-
Notifications
You must be signed in to change notification settings - Fork 439
/
Collectiondb.cpp
4250 lines (3621 loc) · 125 KB
/
Collectiondb.cpp
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
#include "gb-include.h"
#include "Collectiondb.h"
//#include "CollectionRec.h"
#include "Xml.h"
#include "Url.h"
#include "Loop.h"
#include "Spider.h" // for calling SpiderLoop::collectionsUpdated()
#include "Posdb.h"
//#include "Indexdb.h"
#include "Datedb.h"
#include "Titledb.h"
//#include "Revdb.h"
//#include "Sections.h"
#include "Placedb.h"
#include "Tagdb.h"
#include "Catdb.h"
#include "Tfndb.h"
#include "Spider.h"
//#include "Checksumdb.h"
#include "Clusterdb.h"
#include "Spider.h"
#include "Repair.h"
#include "Users.h"
#include "Parms.h"
void testRegex ( ) ;
HashTableX g_collTable;
// a global class extern'd in .h file
Collectiondb g_collectiondb;
Collectiondb::Collectiondb ( ) {
m_wrapped = 0;
m_numRecs = 0;
m_numRecsUsed = 0;
m_numCollsSwappedOut = 0;
m_initializing = false;
//m_lastUpdateTime = 0LL;
m_needsSave = false;
// sanity
if ( RDB_END2 >= RDB_END ) return;
log("db: increase RDB_END2 to at least %"INT32" in "
"Collectiondb.h",(int32_t)RDB_END);
char *xx=NULL;*xx=0;
}
// reset rdb
void Collectiondb::reset() {
log(LOG_INFO,"db: resetting collectiondb.");
for ( int32_t i = 0 ; i < m_numRecs ; i++ ) {
if ( ! m_recs[i] ) continue;
mdelete ( m_recs[i], sizeof(CollectionRec), "CollectionRec" );
delete ( m_recs[i] );
m_recs[i] = NULL;
}
m_numRecs = 0;
m_numRecsUsed = 0;
g_collTable.reset();
}
/*
bool Collectiondb::init ( bool isDump ) {
reset();
if ( g_isYippy ) return true;
// reset # of recs
//m_numRecs = 0;
//m_numRecsUsed = 0;
// . now load ALL recs
// . returns false and sets g_errno on error
if ( ! load ( isDump ) ) return false;
// update time
updateTime();
// so we don't save again
m_needsSave = false;
// sanity
if ( RDB_END2 < RDB_END ) {
log("db: increase RDB_END2 to at least %"INT32" in "
"Collectiondb.h",(int32_t)RDB_END);
char *xx=NULL;*xx=0;
}
// if it set g_errno, return false
//if ( g_errno ) return log("admin: Had init error: %s.",
// mstrerror(g_errno));
g_errno = 0;
// otherwise, true, even if reloadList() blocked
return true;
}
*/
extern bool g_inAutoSave;
// . save to disk
// . returns false if blocked, true otherwise
bool Collectiondb::save ( ) {
if ( g_conf.m_readOnlyMode ) return true;
if ( g_inAutoSave && m_numRecsUsed > 20 && g_hostdb.m_hostId != 0 )
return true;
// which collection rec needs a save
for ( int32_t i = 0 ; i < m_numRecs ; i++ ) {
if ( ! m_recs[i] ) continue;
// temp debug message
//logf(LOG_DEBUG,"admin: SAVING collection #%"INT32" ANYWAY",i);
if ( ! m_recs[i]->m_needsSave ) continue;
// if we core in malloc we won't be able to save the
// coll.conf files
if ( m_recs[i]->m_isCustomCrawl &&
g_inMemFunction &&
g_hostdb.m_hostId != 0 )
continue;
//log(LOG_INFO,"admin: Saving collection #%"INT32".",i);
m_recs[i]->save ( );
}
// oh well
return true;
}
///////////
//
// fill up our m_recs[] array based on the coll.*.*/coll.conf files
//
///////////
bool Collectiondb::loadAllCollRecs ( ) {
m_initializing = true;
char dname[1024];
// MDW: sprintf ( dname , "%s/collections/" , g_hostdb.m_dir );
sprintf ( dname , "%s" , g_hostdb.m_dir );
Dir d;
d.set ( dname );
if ( ! d.open ()) return log("admin: Could not load collection config "
"files.");
int32_t count = 0;
char *f;
while ( ( f = d.getNextFilename ( "*" ) ) ) {
// skip if first char not "coll."
