[sql] Import reference version of SQLite 3.10.2.

third_party/sqlite/sqlite-src-3100200/ext/fts5/test/fts5unicode2.test

Index: third_party/sqlite/sqlite-src-3100200/ext/fts5/test/fts5unicode2.test
diff --git a/third_party/sqlite/src/test/fts4unicode.test b/third_party/sqlite/sqlite-src-3100200/ext/fts5/test/fts5unicode2.test
similarity index 75%
copy from third_party/sqlite/src/test/fts4unicode.test
copy to third_party/sqlite/sqlite-src-3100200/ext/fts5/test/fts5unicode2.test
index f237119a1871b250da71515ad124aa2de706f19d..d3ff5128dafc0def5089a0dac74aa7c8c50a73d9 100644
--- a/third_party/sqlite/src/test/fts4unicode.test
+++ b/third_party/sqlite/sqlite-src-3100200/ext/fts5/test/fts5unicode2.test
@@ -11,69 +11,76 @@
 #
 # The tests in this file focus on testing the "unicode" FTS tokenizer.
 #
+# This is a modified copy of FTS4 test file "fts4_unicode.test".
+#
+
+source [file join [file dirname [info script]] fts5_common.tcl]
+set testprefix fts5unicode2
 
-set testdir [file dirname $argv0]
-source $testdir/tester.tcl
-ifcapable !fts3_unicode { finish_test ; return }
-set ::testprefix fts4unicode
+# If SQLITE_ENABLE_FTS5 is defined, omit this file.
+ifcapable !fts5 {
+  finish_test
+  return
+}
 
 proc do_unicode_token_test {tn input res} {
-  set input [string map {' ''} $input]
-  uplevel [list do_execsql_test $tn "
-    SELECT fts3_tokenizer_test('unicode61', 'remove_diacritics=0', '$input');
-  " [list [list {*}$res]]]
+  uplevel [list do_test $tn [list \
+    sqlite3_fts5_tokenize -subst db "unicode61 remove_diacritics 0" $input
+  ] [list {*}$res]]
 }
 
 proc do_unicode_token_test2 {tn input res} {
-  set input [string map {' ''} $input]
-  uplevel [list do_execsql_test $tn "
-    SELECT fts3_tokenizer_test('unicode61', '$input');
-  " [list [list {*}$res]]]
+  uplevel [list do_test $tn [list \
+    sqlite3_fts5_tokenize -subst db "unicode61" $input
+  ] [list {*}$res]]
 }
 
 proc do_unicode_token_test3 {tn args} {
-  set res   [lindex $args end]
-  set sql "SELECT fts3_tokenizer_test('unicode61'"
-  foreach a [lrange $args 0 end-1] {
-    append sql ", '"
-    append sql [string map {' ''} $a]
-    append sql "'"
-  }
-  append sql ")"
-  uplevel [list do_execsql_test $tn $sql [list [list {*}$res]]]
+  set tokenizer [concat unicode61 {*}[lrange $args 0 end-2]]
+  set input [lindex $args end-1]
+  set res [lindex $args end]
+  uplevel [list do_test $tn [list \
+    sqlite3_fts5_tokenize -subst db $tokenizer $input
+  ] [list {*}$res]]
 }
 
-do_unicode_token_test 1.0 {a B c D} {0 a a 1 b B 2 c c 3 d D}
+do_unicode_token_test 1.0 {a B c D} {a a b B c c d D}
 
 do_unicode_token_test 1.1 "\uC4 \uD6 \uDC" \
-    "0 \uE4 \uC4 1 \uF6 \uD6 2 \uFC \uDC"
+    "\uE4 \uC4 \uF6 \uD6 \uFC \uDC"
 
 do_unicode_token_test 1.2 "x\uC4x x\uD6x x\uDCx" \
-    "0 x\uE4x x\uC4x 1 x\uF6x x\uD6x 2 x\uFCx x\uDCx"
+    "x\uE4x x\uC4x x\uF6x x\uD6x x\uFCx x\uDCx"
 
 # 0x00DF is a small "sharp s". 0x1E9E is a capital sharp s.
-do_unicode_token_test 1.3 "\uDF" "0 \uDF \uDF"
-do_unicode_token_test 1.4 "\u1E9E" "0 \uDF \u1E9E"
+do_unicode_token_test 1.3 "\uDF" "\uDF \uDF"
+do_unicode_token_test 1.4 "\u1E9E" "\uDF \u1E9E"
 
 do_unicode_token_test 1.5 "The quick brown fox" {
-  0 the The 1 quick quick 2 brown brown 3 fox fox
+  the The quick quick brown brown fox fox
 }
 do_unicode_token_test 1.6 "The\u00bfquick\u224ebrown\u2263fox" {
-  0 the The 1 quick quick 2 brown brown 3 fox fox
+  the The quick quick brown brown fox fox
 }
 
