crawlrs 0.2.0

crawlrs - Enterprise-grade web scraping and crawling platform built with Rust
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
// Copyright (c) 2025-2026 Kirky.X🌠
// SPDX-License-Identifier: Apache-2.0

//! SearXNG 实例生命周期编排
//!
//! [`SearxngInstanceServiceImpl`] 负责:采集轮次、探活轮次、熔断记账、逐出与晋升、
//! 降级口径。所有外部依赖(登记簿仓储、HTTP 传输)都是注入的端口,测试用真实
//! trait impl 替身驱动(见 `test_support.rs`)。
//!
//! 并发模型:内部状态全部放在 `Arc` 内,`report(Failure)` 触发的晋升通过
//! `tokio::spawn` 异步执行 —— 失败请求路径不被探活阻塞;晋升本身幂等(池按 url 去重)。

use std::collections::HashSet;
use std::sync::atomic::{AtomicBool, Ordering};
use std::sync::Arc;
use std::time::Duration;

use log::{debug, info, warn};
use tokio::sync::Semaphore;

use crate::config::SearxngSearchSettings;
use crate::domain::repositories::searxng_instance_repository::{
    normalize_instance_url, SearxngInstance, SearxngInstanceRepository, SearxngInstanceState,
};
use crate::engines::circuit_breaker::{CircuitBreaker, CircuitConfig};
use crate::search::types::EngineHealth;

use super::collector;
use super::probe::{probe_instance, ProbeOutcome};
use super::{
    BreakerMirror, CollectorReport, HealthReport, InstanceOutcome, InstancePool, PoolMember,
    PoolMemberState, PoolSnapshot, SearxngError, SearxngInstanceService, SearxngTransport,
};

/// 采集轮跨进程互斥锁的固定常量键("searxng\0" 的 big-endian i64)
pub(crate) const COLLECT_ADVISORY_LOCK_KEY: i64 = 0x7365_6172_786E_6700;

/// 熔断键前缀(实例专属,不与 engines 统计口径混用)
const BREAKER_KEY_PREFIX: &str = "searxng:";

/// 晋升并发上限:失败风暴下每条 URL 一个 spawn 会无界增长,
/// 用信号量把同时进行的晋升探测限制在固定数量;拿不到 permit 的晋升直接丢弃
/// (下一轮探活会重新触发,排队只会把风暴转化为任务积压)。
const PROMOTION_MAX_CONCURRENCY: usize = 8;

/// 服务内部共享状态
struct SearxngInstanceInner {
    /// 登记簿仓储(DB 为唯一共享真相)
    repo: Arc<dyn SearxngInstanceRepository>,
    /// HTTP 传输端口(采集 / 探活 / 搜索共用)
    transport: Arc<dyn SearxngTransport>,
    /// `[search.searxng]` 配置快照
    settings: SearxngSearchSettings,
    /// 专属熔断器实例(不与 engines 共享,避免指标标签污染与阈值互相干扰)
    breaker: CircuitBreaker,
    /// 进程内热池
    pool: InstancePool,
    /// 晋升并发护栏(见 [`PROMOTION_MAX_CONCURRENCY`])
    promotion_sem: Arc<Semaphore>,
    /// 启动期热池重建是否已完成(完成前 `health()` 报 Degraded)
    bootstrap_done: AtomicBool,
}

/// SearXNG 实例生命周期服务实现
#[derive(Clone)]
pub struct SearxngInstanceServiceImpl {
    inner: Arc<SearxngInstanceInner>,
}

impl SearxngInstanceServiceImpl {
    /// 构造服务(专属熔断器 + 目标为 `pool_target` 的空热池)
    pub fn new(
        repo: Arc<dyn SearxngInstanceRepository>,
        transport: Arc<dyn SearxngTransport>,
        settings: SearxngSearchSettings,
    ) -> Self {
        let breaker = CircuitBreaker::with_default_config(CircuitConfig {
            failure_threshold: settings.breaker.failure_threshold,
            recovery_timeout: Duration::from_secs(settings.breaker.recovery_timeout_seconds),
            ..CircuitConfig::default()
        });
        let pool = InstancePool::new(settings.pool_target);
        Self {
            inner: Arc::new(SearxngInstanceInner {
                repo,
                transport,
                settings,
                breaker,
                pool,
                promotion_sem: Arc::new(Semaphore::new(PROMOTION_MAX_CONCURRENCY)),
                bootstrap_done: AtomicBool::new(false),
            }),
        }
    }

    /// 配置快照(引擎侧读取超时 / UA 等)
    pub fn settings(&self) -> &SearxngSearchSettings {
        &self.inner.settings
    }

    /// 熔断键:`searxng:<url>`
    fn breaker_key(url: &str) -> String {
        format!("{BREAKER_KEY_PREFIX}{url}")
    }

    /// 从热池轮转取一个可用实例(跳过 `exclude` 与熔断开成员)
    pub fn acquire(&self, exclude: &[String]) -> Option<Arc<PoolMember>> {
        self.inner.pool.acquire(exclude, &|url| {
            !self.inner.breaker.is_open(&Self::breaker_key(url))
        })
    }

    /// 当前池成员数(测试与管理视图共用)
    pub fn pool_len(&self) -> usize {
        self.inner.pool.len()
    }

    /// 标记启动期重建完成(重建任务结束时调用)
    pub fn mark_bootstrap_done(&self) {
        self.inner.bootstrap_done.store(true, Ordering::Release);
    }

    /// 热池当前是否已满(达到 `pool_target`)
    pub fn pool_is_full(&self) -> bool {
        self.inner.pool.is_full()
    }

    /// 追加一个成员进池(启动期重建 / 晋升路径共用)
    fn add_member(&self, instance: SearxngInstance) -> bool {
        self.inner.pool.insert(Arc::new(PoolMember { instance }))
    }

