daaki-message 0.2.0

RFC 5322 email message parser and builder
Documentation
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
417
418
419
420
421
422
423
424
425
426
427
428
429
430
431
432
433
434
435
436
437
438
439
440
441
442
443
444
445
446
447
448
449
450
451
452
453
454
455
456
457
458
459
460
461
462
463
464
465
466
467
468
469
470
471
472
473
474
475
476
477
478
479
480
481
482
483
484
485
486
487
488
489
490
491
492
493
494
495
496
497
498
499
500
501
502
503
504
505
506
507
508
509
510
511
512
513
514
515
516
517
518
519
520
521
522
523
524
525
526
527
528
529
530
531
532
533
534
535
536
537
538
539
540
541
542
543
544
545
546
547
548
549
550
551
552
553
554
555
556
557
558
559
560
561
562
563
564
565
566
567
568
569
570
571
572
573
574
575
576
577
578
579
580
581
582
583
584
585
586
587
588
589
590
591
592
593
594
595
596
597
598
599
600
601
602
603
604
605
606
607
608
609
610
611
612
613
614
615
616
617
618
619
620
621
622
623
624
625
626
627
628
629
630
631
632
633
634
635
636
637
638
639
640
641
642
643
644
645
646
647
648
649
650
651
652
653
654
655
656
657
658
659
660
661
662
663
664
665
666
667
668
669
670
671
672
673
674
675
676
677
678
679
680
681
682
683
684
685
686
687
688
689
690
691
692
693
694
695
696
697
698
699
700
701
702
703
704
705
706
707
708
709
710
711
712
713
714
715
716
717
718
719
720
721
722
723
724
725
726
727
728
729
730
731
732
733
734
735
736
737
738
739
740
741
742
743
744
745
746
747
748
749
750
751
752
753
754
755
756
757
758
759
760
761
762
763
764
765
766
767
768
769
770
771
772
773
774
775
776
777
778
779
780
781
782
783
784
785
786
787
788
789
790
791
792
793
794
795
796
797
798
799
800
801
802
803
804
805
806
807
808
809
810
811
812
813
814
815
816
817
818
819
820
821
822
823
824
825
826
827
828
829
830
831
832
833
834
835
836
837
838
839
840
841
842
843
844
845
846
847
848
849
850
851
852
853
854
855
856
857
858
859
860
861
862
863
864
865
866
867
868
869
870
871
872
873
874
875
876
877
878
879
880
881
882
883
884
885
886
887
888
889
890
891
892
893
894
895
896
897
898
899
900
901
902
903
904
905
906
907
908
909
910
911
912
913
914
915
916
917
918
919
920
921
922
923
924
925
926
927
928
929
930
931
932
933
934
935
936
937
938
939
940
941
942
943
944
945
946
947
948
949
950
951
952
953
954
955
956
957
958
959
960
961
962
963
964
965
966
967
968
969
970
971
972
973
974
975
976
977
978
979
980
981
982
983
984
985
986
987
988
989
990
991
992
993
994
995
996
997
998
999
1000
1001
1002
1003
1004
1005
1006
1007
1008
1009
1010
1011
1012
1013
1014
1015
1016
1017
1018
1019
1020
1021
1022
1023
1024
1025
1026
1027
1028
1029
1030
1031
1032
1033
1034
1035
1036
1037
1038
1039
1040
1041
1042
1043
1044
1045
1046
1047
1048
1049
1050
1051
1052
1053
1054
1055
1056
1057
1058
1059
1060
1061
1062
1063
1064
1065
1066
1067
1068
1069
1070
1071
1072
1073
1074
1075
1076
1077
1078
1079
1080
1081
1082
1083
1084
1085
1086
1087
1088
1089
1090
1091
1092
1093
1094
1095
1096
1097
1098
1099
1100
1101
1102
1103
1104
1105
1106
1107
1108
1109
1110
1111
1112
1113
1114
1115
1116
1117
1118
1119
1120
1121
1122
1123
1124
1125
1126
1127
1128
1129
1130
1131
1132
1133
1134
1135
1136
1137
1138
1139
1140
1141
1142
1143
1144
1145
1146
1147
1148
1149
1150
1151
1152
1153
1154
1155
1156
1157
1158
1159
1160
1161
1162
1163
1164
1165
1166
1167
1168
1169
1170
1171
1172
1173
1174
1175
1176
1177
1178
1179
1180
1181
1182
1183
1184
1185
1186
1187
1188
1189
1190
1191
1192
1193
1194
1195
1196
1197
1198
1199
1200
1201
1202
1203
1204
1205
1206
1207
1208
1209
1210
1211
1212
1213
1214
1215
1216
1217
1218
1219
1220
1221
1222
1223
1224
1225
1226
1227
1228
1229
1230
1231
1232
1233
1234
1235
1236
1237
1238
1239
1240
1241
1242
1243
1244
1245
1246
1247
1248
1249
1250
1251
1252
1253
1254
1255
1256
1257
1258
1259
1260
1261
1262
1263
1264
1265
1266
1267
1268
1269
1270
1271
1272
1273
1274
1275
1276
1277
1278
1279
1280
1281
1282
1283
1284
1285
1286
1287
1288
1289
1290
1291
1292
1293
1294
1295
1296
1297
1298
1299
1300
1301
1302
1303
1304
1305
1306
1307
1308
1309
1310
1311
1312
1313
1314
1315
1316
1317
1318
1319
1320
1321
1322
1323
1324
1325
1326
1327
1328
1329
1330
1331
1332
1333
1334
1335
1336
1337
1338
1339
1340
1341
1342
1343
1344
1345
1346
1347
1348
1349
1350
1351
1352
1353
1354
1355
1356
1357
1358
1359
1360
1361
1362
1363
1364
1365
1366
1367
1368
1369
1370
1371
1372
1373
1374
1375
1376
1377
1378
1379
1380
1381
1382
1383
1384
1385
1386
1387
1388
1389
1390
1391
1392
1393
1394
1395
1396
1397
1398
1399
1400
1401
1402
1403
1404
1405
1406
1407
1408
1409
1410
1411
1412
1413
1414
1415
1416
1417
1418
1419
1420
1421
1422
1423
1424
1425
1426
1427
1428
1429
1430
1431
1432
1433
1434
1435
1436
1437
1438
1439
1440
1441
1442
1443
1444
1445
1446
1447
1448
1449
1450
1451
1452
1453
1454
1455
1456
1457
1458
1459
1460
1461
1462
1463
1464
1465
1466
1467
1468
1469
1470
1471
1472
1473
1474
1475
1476
1477
1478
1479
1480
1481
1482
1483
1484
1485
1486
1487
1488
1489
1490
1491
1492
1493
1494
1495
1496
1497
1498
1499
1500
1501
1502
1503
1504
1505
1506
1507
1508
1509
1510
1511
1512
1513
1514
1515
1516
1517
1518
1519
1520
1521
1522
1523
1524
1525
1526
1527
1528
1529
1530
1531
1532
1533
1534
1535
1536
1537
1538
1539
1540
1541
1542
1543
1544
1545
1546
1547
1548
1549
1550
1551
1552
1553
1554
1555
1556
1557
1558
1559
1560
1561
1562
1563
1564
1565
1566
1567
1568
1569
1570
1571
1572
1573
1574
1575
1576
1577
1578
1579
1580
1581
1582
1583
1584
1585
1586
1587
1588
1589
1590
1591
1592
1593
1594
1595
1596
1597
1598
1599
1600
1601
1602
1603
1604
1605
1606
1607
1608
1609
1610
1611
1612
1613
1614
1615
1616
1617
1618
1619
1620
1621
1622
1623
1624
1625
1626
1627
1628
1629
1630
1631
1632
1633
1634
1635
1636
1637
1638
1639
1640
1641
1642
1643
1644
1645
1646
1647
1648
1649
1650
1651
1652
1653
1654
1655
1656
1657
1658
1659
1660
1661
1662
1663
1664
1665
1666
1667
1668
1669
1670
1671
1672
1673
1674
1675
1676
1677
1678
1679
1680
1681
1682
1683
1684
1685
1686
1687
1688
1689
1690
1691
1692
1693
1694
1695
1696
1697
1698
1699
1700
1701
1702
1703
1704
1705
1706
1707
1708
1709
1710
1711
1712
1713
1714
1715
1716
1717
1718
1719
1720
1721
1722
1723
1724
1725
1726
1727
1728
1729
1730
1731
1732
1733
1734
1735
1736
1737
1738
1739
1740
1741
1742
1743
1744
1745
1746
1747
1748
1749
1750
1751
1752
1753
1754
1755
1756
1757
1758
1759
1760
1761
1762
1763
1764
1765
1766
1767
1768
1769
1770
1771
1772
1773
1774
1775
1776
1777
1778
1779
1780
1781
1782
1783
1784
1785
1786
1787
1788
1789
1790
1791
1792
1793
1794
1795
1796
1797
1798
1799
1800
1801
1802
1803
1804
1805
1806
1807
1808
1809
1810
1811
1812
1813
1814
1815
1816
1817
1818
1819
1820
1821
1822
1823
1824
1825
1826
1827
1828
1829
1830
1831
1832
1833
1834
1835
1836
1837
1838
1839
1840
1841
1842
1843
1844
1845
1846
1847
1848
1849
1850
1851
1852
1853
1854
1855
1856
1857
1858
1859
1860
1861
1862
1863
1864
1865
1866
1867
1868
1869
1870
1871
1872
1873
1874
1875
1876
1877
1878
1879
1880
1881
1882
1883
1884
1885
1886
1887
1888
1889
1890
1891
1892
1893
1894
1895
1896
1897
1898
1899
1900
1901
1902
1903
1904
1905
1906
1907
1908
1909
1910
1911
1912
1913
1914
1915
1916
1917
1918
1919
1920
1921
1922
1923
1924
1925
1926
1927
1928
1929
1930
1931
1932
1933
1934
1935
1936
1937
1938
1939
1940
1941
1942
1943
1944
1945
1946
1947
1948
1949
1950
1951
1952
1953
1954
1955
1956
1957
1958
1959
1960
1961
1962
1963
1964
1965
1966
1967
1968
1969
1970
1971
1972
1973
1974
1975
1976
1977
1978
1979
1980
1981
1982
1983
1984
1985
1986
1987
1988
1989
1990
1991
1992
1993
1994
1995
1996
1997
1998
1999
2000
2001
2002
2003
2004
2005
2006
2007
2008
2009
2010
2011
2012
2013
2014
2015
2016
2017
2018
2019
2020
2021
2022
2023
2024
2025
2026
2027
2028
2029
2030
2031
2032
2033
2034
2035
2036
2037
2038
2039
2040
2041
2042
2043
2044
2045
2046
2047
2048
2049
2050
2051
2052
2053
2054
2055
2056
2057
2058
2059
2060
2061
2062
2063
2064
2065
2066
2067
2068
2069
2070
2071
2072
2073
2074
2075
2076
2077
2078
2079
2080
2081
2082
2083
2084
2085
2086
2087
2088
2089
2090
2091
2092
2093
2094
2095
2096
2097
2098
2099
2100
2101
2102
2103
2104
2105
2106
2107
2108
2109
2110
2111
2112
2113
2114
2115
2116
2117
2118
2119
2120
2121
2122
2123
2124
2125
2126
2127
2128
2129
2130
2131
2132
2133
2134
2135
2136
2137
2138
2139
2140
2141
2142
2143
2144
2145
2146
2147
2148
2149
2150
2151
2152
2153
2154
2155
2156
2157
2158
2159
2160
2161
2162
2163
2164
2165
2166
2167
2168
2169
2170
2171
2172
2173
2174
2175
2176
2177
2178
2179
2180
2181
2182
2183
2184
2185
2186
2187
2188
2189
2190
2191
2192
2193
2194
2195
2196
2197
2198
2199
2200
2201
2202
2203
2204
2205
2206
2207
2208
2209
2210
2211
2212
2213
2214
2215
2216
2217
2218
2219
2220
2221
2222
2223
2224
2225
2226
2227
2228
2229
2230
2231
2232
2233
2234
2235
2236
2237
2238
2239
2240
2241
2242
2243
2244
2245
2246
2247
2248
2249
2250
2251
2252
2253
2254
2255
2256
2257
2258
2259
2260
2261
2262
2263
2264
2265
2266
2267
2268
2269
2270
2271
2272
2273
2274
2275
2276
2277
2278
2279
2280
2281
2282
2283
2284
2285
2286
2287
2288
2289
2290
2291
2292
2293
2294
2295
2296
2297
2298
2299
2300
2301
2302
2303
2304
2305
2306
2307
2308
2309
2310
2311
2312
2313
2314
2315
2316
2317
2318
2319
2320
2321
2322
2323
2324
2325
2326
2327
2328
2329
2330
2331
2332
2333
2334
2335
2336
2337
2338
2339
2340
2341
2342
2343
2344
2345
2346
2347
2348
2349
2350
2351
2352
2353
2354
2355
2356
2357
2358
2359
2360
2361
2362
2363
2364
2365
2366
2367
2368
2369
2370
2371
2372
2373
2374
2375
2376
2377
2378
2379
2380
2381
2382
2383
2384
2385
2386
2387
2388
2389
2390
2391
2392
2393
2394
2395
2396
2397
2398
2399
2400
2401
2402
2403
2404
2405
2406
2407
2408
2409
2410
2411
2412
2413
2414
2415
2416
2417
2418
2419
2420
2421
2422
2423
2424
2425
2426
2427
2428
2429
2430
2431
2432
2433
2434
2435
2436
2437
2438
2439
2440
2441
2442
2443
2444
2445
2446
2447
2448
2449
2450
2451
2452
2453
2454
2455
2456
2457
2458
2459
2460
2461
2462
2463
2464
2465
2466
2467
2468
2469
2470
2471
2472
2473
2474
2475
2476
2477
2478
2479
2480
2481
2482
2483
2484
2485
2486
2487
2488
2489
2490
2491
2492
2493
2494
2495
2496
2497
2498
2499
2500
2501
2502
2503
2504
2505
2506
2507
2508
2509
2510
2511
2512
2513
2514
2515
2516
2517
2518
2519
2520
2521
2522
2523
2524
2525
2526
2527
2528
2529
2530
2531
2532
2533
2534
2535
2536
2537
2538
2539
2540
2541
2542
2543
2544
2545
2546
2547
2548
2549
2550
2551
2552
2553
2554
2555
2556
2557
2558
2559
2560
2561
2562
2563
2564
2565
2566
2567
2568
2569
2570
2571
2572
2573
2574
2575
2576
2577
2578
2579
2580
2581
2582
2583
2584
2585
2586
2587
2588
2589
2590
2591
2592
2593
2594
2595
2596
2597
2598
2599
2600
2601
2602
2603
2604
2605
2606
2607
2608
2609
2610
2611
2612
2613
2614
2615
2616
2617
2618
2619
2620
2621
2622
2623
2624
2625
2626
2627
2628
2629
2630
2631
2632
2633
2634
2635
2636
2637
2638
2639
2640
2641
2642
2643
2644
2645
2646
2647
2648
2649
2650
2651
2652
2653
2654
2655
2656
2657
2658
2659
2660
2661
2662
2663
2664
2665
2666
2667
2668
2669
2670
2671
2672
2673
2674
2675
2676
2677
2678
2679
2680
2681
2682
2683
2684
2685
2686
2687
2688
2689
2690
2691
2692
2693
2694
2695
2696
2697
2698
2699
2700
2701
2702
2703
2704
2705
2706
2707
2708
2709
2710
2711
2712
2713
2714
2715
2716
2717
2718
2719
2720
2721
2722
2723
2724
2725
2726
2727
2728
2729
2730
2731
2732
2733
2734
2735
2736
2737
2738
2739
2740
2741
2742
2743
2744
2745
2746
2747
2748
2749
2750
2751
2752
2753
2754
2755
2756
2757
2758
2759
2760
2761
2762
2763
2764
2765
2766
2767
2768
2769
2770
2771
2772
2773
2774
2775
2776
2777
2778
2779
2780
2781
2782
2783
2784
2785
2786
2787
2788
2789
2790
2791
2792
2793
2794
2795
2796
2797
2798
2799
2800
2801
2802
2803
2804
2805
2806
2807
2808
2809
2810
2811
2812
2813
2814
2815
2816
2817
2818
2819
2820
2821
2822
2823
2824
2825
2826
2827
2828
2829
2830
2831
2832
2833
2834
2835
2836
2837
2838
2839
2840
2841
2842
2843
2844
2845
2846
2847
2848
2849
2850
2851
2852
2853
2854
2855
2856
2857
2858
2859
2860
2861
2862
2863
2864
2865
2866
2867
2868
2869
2870
2871
2872
2873
2874
2875
2876
2877
2878
2879
2880
2881
2882
2883
2884
2885
2886
2887
2888
2889
2890
2891
2892
2893
2894
2895
2896
2897
2898
2899
2900
2901
2902
2903
2904
2905
2906
2907
2908
2909
2910
2911
2912
2913
2914
2915
2916
2917
2918
2919
2920
2921
2922
2923
2924
2925
2926
2927
2928
2929
2930
2931
2932
2933
2934
2935
2936
2937
2938
2939
2940
2941
2942
2943
2944
2945
2946
2947
2948
2949
2950
2951
2952
2953
2954
2955
2956
2957
2958
2959
2960
2961
2962
2963
2964
2965
2966
2967
2968
2969
2970
2971
2972
2973
2974
2975
2976
2977
2978
2979
2980
2981
2982
2983
2984
2985
2986
2987
2988
2989
2990
2991
2992
2993
2994
2995
2996
2997
2998
2999
3000
3001
3002
3003
3004
3005
3006
3007
3008
3009
3010
3011
3012
3013
3014
3015
3016
3017
3018
3019
3020
3021
3022
3023
3024
3025
3026
3027
3028
3029
3030
3031
3032
3033
3034
3035
3036
3037
3038
3039
3040
3041
3042
3043
3044
3045
3046
3047
3048
3049
3050
3051
3052
3053
3054
3055
3056
3057
3058
3059
3060
3061
3062
3063
3064
3065
3066
3067
3068
3069
3070
3071
3072
3073
3074
3075
3076
3077
3078
3079
3080
3081
3082
3083
3084
3085
3086
3087
3088
3089
3090
3091
3092
3093
3094
3095
3096
3097
3098
3099
3100
3101
3102
3103
3104
3105
3106
3107
3108
3109
3110
3111
3112
3113
3114
3115
3116
3117
3118
3119
3120
3121
3122
3123
3124
3125
3126
3127
3128
3129
3130
3131
3132
3133
3134
3135
3136
3137
3138
3139
3140
3141
3142
3143
3144
3145
3146
3147
3148
3149
3150
3151
3152
3153
3154
3155
3156
3157
3158
3159
3160
3161
3162
3163
3164
3165
3166
3167
3168
3169
3170
3171
3172
3173
3174
3175
3176
3177
3178
3179
3180
3181
3182
3183
3184
3185
3186
3187
3188
3189
3190
3191
3192
3193
3194
3195
3196
3197
3198
3199
3200
3201
3202
3203
3204
3205
3206
3207
3208
3209
3210
3211
3212
3213
3214
3215
3216
3217
3218
3219
3220
3221
3222
3223
3224
3225
3226
3227
3228
3229
3230
3231
3232
3233
3234
3235
3236
3237
3238
3239
3240
3241
3242
3243
3244
3245
3246
3247
3248
3249
3250
3251
3252
3253
3254
3255
3256
3257
3258
3259
3260
3261
3262
3263
3264
3265
3266
3267
3268
3269
3270
3271
3272
3273
3274
3275
3276
3277
3278
3279
3280
3281
3282
3283
3284
3285
3286
3287
3288
3289
3290
3291
3292
3293
3294
3295
3296
3297
3298
3299
3300
3301
3302
3303
3304
3305
3306
3307
3308
3309
3310
3311
3312
3313
3314
3315
3316
3317
3318
3319
3320
3321
3322
3323
3324
3325
3326
3327
3328
3329
3330
3331
3332
3333
3334
3335
3336
3337
3338
3339
3340
3341
3342
3343
3344
3345
3346
3347
3348
3349
3350
3351
3352
3353
3354
3355
3356
3357
3358
3359
3360
3361
3362
3363
3364
3365
3366
3367
3368
3369
3370
3371
3372
3373
3374
3375
3376
3377
3378
3379
3380
3381
3382
3383
3384
3385
3386
3387
3388
3389
3390
3391
3392
3393
3394
3395
3396
3397
3398
3399
3400
3401
3402
3403
3404
3405
3406
3407
3408
3409
3410
3411
3412
3413
3414
3415
3416
3417
3418
3419
3420
3421
3422
3423
3424
3425
3426
3427
3428
3429
3430
3431
3432
3433
3434
3435
3436
3437
3438
3439
3440
3441
3442
3443
3444
3445
3446
3447
3448
3449
3450
3451
3452
3453
3454
3455
3456
3457
3458
3459
3460
3461
3462
3463
3464
3465
3466
3467
3468
3469
3470
3471
3472
3473
3474
3475
3476
3477
3478
3479
3480
3481
3482
3483
3484
3485
3486
3487
3488
3489
3490
3491
3492
3493
3494
3495
3496
3497
3498
3499
3500
3501
3502
3503
3504
3505
3506
3507
3508
3509
3510
3511
3512
3513
3514
3515
3516
3517
3518
3519
3520
3521
3522
3523
3524
3525
3526
3527
3528
3529
3530
3531
3532
3533
3534
3535
3536
3537
3538
3539
3540
3541
3542
3543
3544
3545
3546
3547
3548
3549
3550
3551
3552
3553
3554
3555
3556
3557
3558
3559
3560
3561
3562
3563
3564
3565
3566
3567
3568
3569
3570
3571
3572
3573
3574
3575
3576
3577
3578
3579
3580
3581
3582
3583
3584
3585
3586
3587
3588
3589
3590
3591
3592
3593
3594
3595
3596
3597
3598
3599
3600
3601
3602
3603
3604
3605
3606
3607
3608
3609
3610
3611
3612
3613
3614
3615
3616
3617
3618
3619
3620
3621
3622
3623
3624
3625
3626
3627
3628
3629
3630
3631
3632
3633
3634
3635
3636
3637
3638
3639
3640
3641
3642
3643
3644
3645
3646
3647
3648
3649
3650
3651
3652
3653
3654
3655
3656
3657
3658
3659
3660
3661
3662
3663
3664
3665
3666
3667
3668
3669
3670
3671
3672
3673
3674
3675
3676
3677
3678
3679
3680
3681
3682
3683
3684
3685
3686
3687
3688
3689
3690
3691
3692
3693
3694
3695
3696
3697
3698
3699
3700
3701
3702
3703
3704
3705
3706
3707
3708
3709
3710
3711
3712
3713
3714
3715
3716
3717
3718
3719
3720
3721
3722
3723
3724
3725
3726
3727
3728
3729
3730
3731
3732
3733
3734
3735
3736
3737
3738
3739
3740
3741
3742
3743
3744
3745
3746
3747
3748
3749
3750
3751
3752
3753
3754
3755
3756
3757
3758
3759
3760
3761
3762
3763
3764
3765
3766
3767
3768
3769
3770
3771
3772
3773
3774
3775
3776
3777
3778
3779
3780
3781
3782
3783
3784
3785
3786
3787
3788
3789
3790
3791
3792
3793
3794
3795
3796
3797
3798
3799
3800
3801
3802
3803
3804
3805
3806
3807
3808
3809
3810
3811
3812
3813
3814
3815
3816
3817
3818
3819
3820
3821
3822
3823
3824
3825
3826
3827
3828
3829
3830
3831
3832
3833
3834
3835
3836
3837
3838
3839
3840
3841
3842
3843
3844
3845
3846
3847
3848
3849
3850
3851
3852
3853
3854
3855
3856
3857
3858
3859
3860
3861
3862
3863
3864
3865
3866
3867
3868
3869
3870
3871
3872
3873
3874
3875
3876
3877
3878
3879
3880
3881
3882
3883
3884
3885
3886
3887
3888
3889
3890
3891
3892
3893
3894
3895
3896
3897
3898
3899
3900
3901
3902
3903
3904
3905
3906
3907
3908
3909
3910
3911
3912
3913
3914
3915
3916
3917
3918
3919
3920
3921
3922
3923
3924
3925
3926
3927
3928
3929
3930
3931
3932
3933
3934
3935
3936
3937
3938
3939
3940
3941
3942
3943
3944
3945
3946
3947
3948
3949
3950
3951
3952
3953
3954
3955
3956
3957
3958
3959
3960
3961
3962
3963
3964
3965
3966
3967
3968
3969
3970
3971
3972
3973
3974
3975
3976
3977
3978
3979
3980
3981
3982
3983
3984
3985
3986
3987
3988
3989
3990
3991
3992
3993
3994
3995
3996
3997
3998
3999
4000
4001
4002
4003
4004
4005
4006
4007
4008
4009
4010
4011
4012
4013
4014
4015
4016
4017
4018
4019
4020
4021
4022
4023
4024
4025
4026
4027
4028
4029
4030
4031
4032
4033
4034
4035
4036
4037
4038
4039
4040
4041
4042
4043
4044
4045
4046
4047
4048
4049
4050
4051
4052
4053
4054
4055
4056
4057
4058
4059
4060
4061
4062
4063
4064
4065
4066
4067
4068
4069
4070
4071
4072
4073
4074
4075
4076
4077
4078
4079
4080
4081
4082
4083
4084
4085
4086
4087
4088
4089
4090
4091
4092
4093
4094
4095
4096
4097
4098
4099
4100
4101
4102
4103
4104
4105
4106
4107
4108
4109
4110
4111
4112
4113
4114
4115
4116
4117
4118
4119
4120
4121
4122
4123
4124
4125
4126
4127
4128
4129
4130
4131
4132
4133
4134
4135
4136
4137
4138
4139
4140
4141
4142
4143
4144
4145
4146
4147
4148
4149
4150
4151
4152
4153
4154
4155
4156
4157
4158
4159
4160
4161
4162
4163
4164
4165
4166
4167
4168
4169
4170
4171
4172
4173
4174
4175
4176
4177
4178
4179
4180
4181
4182
4183
4184
4185
4186
4187
4188
4189
4190
4191
4192
4193
4194
4195
4196
4197
4198
4199
4200
4201
4202
4203
4204
4205
4206
4207
4208
4209
4210
4211
4212
4213
4214
4215
4216
4217
4218
4219
4220
4221
4222
4223
4224
4225
4226
4227
4228
4229
4230
4231
4232
4233
4234
4235
4236
4237
4238
4239
4240
4241
4242
4243
4244
4245
4246
4247
4248
4249
4250
4251
4252
4253
4254
4255
4256
4257
4258
4259
4260
4261
4262
4263
4264
4265
4266
4267
4268
4269
4270
4271
4272
4273
4274
4275
4276
4277
4278
4279
4280
4281
4282
4283
4284
4285
4286
4287
4288
4289
4290
4291
4292
4293
4294
4295
4296
4297
4298
4299
4300
4301
4302
4303
4304
4305
4306
4307
4308
4309
4310
4311
4312
4313
4314
4315
4316
4317
4318
4319
4320
4321
4322
4323
4324
4325
4326
4327
4328
4329
4330
4331
4332
4333
4334
4335
4336
4337
4338
4339
4340
4341
4342
4343
4344
4345
4346
4347
4348
4349
4350
4351
4352
4353
4354
4355
4356
4357
4358
4359
4360
4361
4362
4363
4364
4365
4366
4367
4368
4369
4370
4371
4372
4373
4374
4375
4376
4377
4378
4379
4380
4381
4382
4383
4384
4385
4386
4387
4388
4389
4390
4391
4392
4393
4394
4395
4396
4397
4398
4399
4400
4401
4402
4403
4404
4405
4406
4407
4408
4409
4410
4411
4412
4413
4414
4415
4416
4417
4418
4419
4420
4421
4422
4423
4424
4425
4426
4427
4428
4429
4430
4431
4432
4433
4434
4435
4436
4437
4438
4439
4440
4441
4442
4443
4444
4445
4446
4447
4448
4449
4450
4451
4452
4453
4454
4455
4456
4457
4458
4459
4460
4461
4462
4463
4464
4465
4466
4467
4468
4469
4470
4471
4472
4473
4474
4475
4476
4477
4478
4479
4480
4481
4482
4483
4484
4485
4486
4487
4488
4489
4490
4491
4492
4493
4494
4495
4496
4497
4498
4499
4500
4501
4502
4503
4504
4505
4506
4507
4508
4509
4510
4511
4512
4513
4514
4515
4516
4517
4518
4519
4520
4521
4522
4523
4524
4525
4526
4527
4528
4529
4530
4531
4532
4533
4534
4535
4536
4537
4538
4539
4540
4541
4542
4543
4544
4545
4546
4547
4548
4549
4550
4551
4552
4553
4554
4555
4556
4557
4558
4559
4560
4561
4562
4563
4564
4565
4566
4567
4568
4569
4570
4571
4572
4573
4574
4575
4576
4577
4578
4579
4580
4581
4582
4583
4584
4585
4586
4587
4588
4589
4590
4591
4592
4593
4594
4595
4596
4597
4598
4599
4600
4601
4602
4603
4604
4605
4606
4607
4608
4609
4610
4611
4612
4613
4614
4615
4616
4617
4618
4619
4620
4621
4622
4623
4624
4625
4626
4627
4628
4629
4630
4631
4632
4633
4634
4635
4636
4637
4638
4639
4640
4641
4642
4643
4644
4645
4646
4647
4648
4649
4650
4651
4652
4653
4654
4655
4656
4657
4658
4659
4660
4661
4662
4663
4664
4665
4666
4667
4668
4669
4670
4671
4672
4673
4674
4675
4676
4677
4678
4679
4680
4681
4682
4683
4684
4685
4686
4687
4688
4689
4690
4691
4692
4693
4694
4695
4696
4697
4698
4699
4700
4701
4702
4703
4704
4705
4706
4707
4708
4709
4710
4711
4712
4713
4714
4715
4716
4717
4718
4719
4720
4721
4722
4723
4724
4725
4726
4727
4728
4729
4730
4731
4732
4733
4734
4735
4736
4737
4738
4739
4740
4741
4742
4743
4744
4745
4746
4747
4748
4749
4750
4751
4752
4753
4754
4755
4756
4757
4758
4759
4760
4761
4762
4763
4764
4765
4766
4767
4768
4769
4770
4771
4772
4773
4774
4775
4776
4777
4778
4779
4780
4781
4782
4783
4784
4785
4786
4787
4788
4789
4790
4791
4792
4793
4794
4795
4796
4797
4798
4799
4800
4801
4802
4803
4804
4805
4806
4807
4808
4809
4810
4811
4812
4813
4814
4815
4816
4817
4818
4819
4820
4821
4822
4823
4824
4825
4826
4827
4828
4829
4830
4831
4832
4833
4834
4835
4836
4837
4838
4839
4840
4841
4842
4843
4844
4845
4846
4847
4848
4849
4850
4851
4852
4853
4854
4855
4856
4857
4858
4859
4860
4861
4862
4863
4864
4865
4866
4867
4868
4869
4870
4871
4872
4873
4874
4875
4876
4877
4878
4879
4880
4881
4882
4883
4884
4885
4886
4887
4888
4889
4890
4891
4892
4893
4894
4895
4896
4897
4898
4899
4900
4901
4902
4903
4904
4905
4906
4907
4908
4909
4910
4911
4912
4913
4914
4915
4916
4917
4918
4919
4920
4921
4922
4923
4924
4925
4926
4927
4928
4929
4930
4931
4932
4933
4934
4935
4936
4937
4938
4939
4940
4941
4942
4943
4944
4945
4946
4947
4948
4949
4950
4951
4952
4953
4954
4955
4956
4957
4958
4959
4960
4961
4962
4963
4964
4965
4966
4967
4968
4969
4970
4971
4972
4973
4974
4975
4976
4977
4978
4979
4980
4981
4982
4983
4984
4985
4986
4987
4988
4989
4990
4991
4992
4993
4994
4995
4996
4997
4998
4999
5000
5001
5002
5003
5004
5005
5006
5007
5008
5009
5010
5011
5012
5013
5014
5015
5016
5017
5018
5019
5020
5021
5022
5023
5024
5025
5026
5027
5028
5029
5030
5031
5032
5033
5034
5035
5036
5037
5038
5039
5040
5041
5042
5043
5044
5045
5046
5047
5048
5049
5050
5051
5052
5053
5054
5055
5056
5057
5058
5059
5060
5061
5062
5063
5064
5065
5066
5067
5068
5069
5070
5071
5072
5073
5074
5075
5076
5077
5078
5079
5080
5081
5082
5083
5084
5085
5086
5087
5088
5089
5090
5091
5092
5093
5094
5095
5096
5097
5098
5099
5100
5101
5102
5103
5104
5105
5106
5107
5108
5109
5110
5111
5112
5113
5114
5115
5116
5117
5118
5119
5120
5121
5122
5123
5124
5125
5126
5127
5128
5129
5130
5131
5132
5133
5134
5135
5136
5137
5138
5139
5140
5141
5142
5143
5144
5145
5146
5147
5148
5149
5150
5151
5152
5153
5154
5155
5156
5157
5158
5159
5160
5161
5162
5163
5164
5165
5166
5167
5168
5169
5170
5171
5172
5173
5174
5175
5176
5177
5178
5179
5180
5181
5182
5183
5184
5185
5186
5187
5188
5189
5190
5191
5192
5193
5194
5195
5196
5197
5198
5199
5200
5201
5202
5203
5204
5205
5206
5207
5208
5209
5210
5211
5212
5213
5214
5215
5216
5217
5218
5219
5220
5221
5222
5223
5224
5225
5226
5227
5228
5229
5230
5231
5232
5233
5234
5235
5236
5237
5238
5239
5240
5241
5242
5243
5244
5245
5246
5247
5248
5249
5250
5251
5252
5253
5254
5255
5256
5257
5258
5259
5260
5261
5262
5263
5264
5265
5266
5267
5268
5269
5270
5271
5272
5273
5274
5275
5276
5277
5278
5279
5280
5281
5282
5283
5284
5285
5286
5287
5288
5289
5290
5291
5292
5293
5294
5295
5296
5297
5298
5299
5300
5301
5302
5303
5304
5305
5306
5307
5308
5309
5310
5311
5312
5313
5314
5315
5316
5317
5318
5319
5320
5321
5322
5323
5324
5325
5326
5327
5328
5329
5330
5331
5332
5333
5334
5335
5336
5337
5338
5339
5340
5341
5342
5343
5344
5345
5346
5347
5348
5349
5350
5351
5352
5353
5354
5355
5356
5357
5358
5359
5360
5361
5362
5363
5364
5365
5366
5367
5368
5369
5370
5371
5372
5373
5374
5375
5376
5377
5378
5379
5380
5381
5382
5383
5384
5385
5386
5387
5388
5389
5390
5391
5392
5393
5394
5395
5396
5397
5398
5399
5400
5401
5402
5403
5404
5405
5406
5407
5408
5409
5410
5411
5412
5413
5414
5415
5416
5417
5418
5419
5420
5421
5422
5423
5424
5425
5426
5427
5428
5429
5430
5431
5432
5433
5434
5435
5436
5437
5438
5439
5440
5441
5442
5443
5444
5445
5446
5447
5448
5449
5450
5451
5452
5453
5454
5455
5456
5457
5458
5459
5460
5461
5462
5463
5464
5465
5466
5467
5468
5469
5470
5471
5472
5473
5474
5475
5476
5477
5478
5479
5480
5481
5482
5483
5484
5485
5486
5487
5488
5489
5490
5491
5492
5493
5494
5495
5496
5497
5498
5499
5500
5501
5502
5503
5504
5505
5506
5507
5508
5509
5510
5511
5512
5513
5514
5515
5516
5517
5518
5519
5520
5521
5522
5523
5524
5525
5526
5527
5528
5529
5530
5531
5532
5533
5534
5535
5536
5537
5538
5539
5540
5541
5542
5543
5544
5545
5546
5547
5548
5549
5550
5551
5552
5553
5554
5555
5556
5557
5558
5559
5560
5561
5562
5563
5564
5565
5566
5567
5568
5569
5570
5571
5572
5573
5574
5575
5576
5577
5578
5579
5580
5581
5582
5583
5584
5585
5586
5587
5588
5589
5590
5591
5592
5593
5594
5595
5596
5597
5598
5599
5600
5601
5602
5603
5604
5605
5606
5607
5608
5609
5610
5611
5612
5613
5614
5615
5616
5617
5618
5619
5620
5621
5622
5623
5624
5625
5626
5627
5628
5629
5630
5631
5632
5633
5634
5635
5636
5637
5638
5639
5640
5641
5642
5643
5644
5645
5646
5647
5648
5649
5650
5651
5652
5653
5654
5655
5656
5657
5658
5659
5660
5661
5662
5663
5664
5665
5666
5667
5668
5669
5670
5671
5672
5673
5674
5675
5676
5677
5678
5679
5680
5681
5682
5683
5684
5685
5686
5687
5688
5689
5690
5691
5692
5693
5694
5695
5696
5697
5698
5699
5700
5701
5702
5703
5704
5705
5706
5707
5708
5709
5710
5711
5712
5713
5714
5715
5716
5717
5718
5719
5720
5721
5722
5723
5724
5725
5726
5727
5728
5729
5730
5731
5732
5733
5734
5735
5736
5737
5738
5739
5740
5741
5742
5743
5744
5745
5746
5747
5748
5749
5750
5751
5752
5753
5754
5755
5756
5757
5758
5759
5760
5761
5762
5763
5764
5765
5766
5767
5768
5769
5770
5771
5772
5773
5774
5775
5776
5777
5778
5779
5780
5781
5782
5783
5784
5785
5786
5787
5788
5789
5790
5791
5792
5793
5794
5795
5796
5797
5798
5799
5800
5801
5802
5803
5804
5805
5806
5807
5808
5809
5810
5811
5812
5813
5814
5815
5816
5817
5818
5819
5820
5821
5822
5823
5824
5825
5826
5827
5828
5829
5830
5831
5832
5833
5834
5835
5836
5837
5838
5839
5840
5841
5842
5843
5844
5845
5846
5847
5848
5849
5850
5851
5852
5853
5854
5855
5856
5857
5858
5859
5860
5861
5862
5863
5864
5865
5866
5867
5868
5869
5870
5871
5872
5873
5874
5875
5876
5877
5878
5879
5880
5881
5882
5883
5884
5885
5886
5887
5888
5889
5890
5891
5892
5893
5894
5895
5896
5897
5898
5899
5900
5901
5902
5903
5904
5905
5906
5907
5908
5909
5910
5911
5912
5913
5914
5915
5916
5917
5918
5919
5920
5921
5922
5923
5924
5925
5926
5927
5928
5929
5930
5931
5932
5933
5934
5935
5936
5937
5938
5939
5940
5941
5942
5943
5944
5945
5946
5947
5948
5949
5950
5951
5952
5953
5954
5955
5956
5957
5958
5959
5960
5961
5962
5963
5964
5965
5966
5967
5968
5969
5970
5971
5972
5973
5974
5975
5976
5977
5978
5979
5980
5981
5982
5983
5984
5985
5986
5987
5988
5989
5990
5991
5992
5993
5994
5995
5996
5997
5998
5999
6000
6001
6002
6003
6004
6005
6006
6007
6008
6009
6010
6011
6012
6013
6014
6015
6016
6017
6018
6019
6020
6021
6022
6023
6024
6025
6026
6027
6028
6029
6030
6031
6032
6033
6034
6035
6036
6037
6038
6039
6040
6041
6042
6043
6044
6045
6046
6047
6048
6049
6050
6051
6052
6053
6054
6055
6056
6057
6058
6059
6060
6061
6062
6063
6064
6065
6066
6067
6068
6069
6070
6071
6072
6073
6074
6075
6076
6077
6078
6079
6080
6081
6082
6083
6084
6085
6086
6087
6088
6089
6090
6091
6092
6093
6094
6095
6096
6097
6098
6099
6100
6101
6102
6103
6104
6105
6106
6107
6108
6109
6110
6111
6112
6113
6114
6115
6116
6117
6118
6119
6120
6121
6122
6123
6124
6125
6126
6127
6128
6129
6130
6131
6132
6133
6134
6135
6136
6137
6138
6139
6140
6141
6142
6143
6144
6145
6146
6147
6148
6149
6150
6151
6152
6153
6154
6155
6156
6157
6158
6159
6160
6161
6162
6163
6164
6165
6166
6167
6168
6169
6170
6171
6172
6173
6174
6175
6176
6177
6178
6179
6180
6181
6182
6183
6184
6185
6186
6187
6188
6189
6190
6191
6192
6193
6194
6195
6196
6197
6198
6199
6200
6201
6202
6203
6204
6205
6206
6207
6208
6209
6210
6211
6212
6213
6214
6215
6216
6217
6218
6219
6220
6221
6222
6223
6224
6225
6226
6227
6228
6229
6230
6231
6232
6233
6234
6235
6236
6237
6238
6239
6240
6241
6242
6243
6244
6245
6246
6247
6248
6249
6250
6251
6252
6253
6254
6255
6256
6257
6258
6259
6260
6261
6262
6263
6264
6265
6266
6267
6268
6269
6270
6271
6272
6273
6274
6275
6276
6277
6278
6279
6280
6281
6282
6283
6284
6285
6286
6287
6288
6289
6290
6291
6292
6293
6294
6295
6296
6297
6298
6299
6300
6301
6302
6303
6304
6305
6306
6307
6308
6309
6310
6311
6312
6313
6314
6315
6316
6317
6318
6319
6320
6321
6322
6323
6324
6325
6326
6327
6328
6329
6330
6331
6332
6333
6334
6335
6336
6337
6338
6339
6340
6341
6342
6343
6344
6345
6346
6347
6348
6349
6350
6351
6352
6353
6354
6355
6356
6357
6358
6359
6360
6361
6362
6363
6364
6365
6366
6367
6368
6369
6370
6371
6372
6373
6374
6375
6376
6377
6378
6379
6380
6381
6382
6383
6384
6385
6386
6387
6388
6389
6390
6391
6392
6393
6394
6395
6396
6397
6398
6399
6400
6401
6402
6403
6404
6405
6406
6407
6408
6409
6410
6411
6412
6413
6414
6415
6416
6417
6418
6419
6420
6421
6422
6423
6424
6425
6426
6427
6428
6429
6430
6431
6432
6433
6434
6435
6436
6437
6438
6439
6440
6441
6442
6443
6444
6445
6446
6447
6448
6449
6450
6451
6452
6453
6454
6455
6456
6457
6458
6459
6460
6461
6462
6463
6464
6465
6466
6467
6468
6469
6470
6471
6472
6473
6474
6475
6476
6477
6478
6479
6480
6481
6482
6483
6484
6485
6486
6487
6488
6489
6490
6491
6492
6493
6494
6495
6496
6497
6498
6499
6500
6501
6502
6503
6504
6505
6506
6507
6508
6509
6510
6511
6512
6513
6514
6515
6516
6517
6518
6519
6520
6521
6522
6523
6524
6525
6526
6527
6528
6529
6530
6531
6532
6533
6534
6535
6536
6537
6538
6539
6540
6541
6542
6543
6544
6545
6546
6547
6548
6549
6550
6551
6552
6553
6554
6555
6556
6557
6558
6559
6560
6561
6562
6563
6564
6565
6566
6567
6568
6569
6570
6571
6572
6573
6574
6575
6576
6577
6578
6579
6580
6581
6582
6583
6584
6585
6586
6587
6588
6589
6590
6591
6592
6593
6594
6595
6596
6597
6598
6599
6600
6601
6602
6603
6604
6605
6606
6607
6608
6609
6610
6611
6612
6613
6614
6615
6616
6617
6618
6619
6620
6621
6622
6623
6624
6625
6626
6627
6628
6629
6630
6631
6632
6633
6634
6635
6636
6637
6638
6639
6640
6641
6642
6643
6644
6645
6646
6647
6648
6649
6650
6651
6652
6653
6654
6655
6656
6657
6658
6659
6660
6661
6662
6663
6664
6665
6666
6667
6668
6669
6670
6671
6672
6673
6674
6675
6676
6677
6678
6679
6680
6681
6682
6683
6684
6685
6686
6687
6688
6689
6690
6691
6692
6693
6694
6695
6696
6697
6698
6699
6700
6701
6702
6703
6704
6705
6706
6707
6708
6709
6710
6711
6712
6713
6714
6715
6716
6717
6718
6719
6720
6721
6722
6723
6724
6725
6726
6727
6728
6729
6730
6731
6732
6733
6734
6735
6736
6737
6738
6739
6740
6741
6742
6743
6744
6745
6746
6747
6748
6749
6750
6751
6752
6753
6754
6755
6756
6757
6758
6759
6760
6761
6762
6763
6764
6765
6766
6767
6768
6769
6770
6771
6772
6773
6774
6775
6776
6777
6778
6779
6780
6781
6782
6783
6784
6785
6786
6787
6788
6789
6790
6791
6792
6793
6794
6795
6796
6797
6798
6799
6800
6801
6802
6803
6804
6805
6806
6807
6808
6809
6810
6811
6812
6813
6814
6815
6816
6817
6818
6819
6820
6821
6822
6823
6824
6825
6826
6827
6828
6829
6830
6831
6832
6833
6834
6835
6836
6837
6838
6839
6840
6841
6842
6843
6844
6845
6846
6847
6848
6849
6850
6851
6852
6853
6854
6855
6856
6857
6858
6859
6860
6861
6862
6863
6864
6865
6866
6867
6868
6869
6870
6871
6872
6873
6874
6875
6876
6877
6878
6879
6880
6881
6882
6883
6884
6885
6886
6887
6888
6889
6890
6891
6892
6893
6894
6895
6896
6897
6898
6899
6900
6901
6902
6903
6904
6905
6906
6907
6908
6909
6910
6911
6912
6913
6914
6915
6916
6917
6918
6919
6920
6921
6922
6923
6924
6925
6926
6927
6928
6929
6930
6931
6932
6933
6934
6935
6936
6937
6938
6939
6940
6941
6942
6943
6944
6945
6946
6947
6948
6949
6950
6951
6952
6953
6954
6955
6956
6957
6958
6959
6960
6961
6962
6963
6964
6965
6966
6967
6968
6969
6970
6971
6972
6973
6974
6975
6976
6977
6978
6979
6980
6981
6982
6983
6984
6985
6986
6987
6988
6989
6990
6991
6992
6993
6994
6995
6996
6997
6998
6999
7000
7001
7002
7003
7004
7005
7006
7007
7008
7009
7010
7011
7012
7013
7014
7015
7016
7017
7018
7019
7020
7021
7022
7023
7024
7025
7026
7027
7028
7029
7030
7031
7032
7033
7034
7035
7036
7037
7038
7039
7040
7041
7042
7043
7044
7045
7046
7047
7048
7049
7050
7051
7052
7053
7054
7055
7056
7057
7058
7059
7060
7061
7062
7063
7064
7065
7066
7067
7068
7069
7070
7071
7072
7073
7074
7075
7076
7077
7078
7079
7080
7081
7082
7083
7084
7085
7086
7087
7088
7089
7090
7091
7092
7093
7094
7095
7096
7097
7098
7099
7100
7101
7102
7103
7104
7105
7106
7107
7108
7109
7110
7111
7112
7113
7114
7115
7116
7117
7118
7119
7120
7121
7122
7123
7124
7125
7126
7127
7128
7129
7130
7131
7132
7133
7134
7135
7136
7137
7138
7139
7140
7141
7142
7143
7144
7145
7146
7147
7148
7149
7150
7151
7152
7153
7154
7155
7156
7157
7158
7159
7160
7161
7162
7163
7164
7165
7166
7167
7168
7169
7170
7171
7172
7173
7174
7175
7176
7177
7178
7179
7180
7181
7182
7183
7184
7185
7186
7187
7188
7189
7190
7191
7192
7193
7194
7195
7196
7197
7198
7199
7200
7201
7202
7203
7204
7205
7206
7207
7208
7209
7210
7211
7212
7213
7214
7215
7216
7217
7218
7219
7220
7221
7222
7223
7224
7225
7226
7227
7228
7229
7230
7231
7232
7233
7234
7235
7236
7237
7238
7239
7240
7241
7242
7243
7244
7245
7246
7247
7248
7249
7250
7251
7252
7253
7254
7255
7256
7257
7258
7259
7260
7261
7262
7263
7264
7265
7266
7267
7268
7269
7270
7271
7272
7273
7274
7275
7276
7277
7278
7279
7280
7281
7282
7283
7284
7285
7286
7287
7288
7289
7290
7291
7292
7293
7294
7295
7296
7297
7298
7299
7300
7301
7302
7303
7304
7305
7306
7307
7308
7309
7310
7311
7312
7313
7314
7315
7316
7317
7318
7319
7320
7321
7322
7323
7324
7325
7326
7327
7328
7329
7330
7331
7332
7333
7334
7335
7336
7337
7338
7339
7340
7341
7342
7343
7344
7345
7346
7347
7348
7349
7350
7351
7352
7353
7354
7355
7356
7357
7358
7359
7360
7361
7362
7363
7364
7365
7366
7367
7368
7369
7370
7371
7372
7373
7374
7375
7376
7377
7378
7379
7380
7381
7382
7383
7384
7385
7386
7387
7388
7389
7390
7391
7392
7393
7394
7395
7396
7397
7398
7399
7400
7401
7402
7403
7404
7405
7406
7407
7408
7409
7410
7411
7412
7413
7414
7415
7416
7417
7418
7419
7420
7421
7422
7423
7424
7425
7426
7427
7428
7429
7430
7431
7432
7433
7434
7435
7436
7437
7438
7439
7440
7441
7442
7443
7444
7445
7446
7447
7448
7449
7450
7451
7452
7453
7454
7455
7456
7457
7458
7459
7460
7461
7462
7463
7464
7465
7466
7467
7468
7469
7470
7471
7472
7473
7474
7475
7476
7477
7478
7479
7480
7481
7482
7483
7484
7485
7486
7487
7488
7489
7490
7491
7492
7493
7494
7495
7496
7497
7498
7499
7500
7501
7502
7503
7504
7505
7506
7507
7508
7509
7510
7511
7512
7513
7514
7515
7516
7517
7518
7519
7520
7521
7522
7523
7524
7525
7526
7527
7528
7529
7530
7531
7532
7533
7534
7535
7536
7537
7538
7539
7540
7541
7542
7543
7544
7545
7546
7547
7548
7549
7550
7551
7552
7553
7554
7555
7556
7557
7558
7559
7560
7561
7562
7563
7564
7565
7566
7567
7568
7569
7570
7571
7572
7573
7574
7575
7576
7577
7578
7579
7580
7581
7582
7583
7584
7585
7586
7587
7588
7589
7590
7591
7592
7593
7594
7595
7596
7597
7598
7599
7600
7601
7602
7603
7604
7605
7606
7607
7608
7609
7610
7611
7612
7613
7614
7615
7616
7617
7618
7619
7620
7621
7622
7623
7624
7625
7626
7627
7628
7629
7630
7631
7632
7633
7634
7635
7636
7637
7638
7639
7640
7641
7642
7643
7644
7645
7646
7647
7648
7649
7650
7651
7652
7653
7654
7655
7656
7657
7658
7659
7660
7661
7662
7663
7664
7665
7666
7667
7668
7669
7670
7671
7672
7673
7674
7675
7676
7677
7678
7679
7680
7681
7682
7683
7684
7685
7686
7687
7688
7689
7690
7691
7692
7693
7694
7695
7696
7697
7698
7699
7700
7701
7702
7703
7704
7705
7706
7707
7708
7709
7710
7711
7712
7713
7714
7715
7716
7717
7718
7719
7720
7721
7722
7723
7724
7725
7726
7727
7728
7729
7730
7731
7732
7733
7734
7735
7736
7737
7738
7739
7740
7741
7742
7743
7744
7745
7746
7747
7748
7749
7750
7751
7752
7753
7754
7755
7756
7757
7758
7759
7760
7761
7762
7763
7764
7765
7766
7767
7768
7769
7770
7771
7772
7773
7774
7775
7776
7777
7778
7779
7780
7781
7782
7783
7784
7785
7786
7787
7788
7789
7790
7791
7792
7793
7794
7795
7796
7797
7798
7799
7800
7801
7802
7803
7804
7805
7806
7807
7808
7809
7810
7811
7812
7813
7814
7815
7816
7817
7818
7819
7820
7821
7822
7823
7824
7825
7826
7827
7828
7829
7830
7831
7832
7833
7834
7835
7836
7837
7838
7839
7840
7841
7842
7843
7844
7845
7846
7847
7848
7849
7850
7851
7852
7853
7854
7855
7856
7857
7858
7859
7860
7861
7862
7863
7864
7865
7866
7867
7868
7869
7870
7871
7872
7873
7874
7875
7876
7877
7878
7879
7880
7881
7882
7883
7884
7885
7886
7887
7888
7889
7890
7891
7892
7893
7894
7895
7896
7897
7898
7899
7900
7901
7902
7903
7904
7905
7906
7907
7908
7909
7910
7911
7912
7913
7914
7915
7916
7917
7918
7919
7920
7921
7922
7923
7924
7925
7926
7927
7928
7929
7930
7931
7932
7933
7934
7935
7936
7937
7938
7939
7940
7941
7942
7943
7944
7945
7946
7947
7948
7949
7950
7951
7952
7953
7954
7955
7956
7957
7958
7959
7960
7961
7962
7963
7964
7965
7966
7967
7968
7969
7970
7971
7972
7973
7974
7975
7976
7977
7978
7979
7980
7981
7982
7983
7984
7985
7986
7987
7988
7989
7990
7991
7992
7993
7994
7995
7996
7997
7998
7999
8000
8001
8002
8003
8004
8005
8006
8007
8008
8009
8010
8011
8012
8013
8014
8015
8016
8017
8018
8019
8020
8021
8022
8023
8024
8025
8026
8027
8028
8029
8030
8031
8032
8033
8034
8035
8036
8037
8038
8039
8040
8041
8042
8043
8044
8045
8046
8047
8048
8049
8050
8051
8052
8053
8054
8055
8056
8057
8058
8059
8060
8061
8062
8063
8064
8065
8066
8067
8068
8069
8070
8071
8072
8073
8074
8075
8076
8077
8078
8079
8080
8081
8082
8083
8084
8085
8086
8087
8088
8089
8090
8091
8092
8093
8094
8095
8096
8097
8098
8099
8100
8101
8102
8103
8104
8105
8106
8107
8108
8109
8110
8111
8112
8113
8114
8115
8116
8117
8118
8119
8120
8121
8122
8123
8124
8125
8126
8127
8128
8129
8130
8131
8132
8133
8134
8135
8136
8137
8138
8139
8140
8141
8142
8143
8144
8145
8146
8147
8148
8149
8150
8151
8152
8153
8154
8155
8156
8157
8158
8159
8160
8161
8162
8163
8164
8165
8166
8167
8168
8169
8170
8171
8172
8173
8174
8175
8176
8177
8178
8179
8180
8181
8182
8183
8184
8185
8186
8187
8188
8189
8190
8191
8192
8193
8194
8195
8196
8197
8198
8199
8200
8201
8202
8203
8204
8205
8206
8207
8208
8209
8210
8211
8212
8213
8214
8215
8216
8217
8218
8219
8220
8221
8222
8223
8224
8225
8226
8227
8228
8229
8230
8231
8232
8233
8234
8235
8236
8237
8238
8239
8240
8241
8242
8243
8244
8245
8246
8247
8248
8249
8250
8251
8252
8253
8254
8255
8256
8257
8258
8259
8260
8261
8262
8263
8264
8265
8266
8267
8268
8269
8270
8271
8272
8273
8274
8275
8276
8277
8278
8279
8280
8281
8282
8283
8284
8285
8286
8287
8288
8289
8290
8291
8292
8293
8294
8295
8296
8297
8298
8299
8300
8301
8302
8303
8304
8305
8306
8307
8308
8309
8310
8311
8312
8313
8314
8315
8316
8317
8318
8319
8320
8321
8322
8323
8324
8325
8326
8327
8328
8329
8330
8331
8332
8333
8334
8335
8336
8337
8338
8339
8340
8341
8342
8343
8344
8345
8346
8347
8348
8349
8350
8351
8352
8353
8354
8355
8356
8357
8358
8359
8360
8361
8362
8363
8364
8365
8366
8367
8368
8369
8370
8371
8372
8373
8374
8375
8376
8377
8378
8379
8380
8381
8382
8383
8384
8385
8386
8387
8388
8389
8390
8391
8392
8393
8394
8395
8396
8397
8398
8399
8400
8401
8402
8403
8404
8405
8406
8407
8408
8409
8410
8411
8412
8413
8414
8415
8416
8417
8418
8419
8420
8421
8422
8423
8424
8425
8426
8427
8428
8429
8430
8431
8432
8433
8434
8435
8436
8437
8438
8439
8440
8441
8442
8443
8444
8445
8446
8447
8448
8449
8450
8451
8452
8453
8454
8455
#![allow(clippy::unwrap_used, clippy::expect_used)]
// Re-import the public entry points from this module.
use super::{parse_email, parse_headers_only};