if ( strncmp ( f , "coll." , 5 ) != 0 ) continue;
// must end on a digit (i.e. coll.main.0)
if ( ! is_digit (f[gbstrlen(f)-1]) ) continue;
// count them
count++;
}
// reset directory for another scan
d.set ( dname );
if ( ! d.open ()) return log("admin: Could not load collection config "
"files.");
// note it
//log(LOG_INFO,"db: loading collection config files.");
// . scan through all subdirs in the collections dir
// . they should be like, "coll.main/" and "coll.mycollection/"
while ( ( f = d.getNextFilename ( "*" ) ) ) {
// skip if first char not "coll."
if ( strncmp ( f , "coll." , 5 ) != 0 ) continue;
// must end on a digit (i.e. coll.main.0)
if ( ! is_digit (f[gbstrlen(f)-1]) ) continue;
// point to collection
char *coll = f + 5;
// NULL terminate at .
char *pp = strchr ( coll , '.' );
if ( ! pp ) continue;
*pp = '\0';
// get collnum
collnum_t collnum = atol ( pp + 1 );
// add it
if ( ! addExistingColl ( coll , collnum ) )
return false;
// swap it out if we got 100+ collections
// if ( count < 100 ) continue;
// CollectionRec *cr = getRec ( collnum );
// if ( cr ) cr->swapOut();
}
// if no existing recs added... add coll.main.0 always at startup
if ( m_numRecs == 0 ) {
log("admin: adding main collection.");
addNewColl ( "main",
0 , // customCrawl ,
NULL,
0 ,
true , // bool saveIt ,
// Parms.cpp reserves this so it can be sure
// to add the same collnum to every shard
0 );
}
m_initializing = false;
// note it
//log(LOG_INFO,"db: Loaded data for %"INT32" collections. Ranging from "
// "collection #0 to #%"INT32".",m_numRecsUsed,m_numRecs-1);
// update the time
//updateTime();
// don't clean the tree if just dumpin
//if ( isDump ) return true;
return true;
}
// after we've initialized all rdbs in main.cpp call this to clean out
// our rdb trees
bool Collectiondb::cleanTrees ( ) {
// remove any nodes with illegal collnums
Rdb *r;
//r = g_indexdb.getRdb();
//r->m_tree.cleanTree ((char **)r->m_bases);
r = g_posdb.getRdb();
//r->m_tree.cleanTree ();//(char **)r->m_bases);
r->m_buckets.cleanBuckets();
//r = g_datedb.getRdb();
//r->m_tree.cleanTree ((char **)r->m_bases);
r = g_titledb.getRdb();
r->m_tree.cleanTree ();//(char **)r->m_bases);
//r = g_revdb.getRdb();
//r->m_tree.cleanTree ((char **)r->m_bases);
//r = g_sectiondb.getRdb();
//r->m_tree.cleanTree ((char **)r->m_bases);
//r = g_checksumdb.getRdb();
//r->m_tree.cleanTree ((char **)r->m_bases);
//r = g_tfndb.getRdb();
//r->m_tree.cleanTree ((char **)r->m_bases);
r = g_spiderdb.getRdb();
r->m_tree.cleanTree ();//(char **)r->m_bases);
r = g_doledb.getRdb();
r->m_tree.cleanTree ();//(char **)r->m_bases);
// success
return true;
}
/*
void Collectiondb::updateTime() {
// get time now in milliseconds
int64_t newTime = gettimeofdayInMilliseconds();
// change it
if ( m_lastUpdateTime == newTime ) newTime++;
// update it
m_lastUpdateTime = newTime;
// we need a save
m_needsSave = true;
}
*/
#include "Statsdb.h"
#include "Cachedb.h"
#include "Syncdb.h"
// same as addOldColl()
bool Collectiondb::addExistingColl ( char *coll, collnum_t collnum ) {
int32_t i = collnum;
// ensure does not already exist in memory
collnum_t oldCollnum = getCollnum(coll);
if ( oldCollnum >= 0 ) {
g_errno = EEXIST;
log("admin: Trying to create collection \"%s\" but "
"already exists in memory. Do an ls on "
"the working dir to see if there are two "
"collection dirs with the same coll name",coll);
char *xx=NULL;*xx=0;
}
// also try by #, i've seen this happen too
CollectionRec *ocr = getRec ( i );
if ( ocr ) {
g_errno = EEXIST;
log("admin: Collection id %i is in use already by "
"%s, so we can not add %s. moving %s to trash."
