test/autoindex1.test

# 2010 April 07
#
# The author disclaims copyright to this source code.  In place of
# a legal notice, here is a blessing:
#
#    May you do good and not evil.
#    May you find forgiveness for yourself and forgive others.
#    May you share freely, never taking more than you give.
#
#*************************************************************************
# This file implements regression tests for SQLite library.  The
# focus of this script is testing automatic index creation logic.
#

set testdir [file dirname $argv0]
source $testdir/tester.tcl

# If the library is not compiled with automatic index support then
# skip all tests in this file.
#
ifcapable {!autoindex} {
  finish_test
  return
}

# Setup for logging 
db close
sqlite3_shutdown
test_sqlite3_log [list lappend ::log]
set ::log [list]
sqlite3 db test.db


# With automatic index turned off, we do a full scan of the T2 table
do_test autoindex1-100 {
  db eval {
    CREATE TABLE t1(a,b);
    INSERT INTO t1 VALUES(1,11);
    INSERT INTO t1 VALUES(2,22);
    INSERT INTO t1 SELECT a+2, b+22 FROM t1;
    INSERT INTO t1 SELECT a+4, b+44 FROM t1;
    CREATE TABLE t2(c,d);
    INSERT INTO t2 SELECT a, 900+b FROM t1;
  }
  db eval {
    PRAGMA automatic_index=OFF;
    SELECT b, d FROM t1 JOIN t2 ON a=c ORDER BY b;
  }
} {11 911 22 922 33 933 44 944 55 955 66 966 77 977 88 988}
do_test autoindex1-101 {
  db status step
} {63}
do_test autoindex1-102 {
  db status autoindex
} {0}

# With autoindex turned on, we build an index once and then use that index
# to find T2 values.
do_test autoindex1-110 {
  db eval {
    PRAGMA automatic_index=ON;
    SELECT b, d FROM t1 JOIN t2 ON a=c ORDER BY b;
  }
} {11 911 22 922 33 933 44 944 55 955 66 966 77 977 88 988}
do_test autoindex1-111 {
  db status step
} {7}
do_test autoindex1-112 {
  db status autoindex
} {7}
do_test autoindex1-113 {
  set ::log
} {SQLITE_WARNING_AUTOINDEX {automatic index on t2(c)}}

db close
sqlite3_shutdown
test_sqlite3_log
sqlite3_initialize
sqlite3 db test.db

# The same test as above, but this time the T2 query is a subquery rather
# than a join.
do_test autoindex1-200 {
  db eval {
    PRAGMA automatic_index=OFF;
    SELECT b, (SELECT d FROM t2 WHERE c=a) FROM t1;
  }
} {11 911 22 922 33 933 44 944 55 955 66 966 77 977 88 988}
do_test autoindex1-201 {
  db status step
} {35}
do_test autoindex1-202 {
  db status autoindex
} {0}
do_test autoindex1-210 {
  db eval {
    PRAGMA automatic_index=ON;
    ANALYZE;
    UPDATE sqlite_stat1 SET stat='10000' WHERE tbl='t1';
    ANALYZE sqlite_master;
    SELECT b, (SELECT d FROM t2 WHERE c=a) FROM t1;
  }
} {11 911 22 922 33 933 44 944 55 955 66 966 77 977 88 988}
do_test autoindex1-211 {
  db status step
} {7}
do_test autoindex1-212 {
  db status autoindex
} {7}


# Modify the second table of the join while the join is in progress
#
do_execsql_test autoindex1-299 {
  UPDATE sqlite_stat1 SET stat='10000' WHERE tbl='t2';
  ANALYZE sqlite_master;
  EXPLAIN QUERY PLAN
  SELECT b, d FROM t1 CROSS JOIN t2 ON (c=a);
} {/AUTOMATIC COVERING INDEX/}
do_test autoindex1-300 {
  set r {}
  db eval {SELECT b, d FROM t1 CROSS JOIN t2 ON (c=a)} {
    lappend r $b $d
    db eval {UPDATE t2 SET d=d+1}
  }
  set r
} {11 911 22 922 33 933 44 944 55 955 66 966 77 977 88 988}
do_test autoindex1-310 {
  db eval {SELECT d FROM t2 ORDER BY d}
} {919 930 941 952 963 974 985 996}