// Re-import wire-layer functions used by tests.
use super::wire::{parse_headers, split_header_body, split_mime_parts};

// Re-import interpret-layer functions used by tests.
#[allow(unused_imports)]
use super::interpret::{
    contains_at_outside_quotes, decode_body, decode_encoded_words, decode_hex_pair,
    decode_q_encoding, decode_quoted_printable, decode_transfer_encoding, extract_comment_text,
    extract_filename, extract_mime_type, extract_param, extract_rfc2231_continuation,
    extract_rfc2231_param, find_closing_quote, find_param_value, hex_digit, is_disposition_type,
    is_inside_quotes, normalize_display_name_phrase, parse_address_list, parse_rfc5322_date,
    parse_single_address, parse_timezone, parse_year, percent_decode, strip_comments,
    strip_outer_quotes, unescape_quoted_string,
};

// Types needed by tests.
use crate::error::Error;
#[allow(unused_imports)]
use crate::types::Address;

#[test]
fn parse_simple_text_email() {
    let raw = b"From: sender@example.com\r\n\
                 To: recipient@example.com\r\n\
                 Subject: Test\r\n\
                 Date: Thu, 13 Feb 2025 15:47:33 +0000\r\n\
                 Message-ID: <abc123@example.com>\r\n\
                 Content-Type: text/plain; charset=utf-8\r\n\
                 \r\n\
                 Hello, World!";

    let parsed = parse_email(raw).unwrap();
    assert_eq!(parsed.from[0].email, "sender@example.com");
    assert_eq!(parsed.to.len(), 1);
    assert_eq!(parsed.to[0].email, "recipient@example.com");
    assert_eq!(parsed.subject.as_deref(), Some("Test"));
    assert_eq!(parsed.message_id.as_deref(), Some("abc123@example.com"));
    assert_eq!(parsed.body_text.as_deref(), Some("Hello, World!"));
    assert!(parsed.body_html.is_none());
    assert!(parsed.attachments.is_empty());
    assert_eq!(parsed.size, raw.len() as u64);
}

#[test]
fn parse_multipart_alternative() {
    let raw = b"From: sender@example.com\r\n\
                 To: recipient@example.com\r\n\
                 Subject: Multi\r\n\
                 Date: Thu, 13 Feb 2025 15:47:33 +0000\r\n\
                 MIME-Version: 1.0\r\n\
                 Content-Type: multipart/alternative; boundary=\"bound42\"\r\n\
                 \r\n\
                 --bound42\r\n\
                 Content-Type: text/plain; charset=utf-8\r\n\
                 \r\n\
                 Plain text body\r\n\
                 --bound42\r\n\
                 Content-Type: text/html; charset=utf-8\r\n\
                 \r\n\
                 <html><body>HTML body</body></html>\r\n\
                 --bound42--";

    let parsed = parse_email(raw).unwrap();
    assert_eq!(parsed.body_text.as_deref(), Some("Plain text body"));
    assert_eq!(
        parsed.body_html.as_deref(),
        Some("<html><body>HTML body</body></html>")
    );
    assert!(parsed.attachments.is_empty());
}

#[test]
fn parse_encoded_words_base64_subject() {
    let raw = b"From: sender@example.com\r\n\
                 Subject: =?UTF-8?B?SGVsbG8gV29ybGQ=?=\r\n\
                 Date: Thu, 13 Feb 2025 15:47:33 +0000\r\n\
                 \r\n\
                 body";

    let parsed = parse_email(raw).unwrap();
    assert_eq!(parsed.subject.as_deref(), Some("Hello World"));
}

#[test]
fn parse_encoded_words_q_subject() {
    let raw = b"From: sender@example.com\r\n\
                 Subject: =?UTF-8?Q?Hello_World?=\r\n\
                 Date: Thu, 13 Feb 2025 15:47:33 +0000\r\n\
                 \r\n\
                 body";

    let parsed = parse_email(raw).unwrap();
    assert_eq!(parsed.subject.as_deref(), Some("Hello World"));
}

#[test]
fn parse_encoded_words_in_display_name() {
    let raw = b"From: =?UTF-8?B?Sm9obiBEb2U=?= <john@example.com>\r\n\
                 Subject: Test\r\n\
                 Date: Thu, 13 Feb 2025 15:47:33 +0000\r\n\
                 \r\n";

    let parsed = parse_email(raw).unwrap();
    assert_eq!(parsed.from[0].name.as_deref(), Some("John Doe"));
    assert_eq!(parsed.from[0].email, "john@example.com");
}

#[test]
fn parse_non_utf8_charset() {
    // ISO-8859-1 encoded subject: "Héllo"
    let raw = b"From: sender@example.com\r\n\
                 Subject: =?ISO-8859-1?Q?H=E9llo?=\r\n\
                 Date: Thu, 13 Feb 2025 15:47:33 +0000\r\n\
                 \r\n";

    let parsed = parse_email(raw).unwrap();
    assert_eq!(parsed.subject.as_deref(), Some("Héllo"));
}

#[test]
fn parse_message_id_strips_brackets() {
    let raw = b"From: a@b.com\r\n\
                 Message-ID: <unique-id@host.com>\r\n\
                 Date: Thu, 13 Feb 2025 15:47:33 +0000\r\n\
                 \r\n";

    let parsed = parse_email(raw).unwrap();
    assert_eq!(parsed.message_id.as_deref(), Some("unique-id@host.com"));
}

#[test]
fn parse_in_reply_to_multiple() {
    let raw = b"From: a@b.com\r\n\
                 In-Reply-To: <first@host> <second@host>\r\n\
                 Date: Thu, 13 Feb 2025 15:47:33 +0000\r\n\
                 \r\n";

    let parsed = parse_email(raw).unwrap();
    // RFC 5322 Section 3.6.4: in-reply-to = "In-Reply-To:" 1*msg-id CRLF
    assert_eq!(parsed.in_reply_to.len(), 2);
    assert_eq!(parsed.in_reply_to[0], "first@host");
    assert_eq!(parsed.in_reply_to[1], "second@host");
}

#[test]
fn parse_references_all_ids() {
    let raw = b"From: a@b.com\r\n\
                 References: <ref1@host> <ref2@host> <ref3@host>\r\n\
                 Date: Thu, 13 Feb 2025 15:47:33 +0000\r\n\
                 \r\n";

    let parsed = parse_email(raw).unwrap();
    // RFC 5322 Section 3.6.4: references = "References:" 1*msg-id CRLF
    // Each message-id should be a separate Vec element.
    assert_eq!(parsed.references.len(), 3);
    assert_eq!(parsed.references[0], "ref1@host");
    assert_eq!(parsed.references[1], "ref2@host");
    assert_eq!(parsed.references[2], "ref3@host");
}

/// RFC 5322 Section 3.6.4: in-reply-to = "In-Reply-To:" 1*msg-id CRLF
/// Multiple message-IDs are valid and must all be captured.
#[test]
fn parse_in_reply_to_all_ids() {
    let raw = b"From: a@b.com\r\n\
                 In-Reply-To: <id1@a.com> <id2@b.com>\r\n\
                 Date: Thu, 13 Feb 2025 15:47:33 +0000\r\n\
                 \r\n";

    let parsed = parse_email(raw).unwrap();
    assert_eq!(parsed.in_reply_to.len(), 2);
    assert_eq!(parsed.in_reply_to[0], "id1@a.com");
    assert_eq!(parsed.in_reply_to[1], "id2@b.com");
}

/// RFC 2046 Section 5.1.1 defines `body-part = MIME-part-headers [CRLF *OCTET]`,
/// so an attachment body may legitimately be zero bytes long. The parser
/// must still surface top-level attachment metadata instead of dropping
/// the message body entirely when no body octets follow the header block.
#[test]
fn parse_empty_top_level_attachment_is_preserved() {
    let raw = b"From: sender@example.com\r\n\
                 Date: Thu, 13 Feb 2025 15:47:33 +0000\r\n\
                 MIME-Version: 1.0\r\n\
                 Content-Type: application/octet-stream\r\n\
                 Content-Disposition: attachment; filename=\"empty.bin\"\r\n\
                 \r\n";

    let parsed = parse_email(raw).unwrap();
    assert_eq!(
        parsed.attachments.len(),
        1,
        "empty top-level attachment must not be dropped"
    );
    assert_eq!(parsed.attachments[0].filename.as_deref(), Some("empty.bin"));
    assert_eq!(parsed.attachments[0].size, Some(0));
    assert_eq!(parsed.attachments[0].section.as_deref(), Some("1"));
}

#[test]
fn parse_date_with_numeric_timezone() {
    let raw = b"From: a@b.com\r\n\
                 Date: Thu, 13 Feb 2025 15:47:33 +0530\r\n\
                 \r\n";

    let parsed = parse_email(raw).unwrap();
    let date = parsed.date.unwrap();
    assert_eq!(date.year, 2025);
    assert_eq!(date.month, 2);
    assert_eq!(date.day, 13);
    assert_eq!(date.hour, 15);
    assert_eq!(date.minute, 47);
    assert_eq!(date.second, 33);
    assert_eq!(date.tz_offset_minutes, 330);
}

#[test]
fn parse_date_named_timezone() {
    let raw = b"From: a@b.com\r\n\
                 Date: Thu, 13 Feb 2025 10:30:00 EST\r\n\
                 \r\n";

    let parsed = parse_email(raw).unwrap();
    let date = parsed.date.unwrap();
    assert_eq!(date.tz_offset_minutes, -300);
}

#[test]
fn parse_address_with_display_name() {
    let raw = b"From: \"John Doe\" <john@example.com>\r\n\
                 Date: Thu, 13 Feb 2025 15:47:33 +0000\r\n\
                 \r\n";

    let parsed = parse_email(raw).unwrap();
    assert_eq!(parsed.from[0].name.as_deref(), Some("John Doe"));
    assert_eq!(parsed.from[0].email, "john@example.com");
}

#[test]
fn parse_multiple_recipients() {
    let raw = b"From: a@b.com\r\n\
                 To: one@x.com, \"Two\" <two@x.com>, three@x.com\r\n\
                 Cc: cc1@x.com, cc2@x.com\r\n\
                 Date: Thu, 13 Feb 2025 15:47:33 +0000\r\n\
                 \r\n";

    let parsed = parse_email(raw).unwrap();
    assert_eq!(parsed.to.len(), 3);
    assert_eq!(parsed.to[1].name.as_deref(), Some("Two"));
    assert_eq!(parsed.cc.len(), 2);
}

#[test]
fn parse_multipart_with_attachment() {
    let raw = b"From: a@b.com\r\n\
                 Date: Thu, 13 Feb 2025 15:47:33 +0000\r\n\
                 MIME-Version: 1.0\r\n\
                 Content-Type: multipart/mixed; boundary=\"mixbound\"\r\n\
                 \r\n\
                 --mixbound\r\n\
                 Content-Type: text/plain\r\n\
                 \r\n\
                 Message body\r\n\
                 --mixbound\r\n\
                 Content-Type: application/pdf\r\n\
                 Content-Disposition: attachment; filename=\"doc.pdf\"\r\n\
                 \r\n\
                 PDF_CONTENT_HERE\r\n\
                 --mixbound--";

    let parsed = parse_email(raw).unwrap();
    assert_eq!(parsed.body_text.as_deref(), Some("Message body"));
    assert_eq!(parsed.attachments.len(), 1);
    assert_eq!(parsed.attachments[0].filename.as_deref(), Some("doc.pdf"));
    assert_eq!(parsed.attachments[0].content_type, "application/pdf");
    assert!(!parsed.attachments[0].is_inline);
    assert_eq!(parsed.attachments[0].section.as_deref(), Some("2"));
}

#[test]
fn parse_inline_attachment() {
    let raw = b"From: a@b.com\r\n\
                 Date: Thu, 13 Feb 2025 15:47:33 +0000\r\n\
                 Content-Type: multipart/mixed; boundary=\"bound\"\r\n\
                 \r\n\
                 --bound\r\n\
                 Content-Type: text/plain\r\n\
                 \r\n\
                 Body\r\n\
                 --bound\r\n\
                 Content-Type: image/png\r\n\
                 Content-Disposition: inline\r\n\
                 Content-ID: <img001>\r\n\
                 \r\n\
                 PNG_DATA\r\n\
                 --bound--";

    let parsed = parse_email(raw).unwrap();
    assert_eq!(parsed.attachments.len(), 1);
    assert!(parsed.attachments[0].is_inline);
    assert_eq!(parsed.attachments[0].content_id.as_deref(), Some("img001"));
}

#[test]
fn parse_headers_only_no_body() {
    let raw = b"From: a@b.com\r\n\
                 Subject: Headers only\r\n\
                 Date: Thu, 13 Feb 2025 15:47:33 +0000\r\n";

    let parsed = parse_email(raw).unwrap();
    assert_eq!(parsed.subject.as_deref(), Some("Headers only"));
    assert!(parsed.body_text.is_none());
    assert!(parsed.body_html.is_none());
}

#[test]
fn parse_empty_input() {
    let result = parse_email(b"");
    assert!(matches!(result, Err(Error::EmptyInput)));
}

#[test]
fn parse_missing_from() {
    let raw = b"Subject: No from\r\n\
                 Date: Thu, 13 Feb 2025 15:47:33 +0000\r\n\
                 \r\n";

    let parsed = parse_email(raw).expect(
        "messages without From should still parse when other headers are usable \
         (consumer robustness / Postel's law)",
    );
    assert!(
        parsed.from.is_empty(),
        "missing From should produce an empty from list, got {:?}",
        parsed.from
    );
    assert_eq!(parsed.subject.as_deref(), Some("No from"));
}

/// Top-level messages with no header fields at all are malformed, but the
/// parser should still surface the body when it is otherwise usable text.
/// This keeps the public API aligned with its documented "partial message"
/// behavior and mirrors how MIME subparts already tolerate empty headers.
#[test]
fn parse_body_only_message_without_headers() {
    let raw = b"Hello from the body only parser path";

    let parsed = parse_email(raw).expect(
        "body-only top-level messages should still parse so consumers can \
         inspect truncated or malformed maildrops",
    );
    assert!(parsed.from.is_empty());
    assert!(parsed.raw_headers.is_empty());
    assert_eq!(
        parsed.body_text.as_deref(),
        Some("Hello from the body only parser path")
    );
    assert_eq!(parsed.size, raw.len() as u64);
}

/// RFC 2046 permits MIME body-parts with an empty header block. Apply the
/// same liberal parsing to malformed top-level messages that begin with a
/// blank line and then body text.
#[test]
fn parse_blank_line_then_body_without_headers() {
    let raw = b"\r\nBody after an empty top-level header block";

    let parsed = parse_email(raw)
        .expect("an empty top-level header block followed by body text should still parse");
    assert!(parsed.from.is_empty());
    assert!(parsed.raw_headers.is_empty());
    assert_eq!(
        parsed.body_text.as_deref(),
        Some("Body after an empty top-level header block")
    );
}

/// RFC 5322 Section 3.6.8 allows optional fields outside the well-known
/// header set. The parser should still surface those fields, even when no
/// standard message headers are present, so consumers can inspect partial
/// or malformed messages.
#[test]
fn parse_custom_headers_without_well_known_headers() {
    let raw = b"X-Trace: 12345\r\n\
                 List-Id: Example List <list.example>\r\n\
                 \r\n\
                 body";

    let parsed = parse_email(raw).expect(
        "messages with only optional fields should still parse so consumers \
         can inspect extra headers and body content",
    );
    assert!(parsed.from.is_empty());
    assert!(parsed.subject.is_none());
    assert_eq!(parsed.body_text.as_deref(), Some("body"));
    assert!(
        parsed
            .extra_headers
            .iter()
            .any(|(name, value)| name == "x-trace" && value == "12345"),
        "X-Trace must be preserved in extra_headers"
    );
    assert!(
        parsed
            .extra_headers
            .iter()
            .any(|(name, value)| name == "list-id" && value == "Example List <list.example>"),
        "List-Id must be preserved in extra_headers"
    );
}

#[test]
fn parse_quoted_printable_body() {
    let raw = b"From: a@b.com\r\n\
                 Date: Thu, 13 Feb 2025 15:47:33 +0000\r\n\
                 Content-Type: text/plain; charset=utf-8\r\n\
                 Content-Transfer-Encoding: quoted-printable\r\n\
                 \r\n\
                 Hello=20World=0D=0ASoft=\r\n break";

    let parsed = parse_email(raw).unwrap();
    assert_eq!(
        parsed.body_text.as_deref(),
        Some("Hello World\r\nSoft break")
    );
}

#[test]
fn parse_base64_body() {
    let raw = b"From: a@b.com\r\n\
                 Date: Thu, 13 Feb 2025 15:47:33 +0000\r\n\
                 Content-Type: text/plain; charset=utf-8\r\n\
                 Content-Transfer-Encoding: base64\r\n\
                 \r\n\
                 SGVsbG8gV29ybGQ=\r\n";

    let parsed = parse_email(raw).unwrap();
    assert_eq!(parsed.body_text.as_deref(), Some("Hello World"));
}

#[test]
fn parse_nested_multipart_section_numbers() {
    let raw = b"From: a@b.com\r\n\
                 Date: Thu, 13 Feb 2025 15:47:33 +0000\r\n\
                 Content-Type: multipart/mixed; boundary=\"outer\"\r\n\
                 \r\n\
                 --outer\r\n\
                 Content-Type: multipart/alternative; boundary=\"inner\"\r\n\
                 \r\n\
                 --inner\r\n\
                 Content-Type: text/plain\r\n\
                 \r\n\
                 Plain\r\n\
                 --inner\r\n\
                 Content-Type: text/html\r\n\
                 \r\n\
                 <b>HTML</b>\r\n\
                 --inner--\r\n\
                 --outer\r\n\
                 Content-Type: application/pdf\r\n\
                 Content-Disposition: attachment; filename=\"f.pdf\"\r\n\
                 \r\n\
                 DATA\r\n\
                 --outer--";

    let parsed = parse_email(raw).unwrap();
    assert_eq!(parsed.body_text.as_deref(), Some("Plain"));
    assert_eq!(parsed.body_html.as_deref(), Some("<b>HTML</b>"));
    assert_eq!(parsed.attachments.len(), 1);
    // Attachment is part 2 of the outer multipart
    assert_eq!(parsed.attachments[0].section.as_deref(), Some("2"));
}

#[test]
fn parse_rfc2231_filename() {
    let raw = b"From: a@b.com\r\n\
                 Date: Thu, 13 Feb 2025 15:47:33 +0000\r\n\
                 Content-Type: multipart/mixed; boundary=\"b\"\r\n\
                 \r\n\
                 --b\r\n\
                 Content-Type: text/plain\r\n\
                 \r\n\
                 Body\r\n\
                 --b\r\n\
                 Content-Type: application/pdf\r\n\
                 Content-Disposition: attachment; filename*=UTF-8''r%C3%A9sum%C3%A9.pdf\r\n\
                 \r\n\
                 DATA\r\n\
                 --b--";

    let parsed = parse_email(raw).unwrap();
    assert_eq!(parsed.attachments.len(), 1);
    assert_eq!(
        parsed.attachments[0].filename.as_deref(),
        Some("résumé.pdf")
    );
}

#[test]
fn parse_raw_headers_preserved() {
    let raw = b"From: a@b.com\r\n\
                 Subject: Test\r\n\
                 Date: Thu, 13 Feb 2025 15:47:33 +0000\r\n\
                 \r\n\
                 Body";

    let parsed = parse_email(raw).unwrap();
    assert!(parsed.raw_headers.contains("From: a@b.com"));
    assert!(parsed.raw_headers.contains("Subject: Test"));
}

#[test]
fn parse_lf_only_line_endings() {
    let raw = b"From: a@b.com\n\
                 Subject: LF\n\
                 Date: Thu, 13 Feb 2025 15:47:33 +0000\n\
                 \n\
                 Body with LF";

    let parsed = parse_email(raw).unwrap();
    assert_eq!(parsed.subject.as_deref(), Some("LF"));
    assert_eq!(parsed.body_text.as_deref(), Some("Body with LF"));
}

#[test]
fn parse_header_continuation_lines() {
    // Continuation line starts with a space (RFC 5322 Section 2.2.3)
    // Can't use `\` line continuation as it strips leading whitespace.
    let raw = b"From: a@b.com\r\nSubject: This is a very long\r\n subject line that wraps\r\nDate: Thu, 13 Feb 2025 15:47:33 +0000\r\n\r\n";

    let parsed = parse_email(raw).unwrap();
    assert_eq!(
        parsed.subject.as_deref(),
        Some("This is a very long subject line that wraps")
    );
}

#[test]
fn parse_garbage_input_best_effort() {
    // Binary garbage — no valid From header → error
    let result = parse_email(b"\x00\x01\x02\x03\xff\xfe");
    assert!(result.is_err());
}

#[test]
fn parse_truncated_multipart() {
    // Multipart with missing closing boundary
    let raw = b"From: a@b.com\r\n\
                 Date: Thu, 13 Feb 2025 15:47:33 +0000\r\n\
                 Content-Type: multipart/mixed; boundary=\"trunc\"\r\n\
                 \r\n\
                 --trunc\r\n\
                 Content-Type: text/plain\r\n\
                 \r\n\
                 Some text here";

    let parsed = parse_email(raw).unwrap();
    assert_eq!(parsed.body_text.as_deref(), Some("Some text here"));
}

#[test]
fn decode_adjacent_encoded_words() {
    // RFC 2047 Section 6.2: whitespace between adjacent encoded words is removed
    let input = "=?UTF-8?B?SGVs?= =?UTF-8?B?bG8=?=";
    let decoded = decode_encoded_words(input);
    assert_eq!(decoded, "Hello");
}

#[test]
fn decode_iso8859_encoded_word() {
    // =?ISO-8859-1?Q?caf=E9?= → "café"
    let input = "=?ISO-8859-1?Q?caf=E9?=";
    let decoded = decode_encoded_words(input);
    assert_eq!(decoded, "café");
}

#[test]
fn parse_date_without_seconds() {
    let dt = parse_rfc5322_date("Thu, 13 Feb 2025 15:47 +0000").unwrap();
    assert_eq!(dt.hour, 15);
    assert_eq!(dt.minute, 47);
    assert_eq!(dt.second, 0);
}

#[test]
fn parse_two_digit_year() {
    let dt = parse_rfc5322_date("13 Feb 99 12:00:00 +0000").unwrap();
    assert_eq!(dt.year, 1999);

    let dt = parse_rfc5322_date("13 Feb 25 12:00:00 +0000").unwrap();
    assert_eq!(dt.year, 2025);
}

#[test]
fn parse_three_digit_year_rfc5322_section_4_3() {
    // RFC 5322 Section 4.3: any 3-digit year should have 1900 added.
    let dt = parse_rfc5322_date("13 Feb 107 12:00:00 +0000").unwrap();
    assert_eq!(
        dt.year, 2007,
        "3-digit year 107 must map to 2007 per RFC 5322 Section 4.3"
    );

    let dt = parse_rfc5322_date("13 Feb 100 12:00:00 +0000").unwrap();
    assert_eq!(
        dt.year, 2000,
        "3-digit year 100 must map to 2000 per RFC 5322 Section 4.3"
    );

    let dt = parse_rfc5322_date("13 Feb 999 12:00:00 +0000").unwrap();
    assert_eq!(
        dt.year, 2899,
        "3-digit year 999 must map to 2899 per RFC 5322 Section 4.3"
    );
}

#[test]
fn parse_two_digit_year_rfc5322_section_4_3_cutoff() {
    // RFC 5322 Section 4.3: 2-digit years 00-49 → +2000, 50-99 → +1900.
    // The cutoff is 50, not 70.

    // Year 50 should map to 1950 (not 2050)
    let dt = parse_rfc5322_date("13 Feb 50 12:00:00 +0000").unwrap();
    assert_eq!(
        dt.year, 1950,
        "2-digit year 50 must map to 1950 per RFC 5322 Section 4.3"
    );

    // Year 69 should map to 1969 (not 2069)
    let dt = parse_rfc5322_date("13 Feb 69 12:00:00 +0000").unwrap();
    assert_eq!(
        dt.year, 1969,
        "2-digit year 69 must map to 1969 per RFC 5322 Section 4.3"
    );

    // Year 49 should map to 2049
    let dt = parse_rfc5322_date("13 Feb 49 12:00:00 +0000").unwrap();
    assert_eq!(
        dt.year, 2049,
        "2-digit year 49 must map to 2049 per RFC 5322 Section 4.3"
    );
}

#[test]
fn parse_non_text_part_is_attachment() {
    let raw = b"From: a@b.com\r\n\
                 Date: Thu, 13 Feb 2025 15:47:33 +0000\r\n\
                 Content-Type: multipart/mixed; boundary=\"b\"\r\n\
                 \r\n\
                 --b\r\n\
                 Content-Type: text/plain\r\n\
                 \r\n\
                 Text\r\n\
                 --b\r\n\
                 Content-Type: image/jpeg\r\n\
                 \r\n\
                 JPEG_DATA\r\n\
                 --b--";

    let parsed = parse_email(raw).unwrap();
    // image/jpeg without explicit disposition should be treated as attachment
    assert_eq!(parsed.attachments.len(), 1);
    assert_eq!(parsed.attachments[0].content_type, "image/jpeg");
}

#[test]
fn parse_windows1252_body() {
    // Windows-1252 body with smart quotes
    let raw = b"From: a@b.com\r\n\
                 Date: Thu, 13 Feb 2025 15:47:33 +0000\r\n\
                 Content-Type: text/plain; charset=windows-1252\r\n\
                 \r\n\
                 \x93Hello\x94"; // Smart double quotes in Windows-1252

    let parsed = parse_email(raw).unwrap();
    let text = parsed.body_text.unwrap();
    assert!(text.contains("Hello"));
    // Smart quotes should be converted to Unicode
    assert!(text.contains('\u{201c}') || text.contains('\u{201d}'));
}

#[test]
fn parse_html_only_body() {
    let raw = b"From: a@b.com\r\n\
                 Date: Thu, 13 Feb 2025 15:47:33 +0000\r\n\
                 Content-Type: text/html; charset=utf-8\r\n\
                 \r\n\
                 <html><body>Hello</body></html>";

    let parsed = parse_email(raw).unwrap();
    assert!(parsed.body_text.is_none());
    assert_eq!(
        parsed.body_html.as_deref(),
        Some("<html><body>Hello</body></html>")
    );
}

#[test]
fn parse_bcc_addresses() {
    let raw = b"From: a@b.com\r\n\
                 To: to@x.com\r\n\
                 Bcc: hidden@x.com, secret@x.com\r\n\
                 Date: Thu, 13 Feb 2025 15:47:33 +0000\r\n\
                 \r\n";

    let parsed = parse_email(raw).unwrap();
    assert_eq!(parsed.bcc.len(), 2);
    assert_eq!(parsed.bcc[0].email, "hidden@x.com");
}

#[test]
fn mime_depth_limit() {
    // Construct a deeply nested multipart that exceeds MAX_MIME_DEPTH
    // Just verify it doesn't stack overflow
    let mut msg = b"From: a@b.com\r\nDate: Thu, 13 Feb 2025 15:47:33 +0000\r\n\
                    Content-Type: multipart/mixed; boundary=\"b0\"\r\n\r\n"
        .to_vec();

    for i in 0..70 {
        msg.extend_from_slice(
            format!(
                "--b{i}\r\nContent-Type: multipart/mixed; boundary=\"b{}\"\r\n\r\n",
                i + 1
            )
            .as_bytes(),
        );
    }
    msg.extend_from_slice(b"--b70\r\nContent-Type: text/plain\r\n\r\nDeep\r\n--b70--\r\n");

    let parsed = parse_email(&msg).unwrap();
    // Should not panic or stack overflow — may not find the body due to depth limit
    assert!(parsed.body_text.is_none() || parsed.body_text.is_some());
}

#[test]
fn parse_reply_to() {
    let raw = b"From: a@b.com\r\n\
                 Reply-To: noreply@example.com, support@example.com\r\n\
                 Date: Thu, 13 Feb 2025 15:47:33 +0000\r\n\
                 \r\n";

    let parsed = parse_email(raw).unwrap();
    assert_eq!(parsed.reply_to.len(), 2);
    assert_eq!(parsed.reply_to[0].email, "noreply@example.com");
    assert_eq!(parsed.reply_to[1].email, "support@example.com");
}

#[test]
fn parse_gb2312_encoded_word() {
    // GB2312 encoded word: "你好" (nǐ hǎo) in base64
    // "你好" in GB2312 is: 0xC4, 0xE3, 0xBA, 0xC3
    let raw = b"From: sender@example.com\r\n\
                 Subject: =?GB2312?B?xOO6ww==?=\r\n\
                 Date: Thu, 13 Feb 2025 15:47:33 +0000\r\n\
                 \r\n";

    let parsed = parse_email(raw).unwrap();
    assert_eq!(parsed.subject.as_deref(), Some("你好"));
}

#[test]
fn parse_content_id_strips_brackets() {
    let raw = b"From: a@b.com\r\n\
                 Date: Thu, 13 Feb 2025 15:47:33 +0000\r\n\
                 Content-Type: multipart/mixed; boundary=\"b\"\r\n\
                 \r\n\
                 --b\r\n\
                 Content-Type: text/plain\r\n\
                 \r\n\
                 Body\r\n\
                 --b\r\n\
                 Content-Type: image/png\r\n\
                 Content-ID: <cid:image001@01D00000.00000000>\r\n\
                 \r\n\
                 PNG\r\n\
                 --b--";

    let parsed = parse_email(raw).unwrap();
    assert_eq!(
        parsed.attachments[0].content_id.as_deref(),
        Some("cid:image001@01D00000.00000000")
    );
}

#[test]
fn parse_attachment_without_filename() {
    // Attachment with Content-Disposition but no filename parameter
    let raw = b"From: a@b.com\r\n\
                 Date: Thu, 13 Feb 2025 15:47:33 +0000\r\n\
                 Content-Type: multipart/mixed; boundary=\"b\"\r\n\
                 \r\n\
                 --b\r\n\
                 Content-Type: text/plain\r\n\
                 \r\n\
                 Body\r\n\
                 --b\r\n\
                 Content-Type: application/octet-stream\r\n\
                 Content-Disposition: attachment\r\n\
                 \r\n\
                 BINARY\r\n\
                 --b--";

    let parsed = parse_email(raw).unwrap();
    assert_eq!(parsed.attachments.len(), 1);
    assert!(parsed.attachments[0].filename.is_none());
    assert_eq!(
        parsed.attachments[0].content_type,
        "application/octet-stream"
    );
    assert!(!parsed.attachments[0].is_inline);
}

#[test]
fn parse_content_type_without_charset_defaults() {
    // No charset parameter — should default to us-ascii/utf-8 handling
    let raw = b"From: a@b.com\r\n\
                 Date: Thu, 13 Feb 2025 15:47:33 +0000\r\n\
                 Content-Type: text/plain\r\n\
                 \r\n\
                 Hello ASCII";

    let parsed = parse_email(raw).unwrap();
    assert_eq!(parsed.body_text.as_deref(), Some("Hello ASCII"));
}

/// A MIME part with no Content-Type header inside a multipart
/// message must default to "text/plain; charset=us-ascii" per RFC 2045
/// Section 5.2.
#[test]
fn parse_mime_part_no_content_type_defaults_to_us_ascii() {
    // Part has Content-Transfer-Encoding but no Content-Type.
    // RFC 2045 Section 5.2: default is "text/plain; charset=us-ascii".
    let raw = b"From: a@b.com\r\n\
                 Date: Thu, 13 Feb 2025 15:47:33 +0000\r\n\
                 Content-Type: multipart/mixed; boundary=\"b\"\r\n\
                 \r\n\
                 --b\r\n\
                 Content-Transfer-Encoding: 7bit\r\n\
                 \r\n\
                 Hello ASCII\r\n\
                 --b--";

    let parsed = parse_email(raw).unwrap();
    assert_eq!(parsed.body_text.as_deref(), Some("Hello ASCII"));
}

/// A MIME part with zero headers (only the blank line
/// separator, no Content-Type or other headers) must still be parsed.
/// This is a valid RFC 2046 construct — the blank line after the
/// boundary delimiter starts the body when there are no part headers.
#[test]
fn parse_mime_part_no_headers_at_all() {
    let raw = b"From: a@b.com\r\n\
                 Date: Thu, 13 Feb 2025 15:47:33 +0000\r\n\
                 Content-Type: multipart/mixed; boundary=\"b\"\r\n\
                 \r\n\
                 --b\r\n\
                 \r\n\
                 Headerless body\r\n\
                 --b--";

    let parsed = parse_email(raw).unwrap();
    // Part with no headers defaults to text/plain per RFC 2045 Section 5.2.
    assert_eq!(parsed.body_text.as_deref(), Some("Headerless body"));
}

#[test]
fn parse_multipart_only_attachments() {
    // No text/plain or text/html parts — only attachments
    let raw = b"From: a@b.com\r\n\
                 Date: Thu, 13 Feb 2025 15:47:33 +0000\r\n\
                 Content-Type: multipart/mixed; boundary=\"b\"\r\n\
                 \r\n\
                 --b\r\n\
                 Content-Type: application/pdf\r\n\
                 Content-Disposition: attachment; filename=\"a.pdf\"\r\n\
                 \r\n\
                 PDF1\r\n\
                 --b\r\n\
                 Content-Type: image/png\r\n\
                 Content-Disposition: attachment; filename=\"b.png\"\r\n\
                 \r\n\
                 PNG2\r\n\
                 --b--";

    let parsed = parse_email(raw).unwrap();
    assert!(parsed.body_text.is_none());
    assert!(parsed.body_html.is_none());
    assert_eq!(parsed.attachments.len(), 2);
    assert_eq!(parsed.attachments[0].section.as_deref(), Some("1"));
    assert_eq!(parsed.attachments[1].section.as_deref(), Some("2"));
}

#[test]
fn parse_unknown_charset_body_fallback() {
    // Unknown charset should fall back to UTF-8 lossy conversion
    let raw = b"From: a@b.com\r\n\
                 Date: Thu, 13 Feb 2025 15:47:33 +0000\r\n\
                 Content-Type: text/plain; charset=x-unknown-fake\r\n\
                 \r\n\
                 Plain text in unknown charset";

    let parsed = parse_email(raw).unwrap();
    // encoding_rs falls back to UTF-8 for unknown charsets
    assert!(parsed.body_text.is_some());
    assert!(parsed.body_text.unwrap().contains("Plain text"));
}

#[test]
fn parse_content_id_without_disposition_is_inline() {
    // Part with Content-ID but no Content-Disposition — should be inline
    let raw = b"From: a@b.com\r\n\
                 Date: Thu, 13 Feb 2025 15:47:33 +0000\r\n\
                 Content-Type: multipart/mixed; boundary=\"b\"\r\n\
                 \r\n\
                 --b\r\n\
                 Content-Type: text/plain\r\n\
                 \r\n\
                 Body\r\n\
                 --b\r\n\
                 Content-Type: image/gif\r\n\
                 Content-ID: <img42>\r\n\
                 \r\n\
                 GIF89a\r\n\
                 --b--";

    let parsed = parse_email(raw).unwrap();
    assert_eq!(parsed.attachments.len(), 1);
    assert!(parsed.attachments[0].is_inline);
    assert_eq!(parsed.attachments[0].content_id.as_deref(), Some("img42"));
}

#[test]
fn parse_overlong_subject() {
    // Very long subject line — should not crash or truncate
    let long_subject = "A".repeat(10_000);
    let raw = format!(
        "From: a@b.com\r\n\
         Subject: {long_subject}\r\n\
         Date: Thu, 13 Feb 2025 15:47:33 +0000\r\n\
         \r\n"
    );

    let parsed = parse_email(raw.as_bytes()).unwrap();
    assert_eq!(parsed.subject.as_deref(), Some(long_subject.as_str()));
}

#[test]
fn parse_multiple_from_preserves_all() {
    // RFC 5322 Section 3.6.2: from = "From:" mailbox-list CRLF
    // mailbox-list = mailbox *("," mailbox)
    // All originator mailboxes must be preserved.
    let raw = b"From: Alice <alice@example.com>, Bob <bob@example.com>\r\n\
                Date: Thu, 13 Feb 2025 15:47:33 +0000\r\n\
                \r\n";

    let parsed = parse_email(raw).unwrap();
    assert_eq!(
        parsed.from.len(),
        2,
        "RFC 5322 Section 3.6.2: all From mailboxes must be preserved"
    );
    assert_eq!(parsed.from[0].email, "alice@example.com");
    assert_eq!(parsed.from[0].name.as_deref(), Some("Alice"));
    assert_eq!(parsed.from[1].email, "bob@example.com");
    assert_eq!(parsed.from[1].name.as_deref(), Some("Bob"));
}

#[test]
fn parse_multiple_from_takes_first() {
    // RFC 5322 Section 3.6.2: from = "From:" mailbox-list CRLF
    // All originator mailboxes must be preserved.
    let raw = b"From: first@example.com, second@example.com\r\n\
                 Date: Thu, 13 Feb 2025 15:47:33 +0000\r\n\
                 \r\n";

    let parsed = parse_email(raw).unwrap();
    assert_eq!(parsed.from.len(), 2);
    assert_eq!(parsed.from[0].email, "first@example.com");
    assert_eq!(parsed.from[1].email, "second@example.com");
}

#[test]
fn parse_multipart_no_boundary_param() {
    // multipart/mixed but no boundary parameter — fallback to simple body
    let raw = b"From: a@b.com\r\n\
                 Date: Thu, 13 Feb 2025 15:47:33 +0000\r\n\
                 Content-Type: multipart/mixed\r\n\
                 \r\n\
                 Some text content";

    let parsed = parse_email(raw).unwrap();
    // Should not panic; falls back to treating body as simple text
    assert!(parsed.body_text.is_some());
}

#[test]
fn parse_empty_body_after_headers() {
    // Headers + blank line + no body content
    let raw = b"From: a@b.com\r\n\
                 Subject: Empty body\r\n\
                 Date: Thu, 13 Feb 2025 15:47:33 +0000\r\n\
                 \r\n";

    let parsed = parse_email(raw).unwrap();
    assert_eq!(parsed.subject.as_deref(), Some("Empty body"));
    assert!(parsed.body_text.is_none());
}

#[test]
fn parse_mixed_charset_encoded_words() {
    // Adjacent encoded words with different charsets
    let raw = b"From: a@b.com\r\n\
                 Subject: =?UTF-8?B?SGVsbG8=?= =?ISO-8859-1?Q?_caf=E9?=\r\n\
                 Date: Thu, 13 Feb 2025 15:47:33 +0000\r\n\
                 \r\n";

    let parsed = parse_email(raw).unwrap();
    assert_eq!(parsed.subject.as_deref(), Some("Hello caf\u{e9}"));
}