,(int)i,ocr->m_coll,coll,coll);
SafeBuf cmd;
int64_t now = gettimeofdayInMilliseconds();
cmd.safePrintf ( "mv coll.%s.%i trash/coll.%s.%i.%"UINT64
, coll
,(int)i
, coll
,(int)i
, now );
//log("admin: %s",cmd.getBufStart());
gbsystem ( cmd.getBufStart() );
return true;
}
// create the record in memory
CollectionRec *cr = new (CollectionRec);
if ( ! cr )
return log("admin: Failed to allocated %"INT32" bytes for new "
"collection record for \"%s\".",
(int32_t)sizeof(CollectionRec),coll);
mnew ( cr , sizeof(CollectionRec) , "CollectionRec" );
// set collnum right for g_parms.setToDefault() call just in case
// because before it was calling CollectionRec::reset() which
// was resetting the RdbBases for the m_collnum which was garbage
// and ended up resetting random collections' rdb. but now
// CollectionRec::CollectionRec() sets m_collnum to -1 so we should
// not need this!
//cr->m_collnum = oldCollnum;
// get the default.conf from working dir if there
g_parms.setToDefault( (char *)cr , OBJ_COLL , cr );
strcpy ( cr->m_coll , coll );
cr->m_collLen = gbstrlen ( coll );
cr->m_collnum = i;
// point to this, so Rdb and RdbBase can reference it
coll = cr->m_coll;
//log("admin: loaded old coll \"%s\"",coll);
// load coll.conf file
if ( ! cr->load ( coll , i ) ) {
mdelete ( cr, sizeof(CollectionRec), "CollectionRec" );
log("admin: Failed to load coll.%s.%"INT32"/coll.conf",coll,i);
delete ( cr );
if ( m_recs ) m_recs[i] = NULL;
return false;
}
if ( ! registerCollRec ( cr , false ) ) return false;
// always index spider status docs now for custom crawls
if ( cr->m_isCustomCrawl )
cr->m_indexSpiderReplies = true;
// and don't do link voting, will help speed up
if ( cr->m_isCustomCrawl ) {
cr->m_getLinkInfo = false;
cr->m_computeSiteNumInlinks = false;
// limit each shard to 5 spiders per collection to prevent
// ppl from spidering the web and hogging up resources
cr->m_maxNumSpiders = 5;
// diffbot download docs up to 50MB so we don't truncate
// things like sitemap.xml. but keep regular html pages
// 1MB
cr->m_maxTextDocLen = 1024*1024;
// xml, pdf, etc can be this. 50MB
cr->m_maxOtherDocLen = 50000000;
}
// we need to compile the regular expressions or update the url
// filters with new logic that maps crawlbot parms to url filters
return cr->rebuildUrlFilters ( );
}
// . add a new rec
// . returns false and sets g_errno on error
// . was addRec()
// . "isDump" is true if we don't need to initialize all the rdbs etc
// because we are doing a './gb dump ...' cmd to dump out data from
// one Rdb which we will custom initialize in main.cpp where the dump
// code is. like for instance, posdb.
// . "customCrawl" is 0 for a regular collection, 1 for a simple crawl
// 2 for a bulk job. diffbot terminology.
bool Collectiondb::addNewColl ( char *coll ,
char customCrawl ,
char *cpc ,
int32_t cpclen ,
bool saveIt ,
// Parms.cpp reserves this so it can be sure
// to add the same collnum to every shard
collnum_t newCollnum ) {
//do not send add/del coll request until we are in sync with shard!!
// just return ETRYAGAIN for the parmlist...
// ensure coll name is legit
char *p = coll;
for ( ; *p ; p++ ) {
if ( is_alnum_a(*p) ) continue;
if ( *p == '-' ) continue;
if ( *p == '_' ) continue; // underscore now allowed
break;
}
if ( *p ) {
g_errno = EBADENGINEER;
log("admin: \"%s\" is a malformed collection name because it "
"contains the '%c' character.",coll,*p);
return false;
}
// . scan for holes
// . i is also known as the collection id
//int32_t i = (int32_t)newCollnum;
// no longer fill empty slots because if they do a reset then
// a new rec right away it will be filled with msg4 recs not
// destined for it. Later we will have to recycle some how!!