    /// 引擎健康口径:池满 → Healthy;非空但不足 → Degraded(+warn);空 → Unhealthy。
    /// 启动期重建完成前一律 Degraded(避免伪报可用)。
    pub fn health(&self) -> EngineHealth {
        if !self.inner.bootstrap_done.load(Ordering::Acquire) {
            return EngineHealth::Degraded;
        }
        let len = self.inner.pool.len();
        let target = self.inner.pool.target();
        if len >= target {
            EngineHealth::Healthy
        } else if len > 0 {
            warn!(
                "searxng pool degraded: {len}/{target} instances available \
                 (target is a control-loop goal, not a hard guarantee on third-party instances)"
            );
            EngineHealth::Degraded
        } else {
            EngineHealth::Unhealthy
        }
    }

    /// 管理视图:热池成员 + 熔断镜像态 + 健康口径
    pub async fn pool_snapshot(&self) -> PoolSnapshot {
        let members = self
            .inner
            .pool
            .snapshot()
            .iter()
            .map(|member| {
                let key = Self::breaker_key(member.url());
                let breaker = if self.inner.breaker.is_open(&key) {
                    BreakerMirror::Open
                } else if member.instance.state == SearxngInstanceState::Cooldown {
                    BreakerMirror::HalfOpen
                } else {
                    BreakerMirror::Closed
                };
                PoolMemberState {
                    url: member.instance.url.clone(),
                    state: member.instance.state,
                    latency_ms: member.instance.latency_ms,
                    breaker,
                }
            })
            .collect();
        PoolSnapshot {
            target: self.inner.pool.target(),
            members,
            health: self.health(),
        }
    }

    /// 管理视图:登记簿分页列表(active 优先 + uptime 降序)
    pub async fn list_instances(
        &self,
        limit: u64,
        offset: u64,
    ) -> Result<Vec<SearxngInstance>, SearxngError> {
        self.inner
            .repo
            .list_paginated(limit, offset)
            .await
            .map_err(|e| SearxngError::Repository(e.to_string()))
    }

    /// 请求结果回填:熔断记账 + 时延/失败计数;失败达阈值时逐出并异步晋升替补。
    pub async fn report(&self, url: &str, outcome: InstanceOutcome) {
        let key = Self::breaker_key(url);
        match outcome {
            InstanceOutcome::Success { latency_ms } => {
                self.inner.breaker.record_success(&key).await;
                if let Err(e) = self
                    .inner
                    .repo
                    .record_probe_result(url, true, Some(latency_ms), None)
                    .await
                {
                    warn!("searxng success record failed url={url} error={e}");
                }
                // 刷新池内的时延快照(管理视图展示):仅当成员仍在池中才更新。
                // 若并发 Failure 已把它逐出,回填放弃(逐出决定优先,不回插)。
                self.inner.pool.update_if_present(url, |instance| {
                    instance.latency_ms = Some(latency_ms);
                    instance.consecutive_failures = 0;
                    instance.last_success_at = Some(chrono::Utc::now());
                });
            }
            InstanceOutcome::Failure { error } => {
                self.inner.breaker.record_failure(&key).await;
                let message = error.to_string();
                if let Err(e) = self
                    .inner
                    .repo
                    .record_probe_result(url, false, None, Some(&message))
                    .await
                {
                    warn!("searxng failure record failed url={url} error={e}");
                }

                let failures = match self.inner.repo.get_by_url(url).await {
                    Ok(Some(instance)) => instance.consecutive_failures,
                    Ok(None) => {
                        debug!("searxng reported failure for unknown instance url={url}");
                        return;
                    }
                    Err(e) => {
                        warn!("searxng failure count read failed url={url} error={e}");
                        return;
                    }
                };

                // 熔断开 或 连续失败 ≥2 → 逐出热池 + 转 cooldown + 异步晋升
                if (self.inner.breaker.is_open(&key) || failures >= 2)
                    && self.inner.pool.evict(url).is_some()
                {
                    if let Err(e) = self
                        .inner
                        .repo
                        .transition_state(url, SearxngInstanceState::Cooldown)
                        .await
                    {
                        warn!("searxng cooldown transition failed url={url} error={e}");
                    }
                    let this = self.clone();
                    match self.inner.promotion_sem.clone().try_acquire_owned() {
                        Ok(permit) => {
                            tokio::spawn(async move {
                                let _permit = permit;
                                this.promote_one().await;
                            });
                        }
                        Err(_) => {
                            debug!(
                                "searxng promotion skipped: concurrency limit reached url={url}"
                            );
                        }
                    }
                }
            }
        }
    }

    /// 采集一轮:拉数据集 → 宽容解析 → 准入 → 幂等 upsert → 缺席退役。
    ///
    /// 跨进程防重:执行体由 PG advisory lock 包裹(固定常量键)。拿不到锁时
    /// 本轮跳过(`CollectorReport::skipped()` + `debug!`),不重复采集。
    ///
    /// 失败语义:数据集拉取失败返回 `Err`(含原因)且**登记簿零变更**;
    /// 单条准入拒绝/写入失败只累计 `errors`,不中断整轮(其余实例继续处理)。
    pub async fn collect(&self) -> Result<CollectorReport, SearxngError> {
        let guard = match self
            .inner
            .repo
            .try_acquire_collect_lock(COLLECT_ADVISORY_LOCK_KEY)
            .await
        {
            Ok(Some(guard)) => guard,
            Ok(None) => {
                debug!("searxng collect skipped: advisory lock held by another process");
                return Ok(CollectorReport::skipped());
            }
            Err(e) => {
                warn!("searxng collect lock acquisition failed: {e}");
                return Err(SearxngError::Repository(e.to_string()));
            }
        };

        let result = self.collect_inner().await;
        if let Err(e) = guard.release().await {
            warn!("searxng collect lock release failed: {e}");
        }
        result
    }

