignore_paths() {
                if let Some(err) = builder.add_ignore(path) {
                    ignore_message!("{}", err);
                }
            }
        }
        builder
            .max_depth(self.usize_of("max-depth")?)
            .follow_links(self.is_present("follow"))
            .max_filesize(self.max_file_size()?)
            .threads(threads)
            .same_file_system(self.is_present("one-file-system"))
            .skip_stdout(!self.is_present("files"))
            .overrides(self.overrides()?)
            .types(self.types()?)
            .hidden(!self.hidden())
            .parents(!self.no_ignore_parent())
            .ignore(!self.no_ignore_dot())
            .git_global(!self.no_ignore_vcs() && !self.no_ignore_global())
            .git_ignore(!self.no_ignore_vcs())
            .git_exclude(!self.no_ignore_vcs() && !self.no_ignore_exclude())
            .require_git(!self.is_present("no-require-git"))
            .ignore_case_insensitive(self.ignore_file_case_insensitive());
        if !self.no_ignore() && !self.no_ignore_dot() {
            builder.add_custom_ignore_filename(".rgignore");
        }
        self.sort_by()?.configure_builder_sort(&mut builder);
        Ok(builder)
    }
}

/// Mid level routines for converting command line arguments into various types
/// of data structures.
///
/// Methods are sorted alphabetically.
impl ArgMatches {
    /// Returns the form of binary detection to perform on files that are
    /// implicitly searched via recursive directory traversal.
    fn binary_detection_implicit(&self) -> BinaryDetection {
        let none = self.is_present("text") || self.is_present("null-data");
        let convert =
            self.is_present("binary") || self.unrestricted_count() >= 3;
        if none {
            BinaryDetection::none()
        } else if convert {
            BinaryDetection::convert(b'\x00')
        } else {
            BinaryDetection::quit(b'\x00')
        }
    }

    /// Returns the form of binary detection to perform on files that are
    /// explicitly searched via the user invoking ripgrep on a particular
    /// file or files or stdin.
    ///
    /// In general, this should never be BinaryDetection::quit, since that acts
    /// as a filter (but quitting immediately once a NUL byte is seen), and we
    /// should never filter out files that the user wants to explicitly search.
    fn binary_detection_explicit(&self) -> BinaryDetection {
        let none = self.is_present("text") || self.is_present("null-data");
        if none {
            BinaryDetection::none()
        } else {
            BinaryDetection::convert(b'\x00')
        }
    }

    /// Returns true if the command line configuration implies that a match
    /// can never be shown.
    fn can_never_match(&self, patterns: &[String]) -> bool {
        patterns.is_empty() || self.max_count().ok() == Some(Some(0))
    }

    /// Returns true if and only if case should be ignore.
    ///
    /// If --case-sensitive is present, then case is never ignored, even if
    /// --ignore-case is present.
    fn case_insensitive(&self) -> bool {
        self.is_present("ignore-case") && !self.is_present("case-sensitive")
    }

    /// Returns true if and only if smart case has been enabled.
    ///
    /// If either --ignore-case of --case-sensitive are present, then smart
    /// case is disabled.
    fn case_smart(&self) -> bool {
        self.is_present("smart-case")
            && !self.is_present("ignore-case")
            && !self.is_present("case-sensitive")
    }

    /// Returns the user's color choice based on command line parameters and
    /// environment.
    fn color_choice(&self) -> ColorChoice {
        let preference = match self.value_of_lossy("color") {
            None => "auto".to_string(),
            Some(v) => v,
        };
        if preference == "always" {
            ColorChoice::Always
        } else if preference == "ansi" {
            ColorChoice::AlwaysAnsi
        } else if preference == "auto" {
            if std::io::stdout().is_terminal() || self.is_present("pretty") {
                ColorChoice::Auto
            } else {
                ColorChoice::Never
            }
        } else {
            ColorChoice::Never
        }
    }

    /// Returns the color specifications given by the user on the CLI.
    ///
    /// If the was a problem parsing any of the provided specs, then an error
    /// is returned.
    fn color_specs(&self) -> Result<ColorSpecs> {
        // Start with a default set of color specs.
        let mut specs = default_color_specs();
        for spec_str in self.values_of_lossy_vec("colors") {
            specs.push(spec_str.parse()?);
        }
        Ok(ColorSpecs::new(&specs))
    }

    /// Returns true if and only if column numbers should be shown.
    fn column(&self) -> bool {
        if self.is_present("no-column") {
            return false;
        }
        self.is_present("column") || self.is_present("vimgrep")
    }

