Fossil

Check-in [64a9c81a]
Login

Many hyperlinks are disabled.
Use anonymous login to enable hyperlinks.

Overview
Comment:Add the test-clusters command to verify that all artifacts are reachable through cluster chains. Fix the cluster creator so that it does not create gaps if the number of unclustered entries exceeds 800.
Downloads: Tarball | ZIP archive | SQL archive
Timelines: family | ancestors | descendants | both | trunk
Files: files | file ages | folders
SHA1:64a9c81a222733aebcfb4ec6630280b41361e4fb
User & Date: drh 2010-12-27 21:40:45
Context
2010-12-29
20:20
Allow the "localtime" modifier on date/time functions within ticket report SQL statements. Ticket [bf4f5725f6c90fc2] check-in: 8c3bba8e user: drh tags: trunk
2010-12-27
21:40
Add the test-clusters command to verify that all artifacts are reachable through cluster chains. Fix the cluster creator so that it does not create gaps if the number of unclustered entries exceeds 800. check-in: 64a9c81a user: drh tags: trunk
06:15
Spelling fixes, minor editorial check-in: 6b5c797c user: bharder tags: trunk
Changes
Hide Diffs Unified Diffs Ignore Whitespace Patch

Changes to src/rebuild.c.

425
426
427
428
429
430
431




































































432
433
434
435
436
437
438
    db_close();
    db_open_repository(g.zRepositoryName);
  }
  db_begin_transaction();
  create_cluster();
  db_end_transaction(0);  
}





































































/*
** COMMAND: scrub
** %fossil scrub [--verily] [--force] [REPOSITORY]
**
** The command removes sensitive information (such as passwords) from a
** repository so that the respository can be sent to an untrusted reader.







>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>







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
    db_close();
    db_open_repository(g.zRepositoryName);
  }
  db_begin_transaction();
  create_cluster();
  db_end_transaction(0);  
}

/*
** COMMAND: test-clusters
**
** Verify that all non-private and non-shunned artifacts are accessible
** through the cluster chain.
*/
void test_clusters_cmd(void){
  Bag pending;
  Stmt q;
  int n;
  
  db_find_and_open_repository(0, 2);
  bag_init(&pending);
  db_multi_exec(
    "CREATE TEMP TABLE xdone(x INTEGER PRIMARY KEY);"
    "INSERT INTO xdone SELECT rid FROM unclustered;"
    "INSERT OR IGNORE INTO xdone SELECT rid FROM private;"
    "INSERT OR IGNORE INTO xdone"
         " SELECT blob.rid FROM shun JOIN blob USING(uuid);"
  );
  db_prepare(&q,
    "SELECT rid FROM unclustered WHERE rid IN"
    " (SELECT rid FROM tagxref WHERE tagid=%d)", TAG_CLUSTER
  );
  while( db_step(&q)==SQLITE_ROW ){
    bag_insert(&pending, db_column_int(&q, 0));
  }
  db_finalize(&q);
  while( bag_count(&pending)>0 ){
    Manifest *p;
    int rid = bag_first(&pending);
    int i;
    
    bag_remove(&pending, rid);
    p = manifest_get(rid, CFTYPE_CLUSTER);
    if( p==0 ){
      fossil_fatal("bad cluster: rid=%d", rid);
    }
    for(i=0; i<p->nCChild; i++){
      const char *zUuid = p->azCChild[i];
      int crid = name_to_rid(zUuid);
      if( crid==0 ){
         fossil_warning("cluster (rid=%d) references unknown artifact %s",
                        rid, zUuid);
         continue;
      }
      db_multi_exec("INSERT OR IGNORE INTO xdone VALUES(%d)", crid);
      if( db_exists("SELECT 1 FROM tagxref WHERE tagid=%d AND rid=%d",
                    TAG_CLUSTER, crid) ){
        bag_insert(&pending, crid);
      }
    }
    manifest_destroy(p);
  }
  n = db_int(0, "SELECT count(*) FROM /*scan*/"
                "  (SELECT rid FROM blob EXCEPT SELECT x FROM xdone)");
  if( n==0 ){
    printf("all artifacts reachable through clusters\n");
  }else{
    printf("%d unreachable artifacts:\n", n);
    db_prepare(&q, "SELECT rid, uuid FROM blob WHERE rid NOT IN xdone");
    while( db_step(&q)==SQLITE_ROW ){
      printf("  %3d %s\n", db_column_int(&q,0), db_column_text(&q,1));
    }
    db_finalize(&q);
  }
}

/*
** COMMAND: scrub
** %fossil scrub [--verily] [--force] [REPOSITORY]
**
** The command removes sensitive information (such as passwords) from a
** repository so that the respository can be sent to an untrusted reader.