    /// 采集执行体(调用方必须已持有采集锁)
    async fn collect_inner(&self) -> Result<CollectorReport, SearxngError> {
        let collector = &self.inner.settings.collector;
        let dataset = collector::fetch_dataset(self.inner.transport.as_ref(), collector).await?;

        let mut report = CollectorReport {
            fetched: dataset.instances.len(),
            ..Default::default()
        };

        // 数据集中「出现过」的规范化 url 集合:本轮的缺席判定依据。
        // 用原始键(含未通过准入的条目):元数据不再满足准入 ≠ 已从数据集消失,
        // 不应因此累计缺席计数。
        let mut present: HashSet<String> = HashSet::with_capacity(dataset.instances.len());

        collector::sync_instances(
            self.inner.repo.as_ref(),
            &dataset,
            collector,
            &mut report,
            &mut present,
        )
        .await;
        collector::retire_absent(self.inner.repo.as_ref(), &present, collector, &mut report)
            .await?;

        info!(
            "searxng collect finished fetched={} seen={} admitted_new={} updated={} retired={} errors={}",
            report.fetched,
            report.seen,
            report.admitted_new,
            report.updated,
            report.retired,
            report.errors
        );
        Ok(report)
    }

    /// 探活一轮:热池全体 + 冷却期满的 cooldown + 前 `candidate_batch` 个候选。
    ///
    /// 状态机迁移由 `Self::apply_probe_outcome` 统一承载(探活通过 → active 入池;
    /// 失败 → cooldown 逐出;连续失败达 `failure_retire_threshold` → retired)。
    pub async fn health_cycle(&self) -> HealthReport {
        let mut report = HealthReport::default();
        let batch = self.inner.settings.health.candidate_batch.max(1);

        // 1) 热池全体
        for member in self.inner.pool.snapshot() {
            self.probe_and_apply(&member.instance, &mut report).await;
        }

        // 2) cooldown 冷却期满(熔断镜像已放行)→ 回池探测
        match self
            .inner
            .repo
            .list_by_state(SearxngInstanceState::Cooldown, batch)
            .await
        {
            Ok(cooldowns) => {
                for instance in cooldowns {
                    if self.inner.pool.contains(&instance.url) {
                        continue;
                    }
                    if self
                        .inner
                        .breaker
                        .is_open(&Self::breaker_key(&instance.url))
                    {
                        continue;
                    }
                    self.probe_and_apply(&instance, &mut report).await;
                }
            }
            Err(e) => {
                warn!("searxng cooldown listing failed: {e}");
                report.errors += 1;
            }
        }

        // 3) 候选补齐(按 uptime 降序、latency 升序,池满即停)
        match self.inner.repo.list_candidates(batch).await {
            Ok(candidates) => {
                for instance in candidates {
                    if self.inner.pool.is_full() {
                        break;
                    }
                    if self.inner.pool.contains(&instance.url) {
                        continue;
                    }
                    self.probe_and_apply(&instance, &mut report).await;
                }
            }
            Err(e) => {
                warn!("searxng candidate listing failed: {e}");
                report.errors += 1;
            }
        }

        info!(
            "searxng health cycle finished probed={} promoted={} demoted={} retired={} errors={} pool={}/{}",
            report.probed,
            report.promoted,
            report.demoted,
            report.retired,
            report.errors,
            self.inner.pool.len(),
            self.inner.pool.target()
        );
        report
    }

    /// 异步晋升一个替补:`list_candidates` → 探测一次 → 通过则 active 入池。
    ///
    /// 幂等:池按 url 去重,多路并发晋升最多补满 target。
    async fn promote_one(&self) {
        if self.inner.pool.is_full() {
            return;
        }
        let candidates = match self.inner.repo.list_candidates(4).await {
            Ok(candidates) => candidates,
            Err(e) => {
                warn!("searxng promotion candidate listing failed: {e}");
                return;
            }
        };
        for candidate in candidates {
            if self.inner.pool.is_full() {
                break;
            }
            if self.inner.pool.contains(&candidate.url) {
                continue;
            }
            let mut report = HealthReport::default();
            self.probe_and_apply(&candidate, &mut report).await;
            if report.promoted > 0 {
                break;
            }
        }
    }

    /// 探测一个实例并按状态机落库/入池/逐出。
    async fn probe_and_apply(&self, instance: &SearxngInstance, report: &mut HealthReport) {
        report.probed += 1;
        let outcome = probe_instance(
            self.inner.transport.as_ref(),
            instance,
            &self.inner.settings.health,
        )
        .await;
        let key = Self::breaker_key(&instance.url);

        if outcome.alive {
            self.apply_alive_outcome(instance, outcome, report, &key)
                .await;
            return;
        }
        self.apply_failed_outcome(instance, outcome, report, &key)
            .await;
    }

    /// 存活路径:熔断记成功 → 落探活结果 → 刷新实例(时延/失败计数/成功时间)
    /// → 状态机转 active → 入池或替换池成员。
    async fn apply_alive_outcome(
        &self,
        instance: &SearxngInstance,
        outcome: ProbeOutcome,
        report: &mut HealthReport,
        key: &str,
    ) {
        self.inner.breaker.record_success(key).await;
        if let Err(e) = self
            .inner
            .repo
            .record_probe_result(&instance.url, true, outcome.latency_ms, None)
            .await
        {
            warn!(
                "searxng probe success record failed url={} error={e}",
                instance.url
            );
        }

        let mut refreshed = instance.clone();
        refreshed.latency_ms = outcome.latency_ms;
        refreshed.consecutive_failures = 0;
        refreshed.last_success_at = Some(chrono::Utc::now());

        if refreshed.state != SearxngInstanceState::Active {
            if let Err(e) = self
                .inner
                .repo
                .transition_state(&refreshed.url, SearxngInstanceState::Active)
                .await
            {
                warn!(
                    "searxng active transition failed url={} error={e}",
                    refreshed.url
                );
                return;
            }
            refreshed.state = SearxngInstanceState::Active;
        }

        if !self.inner.pool.contains(&refreshed.url) {
            if self.add_member(refreshed.clone()) {
                report.promoted += 1;
                info!("searxng instance promoted into pool url={}", refreshed.url);
            }
        } else {
            self.inner.pool.replace(Arc::new(PoolMember {
                instance: refreshed,
            }));
        }
    }