//else for ( i = 0 ; i < m_numRecs ; i++ ) if ( ! m_recs[i] ) break;
// right now we #define collnum_t int16_t. so do not breach that!
//if ( m_numRecs < 0x7fff ) {
// // set it
// i = m_numRecs;
// // claim it
// // we don't do it here, because we check i below and
// // increment m_numRecs below.
// //m_numRecs++;
//}
// TODO: scan for holes here...
//else {
if ( newCollnum < 0 ) { char *xx=NULL;*xx=0; }
// ceiling?
//int64_t maxColls = 1LL<<(sizeof(collnum_t)*8);
//if ( i >= maxColls ) {
// g_errno = ENOBUFS;
// return log("admin: Limit of %"INT64" collection reached. "
// "Collection not created.",maxColls);
//}
// if empty... bail, no longer accepted, use "main"
if ( ! coll || !coll[0] ) {
g_errno = EBADENGINEER;
return log("admin: Trying to create a new collection "
"but no collection name provided. Use the \"c\" "
"cgi parameter to specify it.");
}
// or if too big
if ( gbstrlen(coll) > MAX_COLL_LEN ) {
g_errno = ENOBUFS;
return log("admin: Trying to create a new collection "
"whose name \"%s\" of %i chars is longer than the "
"max of %"INT32" chars.",coll,gbstrlen(coll),
(int32_t)MAX_COLL_LEN);
}
// ensure does not already exist in memory
if ( getCollnum ( coll ) >= 0 ) {
g_errno = EEXIST;
log("admin: Trying to create collection \"%s\" but "
"already exists in memory.",coll);
// just let it pass...
g_errno = 0 ;
return true;
}
// MDW: ensure not created on disk since time of last load
char dname[512];
sprintf(dname, "%scoll.%s.%"INT32"/",g_hostdb.m_dir,coll,(int32_t)newCollnum);
DIR *dir = opendir ( dname );
if ( dir ) closedir ( dir );
if ( dir ) {
g_errno = EEXIST;
return log("admin: Trying to create collection %s but "
"directory %s already exists on disk.",coll,dname);
}
// create the record in memory
CollectionRec *cr = new (CollectionRec);
if ( ! cr )
return log("admin: Failed to allocated %"INT32" bytes for new "
"collection record for \"%s\".",
(int32_t)sizeof(CollectionRec),coll);
// register the mem
mnew ( cr , sizeof(CollectionRec) , "CollectionRec" );
// get copy collection
//CollectionRec *cpcrec = NULL;
//if ( cpc && cpc[0] ) cpcrec = getRec ( cpc , cpclen );
//if ( cpc && cpc[0] && ! cpcrec )
// log("admin: Collection \"%s\" to copy config from does not "
// "exist.",cpc);
// set collnum right for g_parms.setToDefault() call
//cr->m_collnum = newCollnum;
// . get the default.conf from working dir if there
// . i think this calls CollectionRec::reset() which resets all of its
// rdbbase classes for its collnum so m_collnum needs to be right
//g_parms.setToDefault( (char *)cr );
// get the default.conf from working dir if there
//g_parms.setToDefault( (char *)cr , OBJ_COLL );
g_parms.setToDefault( (char *)cr , OBJ_COLL , cr );
// put search results back so it doesn't mess up results in qatest123
if ( strcmp(coll,"qatest123") == 0 )
cr->m_sameLangWeight = 20.0;
/*
// the default conf file
char tmp1[1024];
sprintf ( tmp1 , "%sdefault.conf" , g_hostdb.m_dir );
// . set our parms from the file.
// . accepts OBJ_COLLECTIONREC or OBJ_CONF
g_parms.setFromFile ( cr , NULL , tmp1 );
*/
// this will override all
// if ( cpcrec ) {
// // copy it, but not the timedb hashtable, etc.
// int32_t size = (char *)&(cpcrec->m_END_COPY) - (char *)cpcrec;
// // JAB: bad gbmemcpy - no donut!
// // this is not how objects are supposed to be copied!!!
