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
// This file is part of Substrate.

// Copyright (C) Parity Technologies (UK) Ltd.
// SPDX-License-Identifier: Apache-2.0

// Licensed 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.

//! # Remote Externalities
//!
//! An equivalent of `sp_io::TestExternalities` that can load its state from a remote substrate
//! based chain, or a local state snapshot file.

use codec::{Compact, Decode, Encode};
use indicatif::{ProgressBar, ProgressStyle};
use jsonrpsee::{core::params::ArrayParams, http_client::HttpClient};
use log::*;
use serde::de::DeserializeOwned;
use sp_core::{
	hexdisplay::HexDisplay,
	storage::{
		well_known_keys::{is_default_child_storage_key, DEFAULT_CHILD_STORAGE_KEY_PREFIX},
		ChildInfo, ChildType, PrefixedStorageKey, StorageData, StorageKey,
	},
};
use sp_runtime::{
	traits::{Block as BlockT, HashingFor},
	StateVersion,
};
use sp_state_machine::TestExternalities;
use spinners::{Spinner, Spinners};
use std::{
	cmp::{max, min},
	fs,
	ops::{Deref, DerefMut},
	path::{Path, PathBuf},
	sync::Arc,
	time::{Duration, Instant},
};
use substrate_rpc_client::{rpc_params, BatchRequestBuilder, ChainApi, ClientT, StateApi};
use tokio_retry::{strategy::FixedInterval, Retry};

type KeyValue = (StorageKey, StorageData);
type TopKeyValues = Vec<KeyValue>;
type ChildKeyValues = Vec<(ChildInfo, Vec<KeyValue>)>;
type SnapshotVersion = Compact<u16>;

const LOG_TARGET: &str = "remote-ext";
const DEFAULT_HTTP_ENDPOINT: &str = "https://try-runtime.polkadot.io:443";
const SNAPSHOT_VERSION: SnapshotVersion = Compact(4);

/// The snapshot that we store on disk.
#[derive(Decode, Encode)]
struct Snapshot<B: BlockT> {
	snapshot_version: SnapshotVersion,
	state_version: StateVersion,
	// <Vec<Key, (Value, MemoryDbRefCount)>>
	raw_storage: Vec<(Vec<u8>, (Vec<u8>, i32))>,
	// The storage root of the state. This may vary from the storage root in the header, if not the
	// entire state was fetched.
	storage_root: B::Hash,
	header: B::Header,
}

impl<B: BlockT> Snapshot<B> {
	pub fn new(
		state_version: StateVersion,
		raw_storage: Vec<(Vec<u8>, (Vec<u8>, i32))>,
		storage_root: B::Hash,
		header: B::Header,
	) -> Self {
		Self {
			snapshot_version: SNAPSHOT_VERSION,
			state_version,
			raw_storage,
			storage_root,
			header,
		}
	}

	fn load(path: &PathBuf) -> Result<Snapshot<B>, &'static str> {
		let bytes = fs::read(path).map_err(|_| "fs::read failed.")?;
		// The first item in the SCALE encoded struct bytes is the snapshot version. We decode and
		// check that first, before proceeding to decode the rest of the snapshot.
		let snapshot_version = SnapshotVersion::decode(&mut &*bytes)
			.map_err(|_| "Failed to decode snapshot version")?;

		if snapshot_version != SNAPSHOT_VERSION {
			return Err("Unsupported snapshot version detected. Please create a new snapshot.")
		}

		Decode::decode(&mut &*bytes).map_err(|_| "Decode failed")
	}
}

/// An externalities that acts exactly the same as [`sp_io::TestExternalities`] but has a few extra
/// bits and pieces to it, and can be loaded remotely.
pub struct RemoteExternalities<B: BlockT> {
	/// The inner externalities.
	pub inner_ext: TestExternalities<HashingFor<B>>,
	/// The block header which we created this externality env.
	pub header: B::Header,
}

impl<B: BlockT> Deref for RemoteExternalities<B> {
	type Target = TestExternalities<HashingFor<B>>;
	fn deref(&self) -> &Self::Target {
		&self.inner_ext
	}
}

impl<B: BlockT> DerefMut for RemoteExternalities<B> {
	fn deref_mut(&mut self) -> &mut Self::Target {
		&mut self.inner_ext
	}
}

/// The execution mode.
#[derive(Clone)]
pub enum Mode<H> {
	/// Online. Potentially writes to a snapshot file.
	Online(OnlineConfig<H>),
	/// Offline. Uses a state snapshot file and needs not any client config.
	Offline(OfflineConfig),
	/// Prefer using a snapshot file if it exists, else use a remote server.
	OfflineOrElseOnline(OfflineConfig, OnlineConfig<H>),
}

impl<H> Default for Mode<H> {
	fn default() -> Self {
		Mode::Online(OnlineConfig::default())
	}
}

/// Configuration of the offline execution.
///
/// A state snapshot config must be present.
#[derive(Clone)]
pub struct OfflineConfig {
	/// The configuration of the state snapshot file to use. It must be present.
	pub state_snapshot: SnapshotConfig,
}

/// Description of the transport protocol (for online execution).
#[derive(Debug, Clone)]
pub enum Transport {
	/// Use the `URI` to open a new WebSocket connection.
	Uri(String),
	/// Use HTTP connection.
	RemoteClient(HttpClient),
}

impl Transport {
	fn as_client(&self) -> Option<&HttpClient> {
		match self {
			Self::RemoteClient(client) => Some(client),
			_ => None,
		}
	}

	// Build an HttpClient from a URI.
	async fn init(&mut self) -> Result<(), &'static str> {
		if let Self::Uri(uri) = self {
			log::debug!(target: LOG_TARGET, "initializing remote client to {:?}", uri);

			// If we have a ws uri, try to convert it to an http uri.
			// We use an HTTP client rather than WS because WS starts to choke with "accumulated
			// message length exceeds maximum" errors after processing ~10k keys when fetching
			// from a node running a default configuration.
			let uri = if uri.starts_with("ws://") {
				let uri = uri.replace("ws://", "http://");
				log::info!(target: LOG_TARGET, "replacing ws:// in uri with http://: {:?} (ws is currently unstable for fetching remote storage, for more see https://github.com/paritytech/jsonrpsee/issues/1086)", uri);
				uri
			} else if uri.starts_with("wss://") {
				let uri = uri.replace("wss://", "https://");
				log::info!(target: LOG_TARGET, "replacing wss:// in uri with https://: {:?} (ws is currently unstable for fetching remote storage, for more see https://github.com/paritytech/jsonrpsee/issues/1086)", uri);
				uri
			} else {
				uri.clone()
			};
			let http_client = HttpClient::builder()
				.max_request_size(u32::MAX)
				.max_response_size(u32::MAX)
				.request_timeout(std::time::Duration::from_secs(60 * 5))
				.build(uri)
				.map_err(|e| {
					log::error!(target: LOG_TARGET, "error: {:?}", e);
					"failed to build http client"
				})?;