    /// 失败路径:熔断记失败 → 落探活结果 → 逐出池成员 → 连续失败达阈值退役
    /// (cooldown/candidate → retired;active 不允许直迁,先落 cooldown,
    /// 下一轮由 cooldown → retired 完成)。
    async fn apply_failed_outcome(
        &self,
        instance: &SearxngInstance,
        outcome: ProbeOutcome,
        report: &mut HealthReport,
        key: &str,
    ) {
        report.errors += 1;
        let message = outcome.error.unwrap_or_else(|| "probe failed".to_string());
        self.inner.breaker.record_failure(key).await;
        if let Err(e) = self
            .inner
            .repo
            .record_probe_result(&instance.url, false, None, Some(&message))
            .await
        {
            warn!(
                "searxng probe failure record failed url={} error={e}",
                instance.url
            );
        }

        let failures = self
            .inner
            .repo
            .get_by_url(&instance.url)
            .await
            .ok()
            .flatten()
            .map(|i| i.consecutive_failures)
            .unwrap_or(instance.consecutive_failures + 1);

        let evicted = self.inner.pool.evict(&instance.url).is_some();
        if evicted {
            report.demoted += 1;
        }

        // 连续失败达阈值 → 退役(cooldown/candidate → retired;active 不允许直迁,
        // 先落 cooldown,下一轮由 cooldown → retired 完成)
        let retire = failures >= self.inner.settings.health.failure_retire_threshold;
        let target_state = if retire
            && instance
                .state
                .can_transition_to(SearxngInstanceState::Retired)
        {
            report.retired += 1;
            SearxngInstanceState::Retired
        } else {
            SearxngInstanceState::Cooldown
        };
        if instance.state != target_state && instance.state.can_transition_to(target_state) {
            if let Err(e) = self
                .inner
                .repo
                .transition_state(&instance.url, target_state)
                .await
            {
                warn!(
                    "searxng state transition failed url={} to={} error={e}",
                    instance.url, target_state
                );
            }
        }
        debug!(
            "searxng probe failed url={} failures={} state={} error={message}",
            instance.url, failures, target_state
        );
    }

    /// 启动期引导:`seed_urls` 逐条 upsert + 后台重建热池(不阻塞 API 就绪)。
    ///
    /// 重建完成前 [`Self::health`] 返回 `Degraded`(不伪报可用)。
    pub fn spawn_bootstrap(&self) {
        let this = self.clone();
        tokio::spawn(async move {
            let report = this.bootstrap().await;
            info!(
                "searxng bootstrap finished probed={} promoted={} demoted={} errors={} pool={}/{} healthy_bootstrapped=true",
                report.probed,
                report.promoted,
                report.demoted,
                report.errors,
                this.inner.pool.len(),
                this.inner.pool.target()
            );
        });
    }

    /// 启动期重建(可等待版本,测试直接调用):
    ///
    /// 1. `seed_urls` 逐条 `upsert_seed`(非法 URL 记 warn 跳过,不阻断启动);
    /// 2. 取 `active` 与高分 `candidate` 逐个探活,通过者入池至 `pool_target`;
    ///    失败者按状态机落 `cooldown`(或达阈值转 `retired`);
    /// 3. 标记重建完成(此后 `health()` 按三档口径返回)。
    pub async fn bootstrap(&self) -> HealthReport {
        for raw in &self.inner.settings.seed_urls {
            match normalize_instance_url(raw) {
                Ok(url) => match self.inner.repo.upsert_seed(&url).await {
                    Ok(outcome) => {
                        info!("searxng seed upserted url={url} outcome={outcome:?}")
                    }
                    Err(e) => warn!("searxng seed upsert failed url={url} error={e}"),
                },
                Err(e) => warn!("searxng seed url rejected raw={raw} error={e}"),
            }
        }

        let mut report = HealthReport::default();
        let limit = (self.inner.pool.target().max(1) * 2) as u64;
        let mut seeds: Vec<SearxngInstance> = Vec::new();
        match self
            .inner
            .repo
            .list_by_state(SearxngInstanceState::Active, limit)
            .await
        {
            Ok(instances) => seeds.extend(instances),
            Err(e) => {
                warn!("searxng bootstrap active listing failed: {e}");
                report.errors += 1;
            }
        }
        match self.inner.repo.list_candidates(limit).await {
            Ok(instances) => seeds.extend(instances),
            Err(e) => {
                warn!("searxng bootstrap candidate listing failed: {e}");
                report.errors += 1;
            }
        }

        for instance in seeds {
            if self.inner.pool.is_full() {
                break;
            }
            if self.inner.pool.contains(&instance.url) {
                continue;
            }
            self.probe_and_apply(&instance, &mut report).await;
        }

        self.mark_bootstrap_done();
        report
    }
}

#[async_trait::async_trait]
impl SearxngInstanceService for SearxngInstanceServiceImpl {
    fn acquire(&self, exclude: &[String]) -> Option<Arc<PoolMember>> {
        SearxngInstanceServiceImpl::acquire(self, exclude)
    }

    async fn report(&self, url: &str, outcome: InstanceOutcome) {
        SearxngInstanceServiceImpl::report(self, url, outcome).await
    }

    async fn collect(&self) -> Result<CollectorReport, SearxngError> {
        SearxngInstanceServiceImpl::collect(self).await
    }

    async fn health_cycle(&self) -> HealthReport {
        SearxngInstanceServiceImpl::health_cycle(self).await
    }

    async fn pool_snapshot(&self) -> PoolSnapshot {
        SearxngInstanceServiceImpl::pool_snapshot(self).await
    }

    async fn list_instances(
        &self,
        limit: u64,
        offset: u64,
    ) -> Result<Vec<SearxngInstance>, SearxngError> {
        SearxngInstanceServiceImpl::list_instances(self, limit, offset).await
    }

    fn health(&self) -> EngineHealth {
        SearxngInstanceServiceImpl::health(self)
    }
}