#[test]
fn parse_no_date_header() {
    // Message without Date header — date should be None
    let raw = b"From: a@b.com\r\n\
                 Subject: No date\r\n\
                 \r\n\
                 Body";

    let parsed = parse_email(raw).unwrap();
    assert!(parsed.date.is_none());
    assert_eq!(parsed.subject.as_deref(), Some("No date"));
}

#[test]
fn parse_explicit_attachment_text_plain() {
    // text/plain with Content-Disposition: attachment should be attachment, not body
    let raw = b"From: a@b.com\r\n\
                 Date: Thu, 13 Feb 2025 15:47:33 +0000\r\n\
                 Content-Type: multipart/mixed; boundary=\"b\"\r\n\
                 \r\n\
                 --b\r\n\
                 Content-Type: text/plain\r\n\
                 \r\n\
                 Body text\r\n\
                 --b\r\n\
                 Content-Type: text/plain\r\n\
                 Content-Disposition: attachment; filename=\"log.txt\"\r\n\
                 \r\n\
                 Log file content\r\n\
                 --b--";

    let parsed = parse_email(raw).unwrap();
    assert_eq!(parsed.body_text.as_deref(), Some("Body text"));
    assert_eq!(parsed.attachments.len(), 1);
    assert_eq!(parsed.attachments[0].filename.as_deref(), Some("log.txt"));
    assert_eq!(parsed.attachments[0].content_type, "text/plain");
}

#[test]
fn parse_date_negative_timezone() {
    let raw = b"From: a@b.com\r\n\
                 Date: Fri, 14 Feb 2025 09:15:00 -0800\r\n\
                 \r\n";

    let parsed = parse_email(raw).unwrap();
    let date = parsed.date.unwrap();
    assert_eq!(date.tz_offset_minutes, -480);
}

#[test]
fn parse_size_equals_input_length() {
    let raw = b"From: a@b.com\r\nDate: Thu, 13 Feb 2025 15:47:33 +0000\r\n\r\nBody";
    let parsed = parse_email(raw).unwrap();
    assert_eq!(parsed.size, raw.len() as u64);
}

#[test]
fn parse_binary_garbage_returns_error() {
    // Pure binary garbage — no From header → MissingFrom
    let garbage: Vec<u8> = (0..=255_u8).collect();
    let result = parse_email(&garbage);
    assert!(result.is_err());
}

#[test]
fn parse_folded_encoded_word_subject() {
    // Subject with encoded word that spans a folded line
    let raw = b"From: a@b.com\r\nSubject: =?UTF-8?B?SGVsbG8=?=\r\n =?UTF-8?B?V29ybGQ=?=\r\nDate: Thu, 13 Feb 2025 15:47:33 +0000\r\n\r\n";

    let parsed = parse_email(raw).unwrap();
    assert_eq!(parsed.subject.as_deref(), Some("HelloWorld"));
}

// -----------------------------------------------------------------------
// Additional edge case tests
// -----------------------------------------------------------------------

#[test]
fn parse_encoded_word_lowercase_encoding() {
    // RFC 2047: encoding indicator is case-insensitive
    let raw = b"From: sender@example.com\r\n\
                 Subject: =?utf-8?b?SGVsbG8=?= =?utf-8?q?_World?=\r\n\
                 Date: Thu, 13 Feb 2025 15:47:33 +0000\r\n\
                 \r\n";

    let parsed = parse_email(raw).unwrap();
    assert_eq!(parsed.subject.as_deref(), Some("Hello World"));
}

#[test]
fn parse_malformed_encoded_word_passthrough() {
    // Incomplete encoded word should be passed through as literal text
    let raw = b"From: a@b.com\r\n\
                 Subject: =?UTF-8?B?broken\r\n\
                 Date: Thu, 13 Feb 2025 15:47:33 +0000\r\n\
                 \r\n";

    let parsed = parse_email(raw).unwrap();
    // Should not crash; subject contains the literal malformed encoded word
    assert!(parsed.subject.is_some());
    assert!(parsed.subject.unwrap().contains("=?"));
}

#[test]
fn parse_encoded_word_unknown_encoding_type() {
    // Unknown encoding type (not B or Q) — should pass through
    let raw = b"From: a@b.com\r\n\
                 Subject: =?UTF-8?X?data?=\r\n\
                 Date: Thu, 13 Feb 2025 15:47:33 +0000\r\n\
                 \r\n";

    let parsed = parse_email(raw).unwrap();
    assert!(parsed.subject.is_some());
    // Unknown encoding passed through as literal
    assert!(parsed.subject.unwrap().contains("=?"));
}

#[test]
fn parse_utf8_directly_in_headers_rfc6532() {
    // RFC 6532: UTF-8 characters directly in headers (no encoded words)
    let raw = "From: José <jose@example.com>\r\n\
                Subject: Ñoño café\r\n\
                Date: Thu, 13 Feb 2025 15:47:33 +0000\r\n\
                \r\n\
                Body";

    let parsed = parse_email(raw.as_bytes()).unwrap();
    assert_eq!(parsed.subject.as_deref(), Some("Ñoño café"));
    assert_eq!(parsed.from[0].name.as_deref(), Some("José"));
    assert_eq!(parsed.from[0].email, "jose@example.com");
}

#[test]
fn parse_multipart_with_preamble() {
    // RFC 2046 Section 5.1.1: preamble text before the first boundary is ignored
    let raw = b"From: a@b.com\r\n\
                 Date: Thu, 13 Feb 2025 15:47:33 +0000\r\n\
                 Content-Type: multipart/mixed; boundary=\"preamble-test\"\r\n\
                 \r\n\
                 This is the preamble, which should be ignored.\r\n\
                 --preamble-test\r\n\
                 Content-Type: text/plain\r\n\
                 \r\n\
                 Actual body\r\n\
                 --preamble-test--";

    let parsed = parse_email(raw).unwrap();
    assert_eq!(parsed.body_text.as_deref(), Some("Actual body"));
}

#[test]
fn parse_attachment_name_from_content_type() {
    // Filename from Content-Type name= param when Content-Disposition has no filename
    let raw = b"From: a@b.com\r\n\
                 Date: Thu, 13 Feb 2025 15:47:33 +0000\r\n\
                 Content-Type: multipart/mixed; boundary=\"b\"\r\n\
                 \r\n\
                 --b\r\n\
                 Content-Type: text/plain\r\n\
                 \r\n\
                 Body\r\n\
                 --b\r\n\
                 Content-Type: application/pdf; name=\"report.pdf\"\r\n\
                 Content-Disposition: attachment\r\n\
                 \r\n\
                 PDF\r\n\
                 --b--";

    let parsed = parse_email(raw).unwrap();
    assert_eq!(parsed.attachments.len(), 1);
    assert_eq!(
        parsed.attachments[0].filename.as_deref(),
        Some("report.pdf")
    );
}

/// RFC 2183 Section 2.3 lets a part carry a `filename` hint even when the
/// sender omits `Content-Disposition`.  The parser should preserve that as
/// attachment metadata instead of promoting the text part to `body_text`.
#[test]
fn text_plain_name_parameter_is_treated_as_attachment() {
    let raw = b"From: a@b.com\r\n\
                 Date: Thu, 13 Feb 2025 15:47:33 +0000\r\n\
                 Content-Type: text/plain; name=\"notes.txt\"\r\n\
                 Content-Transfer-Encoding: base64\r\n\
                 \r\n\
                 SGVsbG8=\r\n";

    let parsed = parse_email(raw).unwrap();
    assert!(
        parsed.body_text.is_none(),
        "filename-bearing text/plain part should not become body_text"
    );
    assert_eq!(parsed.attachments.len(), 1);
    assert_eq!(parsed.attachments[0].content_type, "text/plain");
    assert_eq!(parsed.attachments[0].filename.as_deref(), Some("notes.txt"));
}

/// Multipart variant of the same consumer-visible bug: a secondary
/// `text/plain; name=...` part should remain an attachment, not be folded
/// into the message body.
#[test]
fn multipart_text_plain_name_parameter_is_treated_as_attachment() {
    let raw = b"From: a@b.com\r\n\
                 Date: Thu, 13 Feb 2025 15:47:33 +0000\r\n\
                 Content-Type: multipart/mixed; boundary=\"b\"\r\n\
                 \r\n\
                 --b\r\n\
                 Content-Type: text/plain; name=\"notes.txt\"\r\n\
                 Content-Transfer-Encoding: base64\r\n\
                 \r\n\
                 SGVsbG8=\r\n\
                 --b--";

    let parsed = parse_email(raw).unwrap();
    assert!(
        parsed.body_text.is_none(),
        "filename-bearing multipart text/plain part should not become body_text"
    );
    assert_eq!(parsed.attachments.len(), 1);
    assert_eq!(parsed.attachments[0].content_type, "text/plain");
    assert_eq!(parsed.attachments[0].filename.as_deref(), Some("notes.txt"));
}

#[test]
fn parse_qp_soft_break_lf_only() {
    // Quoted-printable soft line break with just LF (not CRLF)
    let raw = b"From: a@b.com\r\n\
                 Date: Thu, 13 Feb 2025 15:47:33 +0000\r\n\
                 Content-Type: text/plain; charset=utf-8\r\n\
                 Content-Transfer-Encoding: quoted-printable\r\n\
                 \r\n\
                 Hello=\nWorld";

    let parsed = parse_email(raw).unwrap();
    assert_eq!(parsed.body_text.as_deref(), Some("HelloWorld"));
}

#[test]
fn parse_subject_mixed_encoded_and_plain() {
    // Subject with plain text, encoded word, and more plain text
    let raw = b"From: a@b.com\r\n\
                 Subject: Re: =?UTF-8?B?SGVsbG8=?= there\r\n\
                 Date: Thu, 13 Feb 2025 15:47:33 +0000\r\n\
                 \r\n";

    let parsed = parse_email(raw).unwrap();
    assert_eq!(parsed.subject.as_deref(), Some("Re: Hello there"));
}

#[test]
fn parse_whitespace_only_body() {
    // Body consisting only of whitespace
    let raw = b"From: a@b.com\r\n\
                 Date: Thu, 13 Feb 2025 15:47:33 +0000\r\n\
                 Content-Type: text/plain; charset=utf-8\r\n\
                 \r\n\
                 \r\n  \r\n";

    let parsed = parse_email(raw).unwrap();
    // Whitespace-only body should still be returned as body_text
    assert!(parsed.body_text.is_some());
}

#[test]
fn parse_date_missing_timezone() {
    // Date without timezone — should default to +0000
    let raw = b"From: a@b.com\r\n\
                 Date: 13 Feb 2025 12:00:00\r\n\
                 \r\n";

    let parsed = parse_email(raw).unwrap();
    let date = parsed.date.unwrap();
    assert_eq!(date.year, 2025);
    assert_eq!(date.tz_offset_minutes, 0);
}

#[test]
fn parse_deeply_nested_section_dot_notation() {
    // Verify section numbers use correct dot notation for nested multipart
    let raw = b"From: a@b.com\r\n\
                 Date: Thu, 13 Feb 2025 15:47:33 +0000\r\n\
                 Content-Type: multipart/mixed; boundary=\"outer\"\r\n\
                 \r\n\
                 --outer\r\n\
                 Content-Type: multipart/related; boundary=\"rel\"\r\n\
                 \r\n\
                 --rel\r\n\
                 Content-Type: text/html\r\n\
                 \r\n\
                 <img src=\"cid:img1\">\r\n\
                 --rel\r\n\
                 Content-Type: image/png\r\n\
                 Content-ID: <img1>\r\n\
                 \r\n\
                 PNG_DATA\r\n\
                 --rel--\r\n\
                 --outer\r\n\
                 Content-Type: application/pdf\r\n\
                 Content-Disposition: attachment; filename=\"doc.pdf\"\r\n\
                 \r\n\
                 PDF\r\n\
                 --outer--";

    let parsed = parse_email(raw).unwrap();
    // HTML is section 1.1 of the outer, inline image is 1.2
    assert!(parsed.body_html.is_some());
    assert!(parsed.body_html.unwrap().contains("cid:img1"));
    // Inline image attachment: section 1.2
    let inline_att = parsed
        .attachments
        .iter()
        .find(|a| a.content_type == "image/png")
        .unwrap();
    assert_eq!(inline_att.section.as_deref(), Some("1.2"));
    assert!(inline_att.is_inline);
    // PDF attachment: section 2
    let pdf_att = parsed
        .attachments
        .iter()
        .find(|a| a.content_type == "application/pdf")
        .unwrap();
    assert_eq!(pdf_att.section.as_deref(), Some("2"));
}

#[test]
fn parse_non_ascii_bytes_in_body() {
    // Raw non-ASCII bytes in body without charset declaration
    let raw = b"From: a@b.com\r\n\
                 Date: Thu, 13 Feb 2025 15:47:33 +0000\r\n\
                 Content-Type: text/plain\r\n\
                 \r\n\
                 Hello \xff\xfe world";

    let parsed = parse_email(raw).unwrap();
    // Should not panic — lossy UTF-8 conversion
    assert!(parsed.body_text.is_some());
}

#[test]
fn parse_base64_body_with_line_breaks() {
    // Base64 body with CRLF line breaks in the middle (RFC 2045 Section 6.8)
    let raw = b"From: a@b.com\r\n\
                 Date: Thu, 13 Feb 2025 15:47:33 +0000\r\n\
                 Content-Type: text/plain; charset=utf-8\r\n\
                 Content-Transfer-Encoding: base64\r\n\
                 \r\n\
                 SGVs\r\nbG8g\r\nV29y\r\nbGQ=";

    let parsed = parse_email(raw).unwrap();
    assert_eq!(parsed.body_text.as_deref(), Some("Hello World"));
}

#[test]
fn parse_date_extra_whitespace() {
    // Date with extra whitespace between fields
    let raw = b"From: a@b.com\r\n\
                 Date:  Thu,  13  Feb  2025  15:47:33  +0000 \r\n\
                 \r\n";

    let parsed = parse_email(raw).unwrap();
    let date = parsed.date.unwrap();
    assert_eq!(date.year, 2025);
    assert_eq!(date.month, 2);
    assert_eq!(date.day, 13);
}

#[test]
fn parse_multipart_related_with_inline_images() {
    // multipart/related — common for HTML emails with inline images
    let raw = b"From: a@b.com\r\n\
                 Date: Thu, 13 Feb 2025 15:47:33 +0000\r\n\
                 Content-Type: multipart/related; boundary=\"rel\"\r\n\
                 \r\n\
                 --rel\r\n\
                 Content-Type: text/html\r\n\
                 \r\n\
                 <html><img src=\"cid:logo\"></html>\r\n\
                 --rel\r\n\
                 Content-Type: image/jpeg\r\n\
                 Content-ID: <logo>\r\n\
                 Content-Disposition: inline; filename=\"logo.jpg\"\r\n\
                 \r\n\
                 JPEG_DATA\r\n\
                 --rel--";

    let parsed = parse_email(raw).unwrap();
    assert!(parsed.body_html.is_some());
    assert_eq!(parsed.attachments.len(), 1);
    assert!(parsed.attachments[0].is_inline);
    assert_eq!(parsed.attachments[0].content_id.as_deref(), Some("logo"));
    assert_eq!(parsed.attachments[0].filename.as_deref(), Some("logo.jpg"));
}

/// RFC 2183 Sections 2.1 and 2.3: a `filename` parameter may appear on an
/// explicit `inline` body part, and that parameter does not turn the part
/// into an attachment.
#[test]
fn parse_single_part_text_inline_filename_stays_body_text() {
    let raw = b"From: a@b.com\r\n\
                 Date: Thu, 13 Feb 2025 15:47:33 +0000\r\n\
                 Content-Type: text/plain; charset=utf-8; name=\"notice.txt\"\r\n\
                 Content-Disposition: inline; filename=\"notice.txt\"\r\n\
                 \r\n\
                 Inline notice";

    let parsed = parse_email(raw).unwrap();
    assert_eq!(
        parsed.body_text.as_deref(),
        Some("Inline notice"),
        "RFC 2183 Sections 2.1 and 2.3: explicit inline text/plain with a filename hint must still populate body_text"
    );
    assert!(
        parsed.attachments.is_empty(),
        "explicit inline text/plain with a filename hint must not be reclassified as an attachment"
    );
}

/// RFC 2183 Sections 2.1 and 2.3: the presence of `filename` does not
/// override an explicit `inline` disposition inside multipart bodies.
#[test]
fn parse_multipart_inline_html_filename_stays_body_html() {
    let raw = b"From: a@b.com\r\n\
                 Date: Thu, 13 Feb 2025 15:47:33 +0000\r\n\
                 Content-Type: multipart/related; boundary=\"rel\"\r\n\
                 \r\n\
                 --rel\r\n\
                 Content-Type: text/html; charset=utf-8; name=\"body.html\"\r\n\
                 Content-Disposition: inline; filename=\"body.html\"\r\n\
                 \r\n\
                 <p>Inline html body</p>\r\n\
                 --rel--";

    let parsed = parse_email(raw).unwrap();
    assert_eq!(
        parsed.body_html.as_deref(),
        Some("<p>Inline html body</p>"),
        "RFC 2183 Sections 2.1 and 2.3: explicit inline text/html with a filename hint must still populate body_html"
    );
    assert!(
        parsed.attachments.is_empty(),
        "explicit inline text/html with a filename hint must not be reclassified as an attachment"
    );
}

#[test]
fn parse_minimal_message_from_only() {
    // Absolute minimum valid message: just From header
    let raw = b"From: a@b.com\r\n\r\n";

    let parsed = parse_email(raw).unwrap();
    assert_eq!(parsed.from[0].email, "a@b.com");
    assert!(parsed.subject.is_none());
    assert!(parsed.date.is_none());
    assert!(parsed.body_text.is_none());
}

#[test]
fn parse_multiple_same_headers() {
    // Multiple headers with the same name — should take the first one
    let raw = b"From: first@example.com\r\n\
                 From: second@example.com\r\n\
                 Subject: First\r\n\
                 Subject: Second\r\n\
                 Date: Thu, 13 Feb 2025 15:47:33 +0000\r\n\
                 \r\n";

    let parsed = parse_email(raw).unwrap();
    assert_eq!(parsed.from[0].email, "first@example.com");
    assert_eq!(parsed.subject.as_deref(), Some("First"));
}

#[test]
fn parse_date_all_named_timezones() {
    let test_cases = [
        ("EST", -300),
        ("EDT", -240),
        ("CST", -360),
        ("CDT", -300),
        ("MST", -420),
        ("MDT", -360),
        ("PST", -480),
        ("PDT", -420),
        ("GMT", 0),
        ("UTC", 0),
        ("UT", 0),
    ];
    for (tz_name, expected_offset) in test_cases {
        let raw = format!("From: a@b.com\r\nDate: Thu, 13 Feb 2025 12:00:00 {tz_name}\r\n\r\n");
        let parsed = parse_email(raw.as_bytes()).unwrap();
        let date = parsed.date.unwrap();
        assert_eq!(
            date.tz_offset_minutes, expected_offset,
            "Failed for timezone {tz_name}"
        );
    }
}

#[test]
fn parse_boundary_with_special_chars() {
    // Boundary containing special characters (RFC 2046 allows certain chars)
    let raw = b"From: a@b.com\r\n\
                 Date: Thu, 13 Feb 2025 15:47:33 +0000\r\n\
                 Content-Type: multipart/mixed; boundary=\"----=_Part_123+abc\"\r\n\
                 \r\n\
                 ------=_Part_123+abc\r\n\
                 Content-Type: text/plain\r\n\
                 \r\n\
                 Body text\r\n\
                 ------=_Part_123+abc--";

    let parsed = parse_email(raw).unwrap();
    assert_eq!(parsed.body_text.as_deref(), Some("Body text"));
}

#[test]
fn parse_truncated_base64_body() {
    // Truncated base64 — should not crash, fall back to raw bytes
    let raw = b"From: a@b.com\r\n\
                 Date: Thu, 13 Feb 2025 15:47:33 +0000\r\n\
                 Content-Type: text/plain; charset=utf-8\r\n\
                 Content-Transfer-Encoding: base64\r\n\
                 \r\n\
                 SGVsbG8gV29yb===invalid";

    let parsed = parse_email(raw).unwrap();
    // Should not panic — either decodes partial or falls back
    assert!(parsed.body_text.is_some());
}

#[test]
fn parse_address_group_syntax() {
    // Group address syntax: "Group: addr1, addr2;"
    // Should parse what it can without crashing
    let raw = b"From: sender@example.com\r\n\
                 To: Undisclosed:;\r\n\
                 Date: Thu, 13 Feb 2025 15:47:33 +0000\r\n\
                 \r\n";

    let parsed = parse_email(raw).unwrap();
    // Group syntax with no addresses — To list may be empty
    assert_eq!(parsed.from[0].email, "sender@example.com");
}

#[test]
fn parse_iso2022jp_encoded_word() {
    // ISO-2022-JP encoded word (common in Japanese email)
    // "テスト" in ISO-2022-JP base64
    let raw = b"From: a@b.com\r\n\
                 Subject: =?ISO-2022-JP?B?GyRCJUYlOSVIGyhC?=\r\n\
                 Date: Thu, 13 Feb 2025 15:47:33 +0000\r\n\
                 \r\n";

    let parsed = parse_email(raw).unwrap();
    // encoding_rs should handle ISO-2022-JP
    assert!(parsed.subject.is_some());
    assert_eq!(parsed.subject.as_deref(), Some("テスト"));
}

#[test]
fn parse_multipart_missing_parts_tolerance() {
    // Multipart header declared but body is completely different (partial fetch)
    let raw = b"From: a@b.com\r\n\
                 Date: Thu, 13 Feb 2025 15:47:33 +0000\r\n\
                 Content-Type: multipart/mixed; boundary=\"never-appears\"\r\n\
                 \r\n\
                 This body doesn't contain any boundaries at all.";

    let parsed = parse_email(raw).unwrap();
    // Should not fail and should preserve recoverable text instead of
    // discarding the body when the declared boundary never appears.
    assert_eq!(
        parsed.body_text.as_deref(),
        Some("This body doesn't contain any boundaries at all.")
    );
    assert!(parsed.body_html.is_none());
    assert!(parsed.attachments.is_empty());
}

#[test]
fn parse_nested_multipart_without_boundary_falls_back_to_text() {
    // RFC 2046 Section 5.1.1 requires a boundary parameter for multipart
    // parts. For malformed nested parts, apply the same best-effort
    // fallback used at the top level so text content is not silently lost.
    let raw = b"From: a@b.com\r\n\
                 Date: Thu, 13 Feb 2025 15:47:33 +0000\r\n\
                 Content-Type: multipart/mixed; boundary=\"outer\"\r\n\
                 \r\n\
                 --outer\r\n\
                 Content-Type: multipart/alternative\r\n\
                 \r\n\
                 Inner text that should not disappear.\r\n\
                 --outer--\r\n";

    let parsed = parse_email(raw).unwrap();
    assert_eq!(
        parsed.body_text.as_deref(),
        Some("Inner text that should not disappear."),
        "malformed nested multipart parts should fall back to simple text"
    );
}

#[test]
fn parse_nested_multipart_without_boundary_preserves_outer_section_number() {
    // RFC 2046 Section 5.1.1 requires a boundary parameter for multipart
    // parts. When recovering a malformed nested multipart as a simple part,
    // attachment metadata should still use the outer part's IMAP section
    // number rather than hardcoding the top-level "1".
    let raw = b"From: a@b.com\r\n\
                 Date: Thu, 13 Feb 2025 15:47:33 +0000\r\n\
                 Content-Type: multipart/mixed; boundary=\"outer\"\r\n\
                 \r\n\
                 --outer\r\n\
                 Content-Type: text/plain\r\n\
                 \r\n\
                 First part.\r\n\
                 --outer\r\n\
                 Content-Type: multipart/mixed\r\n\
                 Content-Disposition: attachment; filename=\"nested.txt\"\r\n\
                 \r\n\
                 Recovered nested payload.\r\n\
                 --outer--\r\n";

    let parsed = parse_email(raw).unwrap();
    assert_eq!(parsed.attachments.len(), 1);
    assert_eq!(
        parsed.attachments[0].section.as_deref(),
        Some("2"),
        "recovered malformed nested multipart should keep its outer section number"
    );
}

#[test]
fn parse_encoded_word_in_multiple_header_types() {
    // Encoded words in From display name AND Subject
    let raw = b"From: =?UTF-8?Q?M=C3=BCller?= <mueller@example.com>\r\n\
                 To: =?UTF-8?B?U21pdGg=?= <smith@example.com>\r\n\
                 Subject: =?UTF-8?Q?Caf=C3=A9?=\r\n\
                 Date: Thu, 13 Feb 2025 15:47:33 +0000\r\n\
                 \r\n";

    let parsed = parse_email(raw).unwrap();
    assert_eq!(parsed.from[0].name.as_deref(), Some("Müller"));
    assert_eq!(parsed.to[0].name.as_deref(), Some("Smith"));
    assert_eq!(parsed.subject.as_deref(), Some("Café"));
}

#[test]
fn parse_attachment_size_reflects_part_body() {
    // Verify attachment size field is set correctly
    let raw = b"From: a@b.com\r\n\
                 Date: Thu, 13 Feb 2025 15:47:33 +0000\r\n\
                 Content-Type: multipart/mixed; boundary=\"b\"\r\n\
                 \r\n\
                 --b\r\n\
                 Content-Type: text/plain\r\n\
                 \r\n\
                 Body\r\n\
                 --b\r\n\
                 Content-Type: application/pdf\r\n\
                 Content-Disposition: attachment; filename=\"f.pdf\"\r\n\
                 \r\n\
                 0123456789\r\n\
                 --b--";

    let parsed = parse_email(raw).unwrap();
    assert_eq!(parsed.attachments.len(), 1);
    assert_eq!(parsed.attachments[0].size, Some(10));
}

#[test]
fn parse_unquoted_boundary() {
    // Boundary value without quotes (RFC 2046 allows this)
    let raw = b"From: a@b.com\r\n\
                 Date: Thu, 13 Feb 2025 15:47:33 +0000\r\n\
                 Content-Type: multipart/mixed; boundary=simple_boundary\r\n\
                 \r\n\
                 --simple_boundary\r\n\
                 Content-Type: text/plain\r\n\
                 \r\n\
                 Text\r\n\
                 --simple_boundary--";

    let parsed = parse_email(raw).unwrap();
    assert_eq!(parsed.body_text.as_deref(), Some("Text"));
}

#[test]
fn parse_boundary_ignores_stray_bare_word_after_token() {
    // RFC 2046 Section 5.1.1: boundary is a single MIME token. If a
    // malformed sender appends stray bare text after it, recovery should
    // still honor the actual delimiter that appears in the body.
    let raw = b"From: a@b.com\r\n\
                 Date: Thu, 13 Feb 2025 15:47:33 +0000\r\n\
                 Content-Type: multipart/mixed; boundary=actual unexpected\r\n\
                 \r\n\
                 --actual\r\n\
                 Content-Type: text/plain\r\n\
                 \r\n\
                 Hello\r\n\
                 --actual--";

    let parsed = parse_email(raw).unwrap();
    assert_eq!(
        parsed.body_text.as_deref(),
        Some("Hello"),
        "RFC 2046 Section 5.1.1: stray bare text after boundary token must not \
         prevent multipart parsing when the actual delimiter is still unambiguous"
    );
}

#[test]
fn parse_message_id_without_angle_brackets() {
    // Some broken mailers omit angle brackets on Message-ID.
    // The parser tolerates this and returns the bare value.
    let raw = b"From: a@b.com\r\n\
                 Message-ID: bare-id@host.com\r\n\
                 Date: Thu, 13 Feb 2025 15:47:33 +0000\r\n\
                 \r\n";

    let parsed = parse_email(raw).unwrap();
    assert_eq!(parsed.message_id.as_deref(), Some("bare-id@host.com"));
}

#[test]
fn parse_message_id_without_angle_brackets_requires_msg_id_syntax() {
    // Postel fallback for missing angle brackets must still require a
    // syntactically valid bare msg-id body (RFC 5322 Section 3.6.4).
    let raw = b"From: a@b.com\r\n\
                 Message-ID: not-a-msg-id\r\n\
                 Date: Thu, 13 Feb 2025 15:47:33 +0000\r\n\
                 \r\n";

    let parsed = parse_email(raw).unwrap();
    assert!(
        parsed.message_id.is_none(),
        "invalid bare Message-ID fallback must be ignored, got {:?}",
        parsed.message_id
    );
}

#[test]
fn parse_message_id_with_invalid_bracketed_syntax_is_ignored() {
    // RFC 5322 Section 3.6.4: the text inside angle brackets must still
    // be a syntactically valid msg-id body. Brackets alone do not make
    // an invalid identifier valid.
    let raw = b"From: a@b.com\r\n\
                 Message-ID: <not a msg-id>\r\n\
                 Date: Thu, 13 Feb 2025 15:47:33 +0000\r\n\
                 \r\n";

    let parsed = parse_email(raw).unwrap();
    assert!(
        parsed.message_id.is_none(),
        "invalid bracketed Message-ID must be ignored, got {:?}",
        parsed.message_id
    );
}

#[test]
fn parse_message_id_with_quoted_id_left() {
    // RFC 5322 Section 4.5.4: parsers may recover obsolete `obs-id-left`
    // forms, including quoted local-parts, from inbound Message-ID fields.
    let raw = b"From: a@b.com\r\n\
                 Message-ID: <\"user@inner\"@example.com>\r\n\
                 Date: Thu, 13 Feb 2025 15:47:33 +0000\r\n\
                 \r\n";

    let parsed = parse_email(raw).unwrap();
    assert_eq!(
        parsed.message_id.as_deref(),
        Some("\"user@inner\"@example.com")
    );
}

#[test]
fn parse_message_id_with_obs_cfws_between_atoms() {
    // RFC 5322 Section 4.5.4: `obs-id-left = local-part` and
    // `obs-id-right = domain`, allowing CFWS to appear between atoms.
    // The CFWS is not semantically part of the msg-id and should be
    // ignored during interpretation.
    let raw = b"From: a@b.com\r\n\
                 Message-ID: <foo . bar@example . com>\r\n\
                 Date: Thu, 13 Feb 2025 15:47:33 +0000\r\n\
                 \r\n";

    let parsed = parse_email(raw).unwrap();
    assert_eq!(
        parsed.message_id.as_deref(),
        Some("foo.bar@example.com"),
        "obsolete CFWS inside a bracketed Message-ID must be normalized \
         per RFC 5322 Section 4.5.4"
    );
}

#[test]
fn parse_in_reply_to_without_angle_brackets() {
    // Some broken mailers omit angle brackets on In-Reply-To as well.
    // The parser should recover the bare msg-id bodies so thread
    // reconstruction is still possible.
    let raw = b"From: a@b.com\r\n\
                 In-Reply-To: first@host second@[127.0.0.1]\r\n\
                 Date: Thu, 13 Feb 2025 15:47:33 +0000\r\n\
                 \r\n";

    let parsed = parse_email(raw).unwrap();
    assert_eq!(parsed.in_reply_to, vec!["first@host", "second@[127.0.0.1]"]);
}

#[test]
fn parse_references_without_angle_brackets() {
    // Some broken mailers omit angle brackets on References.
    // Preserve all recoverable bare msg-id bodies in order.
    let raw = b"From: a@b.com\r\n\
                 References: ref1@host ref2@[IPv6:2001:db8::1]\r\n\
                 Date: Thu, 13 Feb 2025 15:47:33 +0000\r\n\
                 \r\n";

    let parsed = parse_email(raw).unwrap();
    assert_eq!(
        parsed.references,
        vec!["ref1@host", "ref2@[IPv6:2001:db8::1]"]
    );
}

#[test]
fn parse_references_filters_invalid_bracketed_msg_ids() {
    // RFC 5322 Section 3.6.4: each bracketed token in References must be
    // a valid msg-id body. Invalid `<...>` tokens must not be surfaced in
    // the parsed threading data.
    let raw = b"From: a@b.com\r\n\
                 References: <good@example.com> <not a msg-id> <also.good@[127.0.0.1]>\r\n\
                 Date: Thu, 13 Feb 2025 15:47:33 +0000\r\n\
                 \r\n";

    let parsed = parse_email(raw).unwrap();
    assert_eq!(
        parsed.references,
        vec!["good@example.com", "also.good@[127.0.0.1]"]
    );
}

#[test]
fn parse_empty_references_header() {
    // References header with no valid message-ids
    let raw = b"From: a@b.com\r\n\
                 References: \r\n\
                 Date: Thu, 13 Feb 2025 15:47:33 +0000\r\n\
                 \r\n";

    let parsed = parse_email(raw).unwrap();
    assert!(parsed.references.is_empty());
}

#[test]
fn parse_large_multipart_many_attachments() {
    // Message with many attachment parts — verify section numbering
    let mut raw = b"From: a@b.com\r\nDate: Thu, 13 Feb 2025 15:47:33 +0000\r\n\
                     Content-Type: multipart/mixed; boundary=\"multi\"\r\n\r\n"
        .to_vec();

    raw.extend_from_slice(b"--multi\r\nContent-Type: text/plain\r\n\r\nBody\r\n");
    for i in 1..=5 {
        raw.extend_from_slice(
            format!(
                "--multi\r\nContent-Type: application/octet-stream\r\n\
                 Content-Disposition: attachment; filename=\"file{i}.bin\"\r\n\r\n\
                 DATA{i}\r\n"
            )
            .as_bytes(),
        );
    }
    raw.extend_from_slice(b"--multi--");

    let parsed = parse_email(&raw).unwrap();
    assert_eq!(parsed.body_text.as_deref(), Some("Body"));
    assert_eq!(parsed.attachments.len(), 5);
    for (i, att) in parsed.attachments.iter().enumerate() {
        assert_eq!(
            att.section.as_deref(),
            Some(&(i + 2).to_string() as &str),
            "Wrong section for attachment {i}"
        );
        assert_eq!(
            att.filename.as_deref(),
            Some(&format!("file{}.bin", i + 1) as &str)
        );
    }
}

#[test]
fn parse_message_id_empty_brackets() {
    // Empty angle brackets should return None
    let raw = b"From: a@b.com\r\n\
                 Message-ID: <>\r\n\
                 Date: Thu, 13 Feb 2025 15:47:33 +0000\r\n\
                 \r\n";

    let parsed = parse_email(raw).unwrap();
    assert!(parsed.message_id.is_none());
}

#[test]
fn parse_message_id_empty_value() {
    // Completely empty Message-ID value should return None
    let raw = b"From: a@b.com\r\n\
                 Message-ID: \r\n\
                 Date: Thu, 13 Feb 2025 15:47:33 +0000\r\n\
                 \r\n";

    let parsed = parse_email(raw).unwrap();
    assert!(parsed.message_id.is_none());
}

#[test]
fn rfc2231_param_boundary_check() {
    // The "filename*=" match must be at a parameter boundary,
    // not embedded in another parameter name like "xfilename*="
    let disposition = "attachment; xfilename*=UTF-8''bad.pdf; filename*=UTF-8''good.pdf";
    let result = extract_rfc2231_param(disposition, "filename");
    assert_eq!(result.as_deref(), Some("good.pdf"));
}

#[test]
fn rfc2231_param_at_start() {
    // Parameter at the very start of the value (no preceding `;`)
    let value = "filename*=UTF-8''test.pdf";
    let result = extract_rfc2231_param(value, "filename");
    assert_eq!(result.as_deref(), Some("test.pdf"));
}

#[test]
fn parse_quoted_display_name_with_comma() {
    // Display name with comma must be in a quoted-string (RFC 5322 Section 3.4)
    let raw = b"From: \"Doe, John\" <john@example.com>\r\n\
                 Date: Thu, 13 Feb 2025 15:47:33 +0000\r\n\
                 \r\n";

    let parsed = parse_email(raw).unwrap();
    assert_eq!(parsed.from[0].name.as_deref(), Some("Doe, John"));
    assert_eq!(parsed.from[0].email, "john@example.com");
}

#[test]
fn parse_quoted_display_name_with_escaped_chars() {
    // Backslash-escaped characters in quoted display name (RFC 5322 Section 3.2.4)
    let raw = b"From: \"John \\\"Doc\\\" Doe\" <john@example.com>\r\n\
                 Date: Thu, 13 Feb 2025 15:47:33 +0000\r\n\
                 \r\n";

    let parsed = parse_email(raw).unwrap();
    assert_eq!(parsed.from[0].name.as_deref(), Some("John \"Doc\" Doe"));
    assert_eq!(parsed.from[0].email, "john@example.com");
}

#[test]
fn unescape_quoted_string_backslash() {
    assert_eq!(unescape_quoted_string("hello"), "hello");
    assert_eq!(unescape_quoted_string("a\\\\b"), "a\\b");
    assert_eq!(unescape_quoted_string("a\\\"b"), "a\"b");
    assert_eq!(unescape_quoted_string("trailing\\"), "trailing\\");
}

#[test]
fn parse_address_list_with_escaped_quotes_in_display_name() {
    // RFC 5322 Section 3.2.4: backslash-escaped quotes within a quoted-string
    // must not prematurely end the quoted context during address list splitting.
    let raw = b"From: a@b.com\r\n\
                 To: \"A\\\"B\" <a@x.com>, c@d.com\r\n\
                 Date: Thu, 13 Feb 2025 15:47:33 +0000\r\n\
                 \r\n";

    let parsed = parse_email(raw).unwrap();
    assert_eq!(
        parsed.to.len(),
        2,
        "Expected 2 To addresses but got {:?}",
        parsed.to
    );
    assert_eq!(parsed.to[0].email, "a@x.com");
    assert_eq!(parsed.to[0].name.as_deref(), Some("A\"B"));
    assert_eq!(parsed.to[1].email, "c@d.com");
}

#[test]
fn parse_rfc2231_continuation_filename() {
    // RFC 2231 Section 3: long filenames split across continuation parameters.
    // filename*0="very_long_"; filename*1="filename.pdf"
    let raw = b"From: a@b.com\r\n\
                 Date: Thu, 13 Feb 2025 15:47:33 +0000\r\n\
                 Content-Type: multipart/mixed; boundary=\"b\"\r\n\
                 \r\n\
                 --b\r\n\
                 Content-Type: text/plain\r\n\
                 \r\n\
                 Body\r\n\
                 --b\r\n\
                 Content-Type: application/pdf\r\n\
                 Content-Disposition: attachment; filename*0=\"very_long_\"; filename*1=\"filename.pdf\"\r\n\
                 \r\n\
                 DATA\r\n\
                 --b--";

    let parsed = parse_email(raw).unwrap();
    assert_eq!(parsed.attachments.len(), 1);
    assert_eq!(
        parsed.attachments[0].filename.as_deref(),
        Some("very_long_filename.pdf"),
        "RFC 2231 continuation filename not reassembled"
    );
}

#[test]
fn parse_rfc2231_continuation_with_charset() {
    // RFC 2231 Section 3+4: continuation with charset encoding.
    // filename*0*=UTF-8''r%C3%A9sum; filename*1*=%C3%A9.pdf
    let raw = b"From: a@b.com\r\n\
                 Date: Thu, 13 Feb 2025 15:47:33 +0000\r\n\
                 Content-Type: multipart/mixed; boundary=\"b\"\r\n\
                 \r\n\
                 --b\r\n\
                 Content-Type: text/plain\r\n\
                 \r\n\
                 Body\r\n\
                 --b\r\n\
                 Content-Type: application/pdf\r\n\
                 Content-Disposition: attachment; filename*0*=UTF-8''r%C3%A9sum; filename*1*=%C3%A9.pdf\r\n\
                 \r\n\
                 DATA\r\n\
                 --b--";

    let parsed = parse_email(raw).unwrap();
    assert_eq!(parsed.attachments.len(), 1);
    assert_eq!(
        parsed.attachments[0].filename.as_deref(),
        Some("résumé.pdf"),
        "RFC 2231 continuation with charset not reassembled"
    );
}

/// commas inside RFC 5322 Section 3.2.2 parenthesized comments
/// must NOT split the address list. The comment `(Doe, John)` is a single
/// unit — the comma inside it is not an address separator.
#[test]
fn parse_address_comment_with_comma() {
    let raw = b"From: sender@example.com\r\n\
                 To: user@example.com (Doe, John), other@example.com\r\n\
                 Date: Thu, 13 Feb 2025 15:47:33 +0000\r\n\
                 \r\n";

    let parsed = parse_email(raw).unwrap();
    assert_eq!(
        parsed.to.len(),
        2,
        "Expected 2 To addresses but got {:?}",
        parsed.to
    );
    assert_eq!(parsed.to[0].email, "user@example.com");
    assert_eq!(
        parsed.to[0].name.as_deref(),
        Some("Doe, John"),
        "Display name from comment should be preserved intact"
    );
    assert_eq!(parsed.to[1].email, "other@example.com");
}

#[test]
fn parse_header_unfolding_preserves_wsp() {
    // RFC 5322 Section 2.2.3: unfolding removes the CRLF but the leading
    // WSP (tab or space) is part of the FWS and must be preserved.
    let raw = b"From: a@b.com\r\nSubject: Hello\r\n\tWorld\r\nDate: Thu, 13 Feb 2025 15:47:33 +0000\r\n\r\n";

    let parsed = parse_email(raw).unwrap();
    // The tab should be preserved between "Hello" and "World"
    assert_eq!(
        parsed.subject.as_deref(),
        Some("Hello\tWorld"),
        "Tab from continuation line should be preserved per RFC 5322 Section 2.2.3"
    );
}

#[test]
fn datetime_to_unix_timestamp() {
    use crate::types::DateTime;

    // 2025-02-13 15:47:33 +0000 → known Unix timestamp
    let dt = DateTime {
        year: 2025,
        month: 2,
        day: 13,
        hour: 15,
        minute: 47,
        second: 33,
        tz_offset_minutes: 0,
    };
    // 2025-02-13T15:47:33Z = 1739461653
    assert_eq!(dt.to_unix_timestamp(), 1_739_461_653);

    // Same instant with +0530 offset (local 21:17:33)
    let dt_offset = DateTime {
        year: 2025,
        month: 2,
        day: 13,
        hour: 21,
        minute: 17,
        second: 33,
        tz_offset_minutes: 330,
    };
    assert_eq!(dt_offset.to_unix_timestamp(), dt.to_unix_timestamp());
}

#[test]
fn datetime_from_unix_timestamp() {
    use crate::types::DateTime;

    let ts = 1_739_461_653_i64; // 2025-02-13T15:47:33Z
    let dt = DateTime::from_unix_timestamp(ts, 0);
    assert_eq!(dt.year, 2025);
    assert_eq!(dt.month, 2);
    assert_eq!(dt.day, 13);
    assert_eq!(dt.hour, 15);
    assert_eq!(dt.minute, 47);
    assert_eq!(dt.second, 33);
    assert_eq!(dt.tz_offset_minutes, 0);

    // With +0530 offset
    let dt_offset = DateTime::from_unix_timestamp(ts, 330);
    assert_eq!(dt_offset.hour, 21);
    assert_eq!(dt_offset.minute, 17);
}

#[test]
fn datetime_round_trip_timestamp() {
    use crate::types::DateTime;

    let dt = DateTime {
        year: 2025,
        month: 12,
        day: 31,
        hour: 23,
        minute: 59,
        second: 59,
        tz_offset_minutes: -480,
    };
    let ts = dt.to_unix_timestamp();
    let restored = DateTime::from_unix_timestamp(ts, -480);
    assert_eq!(dt, restored);
}

#[test]
fn datetime_ord_comparison() {
    use crate::types::DateTime;

    // Same instant in different timezones should be equal
    let utc = DateTime {
        year: 2025,
        month: 1,
        day: 1,
        hour: 12,
        minute: 0,
        second: 0,
        tz_offset_minutes: 0,
    };
    let est = DateTime {
        year: 2025,
        month: 1,
        day: 1,
        hour: 7,
        minute: 0,
        second: 0,
        tz_offset_minutes: -300,
    };
    assert_eq!(utc.cmp(&est), std::cmp::Ordering::Equal);

    // Later timestamp should be greater
    let later = DateTime {
        year: 2025,
        month: 1,
        day: 1,
        hour: 13,
        minute: 0,
        second: 0,
        tz_offset_minutes: 0,
    };
    assert!(later > utc);
}

#[test]
fn datetime_epoch() {
    use crate::types::DateTime;

    let epoch = DateTime::from_unix_timestamp(0, 0);
    assert_eq!(epoch.year, 1970);
    assert_eq!(epoch.month, 1);
    assert_eq!(epoch.day, 1);
    assert_eq!(epoch.hour, 0);
    assert_eq!(epoch.minute, 0);
    assert_eq!(epoch.second, 0);
    assert_eq!(epoch.to_unix_timestamp(), 0);
}

#[test]
fn parse_headers_only_extracts_metadata() {
    let raw = b"From: sender@example.com\r\n\
                 To: recipient@example.com\r\n\
                 Subject: Test\r\n\
                 Date: Thu, 13 Feb 2025 15:47:33 +0000\r\n\
                 Message-ID: <abc123@example.com>\r\n\
                 Content-Type: multipart/mixed; boundary=\"b\"\r\n\
                 \r\n\
                 --b\r\n\
                 Content-Type: text/plain\r\n\
                 \r\n\
                 This body should NOT be parsed\r\n\
                 --b\r\n\
                 Content-Type: application/pdf\r\n\
                 Content-Disposition: attachment; filename=\"doc.pdf\"\r\n\
                 \r\n\
                 PDF_DATA\r\n\
                 --b--";

    let parsed = parse_headers_only(raw).unwrap();

    // Header fields should be populated
    assert_eq!(parsed.from[0].email, "sender@example.com");
    assert_eq!(parsed.to.len(), 1);
    assert_eq!(parsed.subject.as_deref(), Some("Test"));
    assert_eq!(parsed.message_id.as_deref(), Some("abc123@example.com"));
    assert!(parsed.date.is_some());

    // Body fields should be empty (skipped)
    assert!(parsed.body_text.is_none());
    assert!(parsed.body_html.is_none());
    assert!(parsed.attachments.is_empty());
}

#[test]
fn parse_headers_only_empty_input() {
    let result = parse_headers_only(b"");
    assert!(matches!(result, Err(Error::EmptyInput)));
}

#[test]
fn parse_headers_only_missing_from() {
    let raw = b"Subject: No From\r\nDate: Thu, 13 Feb 2025 15:47:33 +0000\r\n\r\n";
    let parsed = parse_headers_only(raw).expect(
        "headers-only parsing should tolerate missing From when the header block \
         is otherwise usable",
    );
    assert!(
        parsed.from.is_empty(),
        "missing From should produce an empty from list, got {:?}",
        parsed.from
    );
    assert_eq!(parsed.subject.as_deref(), Some("No From"));
}

/// `parse_headers_only` should also tolerate malformed messages that have
/// no header fields at all, returning an empty header view instead of a
/// structural error.
#[test]
fn parse_headers_only_body_only_message_without_headers() {
    let raw = b"Body-only message with no header section";

    let parsed = parse_headers_only(raw).expect(
        "headers-only parsing should accept body-only malformed messages so \
         callers can still inspect size and raw header state",
    );
    assert!(parsed.from.is_empty());
    assert!(parsed.raw_headers.is_empty());
    assert!(parsed.subject.is_none());
    assert!(parsed.body_text.is_none());
    assert!(parsed.body_html.is_none());
    assert!(parsed.attachments.is_empty());
    assert_eq!(parsed.size, raw.len() as u64);
}

/// Headers-only parsing should also tolerate header blocks that contain
/// only optional/custom fields (RFC 5322 Section 3.6.8).
#[test]
fn parse_headers_only_custom_headers_without_well_known_headers() {
    let raw = b"X-Trace: 12345\r\nList-Id: Example List <list.example>\r\n\r\n";
    let parsed = parse_headers_only(raw).expect(
        "headers-only parsing should accept header blocks containing only \
         optional fields",
    );
    assert!(parsed.from.is_empty());
    assert!(parsed.subject.is_none());
    assert!(
        parsed
            .extra_headers
            .iter()
            .any(|(name, value)| name == "x-trace" && value == "12345"),
        "X-Trace must be preserved in extra_headers"
    );
    assert!(
        parsed
            .extra_headers
            .iter()
            .any(|(name, value)| name == "list-id" && value == "Example List <list.example>"),
        "List-Id must be preserved in extra_headers"
    );
}