// gbmemcpy ( cr , cpcrec , size);
// }
// set coll id and coll name for coll id #i
strcpy ( cr->m_coll , coll );
cr->m_collLen = gbstrlen ( coll );
cr->m_collnum = newCollnum;
// point to this, so Rdb and RdbBase can reference it
coll = cr->m_coll;
//
// BEGIN NEW CODE
//
//
// get token and crawlname if customCrawl is 1 or 2
//
char *token = NULL;
char *crawl = NULL;
SafeBuf tmp;
// . return true with g_errno set on error
// . if we fail to set a parm right we should force ourselves
// out sync
if ( customCrawl ) {
if ( ! tmp.safeStrcpy ( coll ) ) return true;
token = tmp.getBufStart();
// diffbot coll name format is <token>-<crawlname>
char *h = strchr ( tmp.getBufStart() , '-' );
if ( ! h ) {
log("crawlbot: bad custom collname");
g_errno = EBADENGINEER;
mdelete ( cr, sizeof(CollectionRec), "CollectionRec" );
delete ( cr );
return true;
}
*h = '\0';
crawl = h + 1;
if ( ! crawl[0] ) {
log("crawlbot: bad custom crawl name");
mdelete ( cr, sizeof(CollectionRec), "CollectionRec" );
delete ( cr );
g_errno = EBADENGINEER;
return true;
}
// or if too big!
if ( gbstrlen(crawl) > 30 ) {
log("crawlbot: crawlbot crawl NAME is over 30 chars");
mdelete ( cr, sizeof(CollectionRec), "CollectionRec" );
delete ( cr );
g_errno = EBADENGINEER;
return true;
}
}
//log("parms: added new collection \"%s\"", collName );
cr->m_maxToCrawl = -1;
cr->m_maxToProcess = -1;
if ( customCrawl ) {
// always index spider status docs now
cr->m_indexSpiderReplies = true;
// remember the token
cr->m_diffbotToken.set ( token );
cr->m_diffbotCrawlName.set ( crawl );
// bring this back
cr->m_diffbotApiUrl.set ( "" );
cr->m_diffbotUrlCrawlPattern.set ( "" );
cr->m_diffbotUrlProcessPattern.set ( "" );
cr->m_diffbotPageProcessPattern.set ( "" );
cr->m_diffbotUrlCrawlRegEx.set ( "" );
cr->m_diffbotUrlProcessRegEx.set ( "" );
cr->m_diffbotMaxHops = -1;
cr->m_spiderStatus = SP_INITIALIZING;
// do not spider more than this many urls total.
// -1 means no max.
cr->m_maxToCrawl = 100000;
// do not process more than this. -1 means no max.
cr->m_maxToProcess = 100000;
// -1 means no max
cr->m_maxCrawlRounds = -1;
// diffbot download docs up to 10MB so we don't truncate
// things like sitemap.xml
cr->m_maxTextDocLen = 10000000;
cr->m_maxOtherDocLen = 10000000;
// john wants deduping on by default to avoid
// processing similar pgs
cr->m_dedupingEnabled = true;
// show the ban links in the search results. the
// collection name is cryptographic enough to show that
cr->m_isCustomCrawl = customCrawl;
cr->m_diffbotOnlyProcessIfNewUrl = true;
// default respider to off
cr->m_collectiveRespiderFrequency = 0.0;
//cr->m_restrictDomain = true;
// reset the crawl stats
// always turn off gigabits so &s=1000 can do summary skipping
cr->m_docsToScanForTopics = 0;
// turn off link voting, etc. to speed up
cr->m_getLinkInfo = false;
cr->m_computeSiteNumInlinks = false;
}
// . this will core if a host was dead and then when it came
// back up host #0's parms.cpp told it to add a new coll
cr->m_diffbotCrawlStartTime = getTimeGlobalNoCore();
cr->m_diffbotCrawlEndTime = 0;
// . just the basics on these for now
// . if certain parms are changed then the url filters
// must be rebuilt, as well as possibly the waiting tree!!!
// . need to set m_urlFiltersHavePageCounts etc.
cr->rebuildUrlFilters ( );
cr->m_useRobotsTxt = true;
// reset crawler stats.they should be loaded from crawlinfo.txt
memset ( &cr->m_localCrawlInfo , 0 , sizeof(CrawlInfo) );
memset ( &cr->m_globalCrawlInfo , 0 , sizeof(CrawlInfo) );
// note that
log("colldb: initial revival for %s",cr->m_coll);
// . assume we got some urls ready to spider
// . Spider.cpp will wait SPIDER_DONE_TIME seconds and if it has no
// urls it spidered in that time these will get set to 0 and it
// will send out an email alert if m_sentCrawlDoneAlert is not true.
cr->m_localCrawlInfo.m_hasUrlsReadyToSpider = 1;
cr->m_globalCrawlInfo.m_hasUrlsReadyToSpider = 1;
// set some defaults. max spiders for all priorities in this
// collection. NO, default is in Parms.cpp.