#[cfg(test)]
mod tests {
    use super::*;
    use crate::domain::services::searxng::test_support::{
        shared_repo, test_settings, InMemorySearxngRepo, Scripted, ScriptedTransport,
    };

    fn fixture_payload() -> serde_json::Value {
        let raw = include_str!("../../../../tests/fixtures/searxng/instances_sample.json");
        serde_json::from_str(raw).expect("fixture json")
    }

    fn service(
        repo: Arc<InMemorySearxngRepo>,
        transport: Arc<ScriptedTransport>,
    ) -> SearxngInstanceServiceImpl {
        SearxngInstanceServiceImpl::new(repo, transport, test_settings())
    }

    /// 构造一条 active 实例并放入池(模拟已跑起的热池)
    fn seed_pool(svc: &SearxngInstanceServiceImpl, repo: &Arc<InMemorySearxngRepo>, url: &str) {
        let instance = InMemorySearxngRepo::seed_instance(url, SearxngInstanceState::Active);
        repo.insert_raw(instance.clone());
        svc.add_member(instance);
    }

    fn alive_body() -> serde_json::Value {
        serde_json::json!({ "results": [] })
    }

    // ========== 采集(T006/T007 回归) ==========

    #[tokio::test]
    async fn test_collect_counts_match_registry() {
        let repo = shared_repo();
        let transport = Arc::new(ScriptedTransport::new(vec![Scripted::Json {
            status: 200,
            body: fixture_payload(),
        }]));
        let svc = service(repo.clone(), transport.clone());

        let report = svc.collect().await.expect("collect succeeds");

        // fixture 10 条:通过准入的是 working / api-detail / minimal / unknown-fields /
        // bad-metadata(5 条),其余被拒(api-off / http / low uptime / tor / 内网)。
        assert_eq!(report.fetched, 10);
        assert_eq!(report.seen, 5);
        assert_eq!(report.admitted_new, 5);
        assert_eq!(report.updated, 0);
        assert_eq!(report.retired, 0);
        assert_eq!(report.errors, 5);
        assert_eq!(repo.len(), 5);
        assert_eq!(transport.calls.load(std::sync::atomic::Ordering::SeqCst), 1);
    }

    #[tokio::test]
    async fn test_collect_second_round_updates_not_inserts() {
        let repo = shared_repo();
        let payload = fixture_payload();
        let transport = Arc::new(ScriptedTransport::new(vec![
            Scripted::Json {
                status: 200,
                body: payload.clone(),
            },
            Scripted::Json {
                status: 200,
                body: payload,
            },
        ]));
        let svc = service(repo.clone(), transport);

        svc.collect().await.expect("first round");
        let second = svc.collect().await.expect("second round");

        assert_eq!(second.admitted_new, 0);
        assert_eq!(second.updated, 5);
        assert_eq!(repo.len(), 5, "idempotent upsert must not duplicate rows");
    }

    #[tokio::test]
    async fn test_collect_fetch_failure_makes_zero_repo_calls() {
        let repo = shared_repo();
        let transport = Arc::new(ScriptedTransport::new(vec![Scripted::Network(
            "connection refused".to_string(),
        )]));
        let svc = service(repo.clone(), transport);

        let err = svc.collect().await.expect_err("fetch failure must surface");
        assert!(matches!(err, SearxngError::DatasetFetch(_)), "got: {err}");
        assert!(err.to_string().contains("connection refused"));
        assert_eq!(
            repo.dataset_upserts
                .load(std::sync::atomic::Ordering::SeqCst),
            0,
            "registry must stay untouched when the dataset is unreachable"
        );
        assert!(repo.is_empty());
    }

    #[tokio::test]
    async fn test_collect_non_2xx_dataset_returns_err() {
        let repo = shared_repo();
        let transport = Arc::new(ScriptedTransport::new(vec![Scripted::Json {
            status: 503,
            body: serde_json::Value::Null,
        }]));
        let svc = service(repo, transport);

        let err = svc.collect().await.expect_err("503 must surface");
        assert!(matches!(err, SearxngError::DatasetFetch(_)));
        assert!(err.to_string().contains("503"));
    }

    #[tokio::test]
    async fn test_collect_network_error_is_structured() {
        let repo = shared_repo();
        let transport = Arc::new(ScriptedTransport::new(vec![Scripted::Timeout]));
        let svc = service(repo, transport);

        let err = svc.collect().await.expect_err("timeout must surface");
        assert!(matches!(err, SearxngError::DatasetFetch(_)));
        assert!(err.to_string().contains("timed out"));
    }

    #[tokio::test]
    async fn test_collect_retires_absent_dataset_instance_but_not_seed() {
        let repo = shared_repo();
        repo.insert_raw(InMemorySearxngRepo::seed_instance(
            "https://searx.seed.example",
            SearxngInstanceState::Active,
        ));
        repo.insert_raw(InMemorySearxngRepo::dataset_instance(
            "https://searx.absent.example",
            SearxngInstanceState::Active,
        ));

        let mut settings = test_settings();
        settings.collector.retire_miss_cycles = 0; // 首次缺席即退役
        let transport = Arc::new(ScriptedTransport::new(vec![
            Scripted::Json {
                status: 200,
                body: serde_json::json!({ "instances": {} }),
            },
            Scripted::Json {
                status: 200,
                body: serde_json::json!({ "instances": {} }),
            },
        ]));
        let svc = SearxngInstanceServiceImpl::new(repo.clone(), transport, settings);