/// RFC 5322 Section 2.2.3: when a field body begins only on the first
/// folded continuation line, unfolding removes the CRLF and exactly one
/// leading SP/HTAB serves only as the structural separator before the
/// field body.
#[test]
fn parse_headers_only_extra_header_starting_on_continuation_strips_structural_wsp() {
    let raw = b"From: sender@example.com\r\nDate: Thu, 13 Feb 2025 15:47:33 +0000\r\nX-Note:\r\n hello\r\nX-Indent:\r\n  value\r\n\r\n";

    let parsed = parse_headers_only(raw).unwrap();
    assert!(
        parsed
            .extra_headers
            .iter()
            .any(|(name, value)| name == "x-note" && value == "hello"),
        "first continuation SP must be treated as structural, got {:?}",
        parsed.extra_headers
    );
    assert!(
        parsed
            .extra_headers
            .iter()
            .any(|(name, value)| name == "x-indent" && value == " value"),
        "only one structural SP should be stripped from the first continuation line, got {:?}",
        parsed.extra_headers
    );
}

// -----------------------------------------------------------------------
// extract_param quoted-string handling (RFC 5322 §3.2.4)
// -----------------------------------------------------------------------

#[test]
fn extract_param_unescapes_backslash_in_filename() {
    // RFC 5322 Section 3.2.4: quoted-pair `\\` in a quoted-string represents
    // a literal backslash. extract_param must unescape it.
    let raw = b"From: a@b.com\r\n\
                 Date: Thu, 13 Feb 2025 15:47:33 +0000\r\n\
                 Content-Type: multipart/mixed; boundary=\"b\"\r\n\
                 \r\n\
                 --b\r\n\
                 Content-Type: text/plain\r\n\
                 \r\n\
                 Body\r\n\
                 --b\r\n\
                 Content-Type: application/pdf\r\n\
                 Content-Disposition: attachment; filename=\"path\\\\file.pdf\"\r\n\
                 \r\n\
                 DATA\r\n\
                 --b--";

    let parsed = parse_email(raw).unwrap();
    assert_eq!(parsed.attachments.len(), 1);
    // The filename in the message is `path\\file.pdf` (escaped backslash).
    // After unescaping per RFC 5322 Section 3.2.4, it should be `path\file.pdf`.
    assert_eq!(
        parsed.attachments[0].filename.as_deref(),
        Some("path\\file.pdf"),
        "Backslash in quoted-string filename must be unescaped per RFC 5322 Section 3.2.4"
    );
}

#[test]
fn extract_param_handles_escaped_quote_in_filename() {
    // RFC 5322 Section 3.2.4: quoted-pair `\"` in a quoted-string represents
    // a literal double-quote. extract_param must skip escaped quotes when
    // finding the closing quote, and then unescape the result.
    let raw = b"From: a@b.com\r\n\
                 Date: Thu, 13 Feb 2025 15:47:33 +0000\r\n\
                 Content-Type: multipart/mixed; boundary=\"b\"\r\n\
                 \r\n\
                 --b\r\n\
                 Content-Type: text/plain\r\n\
                 \r\n\
                 Body\r\n\
                 --b\r\n\
                 Content-Type: application/pdf\r\n\
                 Content-Disposition: attachment; filename=\"file\\\"name.pdf\"\r\n\
                 \r\n\
                 DATA\r\n\
                 --b--";

    let parsed = parse_email(raw).unwrap();
    assert_eq!(parsed.attachments.len(), 1);
    // The filename in the message is `file\"name.pdf` (escaped quote).
    // After unescaping, it should be `file"name.pdf`.
    assert_eq!(
        parsed.attachments[0].filename.as_deref(),
        Some("file\"name.pdf"),
        "Escaped quote in quoted-string filename must be handled per RFC 5322 Section 3.2.4"
    );
}

#[test]
fn build_then_parse_filename_with_backslash_round_trip() {
    // Round-trip: build a message with a backslash in the filename, then parse it.
    // The builder escapes `\` → `\\` per RFC 5322 Section 3.2.4.
    // The parser must unescape `\\` → `\` to recover the original filename.
    let email = crate::types::OutgoingEmail {
        from: vec![Address {
            name: None,
            email: "a@b.com".into(),
        }],
        sender: None,
        to: vec![Address {
            name: None,
            email: "to@b.com".into(),
        }],
        cc: vec![],
        bcc: vec![],
        reply_to: vec![],
        date: None,
        subject: "test".into(),
        body_text: Some("Body".into()),
        body_html: None,
        in_reply_to: vec![],
        references: vec![],
        attachments: vec![crate::types::OutgoingAttachment {
            filename: "path\\file.pdf".into(),
            content_type: "application/pdf".into(),
            data: b"data".to_vec(),
            is_inline: false,
            content_id: None,
        }],
        extra_headers: vec![],
    };

    let built = crate::build_message(&email).unwrap();
    let parsed = parse_email(&built.raw).unwrap();

    assert_eq!(parsed.attachments.len(), 1);
    assert_eq!(
        parsed.attachments[0].filename.as_deref(),
        Some("path\\file.pdf"),
        "Round-trip filename with backslash must be preserved"
    );
}

#[test]
fn build_then_parse_filename_with_quote_round_trip() {
    // Round-trip: build a message with a double-quote in the filename.
    let email = crate::types::OutgoingEmail {
        from: vec![Address {
            name: None,
            email: "a@b.com".into(),
        }],
        sender: None,
        to: vec![Address {
            name: None,
            email: "to@b.com".into(),
        }],
        cc: vec![],
        bcc: vec![],
        reply_to: vec![],
        date: None,
        subject: "test".into(),
        body_text: Some("Body".into()),
        body_html: None,
        in_reply_to: vec![],
        references: vec![],
        attachments: vec![crate::types::OutgoingAttachment {
            filename: "file\"name.pdf".into(),
            content_type: "application/pdf".into(),
            data: b"data".to_vec(),
            is_inline: false,
            content_id: None,
        }],
        extra_headers: vec![],
    };

    let built = crate::build_message(&email).unwrap();
    let parsed = parse_email(&built.raw).unwrap();

    assert_eq!(parsed.attachments.len(), 1);
    assert_eq!(
        parsed.attachments[0].filename.as_deref(),
        Some("file\"name.pdf"),
        "Round-trip filename with double-quote must be preserved"
    );
}

// -----------------------------------------------------------------------
// to_ascii_lowercase for non-ASCII safety (RFC 6532)
// -----------------------------------------------------------------------

#[test]
fn extract_param_with_non_ascii_before_param() {
    // RFC 6532 allows UTF-8 directly in headers. If a non-ASCII character
    // that changes byte length under Unicode lowercasing appears before
    // the parameter we're searching for, to_lowercase() would misalign
    // byte offsets. to_ascii_lowercase() preserves byte length.
    //
    // İ (U+0130, 2 bytes UTF-8) lowercases to 'i' + combining dot (3 bytes)
    // under full Unicode rules, but stays 2 bytes under ASCII-only rules.
    //
    // We test extract_param directly with İ before the target parameter.
    let header_value = "attachment; description=\"\u{0130}stanbul\"; filename=\"report.pdf\"";
    let result = extract_param(header_value, "filename");
    assert_eq!(
        result.as_deref(),
        Some("report.pdf"),
        "extract_param must work when non-ASCII chars that change byte length \
         under Unicode lowercasing appear before the target parameter (RFC 6532)"
    );
}

// -----------------------------------------------------------------------
// date field range validation (RFC 5322 Section 3.3)
// -----------------------------------------------------------------------

#[test]
fn parse_date_rejects_invalid_hour() {
    // RFC 5322 Section 3.3: hour is 0-23.
    let raw = b"From: a@b.com\r\n\
                 Date: Thu, 13 Feb 2025 25:00:00 +0000\r\n\
                 \r\n";
    let parsed = parse_email(raw).unwrap();
    // Invalid hour (25) should cause date parsing to return None
    assert!(
        parsed.date.is_none(),
        "Date with hour=25 should be rejected per RFC 5322 Section 3.3"
    );
}

#[test]
fn parse_date_rejects_invalid_minute() {
    // RFC 5322 Section 3.3: minute is 0-59.
    let raw = b"From: a@b.com\r\n\
                 Date: Thu, 13 Feb 2025 12:60:00 +0000\r\n\
                 \r\n";
    let parsed = parse_email(raw).unwrap();
    assert!(
        parsed.date.is_none(),
        "Date with minute=60 should be rejected per RFC 5322 Section 3.3"
    );
}

#[test]
fn parse_date_rejects_invalid_second() {
    // RFC 5322 Section 3.3: second is 0-60 (60 for leap second).
    let raw = b"From: a@b.com\r\n\
                 Date: Thu, 13 Feb 2025 12:00:61 +0000\r\n\
                 \r\n";
    let parsed = parse_email(raw).unwrap();
    assert!(
        parsed.date.is_none(),
        "Date with second=61 should be rejected per RFC 5322 Section 3.3"
    );
}

#[test]
fn parse_date_rejects_invalid_day() {
    // RFC 5322 Section 3.3: day is 1-31.
    let raw = b"From: a@b.com\r\n\
                 Date: Thu, 32 Feb 2025 12:00:00 +0000\r\n\
                 \r\n";
    let parsed = parse_email(raw).unwrap();
    assert!(
        parsed.date.is_none(),
        "Date with day=32 should be rejected per RFC 5322 Section 3.3"
    );
}

#[test]
fn parse_date_rejects_day_zero() {
    // RFC 5322 Section 3.3: day starts at 1.
    let raw = b"From: a@b.com\r\n\
                 Date: Thu, 00 Feb 2025 12:00:00 +0000\r\n\
                 \r\n";
    let parsed = parse_email(raw).unwrap();
    assert!(
        parsed.date.is_none(),
        "Date with day=0 should be rejected per RFC 5322 Section 3.3"
    );
}

#[test]
fn parse_date_allows_leap_second() {
    // RFC 5322 Section 3.3: second 60 is valid (leap second).
    let raw = b"From: a@b.com\r\n\
                 Date: Tue, 30 Jun 2015 23:59:60 +0000\r\n\
                 \r\n";
    let parsed = parse_email(raw).unwrap();
    let date = parsed.date.expect("Leap second (60) should be accepted");
    assert_eq!(date.second, 60);
}

#[test]
fn parse_date_accepts_calendar_invalid_but_syntactically_valid_day() {
    // RFC 5322 Section 3.3 grammar: day = ([FWS] 1*2DIGIT FWS).
    // The grammar allows any day value 1-31 regardless of month.
    // Postel's law: accept calendar-impossible dates like Feb 31 since they
    // are syntactically valid and appear in real-world email.
    let raw = b"From: a@b.com\r\n\
                 Date: Wed, 31 Feb 2025 12:00:00 +0000\r\n\
                 \r\n";
    let parsed = parse_email(raw).unwrap();
    let date = parsed
        .date
        .expect("Calendar-invalid but syntactically valid date (31 Feb) must be accepted");
    assert_eq!(date.day, 31);
    assert_eq!(date.month, 2);
    assert_eq!(date.year, 2025);

    // Also verify 31 Apr, 31 Jun, and 29 Feb in non-leap years
    let dt =
        parse_rfc5322_date("Mon, 31 Apr 2025 12:00:00 +0000").expect("31 Apr must be accepted");
    assert_eq!(dt.day, 31);
    assert_eq!(dt.month, 4);

    let dt =
        parse_rfc5322_date("Mon, 31 Jun 2025 12:00:00 +0000").expect("31 Jun must be accepted");
    assert_eq!(dt.day, 31);
    assert_eq!(dt.month, 6);

    let dt = parse_rfc5322_date("Mon, 29 Feb 2023 12:00:00 +0000")
        .expect("29 Feb in non-leap year must be accepted");
    assert_eq!(dt.day, 29);
    assert_eq!(dt.month, 2);
}

#[test]
fn parse_date_with_comment_between_tokens() {
    // RFC 5322 Section 4.3 (obsolete date syntax) allows CFWS
    // (comments and folding white space) between date tokens.
    // A comment like "(Friday)" between day and month must not cause
    // the date to fail to parse.
    let raw = b"From: a@b.com\r\n\
                 Date: Thu, 13 (February) Feb 2025 15:47:33 +0000\r\n\
                 \r\n";

    let parsed = parse_email(raw).unwrap();
    let date = parsed
        .date
        .expect("Date with CFWS comment must parse per RFC 5322 Section 4.3");
    assert_eq!(date.year, 2025);
    assert_eq!(date.month, 2);
    assert_eq!(date.day, 13);
}

#[test]
fn parse_date_with_trailing_comment() {
    // Common: trailing comment like "(UTC)" after timezone.
    // This already works but we add a test to ensure it stays working.
    let raw = b"From: a@b.com\r\n\
                 Date: Thu, 13 Feb 2025 15:47:33 +0000 (UTC)\r\n\
                 \r\n";

    let parsed = parse_email(raw).unwrap();
    let date = parsed.date.expect("Date with trailing comment must parse");
    assert_eq!(date.year, 2025);
    assert_eq!(date.tz_offset_minutes, 0);
}

#[test]
fn parse_date_with_nested_comments() {
    // RFC 5322 Section 3.2.2: comments can be nested.
    let raw = b"From: a@b.com\r\n\
                 Date: Thu, 13 Feb 2025 15:47:33 +0000 (UTC (nested))\r\n\
                 \r\n";

    let parsed = parse_email(raw).unwrap();
    let date = parsed.date.expect("Date with nested comment must parse");
    assert_eq!(date.year, 2025);
}

#[test]
fn parse_display_name_ending_with_escaped_quote() {
    // RFC 5322 Section 3.2.4 — a display name ending with an
    // escaped quote like `"She said \"hello\""` must parse correctly.
    // The outer quotes are the quoted-string delimiters; the inner `\"`
    // sequences are quoted-pairs that represent literal `"`.
    let raw = b"From: \"She said \\\"hello\\\"\" <she@example.com>\r\n\
                 Date: Thu, 13 Feb 2025 15:47:33 +0000\r\n\
                 \r\n";

    let parsed = parse_email(raw).unwrap();
    assert_eq!(
        parsed.from[0].name.as_deref(),
        Some("She said \"hello\""),
        "Display name ending with escaped quote must be parsed correctly \
         per RFC 5322 Section 3.2.4"
    );
}

#[test]
fn address_from_str_ending_with_escaped_quote() {
    // Same bug in Address::from_str — trim_matches('"') strips too many quotes.
    let addr: Address = "\"She said \\\"hello\\\"\" <she@example.com>"
        .parse()
        .unwrap();
    assert_eq!(
        addr.name.as_deref(),
        Some("She said \"hello\""),
        "Address::from_str must handle display names ending with escaped quotes"
    );
}

#[test]
fn boundary_must_be_at_line_start() {
    // RFC 2046 Section 5.1.1 requires the boundary delimiter
    // to appear at the beginning of a line (preceded by CRLF or at the
    // start of the body). A boundary string appearing mid-line in body
    // content must NOT be treated as a boundary delimiter.
    let raw = b"From: a@b.com\r\n\
                 Date: Thu, 13 Feb 2025 15:47:33 +0000\r\n\
                 Content-Type: multipart/mixed; boundary=\"BOUND\"\r\n\
                 \r\n\
                 --BOUND\r\n\
                 Content-Type: text/plain\r\n\
                 \r\n\
                 This line mentions --BOUND in the middle\r\n\
                 --BOUND--";

    let parsed = parse_email(raw).unwrap();
    // The body should contain the mid-line "--BOUND" as literal text,
    // not split into a second part at that point.
    let text = parsed.body_text.as_deref().unwrap_or("");
    assert!(
        text.contains("--BOUND"),
        "Mid-line boundary must be treated as literal text per RFC 2046 Section 5.1.1, \
         but body_text was: {text:?}"
    );
}

#[test]
fn mime_type_exact_match_not_prefix() {
    // A hypothetical MIME type like "text/plaintext" must NOT be treated
    // as text/plain body. Using starts_with("text/plain") would
    // incorrectly match it; exact equality is required.
    let raw = b"From: a@b.com\r\n\
                 Date: Thu, 13 Feb 2025 15:47:33 +0000\r\n\
                 Content-Type: multipart/mixed; boundary=\"b\"\r\n\
                 \r\n\
                 --b\r\n\
                 Content-Type: text/plaintext\r\n\
                 \r\n\
                 Not really plain text\r\n\
                 --b--";

    let parsed = parse_email(raw).unwrap();
    // text/plaintext is NOT text/plain — it should be treated as an
    // attachment, not extracted as body_text.
    assert!(
        parsed.body_text.is_none(),
        "text/plaintext must not be treated as text/plain body"
    );
    assert_eq!(
        parsed.attachments.len(),
        1,
        "text/plaintext should be treated as an attachment"
    );
}

#[test]
fn parse_single_part_non_text_is_attachment() {
    // A single-part message with Content-Type: image/jpeg should be treated
    // as an attachment, not as body_text (RFC 2046; requirements: "A part is
    // an attachment if... a non-text/non-multipart part").
    let raw = b"From: a@b.com\r\n\
                 Date: Thu, 13 Feb 2025 15:47:33 +0000\r\n\
                 Content-Type: image/jpeg\r\n\
                 Content-Transfer-Encoding: base64\r\n\
                 \r\n\
                 /9j/4AAQSkZJRg==";

    let parsed = parse_email(raw).unwrap();
    assert!(
        parsed.body_text.is_none(),
        "image/jpeg single-part must not populate body_text"
    );
    assert_eq!(
        parsed.attachments.len(),
        1,
        "image/jpeg single-part must be treated as an attachment"
    );
    assert_eq!(parsed.attachments[0].content_type, "image/jpeg");
    assert_eq!(parsed.attachments[0].section.as_deref(), Some("1"));
}

#[test]
fn parse_single_part_application_pdf_is_attachment() {
    // Single-part application/pdf should be an attachment, not body_text.
    let raw = b"From: a@b.com\r\n\
                 Date: Thu, 13 Feb 2025 15:47:33 +0000\r\n\
                 Content-Type: application/pdf; name=\"doc.pdf\"\r\n\
                 Content-Disposition: attachment; filename=\"doc.pdf\"\r\n\
                 Content-Transfer-Encoding: base64\r\n\
                 \r\n\
                 JVBERi0xLjQK";

    let parsed = parse_email(raw).unwrap();
    assert!(
        parsed.body_text.is_none(),
        "application/pdf must not populate body_text"
    );
    assert_eq!(parsed.attachments.len(), 1);
    assert_eq!(parsed.attachments[0].content_type, "application/pdf");
    assert_eq!(parsed.attachments[0].filename.as_deref(), Some("doc.pdf"));
    assert!(!parsed.attachments[0].is_inline);
}

#[test]
fn parse_single_part_text_plain_with_attachment_disposition() {
    // text/plain with Content-Disposition: attachment should be treated as
    // an attachment, not body_text (requirements: "A part is an attachment
    // if it has Content-Disposition: attachment").
    let raw = b"From: a@b.com\r\n\
                 Date: Thu, 13 Feb 2025 15:47:33 +0000\r\n\
                 Content-Type: text/plain; charset=utf-8\r\n\
                 Content-Disposition: attachment; filename=\"log.txt\"\r\n\
                 \r\n\
                 Server log data here";

    let parsed = parse_email(raw).unwrap();
    assert!(
        parsed.body_text.is_none(),
        "text/plain with disposition:attachment must not populate body_text"
    );
    assert_eq!(parsed.attachments.len(), 1);
    assert_eq!(parsed.attachments[0].content_type, "text/plain");
    assert_eq!(parsed.attachments[0].filename.as_deref(), Some("log.txt"));
}

/// Regression: Content-Disposition type matching must be token-exact,
/// not prefix-based.  RFC 2183 Section 2 defines:
///   disposition-type = "inline" / "attachment" / extension-token
/// An extension-token like "attachmentfoo" must NOT be treated as
/// "attachment".  Using `starts_with("attachment")` incorrectly matches
/// such tokens, causing text/plain body content to be classified as an
/// attachment instead of populating `body_text`.
#[test]
fn disposition_type_requires_token_boundary() {
    // Content-Disposition uses a fabricated extension-token "attachmentfoo".
    // This is NOT "attachment" per RFC 2183 Section 2, so the text/plain
    // body must populate body_text, not become an attachment.
    let raw = b"From: a@b.com\r\n\
                 Date: Thu, 13 Feb 2025 15:47:33 +0000\r\n\
                 Content-Type: text/plain; charset=utf-8\r\n\
                 Content-Disposition: attachmentfoo\r\n\
                 \r\n\
                 This is body text";

    let parsed = parse_email(raw).unwrap();
    assert!(
        parsed.body_text.is_some(),
        "text/plain with disposition:attachmentfoo must populate body_text \
         (RFC 2183 Section 2: disposition-type is a complete token, not a prefix)"
    );
    assert_eq!(
        parsed.attachments.len(),
        0,
        "extension-token 'attachmentfoo' must not be classified as 'attachment'"
    );
}

#[test]
fn parse_group_address_empty_undisclosed() {
    // RFC 5322 Section 3.4: group = display-name ":" [group-list] ";"
    // "undisclosed-recipients:;" is a common empty group that must not
    // produce malformed addresses or cause a parse failure.
    let raw = b"From: a@b.com\r\n\
                 To: undisclosed-recipients:;\r\n\
                 Date: Thu, 13 Feb 2025 15:47:33 +0000\r\n\
                 \r\n";

    let parsed = parse_email(raw).unwrap();
    // Empty group — no addresses should be extracted
    assert!(
        parsed.to.is_empty(),
        "empty group undisclosed-recipients:; must produce no addresses, got {:?}",
        parsed.to
    );
}

#[test]
fn parse_group_address_with_members() {
    // RFC 5322 Section 3.4: group with member addresses
    // "friends:a@b.com, c@d.com;" should extract the member addresses.
    let raw = b"From: a@b.com\r\n\
                 To: friends:one@x.com, two@x.com;\r\n\
                 Date: Thu, 13 Feb 2025 15:47:33 +0000\r\n\
                 \r\n";

    let parsed = parse_email(raw).unwrap();
    assert_eq!(
        parsed.to.len(),
        2,
        "group with 2 members must produce 2 addresses, got {:?}",
        parsed.to
    );
    assert_eq!(parsed.to[0].email, "one@x.com");
    assert_eq!(parsed.to[1].email, "two@x.com");
}

#[test]
fn parse_group_address_mixed_with_regular() {
    // Mix of regular addresses and group syntax in same header.
    let raw = b"From: a@b.com\r\n\
                 To: solo@x.com, friends:one@x.com, two@x.com;, last@x.com\r\n\
                 Date: Thu, 13 Feb 2025 15:47:33 +0000\r\n\
                 \r\n";

    let parsed = parse_email(raw).unwrap();
    let emails: Vec<&str> = parsed.to.iter().map(|a| a.email.as_str()).collect();
    assert_eq!(
        emails,
        vec!["solo@x.com", "one@x.com", "two@x.com", "last@x.com"],
        "must extract all 4 addresses from mixed regular+group syntax"
    );
}

#[test]
fn decode_qp_trailing_equals_is_soft_break() {
    // RFC 2045 Section 6.7: a trailing '=' at end-of-data is a soft line
    // break and must be removed, not emitted as a literal '=' character.
    let result = decode_quoted_printable(b"Hello=");
    assert_eq!(
        result, b"Hello",
        "trailing '=' must be treated as soft line break per RFC 2045 Section 6.7"
    );
}

#[test]
fn decode_qp_trailing_equals_cr_is_soft_break() {
    // '=\r' at end-of-data (bare CR without LF) should also be removed
    // as a soft line break.
    let result = decode_quoted_printable(b"Hello=\r");
    assert_eq!(
        result, b"Hello",
        "trailing '=\\r' must be treated as soft line break"
    );
}

#[test]
fn decode_qp_bare_cr_soft_break_mid_data() {
    // RFC 2045 Section 6.7: '=' followed by a line break is a soft break.
    // Per Postel's law, a bare CR (=\r not followed by \n) in the middle
    // of the data must also be recognized as a soft line break, not just
    // when it appears at the end of the data.
    let result = decode_quoted_printable(b"Hello=\rWorld");
    assert_eq!(
        result, b"HelloWorld",
        "'=\\r' followed by non-LF byte must be treated as soft line break"
    );
}

#[test]
fn parse_bare_address_with_trailing_comment() {
    // RFC 5322 Section 3.4.1: addr-spec can be followed by CFWS.
    // RFC 5322 Section 3.2.2: parenthesized text is a comment.
    // A trailing comment like "(Display Name)" after a bare address
    // must be stripped from the email and used as the display name.
    let raw = b"From: sender@example.com\r\n\
                 To: user@example.com (Display Name)\r\n\
                 Date: Thu, 13 Feb 2025 15:47:33 +0000\r\n\
                 \r\n";

    let parsed = parse_email(raw).unwrap();
    assert_eq!(parsed.to.len(), 1);
    assert_eq!(
        parsed.to[0].email, "user@example.com",
        "email must not contain the trailing comment"
    );
    assert_eq!(
        parsed.to[0].name.as_deref(),
        Some("Display Name"),
        "trailing comment should become display name per RFC 5322 Section 3.4.1"
    );
}

#[test]
fn parse_bare_address_with_leading_comment() {
    // RFC 5322 Section 3.2.2 / RFC 822 convention: a parenthesized
    // comment adjacent to a bare addr-spec serves as the display name,
    // whether it appears before or after the address. Both positions
    // are common in the wild; extracting from leading comments mirrors
    // the existing trailing-comment extraction for consistency.
    let raw = b"From: sender@example.com\r\n\
                 To: (Comment) user@example.com\r\n\
                 Date: Thu, 13 Feb 2025 15:47:33 +0000\r\n\
                 \r\n";

    let parsed = parse_email(raw).unwrap();
    assert_eq!(parsed.to.len(), 1);
    assert_eq!(
        parsed.to[0].email, "user@example.com",
        "email must not contain the leading comment"
    );
    assert_eq!(
        parsed.to[0].name.as_deref(),
        Some("Comment"),
        "leading comment should be extracted as display name \
         (RFC 5322 Section 3.2.2 / RFC 822 convention)"
    );
}

#[test]
fn extract_param_skips_quoted_values() {
    // Parameter name appearing inside another parameter's quoted value
    // should not be matched (RFC 5322 Section 3.2.4).
    let header = "text/html; boundary=\"has charset=bad inside\"; charset=utf-8";
    let charset = extract_param(header, "charset");
    assert_eq!(
        charset.as_deref(),
        Some("utf-8"),
        "Should skip match inside quoted boundary value"
    );
}

#[test]
fn extract_param_skips_comment_before_unquoted_value() {
    // RFC 2045 Section 5.1 allows RFC 822 / RFC 5322 comments in
    // structured fields, so CFWS between `=` and an unquoted token
    // must not become part of the parameter value.
    let header = "text/plain; charset=(legacy default)windows-1252";
    let charset = extract_param(header, "charset");
    assert_eq!(
        charset.as_deref(),
        Some("windows-1252"),
        "RFC 2045 Section 5.1 comments before an unquoted value must be ignored"
    );
}

#[test]
fn extract_param_skips_comment_before_quoted_value() {
    // RFC 2045 Section 5.1 + RFC 5322 Section 3.2.2: CFWS comments may
    // also appear before a quoted-string parameter value.
    let header = "attachment; filename=(human note)\"report.pdf\"";
    let filename = extract_param(header, "filename");
    assert_eq!(
        filename.as_deref(),
        Some("report.pdf"),
        "RFC 2045 Section 5.1 comments before a quoted-string value must be ignored"
    );
}

#[test]
fn multipart_part_without_charset_uses_us_ascii_default() {
    // RFC 2045 Section 5.2: default Content-Type is text/plain; charset=us-ascii.
    // A MIME part with text/plain but no charset parameter should default
    // to US-ASCII, consistent with the top-level default.
    //
    // Windows-1252 byte 0x93 is a left double quotation mark (U+201C).
    // encoding_rs maps us-ascii → Windows-1252, so 0x93 decodes to U+201C.
    // Under UTF-8, 0x93 is an invalid byte and produces U+FFFD.
    let raw = b"From: a@b.com\r\n\
                 Date: Thu, 13 Feb 2025 15:47:33 +0000\r\n\
                 Content-Type: multipart/mixed; boundary=\"b\"\r\n\
                 \r\n\
                 --b\r\n\
                 Content-Type: text/plain\r\n\
                 \r\n\
                 Hello \x93World\r\n\
                 --b--";

    let parsed = parse_email(raw).unwrap();
    let text = parsed.body_text.unwrap();

    // With the correct US-ASCII default (mapped to Windows-1252 by encoding_rs),
    // 0x93 should decode to U+201C, not the replacement character U+FFFD.
    assert!(
        text.contains('\u{201c}'),
        "Part with text/plain (no charset) should use US-ASCII default per \
         RFC 2045 Section 5.2, decoding 0x93 as U+201C. Got: {text:?}"
    );
    assert!(
        !text.contains('\u{FFFD}'),
        "Part with text/plain (no charset) should not produce UTF-8 replacement \
         characters. Got: {text:?}"
    );
}

/// header unfolding must preserve trailing whitespace from
/// the first line when followed by a continuation line.
///
/// RFC 5322 Section 2.2.3: "Unfolding is accomplished by simply removing
/// any CRLF that is immediately followed by WSP." Only the CRLF is
/// removed; all other whitespace (including trailing spaces on the first
/// line) must be preserved.
#[test]
fn parse_header_unfold_preserves_trailing_whitespace() {
    // "Subject: Hello  \r\n World" should unfold to "Hello   World"
    // (2 trailing spaces from first line + 1 leading space from continuation = 3 spaces)
    let raw = b"From: a@b.com\r\nSubject: Hello  \r\n World\r\nDate: Thu, 13 Feb 2025 15:47:33 +0000\r\n\r\n";
    let parsed = parse_email(raw).unwrap();
    assert_eq!(
        parsed.subject.as_deref(),
        Some("Hello   World"),
        "Trailing whitespace on first line must be preserved during unfolding \
         (RFC 5322 Section 2.2.3)"
    );
}

/// Single-part body text must not include the trailing CRLF
/// that the message format requires after the body content. The trailing
/// CRLF is a format artifact (RFC 5322 Section 3.5), not semantic content.
/// Without this fix, build→parse round-trip adds a spurious "\r\n" to the
/// body text of single-part messages.
#[test]
fn parse_single_part_body_no_trailing_crlf() {
    // Single-part message: body ends with \r\n (format artifact)
    let raw = b"From: a@b.com\r\n\
                 Date: Thu, 13 Feb 2025 15:47:33 +0000\r\n\
                 Content-Type: text/plain; charset=utf-8\r\n\
                 \r\n\
                 Hello, World!\r\n";

    let parsed = parse_email(raw).unwrap();
    assert_eq!(
        parsed.body_text.as_deref(),
        Some("Hello, World!"),
        "Single-part body text must not include trailing CRLF"
    );
}

/// Single-part HTML body must not include trailing CRLF.
#[test]
fn parse_single_part_html_no_trailing_crlf() {
    let raw = b"From: a@b.com\r\n\
                 Date: Thu, 13 Feb 2025 15:47:33 +0000\r\n\
                 Content-Type: text/html; charset=utf-8\r\n\
                 \r\n\
                 <p>Hello</p>\r\n";

    let parsed = parse_email(raw).unwrap();
    assert_eq!(
        parsed.body_html.as_deref(),
        Some("<p>Hello</p>"),
        "Single-part HTML body must not include trailing CRLF"
    );
}

/// Build→parse round-trip for single-part text must preserve body exactly.
#[test]
fn round_trip_single_part_body_text() {
    let email = crate::types::OutgoingEmail {
        from: vec![crate::types::Address {
            name: None,
            email: "a@b.com".into(),
        }],
        sender: None,
        to: vec![crate::types::Address {
            name: None,
            email: "c@d.com".into(),
        }],
        cc: vec![],
        bcc: vec![],
        reply_to: vec![],
        date: None,
        subject: "Test".into(),
        body_text: Some("Hello, World!".into()),
        body_html: None,
        in_reply_to: vec![],
        references: vec![],
        attachments: vec![],
        extra_headers: vec![],
    };

    let built = crate::build_message(&email).unwrap();
    let parsed = parse_email(&built.raw).unwrap();
    assert_eq!(
        parsed.body_text.as_deref(),
        Some("Hello, World!"),
        "Single-part body text must round-trip without trailing CRLF"
    );
}

/// RFC 2047 encoded words in display names must be decoded
/// AFTER address structure parsing, not before. Decoding before parsing
/// breaks address splitting when the decoded text contains address-
/// significant characters like commas.
///
/// RFC 2047 Section 5 rule (3): encoded-words may appear in a 'phrase'
/// (e.g., display name) within address headers. The encoded word
/// `=?UTF-8?B?Sm9obiwgRG9l?=` decodes to `John, Doe`. If decoded
/// before address parsing, the comma splits the address incorrectly.
#[test]
fn parse_encoded_word_display_name_with_comma() {
    // "John, Doe" base64-encoded as an RFC 2047 encoded word.
    // The comma must NOT split the address — it's part of the name.
    let raw = b"From: =?UTF-8?B?Sm9obiwgRG9l?= <john@example.com>\r\n\
                 Date: Thu, 13 Feb 2025 15:47:33 +0000\r\n\
                 \r\n";

    let parsed = parse_email(raw).unwrap();
    assert_eq!(
        parsed.from[0].name.as_deref(),
        Some("John, Doe"),
        "RFC 2047 encoded display name with comma must be preserved \
         (RFC 2047 Section 5 rule 3): decode AFTER address parsing"
    );
    assert_eq!(parsed.from[0].email, "john@example.com");
}

/// RFC 2045 Section 6.8: "Any characters outside of the base64 alphabet
/// are to be ignored in base64-encoded data." The base64 alphabet is
/// A-Z, a-z, 0-9, +, /, = (padding). Stray non-alphabet characters such
/// as `!`, `#`, `~` must be stripped before decoding, not just whitespace.
#[test]
fn parse_base64_body_ignores_non_alphabet_chars() {
    // "Hello World" = "SGVsbG8gV29ybGQ=" in base64.
    // Insert non-base64 characters (!, #, ~) that are NOT whitespace
    // to verify the decoder strips all non-alphabet bytes.
    let raw = b"From: a@b.com\r\n\
                 Date: Thu, 13 Feb 2025 15:47:33 +0000\r\n\
                 Content-Type: text/plain; charset=utf-8\r\n\
                 Content-Transfer-Encoding: base64\r\n\
                 \r\n\
                 SGVs!bG8#gV29~ybGQ=";

    let parsed = parse_email(raw).unwrap();
    assert_eq!(
        parsed.body_text.as_deref(),
        Some("Hello World"),
        "RFC 2045 Section 6.8: non-alphabet characters must be ignored in base64 data"
    );
}

/// Same as above but for To/Cc headers with multiple recipients.
/// The encoded comma must not create a spurious address split.
#[test]
fn parse_encoded_word_display_name_with_comma_in_to() {
    // Two recipients: first has an encoded comma in the display name,
    // second is a plain address. Must parse as exactly 2 addresses.
    let raw = b"From: sender@example.com\r\n\
                 To: =?UTF-8?B?Sm9obiwgRG9l?= <john@example.com>, other@example.com\r\n\
                 Date: Thu, 13 Feb 2025 15:47:33 +0000\r\n\
                 \r\n";

    let parsed = parse_email(raw).unwrap();
    assert_eq!(
        parsed.to.len(),
        2,
        "Must parse exactly 2 addresses, not 3 (encoded comma is not a separator)"
    );
    assert_eq!(
        parsed.to[0].name.as_deref(),
        Some("John, Doe"),
        "First recipient display name must be 'John, Doe'"
    );
    assert_eq!(parsed.to[0].email, "john@example.com");
    assert_eq!(parsed.to[1].email, "other@example.com");
}

/// building a message with `body_text: None` (and attachments,
/// which forces multipart/mixed with an empty text/plain part) and then
/// parsing it back must yield `body_text: None`, not `Some("")`.
///
/// The builder's `write_text_part` always appends a trailing `\r\n` after
/// the body content (RFC 2046 Section 5.1.1 requires CRLF before boundary).
/// When the body is empty, the MIME part body becomes just `\r\n`, which
/// `decode_body` strips to `""`. The parser must treat an empty decoded
/// body as absent (`None`), not present-but-empty (`Some("")`).
///
/// # References
/// - RFC 2046 Section 5.1.1 (CRLF before boundary delimiter)
/// - RFC 5322 Section 3.5 (message body)
#[test]
fn round_trip_empty_body_text_is_none() {
    // Use attachments to force multipart/mixed, which causes the builder
    // to emit an empty text/plain part when body_text is None.
    let email = crate::types::OutgoingEmail {
        from: vec![crate::types::Address {
            name: None,
            email: "a@b.com".into(),
        }],
        sender: None,
        to: vec![crate::types::Address {
            name: None,
            email: "c@d.com".into(),
        }],
        cc: vec![],
        bcc: vec![],
        reply_to: vec![],
        date: None,
        subject: "Empty body".into(),
        body_text: None,
        body_html: None,
        in_reply_to: vec![],
        references: vec![],
        attachments: vec![crate::types::OutgoingAttachment {
            filename: "test.txt".into(),
            content_type: "text/plain".into(),
            data: b"attachment data".to_vec(),
            is_inline: false,
            content_id: None,
        }],
        extra_headers: vec![],
    };

    let built = crate::build_message(&email).unwrap();
    let parsed = parse_email(&built.raw).unwrap();

    assert_eq!(
        parsed.body_text, None,
        "Empty body_text must round-trip as None, not Some(\"\")"
    );
}

/// Same fix for multipart/alternative: `body_text: Some("")`
/// should parse back as `None` since the decoded content is empty.
///
/// # References
/// - RFC 2046 Section 5.1.1 (CRLF before boundary delimiter)
#[test]
fn round_trip_empty_body_html_in_alternative_is_none() {
    // Both body_text and body_html present forces multipart/alternative.
    // An explicitly empty body_html should parse back as None.
    let email = crate::types::OutgoingEmail {
        from: vec![crate::types::Address {
            name: None,
            email: "a@b.com".into(),
        }],
        sender: None,
        to: vec![crate::types::Address {
            name: None,
            email: "c@d.com".into(),
        }],
        cc: vec![],
        bcc: vec![],
        reply_to: vec![],
        date: None,
        subject: "Text only".into(),
        body_text: Some("Plain text".into()),
        body_html: Some(String::new()),
        in_reply_to: vec![],
        references: vec![],
        attachments: vec![],
        extra_headers: vec![],
    };

    let built = crate::build_message(&email).unwrap();
    let parsed = parse_email(&built.raw).unwrap();

    assert_eq!(
        parsed.body_html, None,
        "Empty body_html must parse as None, not Some(\"\")"
    );
    assert_eq!(
        parsed.body_text.as_deref(),
        Some("Plain text"),
        "body_text must be preserved"
    );
}

#[test]
fn extract_mime_type_strips_rfc5322_comments() {
    // RFC 5322 Section 3.2.2 parenthesized comments may appear
    // in CFWS positions within the Content-Type type/subtype production.
    // extract_mime_type must strip them so that "text/plain (comment)"
    // normalizes to "text/plain", not "text/plain (comment)".
    //
    // Single-part message: the comment in Content-Type caused the MIME
    // comparison to fail, making the parser treat the body as an
    // attachment instead of body_text.
    let raw = b"From: a@b.com\r\n\
                 Date: Thu, 13 Feb 2025 15:47:33 +0000\r\n\
                 Content-Type: text/plain (this is a comment); charset=utf-8\r\n\
                 \r\n\
                 Hello with comment";

    let parsed = parse_email(raw).unwrap();
    assert_eq!(
        parsed.body_text.as_deref(),
        Some("Hello with comment"),
        "Body must be extracted as body_text when Content-Type has an RFC 5322 comment"
    );
    assert!(
        parsed.attachments.is_empty(),
        "No attachments expected for a plain text/plain message with a comment"
    );
}

/// RFC 2046 Section 5.1.5: In a multipart/digest, the default Content-Type
/// for body parts is "message/rfc822", NOT "text/plain; charset=us-ascii".
/// Parts without an explicit Content-Type header must be treated as
/// message/rfc822 and show up as attachments, not `body_text`.
#[test]
fn multipart_digest_default_content_type_is_message_rfc822() {
    let raw = b"From: sender@example.com\r\n\
                 Date: Thu, 13 Feb 2025 15:47:33 +0000\r\n\
                 Subject: Digest\r\n\
                 Content-Type: multipart/digest; boundary=\"digestboundary\"\r\n\
                 \r\n\
                 --digestboundary\r\n\
                 \r\n\
                 From: nested@example.com\r\n\
                 Subject: Nested message\r\n\
                 \r\n\
                 Nested body text\r\n\
                 --digestboundary--\r\n";

    let parsed = parse_email(raw).unwrap();

    // The part has no Content-Type header. In multipart/digest, this means
    // message/rfc822 per RFC 2046 Section 5.1.5 — it should appear as an
    // attachment, NOT as body_text.
    assert!(
        parsed.body_text.is_none(),
        "multipart/digest parts without Content-Type should default to \
         message/rfc822, not text/plain — body_text should be None"
    );
    assert_eq!(
        parsed.attachments.len(),
        1,
        "multipart/digest part should be treated as message/rfc822 attachment"
    );
    assert_eq!(
        parsed.attachments[0].content_type, "message/rfc822",
        "default Content-Type in multipart/digest must be message/rfc822 \
         (RFC 2046 Section 5.1.5)"
    );
}

/// Content-ID with whitespace inside angle brackets must be
/// trimmed after bracket stripping.
///
/// RFC 2392 defines Content-ID as `"<" addr-spec ">"`. Some mailers
/// add whitespace around the addr-spec inside the brackets. After
/// stripping `<` and `>`, the result must be trimmed to produce a
/// clean identifier for matching (e.g., for CID references in HTML).
#[test]
fn content_id_whitespace_inside_brackets_trimmed() {
    // Multipart path (walk_mime_tree)
    let raw = b"From: a@b.com\r\n\
                 Date: Thu, 13 Feb 2025 15:47:33 +0000\r\n\
                 Content-Type: multipart/mixed; boundary=\"b\"\r\n\
                 \r\n\
                 --b\r\n\
                 Content-Type: text/plain\r\n\
                 \r\n\
                 Body\r\n\
                 --b\r\n\
                 Content-Type: image/png\r\n\
                 Content-ID: < cid@example.com >\r\n\
                 \r\n\
                 PNG\r\n\
                 --b--";
    let parsed = parse_email(raw).unwrap();
    assert_eq!(
        parsed.attachments[0].content_id.as_deref(),
        Some("cid@example.com"),
        "Content-ID must be trimmed after bracket stripping (RFC 2392)"
    );

    // Simple body path (extract_simple_body)
    let raw_single = b"From: a@b.com\r\n\
                       Date: Thu, 13 Feb 2025 15:47:33 +0000\r\n\
                       Content-Type: image/png\r\n\
                       Content-ID: < cid2@example.com >\r\n\
                       \r\n\
                       PNG";
    let parsed_single = parse_email(raw_single).unwrap();
    assert_eq!(
        parsed_single.attachments[0].content_id.as_deref(),
        Some("cid2@example.com"),
        "Content-ID in single-part message must be trimmed (RFC 2392)"
    );
}

#[test]
fn parse_headers_only_all_fields_verified() {
    // Verify that parse_headers_only extracts ALL header fields correctly,
    // including cc, bcc, reply_to, in_reply_to, and references.
    let raw = b"From: sender@example.com\r\n\
                 To: to@example.com\r\n\
                 Cc: cc@example.com\r\n\
                 Bcc: bcc@example.com\r\n\
                 Reply-To: reply@example.com\r\n\
                 Subject: Full test\r\n\
                 Date: Thu, 13 Feb 2025 15:47:33 +0000\r\n\
                 Message-ID: <msg1@example.com>\r\n\
                 In-Reply-To: <parent@example.com>\r\n\
                 References: <ref1@example.com> <ref2@example.com>\r\n\
                 \r\n\
                 Body that should be ignored";

    let parsed = parse_headers_only(raw).unwrap();

    assert_eq!(parsed.from[0].email, "sender@example.com");
    assert_eq!(parsed.to.len(), 1);
    assert_eq!(parsed.to[0].email, "to@example.com");
    assert_eq!(parsed.cc.len(), 1);
    assert_eq!(parsed.cc[0].email, "cc@example.com");
    assert_eq!(parsed.bcc.len(), 1);
    assert_eq!(parsed.bcc[0].email, "bcc@example.com");
    assert_eq!(parsed.reply_to.len(), 1);
    assert_eq!(parsed.reply_to[0].email, "reply@example.com");
    assert_eq!(parsed.subject.as_deref(), Some("Full test"));
    assert!(parsed.date.is_some());
    assert_eq!(parsed.message_id.as_deref(), Some("msg1@example.com"));
    assert_eq!(parsed.in_reply_to, vec!["parent@example.com"]);
    assert_eq!(
        parsed.references,
        vec!["ref1@example.com", "ref2@example.com"]
    );

    // Body fields must be empty
    assert!(parsed.body_text.is_none());
    assert!(parsed.body_html.is_none());
    assert!(parsed.attachments.is_empty());
}

#[test]
fn parse_missing_optional_headers_returns_none_or_empty() {
    // Only From is required (RFC 5322 Section 3.6.2). All other fields
    // should gracefully return None or empty when absent.
    let raw = b"From: a@b.com\r\n\r\n";

    let parsed = parse_email(raw).unwrap();

    assert_eq!(parsed.from[0].email, "a@b.com");
    assert!(parsed.subject.is_none());
    assert!(parsed.date.is_none());
    assert!(parsed.message_id.is_none());
    assert!(parsed.in_reply_to.is_empty());
    assert!(parsed.references.is_empty());
    assert!(parsed.to.is_empty());
    assert!(parsed.cc.is_empty());
    assert!(parsed.bcc.is_empty());
    assert!(parsed.reply_to.is_empty());
}

#[test]
fn extract_param_rejects_substring_match() {
    // "filename=" must not match inside "xfilename=" — the boundary check
    // requires `;`, space, tab, or start-of-string before the param name.
    let value = "attachment; xfilename=\"bad.pdf\"; filename=\"good.pdf\"";
    let result = extract_param(value, "filename");
    assert_eq!(
        result.as_deref(),
        Some("good.pdf"),
        "Must not match xfilename as filename"
    );
}

#[test]
fn extract_param_rejects_suffix_only_match() {
    // Edge case: param name appears only as suffix of another param name.
    let value = "attachment; notfilename=\"only.pdf\"";
    let result = extract_param(value, "filename");
    assert!(
        result.is_none(),
        "Must not match 'filename' inside 'notfilename'"
    );
}

// -----------------------------------------------------------------------
// Audit coverage: Group address parsing (RFC 5322 Section 3.4)
// -----------------------------------------------------------------------

/// RFC 5322 Section 3.4: empty group `undisclosed-recipients:;`
/// must produce zero addresses.
#[test]
fn parse_group_address_empty() {
    let addrs = parse_address_list("undisclosed-recipients:;");
    assert!(
        addrs.is_empty(),
        "empty group must produce no addresses, got {addrs:?}"
    );
}

/// RFC 5322 Section 3.4: group with members — audit coverage.
#[test]
fn parse_group_address_with_two_members() {
    let addrs = parse_address_list("Friends: a@x.com, b@x.com;");
    assert_eq!(addrs.len(), 2, "group with 2 members: {addrs:?}");
    assert_eq!(addrs[0].email, "a@x.com");
    assert_eq!(addrs[1].email, "b@x.com");
}