# The next test does a 10-way join on unindexed tables.  Without
# automatic indices, the join will take a long time to complete.
# With automatic indices, it should only take about a second.
#
do_test autoindex1-400 {
  db eval {
    CREATE TABLE t4(a, b);
    INSERT INTO t4 VALUES(1,2);
    INSERT INTO t4 VALUES(2,3);
  }
  for {set n 2} {$n<4096} {set n [expr {$n+$n}]} {
    db eval {INSERT INTO t4 SELECT a+$n, b+$n FROM t4}
  }
  db eval {
    SELECT count(*) FROM t4;
  }
} {4096}
do_test autoindex1-401 {
  db eval {
    SELECT count(*)
      FROM t4 AS x1
      JOIN t4 AS x2 ON x2.a=x1.b
      JOIN t4 AS x3 ON x3.a=x2.b
      JOIN t4 AS x4 ON x4.a=x3.b
      JOIN t4 AS x5 ON x5.a=x4.b
      JOIN t4 AS x6 ON x6.a=x5.b
      JOIN t4 AS x7 ON x7.a=x6.b
      JOIN t4 AS x8 ON x8.a=x7.b
      JOIN t4 AS x9 ON x9.a=x8.b
      JOIN t4 AS x10 ON x10.a=x9.b;
  }
} {4087}

# Ticket [8011086c85c6c404014c947fcf3eb9f42b184a0d] from 2010-07-08
# Make sure automatic indices are not created for the RHS of an IN expression
# that is not a correlated subquery.
#
do_execsql_test autoindex1-500 {
  CREATE TABLE t501(a INTEGER PRIMARY KEY, b);
  CREATE TABLE t502(x INTEGER PRIMARY KEY, y);
  INSERT INTO sqlite_stat1(tbl,idx,stat) VALUES('t501',null,'1000000');
  INSERT INTO sqlite_stat1(tbl,idx,stat) VALUES('t502',null,'1000');
  ANALYZE sqlite_master;
  EXPLAIN QUERY PLAN
  SELECT b FROM t501
   WHERE t501.a IN (SELECT x FROM t502 WHERE y=?);
} {
  0 0 0 {SEARCH TABLE t501 USING INTEGER PRIMARY KEY (rowid=?)} 
  0 0 0 {EXECUTE LIST SUBQUERY 1} 
  1 0 0 {SCAN TABLE t502}
}
do_execsql_test autoindex1-501 {
  EXPLAIN QUERY PLAN
  SELECT b FROM t501
   WHERE t501.a IN (SELECT x FROM t502 WHERE y=t501.b);
} {
  0 0 0 {SCAN TABLE t501} 
  0 0 0 {EXECUTE CORRELATED LIST SUBQUERY 1} 
  1 0 0 {SEARCH TABLE t502 USING AUTOMATIC COVERING INDEX (y=?)}
}
do_execsql_test autoindex1-502 {
  EXPLAIN QUERY PLAN
  SELECT b FROM t501
   WHERE t501.a=123
     AND t501.a IN (SELECT x FROM t502 WHERE y=t501.b);
} {
  0 0 0 {SEARCH TABLE t501 USING INTEGER PRIMARY KEY (rowid=?)} 
  0 0 0 {EXECUTE CORRELATED LIST SUBQUERY 1} 
  1 0 0 {SCAN TABLE t502}
}