			*self = Self::RemoteClient(http_client)
		}

		Ok(())
	}
}

impl From<String> for Transport {
	fn from(uri: String) -> Self {
		Transport::Uri(uri)
	}
}

impl From<HttpClient> for Transport {
	fn from(client: HttpClient) -> Self {
		Transport::RemoteClient(client)
	}
}

/// Configuration of the online execution.
///
/// A state snapshot config may be present and will be written to in that case.
#[derive(Clone)]
pub struct OnlineConfig<H> {
	/// The block hash at which to get the runtime state. Will be latest finalized head if not
	/// provided.
	pub at: Option<H>,
	/// An optional state snapshot file to WRITE to, not for reading. Not written if set to `None`.
	pub state_snapshot: Option<SnapshotConfig>,
	/// The pallets to scrape. These values are hashed and added to `hashed_prefix`.
	pub pallets: Vec<String>,
	/// Transport config.
	pub transport: Transport,
	/// Lookout for child-keys, and scrape them as well if set to true.
	pub child_trie: bool,
	/// Storage entry key prefixes to be injected into the externalities. The *hashed* prefix must
	/// be given.
	pub hashed_prefixes: Vec<Vec<u8>>,
	/// Storage entry keys to be injected into the externalities. The *hashed* key must be given.
	pub hashed_keys: Vec<Vec<u8>>,
}

impl<H: Clone> OnlineConfig<H> {
	/// Return rpc (http) client reference.
	fn rpc_client(&self) -> &HttpClient {
		self.transport
			.as_client()
			.expect("http client must have been initialized by now; qed.")
	}

	fn at_expected(&self) -> H {
		self.at.clone().expect("block at must be initialized; qed")
	}
}

impl<H> Default for OnlineConfig<H> {
	fn default() -> Self {
		Self {
			transport: Transport::from(DEFAULT_HTTP_ENDPOINT.to_owned()),
			child_trie: true,
			at: None,
			state_snapshot: None,
			pallets: Default::default(),
			hashed_keys: Default::default(),
			hashed_prefixes: Default::default(),
		}
	}
}

impl<H> From<String> for OnlineConfig<H> {
	fn from(t: String) -> Self {
		Self { transport: t.into(), ..Default::default() }
	}
}

/// Configuration of the state snapshot.
#[derive(Clone)]
pub struct SnapshotConfig {
	/// The path to the snapshot file.
	pub path: PathBuf,
}

impl SnapshotConfig {
	pub fn new<P: Into<PathBuf>>(path: P) -> Self {
		Self { path: path.into() }
	}
}

impl From<String> for SnapshotConfig {
	fn from(s: String) -> Self {
		Self::new(s)
	}
}

impl Default for SnapshotConfig {
	fn default() -> Self {
		Self { path: Path::new("SNAPSHOT").into() }
	}
}

/// Builder for remote-externalities.
#[derive(Clone)]
pub struct Builder<B: BlockT> {
	/// Custom key-pairs to be injected into the final externalities. The *hashed* keys and values
	/// must be given.
	hashed_key_values: Vec<KeyValue>,
	/// The keys that will be excluded from the final externality. The *hashed* key must be given.
	hashed_blacklist: Vec<Vec<u8>>,
	/// Connectivity mode, online or offline.
	mode: Mode<B::Hash>,
	/// If provided, overwrite the state version with this. Otherwise, the state_version of the
	/// remote node is used. All cache files also store their state version.
	///
	/// Overwrite only with care.
	overwrite_state_version: Option<StateVersion>,
}

impl<B: BlockT> Default for Builder<B> {
	fn default() -> Self {
		Self {
			mode: Default::default(),
			hashed_key_values: Default::default(),
			hashed_blacklist: Default::default(),
			overwrite_state_version: None,
		}
	}
}

// Mode methods
impl<B: BlockT> Builder<B> {
	fn as_online(&self) -> &OnlineConfig<B::Hash> {
		match &self.mode {
			Mode::Online(config) => config,
			Mode::OfflineOrElseOnline(_, config) => config,
			_ => panic!("Unexpected mode: Online"),
		}
	}

	fn as_online_mut(&mut self) -> &mut OnlineConfig<B::Hash> {
		match &mut self.mode {
			Mode::Online(config) => config,
			Mode::OfflineOrElseOnline(_, config) => config,
			_ => panic!("Unexpected mode: Online"),
		}
	}
}

// RPC methods
impl<B: BlockT> Builder<B>
where
	B::Hash: DeserializeOwned,
	B::Header: DeserializeOwned,
{
	const PARALLEL_REQUESTS: usize = 4;
	const BATCH_SIZE_INCREASE_FACTOR: f32 = 1.10;
	const BATCH_SIZE_DECREASE_FACTOR: f32 = 0.50;
	const REQUEST_DURATION_TARGET: Duration = Duration::from_secs(15);
	const INITIAL_BATCH_SIZE: usize = 10;
	// nodes by default will not return more than 1000 keys per request
	const DEFAULT_KEY_DOWNLOAD_PAGE: u32 = 1000;
	const MAX_RETRIES: usize = 12;
	const KEYS_PAGE_RETRY_INTERVAL: Duration = Duration::from_secs(5);

	async fn rpc_get_storage(
		&self,
		key: StorageKey,
		maybe_at: Option<B::Hash>,
	) -> Result<Option<StorageData>, &'static str> {
		trace!(target: LOG_TARGET, "rpc: get_storage");
		self.as_online().rpc_client().storage(key, maybe_at).await.map_err(|e| {
			error!(target: LOG_TARGET, "Error = {:?}", e);
			"rpc get_storage failed."
		})
	}

	/// Get the latest finalized head.
	async fn rpc_get_head(&self) -> Result<B::Hash, &'static str> {
		trace!(target: LOG_TARGET, "rpc: finalized_head");

		// sadly this pretty much unreadable...
		ChainApi::<(), _, B::Header, ()>::finalized_head(self.as_online().rpc_client())
			.await
			.map_err(|e| {
				error!(target: LOG_TARGET, "Error = {:?}", e);
				"rpc finalized_head failed."
			})
	}

	async fn get_keys_single_page(
		&self,
		prefix: Option<StorageKey>,
		start_key: Option<StorageKey>,
		at: B::Hash,
	) -> Result<Vec<StorageKey>, &'static str> {
		self.as_online()
			.rpc_client()
			.storage_keys_paged(prefix, Self::DEFAULT_KEY_DOWNLOAD_PAGE, start_key, Some(at))
			.await
			.map_err(|e| {
				error!(target: LOG_TARGET, "Error = {:?}", e);
				"rpc get_keys failed"
			})
	}

