SQLite

Check-in [d281cb8984]
Login

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

Overview
Comment:Fix a bug in merging FTS language tables for languages other than language 0.
Downloads: Tarball | ZIP archive
Timelines: family | ancestors | descendants | both | fts4-languageid
Files: files | file ages | folders
SHA1: d281cb8984c911a4c0cce2ec299e1351d8e580e4
User & Date: dan 2012-03-02 11:48:50.564
Context
2012-03-02
12:26
Fix the FTS 'optimize' command on multi-lingual databases. (check-in: 65fa693729 user: dan tags: fts4-languageid)
11:48
Fix a bug in merging FTS language tables for languages other than language 0. (check-in: d281cb8984 user: dan tags: fts4-languageid)
2012-03-01
19:44
Add the "languageid=" option to fts4. This code is still largely untested and alsmost certainly buggy. (check-in: bea257f70f user: dan tags: fts4-languageid)
Changes
Unified Diff Ignore Whitespace Patch
Changes to ext/fts3/fts3_write.c.
484
485
486
487
488
489
490
491
492
493
494
495
496
497
498
          getAbsoluteLevel(p, iLangid, iIndex, FTS3_SEGDIR_MAXLEVEL-1)
      );
    }
  }else{
    /* "SELECT * FROM %_segdir WHERE level = ? ORDER BY ..." */
    rc = fts3SqlStmt(p, SQL_SELECT_LEVEL, &pStmt, 0);
    if( rc==SQLITE_OK ){ 
      sqlite3_bind_int(pStmt, 1, iLevel+iIndex*FTS3_SEGDIR_MAXLEVEL);
    }
  }
  *ppStmt = pStmt;
  return rc;
}









|







484
485
486
487
488
489
490
491
492
493
494
495
496
497
498
          getAbsoluteLevel(p, iLangid, iIndex, FTS3_SEGDIR_MAXLEVEL-1)
      );
    }
  }else{
    /* "SELECT * FROM %_segdir WHERE level = ? ORDER BY ..." */
    rc = fts3SqlStmt(p, SQL_SELECT_LEVEL, &pStmt, 0);
    if( rc==SQLITE_OK ){ 
      sqlite3_bind_int(pStmt, 1, getAbsoluteLevel(p, iLangid, iIndex, iLevel));
    }
  }
  *ppStmt = pStmt;
  return rc;
}


Changes to test/fts4langid.test.
20
21
22
23
24
25
26





















27
28
29
30
31
32
33
ifcapable !fts3 {
  finish_test
  return
}

set ::testprefix fts4langid
























do_execsql_test 1.1 {
  CREATE VIRTUAL TABLE t1 USING fts4(a, b, languageid=lang_id);
}

do_execsql_test 1.2 {







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







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
ifcapable !fts3 {
  finish_test
  return
}

set ::testprefix fts4langid

#---------------------------------------------------------------------------
# Test plan:
#
#   1.* - Warm-body tests created for specific purposes during development.
#         Passing these doesn't really prove much.
#
#   2.* - Test that FTS queries only ever return rows associated with
#         the requested language.
#
#   3.* - Test that the 'optimize' and 'rebuild' commands work correctly.
#
#   4.* - Test that if one is provided, the tokenizer xLanguage method
#         is called to configure the tokenizer before tokenizing query
#         or document text.
#
#   5.* - Test the fts4aux table when the associated FTS4 table contains
#         multiple languages.
#
#   6.* - Tests with content= tables. Both where there is a real 
#         underlying content table and where there is not.
#


do_execsql_test 1.1 {
  CREATE VIRTUAL TABLE t1 USING fts4(a, b, languageid=lang_id);
}

do_execsql_test 1.2 {
69
70
71
72
73
74
75
76
































































































































































77
78
do_execsql_test 1.16 {
  SELECT rowid FROM t1 WHERE t1 MATCH 'b' AND lang_id = 1;
} {2}

do_catchsql_test 1.17 {
  INSERT INTO t1(content, lang_id) VALUES('123', -1);
} {1 {constraint failed}}

































































































































































finish_test









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


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
do_execsql_test 1.16 {
  SELECT rowid FROM t1 WHERE t1 MATCH 'b' AND lang_id = 1;
} {2}

do_catchsql_test 1.17 {
  INSERT INTO t1(content, lang_id) VALUES('123', -1);
} {1 {constraint failed}}

