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
|
;;; proof-script.el --- Major mode for proof assistant script files.
;;
;; Copyright (C) 1994-2001 LFCS Edinburgh.
;; Authors: David Aspinall, Yves Bertot, Healfdene Goguen,
;; Thomas Kleymann and Dilip Sequeira
;; License: GPL (GNU GENERAL PUBLIC LICENSE)
;;
;; $Id$
;;
;;
;;; Code:
(require 'proof) ; loader
(require 'proof-syntax) ; utils for manipulating syntax
(require 'span) ; abstraction of overlays/extents
(require 'pg-user) ; user-level commands
(require 'proof-menu) ; menus for script mode
(require 'proof-x-symbol) ; x-symbol (maybe initialize)
(require 'proof-mmm) ; mmm (maybe put on automode list)
;; Nuke some byte-compiler warnings
;; NB: eval-when (compile) is different to eval-when-compile!!
(eval-when (compile)
(proof-try-require 'func-menu)
(require 'comint))
;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;
;;
;; PRIVATE VARIABLES
;;
;; Local variables used by proof-script-mode
;;
;; Scripting variables
(defvar proof-last-theorem-dependencies nil
"Contains the dependencies of the last theorem. A list of strings.
Set in `proof-shell-process-urgent-message'.")
(defvar proof-nesting-depth 0
"Current depth of a nested proof.
Zero means outside a proof, 1 means inside a top-level proof, etc.
This variable is maintained in proof-done-advancing; it is zeroed
in proof-shell-clear-state.")
(defvar proof-element-counters nil
"Table of (name . count) pairs, counting elements in scripting buffer.")
;; Buffer-local variables
(deflocal proof-active-buffer-fake-minor-mode nil
"An indication in the modeline that this is the *active* script buffer")
(deflocal proof-script-buffer-file-name nil
;; NB: if buffer-file-name is nil for some other reason, this may break.
"A copied value of buffer-file-name to cope with `find-alternative-file'.
The `find-alternative-file' function has a nasty habit of setting the
buffer file name to nil before running kill buffer, which breaks PG's
kill buffer hook. This variable is used when buffer-file-name is nil.")
;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;
;;
;; Counting and naming proof elements
;;
(defun proof-next-element-count (idiom)
"Return count for next element of type IDIOM.
This uses and updates `proof-element-counters'."
(let ((next (1+ (or (cdr-safe (assq idiom proof-element-counters)) 0))))
(setq proof-element-counters
(cons (cons idiom next)
(remassq idiom proof-element-counters)))
next))
(defun proof-element-id (idiom number)
"Return a string identifier composed from symbol IDIOM and NUMBER."
(concat (symbol-name idiom) "-" (int-to-string number)))
(defun proof-next-element-id (idiom)
(proof-element-id idiom (proof-next-element-count idiom)))
;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;
;;
;; Configuration of function-menu (aka "fume")
;;
;; FIXME: we would like this code only enabled if the user loads
;; func-menu into Emacs.
;;
(deflocal proof-script-last-entity nil
"Record of last entity found.
A hack for entities that are named in two places, so that `find-next-entity'
doesn't return the same values twice.")
;; FIXME mmw: maybe handle comments/strings by using
;; proof-looking-at-syntactic-context
(defun proof-script-find-next-entity (buffer)
"Find the next entity for function menu in a proof script.
A value for `fume-find-function-name-method-alist' for proof scripts.
Uses `fume-function-name-regexp', which is initialised from
`proof-script-next-entity-regexps', which see."
;; Hopefully this function is fast enough.
(set-buffer buffer)
;; could as well use next-entity-regexps directly since this is
;; not really meant to be used as a general function.
(let ((anyentity (car fume-function-name-regexp)))
(if (proof-re-search-forward anyentity nil t)
;; We've found some interesting entity, but have to find out
;; which one, and where it begins.
(let ((entity (buffer-substring (match-beginning 0) (match-end 0)))
(start (match-beginning 0))
(discriminators (cdr fume-function-name-regexp))
(p (point))
disc res)
(while (and (not res) (setq disc (car-safe discriminators)))
(if (proof-string-match (car disc) entity)
(let*
((items (nth 1 disc))
(items (if (numberp items) (list items) items))
(name ""))
(dolist (item items)
(setq name
(concat name
(substring entity
(match-beginning item)
(match-end item))
" ")))
(cond
((eq (nth 2 disc) 'backward)
(setq start
(or (proof-re-search-backward (nth 3 disc) nil t)
start))
(goto-char p))
((eq (nth 2 disc) 'forward)
(proof-re-search-forward (nth 3 disc))))
(setq res (cons name start)))
(setq discriminators (cdr discriminators))))
res))))
;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;
;;
;; Basic functions for handling the locked and queue regions
;;
;; --------------------------------------------------------------
;;
;; Notes on regions in the scripting buffer. (da)
;;
;; The locked region is covered by a collection of non-overlaping
;; spans (our abstraction of extents/overlays).
;;
;; For an unfinished proof, there is one extent for each command
;; or comment outside a command. For a finished proof, there
;; is one extent for the whole proof.
;;
;; Each command span has a 'type property, one of:
;;
;; 'goalsave A goal..savegoal region in the buffer, a completed proof.
;; 'vanilla Initialised in proof-semis-to-vanillas, for
;; 'comment A comment outside a command.
;; 'proverproc A region closed by the prover, processed outwith PG
;; 'pbp A PBP command inserted automatically into the script
;;
;; All spans except those of type 'comment have a 'cmd property,
;; which is set to a string of its command. This is the
;; text in the buffer stripped of leading whitespace and any comments.
;;
;; ** Variables
(deflocal proof-locked-span nil
"The locked span of the buffer.
Each script buffer has its own locked span, which may be detached
from the buffer.
Proof General allows buffers in other modes also to be locked;
these also have a non-nil value for this variable.")
(deflocal proof-queue-span nil
"The queue span of the buffer. May be detached if inactive or empty.
Each script buffer has its own queue span, although only the active
scripting buffer may have an active queue span.")
;; da: reason for buffer local queue span is because initialisation
;; in proof-init-segmentation can happen when a file is visited.
;; So nasty things might happen if a locked file is visited whilst
;; another buffer has a non-empty queue region being processed.
;; ** Getters and setters
(defun proof-span-read-only (span &optional always)
"Make SPAN be read-only according to `proof-strict-read-only' or ALWAYS."
(if (or always proof-strict-read-only)
(span-read-only span)
(span-read-write span)
(span-write-warning span)))
(defun proof-strict-read-only ()
"Set locked spans in script buffers according to `proof-strict-read-only'."
;; NB: this implements the behaviour that read-only is synchronized
;; in all script buffers to follow the current setting of
;; `proof-strict-read-only'. Another possibility would be to
;; just change for local buffer, while at the same time changing
;; the default/global setting. This would be consistent with
;; behaviour of "expensive" x-symbol/mmm options.
(interactive)
;; NB: might be nice to have queue region non read-only too, but
;; since the queue is constructed ahead of time, that wouldn't
;; work. (Better might be to refactor so that the region is
;; parsed as we go)
(proof-map-buffers
(proof-buffers-in-mode proof-mode-for-script)
(if (span-live-p proof-locked-span)
(proof-span-read-only proof-locked-span))))
(defsubst proof-set-queue-endpoints (start end)
"Set the queue span to be START, END."
(set-span-endpoints proof-queue-span start end))
(defsubst proof-set-locked-endpoints (start end)
"Set the locked span to be START, END."
(set-span-endpoints proof-locked-span start end))
(defsubst proof-detach-queue ()
"Remove the span for the queue region."
(and proof-queue-span (detach-span proof-queue-span)))
(defsubst proof-detach-locked ()
"Remove the span for the locked region."
(and proof-locked-span (detach-span proof-locked-span)))
(defsubst proof-set-queue-start (start)
"Set the queue span to begin at START."
(set-span-start proof-queue-span start))
(defsubst proof-set-locked-end (end)
"Set the end of the locked region to be END.
If END is at or before (point-min), remove the locked region.
Otherwise set the locked region to be from (point-min) to END."
;; FIXME: is it really needed to detach the span here?
(if (>= (point-min) end)
(proof-detach-locked)
(set-span-endpoints
proof-locked-span
(point-min)
(min (point-max) end) ;; safety: sometimes called with end>point-max(?)
)))
;; Reimplemented this to mirror above because of remaining
;; span problen
(defsubst proof-set-queue-end (end)
"Set the queue span to end at END."
(if (or (>= (point-min) end)
(<= end (span-start proof-queue-span)))
(proof-detach-queue)
(set-span-end proof-queue-span end)))
;; ** Initialise spans for a buffer
(defun proof-init-segmentation ()
"Initialise the queue and locked spans in a proof script buffer.
Allocate spans if need be. The spans are detached from the
buffer, so the regions are made empty by this function.
Also clear list of script portions."
;; Initialise queue span, remove it from buffer.
(unless proof-queue-span
(setq proof-queue-span (make-span 1 1))
;; FIXME: span-raise is an GNU hack to make locked span appear.
;; overlays still don't work as well as they did/should pre 99.
(span-raise proof-queue-span))
(set-span-property proof-queue-span 'start-closed t)
(set-span-property proof-queue-span 'end-open t)
(proof-span-read-only proof-queue-span 'always)
(set-span-property proof-queue-span 'face 'proof-queue-face)
(detach-span proof-queue-span)
;; Initialise locked span, remove it from buffer
(unless proof-locked-span
(setq proof-locked-span (make-span 1 1))
(span-raise proof-locked-span))
(set-span-property proof-locked-span 'start-closed t)
(set-span-property proof-locked-span 'end-open t)
(proof-span-read-only proof-locked-span)
(set-span-property proof-locked-span 'face 'proof-locked-face)
(detach-span proof-locked-span)
(setq proof-last-theorem-dependencies nil)
(setq proof-element-counters nil)
(pg-clear-script-portions))
;; ** Restarting and clearing spans
(defun proof-restart-buffers (buffers)
"Remove all extents in BUFFERS and maybe reset `proof-script-buffer'.
No effect on a buffer which is nil or killed. If one of the buffers
is the current scripting buffer, then `proof-script-buffer'
will deactivated."
(mapcar
(lambda (buffer)
(save-excursion
(if (buffer-live-p buffer)
(with-current-buffer buffer
(if proof-active-buffer-fake-minor-mode
(setq proof-active-buffer-fake-minor-mode nil))
(delete-spans (point-min) (point-max) 'type) ;; remove top-level spans
(delete-spans (point-min) (point-max) 'idiom) ;; and embedded spans
(setq pg-script-portions nil) ;; also the record of them
(proof-detach-queue) ;; remove queue and locked
(proof-detach-locked)
(proof-init-segmentation)))
(if (eq buffer proof-script-buffer)
(setq proof-script-buffer nil))))
buffers))
(defun proof-script-buffers-with-spans ()
"Return a list of all buffers with spans.
This is calculated by finding all the buffers with a non-nil
value of proof-locked span."
(let ((bufs-left (buffer-list))
bufs-got)
(dolist (buf bufs-left bufs-got)
(if (with-current-buffer buf proof-locked-span)
(setq bufs-got (cons buf bufs-got))))))
(defun proof-script-remove-all-spans-and-deactivate ()
"Remove all spans from scripting buffers via proof-restart-buffers."
(proof-restart-buffers (proof-script-buffers-with-spans)))
(defun proof-script-clear-queue-spans ()
"If there is an active scripting buffer, remove the queue span from it.
This is a subroutine used in proof-shell-handle-{error,interrupt}."
(if proof-script-buffer
(with-current-buffer proof-script-buffer
(proof-detach-queue)
;; FIXME da: point-max seems a bit excessive here,
;; proof-queue-or-locked-end should be enough.
(delete-spans (proof-locked-end) (point-max) 'type))))
;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;
;;
;; Buffer position functions
;;
(defun proof-unprocessed-begin ()
"Return end of locked region in current buffer or (point-min) otherwise.
The position is actually one beyond the last locked character."
(or
(and proof-locked-span
(span-end proof-locked-span))
(point-min)))
(defun proof-script-end ()
"Return the character beyond the last non-whitespace character.
This is the same position proof-locked-end ends up at when asserting
the script. Works for any kind of buffer."
(save-excursion
(goto-char (point-max))
(skip-chars-backward " \t\n")
(point)))
(defun proof-queue-or-locked-end ()
"Return the end of the queue region, or locked region, or (point-min).
This position should be the first writable position in the buffer.
An appropriate point to move point to (or make sure is displayed)
when a queue of commands is being processed."
(or
;; span-end returns nil if span is detatched
(and proof-queue-span (span-end proof-queue-span))
(and proof-locked-span (span-end proof-locked-span))
(point-min)))
;; FIXME: get rid of/rework this function. Some places expect this to
;; return nil if locked region is empty. Moreover, it confusingly
;; returns the point past the end of the locked region.