        // 第一次:active 不允许直迁 retired → 落 cooldown,不计数退役
        let first = svc.collect().await.expect("first collect");
        assert_eq!(first.retired, 0);
        assert_eq!(
            repo.get("https://searx.absent.example").expect("row").state,
            SearxngInstanceState::Cooldown
        );

        // 第二次:cooldown → retired
        let second = svc.collect().await.expect("second collect");
        assert_eq!(second.retired, 1);
        assert_eq!(
            repo.get("https://searx.absent.example").expect("row").state,
            SearxngInstanceState::Retired
        );
        assert_eq!(
            repo.get("https://searx.seed.example").expect("row").state,
            SearxngInstanceState::Active
        );
    }

    #[tokio::test]
    async fn test_collect_rejects_malformed_payload_without_panicking() {
        let repo = shared_repo();
        let transport = Arc::new(ScriptedTransport::new(vec![Scripted::Json {
            status: 200,
            body: serde_json::json!("not-an-object"),
        }]));
        let svc = service(repo.clone(), transport);

        let err = svc
            .collect()
            .await
            .expect_err("malformed payload must surface");
        assert!(matches!(err, SearxngError::DatasetFetch(_)));
        assert!(repo.is_empty());
    }

    #[tokio::test]
    async fn test_collect_skips_when_lock_is_held() {
        let repo = shared_repo();
        let transport = Arc::new(ScriptedTransport::new(vec![]));
        let svc = service(repo.clone(), transport.clone());

        let guard = repo
            .try_acquire_collect_lock(COLLECT_ADVISORY_LOCK_KEY)
            .await
            .expect("lock attempt")
            .expect("lock is free");

        let report = svc.collect().await.expect("skip is not an error");
        assert!(report.skipped, "held lock must make the round skip");
        assert_eq!(report.fetched, 0);
        assert_eq!(
            transport.calls.load(std::sync::atomic::Ordering::SeqCst),
            0,
            "a skipped round must not touch the dataset"
        );

        guard.release().await.expect("release");

        let transport2 = Arc::new(ScriptedTransport::new(vec![Scripted::Json {
            status: 200,
            body: serde_json::json!({ "instances": {} }),
        }]));
        let svc2 = service(repo, transport2);
        let report2 = svc2.collect().await.expect("run after release");
        assert!(!report2.skipped);
    }

    #[tokio::test]
    async fn test_collect_releases_lock_after_failure() {
        let repo = shared_repo();
        let transport = Arc::new(ScriptedTransport::new(vec![Scripted::Network(
            "boom".to_string(),
        )]));
        let svc = service(repo.clone(), transport);

        let _ = svc.collect().await.expect_err("fetch failure");

        let guard = repo
            .try_acquire_collect_lock(COLLECT_ADVISORY_LOCK_KEY)
            .await
            .expect("lock attempt")
            .expect("lock must be free after a failed round");
        guard.release().await.expect("release");
    }

    #[test]
    fn test_collector_report_skipped_shape() {
        let report = CollectorReport::skipped();
        assert!(report.skipped);
        assert_eq!(report.fetched, 0);
        assert_eq!(report.errors, 0);
    }

    #[test]
    fn test_settings_snapshot_is_accessible() {
        let repo = shared_repo();
        let transport = Arc::new(ScriptedTransport::new(vec![]));
        let svc = service(repo, transport);
        assert_eq!(svc.settings().pool_target, 3);
    }

    // ========== 健康口径(T010) ==========

    #[test]
    fn test_health_is_degraded_before_bootstrap_completes() {
        let svc = service(shared_repo(), Arc::new(ScriptedTransport::new(vec![])));
        assert_eq!(svc.health(), EngineHealth::Degraded);
    }

    #[test]
    fn test_health_three_tiers() {
        let repo = shared_repo();
        let svc = service(repo.clone(), Arc::new(ScriptedTransport::new(vec![])));
        svc.mark_bootstrap_done();

        // 空池 → Unhealthy
        assert_eq!(svc.health(), EngineHealth::Unhealthy);

        // 1/3 → Degraded
        seed_pool(&svc, &repo, "https://a.example");
        assert_eq!(svc.health(), EngineHealth::Degraded);

        // 3/3 → Healthy
        seed_pool(&svc, &repo, "https://b.example");
        seed_pool(&svc, &repo, "https://c.example");
        assert_eq!(svc.health(), EngineHealth::Healthy);
    }

    #[test]
    fn test_engine_health_is_available_semantics() {
        assert!(EngineHealth::Healthy.is_available());
        assert!(EngineHealth::Degraded.is_available());
        assert!(!EngineHealth::Unhealthy.is_available());
    }

    // ========== 熔断与晋升(T010) ==========

    #[tokio::test]
    async fn test_repeated_failures_evict_member_and_promote_candidate() {
        let repo = shared_repo();
        // 池内 1 个成员 + 1 个候补(候补探测返回存活)
        let svc = service(
            repo.clone(),
            Arc::new(ScriptedTransport::repeat(
                Scripted::Json {
                    status: 200,
                    body: alive_body(),
                },
                10,
            )),
        );
        svc.mark_bootstrap_done();
        seed_pool(&svc, &repo, "https://failing.example");
        repo.insert_raw(InMemorySearxngRepo::dataset_instance(
            "https://healthy.example",
            SearxngInstanceState::Candidate,
        ));

        // 连续失败 3 次(默认 failure_threshold=3)→ 熔断开 + 逐出
        for _ in 0..3 {
            svc.report(
                "https://failing.example",
                InstanceOutcome::Failure {
                    error: SearxngError::RateLimited,
                },
            )
            .await;
        }

        assert_eq!(
            repo.get("https://failing.example").expect("row").state,
            SearxngInstanceState::Cooldown
        );