	/// Get keys with `prefix` at `block` in a parallel manner.
	async fn rpc_get_keys_parallel(
		&self,
		prefix: &StorageKey,
		block: B::Hash,
		parallel: usize,
	) -> Result<Vec<StorageKey>, &'static str> {
		/// Divide the workload and return the start key of each chunks. Guaranteed to return a
		/// non-empty list.
		fn gen_start_keys(prefix: &StorageKey) -> Vec<StorageKey> {
			let mut prefix = prefix.as_ref().to_vec();
			let scale = 32usize.saturating_sub(prefix.len());

			// no need to divide workload
			if scale < 9 {
				prefix.extend(vec![0; scale]);
				return vec![StorageKey(prefix)]
			}

			let chunks = 16;
			let step = 0x10000 / chunks;
			let ext = scale - 2;

			(0..chunks)
				.map(|i| {
					let mut key = prefix.clone();
					let start = i * step;
					key.extend(vec![(start >> 8) as u8, (start & 0xff) as u8]);
					key.extend(vec![0; ext]);
					StorageKey(key)
				})
				.collect()
		}

		let start_keys = gen_start_keys(&prefix);
		let start_keys: Vec<Option<&StorageKey>> = start_keys.iter().map(Some).collect();
		let mut end_keys: Vec<Option<&StorageKey>> = start_keys[1..].to_vec();
		end_keys.push(None);

		// use a semaphore to limit max scraping tasks
		let parallel = Arc::new(tokio::sync::Semaphore::new(parallel));
		let builder = Arc::new(self.clone());
		let mut handles = vec![];

		for (start_key, end_key) in start_keys.into_iter().zip(end_keys) {
			let permit = parallel
				.clone()
				.acquire_owned()
				.await
				.expect("semaphore is not closed until the end of loop");

			let builder = builder.clone();
			let prefix = prefix.clone();
			let start_key = start_key.cloned();
			let end_key = end_key.cloned();

			let handle = tokio::spawn(async move {
				let res = builder
					.rpc_get_keys_in_range(&prefix, block, start_key.as_ref(), end_key.as_ref())
					.await;
				drop(permit);
				res
			});

			handles.push(handle);
		}

		parallel.close();

		let keys = futures::future::join_all(handles)
			.await
			.into_iter()
			.filter_map(|res| match res {
				Ok(Ok(keys)) => Some(keys),
				_ => None,
			})
			.flatten()
			.collect::<Vec<StorageKey>>();

		Ok(keys)
	}

	/// Get all keys with `prefix` within the given range at `block`.
	/// Both `start_key` and `end_key` are optional if you want an open-ended range.
	async fn rpc_get_keys_in_range(
		&self,
		prefix: &StorageKey,
		block: B::Hash,
		start_key: Option<&StorageKey>,
		end_key: Option<&StorageKey>,
	) -> Result<Vec<StorageKey>, &'static str> {
		let mut last_key: Option<&StorageKey> = start_key;
		let mut keys: Vec<StorageKey> = vec![];

		loop {
			// This loop can hit the node with very rapid requests, occasionally causing it to
			// error out in CI (https://github.com/paritytech/substrate/issues/14129), so we retry.
			let retry_strategy =
				FixedInterval::new(Self::KEYS_PAGE_RETRY_INTERVAL).take(Self::MAX_RETRIES);
			let get_page_closure =
				|| self.get_keys_single_page(Some(prefix.clone()), last_key.cloned(), block);
			let mut page = Retry::spawn(retry_strategy, get_page_closure).await?;

			// avoid duplicated keys across workloads
			if let (Some(last), Some(end)) = (page.last(), end_key) {
				if last >= end {
					page.retain(|key| key < end);
				}
			}

			let page_len = page.len();
			keys.extend(page);
			last_key = keys.last();

			// scraping out of range or no more matches,
			// we are done either way
			if page_len < Self::DEFAULT_KEY_DOWNLOAD_PAGE as usize {
				log::debug!(target: LOG_TARGET, "last page received: {}", page_len);
				break
			}

			log::debug!(
				target: LOG_TARGET,
				"new total = {}, full page received: {}",
				keys.len(),
				HexDisplay::from(last_key.expect("full page received, cannot be None"))
			);
		}

		Ok(keys)
	}

	/// Fetches storage data from a node using a dynamic batch size.
	///
	/// This function adjusts the batch size on the fly to help prevent overwhelming the node with
	/// large batch requests, and stay within request size limits enforced by the node.
	///
	/// # Arguments
	///
	/// * `client` - An `Arc` wrapped `HttpClient` used for making the requests.
	/// * `payloads` - A vector of tuples containing a JSONRPC method name and `ArrayParams`
	///
	/// # Returns
	///
	/// Returns a `Result` with a vector of `Option<StorageData>`, where each element corresponds to
	/// the storage data for the given method and parameters. The result will be an `Err` with a
	/// `String` error message if the request fails.
	///
	/// # Errors
	///
	/// This function will return an error if:
	/// * The batch request fails and the batch size is less than 2.
	/// * There are invalid batch params.
	/// * There is an error in the batch response.
	///
	/// # Example
	///
	/// ```ignore
	/// use your_crate::{get_storage_data_dynamic_batch_size, HttpClient, ArrayParams};
	/// use std::sync::Arc;
	///
	/// async fn example() {
	///     let client = HttpClient::new();
	///     let payloads = vec![
	///         ("some_method".to_string(), ArrayParams::new(vec![])),
	///         ("another_method".to_string(), ArrayParams::new(vec![])),
	///     ];
	///     let initial_batch_size = 10;
	///
	///     let storage_data = get_storage_data_dynamic_batch_size(client, payloads, batch_size).await;
	///     match storage_data {
	///         Ok(data) => println!("Storage data: {:?}", data),
	///         Err(e) => eprintln!("Error fetching storage data: {}", e),
	///     }
	/// }
	/// ```
	async fn get_storage_data_dynamic_batch_size(
		client: &HttpClient,
		payloads: Vec<(String, ArrayParams)>,
		bar: &ProgressBar,
	) -> Result<Vec<Option<StorageData>>, String> {
		let mut all_data: Vec<Option<StorageData>> = vec![];
		let mut start_index = 0;
		let mut retries = 0usize;
		let mut batch_size = Self::INITIAL_BATCH_SIZE;
		let total_payloads = payloads.len();

		while start_index < total_payloads {
			log::debug!(
				target: LOG_TARGET,
				"Remaining payloads: {} Batch request size: {}",
				total_payloads - start_index,
				batch_size,
			);

			let end_index = usize::min(start_index + batch_size, total_payloads);
			let page = &payloads[start_index..end_index];