-do_unicode_token_test2 1.7  {a B c D} {0 a a 1 b B 2 c c 3 d D}
-do_unicode_token_test2 1.8  "\uC4 \uD6 \uDC" "0 a \uC4 1 o \uD6 2 u \uDC"
+do_unicode_token_test2 1.7  {a B c D} {a a b B c c d D}
+do_unicode_token_test2 1.8  "\uC4 \uD6 \uDC" "a \uC4 o \uD6 u \uDC"
 
 do_unicode_token_test2 1.9  "x\uC4x x\uD6x x\uDCx" \
-    "0 xax x\uC4x 1 xox x\uD6x 2 xux x\uDCx"
+    "xax x\uC4x xox x\uD6x xux x\uDCx"
 
 # Check that diacritics are removed if remove_diacritics=1 is specified.
 # And that they do not break tokens.
-do_unicode_token_test2 1.10 "xx\u0301xx" "0 xxxx xx\u301xx"
+do_unicode_token_test2 1.10 "xx\u0301xx" "xxxx xx\u301xx"
 
 # Title-case mappings work
-do_unicode_token_test 1.11 "\u01c5" "0 \u01c6 \u01c5"
+do_unicode_token_test 1.11 "\u01c5" "\u01c6 \u01c5"
+
+do_unicode_token_test 1.12 "\u00C1abc\u00C2 \u00D1def\u00C3" \
+    "\u00E1abc\u00E2 \u00C1abc\u00C2 \u00F1def\u00E3 \u00D1def\u00C3"
+
+do_unicode_token_test 1.13 "\u00A2abc\u00A3 \u00A4def\u00A5" \
+    "abc abc def def"
 
 #-------------------------------------------------------------------------
 #
@@ -128,7 +135,7 @@ proc mapdoc {doc} {
 }
 
 do_test 2.0 {
-  execsql { CREATE VIRTUAL TABLE t2 USING fts4(tokenize=unicode61, x); }
+  execsql { CREATE VIRTUAL TABLE t2 USING fts5(tokenize=unicode61, x); }
   foreach doc $docs {
     set d [mapdoc $doc]
     execsql { INSERT INTO t2 VALUES($d) }
@@ -166,7 +173,9 @@ foreach {tn query snippet} {
 } {
   do_test 2.$tn {
     set q [mapdoc $query]
-    execsql { SELECT snippet(t2, '[', ']', '...') FROM t2 WHERE t2 MATCH $q }
+    execsql { 
+      SELECT snippet(t2, -1, '[', ']', '...', 15) FROM t2 WHERE t2 MATCH $q 
+    }
   } [list [mapdoc $snippet]]
 }
 
@@ -175,12 +184,12 @@ foreach {tn query snippet} {
 # NULL pointer.
 reset_db
 do_execsql_test 3.1 {
-  CREATE VIRTUAL TABLE t1 USING fts4(tokenize=unicode61, x, y);
+  CREATE VIRTUAL TABLE t1 USING fts5(tokenize=unicode61, x, y);
   INSERT INTO t1 VALUES(NULL, 'a b c');
 }
 
 do_execsql_test 3.2 {
-  SELECT snippet(t1, '[', ']') FROM t1 WHERE t1 MATCH 'b'
+  SELECT snippet(t1, -1, '[', ']', '...', 15) FROM t1 WHERE t1 MATCH 'b'
 } {{a [b] c}}
 
 do_execsql_test 3.3 {
@@ -222,12 +231,16 @@ do_test 4.1 {
   set c "\uFFFEdef"
   set d "\uD800def"
   execsql {
-    CREATE VIRTUAL TABLE t1 USING fts4(tokenize=unicode61, x);
+    CREATE VIRTUAL TABLE t1 USING fts5(tokenize=unicode61, x);
     INSERT INTO t1 VALUES($a);
     INSERT INTO t1 VALUES($b);
     INSERT INTO t1 VALUES($c);
     INSERT INTO t1 VALUES($d);
   }
+
+  execsql "CREATE VIRTUAL TABLE t8 USING fts5(
+      a, b, tokenize=\"unicode61 separators '\uFFFE\uD800\u00BF'\"
+  )"
 } {}
 
 do_test 4.2 {
@@ -256,89 +269,99 @@ do_test 4.3 {
   }
 } {}
 
+do_test 4.4 {
+  sqlite3_exec_hex db {
+    CREATE VIRTUAL TABLE t9 USING fts5(a, b, 
+      tokenize="unicode61 separators '%C09004'"
+    );
+    INSERT INTO t9(a) VALUES('abc%88def %89ghi%90');
+  }
+} {0 {}}
+
+
 #-------------------------------------------------------------------------
 