    /// Returns the before and after contexts from the command line.
    ///
    /// If a context setting was absent, then `0` is returned.
    ///
    /// If there was a problem parsing the values from the user as an integer,
    /// then an error is returned.
    fn contexts(&self) -> Result<(usize, usize)> {
        let both = self.usize_of("context")?.unwrap_or(0);
        let after = self.usize_of("after-context")?.unwrap_or(both);
        let before = self.usize_of("before-context")?.unwrap_or(both);
        Ok((before, after))
    }

    /// Returns the unescaped context separator in UTF-8 bytes.
    ///
    /// If one was not provided, the default `--` is returned.
    /// If --no-context-separator is passed, None is returned.
    fn context_separator(&self) -> Option<Vec<u8>> {
        let nosep = self.is_present("no-context-separator");
        let sep = self.value_of_os("context-separator");
        match (nosep, sep) {
            (true, _) => None,
            (false, None) => Some(b"--".to_vec()),
            (false, Some(sep)) => Some(cli::unescape_os(&sep)),
        }
    }

    /// Returns whether the -c/--count or the --count-matches flags were
    /// passed from the command line.
    ///
    /// If --count-matches and --invert-match were passed in, behave
    /// as if --count and --invert-match were passed in (i.e. rg will
    /// count inverted matches as per existing behavior).
    fn counts(&self) -> (bool, bool) {
        let count = self.is_present("count");
        let count_matches = self.is_present("count-matches");
        let invert_matches = self.is_present("invert-match");
        let only_matching = self.is_present("only-matching");
        if count_matches && invert_matches {
            // Treat `-v --count-matches` as `-v -c`.
            (true, false)
        } else if count && only_matching {
            // Treat `-c --only-matching` as `--count-matches`.
            (false, true)
        } else {
            (count, count_matches)
        }
    }

    /// Parse the dfa-size-limit argument option into a byte count.
    fn dfa_size_limit(&self) -> Result<Option<usize>> {
        let r = self.parse_human_readable_size("dfa-size-limit")?;
        u64_to_usize("dfa-size-limit", r)
    }

    /// Returns the encoding mode to use.
    ///
    /// This only returns an encoding if one is explicitly specified. Otherwise
    /// if set to automatic, the Searcher will do BOM sniffing for UTF-16
    /// and transcode seamlessly. If disabled, no BOM sniffing nor transcoding
    /// will occur.
    fn encoding(&self) -> Result<EncodingMode> {
        if self.is_present("no-encoding") {
            return Ok(EncodingMode::Auto);
        }

        let label = match self.value_of_lossy("encoding") {
            None => return Ok(EncodingMode::Auto),
            Some(label) => label,
        };

        if label == "auto" {
            return Ok(EncodingMode::Auto);
        } else if label == "none" {
            return Ok(EncodingMode::Disabled);
        }

        Ok(EncodingMode::Some(Encoding::new(&label)?))
    }

    /// Return the file separator to use based on the CLI configuration.
    fn file_separator(&self) -> Result<Option<Vec<u8>>> {
        // File separators are only used for the standard grep-line format.
        if self.output_kind() != OutputKind::Standard {
            return Ok(None);
        }

        let (ctx_before, ctx_after) = self.contexts()?;
        Ok(if self.heading() {
            Some(b"".to_vec())
        } else if ctx_before > 0 || ctx_after > 0 {
            self.context_separator()
        } else {
            None
        })
    }

    /// Returns true if and only if matches should be grouped with file name
    /// headings.
    fn heading(&self) -> bool {
        if self.is_present("no-heading") || self.is_present("vimgrep") {
            false
        } else {
            std::io::stdout().is_terminal()
                || self.is_present("heading")
                || self.is_present("pretty")
        }
    }

    /// Returns true if and only if hidden files/directories should be
    /// searched.
    fn hidden(&self) -> bool {
        self.is_present("hidden") || self.unrestricted_count() >= 2
    }

    /// Returns the hyperlink pattern to use. A default pattern suitable
    /// for the current system is used if the value is not set.
    ///
    /// If an invalid pattern is provided, then an error is returned.
    fn hyperlink_config(&self) -> Result<HyperlinkConfig> {
        let mut env = HyperlinkEnvironment::new();
        env.host(hostname(self.value_of_os("hostname-bin")))
            .wsl_prefix(wsl_prefix());
        let fmt: HyperlinkFormat =
            match self.value_of_lossy("hyperlink-format") {
                None => "default".parse().unwrap(),
                Some(format) => match format.parse() {
                    Ok(format) => format,
                    Err(err) => {
                        let msg = format!("invalid hyperlink format: {err}");
                        return Err(msg.into());
                    }
                },
            };
        log::debug!("hyperlink format: {:?}", fmt.to_string());
        Ok(HyperlinkConfig::new(env, fmt))
    }