			// Build the batch request
			let mut batch = BatchRequestBuilder::new();
			for (method, params) in page.iter() {
				batch
					.insert(method, params.clone())
					.map_err(|_| "Invalid batch method and/or params")?;
			}

			let request_started = Instant::now();
			let batch_response = match client.batch_request::<Option<StorageData>>(batch).await {
				Ok(batch_response) => {
					retries = 0;
					batch_response
				},
				Err(e) => {
					if retries > Self::MAX_RETRIES {
						return Err(e.to_string())
					}

					retries += 1;
					let failure_log = format!(
						"Batch request failed ({}/{} retries). Error: {}",
						retries,
						Self::MAX_RETRIES,
						e
					);
					// after 2 subsequent failures something very wrong is happening. log a warning
					// and reset the batch size down to 1.
					if retries >= 2 {
						log::warn!("{}", failure_log);
						batch_size = 1;
					} else {
						log::debug!("{}", failure_log);
						// Decrease batch size by DECREASE_FACTOR
						batch_size =
							(batch_size as f32 * Self::BATCH_SIZE_DECREASE_FACTOR) as usize;
					}
					continue
				},
			};

			let request_duration = request_started.elapsed();
			batch_size = if request_duration > Self::REQUEST_DURATION_TARGET {
				// Decrease batch size
				max(1, (batch_size as f32 * Self::BATCH_SIZE_DECREASE_FACTOR) as usize)
			} else {
				// Increase batch size, but not more than the remaining total payloads to process
				min(
					total_payloads - start_index,
					max(
						batch_size + 1,
						(batch_size as f32 * Self::BATCH_SIZE_INCREASE_FACTOR) as usize,
					),
				)
			};

			log::debug!(
				target: LOG_TARGET,
				"Request duration: {:?} Target duration: {:?} Last batch size: {} Next batch size: {}",
				request_duration,
				Self::REQUEST_DURATION_TARGET,
				end_index - start_index,
				batch_size
			);

			let batch_response_len = batch_response.len();
			for item in batch_response.into_iter() {
				match item {
					Ok(x) => all_data.push(x),
					Err(e) => return Err(e.message().to_string()),
				}
			}
			bar.inc(batch_response_len as u64);

			// Update the start index for the next iteration
			start_index = end_index;
		}

		Ok(all_data)
	}

	/// Synonym of `getPairs` that uses paged queries to first get the keys, and then
	/// map them to values one by one.
	///
	/// This can work with public nodes. But, expect it to be darn slow.
	pub(crate) async fn rpc_get_pairs(
		&self,
		prefix: StorageKey,
		at: B::Hash,
		pending_ext: &mut TestExternalities<HashingFor<B>>,
	) -> Result<Vec<KeyValue>, &'static str> {
		let start = Instant::now();
		let mut sp = Spinner::with_timer(Spinners::Dots, "Scraping keys...".into());
		// TODO We could start downloading when having collected the first batch of keys
		// https://github.com/paritytech/polkadot-sdk/issues/2494
		let keys = self
			.rpc_get_keys_parallel(&prefix, at, Self::PARALLEL_REQUESTS)
			.await?
			.into_iter()
			.collect::<Vec<_>>();
		sp.stop_with_message(format!(
			"โœ… Found {} keys ({:.2}s)",
			keys.len(),
			start.elapsed().as_secs_f32()
		));
		if keys.is_empty() {
			return Ok(Default::default())
		}

		let client = self.as_online().rpc_client();
		let payloads = keys
			.iter()
			.map(|key| ("state_getStorage".to_string(), rpc_params!(key, at)))
			.collect::<Vec<_>>();

		let bar = ProgressBar::new(payloads.len() as u64);
		bar.enable_steady_tick(Duration::from_secs(1));
		bar.set_message("Downloading key values".to_string());
		bar.set_style(
			ProgressStyle::with_template(
				"[{elapsed_precise}] {msg} {per_sec} [{wide_bar}] {pos}/{len} ({eta})",
			)
			.unwrap()
			.progress_chars("=>-"),
		);
		let payloads_chunked = payloads.chunks((payloads.len() / Self::PARALLEL_REQUESTS).max(1));
		let requests = payloads_chunked.map(|payload_chunk| {
			Self::get_storage_data_dynamic_batch_size(client, payload_chunk.to_vec(), &bar)
		});
		// Execute the requests and move the Result outside.
		let storage_data_result: Result<Vec<_>, _> =
			futures::future::join_all(requests).await.into_iter().collect();
		// Handle the Result.
		let storage_data = match storage_data_result {
			Ok(storage_data) => storage_data.into_iter().flatten().collect::<Vec<_>>(),
			Err(e) => {
				log::error!(target: LOG_TARGET, "Error while getting storage data: {}", e);
				return Err("Error while getting storage data")
			},
		};
		bar.finish_with_message("โœ… Downloaded key values");
		println!();

		// Check if we got responses for all submitted requests.
		assert_eq!(keys.len(), storage_data.len());

		let key_values = keys
			.iter()
			.zip(storage_data)
			.map(|(key, maybe_value)| match maybe_value {
				Some(data) => (key.clone(), data),
				None => {
					log::warn!(target: LOG_TARGET, "key {:?} had none corresponding value.", &key);
					let data = StorageData(vec![]);
					(key.clone(), data)
				},
			})
			.collect::<Vec<_>>();

		let mut sp = Spinner::with_timer(Spinners::Dots, "Inserting keys into DB...".into());
		let start = Instant::now();
		pending_ext.batch_insert(key_values.clone().into_iter().filter_map(|(k, v)| {
			// Don't insert the child keys here, they need to be inserted separately with all their
			// data in the load_child_remote function.
			match is_default_child_storage_key(&k.0) {
				true => None,
				false => Some((k.0, v.0)),
			}
		}));
		sp.stop_with_message(format!(
			"โœ… Inserted keys into DB ({:.2}s)",
			start.elapsed().as_secs_f32()
		));
		Ok(key_values)
	}