(defun proof-locked-end ()
"Return end of the locked region of the current buffer.
Only call this from a scripting buffer."
(proof-unprocessed-begin))
;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;
;;
;; Predicates for locked region.
;;
;; These work on any buffer, so that non-script buffers can be locked
;; (as processed files) too.
;;
(defun proof-locked-region-full-p ()
"Non-nil if the locked region covers all the buffer's non-whitespace.
Works on any buffer."
(save-excursion
(goto-char (point-max))
(skip-chars-backward " \t\n")
(>= (proof-unprocessed-begin) (point))))
(defun proof-locked-region-empty-p ()
"Non-nil if the locked region is empty. Works on any buffer."
(eq (proof-unprocessed-begin) (point-min)))
(defun proof-only-whitespace-to-locked-region-p ()
"Non-nil if only whitespace separates char after point from end of locked region.
Point should be after the locked region.
NB: If nil, point is left at first non-whitespace character found.
If non-nil, point is left where it was."
;; NB: this function doesn't quite do what you'd expect, but fixing it
;; breaks proof-assert-until-point and electric-terminator which
;; rely on the side effect. So careful!
;; (unless (eobp)
;; (forward-char))
;; (save-excursion -- no, side effect is expected!
(not (proof-re-search-backward "\\S-" (proof-unprocessed-begin) t)))
(defun proof-in-locked-region-p ()
"Non-nil if point is in locked region. Assumes proof script buffer current."
(< (point) (proof-unprocessed-begin)))
;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;
;;
;; Misc movement functions
;;
(defun proof-goto-end-of-locked (&optional switch)
"Jump to the end of the locked region, maybe switching to script buffer.
If interactive or SWITCH is non-nil, switch to script buffer first."
(interactive)
(proof-with-script-buffer
(if (and (not (get-buffer-window proof-script-buffer))
(or switch (interactive-p)))
(switch-to-buffer proof-script-buffer)
(goto-char (proof-unprocessed-begin)))))
;; Careful: movement can happen when the user is typing, not very nice!
(defun proof-goto-end-of-locked-if-pos-not-visible-in-window ()
"If the end of the locked region is not visible, jump to the end of it.
A possible hook function for `proof-shell-handle-error-or-interrupt-hook'.
Does nothing if there is no active scripting buffer, or if
`proof-follow-mode' is set to 'ignore."
(interactive)
(if (and proof-script-buffer
(not (eq proof-follow-mode 'ignore)))
(let ((pos (with-current-buffer proof-script-buffer
(proof-locked-end)))
(win (get-buffer-window proof-script-buffer t)))
(unless (and win (pos-visible-in-window-p pos))
(proof-goto-end-of-locked t)))))
(defun proof-goto-end-of-locked-on-error-if-pos-not-visible-in-window ()
"As `proof-goto-end-of-locked-if-pos-not-visible-in-window' except not for interrupts.
Intended as a hook function for `proof-shell-handle-error-or-interrupt-hook'."
(interactive)
(cond
((eq proof-shell-error-or-interrupt-seen 'error)
(proof-goto-end-of-locked-if-pos-not-visible-in-window))
((eq proof-shell-error-or-interrupt-seen 'interrupt)
(proof-with-current-buffer-if-exists
proof-script-buffer
;; Give a hint of how to jump to the end of locked, unless visible.
(let ((pos (with-current-buffer proof-script-buffer
(proof-locked-end)))
(win (get-buffer-window proof-script-buffer t)))
(unless (and win (pos-visible-in-window-p pos))
(pg-jump-to-end-hint)))))))
;; Simplified version of above for toolbar follow mode -- which wouldn't
;; work with abouve because of proof-shell-handle-error-or-interrupt-hook[?]
(defun proof-goto-end-of-queue-or-locked-if-not-visible ()
"Jump to the end of the queue region or locked region if it isn't visible.
Assumes script buffer is current"
(unless (pos-visible-in-window-p
(proof-queue-or-locked-end)
(get-buffer-window (current-buffer) t))
(goto-char (proof-queue-or-locked-end))))
;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;
;;
;; Names of proofs (and other elements) in a script
;;
;; Each kind of part ("idiom") in a proof script has its own name space.
;; Visibility within a script is then handled with buffer-invisibility-spec
;; controlling appearance of each idiom.
;;
(defvar pg-idioms '(proof)
"Vector of script element kinds PG is aware of for this prover.")
(defvar pg-visibility-specs nil
"Cache of visibility spec symbols used by PG.
This is used for cleaning `buffer-invisibility-spec' in
`pg-clear-script-portions': it doesn't need to be exactly accurate.")
(deflocal pg-script-portions nil
"Table of lists of symbols naming script portions which have been processed so far.")
(defun pg-clear-script-portions ()
"Clear record of script portion names and types from internal list.
Also clear all visibility specifications."
(setq pg-script-portions nil)
(setq buffer-invisibility-spec
(if (listp buffer-invisibility-spec)
;; FIXME: inefficient
(append
(mapcar (lambda (propellips)
(if (memq (car propellips) pg-visibility-specs)
nil (list propellips)))
buffer-invisibility-spec))
;; NB: if invisibility spec was t, we loose here: we need
;; it to be a list unfortunately, whereas other code may
;; want `t'. That's annoying.
;; (FIXME: should tell GNU/X Emacs developers).
nil)))
(defun pg-add-script-element (elt)
(add-to-list pg-script-portions elt))
(defun pg-remove-script-element (ns id)
(let* ((elts (cdr-safe (assq ns pg-script-portions)))
(newelts (remq id elts)))
(setq pg-script-portions
(if newelts
(cons (cons ns newelts) (remassq ns pg-script-portions))
(remassq ns pg-script-portions)))))
(defsubst pg-visname (namespace id)
"Return a unique symbol made from strings NAMESPACE and unique ID."
(intern (concat namespace ":" id)))
(defun pg-add-element (idiom id span &optional name)
"Add element of type IDIOM with identifier ID, referred to by SPAN.
This records the element in `pg-script-portions' and sets span
properties accordingly.
IDIOM, ID, and optional NAME are all strings.
Identifiers must be unique for a given idiom; the optional NAME
will be recorded as a textual name used instead of ID for users;
NAME does not need to be unique."
(let* ((idiomsym (intern idiom))
(idsym (intern id))
(name (or name id))
(visname (pg-visname idiom id))
(delfn `(lambda () (pg-remove-element (quote ,idiomsym) (quote ,idsym))))
(elts (cdr-safe (assq idiomsym pg-script-portions))))
(if elts
(if (memq id elts)
(proof-debug "Element named " name " (type " idiom ") already in buffer.")
(nconc elts (list idsym)))
(setq pg-script-portions (cons (cons idiomsym (list idsym))
pg-script-portions)))
;; Idiom and ID are stored in the span as symbols; name as a string.
(set-span-property span 'idiom idiomsym)
(set-span-property span 'id idsym)
(set-span-property span 'name name)
(set-span-property span 'span-delete-action delfn)
(set-span-property span 'invisible visname)
;; Bad behaviour if span gets copied: unique ID shouldn't be duplicated.
(set-span-property span 'duplicable nil) ;; NB: not supported in Emacs
;; Nice behaviour in with isearch: open invisible regions temporarily.
(set-span-property span 'isearch-open-invisible
'pg-open-invisible-span)
(set-span-property span 'isearch-open-invisible-temporary
'pg-open-invisible-span)))
(defun pg-open-invisible-span (span &optional invisible)
"Function for `isearch-open-invisible' property."
(let ((idiom (span-property span 'idiom))
(id (span-property span 'id)))
(and idiom id
(if invisible
(pg-make-element-invisible
(symbol-name idiom) id)
(pg-make-element-visible
(symbol-name idiom) (symbol-name id))))))
(defun pg-remove-element (ns idsym)
(pg-remove-script-element ns idsym)
;; We could leave the visibility note, but that may
;; be counterintuitive, so lets remove it.
(pg-make-element-visible (symbol-name ns) (symbol-name idsym))
(pg-redisplay-for-gnuemacs))
(defun pg-make-element-invisible (idiom id)
"Make element ID of type IDIOM invisible, with ellipsis."
(let ((visspec (cons (pg-visname idiom id) t)))
(add-to-list 'buffer-invisibility-spec visspec)
(add-to-list 'pg-visibility-specs visspec)))
(defun pg-make-element-visible (idiom id)
"Make element ID of type IDIOM visible."
(setq buffer-invisibility-spec
(remassq (pg-visname idiom id) buffer-invisibility-spec)))
(defun pg-toggle-element-visibility (idiom id)
"Toggle visibility of script element of type IDIOM, named ID."
(if (and (listp buffer-invisibility-spec)
(assq (pg-visname idiom id) buffer-invisibility-spec))
(pg-make-element-visible idiom id)
(pg-make-element-invisible idiom id))
(pg-redisplay-for-gnuemacs))
(defun pg-redisplay-for-gnuemacs ()
"GNU Emacs requires redisplay for changes in buffer-invisibility-spec."
(if proof-running-on-Emacs21
;; GNU Emacs requires redisplay here to see result
;; (sit-for 0) not enough
(redraw-frame (selected-frame))))
(defun pg-show-all-portions (idiom &optional hide)
"Show or hide all portions of kind IDIOM."
(interactive
(list
(completing-read
(concat "Make " (if current-prefix-arg "in" "") "visible all regions of: ")
(apply 'vector pg-idioms) nil t)
current-prefix-arg))
(let ((elts (cdr-safe (assq (intern idiom) pg-script-portions)))
(alterfn (if hide
(lambda (arg) (pg-make-element-invisible idiom
(symbol-name arg)))
(lambda (arg) (pg-make-element-visible idiom
(symbol-name arg))))))
(mapcar alterfn elts))
(pg-redisplay-for-gnuemacs))
;; Next two could be in pg-user.el. No key-bindings for these.
(defun pg-show-all-proofs ()
"Display all completed proofs in the buffer."
(interactive)
(pg-show-all-portions "proof"))
(defun pg-hide-all-proofs ()
"Hide all completed proofs in the buffer."
(interactive)
(pg-show-all-portions "proof" 'hide))
(defun pg-add-proof-element (name span controlspan)
"Add a nested span proof element."
(let ((proofid (proof-next-element-id 'proof)))
(pg-add-element "proof" proofid span name)
;; Set id in controlspan
(set-span-property controlspan 'id (intern proofid))
;; Make a navigable link between the two spans.
(set-span-property span 'controlspan controlspan)
(set-span-property controlspan 'children
(cons span (span-property controlspan 'children)))
(pg-set-span-helphighlights span 'nohighlight)
(if proof-disappearing-proofs
(pg-make-element-invisible "proof" proofid))))
(defun pg-span-name (span)
"Return a user-level name for SPAN."
(let ((type (span-property span 'type))
(idiom (span-property span 'idiom))
(name (span-property span 'name)))
(cond
(idiom
(concat (upcase-initials (symbol-name idiom))
(if (and name
(not (equal name proof-unnamed-theorem-name)))
(concat "[" name "]"))))
((or (eq type 'proof) (eq type 'goalsave))
(concat "Proof"
(let ((name (span-property span 'name)))
(if name (concat " of " name)))))
((eq type 'comment) "Comment")
((eq type 'vanilla) "Command")
((eq type 'pbp) "PBP command")
((eq type 'proverproc)
"Prover-processed"))))
(defun pg-set-span-helphighlights (span &optional nohighlight)
"Set the help echo message, default highlight, and keymap for SPAN."
(let ((helpmsg (concat (pg-span-name span) ""))) ;; " region"
(set-span-property span 'balloon-help helpmsg)
(if pg-show-hints ;; only message in minibuf if hints on
(set-span-property
span 'help-echo
(substitute-command-keys
(concat
helpmsg
" ("
(if (span-property span 'idiom)
"with point in region, \\[pg-toggle-visibility] to show/hide; ")
"\\[popup-mode-menu] for menu)"))))
(set-span-property span 'keymap pg-span-context-menu-keymap)
(unless nohighlight
(set-span-property span 'mouse-face 'proof-mouse-highlight-face))))
;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;
;;
;; Multiple file handling
;;
(defun proof-complete-buffer-atomic (buffer)
"Make sure BUFFER is marked as completely processed, completing with a single step.
If buffer already contains a locked region, only the remainder of the
buffer is closed off atomically.
This works for buffers which are not in proof scripting mode too,
to allow other files loaded by proof assistants to be marked read-only."
;; NB: this isn't quite right, because not all of the structure in the
;; locked region will be preserved when processing across several
;; files. In particular, the span for a currently open goal should be
;; removed. Keeping the structure is an approximation to make up for
;; the fact that that no structure is created by loading files via the
;; proof assistant. Future ideas: proof assistant could ask Proof
;; General to do the loading, to alleviate file handling there;
;; we could cache meta-data resulting from processing files;
;; or even, could include parsing inside PG.