    /// Returns true if ignore files should be processed case insensitively.
    fn ignore_file_case_insensitive(&self) -> bool {
        self.is_present("ignore-file-case-insensitive")
    }

    /// Return all of the ignore file paths given on the command line.
    fn ignore_paths(&self) -> Vec<PathBuf> {
        let paths = match self.values_of_os("ignore-file") {
            None => return vec![],
            Some(paths) => paths,
        };
        paths.map(|p| Path::new(p).to_path_buf()).collect()
    }

    /// Returns true if and only if ripgrep is invoked in a way where it knows
    /// it search exactly one thing.
    fn is_one_search(&self, paths: &[PathBuf]) -> bool {
        if paths.len() != 1 {
            return false;
        }
        self.is_only_stdin(paths) || paths[0].is_file()
    }

    /// Returns true if and only if we're only searching a single thing and
    /// that thing is stdin.
    fn is_only_stdin(&self, paths: &[PathBuf]) -> bool {
        paths == [Path::new("-")]
    }

    /// Returns true if and only if we should show line numbers.
    fn line_number(&self, paths: &[PathBuf]) -> bool {
        if self.output_kind() == OutputKind::Summary {
            return false;
        }
        if self.is_present("no-line-number") {
            return false;
        }
        if self.output_kind() == OutputKind::JSON {
            return true;
        }

        // A few things can imply counting line numbers. In particular, we
        // generally want to show line numbers by default when printing to a
        // tty for human consumption, except for one interesting case: when
        // we're only searching stdin. This makes pipelines work as expected.
        (std::io::stdout().is_terminal() && !self.is_only_stdin(paths))
            || self.is_present("line-number")
            || self.is_present("column")
            || self.is_present("pretty")
            || self.is_present("vimgrep")
    }

    /// The maximum number of columns allowed on each line.
    ///
    /// If `0` is provided, then this returns `None`.
    fn max_columns(&self) -> Result<Option<u64>> {
        Ok(self.usize_of_nonzero("max-columns")?.map(|n| n as u64))
    }

    /// Returns true if and only if a preview should be shown for lines that
    /// exceed the maximum column limit.
    fn max_columns_preview(&self) -> bool {
        self.is_present("max-columns-preview")
    }

    /// The maximum number of matches permitted.
    fn max_count(&self) -> Result<Option<u64>> {
        Ok(self.usize_of("max-count")?.map(|n| n as u64))
    }

    /// Parses the max-filesize argument option into a byte count.
    fn max_file_size(&self) -> Result<Option<u64>> {
        self.parse_human_readable_size("max-filesize")
    }

    /// Returns whether we should attempt to use memory maps or not.
    fn mmap_choice(&self, paths: &[PathBuf]) -> MmapChoice {
        // SAFETY: Memory maps are difficult to impossible to encapsulate
        // safely in a portable way that doesn't simultaneously negate some of
        // the benfits of using memory maps. For ripgrep's use, we never mutate
        // a memory map and generally never store the contents of memory map
        // in a data structure that depends on immutability. Generally
        // speaking, the worst thing that can happen is a SIGBUS (if the
        // underlying file is truncated while reading it), which will cause
        // ripgrep to abort. This reasoning should be treated as suspect.
        let maybe = unsafe { MmapChoice::auto() };
        let never = MmapChoice::never();
        if self.is_present("no-mmap") {
            never
        } else if self.is_present("mmap") {
            maybe
        } else if paths.len() <= 10 && paths.iter().all(|p| p.is_file()) {
            // If we're only searching a few paths and all of them are
            // files, then memory maps are probably faster.
            maybe
        } else {
            never
        }
    }

    /// Returns true if ignore files should be ignored.
    fn no_ignore(&self) -> bool {
        self.is_present("no-ignore") || self.unrestricted_count() >= 1
    }

    /// Returns true if .ignore files should be ignored.
    fn no_ignore_dot(&self) -> bool {
        self.is_present("no-ignore-dot") || self.no_ignore()
    }

    /// Returns true if local exclude (ignore) files should be ignored.
    fn no_ignore_exclude(&self) -> bool {
        self.is_present("no-ignore-exclude") || self.no_ignore()
    }

    /// Returns true if explicitly given ignore files should be ignored.
    fn no_ignore_files(&self) -> bool {
        // We don't look at no-ignore here because --no-ignore is explicitly
        // documented to not override --ignore-file. We could change this, but
        // it would be a fairly severe breaking change.
        self.is_present("no-ignore-files")
    }