//cr->m_maxNumSpiders = 10;
//cr->m_needsSave = 1;
// start the spiders!
cr->m_spideringEnabled = true;
// override this?
saveIt = true;
//
// END NEW CODE
//
//log("admin: adding coll \"%s\" (new=%"INT32")",coll,(int32_t)isNew);
// MDW: create the new directory
retry22:
if ( ::mkdir ( dname ,
getDirCreationFlags() ) ) {
// S_IRUSR | S_IWUSR | S_IXUSR |
// S_IRGRP | S_IWGRP | S_IXGRP |
// S_IROTH | S_IXOTH ) ) {
// valgrind?
if ( errno == EINTR ) goto retry22;
g_errno = errno;
mdelete ( cr , sizeof(CollectionRec) , "CollectionRec" );
delete ( cr );
return log("admin: Creating directory %s had error: "
"%s.", dname,mstrerror(g_errno));
}
// save it into this dir... might fail!
if ( saveIt && ! cr->save() ) {
mdelete ( cr , sizeof(CollectionRec) , "CollectionRec" );
delete ( cr );
return log("admin: Failed to save file %s: %s",
dname,mstrerror(g_errno));
}
if ( ! registerCollRec ( cr , true ) )
return false;
// add the rdbbases for this coll, CollectionRec::m_bases[]
if ( ! addRdbBasesForCollRec ( cr ) )
return false;
return true;
}
void CollectionRec::setBasePtr ( char rdbId , class RdbBase *base ) {
// if in the process of swapping in, this will be false...
//if ( m_swappedOut ) { char *xx=NULL;*xx=0; }
if ( rdbId < 0 || rdbId >= RDB_END ) { char *xx=NULL;*xx=0; }
// Rdb::deleteColl() will call this even though we are swapped in
// but it calls it with "base" set to NULL after it nukes the RdbBase
// so check if base is null here.
if ( base && m_bases[ (unsigned char)rdbId ]){ char *xx=NULL;*xx=0; }
m_bases [ (unsigned char)rdbId ] = base;
}
RdbBase *CollectionRec::getBasePtr ( char rdbId ) {
if ( rdbId < 0 || rdbId >= RDB_END ) { char *xx=NULL;*xx=0; }
return m_bases [ (unsigned char)rdbId ];
}
static bool s_inside = false;
// . returns NULL w/ g_errno set on error.
// . TODO: ensure not called from in thread, not thread safe
RdbBase *CollectionRec::getBase ( char rdbId ) {
if ( s_inside ) { char *xx=NULL;*xx=0; }
if ( ! m_swappedOut ) return m_bases[(unsigned char)rdbId];
log("cdb: swapin collnum=%"INT32"",(int32_t)m_collnum);
// sanity!
if ( g_threads.amThread() ) { char *xx=NULL;*xx=0; }
s_inside = true;
// turn off quickpoll to avoid getbase() being re-called and
// coring from s_inside being true
int32_t saved = g_conf.m_useQuickpoll;
g_conf.m_useQuickpoll = false;
// load them back in. return NULL w/ g_errno set on error.