(save-excursion
(set-buffer buffer)
(save-excursion ;; prevent point moving if user viewing file
(if (< (proof-unprocessed-begin) (proof-script-end))
(let ((span (make-span (proof-unprocessed-begin)
(proof-script-end)))
dummycmd)
;; Reset queue and locked regions.
(proof-init-segmentation)
;; End of locked region is always end of buffer
(proof-set-locked-end (proof-script-end))
;; Configure the overlay span
(set-span-property span 'type 'proverproc)
;; A dummy command for retraction which examines it
;; FIXME: shouldn't be necessary really
(set-span-property span 'dummycmd "")
(pg-set-span-helphighlights span 'nohighlight))))))
;; FIXME da: cleanup of odd asymmetry here: we have a nice setting for
;; proof-register-possibly-new-processed-file but something much more
;; complicated for retracting, because we allow a hook function
;; to calculate the new included files list.
(defun proof-register-possibly-new-processed-file (file &optional informprover noquestions)
"Register a possibly new FILE as having been processed by the prover.
If INFORMPROVER is non-nil, the proof assistant will be told about this,
to co-ordinate with its internal file-management. (Otherwise we assume
that it is a message from the proof assistant which triggers this call).
In this case, the user will be queried to save some buffers, unless
NOQUESTIONS is non-nil.
No action is taken if the file is already registered.
A warning message is issued if the register request came from the
proof assistant and Emacs has a modified buffer visiting the file."
(let* ((cfile (file-truename file))
(buffer (proof-file-to-buffer cfile)))
(proof-debug (concat "Registering file " cfile
(if (member cfile proof-included-files-list)
" (already registered, no action)." ".")))
(unless (member cfile proof-included-files-list)
(and buffer
(not informprover)
(buffer-modified-p buffer)
(proof-warning (concat "Changes to "
(buffer-name buffer)
" have not been saved!")))
;; Add the new file onto the front of the list
(setq proof-included-files-list
(cons cfile proof-included-files-list))
;; If the file is loaded into a buffer, make sure it is completely locked
(if buffer
(proof-complete-buffer-atomic buffer))
;; Tell the proof assistant, if we should and if we can
(if (and informprover proof-shell-inform-file-processed-cmd)
(progn
;; Markus suggests we should ask if the user wants to save
;; the file now (presumably because the proof assistant
;; might examine the file timestamp, or attempt to visit
;; the file later??).
;; Presumably it would be enough to ask about this file,
;; not all files?
(if (and
proof-query-file-save-when-activating-scripting
(not noquestions))
(unwind-protect
(save-some-buffers)))
;; Tell the prover
(proof-shell-invisible-command
(proof-format-filename proof-shell-inform-file-processed-cmd
cfile)
'wait))))))
(defun proof-inform-prover-file-retracted (rfile)
(if proof-shell-inform-file-retracted-cmd
(proof-shell-invisible-command
(proof-format-filename proof-shell-inform-file-retracted-cmd
rfile)
'wait)))
(defun proof-auto-retract-dependencies (cfile &optional informprover)
"Perhaps automatically retract the (linear) dependencies of CFILE.
If `proof-auto-multiple-files' is nil, no action is taken.
If CFILE does not appear on `proof-included-files-list', no action taken.
Any buffers which are visiting files in `proof-included-files-list'
before CFILE are retracted using proof-protected-process-or-retract.
They are retracted in reverse order.
Since the `proof-included-files-list' is examined, we expect scripting
to be turned off before calling here (because turning it off could
otherwise change `proof-included-files-list').
If INFORMPROVER is non-nil, the proof assistant will be told about this,
using `proof-shell-inform-file-retracted-cmd', to co-ordinate with its
internal file-management.
Files which are not visited by any buffer are not retracted, on the
basis that we may not have the information necessary to retract them
-- spans that cover the buffer with definition/declaration
information. A warning message is given for these cases, since it
could cause inconsistency problems.
NB! Retraction can cause recursive calls of this function.
This is a subroutine for proof-unregister-buffer-file-name."
(if proof-auto-multiple-files
(let ((depfiles (reverse
(cdr-safe
(member cfile (reverse proof-included-files-list)))))
rfile rbuf)
(while (setq rfile (car-safe depfiles))
;; If there's a buffer visiting a dependent file, retract it.
;; We test that the file to retract hasn't been retracted
;; already by a recursive call here. (But since we do retraction
;; in reverse order, this shouldn't happen...)
(if (and (member rfile proof-included-files-list)
(setq rbuf (proof-file-to-buffer rfile)))
(progn
(proof-debug "Automatically retracting " rfile)
(proof-protected-process-or-retract 'retract rbuf)
(setq proof-included-files-list
(delete rfile proof-included-files-list))
;; Tell the proof assistant, if we should and we can.
;; This may be useful if we synchronise the *prover* with
;; PG's management of multiple files. If the *prover*
;; informs PG (better case), then we hope the prover will
;; retract dependent files and we shouldn't use this
;; degenerate (linear dependency) code.
(if informprover
(proof-inform-prover-file-retracted rfile)))
;; If no buffer available, issue a warning that nothing was done
(proof-warning "Not retracting unvisited file " rfile))
(setq depfiles (cdr depfiles))))))
(defun proof-unregister-buffer-file-name (&optional informprover)
"Remove current buffer's filename from the list of included files.
No effect if the current buffer has no file name.
If INFORMPROVER is non-nil, the proof assistant will be told about this,
using `proof-shell-inform-file-retracted-cmd', to co-ordinate with its
internal file-management.
If `proof-auto-multiple-files' is non-nil, any buffers on
`proof-included-files-list' before this one will be automatically
retracted using proof-auto-retract-dependencies."
(if buffer-file-name
(let ((cfile (file-truename
(or buffer-file-name
proof-script-buffer-file-name))))
(proof-debug (concat "Unregistering file " cfile
(if (not (member cfile
proof-included-files-list))
" (not registered, no action)." ".")))
(if (member cfile proof-included-files-list)
(progn
(proof-auto-retract-dependencies cfile informprover)
(setq proof-included-files-list
(delete cfile proof-included-files-list))
;; Tell the proof assistant, if we should and we can.
;; This case may be useful if there is a combined
;; management of multiple files between PG and prover.
(if informprover
(proof-inform-prover-file-retracted cfile)))))))
;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;
;;
;; Activating and Deactivating Scripting
;;
;; The notion of "active scripting buffer" clarifies how
;; scripting across multiple files is handled. Only one
;; script buffer is allowed to be active, and actions are
;; taken when scripting is turned off/on.
;;
(defun proof-protected-process-or-retract (action &optional buffer)
"If ACTION='process, process, If ACTION='retract, retract.
Process or retract the current buffer, which should be the active
scripting buffer, according to ACTION.
Retract buffer BUFFER if set, otherwise use the current buffer.
Gives a message in the minibuffer and busy-waits for the retraction
or processing to complete. If it fails for some reason,
an error is signalled here."
(let ((fn (cond ((eq action 'process) 'proof-process-buffer)
((eq action 'retract) 'proof-retract-buffer)))
(name (cond ((eq action 'process) "Processing")
((eq action 'retract) "Retracting")))
(buf (or buffer (current-buffer))))
(if fn
(unwind-protect
(with-current-buffer buf
(message "%s buffer %s..." name buf)
(funcall fn)
(while proof-shell-busy ; busy wait
(sit-for 1))
(message "%s buffer %s...done." name buf)
(sit-for 0))
;; Test to see if action was successful
(with-current-buffer buf
(or (and (eq action 'retract) (proof-locked-region-empty-p))
(and (eq action 'process) (proof-locked-region-full-p))
(error "%s of %s failed!" name buf)))))))
(defun proof-deactivate-scripting-auto ()
"Deactivate scripting without asking questions or raising errors.
If the locked region is full, register the file as processed.
Otherwise retract it. Errors are ignored"
(ignore-errors
(proof-deactivate-scripting
(proof-with-script-buffer
(if (proof-locked-region-full-p) 'process 'retract)))))
(defun proof-deactivate-scripting (&optional forcedaction)
"Deactivate scripting for the active scripting buffer.
Set `proof-script-buffer' to nil and turn off the modeline indicator.
No action if there is no active scripting buffer.
We make sure that the active scripting buffer either has no locked
region or a full locked region (everything in it has been processed).
If this is not already the case, we question the user whether to
retract or assert, or automatically take the action indicated in the
user option `proof-auto-action-when-deactivating-scripting.'
If the scripting buffer is (or has become) fully processed, and it is
associated with a file, it is registered on
`proof-included-files-list'. Conversely, if it is (or has become)
empty, we make sure that it is *not* registered. This is to be
certain that the included files list behaves as we might expect with
respect to the active scripting buffer, in an attempt to harmonize
mixed scripting and file reading in the prover.
This function either succeeds, fails because the user refused to
process or retract a partly finished buffer, or gives an error message
because retraction or processing failed. If this function succeeds,
then `proof-script-buffer' is nil afterwards.
The optional argument FORCEDACTION overrides the user option
`proof-auto-action-when-deactivating-scripting' and prevents
questioning the user. It is used to make a value for
the `kill-buffer-hook' for scripting buffers, so that when
a scripting buffer is killed it is always retracted."
(interactive)
(if proof-script-buffer
(with-current-buffer proof-script-buffer
;; Examine buffer.
;; We must ensure that the locked region is either
;; empty or full, to make sense for multiple-file
;; scripting. (A proof assistant won't be able to
;; process just part of a file typically; moreover
;; switching between buffers during a proof makes
;; no sense.)
(if (or (proof-locked-region-empty-p)
(proof-locked-region-full-p)
;; Buffer is partly-processed
(let*
((action
(or
forcedaction
proof-auto-action-when-deactivating-scripting
(progn
(save-window-excursion
(unless
;; Test to see whether to display the
;; buffer or not.
;; Could have user option here to avoid switching
;; or maybe borrow similar standard setting
;; save-some-buffers-query-display-buffer
(or
(eq (current-buffer)
(window-buffer (selected-window)))
(eq (selected-window) (minibuffer-window)))
(progn
(unless (one-window-p)
(delete-other-windows))
(switch-to-buffer proof-script-buffer t)))
;; Would be nicer to ask a single question, but
;; a nuisance to define our own dialogue since it
;; doesn't really fit with one of the standard ones.
(cond
((y-or-n-p
(format
"Scripting incomplete in buffer %s, retract? "
proof-script-buffer))
'retract)
((y-or-n-p
(format
"Completely process buffer %s instead? "
proof-script-buffer))
'process)))))))
;; Take the required action
(if action
(proof-protected-process-or-retract action)
;; Give an acknowledgement to user's choice
;; neither to assert or retract.
(message "Scripting still active in %s"
proof-script-buffer)
;; Delay because this can be followed by an error
;; message in proof-activate-scripting when trying
;; to switch to another scripting buffer.
(sit-for 1)
nil)))
;; If we get here, then the locked region is (now) either
;; completely empty or completely full.
(progn
;; We can immediately indicate that there is no active
;; scripting buffer
(setq proof-previous-script-buffer proof-script-buffer)
(setq proof-script-buffer nil)
(if (proof-locked-region-full-p)
;; If locked region is full, make sure that this buffer
;; is registered on the included files list, and
;; let the prover know it can consider it processed.
(if (or buffer-file-name proof-script-buffer-file-name)
(proof-register-possibly-new-processed-file
(or buffer-file-name proof-script-buffer-file-name)
'tell-the-prover
forcedaction)))
(if (proof-locked-region-empty-p)
;; If locked region is empty, make sure this buffer is
;; *off* the included files list.
;; FIXME: probably this isn't necessary: the
;; file should be unregistered by the retract
;; action, or in any case since it was only
;; partly processed.
;; FIXME 2: be careful about automatic
;; multiple file handling here, since it calls
;; for activating scripting elsewhere.
;; We move the onus on unregistering now to
;; the activate-scripting action.
(proof-unregister-buffer-file-name))
;; Turn off Scripting indicator here.
(setq proof-active-buffer-fake-minor-mode nil)
;; Make status of inactive scripting buffer show up
;; FIXME da:
;; not really necessary when called by kill buffer, at least.
(if (fboundp 'redraw-modeline)
(redraw-modeline)
(force-mode-line-update)))))))
(defun proof-activate-scripting (&optional nosaves queuemode)
"Ready prover and activate scripting for the current script buffer.
The current buffer is prepared for scripting. No changes are
necessary if it is already in Scripting minor mode. Otherwise, it
will become the new active scripting buffer, provided scripting
can be switched off in the previous active scripting buffer
with `proof-deactivate-scripting'.
