rlucene 0.1.0-alpha.3

A Rust port of Apache Lucene, a full-text search engine library
Documentation
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
/*
 * Licensed to the Apache Software Foundation (ASF) under one or more
 * contributor license agreements.  See the NOTICE file distributed with
 * this work for additional information regarding copyright ownership.
 * The ASF licenses this file to You under the Apache License, Version 2.0
 * (the "License"); you may not use this file except in compliance with
 * the License.  You may obtain a copy of the License at
 *
 *     http://www.apache.org/licenses/LICENSE-2.0
 *
 * Unless required by applicable law or agreed to in writing, software
 * distributed under the License is distributed on an "AS IS" BASIS,
 * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
 * See the License for the specific language governing permissions and
 * limitations under the License.
 */
use crate::core::codecs::Codec;
use crate::core::codecs::doc_values_consumer::DocValuesConsumer;
use crate::core::codecs::doc_values_format::DocValuesFormat;
use crate::core::codecs::doc_values_producer::DocValuesProducer;
use crate::core::codecs::dummy::dummy_binary_doc_values::DummyBinaryDocValues;
use crate::core::codecs::dummy::dummy_doc_values_skipper::DummyDocValuesSkipper;
use crate::core::codecs::dummy::dummy_numeric_doc_values::DummyNumericDocValues;
use crate::core::codecs::dummy::dummy_sorted_doc_values::DummySortedDocValues;
use crate::core::codecs::dummy::dummy_sorted_numeric_doc_values::DummySortedNumericDocValues;
use crate::core::codecs::dummy::dummy_sorted_set_doc_values::DummySortedSetDocValues;
use crate::core::codecs::field_infos_format::FieldInfosFormat;
use crate::core::index::binary_doc_values::BinaryDocValues;
use crate::core::index::doc_values_field_updates::{
  BinaryDocValuesDVFU, DocValuesFieldIterator, DocValuesFieldIteratorEnum,
  DocValuesFieldUpdatesEnum, MergedFieldIterator, MergedIterator, NumericDocValuesDVFU,
};
use crate::core::index::doc_values_field_updates::{merged_field_iterator, merged_iterator};
use crate::core::index::doc_values_iterator::DocValuesIterator;
use crate::core::index::doc_values_type::DocValuesType;
use crate::core::index::field_info::FieldInfo;
use crate::core::index::field_infos::{FieldInfos, FieldNumbersLock};
use crate::core::index::index_reader::IndexReader;
use crate::core::index::leaf_reader::LeafReader;
use crate::core::index::merge_policy::{MergePolicy, MergeReader, MergeReaderSR};
use crate::core::index::numeric_doc_values::NumericDocValues;
use crate::core::index::pending_deletes::{DocBits, PendingDeletesBase, PendingDeletesEnum};
use crate::core::index::segment_commit_info::SegmentCommitInfo;
use crate::core::index::segment_reader::{DefaultLeafReader, SegmentReader};
use crate::core::index::segment_write_state::SegmentWriteState;
use crate::core::index::sorter::DocMapImpl;
use crate::core::index::{BytesRef, pending_soft_deletes};
use crate::core::search::doc_id_set_iterator::DocIdSetIterator;
use crate::core::search::doc_id_set_iterator::DocIdSetIteratorEnum2;
use crate::core::search::doc_id_set_iterator::NO_MORE_DOCS;
use crate::core::store::IO_CONTEXT_DEFAULT;
use crate::core::store::IOContext;
use crate::core::store::directory::Directory;
use crate::core::store::flush_info::FlushInfo;
use crate::core::store::tracking_directory_wrapper::TrackingDirectoryWrapper;
use crate::core::util::IOUtils;
use crate::core::util::accountable::Accountable;
use crate::core::util::close::{Closeable, CloseableRef};
use crate::core::util::error::lucene_error::{LuceneError, Result};
use crate::core::util::function::Function;
use crate::core::util::info_stream::{InfoStream, NO_OUTPUT_INFO_STREAM};
#[cfg(test)]
use crate::test_framework::core::util::failure_context::{
  ExecutionMethod, ExecutionOwner, ExecutionScope,
};
use parking_lot::Mutex;
use std::collections::{HashMap, HashSet};
use std::fmt;
use std::fmt::{Display, Formatter};
use std::sync::Arc;
use std::sync::atomic::{AtomicI32, AtomicI64, Ordering};

type DocValuesUpdateList = smallvec::SmallVec<[Arc<DocValuesFieldUpdatesEnum>; 1]>;

/// Used by IndexWriter to hold open SegmentReaders (for
/// searching or merging), plus pending deletes and updates,
/// for a given segment
pub(crate) struct ReadersAndUpdates<D>
where
  D: Directory,
{
  // Tracks how many consumers are using this instance:
  ref_count: AtomicI32, // starts at 1
  // the major version this index was created with
  pub(crate) index_created_version_major: i32,
  // Only set if there are doc values updates against this segment, and the index is sorted:
  pub(crate) sort_map: Option<Arc<DocMapImpl>>,
  pub(crate) ram_bytes_used: AtomicI64,
  pub(crate) inner: Mutex<Inner<D>>,
  pub(crate) info_id: Arc<str>,
}

pub(crate) struct Inner<D>
where
  D: Directory,
{
  // Set once (None, and then maybe set, and never set again):
  pub(crate) reader: Option<DefaultLeafReader<D>>,
  // How many further deletions we've done against
  // liveDocs vs when we loaded it or last wrote it:
  pub(crate) pending_deletes: PendingDeletesEnum,
  // Indicates whether this segment is currently being merged. While a segment
  // is merging, all field updates are also registered in the
  // mergingDVUpdates map. Also, calls to writeFieldUpdates merge the
  // updates with mergingDVUpdates.
  // That way, when the segment is done merging, IndexWriter can apply the
  // updates on the merged segment too.
  is_merging: bool,
  // Holds resolved (to docIDs) doc values updates that have not yet been
  // written to the index
  pending_dv_updates: HashMap<Arc<str>, DocValuesUpdateList>,
  // Holds resolved (to docIDs) doc values updates that were resolved while
  // this segment was being merged; at the end of the merge we carry over
  // these updates (remapping their docIDs) to the newly merged segment
  merging_dv_updates: HashMap<Arc<str>, DocValuesUpdateList>,
}