# The following code checks a performance regression reported on the
# mailing list on 2010-10-19.  The problem is that the nRowEst field
# of ephermeral tables was not being initialized correctly and so no
# automatic index was being created for the emphemeral table when it was
# used as part of a join.
#
do_execsql_test autoindex1-600 {
  CREATE TABLE flock_owner(
    owner_rec_id INTEGER CONSTRAINT flock_owner_key PRIMARY KEY,
    flock_no VARCHAR(6) NOT NULL REFERENCES flock (flock_no),
    owner_person_id INTEGER NOT NULL REFERENCES person (person_id),
    owner_change_date TEXT, last_changed TEXT NOT NULL,
    CONSTRAINT fo_owner_date UNIQUE (flock_no, owner_change_date)
  );
  CREATE TABLE sheep (
    Sheep_No char(7) NOT NULL,
    Date_of_Birth char(8),
    Sort_DoB text,
    Flock_Book_Vol char(2),
    Breeder_No char(6),
    Breeder_Person integer,
    Originating_Flock char(6),
    Registering_Flock char(6),
    Tag_Prefix char(9),
    Tag_No char(15),
    Sort_Tag_No integer,
    Breeders_Temp_Tag char(15),
    Sex char(1),
    Sheep_Name char(32),
    Sire_No char(7),
    Dam_No char(7),
    Register_Code char(1),
    Colour char(48),
    Colour_Code char(2),
    Pattern_Code char(8),
    Horns char(1),
    Litter_Size char(1),
    Coeff_of_Inbreeding real,
    Date_of_Registration text,
    Date_Last_Changed text,
    UNIQUE(Sheep_No));
  CREATE INDEX fo_flock_no_index  
              ON flock_owner (flock_no);
  CREATE INDEX fo_owner_change_date_index  
              ON flock_owner (owner_change_date);
  CREATE INDEX fo_owner_person_id_index  
              ON flock_owner (owner_person_id);
  CREATE INDEX sheep_org_flock_index  
           ON sheep (originating_flock);
  CREATE INDEX sheep_reg_flock_index  
           ON sheep (registering_flock);
  EXPLAIN QUERY PLAN
  SELECT x.sheep_no, x.registering_flock, x.date_of_registration
   FROM sheep x LEFT JOIN
       (SELECT s.sheep_no, prev.flock_no, prev.owner_person_id,
       s.date_of_registration, prev.owner_change_date
       FROM sheep s JOIN flock_owner prev ON s.registering_flock =
   prev.flock_no
       AND (prev.owner_change_date <= s.date_of_registration || ' 00:00:00')
       WHERE NOT EXISTS
           (SELECT 'x' FROM flock_owner later
           WHERE prev.flock_no = later.flock_no
           AND later.owner_change_date > prev.owner_change_date
           AND later.owner_change_date <= s.date_of_registration||' 00:00:00')
       ) y ON x.sheep_no = y.sheep_no
   WHERE y.sheep_no IS NULL
   ORDER BY x.registering_flock;
} {
  1 0 0 {SCAN TABLE sheep AS s} 
  1 1 1 {SEARCH TABLE flock_owner AS prev USING INDEX sqlite_autoindex_flock_owner_1 (flock_no=? AND owner_change_date<?)} 
  1 0 0 {EXECUTE CORRELATED SCALAR SUBQUERY 2} 
  2 0 0 {SEARCH TABLE flock_owner AS later USING COVERING INDEX sqlite_autoindex_flock_owner_1 (flock_no=? AND owner_change_date>? AND owner_change_date<?)} 
  0 0 0 {SCAN TABLE sheep AS x USING INDEX sheep_reg_flock_index} 
  0 1 1 {SEARCH SUBQUERY 1 AS y USING AUTOMATIC COVERING INDEX (sheep_no=?)}
}


do_execsql_test autoindex1-700 {
  CREATE TABLE t5(a, b, c);
  EXPLAIN QUERY PLAN SELECT a FROM t5 WHERE b=10 ORDER BY c;
} {
  0 0 0 {SCAN TABLE t5} 
  0 0 0 {USE TEMP B-TREE FOR ORDER BY}
}

