# 2010 September 25 # # 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 tests to verify that the "testable statements" in # the lang_createtable.html document are correct. # set testdir [file dirname $argv0] source $testdir/tester.tcl set ::testprefix e_createtable # Test organization: # # e_createtable-0.*: Test that the syntax diagrams are correct. # # e_createtable-1.*: Test statements related to table and database names, # the TEMP and TEMPORARY keywords, and the IF NOT EXISTS clause. # # e_createtable-2.*: Test "CREATE TABLE AS" statements. # proc do_createtable_tests {nm args} { uplevel do_select_tests [list e_createtable-$nm] $args } #------------------------------------------------------------------------- # This command returns a serialized tcl array mapping from the name of # each attached database to a list of tables in that database. For example, # if the database schema is created with: # # CREATE TABLE t1(x); # CREATE TEMP TABLE t2(x); # CREATE TEMP TABLE t3(x); # # Then this command returns "main t1 temp {t2 t3}". # proc table_list {} { set res [list] db eval { pragma database_list } a { set dbname $a(name) set master $a(name).sqlite_master if {$dbname == "temp"} { set master sqlite_temp_master } lappend res $dbname [ db eval "SELECT DISTINCT tbl_name FROM $master ORDER BY tbl_name" ] } set res } # EVIDENCE-OF: R-47266-09114 -- syntax diagram type-name # do_createtable_tests 0.1.1 -repair { drop_all_tables } { 1 "CREATE TABLE t1(c1 one)" {} 2 "CREATE TABLE t1(c1 one two)" {} 3 "CREATE TABLE t1(c1 one two three)" {} 4 "CREATE TABLE t1(c1 one two three four)" {} 5 "CREATE TABLE t1(c1 one two three four(14))" {} 6 "CREATE TABLE t1(c1 one two three four(14, 22))" {} 7 "CREATE TABLE t1(c1 var(+14, -22.3))" {} 8 "CREATE TABLE t1(c1 var(1.0e10))" {} } do_createtable_tests 0.1.2 -error { near "%s": syntax error } { 1 "CREATE TABLE t1(c1 one(number))" {number} } # EVIDENCE-OF: R-60689-48779 -- syntax diagram column-constraint # do_createtable_tests 0.2.1 -repair { drop_all_tables execsql { CREATE TABLE t2(x PRIMARY KEY) } } { 1.1 "CREATE TABLE t1(c1 text PRIMARY KEY)" {} 1.2 "CREATE TABLE t1(c1 text PRIMARY KEY ASC)" {} 1.3 "CREATE TABLE t1(c1 text PRIMARY KEY DESC)" {} 1.4 "CREATE TABLE t1(c1 text CONSTRAINT cons PRIMARY KEY DESC)" {} 2.1 "CREATE TABLE t1(c1 text NOT NULL)" {} 2.2 "CREATE TABLE t1(c1 text CONSTRAINT nm NOT NULL)" {} 2.3 "CREATE TABLE t1(c1 text NULL)" {} 2.4 "CREATE TABLE t1(c1 text CONSTRAINT nm NULL)" {} 3.1 "CREATE TABLE t1(c1 text UNIQUE)" {} 3.2 "CREATE TABLE t1(c1 text CONSTRAINT un UNIQUE)" {} 4.1 "CREATE TABLE t1(c1 text CHECK(c1!=0))" {} 4.2 "CREATE TABLE t1(c1 text CONSTRAINT chk CHECK(c1!=0))" {} 5.1 "CREATE TABLE t1(c1 text DEFAULT 1)" {} 5.2 "CREATE TABLE t1(c1 text DEFAULT -1)" {} 5.3 "CREATE TABLE t1(c1 text DEFAULT +1)" {} 5.4 "CREATE TABLE t1(c1 text DEFAULT -45.8e22)" {} 5.5 "CREATE TABLE t1(c1 text DEFAULT (1+1))" {} 5.6 "CREATE TABLE t1(c1 text CONSTRAINT \"1 2\" DEFAULT (1+1))" {} 6.1 "CREATE TABLE t1(c1 text COLLATE nocase)" {} 6.2 "CREATE TABLE t1(c1 text CONSTRAINT 'a x' COLLATE nocase)" {} 7.1 "CREATE TABLE t1(c1 REFERENCES t2)" {} 7.2 "CREATE TABLE t1(c1 CONSTRAINT abc REFERENCES t2)" {} 8.1 { CREATE TABLE t1(c1 PRIMARY KEY NOT NULL UNIQUE CHECK(c1 IS 'ten') DEFAULT 123 REFERENCES t1 ); } {} 8.2 { CREATE TABLE t1(c1 REFERENCES t1 DEFAULT 123 CHECK(c1 IS 'ten') UNIQUE NOT NULL PRIMARY KEY ); } {} } # EVIDENCE-OF: R-58169-51804 -- syntax diagram table-constraint # do_createtable_tests 0.3.1 -repair { drop_all_tables execsql { CREATE TABLE t2(x PRIMARY KEY) } } { 1.1 "CREATE TABLE t1(c1, c2, PRIMARY