impl<D> ReadersAndUpdates<D>
where
  D: Directory,
{
  pub(crate) fn new(
    index_created_version_major: i32,
    info_id: Arc<str>,
    pending_deletes: PendingDeletesEnum,
  ) -> Self {
    let inner = Mutex::new(Inner {
      reader: None,
      pending_deletes,
      is_merging: false,
      pending_dv_updates: HashMap::new(),
      merging_dv_updates: HashMap::new(),
    });
    Self {
      ref_count: AtomicI32::new(1),
      index_created_version_major,
      sort_map: None,
      ram_bytes_used: AtomicI64::new(0),
      inner,
      info_id,
    }
  }
  /// Init from a previously opened SegmentReader.
  pub(crate) fn with_reader(
    index_created_version_major: i32,
    reader: DefaultLeafReader<D>,
    info: &SegmentCommitInfo<D>,
    pending_deletes: PendingDeletesEnum,
  ) -> Result<Self> {
    debug_assert!(reader.get_original_segment_info_id() == pending_deletes.get_info_id());
    let info_id = Arc::clone(&reader.original_si_id);
    let v = Self::new(index_created_version_major, info_id, pending_deletes);
    {
      let mut inner = v.inner.lock();
      inner.pending_deletes.on_new_reader(&reader, info)?;
      inner.reader = Some(reader);
    }
    Ok(v)
  }
  pub fn inc_ref(&self) {
    let rc = self.ref_count.fetch_add(1, Ordering::SeqCst) + 1;
    debug_assert!(rc > 1);
  }

  pub fn dec_ref(&self) {
    let rc = self.ref_count.fetch_sub(1, Ordering::SeqCst) - 1;
    debug_assert!(rc >= 0);
  }

  pub fn ref_count(&self) -> i32 {
    let rc = self.ref_count.load(Ordering::SeqCst);
    debug_assert!(rc >= 0);
    rc
  }
  pub(crate) fn get_del_count(&self, info: &SegmentCommitInfo<D>) -> i32
  where
    D: Directory,
  {
    self.inner.lock().pending_deletes.get_del_count(info)
  }

  fn assert_no_dup_gen(
    &self,
    field_updates: &[Arc<DocValuesFieldUpdatesEnum>],
    update: &DocValuesFieldUpdatesEnum,
  ) -> bool {
    let dup = field_updates
      .iter()
      .any(|old_update| old_update.del_gen == update.del_gen);
    debug_assert!(!dup, "duplicate delGen={}", update.del_gen);
    true
  }
  /// Adds a new resolved (meaning it maps docIDs to new values) doc values packet.
  /// We buffer these in RAM and write to disk when too much RAM is used or when a merge needs to kick off, or a commit/refresh.
  pub fn add_dv_update<T>(&self, update: T) -> Result<()>
  where
    T: Into<Arc<DocValuesFieldUpdatesEnum>>,
  {
    let update = update.into();
    let mut inner = self.inner.lock();
    if !update.get_finished() {
      return Err(LuceneError::illegal_argument("call finish first"));
    }

    let field = &update.field;
    let update_bytes = update.ram_bytes_used()?;
    let is_merging = inner.is_merging;

    let field_updates = if let Some(updates) = inner.pending_dv_updates.get_mut(field.as_ref()) {
      updates
    } else {
      inner
        .pending_dv_updates
        .entry(Arc::clone(field))
        .or_default()
    };

    debug_assert!(self.assert_no_dup_gen(field_updates, &update));
    self
      .ram_bytes_used
      .fetch_add(update_bytes, Ordering::SeqCst);

    if is_merging {
      field_updates.push(update.clone());
      let merging_updates = if let Some(updates) = inner.merging_dv_updates.get_mut(field.as_ref())
      {
        updates
      } else {
        inner
          .merging_dv_updates
          .entry(Arc::clone(field))
          .or_default()
      };
      merging_updates.push(update);
    } else {
      field_updates.push(update);
    }
    Ok(())
  }

  pub(crate) fn get_num_dv_updates(&self) -> usize {
    let inner = self.inner.lock();
    let mut count = 0usize;
    for updates in inner.pending_dv_updates.values() {
      count += updates.len();
    }
    count
  }

  pub fn release(&self, sr: &SegmentReader<D>, inner: Option<&Inner<D>>) -> Result<()> {
    let _inner = match inner {
      Some(inner) => inner,
      None => &mut *self.inner.lock(),
    };
    debug_assert!(self.info_id.as_ref() == sr.get_original_segment_info_id());
    sr.dec_ref()?;
    Ok(())
  }

  pub fn delete(
    &self,
    doc_id: i32,
    info: &SegmentCommitInfo<D>,
    inner: Option<&mut Inner<D>>,
  ) -> Result<bool> {
    let inner = match inner {
      Some(inner) => inner,
      None => &mut *self.inner.lock(),
    };
    if inner.reader.is_none() && inner.pending_deletes.must_init_on_delete() {
      get_reader(
        IO_CONTEXT_DEFAULT.as_ref().map_err(Clone::clone)?,
        info,
        inner,
        self.index_created_version_major,
      )?
      .dec_ref()?; // pass a reader to initialize the pending deletes
    }

    inner.pending_deletes.delete(doc_id, info)
  }
  pub fn drop_readers(&self) -> Result<()> {
    let mut inner = self.inner.lock();
    if let Some(reader) = inner.reader.take() {
      reader.dec_ref()?;
    }
    self.dec_ref();
    Ok(())
  }
  /// Returns a ref to a clone. NOTE: you should decRef() the reader when you're done (ie do not call close()).
  pub(crate) fn get_read_only_clone(
    &self,
    context: &IOContext,
    info: &SegmentCommitInfo<D>,
  ) -> Result<Option<DefaultLeafReader<D>>> {
    #[cfg(test)]
    let _execution_scope = ExecutionScope::enter(
      ExecutionOwner::ReadersAndUpdates,
      ExecutionMethod::GetReadOnlyClone,
    );
    let mut inner = self.inner.lock();
    if inner.reader.is_none() {
      get_reader(context, info, &mut inner, self.index_created_version_major)?.dec_ref()?;
      debug_assert!(inner.reader.is_some());
    }