	/// Get the values corresponding to `child_keys` at the given `prefixed_top_key`.
	pub(crate) async fn rpc_child_get_storage_paged(
		client: &HttpClient,
		prefixed_top_key: &StorageKey,
		child_keys: Vec<StorageKey>,
		at: B::Hash,
	) -> Result<Vec<KeyValue>, &'static str> {
		let child_keys_len = child_keys.len();

		let payloads = child_keys
			.iter()
			.map(|key| {
				(
					"childstate_getStorage".to_string(),
					rpc_params![
						PrefixedStorageKey::new(prefixed_top_key.as_ref().to_vec()),
						key,
						at
					],
				)
			})
			.collect::<Vec<_>>();

		let bar = ProgressBar::new(payloads.len() as u64);
		let storage_data =
			match Self::get_storage_data_dynamic_batch_size(client, payloads, &bar).await {
				Ok(storage_data) => storage_data,
				Err(e) => {
					log::error!(target: LOG_TARGET, "batch processing failed: {:?}", e);
					return Err("batch processing failed")
				},
			};

		assert_eq!(child_keys_len, storage_data.len());

		Ok(child_keys
			.iter()
			.zip(storage_data)
			.map(|(key, maybe_value)| match maybe_value {
				Some(v) => (key.clone(), v),
				None => {
					log::warn!(target: LOG_TARGET, "key {:?} had no corresponding value.", &key);
					(key.clone(), StorageData(vec![]))
				},
			})
			.collect::<Vec<_>>())
	}

	pub(crate) async fn rpc_child_get_keys(
		client: &HttpClient,
		prefixed_top_key: &StorageKey,
		child_prefix: StorageKey,
		at: B::Hash,
	) -> Result<Vec<StorageKey>, &'static str> {
		let retry_strategy =
			FixedInterval::new(Self::KEYS_PAGE_RETRY_INTERVAL).take(Self::MAX_RETRIES);
		let mut all_child_keys = Vec::new();
		let mut start_key = None;

		loop {
			let get_child_keys_closure = || {
				let top_key = PrefixedStorageKey::new(prefixed_top_key.0.clone());
				substrate_rpc_client::ChildStateApi::storage_keys_paged(
					client,
					top_key,
					Some(child_prefix.clone()),
					Self::DEFAULT_KEY_DOWNLOAD_PAGE,
					start_key.clone(),
					Some(at),
				)
			};

			let child_keys = Retry::spawn(retry_strategy.clone(), get_child_keys_closure)
				.await
				.map_err(|e| {
					error!(target: LOG_TARGET, "Error = {:?}", e);
					"rpc child_get_keys failed."
				})?;

			let keys_count = child_keys.len();
			if keys_count == 0 {
				break;
			}

			start_key = child_keys.last().cloned();
			all_child_keys.extend(child_keys);

			if keys_count < Self::DEFAULT_KEY_DOWNLOAD_PAGE as usize {
				break;
			}
		}

		debug!(
			target: LOG_TARGET,
			"[thread = {:?}] scraped {} child-keys of the child-bearing top key: {}",
			std::thread::current().id(),
			all_child_keys.len(),
			HexDisplay::from(prefixed_top_key)
		);

		Ok(all_child_keys)
	}
}

impl<B: BlockT> Builder<B>
where
	B::Hash: DeserializeOwned,
	B::Header: DeserializeOwned,
{
	/// Load all of the child keys from the remote config, given the already scraped list of top key
	/// pairs.
	///
	/// `top_kv` need not be only child-bearing top keys. It should be all of the top keys that are
	/// included thus far.
	///
	/// This function concurrently populates `pending_ext`. the return value is only for writing to
	/// cache, we can also optimize further.
	async fn load_child_remote(
		&self,
		top_kv: &[KeyValue],
		pending_ext: &mut TestExternalities<HashingFor<B>>,
	) -> Result<ChildKeyValues, &'static str> {
		let child_roots = top_kv
			.iter()
			.filter(|(k, _)| is_default_child_storage_key(k.as_ref()))
			.map(|(k, _)| k.clone())
			.collect::<Vec<_>>();

		if child_roots.is_empty() {
			info!(target: LOG_TARGET, "๐Ÿ‘ฉโ€๐Ÿ‘ฆ no child roots found to scrape",);
			return Ok(Default::default())
		}

		info!(
			target: LOG_TARGET,
			"๐Ÿ‘ฉโ€๐Ÿ‘ฆ scraping child-tree data from {} top keys",
			child_roots.len(),
		);

		let at = self.as_online().at_expected();

		let client = self.as_online().rpc_client();
		let mut child_kv = vec![];
		for prefixed_top_key in child_roots {
			let child_keys =
				Self::rpc_child_get_keys(client, &prefixed_top_key, StorageKey(vec![]), at).await?;

			let child_kv_inner =
				Self::rpc_child_get_storage_paged(client, &prefixed_top_key, child_keys, at)
					.await?;

			let prefixed_top_key = PrefixedStorageKey::new(prefixed_top_key.clone().0);
			let un_prefixed = match ChildType::from_prefixed_key(&prefixed_top_key) {
				Some((ChildType::ParentKeyId, storage_key)) => storage_key,
				None => {
					log::error!(target: LOG_TARGET, "invalid key: {:?}", prefixed_top_key);
					return Err("Invalid child key")
				},
			};

			let info = ChildInfo::new_default(un_prefixed);
			let key_values =
				child_kv_inner.iter().cloned().map(|(k, v)| (k.0, v.0)).collect::<Vec<_>>();
			child_kv.push((info.clone(), child_kv_inner));
			for (k, v) in key_values {
				pending_ext.insert_child(info.clone(), k, v);
			}
		}

		Ok(child_kv)
	}

	/// Build `Self` from a network node denoted by `uri`.
	///
	/// This function concurrently populates `pending_ext`. the return value is only for writing to
	/// cache, we can also optimize further.
	async fn load_top_remote(
		&self,
		pending_ext: &mut TestExternalities<HashingFor<B>>,
	) -> Result<TopKeyValues, &'static str> {
		let config = self.as_online();
		let at = self
			.as_online()
			.at
			.expect("online config must be initialized by this point; qed.");
		log::info!(target: LOG_TARGET, "scraping key-pairs from remote at block height {:?}", at);

		let mut keys_and_values = Vec::new();
		for prefix in &config.hashed_prefixes {
			let now = std::time::Instant::now();
			let additional_key_values =
				self.rpc_get_pairs(StorageKey(prefix.to_vec()), at, pending_ext).await?;
			let elapsed = now.elapsed();
			log::info!(
				target: LOG_TARGET,
				"adding data for hashed prefix: {:?}, took {:.2}s",
				HexDisplay::from(prefix),
				elapsed.as_secs_f32()
			);
			keys_and_values.extend(additional_key_values);
		}

		for key in &config.hashed_keys {
			let key = StorageKey(key.to_vec());
			log::info!(
				target: LOG_TARGET,
				"adding data for hashed key: {:?}",
				HexDisplay::from(&key)
			);
			match self.rpc_get_storage(key.clone(), Some(at)).await? {
				Some(value) => {
					pending_ext.insert(key.clone().0, value.clone().0);
					keys_and_values.push((key, value));
				},
				None => {
					log::warn!(
						target: LOG_TARGET,
						"no data found for hashed key: {:?}",
						HexDisplay::from(&key)
					);
				},
			}
		}