    /// Returns true if global ignore files should be ignored.
    fn no_ignore_global(&self) -> bool {
        self.is_present("no-ignore-global") || self.no_ignore()
    }

    /// Returns true if parent ignore files should be ignored.
    fn no_ignore_parent(&self) -> bool {
        self.is_present("no-ignore-parent") || self.no_ignore()
    }

    /// Returns true if VCS ignore files should be ignored.
    fn no_ignore_vcs(&self) -> bool {
        self.is_present("no-ignore-vcs") || self.no_ignore()
    }

    /// Determine the type of output we should produce.
    fn output_kind(&self) -> OutputKind {
        if self.is_present("quiet") {
            // While we don't technically print results (or aggregate results)
            // in quiet mode, we still support the --stats flag, and those
            // stats are computed by the Summary printer for now.
            return OutputKind::Summary;
        } else if self.is_present("json") {
            return OutputKind::JSON;
        }

        let (count, count_matches) = self.counts();
        let summary = count
            || count_matches
            || self.is_present("files-with-matches")
            || self.is_present("files-without-match");
        if summary {
            OutputKind::Summary
        } else {
            OutputKind::Standard
        }
    }

    /// Builds the set of glob overrides from the command line flags.
    fn overrides(&self) -> Result<Override> {
        let globs = self.values_of_lossy_vec("glob");
        let iglobs = self.values_of_lossy_vec("iglob");
        if globs.is_empty() && iglobs.is_empty() {
            return Ok(Override::empty());
        }

        let mut builder = OverrideBuilder::new(current_dir()?);
        // Make all globs case insensitive with --glob-case-insensitive.
        if self.is_present("glob-case-insensitive") {
            builder.case_insensitive(true).unwrap();
        }
        for glob in globs {
            builder.add(&glob)?;
        }
        // This only enables case insensitivity for subsequent globs.
        builder.case_insensitive(true).unwrap();
        for glob in iglobs {
            builder.add(&glob)?;
        }
        Ok(builder.build()?)
    }

    /// Return all file paths that ripgrep should search.
    ///
    /// If no paths were given, then this returns an empty list.
    fn paths(&self) -> Vec<PathBuf> {
        let mut paths: Vec<PathBuf> = match self.values_of_os("path") {
            None => vec![],
            Some(paths) => paths.map(|p| Path::new(p).to_path_buf()).collect(),
        };
        // If --file, --files or --regexp is given, then the first path is
        // always in `pattern`.
        if self.is_present("file")
            || self.is_present("files")
            || self.is_present("regexp")
        {
            if let Some(path) = self.value_of_os("pattern") {
                paths.insert(0, Path::new(path).to_path_buf());
            }
        }
        paths
    }

    /// Return the default path that ripgrep should search. This should only
    /// be used when ripgrep is not otherwise given at least one file path
    /// as a positional argument.
    fn path_default(&self) -> PathBuf {
        let file_is_stdin = self
            .values_of_os("file")
            .map_or(false, |mut files| files.any(|f| f == "-"));
        let search_cwd = !cli::is_readable_stdin()
            || (self.is_present("file") && file_is_stdin)
            || self.is_present("files")
            || self.is_present("type-list")
            || self.is_present("pcre2-version");
        if search_cwd {
            Path::new("./").to_path_buf()
        } else {
            Path::new("-").to_path_buf()
        }
    }

    /// Returns the unescaped path separator as a single byte, if one exists.
    ///
    /// If the provided path separator is more than a single byte, then an
    /// error is returned.
    fn path_separator(&self) -> Result<Option<u8>> {
        let sep = match self.value_of_os("path-separator") {
            None => return Ok(None),
            Some(sep) => cli::unescape_os(&sep),
        };
        if sep.is_empty() {
            Ok(None)
        } else if sep.len() > 1 {
            Err(From::from(format!(
                "A path separator must be exactly one byte, but \
                 the given separator is {} bytes: {}\n\
                 In some shells on Windows '/' is automatically \
                 expanded. Use '//' instead.",
                sep.len(),
                cli::escape(&sep),
            )))
        } else {
            Ok(Some(sep[0]))
        }
    }

    /// Returns the byte that should be used to terminate paths.
    ///
    /// Typically, this is only set to `\x00` when the --null flag is provided,
    /// and `None` otherwise.
    fn path_terminator(&self) -> Option<u8> {
        if self.is_present("null") {
            Some(b'\x00')
        } else {
            None
        }
    }

    /// Returns the unescaped field context separator. If one wasn't specified,
    /// then '-' is used as the default.
    fn field_context_separator(&self) -> Vec<u8> {
        match self.value_of_os("field-context-separator") {
            None => b"-".to_vec(),
            Some(sep) => cli::unescape_os(&sep),
        }
    }