Changes to src/xfer.c.

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
** Check to see if the number of unclustered entries is greater than
** 100 and if it is, form a new cluster.  Unclustered phantoms do not
** count toward the 100 total.  And phantoms are never added to a new
** cluster.
*/
void create_cluster(void){
  Blob cluster, cksum;

  Stmt q;
  int nUncl;
  int nRow = 0;


  /* We should not ever get any private artifacts in the unclustered table.
  ** But if we do (because of a bug) now is a good time to delete them. */
  db_multi_exec(
    "DELETE FROM unclustered WHERE rid IN (SELECT rid FROM private)"
  );

  nUncl = db_int(0, "SELECT count(*) FROM unclustered /*scan*/"
                    " WHERE NOT EXISTS(SELECT 1 FROM phantom"
                                      " WHERE rid=unclustered.rid)");
  if( nUncl>=100 ){
    blob_zero(&cluster);

    db_prepare(&q, "SELECT uuid FROM unclustered, blob"
                   " WHERE NOT EXISTS(SELECT 1 FROM phantom"
                   "                   WHERE rid=unclustered.rid)"
                   "   AND unclustered.rid=blob.rid"
                   "   AND NOT EXISTS(SELECT 1 FROM shun WHERE uuid=blob.uuid)"
                   " ORDER BY 1");
    while( db_step(&q)==SQLITE_ROW ){
      blob_appendf(&cluster, "M %s\n", db_column_text(&q, 0));
      nRow++;
      if( nRow>=800 && nUncl>nRow+100 ){
        md5sum_blob(&cluster, &cksum);
        blob_appendf(&cluster, "Z %b\n", &cksum);
        blob_reset(&cksum);
        content_put(&cluster, 0, 0, 0);
        blob_reset(&cluster);
        nUncl -= nRow;
        nRow = 0;

      }
    }
    db_finalize(&q);

    db_multi_exec("DELETE FROM unclustered");



    if( nRow>0 ){
      md5sum_blob(&cluster, &cksum);
      blob_appendf(&cluster, "Z %b\n", &cksum);
      blob_reset(&cksum);
      content_put(&cluster, 0, 0, 0);
      blob_reset(&cluster);
    }







>



>












>













|



>



>
|
>
>
>







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
** Check to see if the number of unclustered entries is greater than
** 100 and if it is, form a new cluster.  Unclustered phantoms do not
** count toward the 100 total.  And phantoms are never added to a new
** cluster.
*/
void create_cluster(void){
  Blob cluster, cksum;
  Blob deleteWhere;
  Stmt q;
  int nUncl;
  int nRow = 0;
  int rid;

  /* We should not ever get any private artifacts in the unclustered table.
  ** But if we do (because of a bug) now is a good time to delete them. */
  db_multi_exec(
    "DELETE FROM unclustered WHERE rid IN (SELECT rid FROM private)"
  );

  nUncl = db_int(0, "SELECT count(*) FROM unclustered /*scan*/"
                    " WHERE NOT EXISTS(SELECT 1 FROM phantom"
                                      " WHERE rid=unclustered.rid)");
  if( nUncl>=100 ){
    blob_zero(&cluster);
    blob_zero(&deleteWhere);
    db_prepare(&q, "SELECT uuid FROM unclustered, blob"
                   " WHERE NOT EXISTS(SELECT 1 FROM phantom"
                   "                   WHERE rid=unclustered.rid)"
                   "   AND unclustered.rid=blob.rid"
                   "   AND NOT EXISTS(SELECT 1 FROM shun WHERE uuid=blob.uuid)"
                   " ORDER BY 1");
    while( db_step(&q)==SQLITE_ROW ){
      blob_appendf(&cluster, "M %s\n", db_column_text(&q, 0));
      nRow++;
      if( nRow>=800 && nUncl>nRow+100 ){
        md5sum_blob(&cluster, &cksum);
        blob_appendf(&cluster, "Z %b\n", &cksum);
        blob_reset(&cksum);
        rid = content_put(&cluster, 0, 0, 0);
        blob_reset(&cluster);
        nUncl -= nRow;
        nRow = 0;
        blob_appendf(&deleteWhere, ",%d", rid);
      }
    }
    db_finalize(&q);
    db_multi_exec(
      "DELETE FROM unclustered WHERE rid NOT IN (0 %s)", 
      blob_str(&deleteWhere)
    );
    blob_reset(&deleteWhere);
    if( nRow>0 ){
      md5sum_blob(&cluster, &cksum);
      blob_appendf(&cluster, "Z %b\n", &cksum);
      blob_reset(&cksum);
      content_put(&cluster, 0, 0, 0);
      blob_reset(&cluster);
    }