		Ok(keys_and_values)
	}

	/// The entry point of execution, if `mode` is online.
	///
	/// initializes the remote client in `transport`, and sets the `at` field, if not specified.
	async fn init_remote_client(&mut self) -> Result<(), &'static str> {
		// First, initialize the http client.
		self.as_online_mut().transport.init().await?;

		// Then, if `at` is not set, set it.
		if self.as_online().at.is_none() {
			let at = self.rpc_get_head().await?;
			log::info!(
				target: LOG_TARGET,
				"since no at is provided, setting it to latest finalized head, {:?}",
				at
			);
			self.as_online_mut().at = Some(at);
		}

		// Then, a few transformation that we want to perform in the online config:
		let online_config = self.as_online_mut();
		online_config.pallets.iter().for_each(|p| {
			online_config
				.hashed_prefixes
				.push(sp_crypto_hashing::twox_128(p.as_bytes()).to_vec())
		});

		if online_config.child_trie {
			online_config.hashed_prefixes.push(DEFAULT_CHILD_STORAGE_KEY_PREFIX.to_vec());
		}

		// Finally, if by now, we have put any limitations on prefixes that we are interested in, we
		// download everything.
		if online_config
			.hashed_prefixes
			.iter()
			.filter(|p| *p != DEFAULT_CHILD_STORAGE_KEY_PREFIX)
			.count() == 0
		{
			log::info!(
				target: LOG_TARGET,
				"since no prefix is filtered, the data for all pallets will be downloaded"
			);
			online_config.hashed_prefixes.push(vec![]);
		}

		Ok(())
	}

	async fn load_header(&self) -> Result<B::Header, &'static str> {
		let retry_strategy =
			FixedInterval::new(Self::KEYS_PAGE_RETRY_INTERVAL).take(Self::MAX_RETRIES);
		let get_header_closure = || {
			ChainApi::<(), _, B::Header, ()>::header(
				self.as_online().rpc_client(),
				Some(self.as_online().at_expected()),
			)
		};
		Retry::spawn(retry_strategy, get_header_closure)
			.await
			.map_err(|_| "Failed to fetch header for block from network")?
			.ok_or("Network returned None block header")
	}

	/// Load the data from a remote server. The main code path is calling into `load_top_remote` and
	/// `load_child_remote`.
	///
	/// Must be called after `init_remote_client`.
	async fn load_remote_and_maybe_save(
		&mut self,
	) -> Result<TestExternalities<HashingFor<B>>, &'static str> {
		let state_version =
			StateApi::<B::Hash>::runtime_version(self.as_online().rpc_client(), None)
				.await
				.map_err(|e| {
					error!(target: LOG_TARGET, "Error = {:?}", e);
					"rpc runtime_version failed."
				})
				.map(|v| v.state_version())?;
		let mut pending_ext = TestExternalities::new_with_code_and_state(
			Default::default(),
			Default::default(),
			self.overwrite_state_version.unwrap_or(state_version),
		);

		// Load data from the remote into `pending_ext`.
		let top_kv = self.load_top_remote(&mut pending_ext).await?;
		self.load_child_remote(&top_kv, &mut pending_ext).await?;

		// If we need to save a snapshot, save the raw storage and root hash to the snapshot.
		if let Some(path) = self.as_online().state_snapshot.clone().map(|c| c.path) {
			let (raw_storage, storage_root) = pending_ext.into_raw_snapshot();
			let snapshot = Snapshot::<B>::new(
				state_version,
				raw_storage.clone(),
				storage_root,
				self.load_header().await?,
			);
			let encoded = snapshot.encode();
			log::info!(
				target: LOG_TARGET,
				"writing snapshot of {} bytes to {:?}",
				encoded.len(),
				path
			);
			std::fs::write(path, encoded).map_err(|_| "fs::write failed")?;

			// pending_ext was consumed when creating the snapshot, need to reinitailize it
			return Ok(TestExternalities::from_raw_snapshot(
				raw_storage,
				storage_root,
				self.overwrite_state_version.unwrap_or(state_version),
			))
		}

		Ok(pending_ext)
	}

	async fn do_load_remote(&mut self) -> Result<RemoteExternalities<B>, &'static str> {
		self.init_remote_client().await?;
		let inner_ext = self.load_remote_and_maybe_save().await?;
		Ok(RemoteExternalities { header: self.load_header().await?, inner_ext })
	}

	fn do_load_offline(
		&mut self,
		config: OfflineConfig,
	) -> Result<RemoteExternalities<B>, &'static str> {
		let mut sp = Spinner::with_timer(Spinners::Dots, "Loading snapshot...".into());
		let start = Instant::now();
		info!(target: LOG_TARGET, "Loading snapshot from {:?}", &config.state_snapshot.path);
		let Snapshot { snapshot_version: _, header, state_version, raw_storage, storage_root } =
			Snapshot::<B>::load(&config.state_snapshot.path)?;

		let inner_ext = TestExternalities::from_raw_snapshot(
			raw_storage,
			storage_root,
			self.overwrite_state_version.unwrap_or(state_version),
		);
		sp.stop_with_message(format!("โœ… Loaded snapshot ({:.2}s)", start.elapsed().as_secs_f32()));

		Ok(RemoteExternalities { inner_ext, header })
	}

	pub(crate) async fn pre_build(mut self) -> Result<RemoteExternalities<B>, &'static str> {
		let mut ext = match self.mode.clone() {
			Mode::Offline(config) => self.do_load_offline(config)?,
			Mode::Online(_) => self.do_load_remote().await?,
			Mode::OfflineOrElseOnline(offline_config, _) => {
				match self.do_load_offline(offline_config) {
					Ok(x) => x,
					Err(_) => self.do_load_remote().await?,
				}
			},
		};

		// inject manual key values.
		if !self.hashed_key_values.is_empty() {
			log::info!(
				target: LOG_TARGET,
				"extending externalities with {} manually injected key-values",
				self.hashed_key_values.len()
			);
			ext.batch_insert(self.hashed_key_values.into_iter().map(|(k, v)| (k.0, v.0)));
		}