Activating a new script buffer may be a good time to ask if the
user wants to save some buffers; this is done if the user
option `proof-query-file-save-when-activating-scripting' is set
and provided the optional argument NOSAVES is non-nil.
The optional argument QUEUEMODE relaxes the test for a
busy proof shell to allow one which has mode QUEUEMODE.
In all other cases, a proof shell busy error is given.
Finally, the hooks `proof-activate-scripting-hook' are run.
This can be a useful place to configure the proof assistant for
scripting in a particular file, for example, loading the
correct theory, or whatever. If the hooks issue commands
to the proof assistant (via `proof-shell-invisible-command')
which result in an error, the activation is considered to
have failed and an error is given."
(interactive)
;; FIXME: the scope of this save-excursion is rather wide.
;; Problems without it however: Use button behaves oddly
;; when process is started already.
;; Where is save-excursion needed?
;; First experiment shows that it's the hooks that cause
;; problem, maybe even the use of proof-cd-sync (can't see why).
(save-excursion
;; FIXME: proof-shell-ready-prover here s
(proof-shell-ready-prover queuemode)
(cond
((not (eq proof-buffer-type 'script))
(error "Must be running in a script buffer!"))
;; If the current buffer is the active one there's nothing to do.
((equal (current-buffer) proof-script-buffer))
;; Otherwise we need to activate a new Scripting buffer.
(t
;; If there's another buffer currently active, we need to
;; deactivate it (also fixing up the included files list).
(if proof-script-buffer
(progn
(proof-deactivate-scripting)
;; Test whether deactivation worked
(if proof-script-buffer
(error
"You cannot have more than one active scripting buffer!"))))
;; Now make sure that this buffer is off the included files
;; list. In case we re-activate scripting in an already
;; completed buffer, it may be that the proof assistant
;; needs to retract some of this buffer's dependencies.
(proof-unregister-buffer-file-name 'tell-the-prover)
;; If automatic retraction happened in the above step, we may
;; have inadvertently activated scripting somewhere else.
;; Better turn it off again. This should succeed trivially.
;; NB: it seems that we could move the first test for an already
;; active buffer here, but it is more subtle: the first
;; deactivation can extend the proof-included-files list, which
;; would affect what retraction was done in
;; proof-unregister-buffer-file-name.
(if proof-script-buffer
(proof-deactivate-scripting))
(assert (null proof-script-buffer)
"Bug in proof-activate-scripting: deactivate failed.")
;; Set the active scripting buffer, and initialise the
;; queue and locked regions if necessary.
(setq proof-script-buffer (current-buffer))
(if (proof-locked-region-empty-p)
;; This removes any locked region that was there, but
;; sometimes we switch on scripting in "full" buffers,
;; so mustn't do this.
(proof-init-segmentation))
;; Turn on the minor mode, make it show up.
(setq proof-active-buffer-fake-minor-mode t)
(if (fboundp 'redraw-modeline)
(redraw-modeline)
(force-mode-line-update))
;; This may be a good time to ask if the user wants to save some
;; buffers. On the other hand, it's jolly annoying to be
;; queried on the active scripting buffer if we've started
;; writing in it. So pretend that one is unmodified, at least
;; (we certainly don't expect the proof assitant to load it)
(if (and
proof-query-file-save-when-activating-scripting
(not nosaves))
(let ((modified (buffer-modified-p)))
(set-buffer-modified-p nil)
(unwind-protect
(save-some-buffers)
(set-buffer-modified-p modified))))
;; Run hooks with a variable which suggests whether or not
;; to block. NB: The hook function may send commands to the
;; process which will re-enter this function, but should exit
;; immediately because scripting has been turned on now.
(if proof-activate-scripting-hook
(let
((activated-interactively (interactive-p)))
;; Clear flag in case no hooks run shell commands
(setq proof-shell-error-or-interrupt-seen nil)
(run-hooks 'proof-activate-scripting-hook)
;; In case the activate scripting functions
;; caused an error in the proof assistant, we'll
;; consider activating scripting to have failed,
;; and raise an error here.
;; (Since this behaviour is intimate with the hook functions,
;; it could be removed and left to implementors of
;; specific instances of PG).
;; FIXME: we could consider simply running the hooks
;; as the last step before turning on scripting properly,
;; provided the hooks don't inspect proof-script-buffer.
(if proof-shell-error-or-interrupt-seen
(progn
(proof-deactivate-scripting) ;; turn it off again!
;; Give an error to prevent further actions.
(error "Proof General: Scripting not activated because of error or interrupt")))))))))
(defun proof-toggle-active-scripting (&optional arg)
"Toggle active scripting mode in the current buffer.
With ARG, turn on scripting iff ARG is positive."
(interactive "P")
;; A little less obvious than it may seem: toggling scripting in the
;; current buffer may involve turning it off in some other buffer
;; first!
(if (if (null arg)
(not (eq proof-script-buffer (current-buffer)))
(> (prefix-numeric-value arg) 0))
(progn
(if proof-script-buffer
(call-interactively 'proof-deactivate-scripting))
(call-interactively 'proof-activate-scripting))
(call-interactively 'proof-deactivate-scripting)))
;;
;; End of activating and deactivating scripting section
;;
;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;
;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;
;;
;; Processing the script management queue -- PART 1: "advancing"
;;
;; The proof-action-list contains a list of (span,command,action)
;; triples. The loop looks like: Execute the command, and if it's
;; successful, do action on span. If the command's not successful, we
;; bounce the rest of the queue and do some error processing.
;;
;; When a span has been processed, it is classified as
;; 'comment, 'goalsave, 'vanilla, etc.
;;
;; The main function for dealing with processed spans is
;; `proof-done-advancing'
(defun proof-done-advancing (span)
"The callback function for assert-until-point."
;; FIXME da: if the buffer dies, this function breaks horribly.
(if (not (span-live-p span))
;; NB: Sometimes this function is called with a destroyed
;; extent as argument. Seems odd.
(proof-debug
"Proof General idiosyncrasy: proof-done-advancing called with a dead span!")
;; otherwise...
(let ((end (span-end span))
(cmd (span-property span 'cmd)))
;; State of spans after advancing:
(proof-set-locked-end end)
;; FIXME: buglet here, can sometimes arrive with queue span already detached.
;; (I think when complete file process is requested during scripting)
(if (span-live-p proof-queue-span)
(proof-set-queue-start end))
(cond
;; CASE 1: Comments just get highlighted
((eq (span-property span 'type) 'comment)
(proof-done-advancing-comment span))
;; CASE 2: Save command seen, now we may amalgamate spans.
((and proof-save-command-regexp
(proof-string-match proof-save-command-regexp cmd)
;; FIXME: would like to get rid of proof-really-save-command-p
;; and use nested goals mechanism instead.
(funcall proof-really-save-command-p span cmd)
(decf proof-nesting-depth) ;; [always non-nil/true]
(if proof-nested-goals-history-p
;; For now, we only support this nesting behaviour:
;; don't amalgamate unless the nesting depth is 0,
;; i.e. we're in a top-level proof.
;; This assumes prover keeps history for nested proofs.
;; (True for Isabelle/Isar).
(eq proof-nesting-depth 0)
t))
(proof-done-advancing-save span))
;; CASE 3: Proof completed one step or more ago, non-save
;; command seen, no nested goals allowed.
;;
;; We make a fake goal-save from any previous
;; goal to the command before the present one.
;;
;; This allows smooth undoing in proofs which have no "qed"
;; statements. If your proof assistant doesn't allow these
;; "unclosed" proofs, then you can safely set
;; proof-completed-proof-behaviour.
((and
proof-shell-proof-completed
(or (and (eq proof-completed-proof-behaviour 'extend)
(>= proof-shell-proof-completed 0))
(and (eq proof-completed-proof-behaviour 'closeany)
(> proof-shell-proof-completed 0))
(and (eq proof-completed-proof-behaviour 'closegoal)
(funcall proof-goal-command-p cmd))))
(proof-done-advancing-autosave span))
;; CASE 4: Some other kind of command (or a nested goal).
(t
(proof-done-advancing-other span))))
;; Finally: state of scripting may have changed now, run hooks.
(run-hooks 'proof-state-change-hook)))
(defun proof-done-advancing-comment (span)
"A subroutine of `proof-done-advancing'."
;; Add an element for a new span, which should span
;; the text of the comment.
;; FIXME: might be better to use regexps for matching
;; start/end of comments, rather than comment-start-skip
(let ((bodyspan (make-span
(+ (length comment-start) (span-start span))
(- (span-end span) (length comment-end))))
(id (proof-next-element-id 'comment)))
(pg-add-element "comment" id bodyspan)
(set-span-property span 'id (intern id))
(set-span-property span 'idiom 'comment)
(pg-set-span-helphighlights span)))
(defun proof-done-advancing-save (span)
"A subroutine of `proof-done-advancing'."
(unless (eq proof-shell-proof-completed 1)
;; We expect saves to succeed only for recently completed proofs.
;; Give a hint to the proof assistant implementor in case something
;; odd is going on. (NB: this is normal for nested proofs, though).
(proof-debug
(format
"PG: save command with proof-shell-proof-completed=%s, proof-nesting-depth=%s"
proof-shell-proof-completed proof-nesting-depth)))
(setq proof-shell-proof-completed nil)
;; FIXME: need subroutine here:
(let ((gspan span) ; putative goal span
(savestart (span-start span))
(saveend (span-end span))
(cmd (span-property span 'cmd))
lev nestedundos nam next)
;; Try to set the name of the theorem from the save
;; (message "%s" cmd) 3.4: remove this message.
(and proof-save-with-hole-regexp
(proof-string-match proof-save-with-hole-regexp cmd)
;; Give a message of a name if one can be determined
(message "%s"
(setq nam
(if (stringp proof-save-with-hole-result)
(replace-match proof-save-with-hole-result nil nil cmd)
(match-string proof-save-with-hole-result cmd)))))
;; Search backwards for matching goal command, deleting spans
;; along the way: they will be amalgamated into a single
;; goal-save region, which corresponds to the prover
;; discarding the proof history.
;; Provers without flat history yet nested proofs (i.e. Coq)
;; need to keep a record of the undo count for nested goalsaves.
;; FIXME: should also remove nested 'idiom spans, perhaps.
(setq lev 1)
(setq nestedundos 0)
(while (and gspan (> lev 0))
(setq next (prev-span gspan 'type))
(delete-span gspan)
(setq gspan next)
(if gspan
(progn
(setq cmd (span-property gspan 'cmd))
(cond
;; Ignore comments [may have null cmd setting]
((eq (span-property gspan 'type) 'comment))
;; Nested goal saves: add in any nestedcmds
((eq (span-property gspan 'type) 'goalsave)
(setq nestedundos
(+ nestedundos 1
(or (span-property gspan 'nestedundos) 0))))
;; Increment depth for a nested save, in case
;; prover supports history of nested proofs
((and proof-nested-goals-history-p
proof-save-command-regexp
(proof-string-match proof-save-command-regexp cmd))
(incf lev))
;; Decrement depth when a goal is hit
((funcall proof-goal-command-p cmd)
(decf lev))
;; Remainder cases: see if command matches something we
;; should count for a global undo
((and proof-nested-undo-regexp
(proof-string-match proof-nested-undo-regexp cmd))
(incf nestedundos))
))))
(if (not gspan)
;; No goal span found! Issue a warning and do nothing more.
(proof-warning
"Proof General: script management confused, couldn't find goal span for save.")
;; If the name isn't set, try to set it from the goal,
;; or as a final desparate attempt, set the name to
;; proof-unnamed-theorem-name (Coq actually uses a default
;; name for unnamed theorems, believe it or not, and issues
;; a name-binding error for two unnamed theorems in a row!).
(setq nam (or nam
(proof-get-name-from-goal gspan)
proof-unnamed-theorem-name))
(proof-make-goalsave gspan (span-end gspan)
savestart saveend nam nestedundos)
;; *** Theorem dependencies ***
(if proof-last-theorem-dependencies
(proof-depends-process-dependencies nam gspan)))))
(defun proof-make-goalsave (gspan goalend savestart saveend nam &optional nestedundos)
"Make new goal-save span, using GSPAN. Subroutine of `proof-done-advancing-save'."
(set-span-end gspan saveend)
(set-span-property gspan 'type 'goalsave)
(set-span-property gspan 'idiom 'proof);; links to nested proof element
(set-span-property gspan 'name nam)
(and nestedundos (set-span-property gspan 'nestedundos nestedundos))
(pg-set-span-helphighlights gspan)
;; Now make a nested span covering the purported body of the
;; proof, and add to buffer-local list of elements.