        // 异步晋升:候补被探测通过后入池(轮询等待,避免假阴性)
        let deadline = tokio::time::Instant::now() + Duration::from_secs(2);
        loop {
            if svc.acquire(&[]).map(|m| m.url().to_string()).as_deref()
                == Some("https://healthy.example")
            {
                break;
            }
            assert!(
                tokio::time::Instant::now() < deadline,
                "candidate should be promoted into the pool"
            );
            tokio::time::sleep(Duration::from_millis(10)).await;
        }

        // 已逐出成员不再被 acquire 返回
        let acquired = svc.acquire(&[]).expect("member");
        assert_ne!(acquired.url(), "https://failing.example");
        assert_eq!(svc.pool_len(), 1);
        assert_eq!(svc.health(), EngineHealth::Degraded);
    }

    /// 晋升并发护栏:permit 耗尽时丢弃晋升,候补保持 Candidate
    #[tokio::test]
    async fn test_promotion_skipped_when_concurrency_limit_reached() {
        let repo = shared_repo();
        let svc = service(
            repo.clone(),
            Arc::new(ScriptedTransport::repeat(
                Scripted::Json {
                    status: 200,
                    body: alive_body(),
                },
                10,
            )),
        );
        svc.mark_bootstrap_done();
        seed_pool(&svc, &repo, "https://failing.example");
        repo.insert_raw(InMemorySearxngRepo::dataset_instance(
            "https://healthy.example",
            SearxngInstanceState::Candidate,
        ));

        // 占满全部 permit:后续晋升必须被丢弃(不排队)
        let mut permits = Vec::new();
        for _ in 0..PROMOTION_MAX_CONCURRENCY {
            permits.push(
                svc.inner
                    .promotion_sem
                    .clone()
                    .try_acquire_owned()
                    .expect("permit"),
            );
        }

        for _ in 0..3 {
            svc.report(
                "https://failing.example",
                InstanceOutcome::Failure {
                    error: SearxngError::RateLimited,
                },
            )
            .await;
        }

        // 给异步任务留出窗口;没有 permit 时不应发生晋升
        tokio::time::sleep(Duration::from_millis(100)).await;
        assert!(
            svc.acquire(&[])
                .map(|m| m.url().to_string())
                .as_deref()
                .is_none_or(|url| url != "https://healthy.example"),
            "promotion must be skipped while the semaphore is exhausted"
        );
        assert_eq!(
            repo.get("https://healthy.example").expect("row").state,
            SearxngInstanceState::Candidate
        );

        drop(permits);
    }

    #[tokio::test]
    async fn test_success_report_resets_failures_and_updates_latency() {
        let repo = shared_repo();
        let svc = service(repo.clone(), Arc::new(ScriptedTransport::new(vec![])));
        svc.mark_bootstrap_done();
        seed_pool(&svc, &repo, "https://a.example");

        svc.report(
            "https://a.example",
            InstanceOutcome::Failure {
                error: SearxngError::BadHttpStatus(500),
            },
        )
        .await;
        assert_eq!(
            repo.get("https://a.example")
                .expect("row")
                .consecutive_failures,
            1
        );
        // 失败 1 次未达逐出阈值(≥2)
        assert_eq!(svc.pool_len(), 1);

        svc.report(
            "https://a.example",
            InstanceOutcome::Success { latency_ms: 42 },
        )
        .await;
        let row = repo.get("https://a.example").expect("row");
        assert_eq!(row.consecutive_failures, 0);
        assert_eq!(row.latency_ms, Some(42));
        let snapshot = svc.pool_snapshot().await;
        assert_eq!(snapshot.members[0].latency_ms, Some(42));
    }

    #[tokio::test]
    async fn test_report_for_unknown_instance_is_ignored() {
        let repo = shared_repo();
        let svc = service(repo.clone(), Arc::new(ScriptedTransport::new(vec![])));
        svc.report(
            "https://unknown.example",
            InstanceOutcome::Failure {
                error: SearxngError::Network("boom".to_string()),
            },
        )
        .await;
        assert_eq!(svc.pool_len(), 0);
    }

    // ========== 探活轮次(T010) ==========

    #[tokio::test]
    async fn test_health_cycle_promotes_candidates_and_respects_target() {
        let repo = shared_repo();
        for i in 0..5 {
            repo.insert_raw(InMemorySearxngRepo::dataset_instance(
                &format!("https://candidate-{i}.example"),
                SearxngInstanceState::Candidate,
            ));
        }
        let svc = service(
            repo.clone(),
            Arc::new(ScriptedTransport::repeat(
                Scripted::Json {
                    status: 200,
                    body: alive_body(),
                },
                50,
            )),
        );
        svc.mark_bootstrap_done();

        let report = svc.health_cycle().await;

        assert_eq!(svc.pool_len(), 3, "pool must stop at pool_target");
        assert_eq!(report.promoted, 3);
        assert_eq!(
            repo.count_by_state(SearxngInstanceState::Active)
                .await
                .expect("count"),
            3
        );
        assert_eq!(svc.health(), EngineHealth::Healthy);
    }

    #[tokio::test]
    async fn test_health_cycle_retires_candidate_after_threshold_failures() {
        let repo = shared_repo();
        repo.insert_raw(InMemorySearxngRepo::dataset_instance(
            "https://bad.example",
            SearxngInstanceState::Candidate,
        ));
        let mut settings = test_settings();
        settings.health.failure_retire_threshold = 1;
        let svc = SearxngInstanceServiceImpl::new(
            repo.clone(),
            Arc::new(ScriptedTransport::repeat(
                Scripted::Json {
                    status: 200,
                    body: serde_json::json!({ "no_results": true }),
                },
                10,
            )),
            settings,
        );
        svc.mark_bootstrap_done();

        let report = svc.health_cycle().await;

        assert_eq!(report.errors, 1);
        assert_eq!(report.retired, 1);
        assert_eq!(
            repo.get("https://bad.example").expect("row").state,
            SearxngInstanceState::Retired
        );
        assert_eq!(svc.pool_len(), 0);
        assert_eq!(svc.health(), EngineHealth::Unhealthy);
    }