-do_unicode_token_test3 5.1 {tokenchars=} {
+breakpoint
+do_unicode_token_test3 5.1 {tokenchars {}} {
   sqlite3_reset sqlite3_column_int
 } {
-  0 sqlite3 sqlite3 
-  1 reset reset 
-  2 sqlite3 sqlite3 
-  3 column column 
-  4 int int
+  sqlite3 sqlite3 
+  reset reset 
+  sqlite3 sqlite3 
+  column column 
+  int int
 }
 
-do_unicode_token_test3 5.2 {tokenchars=_} {
+do_unicode_token_test3 5.2 {tokenchars _} {
   sqlite3_reset sqlite3_column_int
 } {
-  0 sqlite3_reset sqlite3_reset 
-  1 sqlite3_column_int sqlite3_column_int
+  sqlite3_reset sqlite3_reset 
+  sqlite3_column_int sqlite3_column_int
 }
 
-do_unicode_token_test3 5.3 {separators=xyz} {
+do_unicode_token_test3 5.3 {separators xyz} {
   Laotianxhorseyrunszfast
 } {
-  0 laotian Laotian
-  1 horse horse
-  2 runs runs
-  3 fast fast
+  laotian Laotian
+  horse horse
+  runs runs
+  fast fast
 }
 
-do_unicode_token_test3 5.4 {tokenchars=xyz} {
+do_unicode_token_test3 5.4 {tokenchars xyz} {
   Laotianxhorseyrunszfast
 } {
-  0 laotianxhorseyrunszfast Laotianxhorseyrunszfast
+  laotianxhorseyrunszfast Laotianxhorseyrunszfast
 }
 
-do_unicode_token_test3 5.5 {tokenchars=_} {separators=zyx} {
+do_unicode_token_test3 5.5 {tokenchars _} {separators zyx} {
   sqlite3_resetxsqlite3_column_intyhonda_phantom
 } {
-  0 sqlite3_reset sqlite3_reset 
-  1 sqlite3_column_int sqlite3_column_int
-  2 honda_phantom honda_phantom
+  sqlite3_reset sqlite3_reset 
+  sqlite3_column_int sqlite3_column_int
+  honda_phantom honda_phantom
 }
 
-do_unicode_token_test3 5.6 "separators=\u05D1" "abc\u05D1def" {
-  0 abc abc 1 def def
+do_unicode_token_test3 5.6 "separators \u05D1" "abc\u05D1def" {
+  abc abc def def
 }
 
 do_unicode_token_test3 5.7                             \
-  "tokenchars=\u2444\u2445"                            \
-  "separators=\u05D0\u05D1\u05D2"                      \
+  "tokenchars \u2444\u2445"                            \
+  "separators \u05D0\u05D1\u05D2"                      \
   "\u2444fre\u2445sh\u05D0water\u05D2fish.\u2445timer" \
   [list                                                \
-    0 \u2444fre\u2445sh \u2444fre\u2445sh              \
-    1 water water                                      \
-    2 fish fish                                        \
-    3 \u2445timer \u2445timer                          \
+    \u2444fre\u2445sh \u2444fre\u2445sh              \
+    water water                                      \
+    fish fish                                        \
+    \u2445timer \u2445timer                          \
   ]
 
 # Check that it is not possible to add a standalone diacritic codepoint 
 # to either separators or tokenchars.
-do_unicode_token_test3 5.8 "separators=\u0301" \
+do_unicode_token_test3 5.8 "separators \u0301" \
   "hello\u0301world \u0301helloworld"          \
-  "0 helloworld hello\u0301world 1 helloworld helloworld"
+  "helloworld hello\u0301world helloworld helloworld"
 
-do_unicode_token_test3 5.9 "tokenchars=\u0301" \
+do_unicode_token_test3 5.9 "tokenchars \u0301" \
   "hello\u0301world \u0301helloworld"          \
-  "0 helloworld hello\u0301world 1 helloworld helloworld"
+  "helloworld hello\u0301world helloworld helloworld"
 
-do_unicode_token_test3 5.10 "separators=\u0301" \
-  "remove_diacritics=0"                        \
+do_unicode_token_test3 5.10 "separators \u0301" \
+  "remove_diacritics 0"                        \
   "hello\u0301world \u0301helloworld"          \
-  "0 hello\u0301world hello\u0301world 1 helloworld helloworld"
+  "hello\u0301world hello\u0301world helloworld helloworld"
 