		// exclude manual key values.
		if !self.hashed_blacklist.is_empty() {
			log::info!(
				target: LOG_TARGET,
				"excluding externalities from {} keys",
				self.hashed_blacklist.len()
			);
			for k in self.hashed_blacklist {
				ext.execute_with(|| sp_io::storage::clear(&k));
			}
		}

		Ok(ext)
	}
}

// Public methods
impl<B: BlockT> Builder<B>
where
	B::Hash: DeserializeOwned,
	B::Header: DeserializeOwned,
{
	/// Create a new builder.
	pub fn new() -> Self {
		Default::default()
	}

	/// Inject a manual list of key and values to the storage.
	pub fn inject_hashed_key_value(mut self, injections: Vec<KeyValue>) -> Self {
		for i in injections {
			self.hashed_key_values.push(i.clone());
		}
		self
	}

	/// Blacklist this hashed key from the final externalities. This is treated as-is, and should be
	/// pre-hashed.
	pub fn blacklist_hashed_key(mut self, hashed: &[u8]) -> Self {
		self.hashed_blacklist.push(hashed.to_vec());
		self
	}

	/// Configure a state snapshot to be used.
	pub fn mode(mut self, mode: Mode<B::Hash>) -> Self {
		self.mode = mode;
		self
	}

	/// The state version to use.
	pub fn overwrite_state_version(mut self, version: StateVersion) -> Self {
		self.overwrite_state_version = Some(version);
		self
	}

	pub async fn build(self) -> Result<RemoteExternalities<B>, &'static str> {
		let mut ext = self.pre_build().await?;
		ext.commit_all().unwrap();

		info!(
			target: LOG_TARGET,
			"initialized state externalities with storage root {:?} and state_version {:?}",
			ext.as_backend().root(),
			ext.state_version
		);

		Ok(ext)
	}
}

#[cfg(test)]
mod test_prelude {
	pub(crate) use super::*;
	pub(crate) use sp_runtime::testing::{Block as RawBlock, ExtrinsicWrapper, H256 as Hash};
	pub(crate) type Block = RawBlock<ExtrinsicWrapper<Hash>>;

	pub(crate) fn init_logger() {
		sp_tracing::try_init_simple();
	}
}

#[cfg(test)]
mod tests {
	use super::test_prelude::*;

	#[tokio::test]
	async fn can_load_state_snapshot() {
		init_logger();
		Builder::<Block>::new()
			.mode(Mode::Offline(OfflineConfig {
				state_snapshot: SnapshotConfig::new("test_data/test.snap"),
			}))
			.build()
			.await
			.unwrap()
			.execute_with(|| {});
	}

	#[tokio::test]
	async fn can_exclude_from_snapshot() {
		init_logger();

		// get the first key from the snapshot file.
		let some_key = Builder::<Block>::new()
			.mode(Mode::Offline(OfflineConfig {
				state_snapshot: SnapshotConfig::new("test_data/test.snap"),
			}))
			.build()
			.await
			.expect("Can't read state snapshot file")
			.execute_with(|| {
				let key =
					sp_io::storage::next_key(&[]).expect("some key must exist in the snapshot");
				assert!(sp_io::storage::get(&key).is_some());
				key
			});

		Builder::<Block>::new()
			.mode(Mode::Offline(OfflineConfig {
				state_snapshot: SnapshotConfig::new("test_data/test.snap"),
			}))
			.blacklist_hashed_key(&some_key)
			.build()
			.await
			.expect("Can't read state snapshot file")
			.execute_with(|| assert!(sp_io::storage::get(&some_key).is_none()));
	}
}

#[cfg(all(test, feature = "remote-test"))]
mod remote_tests {
	use super::test_prelude::*;
	use std::{env, os::unix::fs::MetadataExt};

	fn endpoint() -> String {
		env::var("TEST_WS").unwrap_or_else(|_| DEFAULT_HTTP_ENDPOINT.to_string())
	}

	#[tokio::test]
	async fn state_version_is_kept_and_can_be_altered() {
		const CACHE: &'static str = "state_version_is_kept_and_can_be_altered";
		init_logger();

		// first, build a snapshot.
		let ext = Builder::<Block>::new()
			.mode(Mode::Online(OnlineConfig {
				transport: endpoint().clone().into(),
				pallets: vec!["Proxy".to_owned()],
				child_trie: false,
				state_snapshot: Some(SnapshotConfig::new(CACHE)),
				..Default::default()
			}))
			.build()
			.await
			.unwrap();

		// now re-create the same snapshot.
		let cached_ext = Builder::<Block>::new()
			.mode(Mode::Offline(OfflineConfig { state_snapshot: SnapshotConfig::new(CACHE) }))
			.build()
			.await
			.unwrap();

		assert_eq!(ext.state_version, cached_ext.state_version);

		// now overwrite it
		let other = match ext.state_version {
			StateVersion::V0 => StateVersion::V1,
			StateVersion::V1 => StateVersion::V0,
		};
		let cached_ext = Builder::<Block>::new()
			.mode(Mode::Offline(OfflineConfig { state_snapshot: SnapshotConfig::new(CACHE) }))
			.overwrite_state_version(other)
			.build()
			.await
			.unwrap();

		assert_eq!(cached_ext.state_version, other);
	}

	#[tokio::test]
	async fn snapshot_block_hash_works() {
		const CACHE: &'static str = "snapshot_block_hash_works";
		init_logger();

		// first, build a snapshot.
		let ext = Builder::<Block>::new()
			.mode(Mode::Online(OnlineConfig {
				transport: endpoint().clone().into(),
				pallets: vec!["Proxy".to_owned()],
				child_trie: false,
				state_snapshot: Some(SnapshotConfig::new(CACHE)),
				..Default::default()
			}))
			.build()
			.await
			.unwrap();

		// now re-create the same snapshot.
		let cached_ext = Builder::<Block>::new()
			.mode(Mode::Offline(OfflineConfig { state_snapshot: SnapshotConfig::new(CACHE) }))
			.build()
			.await
			.unwrap();

		assert_eq!(ext.header.hash(), cached_ext.header.hash());
	}

	#[tokio::test]
	async fn child_keys_are_loaded() {
		const CACHE: &'static str = "snapshot_retains_storage";
		init_logger();

		// create an ext with children keys
		let mut child_ext = Builder::<Block>::new()
			.mode(Mode::Online(OnlineConfig {
				transport: endpoint().clone().into(),
				pallets: vec!["Proxy".to_owned()],
				child_trie: true,
				state_snapshot: Some(SnapshotConfig::new(CACHE)),
				..Default::default()
			}))
			.build()
			.await
			.unwrap();