do_execsql_test 1.18 {
  DROP TABLE t1;
  CREATE VIRTUAL TABLE t1 USING fts4(languageid=lang_id);
  INSERT INTO t1(content, lang_id) VALUES('A', 13);
  INSERT INTO t1(content, lang_id) VALUES('B', 13);
  INSERT INTO t1(content, lang_id) VALUES('C', 13);
  INSERT INTO t1(content, lang_id) VALUES('D', 13);
  INSERT INTO t1(content, lang_id) VALUES('E', 13);
  INSERT INTO t1(content, lang_id) VALUES('F', 13);
  INSERT INTO t1(content, lang_id) VALUES('G', 13);
  INSERT INTO t1(content, lang_id) VALUES('H', 13);
  INSERT INTO t1(content, lang_id) VALUES('I', 13);
  INSERT INTO t1(content, lang_id) VALUES('J', 13);
  INSERT INTO t1(content, lang_id) VALUES('K', 13);
  INSERT INTO t1(content, lang_id) VALUES('L', 13);
  INSERT INTO t1(content, lang_id) VALUES('M', 13);
  INSERT INTO t1(content, lang_id) VALUES('N', 13);
  INSERT INTO t1(content, lang_id) VALUES('O', 13);
  INSERT INTO t1(content, lang_id) VALUES('P', 13);
  INSERT INTO t1(content, lang_id) VALUES('Q', 13);
  INSERT INTO t1(content, lang_id) VALUES('R', 13);
  INSERT INTO t1(content, lang_id) VALUES('S', 13);
  SELECT rowid FROM t1 WHERE t1 MATCH 'A';
} {}


#-------------------------------------------------------------------------
# Test cases 2.*
#

proc build_multilingual_db_1 {db} {
  $db eval { CREATE VIRTUAL TABLE t2 USING fts4(x, y, languageid=l) }

  set xwords [list zero one two three four five six seven eight nine ten]
  set ywords [list alpha beta gamma delta epsilon zeta eta theta iota kappa]

  for {set i 0} {$i < 1000} {incr i} {
    set iLangid [expr $i%9]
    set x ""
    set y ""

    set x [list]
    lappend x [lindex $xwords [expr ($i / 1000) % 10]]
    lappend x [lindex $xwords [expr ($i / 100)  % 10]]
    lappend x [lindex $xwords [expr ($i / 10)   % 10]]
    lappend x [lindex $xwords [expr ($i / 1)   % 10]]

    set y [list]
    lappend y [lindex $ywords [expr ($i / 1000) % 10]]
    lappend y [lindex $ywords [expr ($i / 100)  % 10]]
    lappend y [lindex $ywords [expr ($i / 10)   % 10]]
    lappend y [lindex $ywords [expr ($i / 1)   % 10]]

    $db eval { INSERT INTO t2(docid, x, y, l) VALUES($i, $x, $y, $iLangid) }
  }
}

proc rowid_list_set_langid {langid} {
  set ::rowid_list_langid $langid
}
proc rowid_list {pattern} {
  set langid $::rowid_list_langid
  set res [list]
  db eval {SELECT docid, x, y FROM t2 WHERE l = $langid ORDER BY docid ASC} {
    if {[string match "*$pattern*" $x] || [string match "*$pattern*" $y]} {
      lappend res $docid
    }
  }
  return $res
}

proc or_merge_list {list1 list2} {
  set res [list]

  set i1 0
  set i2 0

  set n1 [llength $list1]
  set n2 [llength $list2]

  while {$i1 < $n1 && $i2 < $n2} {
    set e1 [lindex $list1 $i1]
    set e2 [lindex $list2 $i2]

    if {$e1==$e2} {
      lappend res $e1
      incr i1
      incr i2
    } elseif {$e1 < $e2} {
      lappend res $e1
      incr i1
    } else {
      lappend res $e2
      incr i2
    }
  }

  concat $res [lrange $list1 $i1 end] [lrange $list2 $i2 end]
}

proc or_merge_lists {args} {
  set res [lindex $args 0]
  for {set i 1} {$i < [llength $args]} {incr i} {
    set res [or_merge_list $res [lindex $args $i]]
  }
  set res
}

proc and_merge_list {list1 list2} {
  foreach i $list2 { set a($i) 1 }
  set res [list]
  foreach i $list1 {
    if {[info exists a($i)]} {lappend res $i}
  }
  set res
}


proc and_merge_lists {args} {
  set res [lindex $args 0]
  for {set i 1} {$i < [llength $args]} {incr i} {
    set res [and_merge_list $res [lindex $args $i]]
  }
  set res
}

proc filter_list {list langid} {
  set res [list]
  foreach i $list {
    if {($i % 9) == $langid} {lappend res $i}
  }
  set res
}

do_test 2.0 { 
  reset_db
  build_multilingual_db_1 db
} {}

proc do_test_2.1 {tn query res_script} {
  for {set langid 0} {$langid < 10} {incr langid} {
    rowid_list_set_langid $langid
    set res [eval $res_script]

    set actual [
      execsql {SELECT docid FROM t2 WHERE t2 MATCH $query AND l = $langid}
    ]
    do_test 2.1.$tn.$langid [list set {} $actual] $res
  }
}

do_test_2.1 1  {delta}          { rowid_list delta }
do_test_2.1 2  {"zero one two"} { rowid_list "zero one two" }
do_test_2.1 3  {zero one two} {
  and_merge_lists [rowid_list zero] [rowid_list one] [rowid_list two]
}
do_test_2.1 4  {"zero one" OR "one two"} {
  or_merge_lists [rowid_list "zero one"] [rowid_list "one two"]
}

finish_test