-do_unicode_token_test3 5.11 "tokenchars=\u0301" \
-  "remove_diacritics=0"                         \
+do_unicode_token_test3 5.11 "tokenchars \u0301" \
+  "remove_diacritics 0"                         \
   "hello\u0301world \u0301helloworld"           \
-  "0 hello\u0301world hello\u0301world 1 helloworld helloworld"
-
+  "hello\u0301world hello\u0301world helloworld helloworld"
 
 #-------------------------------------------------------------------------
 
 proc do_tokenize {tokenizer txt} {
   set res [list]
-  foreach {a b c} [db one {SELECT fts3_tokenizer_test($tokenizer, $txt)}] {
+  foreach {b c} [sqlite3_fts5_tokenize -subst db $tokenizer $txt] {
     lappend res $b
   }
   set res
@@ -357,7 +380,7 @@ proc do_isspace_test {tn tokenizer lCp} {
 }
 
 set tokenizers [list unicode61]
-ifcapable icu { lappend tokenizers icu }
+#ifcapable icu { lappend tokenizers icu }
 
 # Some tests to check that the tokenizers can both identify white-space 
 # codepoints. All codepoints tested below are of type "Zs" in the
@@ -389,6 +412,7 @@ foreach T $tokenizers {
   do_isspace_test 6.$T.23 $T   {8287 12288}
 }
 
+
 #-------------------------------------------------------------------------
 # Test that the private use ranges are treated as alphanumeric.
 #
@@ -396,8 +420,8 @@ foreach {tn1 c} {
   1 \ue000 2 \ue001 3 \uf000 4 \uf8fe 5 \uf8ff
 } {
   foreach {tn2 config res} {
-    1 ""             "0 hello*world hello*world"
-    2 "separators=*" "0 hello hello 1 world world"
+    1 ""             "hello*world hello*world"
+    2 "separators *" "hello hello world world"
   } {
     set config [string map [list * $c] $config]
     set input  [string map [list * $c] "hello*world"]
@@ -415,7 +439,9 @@ foreach {tn1 c} {
 # 00F6;LATIN SMALL LETTER O WITH DIAERESIS
 #
 do_execsql_test 8.1.1 "
-  CREATE VIRTUAL TABLE t3 USING fts4(tokenize=unicode61 'remove_diacritics=1');
+  CREATE VIRTUAL TABLE t3 USING fts5(
+    content, tokenize='unicode61 remove_diacritics 1'
+  );
   INSERT INTO t3 VALUES('o');
   INSERT INTO t3 VALUES('a');
   INSERT INTO t3 VALUES('O');
@@ -426,24 +452,27 @@ do_execsql_test 8.1.1 "
   INSERT INTO t3 VALUES('\xE4');
 "
 do_execsql_test 8.1.2 {
-  SELECT rowid FROM t3 WHERE t3 MATCH 'o';
+  SELECT rowid FROM t3 WHERE t3 MATCH 'o' ORDER BY rowid ASC;
 } {1 3 5 7}
 do_execsql_test 8.1.3 {
-  SELECT rowid FROM t3 WHERE t3 MATCH 'a';
+  SELECT rowid FROM t3 WHERE t3 MATCH 'a' ORDER BY rowid ASC;
 } {2 4 6 8}
 do_execsql_test 8.2.1 {
-  CREATE VIRTUAL TABLE t4 USING fts4(tokenize=unicode61 "remove_diacritics=0");
-  INSERT INTO t4 SELECT * FROM t3;
+  CREATE VIRTUAL TABLE t4 USING fts5(
+    content, tokenize='unicode61 remove_diacritics 0'
+  );
+  INSERT INTO t4 SELECT * FROM t3 ORDER BY rowid ASC;
 }
 do_execsql_test 8.2.2 {
-  SELECT rowid FROM t4 WHERE t4 MATCH 'o';
+  SELECT rowid FROM t4 WHERE t4 MATCH 'o' ORDER BY rowid ASC;
 } {1 3}
 do_execsql_test 8.2.3 {
-  SELECT rowid FROM t4 WHERE t4 MATCH 'a';
+  SELECT rowid FROM t4 WHERE t4 MATCH 'a' ORDER BY rowid ASC;
 } {2 4}
 
 #-------------------------------------------------------------------------
 #
+if 0 {
 foreach {tn sql} {
   1 {
     CREATE VIRTUAL TABLE t5 USING fts4(tokenize=unicode61 [tokenchars= .]);
@@ -555,4 +584,6 @@ do_execsql_test 11.1 {
   berlin@street sydney.road
 }
 
+}
+
 finish_test