(let ((proofbodyspan
(make-span goalend (if proof-script-integral-proofs
saveend savestart))))
(pg-add-proof-element nam proofbodyspan gspan)))
(defun proof-get-name-from-goal (gspan)
"Try to return a goal name from GSPAN. Subroutine of `proof-done-advancing-save'."
(let ((cmdstr (span-property gspan 'cmd)))
(and proof-goal-with-hole-regexp
(proof-string-match proof-goal-with-hole-regexp cmdstr)
(if (stringp proof-goal-with-hole-result)
(replace-match proof-goal-with-hole-result nil nil cmdstr)
(match-string proof-goal-with-hole-result cmdstr)))))
;; FIXME: this next function should be more like proof-done-advancing-save,
;; perhaps simplifying the proof-completed-proof-behaviour functionality,
;; which isn't seriously used in any prover. At the moment the behaviour
;; here is incomplete compared with proof-done-advancing-save.
;; NB: in this function we assume non-nested proofs, which explains
;; some of the logic. There is no attempt to fix up proof-nesting-depth.
;; NB: 'extend behaviour is not currently compatible with appearance of
;; save commands, since proof-done-advancing-save allow for goalspan
;; already existing.
(defun proof-done-advancing-autosave (span)
"A subroutine of `proof-done-advancing'."
;; In the extend case, the context of proof grows until hit a save
;; or new goal.
(if (eq proof-completed-proof-behaviour 'extend)
(incf proof-shell-proof-completed)
(setq proof-shell-proof-completed nil))
(let* ((swallow (eq proof-completed-proof-behaviour 'extend))
(gspan (if swallow span (prev-span span 'type)))
(newend (if swallow (span-end span) (span-start span)))
(cmd (span-property span 'cmd))
(newgoal (funcall proof-goal-command-p cmd))
nam hitsave dels ncmd)
;; Search backwards to see if we can find a previous goal
;; before a save or the start of the buffer.
;; FIXME: this should really do the work done in
;; proof-done-advancing-save above, too, with nested undos, etc.
(while ;; YUK!
(and
gspan
(or
(eq (span-property gspan 'type) 'comment)
;;(eq (span-property gspan 'type) 'proverproc) ;; NB: not necess currently
(and
(setq ncmd (span-property gspan 'cmd))
(not (funcall proof-goal-command-p (setq cmd ncmd)))
(not
(and proof-save-command-regexp
(proof-string-match proof-save-command-regexp cmd)
(funcall proof-really-save-command-p span cmd)
(setq hitsave t))))))
(setq dels (cons gspan dels))
(setq gspan (prev-span gspan 'type)))
(cond
((or hitsave (null gspan))
(proof-debug
"Proof General strangeness: unclosed proof completed, but couldn't find its start!")
(pg-set-span-helphighlights span))
((and swallow newgoal)
;; If extending the region, goalsave already there; just highlight new region
(setq proof-shell-proof-completed nil)
(pg-set-span-helphighlights span))
(t
;; If, search back through spans, we haven't hit a save or the
;; start of the buffer, we make a fake goal-save region.
;; Delete spans between the previous goal and new command
(mapcar 'delete-span dels)
;; Try to set the name from the goal... [as above]
(setq nam (or (proof-get-name-from-goal gspan)
proof-unnamed-theorem-name))
;; NB: if extending an already closed region, ought to delete
;; the body and extend that too: currently we make multiple nested
;; bodies, a bit messy.
;; (NB: savestart used for nested region: here use saveend)
(proof-make-goalsave gspan
(+ (span-start gspan)
(length (or (span-property-safe gspan 'cmd))))
newend newend nam)))))
(defun proof-done-advancing-other (span)
;; We might add hidable regions also for commands: the problem
;; is that they have no natural surrounding region, so makes
;; it difficult to define a region for revealing again.
;; [ best solution would be to support clicking on ellipsis]
(if (funcall proof-goal-command-p (span-property span 'cmd))
(incf proof-nesting-depth))
(if proof-shell-proof-completed
(incf proof-shell-proof-completed))
(pg-set-span-helphighlights span))
;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;
;;
;; Parsing functions for parsing commands in script
;;
;; Command parsing is suprisingly subtle with various possibilities of
;; command syntax (terminated, not terminated, or lisp-style); whether
;; or not PG silently ignores comments, etc.
;; FIXME: currently there's several sets of functions. We need to be
;; more general and a bit more clear, but the latest versions are a
;; much better attempt.
(defun proof-segment-up-to-parser (pos &optional next-command-end)
"Parse the script buffer from end of locked to POS.
Return a list of (type, string, int) tuples (in reverse order).
Each tuple denotes the command and the position of its final character,
type is one of 'comment, or 'cmd.
The behaviour around comments is set by
`proof-script-fly-past-comments', which see.
This version is used when `proof-script-parse-function' is set,
to the function which parses the script segment by segment."
(save-excursion
(let* ((start (goto-char (proof-queue-or-locked-end)))
(cur (1- start))
(seg t)
prevtype realstart cmdseen segs)
;; Keep parsing until:
;; - we fail to find a segment (seg = nil)
;; - we go beyond the stop point (cur >= end)
;; - unless we're flying past comments, in which case
;; wait for a command (cmdseen<>nil)
(while (and seg
(or (< cur pos)
(and proof-script-fly-past-comments
(not cmdseen))))
;; Skip whitespace before this element
(skip-chars-forward " \t\n")
(setq realstart (point))
(let* ((type (funcall proof-script-parse-function)))
(setq seg nil)
(cond
((eq type 'comment)
(setq seg (list 'comment "" (point))))
((eq type 'cmd)
(setq cmdseen t)
(setq seg (list
'cmd
(buffer-substring realstart (point))
(point))))
((null type)) ; nothing left in buffer
(t
(error
"Proof-segment-up-to-parser: bad TYPE value from proof-script-parse-function.")))
;;
(if seg
(progn
;; Add the new segment, coalescing comments if
;; the user likes it that way. I first made
;; coalescing a separate configuration option, but
;; it works well used in tandem with the fly-past
;; behaviour.
(if (and proof-script-fly-past-comments
(eq type 'comment)
(eq prevtype 'comment))
(setq segs (cons seg (cdr segs)))
(setq segs (cons seg segs)))
;; Update state
(setq cur (point))
(setq prevtype type)))))
;; Return segment list
segs)))
(defun proof-script-generic-parse-find-comment-end ()
"Find the end of the comment point is at the start of. Nil if not found."
(let ((notout t))
;; Find end of comment (NB: doesn't undertand nested comments)
(while (and notout (re-search-forward
proof-script-comment-end-regexp nil 'movetolimit))
(setq notout (proof-buffer-syntactic-context)))
(not (proof-buffer-syntactic-context))))
(defun proof-script-generic-parse-cmdend ()
"Used for `proof-script-parse-function' if `proof-script-command-end-regexp' is set."
(if (looking-at proof-script-comment-start-regexp)
;; Handle comments
(if (proof-script-generic-parse-find-comment-end) 'comment)
;; Handle non-comments: assumed to be commands
(let (foundend)
;; Find end of command
(while (and (setq foundend
(progn
(and
(re-search-forward proof-script-command-end-regexp nil t)
(or (match-beginning 1) ;; optional start of white space
(match-end 0)))))
(proof-buffer-syntactic-context))
;; inside a string or comment before the command end
)
(if (and foundend
(goto-char foundend) ; move to command end
(not (proof-buffer-syntactic-context)))
;; Found command end outside string/comment
'cmd
;; Didn't find command end
nil))))
(defun proof-script-generic-parse-cmdstart ()
"For `proof-script-parse-function' if `proof-script-command-start-regexp' is set."
;; This was added for the fine-grained command structure of Isar
;;
;; It's is a lot more involved than the case of just scanning for
;; the command end; we have to find two successive command starts
;; and go backwards from the second. This coalesces comments
;; following commands with commands themselves, and sends them to
;; the prover (only case where it does). It's needed particularly
;; for Isar's text command (text {* foo *}) so we can define the
;; buffer syntax for text as a comment.
;;
;; To avoid doing that, we would need to scan also for comments but
;; it would be difficult to distinguish between:
;; complete command (* that's it *)
;; and
;; complete (* almost *) command
;;
;; Maybe the second case should be disallowed in command-start regexp case?
;;
;; Another improvement idea might be to take into account both
;; command starts *and* ends, but let's leave that for another day.
;;
;; NB: proof-script-comment-start-regexp doesn't need to be the same
;; as (reqexp-quote comment-start).
(if (looking-at proof-script-comment-start-regexp)
;; Find end of comment
(if (proof-script-generic-parse-find-comment-end) 'comment)
;; Handle non-comments: assumed to be commands
(if (looking-at proof-script-command-start-regexp)
(progn
;; We've got at least the beginnings of a command, skip past it
(goto-char (match-end 0))
(let (foundstart)
;; Find next command start
(while (and (setq
foundstart
(and
(re-search-forward proof-script-command-start-regexp
nil 'movetolimit)
(and (match-beginning 0)
;; jiggery pokery here is to move outside a
;; comment in case a comment start is considered to
;; be a command start (for non fly-past behaviour)
(goto-char (match-beginning 0)))))
(proof-buffer-syntactic-context)
(goto-char (1+ (point))))
;; loop while in a string/comment before the next command start
)
(if (not (proof-buffer-syntactic-context)) ; not inside a comment/string
(if foundstart ; found a second command start
(progn
(goto-char foundstart) ; beginning of command start
(skip-chars-backward " \t\n") ; end of previous command
'cmd)
(if (eq (point) (point-max)) ; At the end of the buffer
(progn
(skip-chars-backward " \t\n") ; benefit of the doubt, let
'cmd)))) ; the PA moan if it's incomplete
;; Return nil in other cases, no complete command found
)))))
(defun proof-script-generic-parse-sexp ()
"Used for `proof-script-parse-function' if `proof-script-sexp-commands' is set."
;; Usual treatment of comments
(if (looking-at proof-script-comment-start-regexp)
;; Find end of comment
(if (proof-script-generic-parse-find-comment-end) 'comment)
(let* ((parse-sexp-ignore-comments t) ; gobble comments into commands
(end (scan-sexps (point) 1)))
(if end
(progn (goto-char end) 'cmd)))))
;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;
;;
;; Old parsing functions written for v3.2 (still used in Isar)
;;
;; NB: compared with previous (even older) code,
;; (1) this doesn't treat comments quite so well, but parsing
;; should be rather more efficient.
;; (2) comments are treated less like commands, and are only
;; coloured blue "in passing" when commands are sent.
;; However, undo still process comments step-by-step.
(defun proof-cmdstart-add-segment-for-cmd (comstart prev)
(let ((tmp (point))
(commentre (concat "[ \t\n]*" proof-script-comment-start-regexp))
(commentend (concat proof-script-comment-end-regexp "[ \t\n]*" ))) ;; FIXME: used?
;; Find end of previous command...
(goto-char comstart)
;; Special hack: terminal char is included in a command, if set.
(if (and proof-terminal-char
(looking-at
(regexp-quote (char-to-string proof-terminal-char))))
(goto-char (1+ (point)))
(skip-chars-backward " \t\n"))
(let* ((prev-no-blanks
(save-excursion
(goto-char prev)
(skip-chars-forward " \t\n")
(point)))
(comend (point))
(bufstr (buffer-substring prev-no-blanks comend))
(type
(save-excursion
;; The behaviour here is a bit odd: this is a
;; half-hearted attempt to strip comments as we send
;; text to the proof assistant, but it breaks when we
;; have certain bad input: (* foo *) blah (* bar *) cmd
;; We check for the case of a comment spanning the
;; *whole* substring, otherwise send the (defective)
;; text as if it were a proper command anyway. This
;; shouldn't cause problems: the proof assistant is
;; certainly capable of skipping comments itself, and
;; the situation should cause an error. (If it were
;; accepted it could upset the undo behaviour)
(goto-char prev-no-blanks)
;; Update: PG 3.4: try to deal with sequences of
;; comments as well, since previous behaviour breaks
;; Isar, in fact, since repeated comments get
;; categorized as commands, breaking sync.
(if (and
(looking-at commentre)
(re-search-forward proof-script-comment-end-regexp)
(progn
(while (looking-at commentre)
(re-search-forward proof-script-comment-end-regexp))
(>= (point) comend)))
'comment 'cmd)))
(string (if (eq type 'comment) "" bufstr)))
(setq prev (point))
(goto-char tmp)
;; Return cons of new prev and the new segment
;; NB: Command string excludes whitespace, span includes it.
(cons prev (list type string prev)))))
(defun proof-segment-up-to-cmdstart (pos &optional next-command-end)
"Parse the script buffer from end of locked to POS.
Return a list of (type, string, int) tuples.
Each tuple denotes the command and the position of its terminator,
type is one of 'comment, or 'cmd.
If optional NEXT-COMMAND-END is non-nil, we include the command
which continues past POS, if any. (NOT IMPLEMENTED IN THIS VERSION).