if ( ! g_collectiondb.addRdbBasesForCollRec ( this ) ) {
log("coll: error swapin: %s",mstrerror(g_errno));
g_conf.m_useQuickpoll = saved;
s_inside = false;
return NULL;
}
g_conf.m_useQuickpoll = saved;
s_inside = false;
g_collectiondb.m_numCollsSwappedOut--;
m_swappedOut = false;
log("coll: swapin was successful for collnum=%"INT32"",(int32_t)m_collnum);
return m_bases[(unsigned char)rdbId];
}
bool CollectionRec::swapOut ( ) {
if ( m_swappedOut ) return true;
log("cdb: swapout collnum=%"INT32"",(int32_t)m_collnum);
// free all RdbBases in each rdb
for ( int32_t i = 0 ; i < g_process.m_numRdbs ; i++ ) {
Rdb *rdb = g_process.m_rdbs[i];
// this frees all the RdbBase::m_files and m_maps for the base
rdb->resetBase ( m_collnum );
}
// now free each base itself
for ( int32_t i = 0 ; i < g_process.m_numRdbs ; i++ ) {
RdbBase *base = m_bases[i];
if ( ! base ) continue;
mdelete (base, sizeof(RdbBase), "Rdb Coll");
delete (base);
m_bases[i] = NULL;
}
m_swappedOut = true;
g_collectiondb.m_numCollsSwappedOut++;
return true;
}
// . called only by addNewColl() and by addExistingColl()
bool Collectiondb::registerCollRec ( CollectionRec *cr , bool isNew ) {
// add m_recs[] and to hashtable
if ( ! setRecPtr ( cr->m_collnum , cr ) )
return false;
return true;
}
// swap it in
bool Collectiondb::addRdbBaseToAllRdbsForEachCollRec ( ) {
for ( int32_t i = 0 ; i < m_numRecs ; i++ ) {
CollectionRec *cr = m_recs[i];
if ( ! cr ) continue;
// skip if swapped out
if ( cr->m_swappedOut ) continue;
// add rdb base files etc. for it
addRdbBasesForCollRec ( cr );
}
// now clean the trees. moved this into here from
// addRdbBasesForCollRec() since we call addRdbBasesForCollRec()
// now from getBase() to load on-demand for saving memory
cleanTrees();
return true;
}
bool Collectiondb::addRdbBasesForCollRec ( CollectionRec *cr ) {
char *coll = cr->m_coll;
//////
//
// if we are doing a dump from the command line, skip this stuff
//
//////
if ( g_dumpMode ) return true;
// tell rdbs to add one, too
//if ( ! g_indexdb.getRdb()->addRdbBase1 ( coll ) ) goto hadError;
if ( ! g_posdb.getRdb()->addRdbBase1 ( coll ) ) goto hadError;
//if ( ! g_datedb.getRdb()->addRdbBase1 ( coll ) ) goto hadError;
if ( ! g_titledb.getRdb()->addRdbBase1 ( coll ) ) goto hadError;
//if ( ! g_revdb.getRdb()->addRdbBase1 ( coll ) ) goto hadError;
//if ( ! g_sectiondb.getRdb()->addRdbBase1 ( coll ) ) goto hadError;
if ( ! g_tagdb.getRdb()->addRdbBase1 ( coll ) ) goto hadError;
//if ( ! g_catdb.getRdb()->addRdbBase1 ( coll ) ) goto hadError;
//if ( ! g_checksumdb.getRdb()->addRdbBase1 ( coll ) ) goto hadError;
//if ( ! g_tfndb.getRdb()->addRdbBase1 ( coll ) ) goto hadError;
if ( ! g_clusterdb.getRdb()->addRdbBase1 ( coll ) ) goto hadError;
if ( ! g_linkdb.getRdb()->addRdbBase1 ( coll ) ) goto hadError;
if ( ! g_spiderdb.getRdb()->addRdbBase1 ( coll ) ) goto hadError;
if ( ! g_doledb.getRdb()->addRdbBase1 ( coll ) ) goto hadError;
// now clean the trees
//cleanTrees();
// debug message
//log ( LOG_INFO, "db: verified collection \"%s\" (%"INT32").",
// coll,(int32_t)cr->m_collnum);
// tell SpiderCache about this collection, it will create a
// SpiderCollection class for it.
//g_spiderCache.reset1();
// success
return true;
hadError:
log("db: error registering coll: %s",mstrerror(g_errno));
return false;
}
/*
bool Collectiondb::isAdmin ( HttpRequest *r , TcpSocket *s ) {
if ( r->getLong("admin",1) == 0 ) return false;
if ( g_conf.isMasterAdmin ( s , r ) ) return true;
char *c = r->getString ( "c" );
CollectionRec *cr = getRec ( c );
if ( ! cr ) return false;
return g_users.hasPermission ( r , PAGE_SEARCH );
//return cr->hasPermission ( r , s );
}
void savingCheckWrapper1 ( int fd , void *state ) {
WaitEntry *we = (WaitEntry *)state;
// no state?
if ( ! we ) { log("colldb: we1 is null"); return; }
// unregister too
g_loop.unregisterSleepCallback ( state,savingCheckWrapper1 );
// if it blocked again i guess tree is still saving
if ( ! g_collectiondb.resetColl ( we->m_coll ,
we ,
we->m_purgeSeeds))
return;
// all done
we->m_callback ( we->m_state );
}
void savingCheckWrapper2 ( int fd , void *state ) {
WaitEntry *we = (WaitEntry *)state;
// no state?