    // force new liveDocs
    if let Some(live_docs) = inner.pending_deletes.get_live_docs() {
      let hard_live_docs = inner.pending_deletes.get_hard_live_docs();
      let sr = SegmentReader::new_from_reader(
        info,
        inner
          .reader
          .as_ref()
          .ok_or_else(|| LuceneError::illegal_state("reader was not initialized"))?,
        Some(live_docs),
        hard_live_docs,
        inner.pending_deletes.num_docs(info)?,
        true,
      )?;
      return Ok(Some(Arc::new(sr)));
    }
    {
      // liveDocs == None and reader != None. That can only be if there are no deletes
      let r = inner
        .reader
        .as_ref()
        .ok_or_else(|| LuceneError::illegal_state("reader was not initialized"))?;
      debug_assert!(r.get_live_docs()?.is_none());
      r.inc_ref()?;
      Ok(inner.reader.clone())
    }
  }

  pub(crate) fn num_deletes_to_merge<P>(
    &self,
    policy: &P,
    info: &SegmentCommitInfo<D>,
  ) -> Result<i32>
  where
    P: MergePolicy<D> + ?Sized,
  {
    let del_count = {
      let mut inner = self.inner.lock();
      let is_soft = matches!(inner.pending_deletes, PendingDeletesEnum::Soft(_));
      if is_soft {
        let dv_gen = inner.pending_deletes.dv_gen()?;
        let (reader, field_infos, on_new_reader) = if dv_gen == -2 {
          let field_infos = pending_soft_deletes::read_field_infos(info)?;
          let field_info = field_infos.field_info_by_name(inner.pending_deletes.field()?)?;
          let on_new_reader = pending_soft_deletes::do_on_new_reader(field_info);
          let reader = if on_new_reader {
            Some(self.get_latest_read(info, &mut inner, self.index_created_version_major)?)
          } else {
            None
          };
          (reader, Some(field_infos), on_new_reader)
        } else {
          (None, None, false)
        };
        inner.pending_deletes.num_deletes_to_merge(
          info,
          reader.as_ref(),
          field_infos,
          on_new_reader,
        )?;
      } else {
        inner
          .pending_deletes
          .num_deletes_to_merge(info, None, None, false)?;
      };
      inner.pending_deletes.get_del_count(info)
    };

    policy.num_deletes_to_merge(info, del_count, &|| {
      let mut inner = self.inner.lock();
      self.get_latest_read(info, &mut inner, self.index_created_version_major)
    })
  }

  /// Returns a snapshot of the live docs.
  pub fn get_live_docs(&self) -> Option<DocBits> {
    let mut inner = self.inner.lock();
    inner.pending_deletes.get_live_docs()
  }

  /// Returns the live-docs bits excluding documents that are not live due to soft-deletes.
  pub fn get_hard_live_docs(&self) -> Option<DocBits> {
    let mut inner = self.inner.lock();
    inner.pending_deletes.get_hard_live_docs()
  }
  pub fn drop_changes(&self) {
    // Discard (don't save) changes when we are dropping
    // the reader; this is used only on the sub-readers
    // after a successful merge.  If deletes had
    // accumulated on those sub-readers while the merge
    // is running, by now we have carried forward those
    // deletes onto the newly merged segment, so we can
    // discard them on the sub-readers:
    let mut inner = self.inner.lock();
    inner.pending_deletes.drop_changes();
    self.drop_merging_updates(Some(&mut inner));
  }
  // Commit live docs (writes new _X_N.del files) and field updates (writes new
  // _X_N updates files) to the directory; returns true if it wrote any file
  // and false if there were no new deletes or updates to write:
  pub fn write_live_docs<D1>(&self, dir: &D1, info: &mut SegmentCommitInfo<D>) -> Result<bool>
  where
    D: Directory,
    D1: Directory,
  {
    let mut inner = self.inner.lock();
    inner.pending_deletes.write_live_docs(dir, info)
  }

  #[allow(clippy::too_many_arguments)]
  pub fn handle_dv_updates<D1, F, IS>(
    &self,
    infos: &FieldInfos,
    dir: D1,
    dv_format: &F,
    inner: &mut Inner<D>,
    reader: &SegmentReader<D>,
    field_files: &mut HashMap<i32, HashSet<String>>,
    max_del_gen: i64,
    info_stream: &IS,
    info: &mut SegmentCommitInfo<D>,
  ) -> Result<()>
  where
    F: DocValuesFormat,
    D1: Directory,
    IS: InfoStream,
  {
    let mut updates_to_apply = smallvec::SmallVec::<[&Arc<DocValuesFieldUpdatesEnum>; 4]>::new();
    for (field, updates) in inner.pending_dv_updates.iter() {
      let ty = updates[0].type_;
      debug_assert!(
        matches!(ty, DocValuesType::Numeric | DocValuesType::Binary),
        "unsupported type: {:?}",
        ty
      );

      updates_to_apply.clear();
      let mut bytes: i64 = 0;

      for update in updates {
        if update.del_gen <= max_del_gen {
          // safe to apply this one
          bytes += update.ram_bytes_used()?;
          updates_to_apply.push(update);
        }
      }

      if updates_to_apply.is_empty() {
        // nothing to apply yet
        continue;
      }

      if info_stream.is_enabled("BD") {
        info_stream.message(
          "BD",
          &format!(
            "now write {} pending numeric DV updates for field={}, seg={}, bytes={:.3} MB",
            updates_to_apply.len(),
            field,
            info,
            (bytes as f64) / 1024.0 / 1024.0
          ),
        )?;
      }

      let next_doc_values_gen = info.get_next_doc_values_gen();
      let segment_suffix =
        crate::core::index::index_file_names::base36(next_doc_values_gen, &mut [0; 14]).to_owned();
      let updates_context = IOContext::with_flush(FlushInfo::new(info.info.max_doc()?, bytes))?;

      let field_info = infos
        .field_info_by_name(field)?
        .ok_or_else(|| LuceneError::illegal_argument("fieldInfo is None"))?;
      field_info.set_doc_values_gen(next_doc_values_gen)?;

      let field_infos = Arc::new(FieldInfos::new(vec![field_info.clone()])?);

      let tracking_dir = TrackingDirectoryWrapper::new(&dir);

      let state = SegmentWriteState::with_suffix(
        Arc::clone(&NO_OUTPUT_INFO_STREAM),
        &tracking_dir,
        field_infos,
        &updates_context,
        segment_suffix,
      );