/// RFC 5322 Section 3.4: multiple groups in a single header value.
#[test]
fn parse_multiple_groups_and_solo() {
    let addrs = parse_address_list("Team A: a1@x.com, a2@x.com;, Team B: b1@x.com;, solo@x.com");
    assert_eq!(addrs.len(), 4, "2 groups + 1 solo: {addrs:?}");
    assert_eq!(addrs[0].email, "a1@x.com");
    assert_eq!(addrs[1].email, "a2@x.com");
    assert_eq!(addrs[2].email, "b1@x.com");
    assert_eq!(addrs[3].email, "solo@x.com");
}

/// Address with parenthesized comment containing commas must not
/// split on those commas (RFC 5322 Section 3.2.2) — audit coverage.
#[test]
fn parse_address_comment_with_comma_audit() {
    let addrs = parse_address_list("user@x.com (Last, First), other@x.com");
    assert_eq!(
        addrs.len(),
        2,
        "comma inside comment must not split: {addrs:?}"
    );
    assert_eq!(addrs[0].email, "user@x.com");
    assert_eq!(addrs[1].email, "other@x.com");
}

// -----------------------------------------------------------------------
// Audit coverage: RFC 2231 Section 3 — continuation edge cases
// -----------------------------------------------------------------------

/// RFC 2231 Section 3: continuation indices start at 0 and gaps are not
/// allowed. When a later section is missing, preserve only the contiguous
/// prefix instead of fabricating bytes from the trailing fragment.
#[test]
fn rfc2231_continuation_gap_collects_available() {
    let header = "attachment; filename*0=\"hello\"; filename*2=\"skipped\"";
    let result = extract_rfc2231_continuation(header, "filename");
    assert_eq!(
        result.as_deref(),
        Some("hello"),
        "RFC 2231 Section 3: reassembly must stop at the first gap"
    );
}

/// RFC 2231 Section 3 forbids wider gaps as well. Even when the scanner
/// sees a much later section, reassembly must stop once index 1 is absent.
#[test]
fn rfc2231_continuation_double_gap_collected() {
    let header = "attachment; filename*0=\"abc\"; filename*3=\"xyz\"";
    let result = extract_rfc2231_continuation(header, "filename");
    assert_eq!(
        result.as_deref(),
        Some("abc"),
        "RFC 2231 Section 3: reassembly must ignore sections after a gap"
    );
}

/// RFC 2231 Section 3: single continuation section produces the value.
#[test]
fn rfc2231_continuation_single_section() {
    let header = "attachment; filename*0=\"report.pdf\"";
    let result = extract_rfc2231_continuation(header, "filename");
    assert_eq!(result.as_deref(), Some("report.pdf"));
}

/// RFC 2231 Section 3: continuation indices start at 0. A parameter that
/// begins at `*1` is malformed and must not be reassembled into a value.
#[test]
fn rfc2231_continuation_requires_section_zero() {
    let header = "attachment; filename*1=\"tail.txt\"";
    let result = extract_rfc2231_continuation(header, "filename");
    assert_eq!(
        result, None,
        "RFC 2231 Section 3: continuation reassembly must require section 0"
    );
}

/// RFC 2231 Section 3: gaps are not allowed. To avoid fabricating bytes
/// that were never sent, the parser should preserve only the contiguous
/// prefix that starts at `*0`.
#[test]
fn rfc2231_continuation_stops_at_first_gap() {
    let header = "attachment; filename*0=\"hello\"; filename*2=\"tail\"";
    let result = extract_rfc2231_continuation(header, "filename");
    assert_eq!(
        result.as_deref(),
        Some("hello"),
        "RFC 2231 Section 3: reassembly must stop at the first missing index"
    );
}

/// RFC 2231 Section 4 + RFC 2045 Section 5.2: when no charset is declared
/// in continuation parameters, the default per the RFCs would be US-ASCII.
/// We intentionally default to UTF-8 as a Postel's law accommodation (see
/// the inline comment in `extract_rfc2231_continuation`). This test verifies
/// that plain (non-charset-encoded) continuation parameters without any
/// charset declaration decode correctly — ASCII values must round-trip
/// identically since US-ASCII is a strict subset of UTF-8.
#[test]
fn rfc2231_continuation_no_charset_defaults_to_utf8() {
    // Plain continuation sections (no `*` suffix, no charset declaration).
    // The parser must reassemble them using the UTF-8 default.
    let raw = b"From: a@b.com\r\n\
                 Date: Thu, 13 Feb 2025 15:47:33 +0000\r\n\
                 Content-Type: multipart/mixed; boundary=\"b\"\r\n\
                 \r\n\
                 --b\r\n\
                 Content-Type: text/plain\r\n\
                 \r\n\
                 Body\r\n\
                 --b\r\n\
                 Content-Type: application/octet-stream\r\n\
                 Content-Disposition: attachment; filename*0=\"annual_\"; filename*1=\"report_\"; filename*2=\"2025.pdf\"\r\n\
                 \r\n\
                 DATA\r\n\
                 --b--";

    let parsed = parse_email(raw).unwrap();
    assert_eq!(parsed.attachments.len(), 1);
    assert_eq!(
        parsed.attachments[0].filename.as_deref(),
        Some("annual_report_2025.pdf"),
        "RFC 2231 continuation without charset should decode as UTF-8"
    );
}

// -----------------------------------------------------------------------
// Audit coverage: Base64 with embedded whitespace (RFC 2045 Section 6.8)
// -----------------------------------------------------------------------

/// RFC 2045 Section 6.8: "Any characters outside of the base64 alphabet
/// are to be ignored in base64-encoded data." Spaces/tabs within base64
/// lines must be stripped before decoding.
#[test]
fn base64_with_embedded_spaces() {
    // "Hello World" = SGVsbG8gV29ybGQ=, with spaces injected
    let data = b"SGVs bG8g V29y bGQ=";
    let decoded = decode_transfer_encoding(data, "base64");
    assert_eq!(
        std::str::from_utf8(&decoded).unwrap(),
        "Hello World",
        "base64 decoder must strip non-alphabet characters (RFC 2045 Section 6.8)"
    );
}

/// RFC 2045 Section 6.8: tabs within base64-encoded data must be ignored.
#[test]
fn base64_with_tabs() {
    let data = b"SGVs\tbG8g\tV29ybGQ=";
    let decoded = decode_transfer_encoding(data, "base64");
    assert_eq!(
        std::str::from_utf8(&decoded).unwrap(),
        "Hello World",
        "base64 decoder must strip tabs (RFC 2045 Section 6.8)"
    );
}

// -----------------------------------------------------------------------
// Audit coverage: Quoted-printable edge cases (RFC 2045 Section 6.7)
// -----------------------------------------------------------------------

/// RFC 2045 Section 6.7: trailing `=` at end of data is a soft line
/// break — it should be stripped, not produce a literal `=`.
#[test]
fn qp_trailing_equals_stripped() {
    let data = b"Hello=";
    let decoded = decode_quoted_printable(data);
    assert_eq!(
        std::str::from_utf8(&decoded).unwrap(),
        "Hello",
        "trailing '=' is a soft break (RFC 2045 Section 6.7)"
    );
}

/// RFC 2045 Section 6.7: `=` followed by non-hex characters is malformed.
/// Postel's law: pass through the literal `=` and the following characters.
#[test]
fn qp_malformed_hex_passthrough() {
    let data = b"Hello=ZZ World";
    let decoded = decode_quoted_printable(data);
    assert_eq!(
        std::str::from_utf8(&decoded).unwrap(),
        "Hello=ZZ World",
        "malformed =ZZ must pass through literally (Postel's law)"
    );
}

// -----------------------------------------------------------------------
// Audit coverage: RFC 2047 Q-encoding edge cases
// -----------------------------------------------------------------------

/// RFC 2047 Section 4.2: `=` followed by non-hex in Q-encoding should
/// pass through the literal `=` (Postel's law — decoders accept gracefully).
#[test]
fn q_encoding_malformed_hex_passthrough() {
    let decoded = decode_q_encoding("Hello=ZZWorld");
    assert_eq!(
        std::str::from_utf8(&decoded).unwrap(),
        "Hello=ZZWorld",
        "malformed =ZZ in Q-encoding must pass through literally"
    );
}

/// RFC 2047 Section 4.2: `=` at end of Q-encoded text with no following
/// hex digits should pass through as literal.
#[test]
fn q_encoding_trailing_equals() {
    let decoded = decode_q_encoding("Hello=");
    assert_eq!(
        std::str::from_utf8(&decoded).unwrap(),
        "Hello=",
        "trailing '=' in Q-encoding must pass through literally"
    );
}

/// RFC 2047 Sections 2 and 4.2 require a complete `=XX` escape inside a
/// Q-encoded word. A malformed encoded-word must be left literal rather
/// than partially decoded into a different header value.
#[test]
fn malformed_q_encoded_word_stays_literal() {
    let input = b"From: test@example.com\r\nSubject: =?utf-8?Q?a=b?=\r\n\r\nbody";
    let parsed = parse_email(input).unwrap();
    assert_eq!(
        parsed.subject.as_deref(),
        Some("=?utf-8?Q?a=b?="),
        "malformed RFC 2047 Q-encoded words must remain literal"
    );
}

// -----------------------------------------------------------------------
// Audit coverage: multipart/digest default Content-Type
// -----------------------------------------------------------------------

/// RFC 2046 Section 5.1.5: parts inside multipart/digest without an
/// explicit Content-Type must default to message/rfc822 (NOT text/plain).
/// Verify via a full `parse_email` round-trip.
#[test]
fn multipart_digest_default_content_type_full_email() {
    let raw = b"From: a@b.com\r\n\
                 Date: Thu, 13 Feb 2025 15:47:33 +0000\r\n\
                 Content-Type: multipart/digest; boundary=\"dg\"\r\n\
                 \r\n\
                 --dg\r\n\
                 \r\n\
                 From: nested@example.com\r\n\
                 Subject: Inner\r\n\
                 \r\n\
                 Inner body\r\n\
                 --dg--";

    let parsed = parse_email(raw).unwrap();
    // Part has no Content-Type → default is message/rfc822 in digest
    assert!(
        parsed.body_text.is_none(),
        "digest part must NOT be treated as text/plain"
    );
    assert_eq!(parsed.attachments.len(), 1);
    assert_eq!(parsed.attachments[0].content_type, "message/rfc822");
}

// -----------------------------------------------------------------------
// Coverage: header line starting with space (L208)
// -----------------------------------------------------------------------

/// RFC 5322 Section 2.2.3: a raw message starting with a space (a
/// continuation line before any header name) should be silently
/// skipped by the header parser, not crash or produce garbage.
#[test]
fn parse_headers_leading_space_skipped() {
    // The first line starts with a space, which is a continuation line
    // with no preceding header name. It must be silently ignored.
    let raw = b" continuation without header\r\nFrom: a@b.com\r\nDate: Thu, 13 Feb 2025 15:47:33 +0000\r\n\r\nBody";
    let parsed = parse_email(raw).unwrap();
    assert_eq!(parsed.from[0].email, "a@b.com");
    assert_eq!(parsed.body_text.as_deref(), Some("Body"));
}

/// A message starting with a tab (another form of continuation WSP)
/// before any header name must also be handled gracefully.
#[test]
fn parse_headers_leading_tab_skipped() {
    let raw = b"\tcontinuation without header\r\nFrom: a@b.com\r\nDate: Thu, 13 Feb 2025 15:47:33 +0000\r\n\r\n";
    let parsed = parse_email(raw).unwrap();
    assert_eq!(parsed.from[0].email, "a@b.com");
}

// -----------------------------------------------------------------------
// Coverage: RFC 2047 decoding edge cases (L491, L542, L550, L558-560)
// -----------------------------------------------------------------------

/// RFC 2047 Section 2: an encoded word with a bad base64 payload
/// should fail decoding and the literal `=?...?=` token should pass
/// through (Postel's law).
#[test]
fn encoded_word_bad_base64_passthrough() {
    // "=====" is not valid base64
    let input = "=?UTF-8?B?=====?=";
    let result = decode_encoded_words(input);
    // Should pass through as literal since base64 decode fails
    assert!(
        result.contains("=?"),
        "Bad base64 encoded word should pass through literally, got: {result:?}"
    );
}

/// RFC 2047 Section 2: an encoded word with an unknown charset should
/// still decode — `decode_charset` falls back to UTF-8 via `encoding_rs`.
#[test]
fn encoded_word_unknown_charset_fallback() {
    // "Hello" in base64 is "SGVsbG8="
    let input = "=?x-nonexistent-charset?B?SGVsbG8=?=";
    let result = decode_encoded_words(input);
    // encoding_rs falls back to UTF-8 for unknown charsets, so this
    // should decode the base64 payload and return something containing "Hello"
    assert!(
        result.contains("Hello"),
        "Unknown charset should fall back to UTF-8, got: {result:?}"
    );
}

/// RFC 2047: a truncated encoded word (missing closing `?=`) should
/// emit the literal `=?` prefix and continue.
#[test]
fn encoded_word_truncated_no_closing() {
    let input = "Start =?UTF-8?B?SGVsbG8= End";
    let result = decode_encoded_words(input);
    // The token is missing the closing "?=", so it cannot be decoded.
    // The "=?" should be emitted literally and parsing continues.
    assert!(
        result.contains("=?"),
        "Truncated encoded word should pass through, got: {result:?}"
    );
}

// -----------------------------------------------------------------------
// Coverage: RFC 2231 parameter continuation (L586, L592, L594)
// -----------------------------------------------------------------------

/// RFC 2231 Section 3: continuation parameters with mixed encoded
/// and plain sections. `name*0*=charset'lang'...; name*1=plain`
#[test]
fn rfc2231_continuation_mixed_encoded_and_plain() {
    let header = "attachment; filename*0*=UTF-8''r%C3%A9sum; filename*1=\"e.pdf\"";
    let result = extract_rfc2231_continuation(header, "filename");
    // Section 0 is encoded: r%C3%A9sum → "résum" (UTF-8)
    // Section 1 is plain: "e.pdf"
    assert_eq!(
        result.as_deref(),
        Some("r\u{e9}sume.pdf"),
        "RFC 2231 mixed encoded/plain continuation should reassemble correctly"
    );
}

/// RFC 2231 Section 3: continuation with three sections.
#[test]
fn rfc2231_continuation_three_sections() {
    let header =
        "attachment; filename*0=\"part1_\"; filename*1=\"part2_\"; filename*2=\"part3.pdf\"";
    let result = extract_rfc2231_continuation(header, "filename");
    assert_eq!(result.as_deref(), Some("part1_part2_part3.pdf"));
}

// -----------------------------------------------------------------------
// Coverage: extract_quoted_value / quoted-string edge cases
// (L605, L614-616, L619-621, L625, L634, L636, L647)
// -----------------------------------------------------------------------

/// RFC 5322 Section 3.2.4: an unterminated quoted-string (no closing
/// `"`) in a parameter value — the parser should read to end of string.
#[test]
fn extract_param_unterminated_quoted_value() {
    let header = "text/plain; charset=\"utf-8";
    let result = extract_param(header, "charset");
    // The closing quote is missing; find_closing_quote returns s.len(),
    // so the entire remainder becomes the value.
    assert_eq!(
        result.as_deref(),
        Some("utf-8"),
        "Unterminated quoted-string should extract to end of string"
    );
}

/// RFC 5322 Section 3.2.4: quoted-string with escaped characters.
#[test]
fn extract_param_quoted_with_backslash_escape() {
    let header = "attachment; filename=\"file\\\\name.txt\"";
    let result = extract_param(header, "filename");
    assert_eq!(
        result.as_deref(),
        Some("file\\name.txt"),
        "Backslash escape in quoted param value must be unescaped"
    );
}

/// RFC 5322 Section 3.2.4: empty quoted-string should return None.
#[test]
fn extract_param_empty_quoted_value() {
    let header = "attachment; filename=\"\"";
    let result = extract_param(header, "filename");
    assert!(
        result.is_none(),
        "Empty quoted-string value should return None, got: {result:?}"
    );
}

/// Regression test: an unterminated quote in a MIME parameter value
/// must NOT absorb subsequent parameters.  When the closing `"` is
/// missing, fall back to `;` as the value terminator — just like
/// the unquoted path — so later parameters remain discoverable.
///
/// # References
/// - RFC 5322 Section 3.2.4 (quoted-string)
/// - Postel's law (RFC 1122 Section 1.2.2)
#[test]
fn extract_param_unterminated_quote_does_not_absorb_next_param() {
    // charset has an unterminated quote; name= follows after ;
    let header = "text/plain; charset=\"utf-8; name=test.txt";
    let charset = extract_param(header, "charset");
    assert_eq!(
        charset.as_deref(),
        Some("utf-8"),
        "unterminated quote must fall back to ';' terminator \
         (RFC 5322 Section 3.2.4, Postel's law)"
    );
    // The name param must still be independently extractable
    let name = extract_param(header, "name");
    assert_eq!(
        name.as_deref(),
        Some("test.txt"),
        "parameter after unterminated quote must still be accessible"
    );
}

/// `extract_comment_text` with nested parentheses (RFC 5322 Section 3.2.2).
#[test]
fn extract_comment_text_nested_parens() {
    let result = extract_comment_text("(outer (inner) text)");
    assert_eq!(
        result.as_deref(),
        Some("outer (inner) text"),
        "Nested parens should be included in comment text"
    );
}

/// `extract_comment_text` with escaped characters.
#[test]
fn extract_comment_text_escaped_chars() {
    let result = extract_comment_text("(hello \\(world\\))");
    assert_eq!(
        result.as_deref(),
        Some("hello (world)"),
        "Escaped parens inside comments should be unescaped"
    );
}

/// `extract_comment_text` with empty parens returns None.
#[test]
fn extract_comment_text_empty() {
    let result = extract_comment_text("()");
    assert!(result.is_none(), "Empty comment should return None");
}

/// `extract_comment_text` with non-paren input returns None.
#[test]
fn extract_comment_text_no_paren() {
    let result = extract_comment_text("not a comment");
    assert!(
        result.is_none(),
        "Non-parenthesized input should return None"
    );
}

// -----------------------------------------------------------------------
// Coverage: comment stripping (L720, L1593-1604)
// -----------------------------------------------------------------------

/// RFC 5322 Section 3.2.2: `strip_comments` must handle nested comments,
/// escaped characters inside comments, and preserve text outside.
#[test]
fn strip_comments_nested_and_escaped() {
    // Nested comment with escaped paren inside
    let result = strip_comments("Hello (outer (inner) comment) World");
    assert_eq!(result, "Hello  World");

    // Escaped paren inside a comment — the `\)` is an escape, not end of comment
    let result = strip_comments("Hello (comment with \\) escaped) World");
    assert_eq!(result, "Hello  World");

    // Escaped backslash outside a comment should be preserved
    let result = strip_comments("Hello \\\\ World");
    assert_eq!(result, "Hello \\\\ World");

    // Backslash-escaped character inside comment should be consumed
    let result = strip_comments("Before (escaped \\( paren) After");
    assert_eq!(result, "Before  After");
}

/// RFC 5322 Section 3.2.2: escaped backslash outside a comment is preserved.
#[test]
fn strip_comments_escaped_outside_comment() {
    let result = strip_comments("no \\(comment\\) here");
    // The `\(` is outside a comment, so the backslash is preserved along
    // with the `(` which then opens a comment (since \ outside comment
    // is just pushed). Let's verify the actual behavior:
    // `\` sets escaped=true, depth==0 so push `\`. Then `(` is the
    // escaped char, pushed. Wait — let me re-read the code.
    // Actually: `\` is not the backslash case in strip_comments because
    // `\\` case only triggers when c=='\\'. Let me trace:
    // 'n','o',' ' — pushed (depth 0)
    // '\\' — escaped=true, depth==0 so push '\\'
    // '(' — escaped char: escaped=false, depth==0 so push '('
    // 'c','o','m','m','e','n','t' — pushed
    // '\\' — escaped=true, depth==0 so push '\\'
    // ')' — escaped char: escaped=false, depth==0 so push ')'
    // ' ','h','e','r','e' — pushed
    // Result: "no \\(comment\\) here" — no comments were stripped
    assert_eq!(
        result, "no \\(comment\\) here",
        "Escaped parens outside comments should not open/close comments"
    );
}

// -----------------------------------------------------------------------
// Coverage: date parsing edge cases (L767, L808, L834, L867)
// -----------------------------------------------------------------------

/// RFC 5322 Section 3.3: date with too few parts should return None.
#[test]
fn parse_date_too_few_parts() {
    // Only day and month, no year or time
    assert!(
        parse_rfc5322_date("13 Feb").is_none(),
        "Date with too few parts should return None"
    );
}

/// RFC 5322 Section 3.3: date with time field that has no colon
/// (not a valid time) should return None.
#[test]
fn parse_date_time_no_colon() {
    assert!(
        parse_rfc5322_date("13 Feb 2025 1547 +0000").is_none(),
        "Time without colon should return None"
    );
}

/// RFC 5322 Section 3.3: unknown month abbreviation returns None.
#[test]
fn parse_date_unknown_month() {
    assert!(
        parse_rfc5322_date("13 Foo 2025 12:00:00 +0000").is_none(),
        "Unknown month name should return None"
    );
}

/// RFC 5322 Section 3.3: completely malformed date string.
#[test]
fn parse_date_completely_malformed() {
    assert!(parse_rfc5322_date("not a date at all").is_none());
    assert!(parse_rfc5322_date("").is_none());
    assert!(parse_rfc5322_date("   ").is_none());
}

/// RFC 5322 Section 4.3: unknown timezone abbreviation defaults to +0000.
#[test]
fn parse_date_unknown_timezone_defaults_zero() {
    let dt = parse_rfc5322_date("13 Feb 2025 12:00:00 ZULU").unwrap();
    assert_eq!(
        dt.tz_offset_minutes, 0,
        "Unknown timezone abbreviation should default to +0000"
    );
}

/// RFC 5322 Section 3.3: non-numeric day should return None.
#[test]
fn parse_date_non_numeric_day() {
    assert!(
        parse_rfc5322_date("XX Feb 2025 12:00:00 +0000").is_none(),
        "Non-numeric day should return None"
    );
}

/// RFC 5322 Section 3.3: non-numeric year should return None.
#[test]
fn parse_date_non_numeric_year() {
    assert!(
        parse_rfc5322_date("13 Feb XXXX 12:00:00 +0000").is_none(),
        "Non-numeric year should return None"
    );
}

// -----------------------------------------------------------------------
// Coverage: boundary detection edge cases
// (L918, L954, L1003-1006, L1022-1023)
// -----------------------------------------------------------------------

/// RFC 2046 Section 5.1.1: boundary preceded by `\n` without `\r`
/// (bare LF line ending) should still be recognized.
#[test]
fn split_mime_parts_lf_only_boundaries() {
    let body = b"--boundary\nContent-Type: text/plain\n\nPart 1\n--boundary\nContent-Type: text/plain\n\nPart 2\n--boundary--";
    let parts = split_mime_parts(body, "boundary");
    assert_eq!(
        parts.len(),
        2,
        "Should find 2 parts with LF-only boundaries"
    );
}

/// RFC 2046 Section 5.1.1: boundary at the very start of body
/// (no preceding newline needed since it's position 0).
#[test]
fn split_mime_parts_boundary_at_start() {
    let body = b"--b\r\nContent-Type: text/plain\r\n\r\nOnly part\r\n--b--";
    let parts = split_mime_parts(body, "b");
    assert_eq!(
        parts.len(),
        1,
        "Should find 1 part when boundary is at start"
    );
    let text = String::from_utf8_lossy(parts[0]);
    assert!(text.contains("Only part"));
}

/// Boundary appearing mid-line must be ignored (RFC 2046 Section 5.1.1).
#[test]
fn split_mime_parts_midline_boundary_ignored() {
    let body = b"--b\r\nContent-Type: text/plain\r\n\r\nText mentioning --b in the middle\r\n--b--";
    let parts = split_mime_parts(body, "b");
    assert_eq!(parts.len(), 1, "Mid-line boundary must not split");
    let text = String::from_utf8_lossy(parts[0]);
    assert!(
        text.contains("--b in the middle"),
        "Mid-line boundary text should be preserved"
    );
}

/// Boundary line with trailing whitespace (spaces/tabs after the
/// boundary marker) must still be recognized (RFC 2046 Section 5.1.1:
/// "...followed by LWSP").
#[test]
fn split_mime_parts_boundary_with_trailing_whitespace() {
    let body = b"--b  \t\r\nContent-Type: text/plain\r\n\r\nBody text\r\n--b--";
    let parts = split_mime_parts(body, "b");
    assert_eq!(
        parts.len(),
        1,
        "Boundary with trailing whitespace should be recognized"
    );
}

/// Boundary not preceded by newline at a non-zero position must be
/// skipped (mid-line match).
#[test]
fn split_mime_parts_boundary_not_at_line_start_skipped() {
    // Body has the delimiter string embedded in content, not at line start
    let body = b"--bound\r\n\r\nSome text has --bound embedded\r\n--bound--";
    let parts = split_mime_parts(body, "bound");
    assert_eq!(parts.len(), 1);
    let text = String::from_utf8_lossy(parts[0]);
    assert!(text.contains("--bound embedded"));
}

// -----------------------------------------------------------------------
// Coverage: Content-Transfer-Encoding quoted value (L1130)
// -----------------------------------------------------------------------

/// RFC 2045 Section 6.1: CTE is formally a token, but some non-conformant
/// mailers quote it (e.g., `"base64"`). Per Postel's law we strip the
/// quotes and decode normally.
#[test]
fn parse_quoted_transfer_encoding() {
    let raw = b"From: a@b.com\r\n\
                 Date: Thu, 13 Feb 2025 15:47:33 +0000\r\n\
                 Content-Type: text/plain; charset=utf-8\r\n\
                 Content-Transfer-Encoding: \"base64\"\r\n\
                 \r\n\
                 SGVsbG8gV29ybGQ=\r\n";
    let parsed = parse_email(raw).unwrap();
    // Quoted CTE value must be decoded just like the unquoted form.
    assert_eq!(
        parsed.body_text.as_deref(),
        Some("Hello World"),
        "Quoted CTE 'base64' should decode the body correctly"
    );
}

/// Same as above but for quoted-printable CTE.
#[test]
fn parse_quoted_transfer_encoding_qp() {
    let raw = b"From: a@b.com\r\n\
                 Date: Thu, 13 Feb 2025 15:47:33 +0000\r\n\
                 Content-Type: text/plain; charset=utf-8\r\n\
                 Content-Transfer-Encoding: \"quoted-printable\"\r\n\
                 \r\n\
                 Hello=20World\r\n";
    let parsed = parse_email(raw).unwrap();
    assert_eq!(
        parsed.body_text.as_deref(),
        Some("Hello World"),
        "Quoted CTE 'quoted-printable' should decode the body correctly"
    );
}

/// Content-Transfer-Encoding with leading/trailing whitespace must be
/// handled (the `.trim()` call in `decode_transfer_encoding`).
#[test]
fn parse_transfer_encoding_with_whitespace() {
    let raw = b"From: a@b.com\r\n\
                 Date: Thu, 13 Feb 2025 15:47:33 +0000\r\n\
                 Content-Type: text/plain; charset=utf-8\r\n\
                 Content-Transfer-Encoding:  base64 \r\n\
                 \r\n\
                 SGVsbG8gV29ybGQ=\r\n";
    let parsed = parse_email(raw).unwrap();
    assert_eq!(
        parsed.body_text.as_deref(),
        Some("Hello World"),
        "CTE with whitespace should still decode correctly"
    );
}

/// RFC 2045 Section 6.1: CTE is a single token, but some broken mailers
/// append trailing parameters or comments (e.g., `base64; name=foo` or
/// `base64 (standard)`).  The parser must extract only the first token
/// before any semicolon or whitespace so decoding still works.
#[test]
fn parse_transfer_encoding_with_trailing_garbage() {
    let raw = b"From: a@b.com\r\n\
                 Date: Thu, 13 Feb 2025 15:47:33 +0000\r\n\
                 Content-Type: text/plain; charset=utf-8\r\n\
                 Content-Transfer-Encoding: base64; name=foo\r\n\
                 \r\n\
                 SGVsbG8gV29ybGQ=\r\n";
    let parsed = parse_email(raw).unwrap();
    assert_eq!(
        parsed.body_text.as_deref(),
        Some("Hello World"),
        "CTE with trailing semicolon+params should still decode as base64"
    );
}

/// Same trailing-garbage issue but with a parenthesized comment after
/// the encoding token, e.g., `quoted-printable (standard)`.
#[test]
fn parse_transfer_encoding_with_trailing_comment() {
    let raw = b"From: a@b.com\r\n\
                 Date: Thu, 13 Feb 2025 15:47:33 +0000\r\n\
                 Content-Type: text/plain; charset=utf-8\r\n\
                 Content-Transfer-Encoding: quoted-printable (standard)\r\n\
                 \r\n\
                 Hello=20World\r\n";
    let parsed = parse_email(raw).unwrap();
    assert_eq!(
        parsed.body_text.as_deref(),
        Some("Hello World"),
        "CTE with trailing comment should still decode as quoted-printable"
    );
}

/// Postel's law (RFC 1122 Section 1.2.2): tolerate non-conformant
/// senders that both quote the CTE token and append trailing junk after
/// it. RFC 2045 Section 6.1 defines CTE as a single unquoted token, but
/// in practice some generators emit forms like `"base64"; name=foo`.
#[test]
fn parse_quoted_transfer_encoding_with_trailing_garbage() {
    let raw = b"From: a@b.com\r\n\
                 Date: Thu, 13 Feb 2025 15:47:33 +0000\r\n\
                 Content-Type: text/plain; charset=utf-8\r\n\
                 Content-Transfer-Encoding: \"base64\"; name=foo\r\n\
                 \r\n\
                 SGVsbG8gV29ybGQ=\r\n";
    let parsed = parse_email(raw).unwrap();
    assert_eq!(
        parsed.body_text.as_deref(),
        Some("Hello World"),
        "Quoted CTE with trailing garbage should still decode as base64"
    );
}

// -----------------------------------------------------------------------
// Coverage: quoted-printable decoding edge cases
// (L1183-1184, L1193)
// -----------------------------------------------------------------------

/// RFC 2045 Section 6.7: soft line break `=\r\n` must be removed,
/// joining the lines without inserting any character.
#[test]
fn qp_soft_line_break_crlf() {
    let data = b"Hello=\r\n World";
    let decoded = decode_quoted_printable(data);
    assert_eq!(
        std::str::from_utf8(&decoded).unwrap(),
        "Hello World",
        "=\\r\\n soft break should be removed (RFC 2045 Section 6.7)"
    );
}

/// RFC 2045 Section 6.7: soft line break `=\n` (LF only, no CR)
/// must also be removed (Postel's law: accept bare LF).
#[test]
fn qp_soft_line_break_lf_only() {
    let data = b"Hello=\nWorld";
    let decoded = decode_quoted_printable(data);
    assert_eq!(
        std::str::from_utf8(&decoded).unwrap(),
        "HelloWorld",
        "=\\n soft break should be removed"
    );
}

/// RFC 2045 Section 6.7: `=\n` near end of data (only 2 bytes left)
/// must be treated as a soft break.
#[test]
fn qp_soft_break_lf_at_end() {
    let data = b"Hi=\n";
    let decoded = decode_quoted_printable(data);
    assert_eq!(
        std::str::from_utf8(&decoded).unwrap(),
        "Hi",
        "=\\n at end of data should be a soft break"
    );
}

/// RFC 2045 Section 6.7: invalid hex after `=` should pass through
/// the `=` as a literal (Postel's law). Only valid hex pairs are decoded.
#[test]
fn qp_invalid_hex_passthrough() {
    let data = b"=GG=4F=4B";
    let decoded = decode_quoted_printable(data);
    // =GG is invalid hex → pass through literally; =4F=4B → "OK"
    assert_eq!(
        std::str::from_utf8(&decoded).unwrap(),
        "=GGOK",
        "Invalid hex =GG should pass through, valid =4F=4B should decode"
    );
}

// -----------------------------------------------------------------------
// Coverage: base64 content length edge cases (L1285, L1343)
// -----------------------------------------------------------------------

/// RFC 2045 Section 6.8: empty base64 body should produce empty output.
#[test]
fn base64_empty_body() {
    let decoded = decode_transfer_encoding(b"", "base64");
    assert!(
        decoded.is_empty(),
        "Empty base64 input should produce empty output"
    );
}

/// RFC 2045 Section 6.8: base64 body that's only whitespace should
/// produce empty output after stripping non-alphabet chars.
#[test]
fn base64_whitespace_only() {
    let decoded = decode_transfer_encoding(b"  \r\n  \r\n", "base64");
    assert!(
        decoded.is_empty(),
        "Whitespace-only base64 input should produce empty output"
    );
}

/// `find_closing_quote`: unterminated quoted-string returns string length.
#[test]
fn find_closing_quote_unterminated() {
    assert_eq!(find_closing_quote("no closing quote here"), 21);
}

/// `find_closing_quote`: escaped quote is skipped.
#[test]
fn find_closing_quote_skips_escaped() {
    // `\\\"` is an escaped quote, should be skipped; real close is at end
    assert_eq!(find_closing_quote("hello\\\"world\""), 12);
}

// -----------------------------------------------------------------------
// Coverage: multipart boundary search paths
// (L1361-1362, L1400, L1404, L1429-1430, L1478-1479, L1544-1545,
//  L1550, L1552, L1567)
// -----------------------------------------------------------------------

/// Multiple parts with CRLF before boundaries — exercise the CRLF
/// stripping logic in `split_mime_parts` (pos >= 2 and body[pos-2] == CR).
#[test]
fn multipart_crlf_before_boundary() {
    let raw = b"From: a@b.com\r\n\
                 Date: Thu, 13 Feb 2025 15:47:33 +0000\r\n\
                 Content-Type: multipart/mixed; boundary=\"mp\"\r\n\
                 \r\n\
                 --mp\r\n\
                 Content-Type: text/plain\r\n\
                 \r\n\
                 Part A\r\n\
                 --mp\r\n\
                 Content-Type: text/html\r\n\
                 \r\n\
                 <b>Part B</b>\r\n\
                 --mp--";
    let parsed = parse_email(raw).unwrap();
    assert_eq!(parsed.body_text.as_deref(), Some("Part A"));
    assert_eq!(parsed.body_html.as_deref(), Some("<b>Part B</b>"));
}

/// Multiple parts with bare LF before boundaries — exercise the LF-only
/// stripping logic (pos >= 1 && body[pos-1] == LF, but pos < 2 or
/// body[pos-2] != CR).
#[test]
fn multipart_lf_only_before_boundary() {
    let raw = b"From: a@b.com\nDate: Thu, 13 Feb 2025 15:47:33 +0000\nContent-Type: multipart/mixed; boundary=\"mp\"\n\n--mp\nContent-Type: text/plain\n\nPart A\n--mp\nContent-Type: text/html\n\n<b>Part B</b>\n--mp--";
    let parsed = parse_email(raw).unwrap();
    assert_eq!(parsed.body_text.as_deref(), Some("Part A"));
    assert_eq!(parsed.body_html.as_deref(), Some("<b>Part B</b>"));
}

/// Truncated multipart: no closing boundary — the trailing content
/// should still be captured as a part.
#[test]
fn multipart_truncated_no_closing() {
    let body = b"--bnd\r\n\r\nFirst part\r\n--bnd\r\n\r\nSecond part with no closing boundary";
    let parts = split_mime_parts(body, "bnd");
    assert_eq!(
        parts.len(),
        2,
        "Should find 2 parts even without closing boundary"
    );
    let text2 = String::from_utf8_lossy(parts[1]);
    assert!(text2.contains("Second part"));
}

// -----------------------------------------------------------------------
// Coverage: comment removal with escapes (L1593-1604)
// -----------------------------------------------------------------------

/// RFC 5322 Section 3.2.2: backslash inside a comment escapes the
/// next character. `\(` inside a comment should NOT increment depth.
#[test]
fn strip_comments_escaped_paren_inside_comment() {
    let result = strip_comments("Before (escaped \\) paren) After");
    // The `\)` inside the comment is an escape: the `)` is consumed but
    // doesn't decrement depth. The real closing `)` comes later.
    assert_eq!(
        result.trim(),
        "Before  After",
        "Escaped close-paren inside comment must not end the comment"
    );
}

/// RFC 5322 Section 3.2.2: escaped open-paren inside a comment
/// should NOT increment the nesting depth.
#[test]
fn strip_comments_escaped_open_paren_inside_comment() {
    let result = strip_comments("X (comment \\( not nested) Y");
    assert_eq!(
        result.trim(),
        "X  Y",
        "Escaped open-paren inside comment must not increase nesting depth"
    );
}

/// RFC 5322 Section 3.2.2 / 3.2.4: parentheses inside a quoted-string
/// are literal characters — they must NOT be treated as comment delimiters.
#[test]
fn strip_comments_respects_quoted_strings() {
    let result = strip_comments("\"value (not a comment)\" rest");
    assert_eq!(
        result, "\"value (not a comment)\" rest",
        "Parentheses inside quoted-string must not be treated as comments"
    );
}

/// RFC 5322 Section 3.2.4: parentheses inside a quoted-string with
/// surrounding real comments should be preserved inside quotes.
#[test]
fn strip_comments_quoted_string_with_real_comment() {
    let result = strip_comments("\"value (literal)\" (real comment) end");
    assert_eq!(
        result, "\"value (literal)\"  end",
        "Real comment stripped, quoted parens preserved"
    );
}

// -----------------------------------------------------------------------
// Coverage: hex_digit lowercase (L1632)
// -----------------------------------------------------------------------

/// Lowercase hex digits a-f must be recognized in QP and percent decoding.
#[test]
fn hex_digit_lowercase() {
    assert_eq!(hex_digit(b'a'), Some(10));
    assert_eq!(hex_digit(b'f'), Some(15));
    assert_eq!(hex_digit(b'c'), Some(12));
    // Also verify uppercase and digits still work
    assert_eq!(hex_digit(b'A'), Some(10));
    assert_eq!(hex_digit(b'F'), Some(15));
    assert_eq!(hex_digit(b'0'), Some(0));
    assert_eq!(hex_digit(b'9'), Some(9));
    // Invalid
    assert_eq!(hex_digit(b'g'), None);
    assert_eq!(hex_digit(b'G'), None);
    assert_eq!(hex_digit(b' '), None);
}

/// QP decoding with lowercase hex digits (RFC 2045 Section 6.7).
#[test]
fn qp_lowercase_hex_digits() {
    // =c3=a9 is UTF-8 for U+00E9 (e-acute) in UTF-8 encoding
    let data = b"caf=c3=a9";
    let decoded = decode_quoted_printable(data);
    assert_eq!(decoded, b"caf\xc3\xa9");
    let text = String::from_utf8_lossy(&decoded);
    assert_eq!(
        text, "caf\u{e9}",
        "Lowercase hex digits in QP should decode correctly (RFC 2045 Section 6.7)"
    );
}

/// `decode_hex_pair` with lowercase hex digits.
#[test]
fn decode_hex_pair_lowercase() {
    assert_eq!(decode_hex_pair(b'f', b'f'), Some(0xFF));
    assert_eq!(decode_hex_pair(b'a', b'0'), Some(0xA0));
    assert_eq!(decode_hex_pair(b'0', b'a'), Some(0x0A));
}

// -----------------------------------------------------------------------
// Coverage: parse_single_address edge cases
// -----------------------------------------------------------------------

/// Address with empty angle brackets (no email inside `<>`) should
/// return None.
#[test]
fn parse_single_address_empty_angle_brackets() {
    let result = parse_single_address("Display Name <>");
    assert!(
        result.is_none(),
        "Empty angle brackets should not produce an address"
    );
}

/// Address with `>` before `<` (malformed) should not match the
/// angle bracket path.
#[test]
fn parse_single_address_reversed_angles() {
    let result = parse_single_address(">bad<user@example.com");
    // rfind('<') finds position of '<', rfind('>') finds position of last '>'
    // angle_end (0) < angle_start (4), so angle bracket path fails.
    // Falls through to bare email check (contains '@').
    assert!(result.is_some());
}

/// Input without `@` and without angle brackets should return None.
#[test]
fn parse_single_address_no_at_no_brackets() {
    let result = parse_single_address("just plain text");
    assert!(
        result.is_none(),
        "Text without @ or <> should not produce an address"
    );
}

// -----------------------------------------------------------------------
// Coverage: is_inside_quotes
// -----------------------------------------------------------------------

/// `is_inside_quotes` must correctly track escaped quotes.
#[test]
fn is_inside_quotes_with_escapes() {
    // Position 15 is after the escaped quote (\"), still inside quotes
    assert!(is_inside_quotes("\"hello \\\" world\"end", 15));
    // Position 0 is before any quotes — not inside
    assert!(!is_inside_quotes("\"hello\"", 0));
    // Position after closing quote — not inside
    assert!(!is_inside_quotes("\"hello\" world", 8));
}

/// RFC 5322 Section 3.2.4: `quoted-pair` (backslash escape) is only valid
/// inside `quoted-string` and `comment`, not outside them. A backslash
/// outside quotes must be treated as a literal character, not as an escape
/// that consumes the next byte.
#[test]
fn is_inside_quotes_backslash_outside_quotes_is_literal() {
    // Input: x=\"y"z"  — backslash is outside quotes, so it does NOT
    // escape the first `"`. The first `"` opens a quoted string, `y`
    // is inside, the second `"` closes it, and `z` is outside.
    // Position of `z` (index 5) should NOT be inside quotes.
    let s = r#"x=\"y"z""#;
    // Byte layout: x=\" y " z "
    //              0123 4 5 6 7
    // After the fix: `\` at index 2 is literal (outside quotes).
    // `"` at index 3 opens a quoted string, `y` is inside, `"` at
    // index 5 closes it. `z` at index 6 is outside.
    assert!(
        !is_inside_quotes(s, 6),
        "RFC 5322 Section 3.2.4: backslash outside quotes must not escape \
         the next character — position 6 ('z') should be outside quotes"
    );
}

// -----------------------------------------------------------------------
// Coverage: strip_outer_quotes
// -----------------------------------------------------------------------

/// `strip_outer_quotes` must not strip when input is too short.
#[test]
fn strip_outer_quotes_short_input() {
    assert_eq!(strip_outer_quotes("\""), "\"");
    assert_eq!(strip_outer_quotes(""), "");
    assert_eq!(strip_outer_quotes("x"), "x");
}

/// `strip_outer_quotes` must not strip when only one side has quotes.
#[test]
fn strip_outer_quotes_one_sided() {
    assert_eq!(strip_outer_quotes("\"hello"), "\"hello");
    assert_eq!(strip_outer_quotes("hello\""), "hello\"");
}

// -----------------------------------------------------------------------
// Coverage: split_header_body edge case — starts with \n
// -----------------------------------------------------------------------

/// RFC 2046: a MIME part starting with bare `\n` (no headers).
#[test]
fn split_header_body_starts_with_lf() {
    let (headers, body) = split_header_body(b"\nBody text here");
    assert!(
        headers.is_empty(),
        "Headers should be empty when input starts with \\n"
    );
    assert_eq!(body, b"Body text here");
}

/// RFC 2046: a MIME part starting with `\r\n` (no headers).
#[test]
fn split_header_body_starts_with_crlf() {
    let (headers, body) = split_header_body(b"\r\nBody text here");
    assert!(
        headers.is_empty(),
        "Headers should be empty when input starts with \\r\\n"
    );
    assert_eq!(body, b"Body text here");
}

// -----------------------------------------------------------------------
// Coverage: CTE decode_body stripping trailing LF (L1130)
// -----------------------------------------------------------------------

/// `decode_body` strips a trailing `\n` (bare LF without CR).
#[test]
fn decode_body_strips_trailing_lf_only() {
    let result = decode_body(b"Hello\n", "", "text/plain; charset=utf-8");
    assert_eq!(result, "Hello", "Trailing bare LF should be stripped");
}

/// `decode_body` with no trailing newline returns content as-is.
#[test]
fn decode_body_no_trailing_newline() {
    let result = decode_body(b"Hello", "", "text/plain; charset=utf-8");
    assert_eq!(
        result, "Hello",
        "No trailing newline should leave content unchanged"
    );
}

/// Percent decoding with lowercase hex (RFC 2231 / RFC 3986).
#[test]
fn percent_decode_lowercase_hex() {
    let decoded = percent_decode("%c3%a9");
    // %c3%a9 is UTF-8 for U+00E9 (e-acute)
    assert_eq!(decoded, vec![0xC3, 0xA9]);
}

/// Percent decoding with invalid hex passes through literally.
#[test]
fn percent_decode_invalid_hex() {
    let decoded = percent_decode("%ZZ");
    assert_eq!(decoded, b"%ZZ");
}

/// Percent decoding with truncated sequence passes through.
#[test]
fn percent_decode_truncated() {
    let decoded = percent_decode("hello%2");
    assert_eq!(decoded, b"hello%2");
}

// -----------------------------------------------------------------------
// Coverage: address with colon that looks like group but has @
// -----------------------------------------------------------------------

/// RFC 5322 Section 3.4: a colon in an address that contains `@`
/// should NOT be treated as group syntax (heuristic).
#[test]
fn parse_address_colon_with_at_sign() {
    // "user:tag@example.com" has a colon but also @ — the colon
    // should be treated as part of the local-part, not group syntax.
    let addrs = parse_address_list("user:tag@example.com");
    // The heuristic checks if current.trim().contains('@') when seeing ':'.
    // At the point of ':', current is "user" (no @), so it enters group mode.
    // Then "tag@example.com" is parsed as a group member.
    assert!(!addrs.is_empty(), "Should parse at least one address");
}

/// RFC 2231 Section 4 + Postel's law: non-conformant mailers may
/// quote RFC 2231 extended values (`filename*="utf-8''%C3%A9.txt"`).
/// The parser must strip outer quotes and decode correctly.
#[test]
fn extract_rfc2231_param_handles_quoted_value() {
    // A non-conformant mailer wraps the RFC 2231 value in double-quotes.
    // The correct unquoted form is: filename*=utf-8''caf%C3%A9.txt
    let disposition = "attachment; filename*=\"utf-8''caf%C3%A9.txt\"";
    let filename = extract_filename(disposition, "application/octet-stream");
    assert_eq!(
        filename.as_deref(),
        Some("café.txt"),
        "RFC 2231 filename with outer quotes must be decoded correctly \
         (Postel's law: be liberal in what you accept)"
    );
}

/// RFC 2231 Section 4 / Postel's law: when a non-conformant mailer wraps
/// an RFC 2231 extended value in double-quotes AND the value contains a
/// semicolon (either percent-encoded `%3B` that decodes to `;`, or a raw
/// `;` inside quotes), the parser must not truncate the value at the
/// semicolon. The closing quote must be found first to delimit the value.
#[test]
fn extract_rfc2231_param_quoted_value_with_semicolon() {
    // Quoted value with a literal semicolon: a non-conformant mailer
    // wraps the RFC 2231 value in quotes and includes a raw semicolon
    // inside the quoted value.
    let disposition = "attachment; filename*=\"UTF-8''file;name.txt\"";
    let result = extract_rfc2231_param(disposition, "filename");
    assert_eq!(
        result.as_deref(),
        Some("file;name.txt"),
        "RFC 2231 quoted value containing a literal semicolon must not \
         be truncated at the semicolon (Postel's law)"
    );
}

/// RFC 2231 Section 4 + Postel's law: a malformed quoted RFC 2231
/// value with no closing quote must not swallow subsequent
/// parameters. The parser should stop at the next semicolon so later
/// parameters remain outside the filename.
#[test]
fn extract_rfc2231_param_unterminated_quoted_value_stops_at_separator() {
    let disposition = "attachment; filename*=\"UTF-8''report.txt; size=123";
    let result = extract_rfc2231_param(disposition, "filename");
    assert_eq!(
        result.as_deref(),
        Some("report.txt"),
        "unterminated quoted RFC 2231 values must stop at the next \
         parameter separator instead of swallowing later parameters"
    );
}