This version is used when `proof-script-command-start-regexp' is set."
(save-excursion
(let* (alist prev cmdfnd startpos comstart)
(goto-char (proof-queue-or-locked-end))
(setq prev (point))
(skip-chars-forward " \t\n")
(setq startpos (point))
(while
(and
(proof-re-search-forward proof-script-command-start-regexp
nil t) ; search for next command
(setq comstart (match-beginning 0)); save command start
(or (save-excursion
(goto-char comstart)
;; continue if inside (or at start of) comment/string
;; NB: this causes the "fly past comments behaviour",
;; it's a bit tricky to remove for Isar because some
;; Isar commands are manifestly exactly a command
;; followed by a comment (e.g. text {* foo *}).
(proof-looking-at-syntactic-context))
(progn ; or, if found command...
(setq cmdfnd
(> comstart startpos)); ignore first match
(<= prev pos))))
(if cmdfnd
(progn
(let ((prevseg (proof-cmdstart-add-segment-for-cmd comstart prev)))
(setq prev (car prevseg))
(setq alist (cons (cdr prevseg) alist))
(setq cmdfnd nil)))))
;; End of parse; see if we found some text at the end of the
;; buffer which could be a command. (NB: With a regexp for
;; start of commands only, we can't be sure it is a complete
;; command).
(if (and comstart ; previous command was found
(<= prev pos) ; last command within range
(goto-char (point-max))
(setq comstart (point)) ; pretend there's another cmd here
(not (proof-buffer-syntactic-context))) ; buffer ends well
(setq alist
(cons (cdr (proof-cmdstart-add-segment-for-cmd comstart prev))
alist)))
; (if (and cmdfnd next-command-end)
; (funcall add-segment-for-cmd))
;; Return resulting list
alist)))
;; FIXME: this needs fixing to include final comment in buffer
;; if there is one!!
(defun proof-segment-up-to-cmdend (pos &optional next-command-end)
"Parse the script buffer from end of locked to POS.
Return a list of (type, string, int) tuples.
Each tuple denotes the command and the position of its terminator,
type is one of 'comment, or 'cmd. 'unclosed-comment may be consed onto
the start if the segment finishes with an unclosed comment.
If optional NEXT-COMMAND-END is non-nil, we include the command
which continues past POS, if any.
This version is used when `proof-script-command-end-regexp' is set."
(save-excursion
(let*
((commentre (concat "[ \t\n]*" proof-script-comment-start-regexp))
(add-segment-for-cmd ; local function: advances "prev"
(lambda ()
(let ((cmdend (point)) start)
(goto-char prev)
;; String may start with comments, let's strip them off
(while
(and
(setq start (point))
(proof-re-search-forward commentre cmdend t)
(or (eq (match-beginning 0) start)
;; In case a comment inside a command was found, make
;; sure we're at the start of the cmd before exiting
(progn (goto-char start) nil)))
;; Look for the end of the comment
;; (FIXME: ignore nested comments here, we should
;; have a consistent policy!)
(unless
(if (progn
(goto-char (or (match-end 1) (match-beginning 0)))
(forward-comment))
(proof-re-search-forward
proof-script-comment-end-regexp cmdend t))
(error
"PG error: proof-segment-up-to-cmd-end didn't find comment end"))
(setq alist (cons (list 'comment "" (point)) alist)))
;; There should be something left: a command.
(skip-chars-forward " \t\n")
(setq alist (cons (list 'cmd
(buffer-substring
(point) cmdend)
cmdend) alist))
(setq prev cmdend)
(goto-char cmdend))))
alist prev cmdfnd startpos tmp)
(goto-char (proof-queue-or-locked-end))
(setq prev (point))
(skip-chars-forward " \t\n")
(setq startpos (point))
(while
(and
(proof-re-search-forward proof-script-command-end-regexp
nil t) ; search for next command
(or (proof-buffer-syntactic-context) ; continue if inside comment/string
(progn ; or, if found command...
(setq cmdfnd t)
(<= (point) pos))))
(if cmdfnd (progn
(funcall add-segment-for-cmd)
(setq cmdfnd nil))))
;; End of parse; if we found a command past POS maybe add it.
;; FIXME: also, if we found a *comment* maybe add it!
(if cmdfnd ; (and cmdfnd next-command-end)
(funcall add-segment-for-cmd))
;; Return resulting list
alist)))
(defun proof-semis-to-vanillas (semis &optional callback-fn)
"Convert a sequence of terminator positions to a set of vanilla extents.
Proof terminator positions SEMIS has the form returned by
the function proof-segment-up-to.
Set the callback to CALLBACK-FN or 'proof-done-advancing by default."
(let ((ct (proof-queue-or-locked-end)) span alist semi)
(while (not (null semis))
(setq semi (car semis)
span (make-span ct (nth 2 semi))
ct (nth 2 semi))
(if (eq (car (car semis)) 'cmd)
(progn
(set-span-property span 'type 'vanilla)
(set-span-property span 'cmd (nth 1 semi))
(setq alist (cons (list span (nth 1 semi)
(or callback-fn 'proof-done-advancing))
alist)))
(set-span-property span 'type 'comment)
(setq alist (cons (list span proof-no-command 'proof-done-advancing)
alist)))
(setq semis (cdr semis)))
(nreverse alist)))
;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;
;;
;; Assert-until-point and friends.
;;
;; These functions parse some region of the script buffer into
;; commands, and add the commands into the queue.
;;
;; First: two commands for moving forwards in proof scripts. Moving
;; forward for a "new" command may insert spaces or new lines. Moving
;; forward for the "next" command does not.
(defun proof-script-new-command-advance ()
"Move point to a nice position for a new command.
Assumes that point is at the end of a command."
(interactive)
; FIXME: pending improvement.2, needs a fix here.
; (if (eq (proof-locked-end) (point))
; ;; A hack to fix problem that the locked span
; ;; is [ ) so sometimes inserting at the end
; ;; tries to extend it, giving "read only" error.
; (if (> (point-max) (proof-locked-end))
; (progn
; (goto-char (1+ (proof-locked-end)))
; (backward-char))))
(if proof-one-command-per-line
;; One command per line: move to next new line, creating one if
;; at end of buffer or at the start of a blank line. (This has
;; the pleasing effect that blank regions of the buffer are
;; automatically extended when inserting new commands).
; unfortunately if we're not at the end of a line to start,
; it skips past stuff to the end of the line! don't want
; that.
; (cond
; ((eq (forward-line) 1)
; (newline))
; ((eolp)
; (newline)
; (forward-line -1)))
(newline)
;; Multiple commands per line: skip spaces at point, and insert
;; the 1/0 number of spaces that were skipped in front of point
;; (at least one). This has the pleasing effect that the spacing
;; policy of the current line is copied: e.g. <command>;
;; <command>; Tab columns don't work properly, however. Instead
;; of proof-one-command-per-line we could introduce a
;; "proof-command-separator" to improve this.
(let ((newspace (max (skip-chars-forward " \t") 1))
(p (point)))
(if proof-script-command-separator
(insert proof-script-command-separator)
(insert-char ?\040 newspace)
(goto-char p)))))
(defun proof-script-next-command-advance ()
"Move point to the beginning of the next command if it's nearby.
Assumes that point is at the end of a command."
(interactive)
;; skip whitespace on this line
(skip-chars-forward " \t")
(if (and proof-one-command-per-line (eolp))
;; go to the next line if we have one command per line
(forward-line)))
;; NB: the "interactive" variant is so that we get a simple docstring.
(defun proof-assert-until-point-interactive ()
"Process the region from the end of the locked-region until point.
Default action if inside a comment is just process as far as the start of
the comment."
(interactive)
(proof-assert-until-point))
;; Assert until point - We actually use this to implement the
;; assert-until-point, electric terminator keypress, and
;; goto-command-end. In different cases we want different things, but
;; usually the information (e.g. are we inside a comment) isn't
;; available until we've actually run proof-segment-up-to (point),
;; hence all the different options when we've done so.
;; FIXME da: this command doesn't behave as the doc string says when
;; inside comments. Also is unhelpful at the start of commands, and
;; in the locked region. I prefer the new version below.
;; FIXME: get rid of duplication between proof-assert-next-command and
;; proof-assert-until-point. Get rid of ignore process nonsense.
;; FIXME: get rid of unclosed-comment-fun nonsense. It's used
;; in the electric terminator function, but we should probably
;; use something else for that!
(defun proof-assert-until-point (&optional unclosed-comment-fun
ignore-proof-process-p)
"Process the region from the end of the locked-region until point.
Default action if inside a comment is just process as far as the start of
the comment.
If you want something different, put it inside
UNCLOSED-COMMENT-FUN. If IGNORE-PROOF-PROCESS-P is set, no commands
will be added to the queue and the buffer will not be activated for
scripting."
(unless ignore-proof-process-p
(proof-activate-scripting nil 'advancing))
(let ((semis))
(save-excursion
;; Give error if no non-whitespace between point and end of
;; locked region.
(if (proof-only-whitespace-to-locked-region-p)
(error "At the end of the locked region already, there's nothing to do to!"))
;; NB: (point) has now been moved backwards to first non-whitespace char.
(setq semis (proof-segment-up-to (point))))
(if (and unclosed-comment-fun (eq 'unclosed-comment (car semis)))
(funcall unclosed-comment-fun)
(if (eq 'unclosed-comment (car semis)) (setq semis (cdr semis)))
(if (and (not ignore-proof-process-p) (null semis))
;; This is another case that there's nothing to do: maybe
;; because inside a string or something.
(if (eq unclosed-comment-fun 'proof-electric-term-incomment-fn)
;; With electric terminator, we shouldn't give an error, but
;; should still insert and pretend it was as if a comment.
(proof-electric-term-incomment-fn)
(error "I can't find any complete commands to process!"))
(goto-char (nth 2 (car semis)))
(and (not ignore-proof-process-p)
(let ((vanillas (proof-semis-to-vanillas (nreverse semis))))
(proof-extend-queue (point) vanillas)))))))
;; da: This is my alternative version of the above.
;; It works from the locked region too.
;; I find it more convenient to assert up to the current command (command
;; point is inside), and move to the next command.
;; This means proofs can be easily replayed with point at the start
;; of lines. Above function gives stupid "nothing to do error." when
;; point is on the start of line or in the locked region.
;; FIXME: behaviour inside comments may be odd at the moment. (it
;; doesn't behave as docstring suggests, same prob as
;; proof-assert-until-point)
;; FIXME: polish the undo behaviour and quit behaviour of this
;; command (should inhibit quit somewhere or other).
(defun proof-assert-next-command
(&optional unclosed-comment-fun ignore-proof-process-p
dont-move-forward for-new-command)
"Process until the end of the next unprocessed command after point.
If inside a comment, just process until the start of the comment.
If you want something different, put it inside UNCLOSED-COMMENT-FUN.
If IGNORE-PROOF-PROCESS-P is set, no commands will be added to the queue.
Afterwards, move forward to near the next command afterwards, unless
DONT-MOVE-FORWARD is non-nil. If FOR-NEW-COMMAND is non-nil,
a space or newline will be inserted automatically."
(interactive)
(unless ignore-proof-process-p
(proof-activate-scripting nil 'advancing))
(or ignore-proof-process-p
(if (proof-locked-region-full-p)
(error "Locked region is full, no more commands to do!")))
(let ((semis))
(save-excursion
;; CHANGE from old proof-assert-until-point: don't bother check
;; for non-whitespace between locked region and point.
;; CHANGE: ask proof-segment-up-to to scan until command end
;; (which it used to do anyway, except in the case of a comment)
(setq semis (proof-segment-up-to (point) t)))
;; old code:
;;(if (not (re-search-backward "\\S-" (proof-unprocessed-begin) t))
;; (progn (goto-char pt)
;; (error "I don't know what I should be doing in this buffer!")))
;; (setq semis (proof-segment-up-to (point))))
(if (and unclosed-comment-fun (eq 'unclosed-comment (car-safe semis)))
(funcall unclosed-comment-fun)
(if (eq 'unclosed-comment (car-safe semis))
(setq semis (cdr semis)))
(if (and (not ignore-proof-process-p) (null semis))
(error "I can't see any complete commands to process!"))
(if (nth 2 (car semis))
(goto-char (nth 2 (car semis))))
(if (not ignore-proof-process-p)
(let ((vanillas (proof-semis-to-vanillas (nreverse semis))))
(proof-extend-queue (point) vanillas)))
;; This is done after the queuing to be polite: it means the
;; spacing policy enforced here is not put into the locked
;; region so the user can re-edit.
(if (not dont-move-forward)
(if for-new-command
(proof-script-new-command-advance)
(proof-script-next-command-advance))))))
(defun proof-goto-point ()
"Assert or retract to the command at current position.