      {
        let mut fields_consumer = dv_format.fields_consumer(&state, &info.info)?;

        let write_result =
          std::panic::catch_unwind(std::panic::AssertUnwindSafe(|| -> Result<()> {
            let update_supplier = FunctionImpl::new(field_info, &updates_to_apply);

            inner.pending_deletes.on_doc_values_update(
              field_info,
              update_supplier.apply_legacy(field_info)?,
              info,
            )?;
            if ty == DocValuesType::Binary {
              let v = DocValuesProducerBinary::new(update_supplier, field, reader, field_info);
              fields_consumer.add_binary_field(&state, &info.info, field_info, &v)?
            } else {
              let v = DocValuesProducerNumeric::new(update_supplier, field, reader, field_info);
              fields_consumer.add_numeric_field(&state, &info.info, field_info, &v)?;
            }
            Ok(())
          }));
        let close_result =
          std::panic::catch_unwind(std::panic::AssertUnwindSafe(|| fields_consumer.close()));
        IOUtils::use_or_suppress_caught_result(write_result, close_result)?;
      }

      info.advance_doc_values_gen();
      debug_assert!(!field_files.contains_key(&field_info.number));
      field_files.insert(field_info.number, state.directory.take_created_files());
    }
    Ok(())
  }

  fn write_field_infos_gen<D1, F>(
    &self,
    field_infos: &FieldInfos,
    dir: D1,
    infos_format: &F,
    info: &mut SegmentCommitInfo<D>,
  ) -> Result<HashSet<String>>
  where
    F: FieldInfosFormat,
    D1: Directory,
  {
    let next_field_infos_gen = info.get_next_field_infos_gen();
    let mut suffix_digits = [0; 14];
    let segment_suffix =
      crate::core::index::index_file_names::base36(next_field_infos_gen, &mut suffix_digits);
    // we write approximately that many bytes (based on Lucene46DVF):
    // HEADER + FOOTER: 40
    // 90 bytes per-field (over estimating long name and attributes map)
    let est_infos_size = 40 + 90 * (field_infos.size() as i64);
    // IOContext for a flush with estimated size
    let flush_info = FlushInfo::new(info.info.max_doc()?, est_infos_size);
    let infos_context = IOContext::with_flush(flush_info)?;
    // separately also track which files were created for this gen
    let tracking_dir = TrackingDirectoryWrapper::new(dir);
    infos_format.write(
      &tracking_dir,
      &info.info,
      segment_suffix,
      field_infos,
      &infos_context,
    )?;
    info.advance_field_infos_gen();
    Ok(tracking_dir.take_created_files())
  }
  pub fn write_field_updates<D1, IS>(
    &self,
    dir: D1,
    field_numbers: &FieldNumbersLock,
    max_del_gen: i64,
    info_stream: &IS,
    info: Option<&mut SegmentCommitInfo<D>>,
  ) -> Result<bool>
  where
    D1: Directory,
    IS: InfoStream,
  {
    let mut inner = self.inner.lock();
    let start_time_ns = std::time::Instant::now();

    let mut any = false;
    'outer: for updates in inner.pending_dv_updates.values() {
      for update in updates {
        if update.del_gen <= max_del_gen && update.any() {
          any = true;
          break 'outer;
        }
      }
    }
    if !any {
      // no updates
      return Ok(false);
    }

    let info = info.ok_or_else(|| LuceneError::illegal_state("SegmentCommitInfo missing"))?;
    let mut new_dv_files = HashMap::with_capacity(inner.pending_dv_updates.len());
    let mut field_infos_files: Option<HashSet<String>> = None;
    let mut field_infos = FieldInfos::default();

    // Do this so we can delete any created files on
    // error; this saves all codecs from having to do it:
    let tracking_dir = TrackingDirectoryWrapper::new(&dir);

    let mut success = false;
    let result = std::panic::catch_unwind(std::panic::AssertUnwindSafe(|| -> Result<()> {
      let codec = info.info.get_codec()?.clone();

      let reader = if let Some(reader) = inner.reader.as_ref() {
        Arc::clone(reader)
      } else {
        let reader = Arc::new(SegmentReader::new(
          info,
          self.index_created_version_major,
          &IOContext::read_once_io_context()?,
        )?);
        inner.pending_deletes.on_new_reader(&reader, info)?;
        reader
      };

      let result = std::panic::catch_unwind(std::panic::AssertUnwindSafe(|| -> Result<()> {
        // clone FieldInfos so that we can update their dvGen separately from
        // the reader's infos and write them to a new fieldInfos_gen file.
        let mut max_field_number: i32 = -1;
        let reader_field_infos = reader.get_field_infos()?;
        let mut updated_infos = Vec::with_capacity(reader_field_infos.size());

        for fi in reader_field_infos.iter() {
          // cannot use builder.add(fi) because it does not preserve
          // the local field number. Field numbers can be different from
          // the global ones if the segment was created externally (and added to
          // this index with `IndexWriter::add_indexes(Directory)`.)
          updated_infos.push(Arc::new(clone_field_info(fi, fi.number)?));
          max_field_number = max_field_number.max(fi.number);
        }

        // create new fields with the right DV type for updates whose field doesn't yet exist
        for updates in inner.pending_dv_updates.values() {
          if let Some(update) = updates.first() {
            let field = &update.field;
            if let Some(fi) = reader_field_infos.field_info_by_name(field)? {
              // the field already exists in this segment
              debug_assert_eq!(*fi.get_doc_values_type(), update.type_);
            } else {
              // the field is not present in this segment so we clone the global field
              // (which is guaranteed to exist) and remaps its field number locally.
              let fi = field_numbers.lock().construct_field_info(
                field,
                update.type_,
                max_field_number + 1,
              )?;
              if let Some(fi) = fi {
                max_field_number += 1;
                updated_infos.push(Arc::new(fi));
              } else {
                debug_assert!(false);
              }
            }
          }
        }

        field_infos = FieldInfos::new(updated_infos)?;

        let dv_format = codec.doc_values_format();

        self.handle_dv_updates(
          &field_infos,
          &tracking_dir,
          &dv_format,
          &mut inner,
          &reader,
          &mut new_dv_files,
          max_del_gen,
          info_stream,
          info,
        )?;

        let files = self.write_field_infos_gen(
          &field_infos,
          &tracking_dir,
          &codec.field_infos_format(),
          info,
        )?;
        field_infos_files = Some(files);