		// create an ext without children keys
		let mut ext = Builder::<Block>::new()
			.mode(Mode::Online(OnlineConfig {
				transport: endpoint().clone().into(),
				pallets: vec!["Proxy".to_owned()],
				child_trie: false,
				state_snapshot: Some(SnapshotConfig::new(CACHE)),
				..Default::default()
			}))
			.build()
			.await
			.unwrap();

		// there should be more keys in the child ext.
		assert!(
			child_ext.as_backend().backend_storage().keys().len() >
				ext.as_backend().backend_storage().keys().len()
		);
	}

	#[tokio::test]
	async fn offline_else_online_works() {
		const CACHE: &'static str = "offline_else_online_works_data";
		init_logger();
		// this shows that in the second run, we use the remote and create a snapshot.
		Builder::<Block>::new()
			.mode(Mode::OfflineOrElseOnline(
				OfflineConfig { state_snapshot: SnapshotConfig::new(CACHE) },
				OnlineConfig {
					transport: endpoint().clone().into(),
					pallets: vec!["Proxy".to_owned()],
					child_trie: false,
					state_snapshot: Some(SnapshotConfig::new(CACHE)),
					..Default::default()
				},
			))
			.build()
			.await
			.unwrap()
			.execute_with(|| {});

		// this shows that in the second run, we are not using the remote
		Builder::<Block>::new()
			.mode(Mode::OfflineOrElseOnline(
				OfflineConfig { state_snapshot: SnapshotConfig::new(CACHE) },
				OnlineConfig {
					transport: "ws://non-existent:666".to_owned().into(),
					..Default::default()
				},
			))
			.build()
			.await
			.unwrap()
			.execute_with(|| {});

		let to_delete = std::fs::read_dir(Path::new("."))
			.unwrap()
			.into_iter()
			.map(|d| d.unwrap())
			.filter(|p| p.path().file_name().unwrap_or_default() == CACHE)
			.collect::<Vec<_>>();

		assert!(to_delete.len() == 1);
		std::fs::remove_file(to_delete[0].path()).unwrap();
	}

	#[tokio::test]
	async fn can_build_one_small_pallet() {
		init_logger();
		Builder::<Block>::new()
			.mode(Mode::Online(OnlineConfig {
				transport: endpoint().clone().into(),
				pallets: vec!["Proxy".to_owned()],
				child_trie: false,
				..Default::default()
			}))
			.build()
			.await
			.unwrap()
			.execute_with(|| {});
	}

	#[tokio::test]
	async fn can_build_few_pallet() {
		init_logger();
		Builder::<Block>::new()
			.mode(Mode::Online(OnlineConfig {
				transport: endpoint().clone().into(),
				pallets: vec!["Proxy".to_owned(), "Multisig".to_owned()],
				child_trie: false,
				..Default::default()
			}))
			.build()
			.await
			.unwrap()
			.execute_with(|| {});
	}

	#[tokio::test(flavor = "multi_thread")]
	async fn can_create_snapshot() {
		const CACHE: &'static str = "can_create_snapshot";
		init_logger();

		Builder::<Block>::new()
			.mode(Mode::Online(OnlineConfig {
				transport: endpoint().clone().into(),
				state_snapshot: Some(SnapshotConfig::new(CACHE)),
				pallets: vec!["Proxy".to_owned()],
				child_trie: false,
				..Default::default()
			}))
			.build()
			.await
			.unwrap()
			.execute_with(|| {});

		let to_delete = std::fs::read_dir(Path::new("."))
			.unwrap()
			.into_iter()
			.map(|d| d.unwrap())
			.filter(|p| p.path().file_name().unwrap_or_default() == CACHE)
			.collect::<Vec<_>>();

		assert!(to_delete.len() == 1);
		let to_delete = to_delete.first().unwrap();
		assert!(std::fs::metadata(to_delete.path()).unwrap().size() > 1);
		std::fs::remove_file(to_delete.path()).unwrap();
	}

	#[tokio::test]
	async fn can_create_child_snapshot() {
		const CACHE: &'static str = "can_create_child_snapshot";
		init_logger();
		Builder::<Block>::new()
			.mode(Mode::Online(OnlineConfig {
				transport: endpoint().clone().into(),
				state_snapshot: Some(SnapshotConfig::new(CACHE)),
				pallets: vec!["Crowdloan".to_owned()],
				child_trie: true,
				..Default::default()
			}))
			.build()
			.await
			.unwrap()
			.execute_with(|| {});

		let to_delete = std::fs::read_dir(Path::new("."))
			.unwrap()
			.into_iter()
			.map(|d| d.unwrap())
			.filter(|p| p.path().file_name().unwrap_or_default() == CACHE)
			.collect::<Vec<_>>();

		assert!(to_delete.len() == 1);
		let to_delete = to_delete.first().unwrap();
		assert!(std::fs::metadata(to_delete.path()).unwrap().size() > 1);
		std::fs::remove_file(to_delete.path()).unwrap();
	}

	#[tokio::test]
	async fn can_build_big_pallet() {
		if std::option_env!("TEST_WS").is_none() {
			return
		}
		init_logger();
		Builder::<Block>::new()
			.mode(Mode::Online(OnlineConfig {
				transport: endpoint().clone().into(),
				pallets: vec!["Staking".to_owned()],
				child_trie: false,
				..Default::default()
			}))
			.build()
			.await
			.unwrap()
			.execute_with(|| {});
	}

	#[tokio::test]
	async fn can_fetch_all() {
		if std::option_env!("TEST_WS").is_none() {
			return
		}
		init_logger();
		Builder::<Block>::new()
			.mode(Mode::Online(OnlineConfig {
				transport: endpoint().clone().into(),
				..Default::default()
			}))
			.build()
			.await
			.unwrap()
			.execute_with(|| {});
	}

	#[tokio::test]
	async fn can_fetch_in_parallel() {
		init_logger();

		let mut builder = Builder::<Block>::new().mode(Mode::Online(OnlineConfig {
			transport: endpoint().clone().into(),
			..Default::default()
		}));
		builder.init_remote_client().await.unwrap();

		let at = builder.as_online().at.unwrap();

		let prefix = StorageKey(vec![13]);
		let paged = builder.rpc_get_keys_in_range(&prefix, at, None, None).await.unwrap();
		let para = builder.rpc_get_keys_parallel(&prefix, at, 4).await.unwrap();
		assert_eq!(paged, para);

		let prefix = StorageKey(vec![]);
		let paged = builder.rpc_get_keys_in_range(&prefix, at, None, None).await.unwrap();
		let para = builder.rpc_get_keys_parallel(&prefix, at, 8).await.unwrap();
		assert_eq!(paged, para);
	}
}