Calls proof-assert-until-point or proof-retract-until-point as
appropriate."
(interactive)
(if (<= (proof-queue-or-locked-end) (point))
;; This asserts only until the next command before point and
;; does nothing if whitespace between point and locked region.
(proof-assert-until-point)
(proof-retract-until-point)))
;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;
;;
;; PBP call-backs
;;
(defun proof-insert-pbp-command (cmd)
"Insert CMD into the proof queue."
(proof-activate-scripting)
(let (span)
(proof-goto-end-of-locked)
(insert cmd)
(setq span (make-span (proof-locked-end) (point)))
(set-span-property span 'type 'pbp)
(set-span-property span 'cmd cmd)
(proof-start-queue (proof-unprocessed-begin) (point)
(list (list span cmd 'proof-done-advancing)))))
;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;
;;
;; Processing the script management queue -- PART 2: retracting
;;
;; Most of the hard work (computing the commands to do the retraction)
;; is implemented in the customisation module (lego.el or coq.el), so
;; code here is fairly straightforward.
;; FIXME: we need to adjust proof-nesting-depth appropriately here.
;; It would help to know the type of retraction which has just
;; occurred: a kill-proof may be assumed to set nesting depth
;; to zero; an undo sequence may alter it some other way.
;; FIXME FIXME: at the moment, the adjustment is made in
;; the wrong place!!
(defun proof-done-retracting (span)
"Callback for proof-retract-until-point.
We update display after proof process has reset its state.
See also the documentation for `proof-retract-until-point'.
Optionally delete the region corresponding to the proof sequence.
After an undo, we clear the proof completed flag. The rationale
is that undoing never leaves prover in a \"proof just completed\"
state, which is true for some proof assistants (but probably not
others)."
;; FIXME: need to fixup proof-nesting-depth
(setq proof-shell-proof-completed nil)
(if (span-live-p span)
(let ((start (span-start span))
(end (span-end span))
(kill (span-property span 'delete-me)))
;; da: check for empty region seems odd here?
;; [prevents regions from being detached in set-locked-end]
(unless (proof-locked-region-empty-p)
(proof-set-locked-end start)
(proof-set-queue-end start))
(delete-spans start end 'type)
(delete-spans start end 'idiom)
(delete-span span)
(if kill (kill-region start end))))
;; State of scripting may have changed now
(run-hooks 'proof-state-change-hook))
(defun proof-setup-retract-action (start end proof-command delete-region)
"Make span from START to END which corresponds to retraction.
Returns retraction action destined for proof shell queue, and make span.
Action holds PROOF-COMMAND and `proof-done-retracting' callback.
Span deletion property set to flag DELETE-REGION."
(let ((span (make-span start end)))
(set-span-property span 'delete-me delete-region)
(list (list span proof-command 'proof-done-retracting))))
(defun proof-last-goal-or-goalsave ()
(save-excursion
(let ((span (span-at-before (proof-locked-end) 'type)))
(while (and span
(not (eq (span-property span 'type) 'goalsave))
(or (eq (span-property span 'type) 'proof)
(eq (span-property span 'type) 'comment)
(eq (span-property span 'type) 'proverproc)
(not (funcall proof-goal-command-p
(span-property span 'cmd)))))
(setq span (prev-span span 'type)))
span)))
;;
;; NB: Should carefully explain/document this behaviour somewhere.
;; The undo is three-phase:
;; undo-cmd - ... - undo-cmd within proof
;; kill proof exit proof
;; forget-to-declaration forget target span
;;
;; It turns out that this behaviour is not quite right for Coq.
;; It might be simpler to just use a single undo/forget
;; command, which is called in all cases.
;;
(defun proof-retract-target (target delete-region)
"Retract the span TARGET and delete it if DELETE-REGION is non-nil.
Notice that this necessitates retracting any spans following TARGET,
up to the end of the locked region."
(let ((end (proof-unprocessed-begin))
(start (span-start target))
(span (proof-last-goal-or-goalsave))
actions)
;; NB: first section only entered if proof-kill-goal-command is
;; non-nill. Otherwise we expect proof-find-and-forget-fn to do
;; all relevent work for arbitrary retractions. FIXME: clean up
;; Examine the last span in the locked region.
;; If the last goal or save span is not a proof or
;; prover processed file, we examine to see how to remove it.
(if (and span proof-kill-goal-command
(not (or
(memq (span-property span 'type)
'(goalsave proverproc)))))
;; If the goal or goalsave span ends before the target span,
;; then we are retracting within the last unclosed proof,
;; and the retraction just amounts to a number of undo
;; steps.
;; FIXME: really, there shouldn't be more work to do: so
;; why call proof-find-and-forget-fn later?
(if (< (span-end span) (span-end target))
(progn
;; Skip comment/non-undoable spans at and immediately following target
(setq span target)
(while (and span
(memq (span-property span 'type) '(comment proverproc)))
(setq span (next-span span 'type)))
;; Calculate undos for the current open segment
;; of proof commands
(setq actions (proof-setup-retract-action
start end
(if (null span) proof-no-command
(funcall proof-count-undos-fn span))
delete-region)
end start))
;; Otherwise, start the retraction by killing off the
;; currently active goal.
;; FIXME: and couldn't we move the end upwards?
;; FIXME: hack proof-nesting-depth here. This is
;; in the wrong place: it should be done *after* the
;; retraction has succeeded.
(setq proof-nesting-depth (1- proof-nesting-depth))
(setq actions
(proof-setup-retract-action (span-start span) end
proof-kill-goal-command
delete-region)
end (span-start span))))
;; Check the start of the target span lies before the end
;; of the locked region (should always be true since we don't
;; make spans outside the locked region at the moment)...
;; But end may have moved backwards above: this just checks whether
;; there is more retraction to be done.
(if (> end start)
(setq actions
;; Append a retract action to clear the entire
;; start-end region. Rely on proof-find-and-forget-fn
;; to calculate a command which "forgets" back to
;; the first definition, declaration, or whatever
;; that comes after the target span.
;; FIXME: originally this assumed a linear context,
;; and that forgetting the first thing forgets all
;; subsequent ones. it might be more general to
;; allow *several* commands, and even queue these
;; separately for each of the spans following target
;; which are concerned.
(nconc actions (proof-setup-retract-action
start end
(funcall proof-find-and-forget-fn target)
delete-region))))
(proof-start-queue (min start end) (proof-locked-end) actions)))
;; FIXME da: I would rather that this function moved point to
;; the start of the region retracted?
;; FIXME da: Maybe retraction to the start of
;; a file should remove it from the list of included files?
;; NB: the "interactive" variant is so that we get a simple docstring.
(defun proof-retract-until-point-interactive (&optional delete-region)
"Tell the proof process to retract until point.
If invoked outside a locked region, undo the last successfully processed
command. If called with a prefix argument (DELETE-REGION non-nil), also
delete the retracted region from the proof-script."
(interactive "P")
(proof-retract-until-point delete-region))
(defun proof-retract-until-point (&optional delete-region)
"Set up the proof process for retracting until point.
In particular, set a flag for the filter process to call
`proof-done-retracting' after the proof process has successfully
reset its state.
If DELETE-REGION is non-nil, delete the region in the proof script
corresponding to the proof command sequence.
If invoked outside a locked region, undo the last successfully processed
command."
(if (proof-locked-region-empty-p)
(error "No locked region")
;; Make sure we're ready: either not busy, or already advancing/retracting.
(proof-activate-scripting)
(let ((span (span-at (point) 'type)))
;; If no span at point, retracts the last span in the buffer.
(unless span
(proof-goto-end-of-locked) ; NB: this moves point
(backward-char)
(setq span (span-at (point) 'type)))
(proof-retract-target span delete-region))))
;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;
;;
;; Proof General scripting mode definition, part 1.
;;
(eval-and-compile ; to define vars
;;; NB: autoload tag below doesn't work for d-d-m, autoload is in proof.el
;;;###autoload
(define-derived-mode proof-mode fundamental-mode
proof-general-name
"Proof General major mode class for proof scripts.
\\{proof-mode-map}"
(setq proof-buffer-type 'script)
;; font-lock-keywords isn't automatically buffer-local in Emacs 21.2
(make-local-variable 'font-lock-keywords)
;; During write-file it can happen that we re-set the mode for
;; the currently active scripting buffer. The user might also
;; do this for some reason. We could maybe let
;; this pass through, but it seems safest to treat it as
;; a kill buffer operation (retract and clear spans).
;; (NB: other situations seem to cause double successive calls to
;; proof-mode).
(if (eq (current-buffer) proof-script-buffer)
(proof-script-kill-buffer-fn))
;; We set hook functions here rather than in proof-config-done so
;; that they can be adjusted by prover specific code if need be.
(proof-script-set-buffer-hooks)
(make-local-hook 'after-set-visited-file-name-hooks)
(add-hook 'after-set-visited-file-name-hooks 'proof-script-set-visited-file-name))
(make-local-hook 'proof-activate-scripting-hook)
(add-hook 'proof-activate-scripting-hook 'proof-cd-sync nil t))
;; NB: proof-mode-map declared by define-derived-mode above
(proof-menu-define-keys proof-mode-map)
(defun proof-script-set-visited-file-name ()
"Called when visited file name is changed.
This is a hook function for `after-set-visited-file-name-hooks'.
For some provers, the file from which script commands are being
processed may be important, and if it is changed with C-x C-w, for
example, we might have to retract the contents or inform the proof
assistant of the new name. This should be done by adding
additional functions to `after-set-visited-file-name-hooks'.
At the least, we need to set the buffer local hooks again
with `proof-script-set-buffer-hooks' which is what this function does,
as well as setting `proof-script-buffer-file-name' (which see).
This hook also gives a warning in case this is the active scripting buffer."
(setq proof-script-buffer-file-name buffer-file-name)
(if (eq (current-buffer) proof-script-buffer)
(proof-warning
"Active scripting buffer changed name; synchronization risked if prover tracks filenames!"))
(proof-script-set-buffer-hooks))
(defun proof-script-set-buffer-hooks ()
"Set the hooks for a proof script buffer.
The hooks set here are cleared by `write-file', so we use this function
to restore them using `after-set-visited-file-name-hooks'."
(make-local-hook 'kill-buffer-hook)
(add-hook 'kill-buffer-hook 'proof-script-kill-buffer-fn t t)
;; Reverting buffer is same as killing it as far as PG is concerned
(make-local-hook 'before-revert-hook)
(add-hook 'before-revert-hook 'proof-script-kill-buffer-fn t t))
(defun proof-script-kill-buffer-fn ()
"Value of `kill-buffer-hook' for proof script buffers.
Clean up before a script buffer is killed.
If killing the active scripting buffer, run proof-deactivate-scripting-auto.
Otherwise just do proof-restart-buffers to delete some spans from memory."
;; Deactivate scripting in the current buffer if need be, forcing
;; automatic retraction if the buffer is not fully processed.
(if (eq (current-buffer) proof-script-buffer)
(proof-deactivate-scripting-auto))
(proof-restart-buffers (list (current-buffer)))
;; Hide away goals, response, and tracing. This is a hack because
;; otherwise we can lead the user to frustration with the
;; dedicated windows nonsense.
(proof-map-buffers
(list proof-goals-buffer proof-response-buffer proof-trace-buffer)
(bury-buffer (current-buffer))))
;; Notes about how to deal with killing/reverting/renaming buffers:
;; (As of XEmacs 21.1.9, see files.el)
;;
;; Killing: easy, set kill-buffer-hook
;; Reverting: ditto, set before-revert-hook to do same as kill.
;; Renaming (write-file): much tricker. Code for write-file does
;; several odd things. It kills off local hook functions, calls
;; `after-set-visited-file-name-hooks' right at the end to give the
;; chance to restore them, but then tends to automatically (re-)set
;; the mode anyway.
;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;
;;
;; Proof General scripting mode definition - part 2
;;
;; The functions proof-config-done[-related] are called after the
;; derived mode has made its settings.
;; The callback *-config-done mechanism is an irritating hack - there
;; should be some elegant mechanism for computing constants after the
;; child has configured. Should petition the author of "derived-mode"
;; about this!
(defun proof-config-done-related ()
"Finish setup of Proof General scripting and related modes.
This is a subroutine of `proof-config-done'.
This is intended for proof assistant buffers which are similar to
script buffers, but for which scripting is not enabled. In
particular, we: lock the buffer if it appears on
`proof-included-files-list'; configure font-lock support from
`font-lock-keywords'; maybe turn on X-Symbol minor mode.
This is used for Isabelle theory files, which share some scripting
mode features, but are only ever processed atomically by the proof
assistant."
(setq proof-script-buffer-file-name buffer-file-name)
;; Has buffer already been processed?