/// RFC 2231 Section 4 defines `filename*=` as an ext-value with valid
/// percent-encoding. When the extended value is malformed and unusable,
/// a valid plain `filename=` fallback must remain usable per Postel's law.
#[test]
fn extract_filename_malformed_rfc2231_param_falls_back_to_plain_filename() {
    let disposition = "attachment; filename*=UTF-8''report%ZZ.txt; filename=\"report.txt\"";
    let filename = extract_filename(disposition, "application/octet-stream");
    assert_eq!(
        filename.as_deref(),
        Some("report.txt"),
        "malformed RFC 2231 filename* must not override a valid plain filename fallback"
    );
}

/// RFC 5322 Section 3.4: group syntax `display-name ":" [group-list] ";"`
/// must be recognized even when the display-name is a quoted-string
/// containing `@`. The `@` inside quotes is not an addr-spec indicator.
#[test]
fn group_address_with_at_in_quoted_display_name() {
    let input = r#""user@host": addr@example.com;"#;
    let addrs = parse_address_list(input);
    assert_eq!(
        addrs.len(),
        1,
        "group with quoted display-name containing '@' must parse the member address"
    );
    assert_eq!(addrs[0].email, "addr@example.com");
}

/// RFC 5322 Section 4.4: obsolete source-route addresses like
/// `<@host1,@host2:user@domain>` must have the route prefix stripped,
/// yielding `user@domain` as the addr-spec.
#[test]
fn obs_route_stripped_from_angle_bracket_address() {
    // Single-hop route
    let addrs = parse_address_list("<@relay:user@example.com>");
    assert_eq!(addrs.len(), 1);
    assert_eq!(addrs[0].email, "user@example.com");

    // Multi-hop route
    let addrs = parse_address_list("<@hop1,@hop2:alice@domain.org>");
    assert_eq!(addrs.len(), 1);
    assert_eq!(addrs[0].email, "alice@domain.org");

    // With display name
    let addrs = parse_address_list("Alice <@relay:alice@example.com>");
    assert_eq!(addrs.len(), 1);
    assert_eq!(addrs[0].name.as_deref(), Some("Alice"));
    assert_eq!(addrs[0].email, "alice@example.com");
}

/// RFC 2231 Section 4: the first encoded continuation section (`param*0*=`)
/// must follow the `charset'language'value` format. When the value lacks the
/// required apostrophes, the section is malformed and must be treated as
/// non-encoded (raw bytes used as-is) rather than percent-decoded.
#[test]
fn rfc2231_continuation_malformed_first_encoded_section_not_percent_decoded() {
    // filename*0*= without charset'language' apostrophes — malformed per RFC 2231.
    // The raw value "hello%20world.txt" must be kept as-is, not percent-decoded
    // to "hello world.txt".
    let header = "attachment; filename*0*=hello%20world.txt";
    let result = extract_rfc2231_continuation(header, "filename");
    assert_eq!(
        result.as_deref(),
        Some("hello%20world.txt"),
        "malformed charset'language'value must not be percent-decoded"
    );
}

/// RFC 2231 Sections 3-4 do not define duplicate continuation sections,
/// but Postel's law says a malformed encoded duplicate must not override
/// an earlier plain section that is already usable.
#[test]
fn rfc2231_continuation_malformed_encoded_duplicate_does_not_override_plain() {
    let header =
        "attachment; filename*0=\"report\"; filename*0*=broken%20value; filename*1=\".txt\"";
    let result = extract_rfc2231_continuation(header, "filename");
    assert_eq!(
        result.as_deref(),
        Some("report.txt"),
        "malformed encoded duplicate section must not override the first valid plain section"
    );
}

/// RFC 5322 Section 3.3: zone = ("+" / "-") 4DIGIT where the digits
/// represent HHMM.  Offsets with hours > 23 or minutes > 59 are
/// semantically invalid and would produce a >4-digit zone field when
/// formatted back via `to_rfc5322_string()`.  They must be treated as
/// unknown and map to +0000 per RFC 5322 Section 4.3.
#[test]
fn test_invalid_timezone_offset_defaults_to_zero() {
    let input = b"From: test@example.com\r\nDate: Thu, 13 Feb 2025 12:00:00 +9999\r\n\r\n";
    let email = parse_email(input).unwrap();
    // RFC 5322 Section 4.3: unknown/invalid timezone -> +0000
    assert_eq!(email.date.unwrap().tz_offset_minutes, 0);
}

/// RFC 5322 Section 3.3: valid edge-case timezone offsets must still
/// parse correctly after the range check is added.
#[test]
fn test_valid_edge_case_timezone_offsets() {
    // +1400 (Line Islands, max real-world offset) -> 14*60 + 0 = 840
    let input = b"From: test@example.com\r\nDate: Thu, 13 Feb 2025 12:00:00 +1400\r\n\r\n";
    let email = parse_email(input).unwrap();
    assert_eq!(email.date.unwrap().tz_offset_minutes, 840);

    // -1200 (Baker Island, min real-world offset) -> -(12*60 + 0) = -720
    let input = b"From: test@example.com\r\nDate: Thu, 13 Feb 2025 12:00:00 -1200\r\n\r\n";
    let email = parse_email(input).unwrap();
    assert_eq!(email.date.unwrap().tz_offset_minutes, -720);

    // +0000 (UTC) -> 0
    let input = b"From: test@example.com\r\nDate: Thu, 13 Feb 2025 12:00:00 +0000\r\n\r\n";
    let email = parse_email(input).unwrap();
    assert_eq!(email.date.unwrap().tz_offset_minutes, 0);
}

/// RFC 5322 Section 3.3 defines a numeric zone as exactly sign + 4DIGIT.
/// Overlong offsets such as `+12345` are malformed and must not be
/// partially interpreted as `+1234`; they degrade to unknown local time
/// (`+0000`) under the parser's Postel-style fallback.
#[test]
fn test_overlong_timezone_offset_defaults_to_zero() {
    let input = b"From: test@example.com\r\nDate: Thu, 13 Feb 2025 12:00:00 +12345\r\n\r\n";
    let email = parse_email(input).unwrap();
    assert_eq!(
        email.date.unwrap().tz_offset_minutes,
        0,
        "zone = sign 4DIGIT, so +12345 must not be truncated to +1234 \
         (RFC 5322 Section 3.3)"
    );
}

/// RFC 5322 Section 4.3: military single-letter time zones (A–I, K–Y)
/// "SHOULD all be considered equivalent to '-0000'" because their
/// historical usage is unreliable.  Named civil zones (EST, PDT, etc.)
/// retain their well-known offsets.
#[test]
fn test_military_timezone_codes_treated_as_unknown() {
    // Military zone 'A' was nominally +0100, but RFC 5322 Section 4.3
    // says it should be treated as -0000 (offset 0).
    let input = b"From: test@example.com\r\n\
                  Date: Thu, 13 Feb 2025 12:00:00 A\r\n\r\n";
    let email = parse_email(input).unwrap();
    assert_eq!(
        email.date.unwrap().tz_offset_minutes,
        0,
        "military zone 'A' must be treated as -0000 (RFC 5322 Section 4.3)"
    );

    // Military zone 'N' was nominally -0100.
    let input = b"From: test@example.com\r\n\
                  Date: Thu, 13 Feb 2025 12:00:00 N\r\n\r\n";
    let email = parse_email(input).unwrap();
    assert_eq!(
        email.date.unwrap().tz_offset_minutes,
        0,
        "military zone 'N' must be treated as -0000 (RFC 5322 Section 4.3)"
    );

    // Military zone 'M' was nominally +1200.
    let input = b"From: test@example.com\r\n\
                  Date: Thu, 13 Feb 2025 12:00:00 M\r\n\r\n";
    let email = parse_email(input).unwrap();
    assert_eq!(
        email.date.unwrap().tz_offset_minutes,
        0,
        "military zone 'M' must be treated as -0000 (RFC 5322 Section 4.3)"
    );

    // Named civil zones like EDT still have well-known meanings.
    let input = b"From: test@example.com\r\n\
                  Date: Thu, 13 Feb 2025 12:00:00 EDT\r\n\r\n";
    let email = parse_email(input).unwrap();
    assert_eq!(
        email.date.unwrap().tz_offset_minutes,
        -240,
        "EDT is a well-known civil zone and should still resolve to -0400"
    );
}

#[test]
fn spec_audit_date_accepts_calendar_invalid_days() {
    // RFC 5322 Section 3.3: day = 1*2DIGIT, grammar allows 1-31 for any month.
    // Postel's law: calendar-impossible dates are syntactically valid and
    // appear in real-world email, so we accept them rather than rejecting.
    assert!(parse_rfc5322_date("Thu, 31 Feb 2025 12:00:00 +0000").is_some());
    assert!(parse_rfc5322_date("Mon, 31 Apr 2025 12:00:00 +0000").is_some());
    assert!(parse_rfc5322_date("Mon, 31 Jun 2025 12:00:00 +0000").is_some());
    assert!(parse_rfc5322_date("Mon, 29 Feb 2023 12:00:00 +0000").is_some()); // non-leap
    assert!(parse_rfc5322_date("Thu, 29 Feb 2024 12:00:00 +0000").is_some()); // leap year
    assert!(parse_rfc5322_date("Tue, 28 Feb 2023 12:00:00 +0000").is_some());
}

#[test]
fn spec_audit_date_accepts_mismatched_day_of_week() {
    // RFC 5322 Section 3.3: when day-of-week is present, it MUST be the
    // day implied by the date. However, this is a generation constraint.
    // Per Postel's law (RFC 1122 Section 1.2.2), the parser accepts
    // mismatched day-of-week since real-world mailers frequently get it
    // wrong. The numeric date fields are authoritative.
    // 2025-02-13 is Thursday, not Monday — parser must still accept it.
    let result = parse_rfc5322_date("Mon, 13 Feb 2025 12:00:00 +0000");
    assert!(
        result.is_some(),
        "Parser must accept dates with incorrect day-of-week (Postel's law)"
    );
    let dt = result.unwrap();
    assert_eq!(dt.year, 2025);
    assert_eq!(dt.month, 2);
    assert_eq!(dt.day, 13);
}

#[test]
fn spec_audit_encoded_word_in_quoted_string_not_decoded() {
    // RFC 2047 Section 5: encoded-words MUST NOT appear in quoted-strings.
    let raw = b"From: \"=?UTF-8?B?SGVsbG8=?=\" <test@example.com>\r\n\
                Date: Thu, 13 Feb 2025 15:47:33 +0000\r\n\r\n";
    let parsed = parse_email(raw).unwrap();
    assert_eq!(
        parsed.from[0].name.as_deref(),
        Some("=?UTF-8?B?SGVsbG8=?="),
        "Encoded words inside quoted-strings must NOT be decoded (RFC 2047 Section 5)"
    );
}

#[test]
fn spec_audit_encoded_word_in_unquoted_phrase_decoded() {
    // RFC 2047 Section 5: encoded-words in phrase context ARE decoded.
    let raw = b"From: =?UTF-8?B?SGVsbG8=?= <test@example.com>\r\n\
                Date: Thu, 13 Feb 2025 15:47:33 +0000\r\n\r\n";
    let parsed = parse_email(raw).unwrap();
    assert_eq!(parsed.from[0].name.as_deref(), Some("Hello"));
}

#[test]
fn spec_audit_overlong_encoded_word_decoded() {
    // RFC 2047 Section 2: the 75-character limit is a generation constraint.
    // Per Postel's law (RFC 1122 Section 1.2.2), overlong encoded words are
    // decoded since real-world mailers frequently exceed the limit.
    let raw = b"Subject: =?UTF-8?Q?AAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAA?=\r\n\
                From: test@example.com\r\n\
                Date: Thu, 13 Feb 2025 15:47:33 +0000\r\n\r\n";
    let parsed = parse_email(raw).unwrap();
    assert_eq!(
        parsed.subject.as_deref(),
        Some("AAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAA"),
        "Overlong encoded-word must be decoded (Postel's law, RFC 2047 Section 6.1)"
    );
}

#[test]
fn extract_param_case_insensitive_param_name() {
    // RFC 2045 Section 5.1: parameter names are case-insensitive.
    // The header value is lowered internally, so an uppercase param_name
    // must also be lowered to match.
    let header = "text/plain; charset=utf-8";
    assert_eq!(
        extract_param(header, "CHARSET"),
        Some("utf-8".to_string()),
        "extract_param must match param_name case-insensitively (RFC 2045 Section 5.1)"
    );
    // Mixed case should also work
    assert_eq!(
        extract_param(header, "Charset"),
        Some("utf-8".to_string()),
        "extract_param must match mixed-case param_name (RFC 2045 Section 5.1)"
    );
}

#[test]
fn extract_rfc2231_param_case_insensitive_param_name() {
    // RFC 2045 Section 5.1: parameter names are case-insensitive.
    let header = "attachment; filename*=utf-8''hello%20world.txt";
    assert_eq!(
        extract_rfc2231_param(header, "FILENAME"),
        Some("hello world.txt".to_string()),
        "extract_rfc2231_param must match param_name case-insensitively (RFC 2045 Section 5.1)"
    );
}

#[test]
fn extract_rfc2231_continuation_case_insensitive_param_name() {
    // RFC 2045 Section 5.1: parameter names are case-insensitive.
    let header = "attachment; filename*0=\"hello \"; filename*1=\"world.txt\"";
    assert_eq!(
        extract_rfc2231_continuation(header, "FILENAME"),
        Some("hello world.txt".to_string()),
        "extract_rfc2231_continuation must match param_name case-insensitively \
         (RFC 2045 Section 5.1)"
    );
}

/// RFC 2231 Section 4: `decode_body()` must recognize the RFC 2231
/// encoded charset form `charset*=charset'lang'value` when decoding
/// body parts. Without this, the body falls back to US-ASCII and
/// non-ASCII content is garbled.
#[test]
fn decode_body_rfc2231_charset_star_encoded() {
    // Body is UTF-8 encoded "café" (c3 a9 for é)
    let body = b"caf\xc3\xa9";
    let ct = "text/plain; charset*=UTF-8''utf-8";
    let result = decode_body(body, "8bit", ct);
    assert_eq!(
        result, "caf\u{e9}",
        "RFC 2231 Section 4: charset*=UTF-8''utf-8 must be recognized \
         by decode_body so non-ASCII bodies are decoded correctly"
    );
}

/// RFC 2231 Section 3: `decode_body()` must recognize the RFC 2231
/// continuation form `charset*0=...; charset*1=...` when decoding
/// body parts. Without this, the charset is not assembled and
/// non-ASCII content is garbled.
#[test]
fn decode_body_rfc2231_charset_continuation() {
    // Body is UTF-8 encoded "café" (c3 a9 for é)
    let body = b"caf\xc3\xa9";
    let ct = "text/plain; charset*0=ut; charset*1=f-8";
    let result = decode_body(body, "8bit", ct);
    assert_eq!(
        result, "caf\u{e9}",
        "RFC 2231 Section 3: charset*0=ut; charset*1=f-8 must be \
         assembled and recognized by decode_body"
    );
}

/// RFC 2045 Section 5.1: after stripping comments from the type/subtype,
/// whitespace around the `/` separator must be normalized. Input like
/// `text (comment) /plain` or `text/ plain` should yield `"text/plain"`,
/// not `"text  /plain"` or `"text/ plain"`.
#[test]
fn extract_mime_type_normalizes_whitespace_around_slash() {
    // Whitespace left by comment stripping before the slash
    assert_eq!(
        extract_mime_type("text (comment) /plain"),
        "text/plain",
        "whitespace before '/' (left by comment stripping) must be removed"
    );
    // Whitespace after the slash
    assert_eq!(
        extract_mime_type("text/ plain"),
        "text/plain",
        "whitespace after '/' must be removed"
    );
    // Whitespace on both sides of the slash
    assert_eq!(
        extract_mime_type("text / plain"),
        "text/plain",
        "whitespace around '/' must be removed"
    );
    // Normal input without extra whitespace — must still work
    assert_eq!(extract_mime_type("text/plain; charset=utf-8"), "text/plain");
}

/// RFC 2231 Section 3: non-contiguous continuation indices are malformed.
/// Preserve only the contiguous prefix beginning at `*0`.
#[test]
fn extract_rfc2231_continuation_non_contiguous_sections() {
    let header = "attachment; filename*0=\"hello \"; filename*2=\"world.txt\"";
    let result = extract_rfc2231_continuation(header, "filename");
    assert_eq!(
        result,
        Some("hello ".to_string()),
        "RFC 2231 Section 3: reassembly must stop before section 2 when section 1 is missing"
    );
}

/// RFC 2231 Section 3 still forbids very large gaps. The scanner may see
/// a later segment, but reassembly must retain only the contiguous prefix.
#[test]
fn rfc2231_gap_larger_than_three() {
    let header = "attachment; filename*0=\"hello\"; filename*5=\"world\"";
    let result = extract_rfc2231_continuation(header, "filename");
    assert_eq!(
        result,
        Some("hello".to_string()),
        "RFC 2231 Section 3: reassembly must stop at the first missing section"
    );
}

/// RFC 2231 Section 3: when section 0 is plain (non-encoded) and a later
/// section is the first charset-encoded one, the charset must be extracted
/// from that later section. Previously, charset extraction only triggered
/// for section 0 or when `sections.is_empty()`, so a plain section 0
/// prevented charset extraction from section 1.
#[test]
fn rfc2231_continuation_charset_from_non_first_encoded_section() {
    // Section *0 is plain, section *1* is the first encoded section with charset.
    let header = "attachment; filename*0=\"hello \"; filename*1*=UTF-8''w%C3%B6rld.txt";
    let result = extract_rfc2231_continuation(header, "filename");
    assert_eq!(
        result,
        Some("hello wörld.txt".to_string()),
        "RFC 2231 Section 3: charset from the first encoded section (not section 0) \
         must be used to decode percent-encoded bytes"
    );
}

/// RFC 2231 Section 3 + Postel's law (RFC 1122 Section 1.2.2):
/// `find_param_value` must tolerate optional whitespace around `=` in
/// continuation parameters, just as `extract_param` does for simple
/// parameters.
#[test]
fn rfc2231_continuation_tolerates_spaces_around_eq() {
    // Spaces around '=' in continuation parameters — non-conformant but
    // seen in the wild. Postel's law says we should accept this.
    let header = "attachment; filename*0 = \"hello\"; filename*1 = \"world.txt\"";
    assert_eq!(
        extract_rfc2231_continuation(header, "filename"),
        Some("helloworld.txt".to_string()),
        "RFC 2231 Section 3 + Postel's law: spaces around '=' in continuation \
         parameters must be tolerated"
    );
}

/// RFC 2046 Section 5.1.4: in multipart/alternative, the LAST matching
/// part is the "best choice". The parser must prefer the last text/html
/// (and last text/plain) in an alternative container, not the first.
#[test]
fn multipart_alternative_prefers_last_html() {
    let raw = b"From: sender@example.com\r\n\
Subject: test\r\n\
Date: Thu, 13 Feb 2025 15:47:33 +0000\r\n\
Content-Type: multipart/alternative; boundary=\"alt\"\r\n\
\r\n\
--alt\r\n\
Content-Type: text/plain\r\n\
\r\n\
plain text\r\n\
--alt\r\n\
Content-Type: text/html\r\n\
\r\n\
<p>first html</p>\r\n\
--alt\r\n\
Content-Type: text/html\r\n\
\r\n\
<p>second html - better version</p>\r\n\
--alt--";
    let parsed = parse_email(raw).unwrap();
    assert_eq!(
        parsed.body_html.as_deref(),
        Some("<p>second html - better version</p>"),
        "RFC 2046 Section 5.1.4: should prefer LAST text/html in multipart/alternative"
    );
}

/// RFC 2046 Section 5.1.4: same principle for text/plain — if there are
/// multiple text/plain parts in multipart/alternative, the last is preferred.
#[test]
fn multipart_alternative_prefers_last_plain() {
    let raw = b"From: sender@example.com\r\n\
Subject: test\r\n\
Date: Thu, 13 Feb 2025 15:47:33 +0000\r\n\
Content-Type: multipart/alternative; boundary=\"alt\"\r\n\
\r\n\
--alt\r\n\
Content-Type: text/plain\r\n\
\r\n\
first plain\r\n\
--alt\r\n\
Content-Type: text/plain\r\n\
\r\n\
second plain - better version\r\n\
--alt\r\n\
Content-Type: text/html\r\n\
\r\n\
<p>html version</p>\r\n\
--alt--";
    let parsed = parse_email(raw).unwrap();
    assert_eq!(
        parsed.body_text.as_deref(),
        Some("second plain - better version"),
        "RFC 2046 Section 5.1.4: should prefer LAST text/plain in multipart/alternative"
    );
}

/// RFC 2046 Section 5.1.1: boundary delimiter `--{boundary}` must be
/// complete — a longer string that merely starts with the boundary
/// (e.g. `--abcdef` when boundary is `abc`) is NOT a valid delimiter.
#[test]
fn boundary_not_matched_as_prefix() {
    let raw = b"From: a@b.com\r\nDate: Thu, 13 Feb 2025 15:47:33 +0000\r\nContent-Type: multipart/mixed; boundary=\"abc\"\r\n\r\n--abc\r\nContent-Type: text/plain\r\n\r\nBody text here\r\n--abcdef\r\nThis is NOT a boundary\r\n--abc--";
    let parsed = parse_email(raw).unwrap();
    assert_eq!(
        parsed.body_text.as_deref(),
        Some("Body text here\r\n--abcdef\r\nThis is NOT a boundary")
    );
}

/// RFC 2046 Section 5.1.1: the closing delimiter line is
/// `--boundary-- [LWSP] CRLF`. Extra non-whitespace text after the final
/// `--` is invalid and must be treated as part data, not as the closing
/// boundary.
#[test]
fn closing_boundary_requires_valid_terminator() {
    let raw = b"From: a@b.com\r\nDate: Thu, 13 Feb 2025 15:47:33 +0000\r\nContent-Type: multipart/mixed; boundary=\"abc\"\r\n\r\n--abc\r\nContent-Type: text/plain\r\n\r\nBody text here\r\n--abc--junk\r\nThis is NOT a closing boundary\r\n--abc--";
    let parsed = parse_email(raw).unwrap();
    assert_eq!(
        parsed.body_text.as_deref(),
        Some("Body text here\r\n--abc--junk\r\nThis is NOT a closing boundary")
    );
}

/// RFC 2045 Section 5.2: a Content-Type without a `/` (e.g. `text`)
/// is invalid and should default to `text/plain`.
#[test]
fn invalid_content_type_defaults_to_text_plain() {
    let raw = b"From: a@b.com\r\nDate: Thu, 13 Feb 2025 15:47:33 +0000\r\nContent-Type: text\r\n\r\nBody text";
    let parsed = parse_email(raw).unwrap();
    assert_eq!(parsed.body_text.as_deref(), Some("Body text"));
}

/// RFC 2047 Section 6.2: whitespace between adjacent encoded-words
/// is collapsed only when BOTH tokens are valid encoded words.
#[test]
fn whitespace_preserved_before_invalid_encoded_word() {
    let input = "=?UTF-8?B?SGVsbG8=?= =?broken";
    let decoded = decode_encoded_words(input);
    assert_eq!(decoded, "Hello =?broken");
}

/// RFC 2047 Section 5: in `*text` headers such as `Subject`, an
/// encoded-word MUST be separated from adjacent text by linear
/// whitespace. A glued token must stay literal instead of being decoded.
#[test]
fn encoded_word_without_lwsp_boundary_stays_literal_in_subject() {
    let raw = b"From: sender@example.com\r\n\
                 Date: Thu, 13 Feb 2025 15:47:33 +0000\r\n\
                 Subject: Prefix=?UTF-8?Q?caf=C3=A9?=Suffix\r\n\
                 \r\n\
                 Body\r\n";

    let parsed = parse_email(raw).unwrap();
    assert_eq!(
        parsed.subject.as_deref(),
        Some("Prefix=?UTF-8?Q?caf=C3=A9?=Suffix"),
        "RFC 2047 Section 5: Subject encoded-words must not decode \
         when they are glued to adjacent text without linear whitespace"
    );
}

/// RFC 2047 Section 5: within a `phrase`, an encoded-word MUST be
/// separated from adjacent words or specials by linear whitespace. A
/// display-name token glued to preceding text must remain literal.
#[test]
fn encoded_word_without_lwsp_boundary_stays_literal_in_phrase() {
    let raw = b"From: Prefix=?UTF-8?Q?caf=C3=A9?= <sender@example.com>\r\n\
                 To: recipient@example.com\r\n\
                 Date: Thu, 13 Feb 2025 15:47:33 +0000\r\n\
                 \r\n\
                 Body\r\n";

    let parsed = parse_email(raw).unwrap();
    assert_eq!(
        parsed.from[0].name.as_deref(),
        Some("Prefix=?UTF-8?Q?caf=C3=A9?="),
        "RFC 2047 Section 5: phrase encoded-words must not decode \
         when they are glued to adjacent words without linear whitespace"
    );
}

/// RFC 2047 Section 2: encoded-text must not contain spaces.
/// A space in the encoded-text makes the token an invalid encoded
/// word, so it must be passed through as literal text.
#[test]
fn encoded_word_base64_with_space_in_text() {
    // A space inside B-encoded-text is treated as a fold artifact
    // (RFC 5322 Section 2.2.3) and stripped before base64 decoding,
    // per Postel's law (RFC 1122 Section 1.2.2).
    let input = "=?UTF-8?B?SGVs bG8=?=";
    let decoded = decode_encoded_words(input);
    assert_eq!(
        decoded, "Hello",
        "space in B-encoded-text should be stripped as fold artifact"
    );
}

/// RFC 2046 Section 5.1.1 / Postel's law: bare CR line endings
/// should be accepted as boundary delimiters. The parser already
/// accepts bare LF (see `split_mime_parts_lf_only_boundaries`);
/// bare CR is the remaining gap.
#[test]
fn split_mime_parts_bare_cr_boundaries() {
    // Boundary preceded by bare CR (\r without \n).
    // Mirrors `split_mime_parts_lf_only_boundaries` but with \r.
    let body = b"--boundary\r\nContent-Type: text/plain\r\n\r\nPart 1\r--boundary\r\nContent-Type: text/plain\r\n\r\nPart 2\r--boundary--";
    let parts = split_mime_parts(body, "boundary");
    assert_eq!(
        parts.len(),
        2,
        "Should find 2 parts with bare-CR-only boundaries"
    );
}

/// RFC 5322 Section 3.4: an unmatched `>` in a malformed address list
/// must not cause `angle_depth` to go negative, which would prevent
/// subsequent commas from being recognized as address separators.
/// Regression: `parse_address_list` unconditionally decremented
/// `angle_depth` on `>`, allowing it to go negative.
#[test]
fn regression_unmatched_angle_bracket_does_not_break_separator() {
    // The stray `>` before the comma must not merge the two addresses
    // into one. The second address must still be recognized.
    let addrs = parse_address_list("bad> addr, user@example.com");
    assert!(
        addrs.iter().any(|a| a.email == "user@example.com"),
        "unmatched '>' must not prevent comma from splitting addresses: {addrs:?}"
    );
}

/// RFC 5322 Section 3.2.2: a parenthesized comment appearing BEFORE a
/// bare addr-spec should be extracted as the display name (common RFC 822
/// convention used by some mailers).
/// Regression: `parse_single_address` only extracted display names from
/// trailing comments, silently dropping leading comments.
#[test]
fn regression_leading_comment_display_name_extracted() {
    let addr = parse_single_address("(John Doe) user@example.com");
    let addr = addr.expect("should parse successfully");
    assert_eq!(addr.email, "user@example.com");
    assert_eq!(
        addr.name.as_deref(),
        Some("John Doe"),
        "leading comment should be used as display name (RFC 5322 Section 3.2.2)"
    );
}

/// RFC 2047 Section 5 rule (2): encoded-words may appear in comments.
/// A leading comment containing an encoded-word should be decoded.
#[test]
fn regression_leading_comment_rfc2047_decoded() {
    let addr = parse_single_address("(=?UTF-8?B?Sm9obg==?=) user@example.com");
    let addr = addr.expect("should parse successfully");
    assert_eq!(addr.email, "user@example.com");
    assert_eq!(
        addr.name.as_deref(),
        Some("John"),
        "encoded-words in leading comment must be decoded (RFC 2047 Section 5 rule 2)"
    );
}

/// RFC 5322 Section 3.2.2: comments inside a `display-name` are CFWS and
/// must not appear in the semantic display name returned to callers.
#[test]
fn regression_name_addr_comment_not_retained_in_display_name() {
    let raw = b"From: John (Boss) Doe <john@example.com>\r\n\
                To: recipient@example.com\r\n\
                Subject: Commented name\r\n\
                \r\n\
                body";

    let parsed = parse_email(raw).expect("message should parse successfully");
    assert_eq!(parsed.from.len(), 1);
    assert_eq!(
        parsed.from[0].name.as_deref(),
        Some("John Doe"),
        "display-name comments are CFWS and must not be retained"
    );
}

/// RFC 5322 Section 3.2.5: a display-name is a `phrase` made of one or
/// more `word`s, and each `word` may be a quoted-string. Mixed phrases
/// such as `\"John\" Doe <...>` must therefore unquote only the quoted
/// word, not preserve the literal quote characters.
#[test]
fn regression_name_addr_mixed_phrase_unquotes_quoted_word() {
    let raw = b"From: \"John\" Doe <john@example.com>\r\n\
                To: recipient@example.com\r\n\
                Subject: Mixed phrase\r\n\
                \r\n\
                body";

    let parsed = parse_email(raw).expect("message should parse successfully");
    assert_eq!(parsed.from.len(), 1);
    assert_eq!(
        parsed.from[0].name.as_deref(),
        Some("John Doe"),
        "quoted-string words in a display-name phrase must be unquoted"
    );
}

// ===== Edge-case bug-hunting tests =====

/// RFC 2047 Section 4.2: in Q-encoding, underscore represents space.
/// Verify `=?utf-8?Q?hello_world?=` decodes to "hello world", not `hello_world`.
#[test]
fn edge_q_encoding_underscore_is_space() {
    let result = decode_encoded_words("=?utf-8?Q?hello_world?=");
    assert_eq!(
        result, "hello world",
        "RFC 2047 Section 4.2: underscore in Q-encoding must decode to space"
    );
}

/// RFC 2047 Section 6.2: whitespace between adjacent encoded words
/// must be collapsed, even when the encoded words use different charsets.
#[test]
fn edge_adjacent_encoded_words_collapse_whitespace_different_charsets() {
    // "Hello" in UTF-8 B-encoding, then "World" in ISO-8859-1 B-encoding
    // SGVsbG8= is base64("Hello"), V29ybGQ= is base64("World")
    let result = decode_encoded_words("=?utf-8?B?SGVsbG8=?=   =?iso-8859-1?B?V29ybGQ=?=");
    assert_eq!(
        result, "HelloWorld",
        "RFC 2047 Section 6.2: whitespace between adjacent encoded words \
         with different charsets must be collapsed"
    );
}

/// RFC 2046 Section 5.1.1: boundary delimiters must appear at the
/// beginning of a line. A boundary string appearing mid-line in the
/// content must NOT split the part.
#[test]
fn edge_boundary_in_content_not_at_line_start() {
    let raw = b"From: a@b.com\r\n\
                 Date: Thu, 13 Feb 2025 15:47:33 +0000\r\n\
                 Content-Type: multipart/mixed; boundary=\"BOUND\"\r\n\
                 \r\n\
                 --BOUND\r\n\
                 Content-Type: text/plain\r\n\
                 \r\n\
                 This has --BOUND in the middle\r\n\
                 --BOUND--";
    let parsed = parse_email(raw).unwrap();
    assert_eq!(
        parsed.body_text.as_deref(),
        Some("This has --BOUND in the middle"),
        "Boundary mid-line must not split the part (RFC 2046 Section 5.1.1)"
    );
}

/// RFC 5322 Section 3.2.4: semicolons inside quoted parameter values
/// must not split parameters. `charset="utf-8; extra"` should extract
/// the whole quoted string as the charset value.
#[test]
fn edge_semicolon_in_quoted_param_value() {
    let ct = "text/plain; charset=\"utf-8; extra\"";
    let result = extract_param(ct, "charset");
    assert_eq!(
        result.as_deref(),
        Some("utf-8; extra"),
        "Semicolons inside quoted parameter values must not split params \
         (RFC 5322 Section 3.2.4)"
    );
}

/// RFC 2047: `=20` in Q-encoding is a hex-encoded space. A subject
/// consisting entirely of encoded spaces should decode to spaces.
#[test]
fn edge_subject_only_whitespace_after_decode() {
    let result = decode_encoded_words("=?utf-8?Q?=20=20?=");
    assert_eq!(
        result, "  ",
        "RFC 2047 Section 4.2: =20 in Q-encoding must decode to space"
    );
}

/// RFC 5322 Section 2.2.3: header lines over 998 characters must be
/// handled correctly via continuation line unfolding.
#[test]
fn edge_very_long_header_unfolded() {
    let long_value = "x".repeat(1000);
    let raw = format!(
        "From: a@b.com\r\n\
         Date: Thu, 13 Feb 2025 15:47:33 +0000\r\n\
         Received: {long_value}\r\n\
         \r\n"
    );
    let parsed = parse_email(raw.as_bytes()).unwrap();
    assert!(
        parsed.raw_headers.contains(&long_value),
        "Very long unfolded header line (>998 chars) must be preserved"
    );
}

/// RFC 5322 Section 4.3: CFWS (comments and folding white space) may
/// appear between most date-time tokens. A date with comments before the
/// day-of-week comma and between other tokens must still parse.
#[test]
fn edge_date_cfws_before_day() {
    let raw = b"From: a@b.com\r\n\
                 Date: (Mon, ) 13 Feb (year) 2025 15:47:33 +0000\r\n\
                 \r\n";
    let parsed = parse_email(raw).unwrap();
    let date = parsed
        .date
        .expect("Date with CFWS in various positions must parse per RFC 5322 Section 4.3");
    assert_eq!(date.day, 13);
    assert_eq!(date.month, 2);
    assert_eq!(date.year, 2025);
    assert_eq!(date.hour, 15);
    assert_eq!(date.minute, 47);
    assert_eq!(date.second, 33);
}

/// RFC 2046 Section 5.1: nested message/rfc822 inside multipart/mixed.
/// The inner message itself contains a multipart/mixed. Section numbers
/// must be assigned correctly per RFC 3501 Section 6.4.5.
#[test]
fn edge_nested_message_rfc822() {
    let inner_msg = "From: inner@example.com\r\n\
                     Date: Thu, 13 Feb 2025 15:47:33 +0000\r\n\
                     Content-Type: multipart/mixed; boundary=\"inner-bound\"\r\n\
                     \r\n\
                     --inner-bound\r\n\
                     Content-Type: text/plain\r\n\
                     \r\n\
                     Inner text\r\n\
                     --inner-bound\r\n\
                     Content-Type: application/pdf\r\n\
                     Content-Disposition: attachment; filename=\"inner.pdf\"\r\n\
                     \r\n\
                     PDF-data\r\n\
                     --inner-bound--";

    let raw = format!(
        "From: outer@example.com\r\n\
         Date: Thu, 13 Feb 2025 15:47:33 +0000\r\n\
         Content-Type: multipart/mixed; boundary=\"outer-bound\"\r\n\
         \r\n\
         --outer-bound\r\n\
         Content-Type: text/plain\r\n\
         \r\n\
         Outer text\r\n\
         --outer-bound\r\n\
         Content-Type: message/rfc822\r\n\
         \r\n\
         {inner_msg}\r\n\
         --outer-bound--"
    );
    let parsed = parse_email(raw.as_bytes()).unwrap();
    // Should parse without panicking and extract body text
    assert!(
        parsed.body_text.is_some(),
        "Nested message/rfc822 must be parsed without error"
    );
}

/// RFC 2047 Section 6.2: two adjacent encoded words with only a single
/// space between them must have the intervening whitespace collapsed.
#[test]
fn edge_adjacent_encoded_words_single_space() {
    let result = decode_encoded_words("=?utf-8?Q?A?= =?utf-8?Q?B?=");
    assert_eq!(
        result, "AB",
        "RFC 2047 Section 6.2: single space between adjacent encoded words must be collapsed"
    );
}

/// RFC 2047 Section 6.2: whitespace between an encoded word and a
/// non-encoded word must be preserved.
#[test]
fn edge_encoded_word_followed_by_plain_text() {
    let result = decode_encoded_words("=?utf-8?Q?Hello?= World");
    assert_eq!(
        result, "Hello World",
        "RFC 2047 Section 6.2: whitespace between encoded word and plain text must be preserved"
    );
}

/// RFC 2047 Q-encoding: multiple underscores should each become a space.
#[test]
fn edge_q_encoding_multiple_underscores() {
    let result = decode_encoded_words("=?utf-8?Q?a__b___c?=");
    assert_eq!(
        result, "a  b   c",
        "RFC 2047 Section 4.2: each underscore in Q-encoding represents one space"
    );
}

/// RFC 2047: an encoded-word with empty encoded text `=?utf-8?Q??=`
/// should decode to an empty string, not cause a panic.
#[test]
fn edge_encoded_word_empty_payload() {
    // RFC 2047 Section 2: encoded-text = 1*<Any printable ASCII ...>
    // Empty encoded-text is malformed; the encoded-word must be left as
    // literal text rather than decoded.
    let result = decode_encoded_words("=?utf-8?Q??=");
    assert_eq!(
        result, "=?utf-8?Q??=",
        "RFC 2047: empty Q-encoded payload is malformed and must be left as literal text"
    );
}

/// RFC 2047 Section 2: an encoded-word with empty B-encoded payload is
/// malformed (encoded-text = 1*<...>) and must be left as literal text.
#[test]
fn edge_encoded_word_empty_base64_payload() {
    let result = decode_encoded_words("=?utf-8?B??=");
    assert_eq!(
        result, "=?utf-8?B??=",
        "RFC 2047: empty B-encoded payload is malformed and must be left as literal text"
    );
}

/// RFC 2047: an encoded-word with an unknown encoding letter (not Q or B)
/// should be left as-is (not decoded).
#[test]
fn edge_encoded_word_unknown_encoding() {
    let result = decode_encoded_words("=?utf-8?X?test?=");
    // Should either pass through literally or produce "=?" + rest
    assert!(
        result.contains("utf-8") || result.contains("test"),
        "Unknown encoding letter should not cause a panic"
    );
}

/// RFC 2046 Section 5.1.1: a multipart boundary delimiter must appear
/// at the beginning of a line. Verify that the boundary `abc` does NOT
/// match `--abcdef` (a longer string that starts with the boundary).
#[test]
fn edge_boundary_prefix_not_false_match() {
    let raw = b"From: a@b.com\r\n\
                 Date: Thu, 13 Feb 2025 15:47:33 +0000\r\n\
                 Content-Type: multipart/mixed; boundary=\"abc\"\r\n\
                 \r\n\
                 --abc\r\n\
                 Content-Type: text/plain\r\n\
                 \r\n\
                 --abcdef is not a real boundary\r\n\
                 --abc--";
    let parsed = parse_email(raw).unwrap();
    assert!(
        parsed
            .body_text
            .as_deref()
            .unwrap_or("")
            .contains("--abcdef"),
        "Boundary 'abc' must not falsely match '--abcdef' (RFC 2046 Section 5.1.1)"
    );
}

/// RFC 5322 Section 2.2.3: header unfolding with tab continuation.
/// A Subject header split across lines with a tab continuation character
/// should be correctly unfolded.
#[test]
fn edge_header_unfolding_with_tab() {
    let raw = b"From: a@b.com\r\n\
                 Date: Thu, 13 Feb 2025 15:47:33 +0000\r\n\
                 Subject: Hello\r\n\
                 \tWorld\r\n\
                 \r\n";
    let parsed = parse_email(raw).unwrap();
    // The unfolded subject should have the tab preserved per RFC 5322 Section 2.2.3
    assert!(
        parsed.subject.as_ref().unwrap().contains("Hello"),
        "Subject must contain 'Hello' after unfolding"
    );
    assert!(
        parsed.subject.as_ref().unwrap().contains("World"),
        "Subject must contain 'World' from continuation line"
    );
}

/// RFC 2046 Section 5.1.1: a completely empty multipart body (no parts
/// between boundary delimiters) should not panic.
#[test]
fn edge_empty_multipart_body() {
    let raw = b"From: a@b.com\r\n\
                 Date: Thu, 13 Feb 2025 15:47:33 +0000\r\n\
                 Content-Type: multipart/mixed; boundary=\"b\"\r\n\
                 \r\n\
                 --b\r\n\
                 --b--";
    let parsed = parse_email(raw).unwrap();
    // Should not panic. Body text may be None or empty.
    // The important thing is graceful handling.
    assert!(
        parsed.body_text.is_none() || parsed.body_text.as_deref() == Some(""),
        "Empty multipart body should produce no body text"
    );
}

/// RFC 2047 Section 4.2: Q-encoding `=` (the literal equals sign) requires
/// hex encoding as `=3D`. Verify that a bare `=` in Q-encoding doesn't
/// break the decoder.
#[test]
fn edge_q_encoding_bare_equals() {
    // A bare `=` not followed by two hex digits — should be passed through
    // or handled gracefully (not panic).
    let result = decode_encoded_words("=?utf-8?Q?a=b?=");
    // The `=b` is an incomplete hex sequence. The decoder should not panic.
    assert!(
        !result.is_empty(),
        "Bare equals in Q-encoding should not cause panic"
    );
}

/// RFC 2047 Section 6.2: three consecutive encoded words with spaces
/// between each pair — all intervening whitespace should be collapsed.
#[test]
fn edge_three_adjacent_encoded_words() {
    let result = decode_encoded_words("=?utf-8?Q?A?= =?utf-8?Q?B?= =?utf-8?Q?C?=");
    assert_eq!(
        result, "ABC",
        "RFC 2047 Section 6.2: whitespace between all three adjacent \
         encoded words must be collapsed"
    );
}

/// RFC 5322 Section 3.3: a two-digit year between 50-99 should have
/// 1900 added per the obsolete year rule (RFC 5322 Section 4.3).
#[test]
fn edge_date_two_digit_year_obsolete() {
    let date = parse_rfc5322_date("13 Feb 99 15:47:33 +0000");
    let date = date.expect("Two-digit year 99 must parse per RFC 5322 Section 4.3");
    assert_eq!(
        date.year, 1999,
        "RFC 5322 Section 4.3: two-digit year 99 must map to 1999"
    );
}

/// RFC 5322 Section 4.3: a two-digit year between 00-49 should have
/// 2000 added.
#[test]
fn edge_date_two_digit_year_2000s() {
    let date = parse_rfc5322_date("13 Feb 05 15:47:33 +0000");
    let date = date.expect("Two-digit year 05 must parse per RFC 5322 Section 4.3");
    assert_eq!(
        date.year, 2005,
        "RFC 5322 Section 4.3: two-digit year 05 must map to 2005"
    );
}

/// RFC 5322 Section 2.2.3: folding white space (CRLF followed by SP or
/// HTAB) within a date header body must be normalised to a single space
/// before parsing. RFC 5322 Section 4.3 allows CFWS between all tokens
/// in the obsolete date-time syntax.
#[test]
fn parse_rfc5322_date_fws_space_folding() {
    let dt = parse_rfc5322_date("Thu, 01 Jan\r\n 2015 12:00:00 +0000");
    let dt = dt.expect("RFC 5322 Section 2.2.3: FWS (CRLF SP) within date must be normalised");
    assert_eq!(dt.year, 2015);
    assert_eq!(dt.month, 1);
    assert_eq!(dt.day, 1);
    assert_eq!(dt.hour, 12);
    assert_eq!(dt.minute, 0);
    assert_eq!(dt.second, 0);
    assert_eq!(dt.tz_offset_minutes, 0);
}

/// RFC 5322 Section 2.2.3: folding white space may use CRLF followed by
/// HTAB instead of a space.
#[test]
fn parse_rfc5322_date_fws_tab_folding() {
    let dt = parse_rfc5322_date("Thu, 01 Jan\r\n\t2015 12:00:00 +0000");
    let dt = dt.expect("RFC 5322 Section 2.2.3: FWS (CRLF HTAB) within date must be normalised");
    assert_eq!(dt.year, 2015);
    assert_eq!(dt.month, 1);
    assert_eq!(dt.day, 1);
    assert_eq!(dt.hour, 12);
    assert_eq!(dt.minute, 0);
    assert_eq!(dt.second, 0);
    assert_eq!(dt.tz_offset_minutes, 0);
}

/// RFC 5322 Section 4.3: the obsolete date-time syntax allows CFWS
/// between almost all tokens. Multiple FWS sequences in a single date
/// header must all be normalised.
#[test]
fn parse_rfc5322_date_fws_multiple_folds() {
    let dt = parse_rfc5322_date("Thu,\r\n 01\r\n Jan\r\n 2015\r\n 12:00:00\r\n +0000");
    let dt =
        dt.expect("RFC 5322 Section 4.3: multiple FWS sequences in date must all be normalised");
    assert_eq!(dt.year, 2015);
    assert_eq!(dt.month, 1);
    assert_eq!(dt.day, 1);
    assert_eq!(dt.hour, 12);
    assert_eq!(dt.minute, 0);
    assert_eq!(dt.second, 0);
    assert_eq!(dt.tz_offset_minutes, 0);
}

/// RFC 5322 Section 3.3 / Postel's law: some non-conformant mailers emit
/// a comma after the month name (e.g., `"Thu, 13 Feb, 2025 15:47:33 +0000"`).
/// The parser must strip trailing punctuation from the month token before
/// matching.
#[test]
fn parse_rfc5322_date_month_trailing_comma() {
    let dt = parse_rfc5322_date("Thu, 13 Feb, 2025 15:47:33 +0000");
    let dt = dt.expect("Postel's law: trailing comma after month must be tolerated");
    assert_eq!(dt.day, 13);
    assert_eq!(dt.month, 2);
    assert_eq!(dt.year, 2025);
    assert_eq!(dt.hour, 15);
    assert_eq!(dt.minute, 47);
    assert_eq!(dt.second, 33);
    assert_eq!(dt.tz_offset_minutes, 0);
}

/// RFC 5322 Section 3.3 / Postel's law: some non-conformant mailers emit
/// a period after the abbreviated month name (e.g., `"13 Feb. 2025 12:00:00 +0000"`).
/// The parser must strip trailing punctuation from the month token.
#[test]
fn parse_rfc5322_date_month_trailing_period() {
    let dt = parse_rfc5322_date("13 Feb. 2025 12:00:00 +0000");
    let dt = dt.expect("Postel's law: trailing period after month must be tolerated");
    assert_eq!(dt.day, 13);
    assert_eq!(dt.month, 2);
    assert_eq!(dt.year, 2025);
    assert_eq!(dt.hour, 12);
    assert_eq!(dt.minute, 0);
    assert_eq!(dt.second, 0);
    assert_eq!(dt.tz_offset_minutes, 0);
}

/// RFC 5322 Section 3.3: `date-time = [ day-of-week "," ] date time CFWS`.
/// A date without day-of-week but with a comma after the month
/// (non-conformant but seen in the wild) should still parse.
/// The parser must not mistake the pre-comma text for a day-of-week.
#[test]
fn parse_date_no_dow_comma_after_month() {
    let result = parse_rfc5322_date("13 Feb, 2025 12:00:00 +0000");
    assert!(
        result.is_some(),
        "date without DOW but with comma after month should parse"
    );
    let dt = result.unwrap();
    assert_eq!(dt.day, 13);
    assert_eq!(dt.month, 2);
    assert_eq!(dt.year, 2025);
    assert_eq!(dt.hour, 12);
    assert_eq!(dt.minute, 0);
    assert_eq!(dt.second, 0);
}

/// Ensure normal day-of-week parsing still works after the DOW-detection fix.
/// RFC 5322 Section 3.3: `day-of-week = ([FWS] day-name)`.
#[test]
fn parse_date_with_dow_still_works() {
    let result = parse_rfc5322_date("Thu, 13 Feb 2025 12:00:00 +0000");
    assert!(result.is_some());
    let dt = result.unwrap();
    assert_eq!(dt.day, 13);
    assert_eq!(dt.month, 2);
    assert_eq!(dt.year, 2025);
}