        Ok(())
      }));

      if !matches!(
        inner.reader.as_ref(),
        Some(pooled_reader) if Arc::ptr_eq(&reader, pooled_reader)
      ) {
        reader.close()?;
      }

      unwrap_caught_result!(result)?;
      success = true;
      Ok(())
    }));

    if !success {
      info.advance_next_write_field_infos_gen();
      info.advance_next_write_doc_values_gen();
      IOUtils::delete_files_ignoring_exceptions(&dir, tracking_dir.take_created_files());
    }
    unwrap_caught_result!(result)?;
    // Prune the now-written DV updates:
    let mut bytes_freed: i64 = 0;
    for updates in inner.pending_dv_updates.values_mut() {
      let mut upto = 0;
      for read in 0..updates.len() {
        if updates[read].del_gen > max_del_gen {
          updates.swap(upto, read);
          upto += 1;
        } else {
          bytes_freed += updates[read].ram_bytes_used()?;
        }
      }
      updates.truncate(upto);
    }
    inner
      .pending_dv_updates
      .retain(|_, updates| !updates.is_empty());

    let prev = self.ram_bytes_used.fetch_sub(bytes_freed, Ordering::SeqCst);
    let bytes_now = prev - bytes_freed;
    debug_assert!(bytes_now >= 0, "ram_bytes_used should not go negative");
    // writing field updates succeeded
    debug_assert!(field_infos_files.is_some());
    let field_infos_files = field_infos_files
      .take()
      .ok_or_else(|| LuceneError::illegal_state("field infos files were not written"))?;
    info.set_owned_field_infos_files(field_infos_files);
    // update the doc-values updates files. the files map each field to its set
    // of files, hence we copy from the existing map all fields w/ updates that
    // were not updated in this session, and add new mappings for fields that
    // were updated now.
    debug_assert!(!new_dv_files.is_empty());

    let existing_dv_files = info.take_doc_values_updates_files();
    for (field_num, files_set) in existing_dv_files {
      new_dv_files.entry(field_num).or_insert(files_set);
    }
    info.set_doc_values_updates_files(std::borrow::Cow::Owned(new_dv_files));
    // if there is a reader open, reopen it to reflect the updates
    if inner.reader.is_some() {
      swap_new_reader_with_latest_live_docs(&mut inner, info)?;
    }
    if info_stream.is_enabled("BD") {
      let message = format!(
        "done write field updates for seg={}; took {:.3}s; new files: {:?}",
        info,
        start_time_ns.elapsed().as_secs_f64(),
        info.get_doc_values_updates_files(),
      );
      info_stream.message("BD", &message)?;
    }
    Ok(true)
  }

  pub(crate) fn set_is_merging(&self) {
    let mut inner = self.inner.lock();
    if !inner.is_merging {
      inner.is_merging = true;
      debug_assert!(inner.merging_dv_updates.is_empty());
    }
  }

  #[cfg(test)]
  pub(crate) fn is_merging(&self) -> bool {
    let inner = self.inner.lock();
    inner.is_merging
  }

  /// Returns a reader for merge, with the latest doc values updates and deletions.
  pub(crate) fn get_reader_for_merge<F>(
    &self,
    context: &IOContext,
    info: &SegmentCommitInfo<D>,
    reader_consumer: F,
  ) -> Result<MergeReaderSR<D>>
  where
    F: FnOnce(&MergeReaderSR<D>) -> Result<()>,
  {
    // We must carry over any still-pending DV updates because they were not
    // successfully written, e.g. because there was a hole in the delGens,
    // or they arrived after we wrote all DVs for merge but before we set
    // isMerging here:
    let mut inner = self.inner.lock();
    let Inner {
      pending_dv_updates,
      merging_dv_updates,
      ..
    } = &mut *inner;

    for (field, updates) in pending_dv_updates.iter() {
      let entry = match merging_dv_updates.get_mut(field) {
        Some(entry) => {
          entry.reserve(updates.len());
          entry
        },
        None => merging_dv_updates
          .entry(field.clone())
          .or_insert_with(|| DocValuesUpdateList::with_capacity(updates.len())),
      };
      entry.extend(updates.iter().cloned());
    }

    let mut reader = get_reader(context, info, &mut inner, self.index_created_version_major)?;
    if inner.pending_deletes.needs_refresh(reader.as_ref(), info)?
      || reader.get_segment_info().get_del_gen() != info.get_del_gen()
    {
      // beware of zombies:
      debug_assert!(inner.pending_deletes.get_live_docs().is_some());
      reader = Arc::new(create_new_reader_with_latest_live_docs(
        &mut inner,
        Some(reader.as_ref()),
        info,
      )?);
    }

    debug_assert!(
      inner
        .pending_deletes
        .verify_doc_counts(reader.as_ref(), info)?
    );

    let merge_reader = MergeReader::new(reader, inner.pending_deletes.get_hard_live_docs());
    reader_consumer(&merge_reader)?;
    Ok(merge_reader)
  }

  /// Drops all merging updates.
  /// Called from IndexWriter after this segment finished merging (whether successfully or not).
  pub fn drop_merging_updates(&self, inner: Option<&mut Inner<D>>) {
    let inner = match inner {
      Some(inner) => inner,
      None => &mut *self.inner.lock(),
    };
    inner.merging_dv_updates.clear();
    inner.is_merging = false;
  }
  pub fn get_merging_dv_updates(&self) -> HashMap<Arc<str>, DocValuesUpdateList> {
    // We must atomically (in single sync'd block) clear isMerging when we return the DV updates
    // otherwise we can lose updates:
    let mut inner = self.inner.lock();
    inner.is_merging = false;
    inner.merging_dv_updates.clone()
  }
  pub fn is_fully_deleted(&self, info: &SegmentCommitInfo<D>) -> Result<bool> {
    let mut inner = self.inner.lock();
    let is_soft = matches!(inner.pending_deletes, PendingDeletesEnum::Soft(_));
    if is_soft {
      let dv_gen = inner.pending_deletes.dv_gen()?;
      let (reader, field_infos, on_new_reader) = if dv_gen == -2 {
        let field_infos = pending_soft_deletes::read_field_infos(info)?;
        let field_info = field_infos.field_info_by_name(inner.pending_deletes.field()?)?;
        let on_new_reader = pending_soft_deletes::do_on_new_reader(field_info);
        let reader = if on_new_reader {
          Some(self.get_latest_read(info, &mut inner, self.index_created_version_major)?)
        } else {
          None
        };
        (reader, Some(field_infos), on_new_reader)
      } else {
        (None, None, false)
      };
      inner
        .pending_deletes
        .is_fully_deleted(info, reader.as_ref(), field_infos, on_new_reader)
    } else {
      inner
        .pending_deletes
        .is_fully_deleted(info, None, None, false)
    }
  }
  pub(crate) fn keep_fully_deleted_segment<P>(
    &self,
    merge_policy: &P,
    info: &SegmentCommitInfo<D>,
  ) -> Result<bool>
  where
    P: MergePolicy<D> + ?Sized,
  {
    merge_policy.keep_fully_deleted_segment(|| {
      let mut inner = self.inner.lock();
      self.get_latest_read(info, &mut inner, self.index_created_version_major)
    })
  }
  pub(crate) fn get_info_id(&self) -> &str {
    &self.info_id
  }
  pub(crate) fn get_latest_read(
    &self,
    info: &SegmentCommitInfo<D>,
    inner: &mut Inner<D>,
    index_created_version_major: i32,
  ) -> Result<DefaultLeafReader<D>>
  where
    D: Directory,
  {
    if inner.reader.is_none() {
      // get a reader and dec the ref right away we just make sure we have a reader
      get_reader(
        IO_CONTEXT_DEFAULT.as_ref().map_err(Clone::clone)?,
        info,
        inner,
        index_created_version_major,
      )?
      .dec_ref()?;
    }
    // we should take the reader out of the struct temporarily, cause borrow check
    // it is safe take reader under lock because we put it back right away
    let needs_refresh = {
      let reader = inner
        .reader
        .as_ref()
        .ok_or_else(|| LuceneError::illegal_state("reader was not initialized"))?;
      inner.pending_deletes.needs_refresh(reader, info)?
    };
    if needs_refresh {
      // we have a reader but its live-docs are out of sync. let's create a temporary one that we
      // never share
      swap_new_reader_with_latest_live_docs(inner, info)?;
    }
    // put reader back and return a clone
    let reader = inner
      .reader
      .as_ref()
      .ok_or_else(|| LuceneError::illegal_state("reader was not initialized"))?;
    Ok(Arc::clone(reader))
  }
}