;; NB: call to file-truename is needed for GNU Emacs which
;; chooses to make buffer-file-truename abbreviate-file-name
;; form of file-truename.
(and buffer-file-truename
(member (file-truename buffer-file-truename)
proof-included-files-list)
(proof-complete-buffer-atomic (current-buffer)))
;; calculate some strings and regexps for searching
(setq proof-terminal-string
(if proof-terminal-char
(char-to-string proof-terminal-char)
""))
(make-local-variable 'comment-start)
(setq comment-start (concat proof-script-comment-start " "))
(make-local-variable 'comment-end)
(setq comment-end (concat " " proof-script-comment-end))
(unless proof-script-comment-start-regexp
(setq proof-script-comment-start-regexp (regexp-quote proof-script-comment-start)))
(unless proof-script-comment-end-regexp
(setq proof-script-comment-end-regexp (regexp-quote proof-script-comment-end)))
(make-local-variable 'comment-start-skip)
(setq comment-start-skip
(concat proof-script-comment-start-regexp "+\\s_?"))
;;
;; Fontlock support.
;;
;; Assume font-lock case folding follows proof-case-fold-search
(proof-font-lock-configure-defaults 'autofontify proof-case-fold-search))
;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;
;;
;; Generic defaults for hooks, based on regexps.
;;
;; The next step is to use proof-stringfn-match scheme more widely, to
;; allow settings which are string or fn, so we don't need both regexp
;; and function hooks, and so that the other hooks can be functions too.
(defun proof-generic-goal-command-p (str)
"Is STR a goal? Decide by matching with `proof-goal-command-regexp'."
(proof-string-match-safe proof-goal-command-regexp str))
(defun proof-generic-state-preserving-p (cmd)
"Is CMD state preserving? Match on `proof-non-undoables-regexp'."
;; FIXME: logic here is not quite finished: proof-non-undoables are
;; certainly not state preserving, but so are a bunch more things,
;; i.e. ordinary proof commands which may appear in proof scripts.
;; Might be better to add positive and negative regexps for
;; state-preserving tests (only one of which needs to be set).
(not (proof-string-match-safe proof-non-undoables-regexp cmd)))
(defun proof-generic-count-undos (span)
"Count number of undos in a span, return command needed to undo that far.
Command is set using `proof-undo-n-times-cmd'.
A default value for `proof-count-undos-fn'.
For this function to work properly, you must configure
`proof-undo-n-times-cmd' and `proof-ignore-for-undo-count'."
(let
((case-fold-search proof-case-fold-search)
(ct 0) str i)
(while span
(setq str (span-property span 'cmd))
(cond ((eq (span-property span 'type) 'vanilla)
(unless (proof-stringfn-match proof-ignore-for-undo-count str)
(incf ct)))
((eq (span-property span 'type) 'pbp)
(setq i 0)
(while (< i (length str))
(if (= (aref str i) proof-terminal-char) (incf ct))
(incf i))))
(setq span (next-span span 'type)))
(if (= ct 0)
proof-no-command
(cond ((stringp proof-undo-n-times-cmd)
(format proof-undo-n-times-cmd ct))
((functionp proof-undo-n-times-cmd)
(funcall proof-undo-n-times-cmd ct))))))
(defun proof-generic-find-and-forget (span)
"Calculate a forget/undo command to forget back to SPAN.
This is a long-range forget: we know that there is no
open goal at the moment, so forgetting involves unbinding
declarations, etc, rather than undoing proof steps.
This generic implementation assumes it is enough to find the
nearest following span with a `name' property, and retract
that using `proof-forget-id-command' with the given name.
If this behaviour is not correct, you must customize the function
with something different."
;; Modelled on Isar's find-and-forget function, but less
;; general at the moment: will only issue one und command.
;; FIXME: would be much cleaner to wrap up the undo behaviour
;; also within proofs in this function.
(let (str ans typ name answers)
(while span
(setq ans nil)
(setq str (span-property span 'cmd))
(setq typ (span-property span 'type))
(cond
;; comment, diagnostic, prover processed, nested proof command: skip
((or (eq typ 'comment)
(eq typ 'proverproc)
(eq typ 'proof)
(and proof-ignore-for-undo-count cmd
(proof-string-match proof-ignore-for-undo-count cmd))))
;; some named element: use generic forget-id function; finish.
((setq name (span-property span 'name))
(setq ans (format proof-forget-id-command name))
(setq span nil)))
(if ans (setq answers (cons ans answers)))
(if span (setq span (next-span span 'type))))
(if (null answers) proof-no-command (apply 'concat answers))))
;;
;; End of new generic functions
;;
;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;
;;
;; Sanity checks on important settings
;;
(defconst proof-script-important-settings
'(proof-script-comment-start ;
proof-script-comment-end
proof-save-command-regexp
; proof-goal-command-regexp ; not needed if proof-goal-command-p is set
; proof-goal-with-hole-regexp ; non-essential?
; proof-save-with-hole-regexp ; non-essential?
; proof-showproof-command ; non-essential
; proof-goal-command ; non-essential
; proof-save-command ; do
; proof-kill-goal-command ; do
))
(defun proof-config-done ()
"Finish setup of Proof General scripting mode.
Call this function in the derived mode for the proof assistant to
finish setup which depends on specific proof assistant configuration."
;; Common configuration for shared script/other related buffers.
(proof-config-done-related)
;; Preamble: make this mode class "pg-sticky" so that renaming file
;; to something different doesn't change the mode, no matter what
;; the filename. This is a hack so that write-file will work:
;; otherwise Emacs insists (XEmacs 21.1.9 onwards) on re-setting the
;; mode, which leads to problems with synchronization and losing
;; extents. (Attempt to catch this in proof-mode by looking for
;; active scripting buffer fails; perhaps because of kill buffer
;; function) [NB: could do this at top level at load time]
(put major-mode 'mode-class 'pg-sticky)
;; Make X-symbol ignore that we've asked for fixed mode
(put major-mode 'x-symbol-mode-disable 'ignore)
;; Now, define some values if they aren't defined already.
(unless proof-mode-for-script
(setq proof-mode-for-script major-mode))
(if (and proof-non-undoables-regexp
(not proof-ignore-for-undo-count))
(setq proof-ignore-for-undo-count
proof-non-undoables-regexp))
;; Give warnings if some crucial settings haven't been made
(dolist (sym proof-script-important-settings)
(proof-warn-if-unset "proof-config-done" sym))
;; Additional key definitions which depend on configuration for
;; specific proof assistant.
;; FIXME da: generalize here. Might have electric terminator for
;; other parsing mechanisms too, using new proof-script-parse-function
;; Could use a default terminal char
(if proof-terminal-char
(progn
(define-key proof-mode-map
(vconcat [(control c)] (vector proof-terminal-char))
'proof-electric-terminator-toggle)
(define-key proof-mode-map (vector proof-terminal-char)
'proof-electric-terminator)))
;; It's ugly, but every script buffer has a local copy changed in
;; sync by the fn proof-electric-terminator-enable
(setq proof-electric-terminator proof-electric-terminator-enable)
(make-local-variable 'indent-line-function)
(setq indent-line-function 'proof-indent-line)
;; Toolbar and scripting menu
;; NB: autloads proof-toolbar, which defines proof-toolbar-scripting-menu.
(proof-toolbar-setup)
;; Menus: the Proof-General and the specific menu
(proof-menu-define-main)
(proof-menu-define-specific)
(easy-menu-add proof-mode-menu proof-mode-map)
(easy-menu-add proof-assistant-menu proof-mode-map)
;; Define parsing functions
(proof-setup-parsing-mechanism)
;; Setup imenu and/or func-menu.
(proof-setup-imenu)
(proof-setup-func-menu)
;; Offer to save script mode buffers which have no files,
;; in case Emacs is exited accidently.
(or (buffer-file-name)
(setq buffer-offer-save t))
;; Localise the invisibility glyph (XEmacs only):
(let ((img (proof-get-image "hiddenproof" t "<proof>")))
(cond
((and img proof-running-on-XEmacs)
(set-glyph-image invisible-text-glyph img (current-buffer)))))
(if (proof-ass x-symbol-enable)
(proof-x-symbol-enable))
;; Finally, make sure the user has been welcomed!
;; [NB: this doesn't work well, can get zapped by loading messages]
(proof-splash-message))
;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;
;;
;; Subroutines of proof-config-done
;;
(defun proof-setup-parsing-mechanism ()
"Choose parsing mechanism according to different kinds of script syntax.
Choice of function depends on configuration setting."
(unless (fboundp 'proof-segment-up-to)
(if proof-script-use-old-parser
;; Configuration for using old parsing mechanism.
(cond
(proof-script-parse-function ;; still allowed to override in 3.2
(defalias 'proof-segment-up-to 'proof-segment-up-to-parser))
;; 3.2 mechanism here
(proof-script-command-start-regexp
(defalias 'proof-segment-up-to 'proof-segment-up-to-cmdstart))
(t
(defalias 'proof-segment-up-to 'proof-segment-up-to-cmdend)
(unless proof-script-command-end-regexp
(proof-warn-if-unset "proof-config-done" 'proof-terminal-char)
(setq proof-script-command-end-regexp
(if proof-terminal-char
(regexp-quote (char-to-string proof-terminal-char))
"$")))))
;; Configuration for using new parsing (3.3 and later; default in 3.5)
(progn
(defalias 'proof-segment-up-to 'proof-segment-up-to-parser)
(cond
(proof-script-parse-function
;; already set, nothing to do
)
(proof-script-sexp-commands
(setq proof-script-parse-function 'proof-script-generic-parse-sexp))
(proof-script-command-start-regexp
(setq proof-script-parse-function 'proof-script-generic-parse-cmdstart))
((or proof-script-command-end-regexp proof-terminal-char)
(setq proof-script-parse-function 'proof-script-generic-parse-cmdend)
(unless proof-script-command-end-regexp
(proof-warn-if-unset "probof-config-done" 'proof-terminal-char)
(setq proof-script-command-end-regexp
(if proof-terminal-char
(regexp-quote (char-to-string proof-terminal-char))
"$"))))
(t
(error "Configuration error: must set `proof-terminal-char' or one of its friends"))))
))) ; default if nothing set is EOL.
(defun proof-setup-imenu ()
"Setup a default for imenu."
(unless (and (boundp 'imenu-generic-expression)
imenu-generic-expression)
(set (make-local-variable 'imenu-generic-expression)
(delq nil
(list
(if proof-goal-with-hole-regexp
(list nil proof-goal-with-hole-regexp
proof-goal-with-hole-result))
(if proof-save-with-hole-regexp
(list "Saves" proof-save-with-hole-regexp
proof-save-with-hole-result)))))))
(defun proof-setup-func-menu ()
"Configure func-menu for a proof script buffer."
;; NB: Ideal place for this and similar stuff would be in something
;; evaluated at top level after defining the derived mode: normally
;; we wouldn't repeat this each time the mode function is run, so we
;; wouldn't need "pushnew").
(if (proof-try-require 'func-menu)
(progn
;; FIXME: we'd like to let the library load later, but
;; it's a bit tricky: this stuff doesn't seem to work
;; in an eval-after-load body (local vars?).
(unless proof-script-next-entity-regexps ; unless already set
;; Try to calculate a useful default value.
;; FIXME: this is rather complicated! The use of the regexp
;; variables needs sorting out.
(customize-set-variable
'proof-script-next-entity-regexps
(let ((goal-discrim
;; Goal discriminator searches forward for matching
;; save if the regexp is set.
(if proof-goal-with-hole-regexp
(if proof-save-command-regexp
(list
proof-goal-with-hole-regexp 2
'forward proof-save-command-regexp)
(list proof-goal-with-hole-regexp 2))))
;; Save discriminator searches backward for matching
;; goal if the regexp is set.
(save-discrim
(if proof-save-with-hole-regexp
(if proof-goal-command-regexp
(list
proof-save-with-hole-regexp 2
'backward proof-goal-command-regexp)
(list proof-save-with-hole-regexp 2)))))
(cond
((and proof-goal-with-hole-regexp proof-save-with-hole-regexp)
(list
(proof-regexp-alt
proof-goal-with-hole-regexp
proof-save-with-hole-regexp) goal-discrim save-discrim))
(proof-goal-with-hole-regexp
(list proof-goal-with-hole-regexp goal-discrim))
(proof-save-with-hole-regexp
(list proof-save-with-hole-regexp save-discrim))))))
(if proof-script-next-entity-regexps
;; Enable func-menu for this mode if regexps set now
(progn
(pushnew
(cons major-mode 'proof-script-next-entity-regexps)
fume-function-name-regexp-alist)
(pushnew
(cons major-mode proof-script-find-next-entity-fn)
fume-find-function-name-method-alist))))))
(provide 'proof-script)
;;; proof-script.el ends here
|