sqlite/test/fts4aa.test

# 2010 February 02
#
# 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 the FTS4 module.
#
#

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

# If SQLITE_ENABLE_FTS3 is defined, omit this file.
ifcapable !fts3 {
  finish_test
  return
}

# Create the fts_kjv_genesis procedure which fills and FTS3/4 table with
# the complete text of the Book of Genesis.
#
source $testdir/genesis.tcl

# The following is a list of queries to perform against the above
# FTS3/FTS4 database.  We will be trying these queries in various
# configurations to ensure that they always return the same answers.
#
set fts4aa_queries {
  {abraham}
  {the king}
  {"the king"}
  {abraham OR joseph}
  {ab* OR jos*}
  {lived t*}
  {spake hebrew}
  {melchizedek}
  {t* melchizedek}
  {melchizedek t*}
}
unset -nocomplain fts4aa_res

# Set up the baseline results
#
do_test fts4aa-1.0 {
  db eval {
    CREATE VIRTUAL TABLE t1 USING fts4(words, tokenize porter);
  }
  fts_kjv_genesis
  foreach q $::fts4aa_queries {
    set r [db eval {SELECT docid FROM t1 WHERE words MATCH $q ORDER BY docid}]
    set ::fts4aa_res($q) $r
  }
} {}

# Legacy test cases
#
do_test fts4aa-1.1 {
  db eval {
    SELECT docid FROM t1 EXCEPT SELECT docid FROM t1_docsize
  }
} {}
do_test fts4aa-1.2 {
  db eval {
    SELECT docid FROM t1_docsize EXCEPT SELECT docid FROM t1
  }
} {}

proc mit {blob} {
  set scan(littleEndian) i*
  set scan(bigEndian) I*
  binary scan $blob $scan($::tcl_platform(byteOrder)) r
  return $r
}
db func mit mit

do_test fts4aa-1.3 {
  db eval {
    SELECT docid, mit(matchinfo(t1, 'pcxnal')) FROM t1 WHERE t1 MATCH 'melchizedek';
  }
} {1014018 {1 1 1 1 1 1533 25 20}}
do_test fts4aa-1.4 {
  db eval {
    SELECT docid, mit(matchinfo(t1, 'pcxnal')) FROM t1
     WHERE t1 MATCH 'spake hebrew'
     ORDER BY docid;
  }
} {1039014 {2 1 1 40 40 1 6 6 1533 25 42} 1039017 {2 1 1 40 40 1 6 6 1533 25 26}}
do_test fts4aa-1.5 {
  db eval {
    SELECT docid, mit(matchinfo(t1, 'pcxnal')) FROM t1
     WHERE t1 MATCH 'laban overtook jacob'
     ORDER BY docid;
  }
} {1031025 {3 1 2 54 46 1 3 3 2 181 160 1533 25 24}}

do_test fts4aa-1.6 {
  db eval {
    DELETE FROM t1 WHERE docid!=1050026;
    SELECT hex(size) FROM t1_docsize;
    SELECT hex(value) FROM t1_stat;
  }
} {17 01176F}

do_test fts4aa-1.7 {
  db eval {
    SELECT docid FROM t1 EXCEPT SELECT docid FROM t1_docsize
  }
} {}
do_test fts4aa-1.8 {
  db eval {
    SELECT docid FROM t1_docsize EXCEPT SELECT docid FROM t1
  }
} {}
ifcapable fts4_deferred {
  do_test fts4aa-1.9 {
    # Note: Token 'in' is being deferred in the following query. 
    db eval {
      SELECT docid, mit(matchinfo(t1, 'pcxnal')) FROM t1
       WHERE t1 MATCH 'joseph died in egypt'
       ORDER BY docid;
    }
  } {1050026 {4 1 1 1 1 1 1 1 2 1 1 1 1 1 1 23 23}}
}

# Should get the same search results from FTS3
#
do_test fts4aa-2.0 {
  db eval {
    DROP TABLE t1;
    CREATE VIRTUAL TABLE t1 USING fts3(words, tokenize porter);
  }
  fts_kjv_genesis
} {}
unset -nocomplain ii
set ii 0
foreach {q r} [array get fts4aa_res] {
  incr ii
  do_test fts4aa-2.$ii {
    db eval {SELECT docid FROM t1 WHERE words MATCH $::q ORDER BY docid}
  } $r
}

# Should get the same search results when the page size is very large
#
do_test fts4aa-3.0 {
  db close
  forcedelete test.db
  sqlite3 db test.db
  db eval {
    PRAGMA page_size=65536;
    CREATE VIRTUAL TABLE t1 USING fts4(words, tokenize porter);
  }
  fts_kjv_genesis
} {}
unset -nocomplain ii
set ii 0
foreach {q r} [array get fts4aa_res] {
  incr ii
  do_test fts4aa-3.$ii {
    db eval {SELECT docid FROM t1 WHERE words MATCH $::q ORDER BY docid}
  } $r
}

# Should get the same search results when an authorizer prevents
# all PRAGMA statements.
#
proc no_pragma_auth {code arg1 arg2 arg3 arg4} {
  if {$code=="SQLITE_PRAGMA"} {return SQLITE_DENY}
  return SQLITE_OK;
}
do_test fts4aa-4.0 {
  db auth ::no_pragma_auth
  db eval {
    DROP TABLE t1;
    CREATE VIRTUAL TABLE t1 USING fts4(words, tokenize porter);
  }
  fts_kjv_genesis
} {}
unset -nocomplain ii
set ii 0
foreach {q r} [array get fts4aa_res] {
  incr ii
  do_test fts4aa-4.$ii {
    db eval {SELECT docid FROM t1 WHERE words MATCH $::q ORDER BY docid}
  } $r
}

finish_test