    /// Returns the unescaped field match separator. If one wasn't specified,
    /// then ':' is used as the default.
    fn field_match_separator(&self) -> Vec<u8> {
        match self.value_of_os("field-match-separator") {
            None => b":".to_vec(),
            Some(sep) => cli::unescape_os(&sep),
        }
    }

    /// Get a sequence of all available patterns from the command line.
    /// This includes reading the -e/--regexp and -f/--file flags.
    ///
    /// If any pattern is invalid UTF-8, then an error is returned.
    fn patterns(&self) -> Result<Vec<String>> {
        if self.is_present("files") || self.is_present("type-list") {
            return Ok(vec![]);
        }
        let mut seen = HashSet::new();
        let mut pats = vec![];
        let mut add = |pat: String| {
            if !seen.contains(&pat) {
                seen.insert(pat.clone());
                pats.push(pat);
            }
        };
        match self.values_of_os("regexp") {
            None => {
                if self.values_of_os("file").is_none() {
                    if let Some(os_pat) = self.value_of_os("pattern") {
                        add(self.pattern_from_os_str(os_pat)?);
                    }
                }
            }
            Some(os_pats) => {
                for os_pat in os_pats {
                    add(self.pattern_from_os_str(os_pat)?);
                }
            }
        }
        if let Some(paths) = self.values_of_os("file") {
            for path in paths {
                if path == "-" {
                    let it = cli::patterns_from_stdin()?
                        .into_iter()
                        .map(|p| self.pattern_from_string(p));
                    for pat in it {
                        add(pat);
                    }
                } else {
                    let it = cli::patterns_from_path(path)?
                        .into_iter()
                        .map(|p| self.pattern_from_string(p));
                    for pat in it {
                        add(pat);
                    }
                }
            }
        }
        Ok(pats)
    }

    /// Converts an OsStr pattern to a String pattern. The pattern is escaped
    /// if -F/--fixed-strings is set.
    ///
    /// If the pattern is not valid UTF-8, then an error is returned.
    fn pattern_from_os_str(&self, pat: &OsStr) -> Result<String> {
        let s = cli::pattern_from_os(pat)?;
        Ok(self.pattern_from_str(s))
    }

    /// Converts a &str pattern to a String pattern. The pattern is escaped
    /// if -F/--fixed-strings is set.
    fn pattern_from_str(&self, pat: &str) -> String {
        self.pattern_from_string(pat.to_string())
    }

    /// Applies additional processing on the given pattern if necessary
    /// (such as escaping meta characters or turning it into a line regex).
    fn pattern_from_string(&self, pat: String) -> String {
        if pat.is_empty() {
            // This would normally just be an empty string, which works on its
            // own, but if the patterns are joined in a set of alternations,
            // then you wind up with `foo|`, which is currently invalid in
            // Rust's regex engine.
            "(?:)".to_string()
        } else {
            pat
        }
    }

    /// Returns the preprocessor command if one was specified.
    fn preprocessor(&self) -> Option<PathBuf> {
        let path = match self.value_of_os("pre") {
            None => return None,
            Some(path) => path,
        };
        if path.is_empty() {
            return None;
        }
        Some(Path::new(path).to_path_buf())
    }

    /// Builds the set of globs for filtering files to apply to the --pre
    /// flag. If no --pre-globs are available, then this always returns an
    /// empty set of globs.
    fn preprocessor_globs(&self) -> Result<Override> {
        let globs = self.values_of_lossy_vec("pre-glob");
        if globs.is_empty() {
            return Ok(Override::empty());
        }
        let mut builder = OverrideBuilder::new(current_dir()?);
        for glob in globs {
            builder.add(&glob)?;
        }
        Ok(builder.build()?)
    }

    /// Parse the regex-size-limit argument option into a byte count.
    fn regex_size_limit(&self) -> Result<Option<usize>> {
        let r = self.parse_human_readable_size("regex-size-limit")?;
        u64_to_usize("regex-size-limit", r)
    }

    /// Returns the replacement string as UTF-8 bytes if it exists.
    fn replacement(&self) -> Option<Vec<u8>> {
        self.value_of_lossy("replace").map(|s| s.into_bytes())
    }