if ( ! we ) { log("colldb: we2 is null"); return; }
// unregister too
g_loop.unregisterSleepCallback ( state,savingCheckWrapper2 );
// if it blocked again i guess tree is still saving
if ( ! g_collectiondb.deleteRec ( we->m_coll , we ) ) return;
// all done
we->m_callback ( we->m_state );
}
*/
/*
// delete all records checked in the list
bool Collectiondb::deleteRecs ( HttpRequest *r ) {
for ( int32_t i = 0 ; i < r->getNumFields() ; i++ ) {
char *f = r->getField ( i );
if ( strncmp ( f , "del" , 3 ) != 0 ) continue;
char *coll = f + 3;
//if ( ! is_digit ( f[3] ) ) continue;
//int32_t h = atol ( f + 3 );
deleteRec ( coll , NULL );
}
return true;
}
*/
/*
// . delete a collection
// . this uses blocking unlinks, may make non-blocking later
// . returns false if blocked, true otherwise
bool Collectiondb::deleteRec ( char *coll , WaitEntry *we ) {
// force on for now
//deleteTurkdb = true;
// no spiders can be out. they may be referencing the CollectionRec
// in XmlDoc.cpp... quite likely.
//if ( g_conf.m_spideringEnabled ||
// g_spiderLoop.m_numSpidersOut > 0 ) {
// log("admin: Can not delete collection while "
// "spiders are enabled or active.");
// return false;
//}
// ensure it's not NULL
if ( ! coll ) {
log(LOG_LOGIC,"admin: Collection name to delete is NULL.");
g_errno = ENOTFOUND;
return true;
}
// find the rec for this collection
collnum_t collnum = getCollnum ( coll );
return deleteRec2 ( collnum , we );
}
*/
// if there is an outstanding disk read thread or merge thread then
// Spider.cpp will handle the delete in the callback.
// this is now tryToDeleteSpiderColl in Spider.cpp
/*
void Collectiondb::deleteSpiderColl ( SpiderColl *sc ) {
sc->m_deleteMyself = true;
// if not currently being accessed nuke it now
if ( ! sc->m_msg5.m_waitingForList &&
! sc->m_msg5b.m_waitingForList &&
! sc->m_msg1.m_mcast.m_inUse ) {
mdelete ( sc, sizeof(SpiderColl),"nukecr2");
delete ( sc );
return;
}
}
*/
/// this deletes the collection, not just part of a reset.
bool Collectiondb::deleteRec2 ( collnum_t collnum ) { //, WaitEntry *we ) {
// do not allow this if in repair mode
if ( g_repair.isRepairActive() && g_repair.m_collnum == collnum ) {
log("admin: Can not delete collection while in repair mode.");
g_errno = EBADENGINEER;
return true;
}
// bitch if not found
if ( collnum < 0 ) {
g_errno = ENOTFOUND;
log(LOG_LOGIC,"admin: Collection #%"INT32" is bad, "
"delete failed.",(int32_t)collnum);
return true;
}
CollectionRec *cr = m_recs [ collnum ];
if ( ! cr ) {
log("admin: Collection id problem. Delete failed.");
g_errno = ENOTFOUND;
return true;
}
if ( g_process.isAnyTreeSaving() ) {
// note it
log("admin: tree is saving. waiting2.");
// all done
return false;
}
// spiders off
//if ( cr->m_spiderColl &&
// cr->m_spiderColl->getTotalOutstandingSpiders() > 0 ) {
// log("admin: Can not delete collection while "
// "spiders are outstanding for collection. Turn off "
// "spiders and wait for them to exit.");
// return false;
//}
char *coll = cr->m_coll;
// note it
log(LOG_INFO,"db: deleting coll \"%s\" (%"INT32")",coll,
(int32_t)cr->m_collnum);
// we need a save
m_needsSave = true;
// nuke doleiptable and waintree and waitingtable
/*
SpiderColl *sc = g_spiderCache.getSpiderColl ( collnum );
sc->m_waitingTree.clear();
sc->m_waitingTable.clear();
sc->m_doleIpTable.clear();
g_spiderLoop.m_lockTable.clear();
g_spiderLoop.m_lockCache.clear(0);
sc->m_lastDownloadCache.clear(collnum);