impl<D> Display for ReadersAndUpdates<D>
where
  D: Directory,
{
  fn fmt(&self, f: &mut Formatter<'_>) -> fmt::Result {
    let inner = self.inner.lock();
    write!(
      f,
      "ReadersAndLiveDocs(seg={}, PendingDeletesEnum={})",
      self.info_id, inner.pending_deletes
    )
  }
}
pub(crate) fn create_new_reader_with_latest_live_docs<'a, D>(
  inner: &'a mut Inner<D>, // The mutable borrow proves exclusive access to the writer state.
  reader: Option<&'a SegmentReader<D>>,
  info: &SegmentCommitInfo<D>,
) -> Result<SegmentReader<D>>
where
  D: Directory,
{
  let reader = match reader {
    Some(r) => r,
    None => inner
      .reader
      .as_ref()
      .ok_or_else(|| LuceneError::illegal_state("reader was not initialized"))?
      .as_ref(),
  };

  let new_reader = SegmentReader::new_from_reader(
    info,
    reader,
    inner.pending_deletes.get_live_docs(),
    inner.pending_deletes.get_hard_live_docs(),
    inner.pending_deletes.num_docs(info)?,
    true,
  )?;

  let mut success = false;
  let res = std::panic::catch_unwind(std::panic::AssertUnwindSafe(|| -> Result<()> {
    inner.pending_deletes.on_new_reader(&new_reader, info)?;
    reader.dec_ref()?;
    success = true;
    Ok(())
  }));

  if !success {
    new_reader.dec_ref()?;
  }
  unwrap_caught_result!(res)?;
  Ok(new_reader)
}
fn swap_new_reader_with_latest_live_docs<D>(
  inner: &mut Inner<D>,
  info: &SegmentCommitInfo<D>,
) -> Result<()>
where
  D: Directory,
{
  inner.reader = Some(Arc::new(create_new_reader_with_latest_live_docs(
    inner, None, info,
  )?));
  Ok(())
}
/// Returns a [`SegmentReader`].
pub(crate) fn get_reader<D>(
  context: &IOContext,
  info: &SegmentCommitInfo<D>,
  inner: &mut Inner<D>,
  index_created_version_major: i32,
) -> Result<DefaultLeafReader<D>>
where
  D: Directory,
{
  if inner.reader.is_none() {
    let reader = SegmentReader::new(info, index_created_version_major, context)?;
    inner.pending_deletes.on_new_reader(&reader, info)?;
    inner.reader = Some(Arc::new(reader));
  }
  // Ref for caller
  let reader = inner
    .reader
    .as_ref()
    .ok_or_else(|| LuceneError::illegal_state("reader was not initialized"))?;
  reader.inc_ref()?;
  Ok(reader.clone())
}

enum CurrentSource {
  OnDisk,
  Update,
}
/// Merges current on-disk doc values with incoming updates, giving the updates precedence.
/// in other words the values of the update always wins over the on-disk version.
struct MergedDocValues<DI> {
  // merged docID
  doc_id_out: i32,
  // docID from our original doc values
  doc_id_on_disk: i32,
  // docID from our updates
  update_doc_id: i32,

  on_disk_doc_values: Option<DI>,
  update_doc_values: DocIdSetIteratorEnum2<
    BinaryDocValuesDVFU<MergedFieldIterator>,
    NumericDocValuesDVFU<MergedFieldIterator>,
  >,
  current_values_supplier: Option<CurrentSource>,
}
impl<DI> MergedDocValues<DI> {
  pub fn new(
    on_disk_doc_values: Option<DI>,
    update_doc_values: DocIdSetIteratorEnum2<
      BinaryDocValuesDVFU<MergedFieldIterator>,
      NumericDocValuesDVFU<MergedFieldIterator>,
    >,
  ) -> Self {
    Self {
      doc_id_out: -1,
      doc_id_on_disk: -1,
      update_doc_id: -1,
      on_disk_doc_values,
      update_doc_values,
      current_values_supplier: None,
    }
  }
}
impl<DI> crate::core::search::doc_id_set_iterator::DocIdSetIteratorExtensions
  for MergedDocValues<DI>
where
  DI: DocValuesIterator,
{
}
impl<DI> crate::core::search::doc_id_set_iterator::BitSetIteratorAccess for MergedDocValues<DI> where
  DI: DocValuesIterator
{
}

impl<DI> DocIdSetIterator for MergedDocValues<DI>
where
  DI: DocValuesIterator,
{
  fn doc_id(&self) -> i32 {
    self.doc_id_out
  }

  fn next_doc(&mut self) -> Result<i32> {
    let mut has_value = false;

    while !has_value {
      if self.doc_id_on_disk == self.doc_id_out {
        match self.on_disk_doc_values.as_mut() {
          Some(dv) => {
            self.doc_id_on_disk = dv.next_doc()?;
          },
          None => {
            self.doc_id_on_disk = NO_MORE_DOCS;
          },
        }
      }