/// Both a proper DOW comma and a trailing month comma — the parser must
/// strip the DOW correctly and then tolerate the month-trailing comma.
/// RFC 5322 Section 3.3.
#[test]
fn parse_date_dow_comma_month_comma() {
    let result = parse_rfc5322_date("Thu, 13 Feb, 2025 15:47:33 +0000");
    assert!(result.is_some());
    let dt = result.unwrap();
    assert_eq!(dt.day, 13);
    assert_eq!(dt.month, 2);
    assert_eq!(dt.year, 2025);
}

/// RFC 2047 Section 4.2: Q-encoding with mixed `=XX` hex and underscore
/// characters. `=48=65=6C=6C=6F_=57=6F=72=6C=64` decodes to "Hello World".
#[test]
fn edge_q_encoding_mixed_hex_and_underscore() {
    let result = decode_encoded_words("=?utf-8?Q?=48=65=6C=6C=6F_=57=6F=72=6C=64?=");
    assert_eq!(
        result, "Hello World",
        "RFC 2047 Section 4.2: mixed hex encoding and underscore in Q-encoding"
    );
}

/// Multipart message where the text/plain part has
/// Content-Transfer-Encoding: base64.
#[test]
fn edge_multipart_base64_text_part() {
    // "Hello World" in base64 = "SGVsbG8gV29ybGQ="
    let raw = b"From: a@b.com\r\n\
                 Date: Thu, 13 Feb 2025 15:47:33 +0000\r\n\
                 Content-Type: multipart/mixed; boundary=\"b\"\r\n\
                 \r\n\
                 --b\r\n\
                 Content-Type: text/plain; charset=utf-8\r\n\
                 Content-Transfer-Encoding: base64\r\n\
                 \r\n\
                 SGVsbG8gV29ybGQ=\r\n\
                 --b--";
    let parsed = parse_email(raw).unwrap();
    assert_eq!(
        parsed.body_text.as_deref(),
        Some("Hello World"),
        "Base64-encoded text/plain in multipart must be decoded correctly"
    );
}

/// RFC 2047: encoded-word with unknown charset should decode gracefully
/// (not panic), falling back to lossy UTF-8 conversion.
#[test]
fn edge_encoded_word_unknown_charset() {
    let result = decode_encoded_words("=?x-unknown?B?SGVsbG8=?=");
    // encoding_rs falls back to UTF-8 for unknown charsets
    assert_eq!(
        result, "Hello",
        "Unknown charset in encoded-word should fall back gracefully"
    );
}

/// Parse a From header where the display name contains an `@` sign
/// inside quotes — the `@` must NOT be treated as an address separator.
#[test]
fn edge_from_display_name_with_at_sign() {
    let raw = b"From: \"user@company\" <real@example.com>\r\n\
                 Date: Thu, 13 Feb 2025 15:47:33 +0000\r\n\
                 \r\n";
    let parsed = parse_email(raw).unwrap();
    assert_eq!(
        parsed.from[0].email, "real@example.com",
        "Email must be extracted from angle brackets, not from quoted display name"
    );
    assert_eq!(
        parsed.from[0].name.as_deref(),
        Some("user@company"),
        "Display name containing '@' in quotes must be preserved"
    );
}

/// Parse a message with no blank line between headers and body —
/// the entire content should be treated as headers only (RFC 5322
/// Section 2.1 requires the blank line separator).
#[test]
fn edge_no_blank_line_separator() {
    let raw = b"From: a@b.com\r\n\
                 Date: Thu, 13 Feb 2025 15:47:33 +0000\r\n\
                 Subject: Test\r\n\
                 This is body text without blank line";
    let parsed = parse_email(raw).unwrap();
    // "This is body text..." has no colon and isn't a continuation line,
    // so it should be silently skipped by the header parser.
    // The body should be empty since there's no blank-line separator.
    // Or the parser might treat "This..." as beginning of body.
    // Either way, it must not panic.
    assert_eq!(parsed.subject.as_deref(), Some("Test"));
}

/// Encoded-word in Subject with charset switching between ISO-8859-1
/// and UTF-8 — both parts must be decoded correctly and whitespace
/// between them collapsed (RFC 2047 Section 6.2).
#[test]
fn edge_encoded_word_charset_switch() {
    // "café" in ISO-8859-1: 0x63 0x61 0x66 0xE9 → base64 "Y2Fm6Q=="
    // "résumé" in UTF-8: base64 "csOpc3Vtw6k="
    let result = decode_encoded_words("=?iso-8859-1?B?Y2Fm6Q==?= =?utf-8?B?csOpc3Vtw6k=?=");
    assert_eq!(
        result, "caférésumé",
        "RFC 2047 Section 6.2: encoded words with different charsets must \
         decode correctly with whitespace collapsed"
    );
}

/// Multiple From: headers — only the first should be used (RFC 5322
/// Section 3.6 says From occurs exactly once, but per Postel's law
/// we should use the first and not error).
#[test]
fn edge_duplicate_from_header() {
    let raw = b"From: first@example.com\r\n\
                 From: second@example.com\r\n\
                 Date: Thu, 13 Feb 2025 15:47:33 +0000\r\n\
                 \r\n";
    let parsed = parse_email(raw).unwrap();
    assert_eq!(
        parsed.from[0].email, "first@example.com",
        "First From header should be used when duplicates exist"
    );
}

/// Content-Type header with a boundary parameter that contains
/// a double-quote inside a quoted-string. The parser must correctly
/// extract the boundary using backslash-escaping (RFC 5322 Section 3.2.4).
#[test]
fn edge_boundary_with_escaped_quote() {
    // boundary parameter = "a\"b" — the boundary is actually: a"b
    let ct = r#"multipart/mixed; boundary="a\"b""#;
    let boundary = extract_param(ct, "boundary");
    assert_eq!(
        boundary.as_deref(),
        Some("a\"b"),
        "Boundary with escaped quote must be correctly unescaped \
         (RFC 5322 Section 3.2.4)"
    );
}

/// NUL byte (0x00) in header value should not cause the parser to panic
/// or produce incorrect results. Per Postel's law, graceful handling
/// is preferred.
#[test]
fn edge_nul_byte_in_subject() {
    let mut raw = Vec::new();
    raw.extend_from_slice(b"From: a@b.com\r\n");
    raw.extend_from_slice(b"Date: Thu, 13 Feb 2025 15:47:33 +0000\r\n");
    raw.extend_from_slice(b"Subject: Hello\x00World\r\n");
    raw.extend_from_slice(b"\r\n");
    let parsed = parse_email(&raw).unwrap();
    // NUL byte handling is implementation-defined; we just verify
    // the parser doesn't panic.
    assert!(
        parsed.subject.is_some(),
        "Subject with NUL byte must still be extracted (Postel's law)"
    );
}

/// Address with group syntax: `Group: addr1@a.com, addr2@b.com;`
/// The addresses inside the group should be extracted.
#[test]
fn edge_address_group_syntax() {
    let raw = b"From: sender@example.com\r\n\
                 Date: Thu, 13 Feb 2025 15:47:33 +0000\r\n\
                 To: Friends: alice@a.com, bob@b.com;\r\n\
                 \r\n";
    let parsed = parse_email(raw).unwrap();
    assert!(
        parsed.to.len() >= 2,
        "Group syntax 'Friends: alice, bob;' must extract both addresses. \
         Got: {:?}",
        parsed.to
    );
}

/// RFC 2046 Section 5.1.1: multipart/mixed with boundary that contains
/// characters valid in boundary but unusual: `+`, `/`, `=`.
#[test]
fn edge_boundary_special_chars() {
    let raw = b"From: a@b.com\r\n\
                 Date: Thu, 13 Feb 2025 15:47:33 +0000\r\n\
                 Content-Type: multipart/mixed; boundary=\"a+b/c=d\"\r\n\
                 \r\n\
                 --a+b/c=d\r\n\
                 Content-Type: text/plain\r\n\
                 \r\n\
                 body text\r\n\
                 --a+b/c=d--";
    let parsed = parse_email(raw).unwrap();
    assert_eq!(
        parsed.body_text.as_deref(),
        Some("body text"),
        "Boundary with special chars +/= must be handled correctly \
         (RFC 2046 Section 5.1.1)"
    );
}

/// Postel's law (RFC 1122 §1.2.2): tolerate whitespace around `=` in
/// MIME parameters. Some non-conformant mailers emit `charset = utf-8`
/// instead of `charset=utf-8`.
#[test]
fn parse_param_space_before_equals() {
    // Use UTF-8 encoded content (café) so that falling back to the
    // default US-ASCII charset would produce a different (wrong) result,
    // proving the charset parameter was actually extracted.
    let body_bytes: &[u8] = "café".as_bytes(); // 5 bytes: 63 61 66 c3 a9
    let mut raw = b"From: a@b.com\r\n\
                    Date: Thu, 13 Feb 2025 15:47:33 +0000\r\n\
                    Content-Type: text/plain; charset =utf-8\r\n\
                    \r\n"
        .to_vec();
    raw.extend_from_slice(body_bytes);
    let parsed = parse_email(&raw).unwrap();
    assert_eq!(
        parsed.body_text.as_deref(),
        Some("café"),
        "charset parameter with space before '=' must be recognized"
    );
}

#[test]
fn parse_param_spaces_around_equals() {
    let body_bytes: &[u8] = "café".as_bytes();
    let mut raw = b"From: a@b.com\r\n\
                    Date: Thu, 13 Feb 2025 15:47:33 +0000\r\n\
                    Content-Type: text/plain; charset = utf-8\r\n\
                    \r\n"
        .to_vec();
    raw.extend_from_slice(body_bytes);
    let parsed = parse_email(&raw).unwrap();
    assert_eq!(
        parsed.body_text.as_deref(),
        Some("café"),
        "charset with spaces around '=' must be parsed (Postel's law)"
    );
}

#[test]
fn parse_boundary_space_around_equals() {
    let raw = b"From: a@b.com\r\n\
                Date: Thu, 13 Feb 2025 15:47:33 +0000\r\n\
                Content-Type: multipart/mixed; boundary = \"bound42\"\r\n\
                \r\n\
                --bound42\r\n\
                Content-Type: text/plain\r\n\
                \r\n\
                Body text\r\n\
                --bound42--";
    let parsed = parse_email(raw).unwrap();
    assert_eq!(
        parsed.body_text.as_deref(),
        Some("Body text"),
        "boundary with spaces around '=' must be parsed (Postel's law)"
    );
}

#[test]
fn parse_rfc2231_param_space_around_equals() {
    let raw = b"From: a@b.com\r\n\
                Date: Thu, 13 Feb 2025 15:47:33 +0000\r\n\
                Content-Type: multipart/mixed; boundary=\"b\"\r\n\
                \r\n\
                --b\r\n\
                Content-Type: text/plain\r\n\
                \r\n\
                Body\r\n\
                --b\r\n\
                Content-Type: application/pdf\r\n\
                Content-Disposition: attachment; filename* = UTF-8''r%C3%A9sum%C3%A9.pdf\r\n\
                \r\n\
                DATA\r\n\
                --b--";
    let parsed = parse_email(raw).unwrap();
    assert_eq!(
        parsed.attachments[0].filename.as_deref(),
        Some("résumé.pdf"),
        "RFC 2231 filename* with space around '=' must be parsed (Postel's law)"
    );
}

/// RFC 5322 Section 3.6.2: the Sender field specifies the mailbox of the
/// agent responsible for the actual transmission of the message. It
/// contains exactly one mailbox (not a mailbox-list like From).
#[test]
fn parse_sender_header() {
    let raw = b"From: author@example.com\r\n\
                 Sender: transmitter@example.com\r\n\
                 To: recipient@example.com\r\n\
                 Subject: Test Sender\r\n\
                 Date: Thu, 13 Feb 2025 15:47:33 +0000\r\n\
                 \r\n\
                 Body";
    let parsed = parse_email(raw).unwrap();
    assert_eq!(
        parsed.sender.as_ref().map(|a| a.email.as_str()),
        Some("transmitter@example.com"),
        "Sender header should be parsed into parsed.sender"
    );
    assert_eq!(
        parsed.sender.as_ref().and_then(|a| a.name.as_deref()),
        None,
        "Sender with bare address should have no display name"
    );
}

/// RFC 5322 Section 3.6.2: Sender is optional. When absent,
/// parsed.sender should be None.
#[test]
fn parse_sender_header_absent() {
    let raw = b"From: author@example.com\r\n\
                 To: recipient@example.com\r\n\
                 Subject: No Sender\r\n\
                 Date: Thu, 13 Feb 2025 15:47:33 +0000\r\n\
                 \r\n\
                 Body";
    let parsed = parse_email(raw).unwrap();
    assert!(
        parsed.sender.is_none(),
        "Missing Sender header should yield None"
    );
}

/// RFC 5322 Section 3.6.2: Sender with display name.
#[test]
fn parse_sender_header_with_display_name() {
    let raw = b"From: author@example.com\r\n\
                 Sender: Secretary <secretary@example.com>\r\n\
                 To: recipient@example.com\r\n\
                 Subject: Test\r\n\
                 Date: Thu, 13 Feb 2025 15:47:33 +0000\r\n\
                 \r\n\
                 Body";
    let parsed = parse_email(raw).unwrap();
    let sender = parsed.sender.as_ref().unwrap();
    assert_eq!(sender.email, "secretary@example.com");
    assert_eq!(sender.name.as_deref(), Some("Secretary"));
}

/// RFC 5322 Section 3.6.2: `parse_headers_only` should also extract Sender.
#[test]
fn parse_headers_only_extracts_sender() {
    let raw = b"From: author@example.com\r\n\
                 Sender: transmitter@example.com\r\n\
                 To: recipient@example.com\r\n\
                 Subject: Test\r\n\
                 Date: Thu, 13 Feb 2025 15:47:33 +0000\r\n\
                 \r\n\
                 Body";
    let parsed = parse_headers_only(raw).unwrap();
    assert_eq!(
        parsed.sender.as_ref().map(|a| a.email.as_str()),
        Some("transmitter@example.com"),
        "parse_headers_only should also extract Sender"
    );
}

/// RFC 5322 Section 3.6.8: optional fields (headers not defined in the
/// spec) must be accessible via `extra_headers` without re-parsing
/// `raw_headers`.
#[test]
fn parse_extra_headers_accessible() {
    let raw = b"From: sender@example.com\r\n\
                 To: recipient@example.com\r\n\
                 Subject: Test\r\n\
                 X-Mailer: TestMailer\r\n\
                 List-Unsubscribe: <mailto:unsub@example.com>\r\n\
                 DKIM-Signature: v=1; a=rsa-sha256; d=example.com\r\n\
                 \r\n\
                 Body text\r\n";

    let parsed = parse_email(raw).unwrap();

    // Extra headers should contain all non-well-known headers
    assert!(
        parsed
            .extra_headers
            .iter()
            .any(|(k, v)| k == "x-mailer" && v == "TestMailer"),
        "extra_headers must contain X-Mailer (RFC 5322 Section 3.6.8)"
    );
    assert!(
        parsed
            .extra_headers
            .iter()
            .any(|(k, v)| k == "list-unsubscribe" && v == "<mailto:unsub@example.com>"),
        "extra_headers must contain List-Unsubscribe (RFC 5322 Section 3.6.8)"
    );
    assert!(
        parsed
            .extra_headers
            .iter()
            .any(|(k, v)| k == "dkim-signature" && v == "v=1; a=rsa-sha256; d=example.com"),
        "extra_headers must contain DKIM-Signature (RFC 5322 Section 3.6.8)"
    );

    // Well-known headers must NOT appear in extra_headers
    assert!(
        !parsed
            .extra_headers
            .iter()
            .any(|(k, _)| k == "from" || k == "to" || k == "subject"),
        "Well-known headers must not appear in extra_headers"
    );
}

#[test]
fn military_single_letter_timezones_rfc5322_section_4_3() {
    // RFC 5322 Section 4.3: single-letter military time zones (A–I, K–Y)
    // "SHOULD all be considered equivalent to '-0000'" because their
    // historical usage is unreliable.  Z is explicitly UTC.

    // All military single-letter zones (except Z) map to 0 (unknown).
    assert_eq!(
        parse_timezone("A"),
        0,
        "A should be -0000 (RFC 5322 Section 4.3 SHOULD)"
    );
    assert_eq!(
        parse_timezone("M"),
        0,
        "M should be -0000 (RFC 5322 Section 4.3 SHOULD)"
    );
    assert_eq!(
        parse_timezone("N"),
        0,
        "N should be -0000 (RFC 5322 Section 4.3 SHOULD)"
    );
    assert_eq!(
        parse_timezone("Y"),
        0,
        "Y should be -0000 (RFC 5322 Section 4.3 SHOULD)"
    );
    // Z = +0000 = 0 minutes (UTC)
    assert_eq!(
        parse_timezone("Z"),
        0,
        "Z should be +0000 (RFC 5322 Section 4.3)"
    );
    // J is not used per RFC 5322 Section 4.3, falls back to +0000
    assert_eq!(
        parse_timezone("J"),
        0,
        "J is not used and should fall back to +0000 (RFC 5322 Section 4.3)"
    );

    // Verify full date string parsing with military zone A
    let raw = b"From: sender@example.com\r\n\
                To: recipient@example.com\r\n\
                Subject: Test\r\n\
                Date: Thu, 01 Jan 2015 12:00:00 A\r\n\
                \r\n\
                Body\r\n";
    let parsed = crate::parse_email(raw).unwrap();
    let date = parsed.date.expect("date must be present");
    assert_eq!(
        date.tz_offset_minutes, 0,
        "Full date with military zone A should have tz_offset_minutes = 0 \
         (RFC 5322 Section 4.3: military zones are -0000)"
    );
}

#[test]
fn explicit_attachment_with_content_id_is_not_inline() {
    // RFC 2183 Section 2: Content-Disposition: attachment must not be overridden
    // by the presence of Content-ID (RFC 2392). Content-ID is just an identifier,
    // not a disposition indicator.
    let raw = b"From: sender@example.com\r\n\
                To: recipient@example.com\r\n\
                Subject: Test\r\n\
                Content-Type: image/png\r\n\
                Content-Disposition: attachment; filename=\"image.png\"\r\n\
                Content-ID: <img001@example.com>\r\n\
                Content-Transfer-Encoding: base64\r\n\
                \r\n\
                iVBORw0KGgo=\r\n";
    let parsed = crate::parse_email(raw).unwrap();
    assert_eq!(parsed.attachments.len(), 1);
    assert!(
        !parsed.attachments[0].is_inline,
        "Content-Disposition: attachment with Content-ID must not be marked inline \
         (RFC 2183 Section 2)"
    );
}

#[test]
fn multipart_explicit_attachment_with_content_id_is_not_inline() {
    // Same bug in multipart context: Content-Disposition: attachment must not be
    // overridden by Content-ID (RFC 2183 Section 2).
    let raw = b"From: sender@example.com\r\n\
                To: recipient@example.com\r\n\
                Subject: Test\r\n\
                Content-Type: multipart/mixed; boundary=\"BOUND\"\r\n\
                \r\n\
                --BOUND\r\n\
                Content-Type: text/plain\r\n\
                \r\n\
                Hello\r\n\
                --BOUND\r\n\
                Content-Type: image/png\r\n\
                Content-Disposition: attachment; filename=\"image.png\"\r\n\
                Content-ID: <img001@example.com>\r\n\
                Content-Transfer-Encoding: base64\r\n\
                \r\n\
                iVBORw0KGgo=\r\n\
                --BOUND--\r\n";
    let parsed = crate::parse_email(raw).unwrap();
    assert_eq!(parsed.attachments.len(), 1);
    assert!(
        !parsed.attachments[0].is_inline,
        "multipart: Content-Disposition: attachment with Content-ID must not be marked inline \
         (RFC 2183 Section 2)"
    );
}

/// RFC 5322 Section 3.2.2 / 3.4: `@` inside a parenthesized comment in
/// the display-name portion of a group address must not prevent group
/// syntax recognition.  Comments are not semantically meaningful and the
/// `@` within them is not an addr-spec indicator.
#[test]
fn group_address_with_at_in_comment() {
    // Comment containing '@' right after the group display-name.
    let addrs = parse_address_list("Group (user@host): member@example.com;");
    assert_eq!(
        addrs.len(),
        1,
        "group with comment containing '@' must parse the member address"
    );
    assert_eq!(addrs[0].email, "member@example.com");

    // Comment with a bare '@' sign (no full address).
    let addrs = parse_address_list("Group (has @ sign): member@example.com;");
    assert_eq!(
        addrs.len(),
        1,
        "group with comment containing bare '@' must parse the member address"
    );
    assert_eq!(addrs[0].email, "member@example.com");

    // Plain address must still work (no regression).
    let addrs = parse_address_list("user@host.com");
    assert_eq!(addrs.len(), 1);
    assert_eq!(addrs[0].email, "user@host.com");

    // Verify that `contains_at_outside_quotes` itself correctly
    // distinguishes between '@' in comments vs. outside them.
    assert!(
        !contains_at_outside_quotes("Group (user@host)"),
        "'@' only inside a comment must return false"
    );
    assert!(
        contains_at_outside_quotes("user@host.com (Display Name)"),
        "'@' outside a comment must return true"
    );
    assert!(
        !contains_at_outside_quotes(r#""user@host""#),
        "'@' only inside a quoted string must return false"
    );
    assert!(
        !contains_at_outside_quotes("Group (escaped \\) user@host)"),
        "'@' after an escaped paren inside a comment must return false"
    );
}

/// Spaces in encoded-text are treated as fold artifacts (RFC 5322
/// Section 2.2.3) and stripped before decoding per Postel's law.
/// HTAB and other control characters still cause rejection.
#[test]
fn regression_encoded_word_space_in_encoded_text_decoded() {
    // Base64 payload with an embedded space — treated as fold artifact.
    let with_space = "=?UTF-8?B?SGVs bG8=?=";
    assert_eq!(
        decode_encoded_words(with_space),
        "Hello",
        "space in B-encoded-text should be stripped as fold artifact"
    );

    // Valid base64 encoded word without spaces must still decode.
    let good_b64 = "=?UTF-8?B?SGVsbG8=?=";
    assert_eq!(
        decode_encoded_words(good_b64),
        "Hello",
        "valid base64 encoded word must decode normally"
    );

    // Valid Q-encoded word without spaces must still decode.
    let good_q = "=?UTF-8?Q?caf=C3=A9?=";
    assert_eq!(
        decode_encoded_words(good_q),
        "caf\u{e9}",
        "valid Q-encoded word must decode normally"
    );

    // Q-encoded word with a space in the encoded-text — treated as fold
    // artifact, stripped before decoding.
    let q_with_space = "=?UTF-8?Q?hel lo?=";
    assert_eq!(
        decode_encoded_words(q_with_space),
        "hello",
        "space in Q-encoded-text should be stripped as fold artifact"
    );
}

/// RFC 2231 Section 3: later high-numbered fragments must not be stitched
/// onto section 0 when every intermediate section is missing.
#[test]
fn rfc2231_continuation_gap_larger_than_ten() {
    let header = "attachment; filename*0=\"part1\"; filename*15=\"part2\"";
    let result = extract_rfc2231_continuation(header, "filename");
    assert_eq!(
        result,
        Some("part1".to_string()),
        "RFC 2231 Section 3: reassembly must stop before section 15 when sections 1-14 are absent"
    );
}

/// RFC 2231 Section 3 does not cap continuation indices at two digits.
/// Long parameter values may legitimately require more than 100 segments,
/// and the parser must reassemble the entire value rather than truncating
/// at section 99.
#[test]
fn rfc2231_continuation_more_than_one_hundred_sections() {
    use std::fmt::Write as _;

    let expected = "a".repeat(5_600);
    let mut header = String::from("attachment");
    for (idx, chunk) in expected.as_bytes().chunks(50).enumerate() {
        let chunk = std::str::from_utf8(chunk).unwrap();
        let _ = write!(header, "; filename*{idx}=\"{chunk}\"");
    }

    let result = extract_rfc2231_continuation(&header, "filename");
    assert_eq!(
        result.as_deref(),
        Some(expected.as_str()),
        "RFC 2231 Section 3: continuation parsing must not truncate after \
         100 sections"
    );
}

/// RFC 5322 Section 3.6 + Postel's law: when broken mailers produce
/// duplicate address-list headers (To, Cc, Bcc, Reply-To), the parser
/// must concatenate all occurrences instead of silently dropping
/// recipients from subsequent headers.
#[test]
fn regression_duplicate_address_headers_concatenated() {
    // Two separate To: headers — a broken but real-world pattern.
    let raw = b"From: sender@example.com\r\n\
                 To: alice@example.com\r\n\
                 To: bob@example.com\r\n\
                 Date: Thu, 13 Feb 2025 15:47:33 +0000\r\n\
                 \r\n";

    let parsed = parse_email(raw).unwrap();
    assert_eq!(
        parsed.to.len(),
        2,
        "both To: headers must be parsed — Postel's law requires concatenation"
    );
    assert_eq!(parsed.to[0].email, "alice@example.com");
    assert_eq!(parsed.to[1].email, "bob@example.com");

    // Two separate Cc: headers.
    let raw_cc = b"From: sender@example.com\r\n\
                    Cc: carol@example.com, dave@example.com\r\n\
                    Cc: eve@example.com\r\n\
                    Date: Thu, 13 Feb 2025 15:47:33 +0000\r\n\
                    \r\n";

    let parsed_cc = parse_email(raw_cc).unwrap();
    assert_eq!(
        parsed_cc.cc.len(),
        3,
        "all Cc: headers must be concatenated"
    );
    assert_eq!(parsed_cc.cc[0].email, "carol@example.com");
    assert_eq!(parsed_cc.cc[1].email, "dave@example.com");
    assert_eq!(parsed_cc.cc[2].email, "eve@example.com");
}

/// Regression test for MSG-001: duplicate From headers lose addresses.
///
/// `extract_from` used `get_header_value` which returns only the first
/// matching header. Duplicate From headers (while technically violating
/// RFC 5322 Section 3.6 SHOULD-at-most-once) are produced by broken
/// mailers; per Postel's law we concatenate addresses from all
/// occurrences, consistent with how To/Cc/Bcc are handled.
#[test]
fn test_duplicate_from_headers_concatenated() {
    let raw = b"From: alice@example.com\r\n\
                 From: bob@example.com\r\n\
                 Date: Thu, 13 Feb 2025 15:47:33 +0000\r\n\
                 \r\n";

    let parsed = parse_email(raw).unwrap();
    assert_eq!(
        parsed.from.len(),
        2,
        "both From: headers must be concatenated"
    );
    assert_eq!(parsed.from[0].email, "alice@example.com");
    assert_eq!(parsed.from[1].email, "bob@example.com");
}

/// Regression test: RFC 2047 Section 5 says encoded-words MUST NOT
/// appear in structured header fields like `Received`,
/// `Authentication-Results`, `DKIM-Signature`, etc.  The parser must
/// not attempt to decode `=?...?=` sequences in these headers.
#[test]
fn structured_headers_not_rfc2047_decoded() {
    let raw = b"From: sender@example.com\r\n\
                 Date: Thu, 13 Feb 2025 15:47:33 +0000\r\n\
                 Received: from =?utf-8?Q?server?= (mx.example.com)\r\n\
                 Authentication-Results: dkim=pass header.d==?utf-8?B?ZXhhbXBsZQ==?=.com\r\n\
                 DKIM-Signature: v=1; d==?utf-8?Q?example?=.com\r\n\
                 \r\n\
                 body";

    let parsed = parse_email(raw).unwrap();

    // Find each structured header and verify the encoded-word syntax
    // is preserved verbatim (not decoded).
    let received = parsed
        .extra_headers
        .iter()
        .find(|(k, _)| k == "received")
        .expect("Received header must be present");
    assert!(
        received.1.contains("=?utf-8?Q?server?="),
        "Received header must NOT be RFC 2047 decoded; got: {:?}",
        received.1
    );

    let auth = parsed
        .extra_headers
        .iter()
        .find(|(k, _)| k == "authentication-results")
        .expect("Authentication-Results header must be present");
    assert!(
        auth.1.contains("=?utf-8?B?ZXhhbXBsZQ==?="),
        "Authentication-Results header must NOT be RFC 2047 decoded; got: {:?}",
        auth.1
    );

    let dkim = parsed
        .extra_headers
        .iter()
        .find(|(k, _)| k == "dkim-signature")
        .expect("DKIM-Signature header must be present");
    assert!(
        dkim.1.contains("=?utf-8?Q?example?="),
        "DKIM-Signature header must NOT be RFC 2047 decoded; got: {:?}",
        dkim.1
    );
}

/// RFC 5322 Section 3.6.6: `Resent-Date` and `Resent-Message-ID`
/// are structured fields, so RFC 2047 Section 5 forbids encoded-word
/// decoding there. The parser must preserve any `=?...?=` sequences
/// literally in `extra_headers`.
#[test]
fn resent_structured_headers_not_rfc2047_decoded() {
    let raw = b"From: sender@example.com\r\n\
                 Date: Thu, 13 Feb 2025 15:47:33 +0000\r\n\
                 Resent-Date: =?utf-8?Q?Thu,_13_Feb_2025_15:47:33_+0000?=\r\n\
                 Resent-Message-ID: =?utf-8?Q?<resent@example.com>?=\r\n\
                 \r\n\
                 body";

    let parsed = parse_email(raw).unwrap();

    let resent_date = parsed
        .extra_headers
        .iter()
        .find(|(k, _)| k == "resent-date")
        .expect("Resent-Date header must be present");
    assert!(
        resent_date
            .1
            .contains("=?utf-8?Q?Thu,_13_Feb_2025_15:47:33_+0000?="),
        "Resent-Date must NOT be RFC 2047 decoded; got: {:?}",
        resent_date.1
    );

    let resent_message_id = parsed
        .extra_headers
        .iter()
        .find(|(k, _)| k == "resent-message-id")
        .expect("Resent-Message-ID header must be present");
    assert!(
        resent_message_id
            .1
            .contains("=?utf-8?Q?<resent@example.com>?="),
        "Resent-Message-ID must NOT be RFC 2047 decoded; got: {:?}",
        resent_message_id.1
    );
}

/// RFC 5322 Section 3.6.6: resent address fields use the same mailbox /
/// address-list syntax as their non-Resent counterparts. RFC 2047
/// Section 5 rule (3) permits encoded-words in `phrase` productions
/// (display names), so the parser must decode them.
#[test]
fn resent_address_headers_rfc2047_decoded_in_extra_headers() {
    let raw = b"From: sender@example.com\r\n\
                 Date: Thu, 13 Feb 2025 15:47:33 +0000\r\n\
                 Resent-From: =?utf-8?Q?Alice_Sender?= <alice@example.com>\r\n\
                 Resent-To: =?utf-8?Q?Bob_Recipient?= <bob@example.com>\r\n\
                 Resent-Cc: =?utf-8?Q?Carol_Copy?= <carol@example.com>\r\n\
                 \r\n\
                 body";

    let parsed = parse_email(raw).unwrap();

    let resent_from = parsed
        .extra_headers
        .iter()
        .find(|(k, _)| k == "resent-from")
        .expect("Resent-From header must be present");
    assert!(
        resent_from.1.contains("Alice Sender"),
        "Resent-From display name must be RFC 2047 decoded; got {:?}",
        resent_from.1
    );

    let resent_to = parsed
        .extra_headers
        .iter()
        .find(|(k, _)| k == "resent-to")
        .expect("Resent-To header must be present");
    assert!(
        resent_to.1.contains("Bob Recipient"),
        "Resent-To display name must be RFC 2047 decoded; got {:?}",
        resent_to.1
    );

    let resent_cc = parsed
        .extra_headers
        .iter()
        .find(|(k, _)| k == "resent-cc")
        .expect("Resent-Cc header must be present");
    assert!(
        resent_cc.1.contains("Carol Copy"),
        "Resent-Cc display name must be RFC 2047 decoded; got {:?}",
        resent_cc.1
    );
}

/// RFC 5322 Section 4.5.6: obsolete `Resent-Reply-To` uses address-list
/// syntax, so RFC 2047 Section 5 rule (3) permits encoded-words in the
/// `phrase` (display name) production. The parser must decode them.
#[test]
fn obsolete_resent_reply_to_rfc2047_decoded_in_extra_headers() {
    let raw = b"From: sender@example.com\r\n\
                 Date: Thu, 13 Feb 2025 15:47:33 +0000\r\n\
                 Resent-Reply-To: =?utf-8?Q?Relay_Desk?= <relay@example.com>\r\n\
                 \r\n\
                 body";

    let parsed = parse_email(raw).unwrap();

    let resent_reply_to = parsed
        .extra_headers
        .iter()
        .find(|(k, _)| k == "resent-reply-to")
        .expect("Resent-Reply-To header must be present");
    assert!(
        resent_reply_to.1.contains("Relay Desk"),
        "Resent-Reply-To display name must be RFC 2047 decoded; got {:?}",
        resent_reply_to.1
    );
}

/// Regression test: domain-literal addresses like `user@[10,0,0,1]`
/// must not be split on the commas inside the brackets.
/// RFC 5321 Section 4.1.3 defines domain-literal as `"[" *dtext "]"`.
/// The address parser must track bracket depth to avoid treating
/// characters inside `[...]` as structural separators.
#[test]
fn domain_literal_with_commas_not_split() {
    let input = "user@[10,0,0,1], other@example.com";
    let addrs = parse_address_list(input);
    assert_eq!(
        addrs.len(),
        2,
        "expected 2 addresses, got {}: {:?}",
        addrs.len(),
        addrs
    );
    assert_eq!(
        addrs[0].email, "user@[10,0,0,1]",
        "first address must preserve domain-literal including commas"
    );
    assert_eq!(addrs[1].email, "other@example.com");
}

/// Regression test: dates with fractional seconds like "45.123" must
/// not silently drop the seconds to 0.  RFC 5322 Section 3.3 does not
/// define fractional seconds, but non-conformant mailers produce them.
/// Per Postel's law we strip the fractional part and keep the integer.
#[test]
fn parse_date_with_fractional_seconds() {
    let raw = b"From: a@b.com\r\n\
                 Date: Thu, 13 Feb 2025 15:30:45.123 +0000\r\n\
                 \r\n";

    let parsed = parse_email(raw).unwrap();
    let date = parsed.date.unwrap();
    assert_eq!(date.year, 2025);
    assert_eq!(date.month, 2);
    assert_eq!(date.day, 13);
    assert_eq!(date.hour, 15);
    assert_eq!(date.minute, 30);
    assert_eq!(
        date.second, 45,
        "fractional seconds must not cause second to drop to 0"
    );
    assert_eq!(date.tz_offset_minutes, 0);
}

#[test]
fn test_rfc2047_empty_fields() {
    // RFC 2047 Section 2 requires:
    //   charset      = token          (token = 1*<...>)
    //   encoding     = token          (token = 1*<...>)
    //   encoded-text = 1*<Any printable ASCII character other than "?" or SPACE>
    // Encoded-words with empty charset, encoding, or encoded-text are
    // malformed and must be left as literal text (not decoded).

    // Empty charset: =??Q?test?=
    let raw_empty_charset = b"From: sender@example.com\r\n\
                 To: recipient@example.com\r\n\
                 Subject: =??Q?test?=\r\n\
                 Date: Thu, 13 Feb 2025 15:47:33 +0000\r\n\
                 \r\n";
    let parsed = parse_email(raw_empty_charset).unwrap();
    assert_eq!(
        parsed.subject.as_deref(),
        Some("=??Q?test?="),
        "empty charset must not be decoded"
    );

    // Empty encoding: =?UTF-8??test?=
    let raw_empty_encoding = b"From: sender@example.com\r\n\
                 To: recipient@example.com\r\n\
                 Subject: =?UTF-8??test?=\r\n\
                 Date: Thu, 13 Feb 2025 15:47:33 +0000\r\n\
                 \r\n";
    let parsed = parse_email(raw_empty_encoding).unwrap();
    assert_eq!(
        parsed.subject.as_deref(),
        Some("=?UTF-8??test?="),
        "empty encoding must not be decoded"
    );

    // Empty encoded-text: =?UTF-8?Q??=
    let raw_empty_text = b"From: sender@example.com\r\n\
                 To: recipient@example.com\r\n\
                 Subject: =?UTF-8?Q??=\r\n\
                 Date: Thu, 13 Feb 2025 15:47:33 +0000\r\n\
                 \r\n";
    let parsed = parse_email(raw_empty_text).unwrap();
    assert_eq!(
        parsed.subject.as_deref(),
        Some("=?UTF-8?Q??="),
        "empty encoded-text must not be decoded"
    );
}

/// RFC 2047 Section 2: encoded-text = 1*<Any printable ASCII character
/// other than "?" or SPACE>.  An encoded-word whose encoded-text
/// contains a literal '?' must be treated as malformed and left as
/// literal text — the '?' would cause `find("?=")` to match
/// prematurely, silently truncating the decoded output.
#[test]
fn test_rfc2047_question_mark_in_encoded_text() {
    // =?UTF-8?Q?hello?world?= — the first '?=' match is between
    // "hello" and "world", which would truncate to "hello" if the
    // '?' in encoded-text is not rejected.
    let raw = b"From: sender@example.com\r\n\
                 To: recipient@example.com\r\n\
                 Subject: =?UTF-8?Q?hello?world?=\r\n\
                 Date: Thu, 13 Feb 2025 15:47:33 +0000\r\n\
                 \r\n";
    let parsed = parse_email(raw).unwrap();
    assert_eq!(
        parsed.subject.as_deref(),
        Some("=?UTF-8?Q?hello?world?="),
        "'?' in encoded-text must cause the entire encoded-word to be \
         left as literal text (RFC 2047 Section 2)"
    );
}

// -----------------------------------------------------------------------
// Edge-case probe tests
// -----------------------------------------------------------------------

/// RFC 2047 Section 4.2: in Q-encoding, underscores represent spaces.
/// Verify that `=?UTF-8?Q?hello_world?=` decodes to "hello world".
#[test]
fn edge_rfc2047_q_encoding_underscore_to_space() {
    let raw = b"From: sender@example.com\r\n\
                 To: recipient@example.com\r\n\
                 Subject: =?UTF-8?Q?hello_world?=\r\n\
                 Date: Thu, 13 Feb 2025 15:47:33 +0000\r\n\
                 \r\n";
    let parsed = parse_email(raw).unwrap();
    assert_eq!(
        parsed.subject.as_deref(),
        Some("hello world"),
        "RFC 2047 Section 4.2: underscore in Q-encoding must decode to space"
    );
}

/// RFC 2047 Section 6.2: whitespace between adjacent encoded words must be
/// collapsed even when the encoded words use different charsets.
/// `=?UTF-8?B?w6k=?= =?ISO-8859-1?Q?=E9?=` should produce two e-acute
/// characters with no space between them.
#[test]
fn edge_rfc2047_adjacent_encoded_words_different_charsets() {
    // UTF-8 base64 for "e-acute" (U+00E9): 0xC3 0xA9 → base64 "w6k="
    // ISO-8859-1 Q-encoding for "e-acute" (0xE9): "=E9"
    let raw = b"From: sender@example.com\r\n\
                 To: recipient@example.com\r\n\
                 Subject: =?UTF-8?B?w6k=?= =?ISO-8859-1?Q?=E9?=\r\n\
                 Date: Thu, 13 Feb 2025 15:47:33 +0000\r\n\
                 \r\n";
    let parsed = parse_email(raw).unwrap();
    assert_eq!(
        parsed.subject.as_deref(),
        Some("\u{00E9}\u{00E9}"),
        "RFC 2047 Section 6.2: whitespace between adjacent encoded words \
         with different charsets must be collapsed"
    );
}

/// RFC 2046 Section 5.1.1: boundary delimiter matching must not treat
/// `--abcdef` as a match for boundary `abc`. The delimiter must be
/// followed by a valid terminator (CR, LF, `-`, SP, HTAB, or EOF).
#[test]
fn edge_multipart_boundary_prefix_of_another() {
    let raw = b"From: sender@example.com\r\n\
                 To: recipient@example.com\r\n\
                 Content-Type: multipart/mixed; boundary=\"abc\"\r\n\
                 Date: Thu, 13 Feb 2025 15:47:33 +0000\r\n\
                 \r\n\
                 --abc\r\n\
                 Content-Type: text/plain; charset=utf-8\r\n\
                 \r\n\
                 Line with --abcdef in the body should not split\r\n\
                 --abc\r\n\
                 Content-Type: text/plain; charset=utf-8\r\n\
                 \r\n\
                 Second part\r\n\
                 --abc--\r\n";
    let parsed = parse_email(raw).unwrap();
    assert_eq!(
        parsed.body_text.as_deref(),
        Some("Line with --abcdef in the body should not split"),
        "RFC 2046 Section 5.1.1: boundary 'abc' must not match '--abcdef' \
         in the body — the delimiter must be followed by a valid terminator"
    );
    // The second text/plain part is an extra beyond body_text, so it
    // becomes an attachment. This proves the boundary delimiter was not
    // prematurely consumed by "--abcdef" in the first part's body.
    assert_eq!(
        parsed.attachments.len(),
        1,
        "Two text/plain parts: first is body_text, second becomes an \
         attachment — this proves boundary splitting was correct"
    );
}

/// RFC 2231 Section 3: continuation parameters with a gap (missing index 1)
/// must stop at the contiguous prefix beginning at section 0.
#[test]
fn edge_rfc2231_continuation_with_gap_missing_index_1() {
    let header = "attachment; filename*0=\"part1\"; filename*2=\"part3\"";
    let result = extract_rfc2231_continuation(header, "filename");
    assert_eq!(
        result.as_deref(),
        Some("part1"),
        "RFC 2231 Section 3: continuation must stop at the first missing index"
    );
}

/// Builder round-trip with a subject containing the literal text "=?"
/// which could be misinterpreted as an RFC 2047 encoded-word marker.
/// The subject `Price is =? 50€` must survive build→parse unchanged.
#[test]
fn edge_builder_roundtrip_subject_with_encoded_word_marker() {
    use crate::build_message;
    use crate::types::OutgoingEmail;

    let subject = "Price is =? 50\u{20AC}";
    let email = OutgoingEmail {
        from: vec![Address {
            name: None,
            email: "sender@example.com".to_string(),
        }],
        sender: None,
        to: vec![Address {
            name: None,
            email: "recipient@example.com".to_string(),
        }],
        cc: vec![],
        bcc: vec![],
        reply_to: vec![],
        date: Some(crate::types::DateTime {
            year: 2025,
            month: 2,
            day: 13,
            hour: 15,
            minute: 47,
            second: 33,
            tz_offset_minutes: 0,
        }),
        subject: subject.to_string(),
        body_text: Some("test".to_string()),
        body_html: None,
        in_reply_to: vec![],
        references: vec![],
        attachments: vec![],
        extra_headers: vec![],
    };
    let built = build_message(&email).unwrap();
    let parsed = parse_email(&built.raw).unwrap();
    assert_eq!(
        parsed.subject.as_deref(),
        Some(subject),
        "Subject containing literal '=?' and non-ASCII must survive \
         build->parse round-trip without being mangled by RFC 2047 encoding"
    );
}

/// RFC 5322 Section 4.3: a two-digit year of 99 in a date header
/// must be interpreted as 1999 (50-99 → +1900).
/// Uses the full `parse_email` path (not just `parse_rfc5322_date`).
#[test]
fn edge_date_two_digit_year_via_parse_email() {
    let raw = b"From: sender@example.com\r\n\
                 To: recipient@example.com\r\n\
                 Subject: Test\r\n\
                 Date: Sat, 13 Feb 99 12:00:00 +0000\r\n\
                 \r\n";
    let parsed = parse_email(raw).unwrap();
    let date = parsed.date.expect("date must parse");
    assert_eq!(
        date.year, 1999,
        "RFC 5322 Section 4.3: two-digit year 99 must map to 1999"
    );
    assert_eq!(date.month, 2);
    assert_eq!(date.day, 13);
}

/// RFC 5322 Section 3.4.1: a bare addr-spec followed by a parenthesized
/// comment should extract the email correctly and use the comment as
/// display name. `user@example.com (John Doe)` → email=user@example.com,
/// name=Some("John Doe").
#[test]
fn edge_address_with_trailing_comment() {
    let raw = b"From: user@example.com (John Doe)\r\n\
                 To: recipient@example.com\r\n\
                 Subject: Test\r\n\
                 Date: Thu, 13 Feb 2025 15:47:33 +0000\r\n\
                 \r\n";
    let parsed = parse_email(raw).unwrap();
    assert_eq!(
        parsed.from[0].email, "user@example.com",
        "RFC 5322 Section 3.4.1: email must be extracted from bare addr-spec \
         before parenthesized comment"
    );
    assert_eq!(
        parsed.from[0].name.as_deref(),
        Some("John Doe"),
        "RFC 822 convention: trailing parenthesized comment should be used \
         as display name"
    );
}

/// RFC 2046 Section 5.1.5: in multipart/digest, parts without an explicit
/// Content-Type header default to message/rfc822, not text/plain.
/// Such a part should not be treated as `body_text`.
#[test]
fn edge_multipart_digest_default_content_type() {
    // Build a multipart/digest with one part that has no Content-Type header.
    // The part body is a complete RFC 5322 message.
    let inner_message = "From: inner@example.com\r\n\
                         To: recipient@example.com\r\n\
                         Subject: Inner\r\n\
                         Date: Thu, 13 Feb 2025 15:47:33 +0000\r\n\
                         \r\n\
                         Inner body text";
    let raw = format!(
        "From: sender@example.com\r\n\
         To: recipient@example.com\r\n\
         Content-Type: multipart/digest; boundary=\"digestbound\"\r\n\
         Date: Thu, 13 Feb 2025 15:47:33 +0000\r\n\
         \r\n\
         --digestbound\r\n\
         \r\n\
         {inner_message}\r\n\
         --digestbound--\r\n"
    );
    let parsed = parse_email(raw.as_bytes()).unwrap();
    // The part defaults to message/rfc822, so it should NOT be treated
    // as body_text (which is for text/plain). It should instead appear
    // as an attachment with content_type "message/rfc822".
    assert!(
        parsed.body_text.is_none(),
        "RFC 2046 Section 5.1.5: parts in multipart/digest without \
         Content-Type must default to message/rfc822, not text/plain — \
         body_text should be None"
    );
    assert_eq!(
        parsed.attachments.len(),
        1,
        "The message/rfc822 part should appear as an attachment"
    );
    assert_eq!(
        parsed.attachments[0].content_type, "message/rfc822",
        "RFC 2046 Section 5.1.5: default Content-Type in multipart/digest \
         is message/rfc822"
    );
}

/// RFC 2045 Section 6.8: base64 encoded data may contain whitespace
/// (line breaks) that must be ignored during decoding.
/// `SGVs\r\nbG8=` should decode to "Hello".
#[test]
fn edge_base64_body_with_embedded_whitespace() {
    let raw = b"From: sender@example.com\r\n\
                 To: recipient@example.com\r\n\
                 Subject: Test\r\n\
                 Date: Thu, 13 Feb 2025 15:47:33 +0000\r\n\
                 Content-Type: text/plain; charset=utf-8\r\n\
                 Content-Transfer-Encoding: base64\r\n\
                 \r\n\
                 SGVs\r\nbG8=";
    let parsed = parse_email(raw).unwrap();
    assert_eq!(
        parsed.body_text.as_deref(),
        Some("Hello"),
        "RFC 2045 Section 6.8: whitespace in base64 body must be ignored \
         during decoding"
    );
}

// -----------------------------------------------------------------------
// Regression: Subject trim_start must not strip RFC 2047 encoded whitespace
// -----------------------------------------------------------------------