    /// Returns the sorting criteria based on command line parameters.
    fn sort_by(&self) -> Result<SortBy> {
        // For backcompat, continue supporting deprecated --sort-files flag.
        if self.is_present("sort-files") {
            return Ok(SortBy::asc(SortByKind::Path));
        }
        let sortby = match self.value_of_lossy("sort") {
            None => match self.value_of_lossy("sortr") {
                None => return Ok(SortBy::none()),
                Some(choice) => SortBy::desc(SortByKind::new(&choice)),
            },
            Some(choice) => SortBy::asc(SortByKind::new(&choice)),
        };
        Ok(sortby)
    }

    /// Returns true if and only if aggregate statistics for a search should
    /// be tracked.
    ///
    /// Generally, this is only enabled when explicitly requested by in the
    /// command line arguments via the --stats flag, but this can also be
    /// enabled implicitly via the output format, e.g., for JSON Lines.
    fn stats(&self) -> bool {
        self.output_kind() == OutputKind::JSON || self.is_present("stats")
    }

    /// When the output format is `Summary`, this returns the type of summary
    /// output to show.
    ///
    /// This returns `None` if the output format is not `Summary`.
    fn summary_kind(&self) -> Option<SummaryKind> {
        let (count, count_matches) = self.counts();
        if self.is_present("quiet") {
            Some(SummaryKind::Quiet)
        } else if count_matches {
            Some(SummaryKind::CountMatches)
        } else if count {
            Some(SummaryKind::Count)
        } else if self.is_present("files-with-matches") {
            Some(SummaryKind::PathWithMatch)
        } else if self.is_present("files-without-match") {
            Some(SummaryKind::PathWithoutMatch)
        } else {
            None
        }
    }

    /// Return the number of threads that should be used for parallelism.
    fn threads(&self) -> Result<usize> {
        if self.sort_by()?.kind != SortByKind::None {
            return Ok(1);
        }
        let threads = self.usize_of("threads")?.unwrap_or(0);
        let available =
            std::thread::available_parallelism().map_or(1, |n| n.get());
        Ok(if threads == 0 { std::cmp::min(12, available) } else { threads })
    }

    /// Builds a file type matcher from the command line flags.
    fn types(&self) -> Result<Types> {
        let mut builder = TypesBuilder::new();
        builder.add_defaults();
        for ty in self.values_of_lossy_vec("type-clear") {
            builder.clear(&ty);
        }
        for def in self.values_of_lossy_vec("type-add") {
            builder.add_def(&def)?;
        }
        for ty in self.values_of_lossy_vec("type") {
            builder.select(&ty);
        }
        for ty in self.values_of_lossy_vec("type-not") {
            builder.negate(&ty);
        }
        builder.build().map_err(From::from)
    }

    /// Returns the number of times the `unrestricted` flag is provided.
    fn unrestricted_count(&self) -> u64 {
        self.occurrences_of("unrestricted")
    }

    /// Returns true if and only if Unicode mode should be enabled.
    fn unicode(&self) -> bool {
        // Unicode mode is enabled by default, so only disable it when
        // --no-unicode is given explicitly.
        !(self.is_present("no-unicode") || self.is_present("no-pcre2-unicode"))
    }

    /// Returns true if and only if file names containing each match should
    /// be emitted.
    fn with_filename(&self, paths: &[PathBuf]) -> bool {
        if self.is_present("no-filename") {
            false
        } else {
            let path_stdin = Path::new("-");
            self.is_present("with-filename")
                || self.is_present("vimgrep")
                || paths.len() > 1
                || paths
                    .get(0)
                    .map_or(false, |p| p != path_stdin && p.is_dir())
        }
    }
}

/// Lower level generic helper methods for teasing values out of clap.
impl ArgMatches {
    /// Like values_of_lossy, but returns an empty vec if the flag is not
    /// present.
    fn values_of_lossy_vec(&self, name: &str) -> Vec<String> {
        self.values_of_lossy(name).unwrap_or_else(Vec::new)
    }

    /// Safely reads an arg value with the given name, and if it's present,
    /// tries to parse it as a usize value.
    ///
    /// If the number is zero, then it is considered absent and `None` is
    /// returned.
    fn usize_of_nonzero(&self, name: &str) -> Result<Option<usize>> {
        let n = match self.usize_of(name)? {
            None => return Ok(None),
            Some(n) => n,
        };
        Ok(if n == 0 { None } else { Some(n) })
    }

    /// Safely reads an arg value with the given name, and if it's present,
    /// tries to parse it as a usize value.
    fn usize_of(&self, name: &str) -> Result<Option<usize>> {
        match self.value_of_lossy(name) {
            None => Ok(None),
            Some(v) => v.parse().map(Some).map_err(From::from),
        }
    }