      if self.update_doc_id == self.doc_id_out {
        self.update_doc_id = self.update_doc_values.next_doc()?;
      }

      if self.doc_id_on_disk < self.update_doc_id {
        // no update to this doc - we use the on-disk values
        self.doc_id_out = self.doc_id_on_disk;
        self.current_values_supplier = Some(CurrentSource::OnDisk);
        has_value = true;
      } else {
        self.doc_id_out = self.update_doc_id;
        if self.doc_id_out != NO_MORE_DOCS {
          self.current_values_supplier = Some(CurrentSource::Update);
          has_value = match self.update_doc_values {
            DocIdSetIteratorEnum2::A(ref mut dv) => dv.iterator.has_value()?,
            DocIdSetIteratorEnum2::B(ref mut dv) => dv.iterator.has_value()?,
          };
        } else {
          has_value = true;
        }
      }
    }
    Ok(self.doc_id_out)
  }

  fn advance(&mut self, _target: i32) -> Result<i32> {
    Err(LuceneError::unsupported_operation(""))
  }

  fn cost(&self) -> Result<i64> {
    self
      .on_disk_doc_values
      .as_ref()
      .ok_or_else(|| LuceneError::illegal_state("on-disk doc values are not available"))?
      .cost()
  }
}

impl<DI> DocValuesIterator for MergedDocValues<DI>
where
  DI: DocValuesIterator,
{
  fn advance_exact(&mut self, _target: i32) -> Result<bool> {
    Err(LuceneError::unsupported_operation(""))
  }
}

struct BinaryDocValuesImpl<DI> {
  merged_doc_values: MergedDocValues<DI>,
}
impl<DI> BinaryDocValuesImpl<DI> {
  fn new(merged_doc_values: MergedDocValues<DI>) -> Self {
    Self { merged_doc_values }
  }
}

impl<DI> DocValuesIterator for BinaryDocValuesImpl<DI>
where
  DI: DocValuesIterator,
{
  fn advance_exact(&mut self, target: i32) -> Result<bool> {
    self.merged_doc_values.advance_exact(target)
  }
}

impl<DI> crate::core::search::doc_id_set_iterator::DocIdSetIteratorExtensions
  for BinaryDocValuesImpl<DI>
where
  DI: DocValuesIterator,
{
}
impl<DI> crate::core::search::doc_id_set_iterator::BitSetIteratorAccess for BinaryDocValuesImpl<DI> where
  DI: DocValuesIterator
{
}

impl<DI> DocIdSetIterator for BinaryDocValuesImpl<DI>
where
  DI: DocValuesIterator,
{
  fn doc_id(&self) -> i32 {
    self.merged_doc_values.doc_id()
  }

  fn next_doc(&mut self) -> Result<i32> {
    self.merged_doc_values.next_doc()
  }

  fn advance(&mut self, target: i32) -> Result<i32> {
    self.merged_doc_values.advance(target)
  }

  fn cost(&self) -> Result<i64> {
    self.merged_doc_values.cost()
  }
}

impl<DI> BinaryDocValues for BinaryDocValuesImpl<DI>
where
  DI: BinaryDocValues,
{
  type Value<'a>
    = crate::core::index::BytesRefValueEnum2<DI::Value<'a>, BytesRef<&'a [u8]>>
  where
    Self: 'a;

  fn binary_value(&self) -> Result<Self::Value<'_>> {
    match self.merged_doc_values.current_values_supplier {
      Some(CurrentSource::OnDisk) => {
        if let Some(dv) = &self.merged_doc_values.on_disk_doc_values {
          dv.binary_value()
            .map(crate::core::index::BytesRefValueEnum2::A)
        } else {
          Err(LuceneError::illegal_state(
            "no on-disk doc values available",
          ))
        }
      },
      Some(CurrentSource::Update) => match self.merged_doc_values.update_doc_values {
        DocIdSetIteratorEnum2::A(ref dv) => dv
          .binary_value()
          .map(crate::core::index::BytesRefValueEnum2::B),
        DocIdSetIteratorEnum2::B(_) => Err(LuceneError::illegal_state(
          "update doc values should be BinaryDocValuesDVFU",
        )),
      },
      None => Err(LuceneError::illegal_state("no current values supplier set")),
    }
  }
}
struct NumericDocValuesImpl<DI> {
  merged_doc_values: MergedDocValues<DI>,
}

impl<DI> NumericDocValuesImpl<DI> {
  fn new(merged_doc_values: MergedDocValues<DI>) -> Self {
    Self { merged_doc_values }
  }
}

impl<DI> DocValuesIterator for NumericDocValuesImpl<DI>
where
  DI: DocValuesIterator,
{
  fn advance_exact(&mut self, target: i32) -> Result<bool> {
    self.merged_doc_values.advance_exact(target)
  }
}

impl<DI> crate::core::search::doc_id_set_iterator::DocIdSetIteratorExtensions
  for NumericDocValuesImpl<DI>
where
  DI: DocValuesIterator,
{
}
impl<DI> crate::core::search::doc_id_set_iterator::BitSetIteratorAccess for NumericDocValuesImpl<DI> where
  DI: DocValuesIterator
{
}

impl<DI> DocIdSetIterator for NumericDocValuesImpl<DI>
where
  DI: DocValuesIterator,
{
  fn doc_id(&self) -> i32 {
    self.merged_doc_values.doc_id()
  }

  fn next_doc(&mut self) -> Result<i32> {
    self.merged_doc_values.next_doc()
  }

  fn advance(&mut self, target: i32) -> Result<i32> {
    self.merged_doc_values.advance(target)
  }

  fn cost(&self) -> Result<i64> {
    self.merged_doc_values.cost()
  }
}

impl<DI> NumericDocValues for NumericDocValuesImpl<DI>
where
  DI: NumericDocValues,
{
  fn long_value(&mut self) -> Result<i64> {
    match self.merged_doc_values.current_values_supplier {
      Some(CurrentSource::OnDisk) => {
        if let Some(dv) = &mut self.merged_doc_values.on_disk_doc_values {
          dv.long_value()
        } else {
          Err(LuceneError::illegal_state(
            "no on-disk doc values available",
          ))
        }
      },
      Some(CurrentSource::Update) => match self.merged_doc_values.update_doc_values {
        DocIdSetIteratorEnum2::A(_) => Err(LuceneError::illegal_state(
          "update doc values should be BinaryDocValuesDVFU",
        )),
        DocIdSetIteratorEnum2::B(ref mut dv) => dv.long_value(),
      },
      None => Err(LuceneError::illegal_state("no current values supplier set")),
    }
  }
}