# The following checks a performance issue reported on the sqlite-dev
# mailing list on 2013-01-10
#
do_execsql_test autoindex1-800 {
  CREATE TABLE accounts(
    _id INTEGER PRIMARY KEY AUTOINCREMENT,
    account_name TEXT,
    account_type TEXT,
    data_set TEXT
  );
  CREATE TABLE data(
    _id INTEGER PRIMARY KEY AUTOINCREMENT,
    package_id INTEGER REFERENCES package(_id),
    mimetype_id INTEGER REFERENCES mimetype(_id) NOT NULL,
    raw_contact_id INTEGER REFERENCES raw_contacts(_id) NOT NULL,
    is_read_only INTEGER NOT NULL DEFAULT 0,
    is_primary INTEGER NOT NULL DEFAULT 0,
    is_super_primary INTEGER NOT NULL DEFAULT 0,
    data_version INTEGER NOT NULL DEFAULT 0,
    data1 TEXT,
    data2 TEXT,
    data3 TEXT,
    data4 TEXT,
    data5 TEXT,
    data6 TEXT,
    data7 TEXT,
    data8 TEXT,
    data9 TEXT,
    data10 TEXT,
    data11 TEXT,
    data12 TEXT,
    data13 TEXT,
    data14 TEXT,
    data15 TEXT,
    data_sync1 TEXT,
    data_sync2 TEXT,
    data_sync3 TEXT,
    data_sync4 TEXT 
  );
  CREATE TABLE mimetypes(
    _id INTEGER PRIMARY KEY AUTOINCREMENT,
    mimetype TEXT NOT NULL
  );
  CREATE TABLE raw_contacts(
    _id INTEGER PRIMARY KEY AUTOINCREMENT,
    account_id INTEGER REFERENCES accounts(_id),
    sourceid TEXT,
    raw_contact_is_read_only INTEGER NOT NULL DEFAULT 0,
    version INTEGER NOT NULL DEFAULT 1,
    dirty INTEGER NOT NULL DEFAULT 0,
    deleted INTEGER NOT NULL DEFAULT 0,
    contact_id INTEGER REFERENCES contacts(_id),
    aggregation_mode INTEGER NOT NULL DEFAULT 0,
    aggregation_needed INTEGER NOT NULL DEFAULT 1,
    custom_ringtone TEXT,
    send_to_voicemail INTEGER NOT NULL DEFAULT 0,
    times_contacted INTEGER NOT NULL DEFAULT 0,
    last_time_contacted INTEGER,
    starred INTEGER NOT NULL DEFAULT 0,
    display_name TEXT,
    display_name_alt TEXT,
    display_name_source INTEGER NOT NULL DEFAULT 0,
    phonetic_name TEXT,
    phonetic_name_style TEXT,
    sort_key TEXT,
    sort_key_alt TEXT,
    name_verified INTEGER NOT NULL DEFAULT 0,
    sync1 TEXT,
    sync2 TEXT,
    sync3 TEXT,
    sync4 TEXT,
    sync_uid TEXT,
    sync_version INTEGER NOT NULL DEFAULT 1,
    has_calendar_event INTEGER NOT NULL DEFAULT 0,
    modified_time INTEGER,
    is_restricted INTEGER DEFAULT 0,
    yp_source TEXT,
    method_selected INTEGER DEFAULT 0,
    custom_vibration_type INTEGER DEFAULT 0,
    custom_ringtone_path TEXT,
    message_notification TEXT,
    message_notification_path TEXT
  );
  CREATE INDEX data_mimetype_data1_index ON data (mimetype_id,data1);
  CREATE INDEX data_raw_contact_id ON data (raw_contact_id);
  CREATE UNIQUE INDEX mime_type ON mimetypes (mimetype);
  CREATE INDEX raw_contact_sort_key1_index ON raw_contacts (sort_key);
  CREATE INDEX raw_contact_sort_key2_index ON raw_contacts (sort_key_alt);
  CREATE INDEX raw_contacts_contact_id_index ON raw_contacts (contact_id);
  CREATE INDEX raw_contacts_source_id_account_id_index
      ON raw_contacts (sourceid, account_id);
  ANALYZE sqlite_master;
  INSERT INTO sqlite_stat1
     VALUES('raw_contacts','raw_contact_sort_key2_index','1600 4');
  INSERT INTO sqlite_stat1
     VALUES('raw_contacts','raw_contact_sort_key1_index','1600 4');
  INSERT INTO sqlite_stat1
     VALUES('raw_contacts','raw_contacts_source_id_account_id_index',
            '1600 1600 1600');
  INSERT INTO sqlite_stat1
     VALUES('raw_contacts','raw_contacts_contact_id_index','1600 1');
  INSERT INTO sqlite_stat1 VALUES('mimetypes','mime_type','12 1');
  INSERT INTO sqlite_stat1
     VALUES('data','data_mimetype_data1_index','9819 2455 3');
  INSERT INTO sqlite_stat1 VALUES('data','data_raw_contact_id','9819 7');
  INSERT INTO sqlite_stat1 VALUES('accounts',NULL,'1');
  DROP TABLE IF EXISTS sqlite_stat3;
  ANALYZE sqlite_master;
  
  EXPLAIN QUERY PLAN
  SELECT * FROM 
        data JOIN mimetypes ON (data.mimetype_id=mimetypes._id) 
             JOIN raw_contacts ON (data.raw_contact_id=raw_contacts._id) 
             JOIN accounts ON (raw_contacts.account_id=accounts._id)
   WHERE mimetype_id=10 AND data14 IS NOT NULL;
} {/SEARCH TABLE data .*SEARCH TABLE raw_contacts/}
do_execsql_test autoindex1-801 {
  EXPLAIN QUERY PLAN
  SELECT * FROM 
        data JOIN mimetypes ON (data.mimetype_id=mimetypes._id) 
             JOIN raw_contacts ON (data.raw_contact_id=raw_contacts._id) 
             JOIN accounts ON (raw_contacts.account_id=accounts._id)
   WHERE mimetypes._id=10 AND data14 IS NOT NULL;
} {/SEARCH TABLE data .*SEARCH TABLE raw_contacts/}

finish_test