    /// Parses an argument of the form `[0-9]+(KMG)?`.
    ///
    /// If the aforementioned format is not recognized, then this returns an
    /// error.
    fn parse_human_readable_size(
        &self,
        arg_name: &str,
    ) -> Result<Option<u64>> {
        let size = match self.value_of_lossy(arg_name) {
            None => return Ok(None),
            Some(size) => size,
        };
        Ok(Some(cli::parse_human_readable_size(&size)?))
    }
}

/// The following methods mostly dispatch to the underlying clap methods
/// directly. Methods that would otherwise get a single value will fetch all
/// values and return the last one. (Clap returns the first one.) We only
/// define the ones we need.
impl ArgMatches {
    fn is_present(&self, name: &str) -> bool {
        self.0.is_present(name)
    }

    fn occurrences_of(&self, name: &str) -> u64 {
        self.0.occurrences_of(name)
    }

    fn value_of_lossy(&self, name: &str) -> Option<String> {
        self.0.value_of_lossy(name).map(|s| s.into_owned())
    }

    fn values_of_lossy(&self, name: &str) -> Option<Vec<String>> {
        self.0.values_of_lossy(name)
    }

    fn value_of_os(&self, name: &str) -> Option<&OsStr> {
        self.0.value_of_os(name)
    }

    fn values_of_os(&self, name: &str) -> Option<clap::OsValues<'_>> {
        self.0.values_of_os(name)
    }
}

/// Inspect an error resulting from building a Rust regex matcher, and if it's
/// believed to correspond to a syntax error that another engine could handle,
/// then add a message to suggest the use of the engine flag.
fn suggest(msg: String) -> String {
    if let Some(pcre_msg) = suggest_pcre2(&msg) {
        return pcre_msg;
    }
    msg
}

/// Inspect an error resulting from building a Rust regex matcher, and if it's
/// believed to correspond to a syntax error that PCRE2 could handle, then
/// add a message to suggest the use of -P/--pcre2.
fn suggest_pcre2(msg: &str) -> Option<String> {
    #[cfg(feature = "pcre2")]
    fn suggest(msg: &str) -> Option<String> {
        if !msg.contains("backreferences") && !msg.contains("look-around") {
            None
        } else {
            Some(format!(
                "{}

Consider enabling PCRE2 with the --pcre2 flag, which can handle backreferences
and look-around.",
                msg
            ))
        }
    }

    #[cfg(not(feature = "pcre2"))]
    fn suggest(_: &str) -> Option<String> {
        None
    }

    suggest(msg)
}

fn suggest_multiline(msg: String) -> String {
    if msg.contains("the literal") && msg.contains("not allowed") {
        format!(
            "{}

Consider enabling multiline mode with the --multiline flag (or -U for short).
When multiline mode is enabled, new line characters can be matched.",
            msg
        )
    } else {
        msg
    }
}

/// Convert the result of parsing a human readable file size to a `usize`,
/// failing if the type does not fit.
fn u64_to_usize(arg_name: &str, value: Option<u64>) -> Result<Option<usize>> {
    use std::usize;

    let value = match value {
        None => return Ok(None),
        Some(value) => value,
    };
    if value <= usize::MAX as u64 {
        Ok(Some(value as usize))
    } else {
        Err(From::from(format!("number too large for {}", arg_name)))
    }
}

/// Sorts by an optional parameter.
//
/// If parameter is found to be `None`, both entries compare equal.
fn sort_by_option<T: Ord>(
    p1: &Option<T>,
    p2: &Option<T>,
    reverse: bool,
) -> std::cmp::Ordering {
    match (p1, p2, reverse) {
        (Some(p1), Some(p2), true) => p1.cmp(&p2).reverse(),
        (Some(p1), Some(p2), false) => p1.cmp(&p2),
        _ => std::cmp::Ordering::Equal,
    }
}

/// Returns a clap matches object if the given arguments parse successfully.
///
/// Otherwise, if an error occurred, then it is returned unless the error
/// corresponds to a `--help` or `--version` request. In which case, the
/// corresponding output is printed and the current process is exited
/// successfully.
fn clap_matches<I, T>(args: I) -> Result<clap::ArgMatches<'static>>
where
    I: IntoIterator<Item = T>,
    T: Into<OsString> + Clone,
{
    let err = match app::app().get_matches_from_safe(args) {
        Ok(matches) => return Ok(matches),
        Err(err) => err,
    };
    if err.use_stderr() {
        return Err(err.into());
    }
    // Explicitly ignore any error returned by write!. The most likely error
    // at this point is a broken pipe error, in which case, we want to ignore
    // it and exit quietly.
    //
    // (This is the point of this helper function. clap's functionality for
    // doing this will panic on a broken pipe error.)
    let _ = write!(io::stdout(), "{}", err);
    std::process::exit(0);
}