    #[tokio::test]
    async fn test_health_cycle_demotes_failed_pool_member_to_cooldown() {
        let repo = shared_repo();
        let svc = service(
            repo.clone(),
            Arc::new(ScriptedTransport::repeat(
                Scripted::Json {
                    status: 429,
                    body: serde_json::Value::Null,
                },
                10,
            )),
        );
        svc.mark_bootstrap_done();
        seed_pool(&svc, &repo, "https://throttled.example");

        let report = svc.health_cycle().await;

        assert_eq!(report.demoted, 1);
        assert_eq!(
            repo.get("https://throttled.example").expect("row").state,
            SearxngInstanceState::Cooldown
        );
        assert_eq!(svc.pool_len(), 0);
    }

    #[tokio::test]
    async fn test_health_cycle_revives_cooldown_when_probe_passes() {
        let repo = shared_repo();
        repo.insert_raw(InMemorySearxngRepo::dataset_instance(
            "https://recovered.example",
            SearxngInstanceState::Cooldown,
        ));
        let svc = service(
            repo.clone(),
            Arc::new(ScriptedTransport::repeat(
                Scripted::Json {
                    status: 200,
                    body: alive_body(),
                },
                5,
            )),
        );
        svc.mark_bootstrap_done();

        let report = svc.health_cycle().await;

        assert_eq!(report.promoted, 1);
        assert_eq!(
            repo.get("https://recovered.example").expect("row").state,
            SearxngInstanceState::Active
        );
        assert_eq!(svc.pool_len(), 1);
    }

    // ========== 启动期重建与 seed 引导(T011) ==========

    #[tokio::test]
    async fn test_bootstrap_promotes_seed_that_passes_probe() {
        let repo = shared_repo();
        let mut settings = test_settings();
        settings.seed_urls = vec!["https://seed.example/".to_string()];
        settings.pool_target = 3;
        let svc = SearxngInstanceServiceImpl::new(
            repo.clone(),
            Arc::new(ScriptedTransport::repeat(
                Scripted::Json {
                    status: 200,
                    body: alive_body(),
                },
                5,
            )),
            settings,
        );

        // 重建前:Degraded(不伪报可用)
        assert_eq!(svc.health(), EngineHealth::Degraded);

        let report = svc.bootstrap().await;

        assert_eq!(report.promoted, 1);
        assert_eq!(svc.pool_len(), 1);
        let acquired = svc.acquire(&[]).expect("seed in pool");
        assert_eq!(acquired.url(), "https://seed.example");
        assert_eq!(
            repo.get("https://seed.example").expect("row").state,
            SearxngInstanceState::Active
        );
    }

    #[tokio::test]
    async fn test_bootstrap_keeps_failing_seed_out_of_pool_and_cooldowns_it() {
        let repo = shared_repo();
        let mut settings = test_settings();
        settings.seed_urls = vec!["https://dead-seed.example".to_string()];
        let svc = SearxngInstanceServiceImpl::new(
            repo.clone(),
            Arc::new(ScriptedTransport::repeat(
                Scripted::Json {
                    status: 503,
                    body: serde_json::Value::Null,
                },
                5,
            )),
            settings,
        );

        let report = svc.bootstrap().await;

        assert_eq!(report.promoted, 0);
        assert_eq!(report.errors, 1);
        assert_eq!(svc.pool_len(), 0);
        assert_eq!(
            repo.get("https://dead-seed.example").expect("row").state,
            SearxngInstanceState::Cooldown
        );
        // 重建已完成:空池按三档口径为 Unhealthy
        assert_eq!(svc.health(), EngineHealth::Unhealthy);
    }

    #[tokio::test]
    async fn test_bootstrap_skips_invalid_seed_url_without_blocking() {
        let repo = shared_repo();
        let mut settings = test_settings();
        settings.seed_urls = vec!["http://127.0.0.1:9000".to_string(), "not-a-url".to_string()];
        let svc = SearxngInstanceServiceImpl::new(
            repo.clone(),
            Arc::new(ScriptedTransport::new(vec![])),
            settings,
        );

        let report = svc.bootstrap().await;

        assert_eq!(report.promoted, 0);
        assert!(
            repo.is_empty(),
            "invalid seed urls must not reach the registry"
        );
    }

    #[tokio::test]
    async fn test_trait_object_disabled_shape_is_option_none() {
        // disabled 语义:服务不构造 → Option<Arc<dyn SearxngInstanceService>> 为 None
        // (由 bootstrap/services.rs 装配层断言;这里锁定 trait 可装箱为 trait object)
        let svc: Arc<dyn SearxngInstanceService> = Arc::new(service(
            shared_repo(),
            Arc::new(ScriptedTransport::new(vec![])),
        ));
        assert!(svc.acquire(&[]).is_none());
        let snapshot = svc.pool_snapshot().await;
        assert_eq!(snapshot.target, 3);
        assert!(snapshot.members.is_empty());
    }

    #[tokio::test]
    async fn test_pool_snapshot_reports_breaker_mirror_and_health() {
        let repo = shared_repo();
        let svc = service(repo.clone(), Arc::new(ScriptedTransport::new(vec![])));
        svc.mark_bootstrap_done();
        seed_pool(&svc, &repo, "https://a.example");

        let snapshot = svc.pool_snapshot().await;
        assert_eq!(snapshot.target, 3);
        assert_eq!(snapshot.members.len(), 1);
        assert_eq!(snapshot.members[0].url, "https://a.example");
        assert_eq!(snapshot.members[0].breaker, BreakerMirror::Closed);
        assert_eq!(snapshot.health, EngineHealth::Degraded);
    }
}