/// RFC 2047 Section 6.2 / RFC 5322 Section 3.6.5: leading whitespace
/// that is encoded inside an RFC 2047 encoded-word in the Subject header
/// is intentional content, not a fold artifact.  `trim_start()` after
/// decoding strips it, losing data.
///
/// `=?UTF-8?B?IEhlbGxv?=` encodes " Hello" (leading space + "Hello").
/// The decoded subject must be " Hello", not "Hello".
#[test]
fn regression_subject_trim_preserves_rfc2047_encoded_leading_space() {
    let raw = b"From: a@b.com\r\n\
                 Date: Thu, 13 Feb 2025 15:47:33 +0000\r\n\
                 Subject: =?UTF-8?B?IEhlbGxv?=\r\n\
                 \r\n\
                 body";
    let parsed = parse_email(raw).unwrap();
    assert_eq!(
        parsed.subject.as_deref(),
        Some(" Hello"),
        "RFC 2047 Section 6.2: leading whitespace encoded inside an \
         encoded-word is intentional content, not a fold artifact — \
         trim_start() must not strip it"
    );
}

/// Same issue: a Subject that is entirely encoded whitespace
/// (`=?UTF-8?B?IA==?=` encodes a single space " ") must not be reduced
/// to an empty string.
#[test]
fn regression_subject_trim_preserves_encoded_whitespace_only() {
    let raw = b"From: a@b.com\r\n\
                 Date: Thu, 13 Feb 2025 15:47:33 +0000\r\n\
                 Subject: =?UTF-8?B?IA==?=\r\n\
                 \r\n\
                 body";
    let parsed = parse_email(raw).unwrap();
    assert_eq!(
        parsed.subject.as_deref(),
        Some(" "),
        "RFC 2047 Section 6.2: a Subject consisting entirely of encoded \
         whitespace must not be stripped to empty"
    );
}

// -----------------------------------------------------------------------
// Regression: RFC 2231 Section 5 language tags in encoded-word charsets
// -----------------------------------------------------------------------

/// RFC 2231 Section 5 allows a language tag after the charset in an
/// RFC 2047 encoded-word: `=?charset*language?encoding?text?=`.
/// The `*language` suffix must be stripped before charset lookup,
/// otherwise non-UTF-8 charsets like ISO-8859-1 fail to decode.
#[test]
fn regression_rfc2231_language_tag_iso8859() {
    let raw = b"From: a@b.com\r\n\
                 Date: Thu, 13 Feb 2025 15:47:33 +0000\r\n\
                 Subject: =?ISO-8859-1*de?Q?Pr=FCfung?=\r\n\
                 \r\n\
                 body";
    let parsed = parse_email(raw).unwrap();
    assert_eq!(
        parsed.subject.as_deref(),
        Some("Pr\u{00FC}fung"),
        "RFC 2231 Section 5: language tag in charset field must be \
         stripped before charset lookup — ISO-8859-1*de should decode \
         as ISO-8859-1"
    );
}

/// RFC 2231 Section 5: UTF-8 with a language tag should still decode
/// correctly (this already worked by accident since `encoding_rs` falls
/// back to UTF-8 for unknown labels, but we test it explicitly).
#[test]
fn regression_rfc2231_language_tag_utf8() {
    let raw = b"From: a@b.com\r\n\
                 Date: Thu, 13 Feb 2025 15:47:33 +0000\r\n\
                 Subject: =?UTF-8*en?B?SGVsbG8gV29ybGQ=?=\r\n\
                 \r\n\
                 body";
    let parsed = parse_email(raw).unwrap();
    assert_eq!(
        parsed.subject.as_deref(),
        Some("Hello World"),
        "RFC 2231 Section 5: UTF-8*en should decode as UTF-8"
    );
}

/// RFC 2231 Section 5: an empty language tag (charset followed by `*`
/// with nothing after it) must also be handled — the charset is the
/// part before the `*`.
#[test]
fn regression_rfc2231_empty_language_tag() {
    let raw = b"From: a@b.com\r\n\
                 Date: Thu, 13 Feb 2025 15:47:33 +0000\r\n\
                 Subject: =?ISO-8859-1*?Q?Pr=FCfung?=\r\n\
                 \r\n\
                 body";
    let parsed = parse_email(raw).unwrap();
    assert_eq!(
        parsed.subject.as_deref(),
        Some("Pr\u{00FC}fung"),
        "RFC 2231 Section 5: empty language tag (ISO-8859-1*) must \
         still decode as ISO-8859-1"
    );
}

/// The fix must still strip fold-artifact whitespace: when the builder
/// folds before the first encoded-word, unfolding produces a leading
/// space that is NOT part of the decoded content.
#[test]
fn regression_subject_trim_still_strips_fold_artifact() {
    // Simulate a folded Subject: the raw header value after unfolding
    // has a leading space from the continuation line.
    let raw = b"From: a@b.com\r\n\
                 Date: Thu, 13 Feb 2025 15:47:33 +0000\r\n\
                 Subject:\r\n =?UTF-8?B?SGVsbG8=?=\r\n\
                 \r\n\
                 body";
    let parsed = parse_email(raw).unwrap();
    assert_eq!(
        parsed.subject.as_deref(),
        Some("Hello"),
        "Fold-artifact whitespace before an encoded-word must still be \
         stripped (RFC 5322 Section 2.2.3)"
    );
}

/// Regression: some non-conformant mailers wrap disposition-type in
/// double-quotes (e.g., `"attachment"` instead of `attachment`).
/// RFC 2183 Section 2 defines disposition-type as a bare token, but
/// per Postel's law we should tolerate quoted values.
#[test]
fn is_disposition_type_quoted() {
    // Quoted disposition-type with parameters.
    assert!(
        is_disposition_type("\"attachment\"; filename=\"test.txt\"", "attachment"),
        "quoted attachment with params should match"
    );
    // Quoted disposition-type without parameters.
    assert!(
        is_disposition_type("\"inline\"", "inline"),
        "quoted inline should match"
    );
    // Ensure unquoted tokens still work.
    assert!(
        is_disposition_type("attachment; filename=\"test.txt\"", "attachment"),
        "unquoted attachment should still match"
    );
    assert!(
        is_disposition_type("inline", "inline"),
        "unquoted inline should still match"
    );
}

/// RFC 2183 Section 2 uses a structured field body, so RFC 5322 Section
/// 3.2.2 comments may appear before the disposition token as CFWS.
/// The parser must still classify the part as an attachment.
#[test]
fn parse_email_attachment_with_leading_disposition_comment() {
    let raw = b"From: sender@example.com\r\n\
Date: Thu, 13 Feb 2025 15:47:33 +0000\r\n\
MIME-Version: 1.0\r\n\
Content-Type: text/plain; charset=us-ascii\r\n\
Content-Disposition: (saved copy) attachment; filename=\"notes.txt\"\r\n\
\r\n\
Quarterly report\r\n";

    let parsed = parse_email(raw).expect("message with attachment comment must parse");

    assert!(
        parsed.body_text.is_none(),
        "Content-Disposition CFWS before attachment must not demote the part to body text"
    );
    assert_eq!(
        parsed.attachments.len(),
        1,
        "attachment with leading disposition comment must still be detected"
    );
    assert_eq!(
        parsed.attachments[0].filename.as_deref(),
        Some("notes.txt"),
        "filename parameter must still be preserved"
    );
    assert!(
        !parsed.attachments[0].is_inline,
        "attachment disposition must not be treated as inline"
    );
}

/// RFC 2183 Section 2.10 allows `Content-Disposition` on the top-level
/// message. Consumers using `parse_headers_only()` still need that header
/// preserved, so it must remain in `extra_headers` even though the parser
/// also interprets it for body classification.
#[test]
fn parse_headers_only_preserves_top_level_content_disposition() {
    let raw = b"From: sender@example.com\r\n\
Date: Thu, 13 Feb 2025 15:47:33 +0000\r\n\
MIME-Version: 1.0\r\n\
Content-Type: text/plain; charset=us-ascii\r\n\
Content-Disposition: attachment; filename=\"message.eml\"\r\n\
\r\n\
Quarterly report\r\n";

    let parsed =
        parse_headers_only(raw).expect("message with top-level Content-Disposition must parse");

    assert!(
        parsed
            .extra_headers
            .iter()
            .any(|(name, value)| {
                name == "content-disposition"
                    && value == "attachment; filename=\"message.eml\""
            }),
        "top-level Content-Disposition must be preserved in extra_headers for header-only consumers"
    );
}

/// RFC 2045 Section 7 allows `Content-ID` on any MIME entity. Consumers
/// using `parse_headers_only()` still need that identifier preserved on the
/// top-level message rather than silently dropped.
#[test]
fn parse_headers_only_preserves_top_level_content_id() {
    let raw = b"From: sender@example.com\r\n\
Date: Thu, 13 Feb 2025 15:47:33 +0000\r\n\
MIME-Version: 1.0\r\n\
Content-Type: text/plain; charset=us-ascii\r\n\
Content-ID: <top-level@example.com>\r\n\
\r\n\
Quarterly report\r\n";

    let parsed = parse_headers_only(raw).expect("message with top-level Content-ID must parse");

    assert!(
        parsed
            .extra_headers
            .iter()
            .any(|(name, value)| name == "content-id" && value == "<top-level@example.com>"),
        "top-level Content-ID must be preserved in extra_headers for header-only consumers"
    );
}

/// Regression: a bare addr-spec with a quoted local-part containing `@`
/// (e.g., `"user@internal"@example.com`) was misparsed because the simple
/// `contains('@')` check found the `@` inside the quoted-string rather
/// than the structural one separating local-part from domain.
/// RFC 5322 Section 3.4.1: `addr-spec = local-part "@" domain`,
/// where `local-part = dot-atom / quoted-string`.
#[test]
fn regression_quoted_local_part_with_at_in_bare_address() {
    // Bare address with quoted local-part containing '@'
    let bare = parse_single_address("\"user@internal\"@example.com");
    assert!(bare.is_some(), "bare quoted-local-part addr should parse");
    let bare = bare.unwrap();
    assert_eq!(
        bare.email, "\"user@internal\"@example.com",
        "email field must contain the full addr-spec including quotes"
    );
    assert!(
        bare.name.is_none(),
        "bare address should have no display name"
    );

    // Angle-bracketed form — this should already work
    let angle = parse_single_address("<\"user@internal\"@example.com>");
    assert!(
        angle.is_some(),
        "angle-bracketed quoted-local-part addr should parse"
    );
    let angle = angle.unwrap();
    assert_eq!(
        angle.email, "\"user@internal\"@example.com",
        "angle-bracketed form must produce the same addr-spec"
    );
    assert!(
        angle.name.is_none(),
        "no display name in angle-bracket-only form"
    );

    // Quoted local-part with '@' but NO domain part — this is NOT a valid
    // addr-spec. The '@' is inside the quoted-string and does not serve
    // as the structural delimiter (RFC 5322 Section 3.4.1).
    let not_addr = parse_single_address("\"user@internal\"");
    assert!(
        not_addr.is_none(),
        "a quoted-string with '@' but no domain must not be treated as an addr-spec"
    );

    // Bare address with quoted local-part containing '@' plus a trailing
    // comment — must correctly extract the display name from the comment
    // and the full addr-spec (RFC 5322 Section 3.2.2).
    let with_comment = parse_single_address("\"user@internal\"@example.com (Display Name)");
    assert!(
        with_comment.is_some(),
        "quoted-local with trailing comment should parse"
    );
    let with_comment = with_comment.unwrap();
    assert_eq!(
        with_comment.email, "\"user@internal\"@example.com",
        "email field must be the full addr-spec"
    );
    assert_eq!(
        with_comment.name.as_deref(),
        Some("Display Name"),
        "display name must be extracted from trailing comment"
    );
}

/// Regression test: an unquoted Content-Type parameter value that was folded
/// mid-token (RFC 5322 Section 2.2.3) gets a space where the fold was after
/// header unfolding. The parser must reassemble the token by stripping the
/// fold-introduced whitespace, since RFC 2045 Section 5.1 forbids spaces
/// within tokens.
#[test]
fn extract_param_folded_mid_token_charset() {
    // After header unfolding, `charset=ut\r\n f-8` becomes `charset=ut f-8`.
    // The parser must recognise that the space is a folding artifact and
    // return "utf-8", not "ut".
    let header_value = "text/plain; charset=ut f-8";
    let result = extract_param(header_value, "charset");
    assert_eq!(
        result.as_deref(),
        Some("utf-8"),
        "folded mid-token value must be reassembled (RFC 5322 Section 2.2.3, \
         RFC 2045 Section 5.1)"
    );
}

/// RFC 5322 Section 2.2.3 unfolding plus RFC 2045 Section 5.1 token syntax:
/// a fold may split an unquoted MIME token even when the continuation
/// fragment is alphabetic-only. The parser must still reassemble the token.
#[test]
fn extract_param_folded_mid_token_boundary_all_alpha() {
    let header_value = "multipart/mixed; boundary=abc def";
    let result = extract_param(header_value, "boundary");
    assert_eq!(
        result.as_deref(),
        Some("abcdef"),
        "alphabetic folded boundary token must be reassembled after unfolding"
    );
}

/// RFC 2045 Section 5.1 inherits RFC 5322 CFWS, so a parameter name may
/// follow a comment without an extra literal space, e.g. `;(note)charset=`.
/// The scanner must treat the comment as a separator rather than requiring
/// the immediately preceding byte to be `;`, SP, or HTAB.
#[test]
fn extract_param_accepts_comment_only_cfws_before_name() {
    let header_value = "text/plain;(note)charset=utf-8";
    let result = extract_param(header_value, "charset");
    assert_eq!(
        result.as_deref(),
        Some("utf-8"),
        "comment-only CFWS before a parameter name must still count as a valid boundary"
    );
}

/// When an unquoted parameter value is followed by whitespace and a bare
/// word that is NOT a parameter name (no `=`), the fold-concatenation
/// heuristic must NOT merge the bare word into the value.
///
/// RFC 2045 Section 5.1: unquoted parameter values are tokens delimited
/// by `;` or whitespace.  A bare word like `unexpected` after the value
/// is stray text, not a fold continuation.
#[test]
fn extract_param_no_false_fold_concat() {
    let ct = "text/plain; charset=utf-8 unexpected; name=test.txt";
    let result = extract_param(ct, "charset");
    assert_eq!(
        result.as_deref(),
        Some("utf-8"),
        "should not concatenate 'unexpected' with 'utf-8'"
    );
}

/// MIME parameters are tokens optionally surrounded by CFWS; trailing
/// comments are not part of the parameter value.
///
/// RFC 2045 Section 5.1 uses RFC 822-style parameter syntax, and RFC 5322
/// Section 3.2.2 defines comments as CFWS. The parser must stop at the
/// token boundary instead of absorbing `(comment)` into the value.
#[test]
fn extract_param_does_not_absorb_trailing_comment() {
    let ct = "text/plain; charset=utf-8 (comment); name=test.txt";
    let result = extract_param(ct, "charset");
    assert_eq!(
        result.as_deref(),
        Some("utf-8"),
        "trailing comment must not be absorbed into MIME token value"
    );
}

/// Regression test: `extract_param` must not match a parameter name that
/// appears inside a CFWS comment (RFC 5322 Section 3.2.2).
///
/// RFC 2045 Section 5.1 inherits CFWS from RFC 822.  A parenthesized
/// comment like `(charset=iso-8859-1)` should be ignored during
/// parameter extraction.
///
/// Before the fix, `extract_param` lacked an `is_inside_comment` check,
/// so it would match parameter names inside comments when preceded by SP.
///
/// # References
/// - RFC 2045 Section 5.1 (Content-Type parameter syntax)
/// - RFC 5322 Section 3.2.2 (CFWS / comment)
#[test]
fn extract_param_ignores_comment_content() {
    // The comment `( charset=iso-8859-1)` contains a false match.
    // The real parameter is `charset=utf-8` after the comment.
    let ct = "text/plain; ( charset=iso-8859-1) charset=utf-8";
    let result = extract_param(ct, "charset");
    assert_eq!(
        result.as_deref(),
        Some("utf-8"),
        "must not match 'charset' inside a parenthesized comment"
    );
}

/// Regression test: nested comments should also be skipped.
///
/// # References
/// - RFC 5322 Section 3.2.2 (nested comments)
#[test]
fn extract_param_ignores_nested_comment() {
    let ct = "text/plain; (outer (charset=iso) comment) charset=utf-8";
    let result = extract_param(ct, "charset");
    assert_eq!(
        result.as_deref(),
        Some("utf-8"),
        "must not match 'charset' inside a nested comment"
    );
}

/// Exercises the full `parse_email` pipeline with a folded charset that
/// affects body decoding. Uses windows-1252 so incorrect charset ("windows"
/// instead of "windows-1252") would produce garbled output for non-ASCII
/// bytes.
#[test]
fn parse_email_folded_charset_mid_token() {
    // windows-1252 byte 0x93 = left double quotation mark U+201C
    // If the parser truncates the charset to "windows" (before the fold-
    // introduced space), encoding_rs won't recognise it and will fall back
    // to UTF-8, producing a replacement character U+FFFD instead.
    let raw: &[u8] = b"From: test@example.com\r\n\
        Date: Thu, 13 Feb 2025 15:47:33 +0000\r\n\
        Content-Type: text/plain; charset=windows\r\n \
        -1252\r\n\
        \r\n\
        \x93Hello\x94";
    let parsed = parse_email(raw).unwrap();
    let body = parsed.body_text.unwrap();
    assert!(
        body.contains('\u{201C}') && body.contains('\u{201D}'),
        "body must be decoded with reassembled charset windows-1252, got: {body:?}"
    );
}

/// Postel's law for MIME parameters must also tolerate a token folded
/// across more than one continuation line. The parser currently
/// reassembles only one fold fragment, which truncates charsets such as
/// `windows-1252` when they are folded twice.
#[test]
fn parse_email_folded_charset_multiple_mid_token_folds() {
    let raw: &[u8] = b"From: test@example.com\r\n\
        Date: Thu, 13 Feb 2025 15:47:33 +0000\r\n\
        Content-Type: text/plain; charset=win\r\n \
        dows-\r\n \
        1252\r\n\
        \r\n\
        \x93Hello\x94";
    let parsed = parse_email(raw).unwrap();
    let body = parsed.body_text.unwrap();
    assert!(
        body.contains('\u{201C}') && body.contains('\u{201D}'),
        "body must be decoded with fully reassembled charset windows-1252, \
         got: {body:?}"
    );
}

/// Regression test: alphabetic continuation fragments can still be valid
/// folded MIME token continuations when the previous fragment ends with a
/// non-alphabetic token character such as `-`.
///
/// RFC 2045 Section 5.1 forbids spaces inside tokens, so the unfolded
/// space here can only be a fold artifact (RFC 5322 Section 2.2.3). The
/// parser must reassemble `koi8-` + `r` into `koi8-r`, not truncate at the
/// fold boundary.
#[test]
fn parse_email_folded_charset_alpha_tail_after_hyphen() {
    let (encoded, _, _) = encoding_rs::KOI8_R.encode("Привет");

    let mut raw = b"From: test@example.com\r\n\
        Date: Thu, 13 Feb 2025 15:47:33 +0000\r\n\
        Content-Type: text/plain; charset=koi8-\r\n \
        r\r\n\
        \r\n"
        .to_vec();
    raw.extend_from_slice(&encoded);

    let parsed = parse_email(&raw).unwrap();
    assert_eq!(
        parsed.body_text.as_deref(),
        Some("Привет"),
        "body must be decoded with reassembled charset koi8-r"
    );
}

#[test]
fn encoded_word_rejects_control_chars_in_encoded_text() {
    // RFC 2047 Section 2: encoded-text must be printable ASCII (33-126)
    // minus '?'. A tab (0x09) in encoded-text should cause the encoded
    // word to be rejected, falling back to literal text.
    let input = b"From: test@example.com\r\nSubject: =?utf-8?B?\tSGVsbG8=?=\r\n\r\n";
    let parsed = parse_email(input).unwrap();
    // The encoded word should NOT be decoded — it should appear as literal text
    assert!(
        parsed.subject.as_deref() != Some("Hello"),
        "encoded word with tab in encoded-text must be rejected (RFC 2047 Section 2), \
         got subject: {:?}",
        parsed.subject
    );
}

#[test]
fn encoded_word_rejects_high_bytes_in_encoded_text() {
    // RFC 2047 Section 2: encoded-text is printable ASCII only (33-126).
    // Byte 0x80 is outside that range.
    let input = b"From: test@example.com\r\nSubject: =?utf-8?B?\x80SGVsbG8=?=\r\n\r\n";
    let parsed = parse_email(input).unwrap();
    assert!(
        parsed.subject.as_deref() != Some("Hello"),
        "encoded word with high byte in encoded-text must be rejected (RFC 2047 Section 2), \
         got subject: {:?}",
        parsed.subject
    );
}

/// Regression test: an RFC 2047 B-encoded word that was split by header
/// folding (RFC 5322 Section 2.2.3) must still be decoded.
///
/// After unfolding, the CRLF is removed but the continuation line's
/// leading whitespace is preserved inside the encoded-text.  The decoder
/// must strip fold-introduced whitespace from B-encoded text before
/// base64 decoding, just as it already does for body base64 decoding.
///
/// # References
/// - RFC 2047 Section 2 (encoded-word structure)
/// - RFC 5322 Section 2.2.3 (header unfolding)
#[test]
fn encoded_word_folded_b_encoding() {
    // "Hello" = base64 "SGVsbG8=" split across fold: "SGVs" + "\r\n " + "bG8="
    let input = b"From: test@example.com\r\nSubject: =?UTF-8?B?SGVs\r\n bG8=?=\r\n\r\nbody";
    let parsed = parse_email(input).unwrap();
    assert_eq!(
        parsed.subject.as_deref(),
        Some("Hello"),
        "B-encoded word split by header folding must decode to 'Hello'"
    );
}

/// Regression test: Q-encoded word split by header folding must also
/// decode correctly.
///
/// # References
/// - RFC 2047 Section 4.2 (Q encoding)
/// - RFC 5322 Section 2.2.3 (header unfolding)
#[test]
fn encoded_word_folded_q_encoding() {
    // "Hello" in Q encoding: "Hel" + "\r\n " + "lo"
    let input = b"From: test@example.com\r\nSubject: =?UTF-8?Q?Hel\r\n lo?=\r\n\r\nbody";
    let parsed = parse_email(input).unwrap();
    assert_eq!(
        parsed.subject.as_deref(),
        Some("Hello"),
        "Q-encoded word split by header folding must decode to 'Hello'"
    );
}

/// RFC 5322 Section 2.2.3 allows folded header lines to continue with
/// either SP or HTAB. RFC 2047 decoders therefore need to tolerate a
/// fold-introduced HTAB inside an encoded-word and strip it before
/// decoding, just as the SP variant is already handled.
#[test]
fn encoded_word_folded_b_encoding_with_tab() {
    // "Hello" = base64 "SGVsbG8=" split across fold: "SGVs" + "\r\n\t" + "bG8="
    let input = b"From: test@example.com\r\nSubject: =?UTF-8?B?SGVs\r\n\tbG8=?=\r\n\r\nbody";
    let parsed = parse_email(input).unwrap();
    assert_eq!(
        parsed.subject.as_deref(),
        Some("Hello"),
        "B-encoded word split by tab folding must decode to 'Hello'"
    );
}

// ========================================================================
// Fuzz regression tests
// ========================================================================

/// Regression: `parse_timezone` used fixed byte-offset slicing (`s[1..3]`,
/// `s[3..5]`) to extract hours and minutes from a timezone string like
/// `+0530`. When the input contained multi-byte UTF-8 characters (e.g.,
/// U+FFFD replacement chars from `from_utf8_lossy` on non-ASCII header
/// bytes), these byte offsets could land mid-character, causing a panic.
///
/// Fixed by using `s.get(1..3)` / `s.get(3..5)` which return `None`
/// on non-char-boundary indices instead of panicking.
///
/// Discovered by cargo-fuzz on the `parse_email` target.
#[test]
fn fuzz_parse_timezone_non_ascii_no_panic() {
    // Minimized crash input: Date header value ends with +0\xff\xff
    // which becomes +0\u{FFFD}\u{FFFD} after lossy UTF-8 conversion.
    let input: &[u8] = &[
        b'F', b'r', b'o', b'm', b':', b' ', b'a', b'@', b'b', b'\n', b'D', b'a', b't', b'e', b':',
        b' ', b'T', b'h', b'u', b',', b' ', b'3', b' ', b'F', b'e', b'b', b' ', b'2', b'0', b'2',
        b'5', b' ', b'1', b'5', b':', b'4', b'3', b' ', b'+', b'0', 0xFF, 0xFF,
    ];
    // Must not panic — should return Ok or Err, never abort.
    let _ = parse_email(input);
}

// ========================================================================
// Property-based invariant tests
// ========================================================================

mod prop_invariants {
    use super::*;
    use proptest::prelude::*;

    proptest! {
        #![proptest_config(ProptestConfig::with_cases(1000))]

        /// Email parser never panics on arbitrary input (Postel's law).
        #[test]
        fn parse_email_never_panics(data in prop::collection::vec(any::<u8>(), 0..1000)) {
            let _ = parse_email(&data);
        }

        /// Headers-only parser never panics on arbitrary input.
        #[test]
        fn parse_headers_only_never_panics(data in prop::collection::vec(any::<u8>(), 0..1000)) {
            let _ = parse_headers_only(&data);
        }

        /// RFC 2047 decoder never panics on arbitrary UTF-8 input.
        #[test]
        fn decode_encoded_words_never_panics(s in ".*") {
            let _ = decode_encoded_words(&s);
        }

        /// Date parser never panics on arbitrary UTF-8 input.
        #[test]
        fn parse_date_never_panics(s in ".*") {
            let _ = parse_rfc5322_date(&s);
        }

        /// Parsed email size matches input length.
        ///
        /// RFC 5322: the `size` field must reflect the raw input byte count
        /// so callers can trust it for quota checks, progress bars, etc.
        #[test]
        fn parsed_size_matches_input(data in prop::collection::vec(any::<u8>(), 0..1000)) {
            if let Ok(parsed) = parse_email(&data) {
                prop_assert_eq!(
                    parsed.size,
                    data.len() as u64,
                    "ParsedEmail.size must match input byte count"
                );
            }
        }

        /// Regression: `extract_param` must not match parameters inside
        /// RFC 5322 Section 3.2.2 comments.
        ///
        /// Bug 823a19e: `extract_param` found `charset` inside
        /// `(charset=fake)` comment.
        ///
        /// # References
        /// - RFC 2045 Section 5.1 (Content-Type parameter syntax)
        /// - RFC 5322 Section 3.2.2 (CFWS / comment)
        #[test]
        fn extract_param_ignores_comments(
            real_value in "[a-z][a-z0-9-]{0,15}",
            decoy_value in "[a-z][a-z0-9-]{0,15}",
        ) {
            let ct = format!("text/plain; (charset={decoy_value}) charset={real_value}");
            let result = extract_param(&ct, "charset");
            prop_assert_eq!(
                result.as_deref(),
                Some(real_value.as_str()),
                "must extract real value, not value inside comment, for: {:?}",
                ct
            );
        }

        /// Regression: backslash in unquoted parameter values must be literal.
        ///
        /// Bug 86922a5: backslash was treated as escape outside quoted strings.
        ///
        /// # References
        /// - RFC 2045 Section 5.1 (Content-Type parameter values)
        #[test]
        fn unquoted_param_preserves_backslash(
            prefix in "[a-z]{1,5}",
            suffix in "[a-z]{1,5}",
        ) {
            let ct = format!("application/octet-stream; filename={prefix}\\{suffix}.txt");
            let result = extract_param(&ct, "filename");
            if let Some(val) = result {
                prop_assert!(
                    val.contains('\\'),
                    "backslash in unquoted value must be preserved: got {:?} from {:?}",
                    val, ct
                );
            }
        }
    }
}

/// Tests that parsers terminate in bounded time on adversarial inputs.
/// These catch infinite loops, excessive recursion, and algorithmic
/// complexity attacks that wouldn't be caught by single-call property tests.
mod stuck_tests {
    use super::*;
    use std::time::{Duration, Instant};

    /// Maximum time any single parse call should take.
    /// If it exceeds this, the parser is effectively "stuck".
    const MAX_PARSE_TIME: Duration = Duration::from_secs(5);

    fn assert_terminates<F: FnOnce()>(name: &str, f: F) {
        let start = Instant::now();
        f();
        let elapsed = start.elapsed();
        assert!(
            elapsed < MAX_PARSE_TIME,
            "{name} took {elapsed:?}, exceeds {MAX_PARSE_TIME:?} — parser may be stuck"
        );
    }

    #[test]
    fn parse_email_repeated_bytes_100kb() {
        let data = vec![0xFFu8; 100_000];
        assert_terminates("100KB 0xFF", || {
            let _ = parse_email(&data);
        });
    }

    #[test]
    fn parse_email_deeply_nested_multipart() {
        // 100 levels of nested multipart — tests recursion depth limits
        let mut data = Vec::new();
        let boundaries: Vec<String> = (0..100).map(|i| format!("boundary{i:03}")).collect();

        // Build headers
        data.extend(
            b"From: a@b.com\r\nContent-Type: multipart/mixed; boundary=\"boundary000\"\r\n\r\n",
        );

        // Nest 100 levels deep
        for i in 0..99 {
            data.extend(format!("--{}\r\n", boundaries[i]).as_bytes());
            data.extend(
                format!(
                    "Content-Type: multipart/mixed; boundary=\"{}\"\r\n\r\n",
                    boundaries[i + 1]
                )
                .as_bytes(),
            );
        }
        // Innermost part
        data.extend(format!("--{}\r\n", boundaries[99]).as_bytes());
        data.extend(b"Content-Type: text/plain\r\n\r\nHello\r\n");
        data.extend(format!("--{}--\r\n", boundaries[99]).as_bytes());

        // Close all boundaries
        for i in (0..99).rev() {
            data.extend(format!("--{}--\r\n", boundaries[i]).as_bytes());
        }

        assert_terminates("100-deep nested multipart", || {
            let _ = parse_email(&data);
        });
    }

    #[test]
    fn parse_email_huge_header_count() {
        // 10,000 headers — tests that header parsing is linear
        let mut data = Vec::new();
        data.extend(b"From: a@b.com\r\n");
        for i in 0..10_000 {
            data.extend(format!("X-Header-{i}: value{i}\r\n").as_bytes());
        }
        data.extend(b"\r\nBody\r\n");
        assert_terminates("10K headers", || {
            let _ = parse_email(&data);
        });
    }

    #[test]
    fn parse_email_long_header_value() {
        // Single header with 1MB value — tests that folded line processing is linear
        let mut data = Vec::new();
        data.extend(b"From: a@b.com\r\nSubject: ");
        // 1MB of 'A' characters with folding every 76 chars
        for chunk in 0..14_000 {
            if chunk > 0 {
                data.extend(b"\r\n ");
            }
            data.extend(vec![b'A'; 76]);
        }
        data.extend(b"\r\n\r\nBody\r\n");
        assert_terminates("1MB folded header", || {
            let _ = parse_email(&data);
        });
    }

    #[test]
    fn decode_encoded_words_pathological_pattern() {
        // Many encoded-word-like patterns that fail to decode —
        // tests that the decoder doesn't backtrack quadratically.
        let mut input = String::new();
        for i in 0..10_000 {
            use std::fmt::Write;
            let _ = write!(input, "=?UTF-8?X?invalid{i}?= ");
        }
        assert_terminates("10K invalid encoded words", || {
            let _ = decode_encoded_words(&input);
        });
    }

    #[test]
    fn parse_headers_non_ascii_name_skipped() {
        // RFC 5322 Section 2.2: header field names are ASCII `ftext`.
        // RFC 6532 extends field bodies, not field names, so malformed
        // UTF-8 names must not be parsed as real headers.
        let input = "X-\u{0130}: value\r\n\r\n";
        let headers = parse_headers(input.as_bytes());
        assert!(headers.is_empty(), "non-ASCII header names must be skipped");
    }

    #[test]
    fn parse_year_four_digit_low_value_not_obs_year() {
        // RFC 5322 Section 4.3: obs-year rules only apply to 2-digit and
        // 3-digit year strings. A 4-digit "0050" must parse as year 50,
        // not year 1950, to allow round-tripping via to_rfc5322_string().
        assert_eq!(parse_year("0050"), Some(50));
        assert_eq!(parse_year("0000"), Some(0));
        assert_eq!(parse_year("0500"), Some(500));
        // Existing 2-digit behavior preserved
        assert_eq!(parse_year("50"), Some(1950));
        assert_eq!(parse_year("99"), Some(1999));
        assert_eq!(parse_year("00"), Some(2000));
        assert_eq!(parse_year("49"), Some(2049));
        // Existing 3-digit behavior preserved
        assert_eq!(parse_year("100"), Some(2000));
        // Existing 4-digit behavior preserved
        assert_eq!(parse_year("2025"), Some(2025));
    }

    #[test]
    fn parse_date_repeated_whitespace() {
        // Date with enormous amounts of whitespace — tests strip_comments
        // and split_whitespace performance.
        let mut input = String::from("Thu, 13 Feb 2025 15:47:33 +0000");
        for _ in 0..100_000 {
            input.push(' ');
        }
        assert_terminates("100K trailing spaces in date", || {
            let _ = parse_rfc5322_date(&input);
        });
    }
}

#[test]
fn duplicate_in_reply_to_headers_concatenated() {
    // RFC 5322 Section 3.6.4 + Postel's law: broken mailers may
    // emit duplicate In-Reply-To headers. Concatenate message-IDs
    // from all occurrences, matching treatment of address headers.
    let raw = b"From: a@b.com\r\n\
                In-Reply-To: <id1@host>\r\n\
                In-Reply-To: <id2@host>\r\n\
                Date: Thu, 13 Feb 2025 15:47:33 +0000\r\n\r\n";
    let parsed = parse_email(raw).unwrap();
    assert_eq!(parsed.in_reply_to, vec!["id1@host", "id2@host"]);
}

#[test]
fn duplicate_references_headers_concatenated() {
    // RFC 5322 Section 3.6.4 + Postel's law: broken mailers may
    // emit duplicate References headers. Concatenate message-IDs
    // from all occurrences, matching treatment of address headers.
    let raw = b"From: a@b.com\r\n\
                References: <ref1@host>\r\n\
                References: <ref2@host>\r\n\
                Date: Thu, 13 Feb 2025 15:47:33 +0000\r\n\r\n";
    let parsed = parse_email(raw).unwrap();
    assert_eq!(parsed.references, vec!["ref1@host", "ref2@host"]);
}

#[test]
fn extract_rfc2231_param_skips_comment_match() {
    // RFC 5322 Section 3.2.2: parameter names inside parenthesized
    // comments must be ignored during extraction.
    // Use a space before the param name inside the comment so it
    // passes the is_param_boundary check (requires ; or whitespace).
    let header = "attachment; ( filename*=UTF-8''wrong.pdf) filename*=UTF-8''correct.pdf";
    let result = extract_rfc2231_param(header, "filename");
    assert_eq!(result.as_deref(), Some("correct.pdf"));
}

#[test]
fn find_param_value_skips_comment_match() {
    // RFC 5322 Section 3.2.2: parameter names inside parenthesized
    // comments must be ignored during extraction.
    // Use a space before the param name inside the comment so it
    // passes the is_param_boundary check (requires ; or whitespace).
    let header = "attachment; ( filename*0=\"wrong\") filename*0=\"correct\"";
    let lower = header.to_ascii_lowercase();
    let result = find_param_value(&lower, header, "filename*0=");
    assert_eq!(result.as_deref(), Some("correct"));
}

/// MSG-001: Parser must accept dates with incorrect day-of-week per Postel's
/// law (RFC 1122 Section 1.2.2). The day-of-week MUST requirement in
/// RFC 5322 Section 3.3 is a generation constraint; real-world mailers
/// frequently compute the wrong weekday.
#[test]
fn regression_msg001_date_accepts_wrong_day_of_week() {
    // 2025-02-13 is Thursday, not Wednesday. The parser must still accept
    // this and return the correct numeric date fields.
    let result = parse_rfc5322_date("Wed, 13 Feb 2025 15:47:33 +0000");
    assert!(
        result.is_some(),
        "Parser must accept dates with incorrect day-of-week (Postel's law)"
    );
    let dt = result.unwrap();
    assert_eq!(dt.year, 2025);
    assert_eq!(dt.month, 2);
    assert_eq!(dt.day, 13);
    assert_eq!(dt.hour, 15);
    assert_eq!(dt.minute, 47);
    assert_eq!(dt.second, 33);
    assert_eq!(dt.tz_offset_minutes, 0);
}

/// MSG-019: Parser must decode overlong encoded words per Postel's law
/// (RFC 1122 Section 1.2.2). The 75-character limit in RFC 2047 Section 2
/// is a generation constraint; real-world mailers (Thunderbird, Asian-locale
/// clients) frequently produce overlong encoded words.
#[test]
fn regression_msg019_overlong_encoded_word_decoded() {
    // Encoded word is 116 chars (> 75 limit). Decodes to:
    // "This is a long UTF-8 subject that will exceed seventy-five characters easily"
    let encoded_word = "=?UTF-8?B?VGhpcyBpcyBhIGxvbmcgVVRGLTggc3ViamVjdCB0aGF0IHdpbGwgZXhjZWVkIHNldmVudHktZml2ZSBjaGFyYWN0ZXJzIGVhc2lseQ==?=";
    assert!(
        encoded_word.len() > 75,
        "Test setup: encoded word must be > 75 chars, got {}",
        encoded_word.len()
    );

    let decoded = decode_encoded_words(encoded_word);
    assert_eq!(
        decoded, "This is a long UTF-8 subject that will exceed seventy-five characters easily",
        "Overlong encoded words must be decoded (Postel's law, RFC 2047 Section 6.1)"
    );
}

/// MSG-100: Resent address headers (`Resent-From`, `Resent-Sender`,
/// `Resent-To`, `Resent-Cc`, `Resent-Bcc`, `Resent-Reply-To`) use the
/// same address-list / mailbox syntax as their non-Resent counterparts
/// (RFC 5322 Section 3.6.6). RFC 2047 Section 5 rule (3) permits
/// encoded-words in `phrase` productions (i.e., display names), so the
/// parser must decode them rather than preserving raw `=?...?=` sequences.
#[test]
fn regression_msg100_resent_address_headers_decode_display_names() {
    // "äö" encoded as UTF-8 base64 = "w6TDtg=="
    let raw = b"From: sender@example.com\r\n\
                 Date: Thu, 13 Feb 2025 15:47:33 +0000\r\n\
                 Resent-From: =?UTF-8?B?w6TDtg==?= <user@example.com>\r\n\
                 Resent-Sender: =?UTF-8?B?w6TDtg==?= <sender2@example.com>\r\n\
                 Resent-To: =?UTF-8?B?w6TDtg==?= <to@example.com>\r\n\
                 Resent-Cc: =?UTF-8?B?w6TDtg==?= <cc@example.com>\r\n\
                 Resent-Bcc: =?UTF-8?B?w6TDtg==?= <bcc@example.com>\r\n\
                 Resent-Reply-To: =?UTF-8?B?w6TDtg==?= <reply@example.com>\r\n\
                 \r\n\
                 body";

    let parsed = parse_email(raw).unwrap();

    for header_name in &[
        "resent-from",
        "resent-sender",
        "resent-to",
        "resent-cc",
        "resent-bcc",
        "resent-reply-to",
    ] {
        let (_, value) = parsed
            .extra_headers
            .iter()
            .find(|(k, _)| k == header_name)
            .unwrap_or_else(|| panic!("{header_name} header must be present in extra_headers"));
        assert!(
            value.contains("\u{e4}\u{f6}"),
            "{header_name} display name must be RFC 2047 decoded to 'äö'; got: {value:?}"
        );
        assert!(
            !value.contains("=?UTF-8?B?"),
            "{header_name} must not contain raw encoded-word syntax; got: {value:?}"
        );
    }
}

/// MSG-119: RFC 2045 Section 6.7 Rule #3 — trailing literal SP/HTAB before
/// a hard line break (CRLF) must be stripped during QP decoding because the
/// encoder MUST NOT represent trailing whitespace at the end of an encoded
/// line. Transport gateways may add trailing whitespace, so the decoder
/// must remove it. Hex-encoded whitespace (`=20`, `=09`) is intentionally
/// preserved because it was explicitly encoded by the sender.
#[test]
fn regression_msg119_qp_strip_trailing_whitespace_before_hard_break() {
    // 1. Trailing literal spaces before CRLF are stripped
    assert_eq!(
        decode_quoted_printable(b"hello   \r\nworld"),
        b"hello\r\nworld",
        "trailing literal spaces before CRLF must be stripped (RFC 2045 Section 6.7 Rule #3)"
    );

    // 2. Trailing literal tabs before CRLF are stripped
    assert_eq!(
        decode_quoted_printable(b"hello\t\t\r\nworld"),
        b"hello\r\nworld",
        "trailing literal tabs before CRLF must be stripped (RFC 2045 Section 6.7 Rule #3)"
    );

    // 3. Hex-encoded =20 before CRLF is preserved (intentionally encoded)
    assert_eq!(
        decode_quoted_printable(b"hello=20\r\nworld"),
        b"hello \r\nworld",
        "hex-encoded =20 before CRLF must be preserved (explicitly encoded by sender)"
    );

    // 4. Trailing literal spaces at end of input (no CRLF) are stripped
    assert_eq!(
        decode_quoted_printable(b"hello   "),
        b"hello",
        "trailing literal spaces at end of input must be stripped (RFC 2045 Section 6.7 Rule #3)"
    );

    // 5. Lines with only whitespace have it stripped
    assert_eq!(
        decode_quoted_printable(b"hello\r\n   \r\nworld"),
        b"hello\r\n\r\nworld",
        "lines containing only whitespace must be stripped (RFC 2045 Section 6.7 Rule #3)"
    );

    // 6. Mixed tabs and spaces before CRLF
    assert_eq!(
        decode_quoted_printable(b"hello \t \r\nworld"),
        b"hello\r\nworld",
        "mixed trailing tabs and spaces before CRLF must be stripped"
    );

    // 7. Hex-encoded =09 (tab) before CRLF is preserved
    assert_eq!(
        decode_quoted_printable(b"hello=09\r\nworld"),
        b"hello\t\r\nworld",
        "hex-encoded =09 before CRLF must be preserved (explicitly encoded by sender)"
    );
}

/// RFC 5322 Section 4.3: obsolete time syntax allows CFWS around the colon
/// separators in time-of-day:
///   obs-hour   = [CFWS] 2DIGIT [CFWS]
///   obs-minute = [CFWS] 2DIGIT [CFWS]
///   obs-second = [CFWS] 2DIGIT [CFWS]
///
/// Whitespace around colons splits the time across multiple whitespace-
/// delimited tokens, e.g. `"09 : 55 : 06"` becomes `["09", ":", "55", ":", "06"]`.
#[test]
fn parse_date_obs_time_whitespace_around_colons() {
    let dt = parse_rfc5322_date("Fri, 21 Nov 1997 09 : 55 : 06 -0600").unwrap();
    assert_eq!(dt.hour, 9);
    assert_eq!(dt.minute, 55);
    assert_eq!(dt.second, 6);
    assert_eq!(dt.tz_offset_minutes, -360);
    assert_eq!(dt.year, 1997);
    assert_eq!(dt.month, 11);
    assert_eq!(dt.day, 21);
}

/// Variant: whitespace only before colons, not after.
/// RFC 5322 Section 4.3 obs-time.
#[test]
fn parse_date_obs_time_whitespace_before_colons() {
    let dt = parse_rfc5322_date("Fri, 21 Nov 1997 09 :55 :06 -0600").unwrap();
    assert_eq!(dt.hour, 9);
    assert_eq!(dt.minute, 55);
    assert_eq!(dt.second, 6);
    assert_eq!(dt.tz_offset_minutes, -360);
}

/// Variant: whitespace only after colons, not before.
/// RFC 5322 Section 4.3 obs-time.
#[test]
fn parse_date_obs_time_whitespace_after_colons() {
    let dt = parse_rfc5322_date("Fri, 21 Nov 1997 09: 55: 06 -0600").unwrap();
    assert_eq!(dt.hour, 9);
    assert_eq!(dt.minute, 55);
    assert_eq!(dt.second, 6);
    assert_eq!(dt.tz_offset_minutes, -360);
}

/// Variant: mixed whitespace placement around colons.
/// RFC 5322 Section 4.3 obs-time.
#[test]
fn parse_date_obs_time_mixed_whitespace() {
    let dt = parse_rfc5322_date("Fri, 21 Nov 1997 09 :55: 06 -0600").unwrap();
    assert_eq!(dt.hour, 9);
    assert_eq!(dt.minute, 55);
    assert_eq!(dt.second, 6);
    assert_eq!(dt.tz_offset_minutes, -360);
}

/// Normal time format (no whitespace around colons) must still work after
/// the obs-time fix. RFC 5322 Section 3.3.
#[test]
fn parse_date_standard_time_still_works_after_obs_fix() {
    let dt = parse_rfc5322_date("21 Nov 1997 09:55:06 -0600").unwrap();
    assert_eq!(dt.hour, 9);
    assert_eq!(dt.minute, 55);
    assert_eq!(dt.second, 6);
    assert_eq!(dt.tz_offset_minutes, -360);
}

/// Obs-time without seconds — RFC 5322 Section 4.3 makes obs-second
/// optional: `obs-time = obs-hour ":" obs-minute [":" obs-second]`.
#[test]
fn parse_date_obs_time_whitespace_no_seconds() {
    let dt = parse_rfc5322_date("Fri, 21 Nov 1997 09 : 55 -0600").unwrap();
    assert_eq!(dt.hour, 9);
    assert_eq!(dt.minute, 55);
    assert_eq!(dt.second, 0);
    assert_eq!(dt.tz_offset_minutes, -360);
}

/// RFC 5322 Section 4.3: "UT" and "GMT" are well-defined as +0000
/// (semantically identical to Universal Time). They must NOT be
/// conflated with unreliable military single-letter zones.
#[test]
fn parse_timezone_ut_gmt_are_well_defined_rfc5322_section_4_3() {
    // RFC 5322 Section 4.3: "'UT' and 'GMT' are indications of
    // 'Universal Time' and 'Greenwich Mean Time', respectively, and
    // are both semantically identical to '+0000'."
    assert_eq!(
        parse_timezone("UT"),
        0,
        "UT is well-defined as +0000 (RFC 5322 Section 4.3)"
    );
    assert_eq!(
        parse_timezone("GMT"),
        0,
        "GMT is well-defined as +0000 (RFC 5322 Section 4.3)"
    );
    // UTC is not in RFC 5322 but is universally understood as +0000.
    assert_eq!(parse_timezone("UTC"), 0, "UTC is well-defined as +0000");
}

/// RFC 5322 Section 4.3: "Z" is defined as +0000 (UTC) in the
/// military zone table and is universally understood. It must be
/// treated as well-defined, unlike the other single-letter zones.
#[test]
fn parse_timezone_z_is_well_defined_utc() {
    assert_eq!(
        parse_timezone("Z"),
        0,
        "Z is well-defined as +0000 (RFC 5322 Section 4.3)"
    );
}

/// RFC 5322 Section 4.3: military single-letter zones (A-I, K-Y)
/// are unreliable and "SHOULD all be considered equivalent to '-0000'".
/// They must map to 0 but are semantically distinct from the
/// well-defined UT/GMT/UTC/Z zones.
#[test]
fn parse_timezone_military_zones_are_unreliable() {
    assert_eq!(
        parse_timezone("A"),
        0,
        "military zone A is unreliable, maps to -0000 (RFC 5322 Section 4.3)"
    );
    assert_eq!(
        parse_timezone("K"),
        0,
        "military zone K is unreliable, maps to -0000 (RFC 5322 Section 4.3)"
    );
    assert_eq!(
        parse_timezone("N"),
        0,
        "military zone N is unreliable, maps to -0000 (RFC 5322 Section 4.3)"
    );
    assert_eq!(
        parse_timezone("Y"),
        0,
        "military zone Y is unreliable, maps to -0000 (RFC 5322 Section 4.3)"
    );
}