/// Attempts to discover the current working directory. This mostly just defers
/// to the standard library, however, such things will fail if ripgrep is in
/// a directory that no longer exists. We attempt some fallback mechanisms,
/// such as querying the PWD environment variable, but otherwise return an
/// error.
fn current_dir() -> Result<PathBuf> {
    let err = match env::current_dir() {
        Err(err) => err,
        Ok(cwd) => return Ok(cwd),
    };
    if let Some(cwd) = env::var_os("PWD") {
        if !cwd.is_empty() {
            return Ok(PathBuf::from(cwd));
        }
    }
    Err(format!(
        "failed to get current working directory: {} \
         --- did your CWD get deleted?",
        err,
    )
    .into())
}

/// Retrieves the hostname that ripgrep should use wherever a hostname is
/// required. Currently, that's just in the hyperlink format.
///
/// This works by first running the given binary program (if present and with
/// no arguments) to get the hostname after trimming leading and trailing
/// whitespace. If that fails for any reason, then it falls back to getting
/// the hostname via platform specific means (e.g., `gethostname` on Unix).
///
/// The purpose of `bin` is to make it possible for end users to override how
/// ripgrep determines the hostname.
fn hostname(bin: Option<&OsStr>) -> Option<String> {
    let Some(bin) = bin else { return platform_hostname() };
    let bin = match grep::cli::resolve_binary(bin) {
        Ok(bin) => bin,
        Err(err) => {
            log::debug!(
                "failed to run command '{bin:?}' to get hostname \
                 (falling back to platform hostname): {err}",
            );
            return platform_hostname();
        }
    };
    let mut cmd = std::process::Command::new(&bin);
    cmd.stdin(std::process::Stdio::null());
    let rdr = match grep::cli::CommandReader::new(&mut cmd) {
        Ok(rdr) => rdr,
        Err(err) => {
            log::debug!(
                "failed to spawn command '{bin:?}' to get \
                 hostname (falling back to platform hostname): {err}",
            );
            return platform_hostname();
        }
    };
    let out = match io::read_to_string(rdr) {
        Ok(out) => out,
        Err(err) => {
            log::debug!(
                "failed to read output from command '{bin:?}' to get \
                 hostname (falling back to platform hostname): {err}",
            );
            return platform_hostname();
        }
    };
    let hostname = out.trim();
    if hostname.is_empty() {
        log::debug!(
            "output from command '{bin:?}' is empty after trimming \
             leading and trailing whitespace (falling back to \
             platform hostname)",
        );
        return platform_hostname();
    }
    Some(hostname.to_string())
}

/// Attempts to get the hostname by using platform specific routines. For
/// example, this will do `gethostname` on Unix and `GetComputerNameExW` on
/// Windows.
fn platform_hostname() -> Option<String> {
    let hostname_os = match grep::cli::hostname() {
        Ok(x) => x,
        Err(err) => {
            log::debug!("could not get hostname: {}", err);
            return None;
        }
    };
    let Some(hostname) = hostname_os.to_str() else {
        log::debug!(
            "got hostname {:?}, but it's not valid UTF-8",
            hostname_os
        );
        return None;
    };
    Some(hostname.to_string())
}

/// Returns a value that is meant to fill in the `{wslprefix}` variable for
/// a user given hyperlink format. A WSL prefix is a share/network like thing
/// that is meant to permit Windows applications to open files stored within
/// a WSL drive.
///
/// If a WSL distro name is unavailable, not valid UTF-8 or this isn't running
/// in a Unix environment, then this returns None.
///
/// See: <https://learn.microsoft.com/en-us/windows/wsl/filesystems>
fn wsl_prefix() -> Option<String> {
    if !cfg!(unix) {
        return None;
    }
    let distro_os = env::var_os("WSL_DISTRO_NAME")?;
    let Some(distro) = distro_os.to_str() else {
        log::debug!(
            "found WSL_DISTRO_NAME={:?}, but value is not UTF-8",
            distro_os
        );
        return None;
    };
    Some(format!("wsl$/{distro}"))
}

/// Tries to assign a timestamp to every `Subject` in the vector to help with
/// sorting Subjects by time.
fn load_timestamps<G>(
    subjects: impl Iterator<Item = Subject>,
    get_time: G,
) -> Vec<(Option<std::time::SystemTime>, Subject)>
where
    G: Fn(&std::fs::Metadata) -> io::Result<std::time::SystemTime>,
{
    subjects
        .map(|s| (s.path().metadata().and_then(|m| get_time(&m)).ok(), s))
        .collect()
}