struct DocValuesProducerBinary<'a, D>
where
  D: Directory,
{
  update_supplier: FunctionImpl<'a>,
  field: &'a str,
  reader: &'a SegmentReader<D>,
  field_info: &'a Arc<FieldInfo>,
}

impl<D> CloseableRef for DocValuesProducerBinary<'_, D> where D: Directory {}

impl<'a, D> DocValuesProducerBinary<'a, D>
where
  D: Directory,
{
  pub fn new(
    update_supplier: FunctionImpl<'a>,
    field: &'a str,
    reader: &'a SegmentReader<D>,
    field_info: &'a Arc<FieldInfo>,
  ) -> Self {
    Self {
      update_supplier,
      field,
      reader,
      field_info,
    }
  }
}

impl<'a, D> DocValuesProducer for DocValuesProducerBinary<'a, D>
where
  D: Directory,
{
  type NumericDocValues = DummyNumericDocValues;
  type BinaryDocValues = BinaryDocValuesImpl<<SegmentReader<D> as LeafReader>::BinaryDocValues>;

  fn get_binary(&self, _field: &Arc<FieldInfo>) -> Result<Self::BinaryDocValues> {
    let iterator = match self.update_supplier.apply(self.field_info)? {
      Some(it) => it,
      None => {
        return Err(LuceneError::illegal_argument(
          "iterator should never None here",
        ));
      },
    };
    let merged_doc_values = MergedDocValues::new(
      self.reader.get_binary_doc_values(self.field)?,
      DocIdSetIteratorEnum2::A(BinaryDocValuesDVFU::new(iterator)),
    );
    Ok(BinaryDocValuesImpl::new(merged_doc_values))
  }

  type SortedDocValues = DummySortedDocValues;
  type SortedNumericDocValues = DummySortedNumericDocValues;
  type SortedSetDocValues = DummySortedSetDocValues;
  type DocValuesSkipper = DummyDocValuesSkipper;
}
struct DocValuesProducerNumeric<'a, D>
where
  D: Directory,
{
  update_supplier: FunctionImpl<'a>,
  field: &'a str,
  reader: &'a SegmentReader<D>,
  field_info: &'a Arc<FieldInfo>,
}

impl<D> CloseableRef for DocValuesProducerNumeric<'_, D> where D: Directory {}

impl<'a, D> DocValuesProducerNumeric<'a, D>
where
  D: Directory,
{
  pub fn new(
    update_supplier: FunctionImpl<'a>,
    field: &'a str,
    reader: &'a SegmentReader<D>,
    field_info: &'a Arc<FieldInfo>,
  ) -> Self {
    Self {
      update_supplier,
      field,
      reader,
      field_info,
    }
  }
}
impl<'a, D> DocValuesProducer for DocValuesProducerNumeric<'a, D>
where
  D: Directory,
{
  type NumericDocValues = NumericDocValuesImpl<<SegmentReader<D> as LeafReader>::NumericDocValues>;
  fn get_numeric(&self, _field: &Arc<FieldInfo>) -> Result<Self::NumericDocValues> {
    let iterator = match self.update_supplier.apply(self.field_info)? {
      Some(it) => it,
      None => {
        return Err(LuceneError::illegal_argument(
          "iterator should never None here",
        ));
      },
    };

    let merged_doc_values = MergedDocValues::new(
      self.reader.get_numeric_doc_values(self.field)?,
      DocIdSetIteratorEnum2::B(NumericDocValuesDVFU::new(iterator)),
    );
    // Merge sort of the original doc values with updated doc values:
    Ok(NumericDocValuesImpl::new(merged_doc_values))
  }
  type BinaryDocValues = DummyBinaryDocValues;
  type SortedDocValues = DummySortedDocValues;
  type SortedNumericDocValues = DummySortedNumericDocValues;
  type SortedSetDocValues = DummySortedSetDocValues;

  type DocValuesSkipper = DummyDocValuesSkipper;
}

struct FunctionImpl<'a> {
  field_info: &'a Arc<FieldInfo>,
  updates_to_apply: &'a [&'a Arc<DocValuesFieldUpdatesEnum>],
}
impl<'a> FunctionImpl<'a> {
  fn new(
    field_info: &'a Arc<FieldInfo>,
    updates_to_apply: &'a [&'a Arc<DocValuesFieldUpdatesEnum>],
  ) -> Self {
    Self {
      field_info,
      updates_to_apply,
    }
  }
}
impl FunctionImpl<'_> {
  // Preserve the original carrier for pending (including soft) delete callbacks.
  fn apply_legacy(
    &self,
    info: &Arc<FieldInfo>,
  ) -> Result<Option<MergedIterator<DocValuesFieldIteratorEnum>>> {
    merged_iterator(self.iterators(info)?)
  }

  fn iterators(&self, info: &Arc<FieldInfo>) -> Result<Vec<DocValuesFieldIteratorEnum>> {
    if !Arc::ptr_eq(info, self.field_info) {
      return Err(LuceneError::illegal_argument(format!(
        "expected field info for field: {} but got: {}",
        self.field_info.name, info.name
      )));
    }

    let mut subs = Vec::with_capacity(self.updates_to_apply.len());
    for v in self.updates_to_apply {
      subs.push(v.iterator()?)
    }
    Ok(subs)
  }
}
impl Function<Arc<FieldInfo>, Option<MergedFieldIterator>> for FunctionImpl<'_> {
  fn apply(&self, info: &Arc<FieldInfo>) -> Result<Option<MergedFieldIterator>> {
    merged_field_iterator(self.iterators(info)?)
  }
}

fn clone_field_info(fi: &FieldInfo, field_number: i32) -> Result<FieldInfo> {
  FieldInfo::new(
    fi.name.to_string(),
    field_number,
    fi.has_term_vectors(),
    fi.omits_norms(),
    fi.has_payloads(),
    *fi.get_index_options(),
    *fi.get_doc_values_type(),
    *fi.doc_values_skip_index_type(),
    fi.get_doc_values_gen(),
    fi.attributes(),
    fi.get_point_dimension_count(),
    fi.get_point_index_dimension_count(),
    fi.get_point_num_bytes(),
    fi.get_vector_dimension(),
    *fi.get_vector_encoding(),
    *fi.get_vector_similarity_function(),
    fi.is_soft_deletes_field(),
    fi.is_parent_field(),
  )
}