KEY(c1))" {} 1.2 "CREATE TABLE t1(c1, c2, PRIMARY KEY(c1, c2))" {} 1.3 "CREATE TABLE t1(c1, c2, PRIMARY KEY(c1, c2) ON CONFLICT IGNORE)" {} 2.1 "CREATE TABLE t1(c1, c2, UNIQUE(c1))" {} 2.2 "CREATE TABLE t1(c1, c2, UNIQUE(c1, c2))" {} 2.3 "CREATE TABLE t1(c1, c2, UNIQUE(c1, c2) ON CONFLICT IGNORE)" {} 3.1 "CREATE TABLE t1(c1, c2, CHECK(c1 IS NOT c2))" {} 4.1 "CREATE TABLE t1(c1, c2, FOREIGN KEY(c1) REFERENCES t2)" {} } # EVIDENCE-OF: R-44826-22243 -- syntax diagram column-def # do_createtable_tests 0.4.1 -repair { drop_all_tables } { 1 {CREATE TABLE t1( col1, col2 TEXT, col3 INTEGER UNIQUE, col4 VARCHAR(10, 10) PRIMARY KEY, "name with spaces" REFERENCES t1 ); } {} } # EVIDENCE-OF: R-45698-45677 -- syntax diagram create-table-stmt # do_createtable_tests 0.5.1 -repair { drop_all_tables execsql { CREATE TABLE t2(a, b, c) } } { 1 "CREATE TABLE t1(a, b, c)" {} 2 "CREATE TEMP TABLE t1(a, b, c)" {} 3 "CREATE TEMPORARY TABLE t1(a, b, c)" {} 4 "CREATE TABLE IF NOT EXISTS t1(a, b, c)" {} 5 "CREATE TEMP TABLE IF NOT EXISTS t1(a, b, c)" {} 6 "CREATE TEMPORARY TABLE IF NOT EXISTS t1(a, b, c)" {} 7 "CREATE TABLE main.t1(a, b, c)" {} 8 "CREATE TEMP TABLE temp.t1(a, b, c)" {} 9 "CREATE TEMPORARY TABLE temp.t1(a, b, c)" {} 10 "CREATE TABLE IF NOT EXISTS main.t1(a, b, c)" {} 11 "CREATE TEMP TABLE IF NOT EXISTS temp.t1(a, b, c)" {} 12 "CREATE TEMPORARY TABLE IF NOT EXISTS temp.t1(a, b, c)" {} 13 "CREATE TABLE t1 AS SELECT * FROM t2" {} 14 "CREATE TEMP TABLE t1 AS SELECT c, b, a FROM t2" {} 15 "CREATE TABLE t1 AS SELECT count(*), max(b), min(a) FROM t2" {} } # EVIDENCE-OF: R-24369-11919 -- syntax diagram foreign-key-clause # # 1: Explicit parent-key columns. # 2: Implicit child-key columns. # # 1: MATCH FULL # 2: MATCH PARTIAL # 3: MATCH SIMPLE # 4: MATCH STICK # 5: # # 1: ON DELETE SET NULL # 2: ON DELETE SET DEFAULT # 3: ON DELETE CASCADE # 4: ON DELETE RESTRICT # 5: ON DELETE NO ACTION # 6: # # 1: ON UPDATE SET NULL # 2: ON UPDATE SET DEFAULT # 3: ON UPDATE CASCADE # 4: ON UPDATE RESTRICT # 5: ON UPDATE NO ACTION # 6: # # 1: NOT DEFERRABLE INITIALLY DEFERRED # 2: NOT DEFERRABLE INITIALLY IMMEDIATE # 3: NOT DEFERRABLE # 4: DEFERRABLE INITIALLY DEFERRED # 5: DEFERRABLE INITIALLY IMMEDIATE # 6: DEFERRABLE # 7: # do_createtable_tests 0.6.1 -repair { drop_all_tables execsql { CREATE TABLE t2(x PRIMARY KEY, y) } execsql { CREATE TABLE t3(i, j, UNIQUE(i, j) ) } } { 11146 { CREATE TABLE t1(a REFERENCES t2(x) MATCH FULL ON DELETE SET NULL ON UPDATE RESTRICT DEFERRABLE )} {} 11412 { CREATE TABLE t1(a REFERENCES t2(x) ON DELETE RESTRICT ON UPDATE SET NULL MATCH FULL NOT DEFERRABLE INITIALLY IMMEDIATE )} {} 12135 { CREATE TABLE t1(a REFERENCES t2(x) MATCH PARTIAL ON DELETE SET NULL ON UPDATE CASCADE DEFERRABLE INITIALLY IMMEDIATE )} {} 12427 { CREATE TABLE t1(a REFERENCES t2(x) MATCH PARTIAL ON DELETE RESTRICT ON UPDATE SET DEFAULT )} {} 12446 { CREATE TABLE t1(a REFERENCES t2(x) MATCH PARTIAL ON DELETE RESTRICT ON UPDATE RESTRICT DEFERRABLE )} {} 12522 { CREATE TABLE t1(a REFERENCES t2(x) MATCH PARTIAL ON DELETE NO ACTION ON UPDATE SET DEFAULT NOT DEFERRABLE INITIALLY IMMEDIATE )} {} 13133 { CREATE TABLE t1(a REFERENCES t2(x) MATCH SIMPLE ON DELETE SET NULL ON UPDATE CASCADE NOT DEFERRABLE )} {} 13216 { CREATE TABLE t1(a REFERENCES t2(x) MATCH SIMPLE ON DELETE SET DEFAULT ON UPDATE SET NULL DEFERRABLE )} {} 13263 { CREATE TABLE t1(a REFERENCES t2(x) MATCH SIMPLE ON DELETE SET DEFAULT NOT DEFERRABLE )} {} 13421 { CREATE TABLE t1(a REFERENCES t2(x) MATCH SIMPLE ON DELETE RESTRICT ON UPDATE SET DEFAULT NOT DEFERRABLE INITIALLY DEFERRED )} {} 13432 { CREATE TABLE t1(a REFERENCES t2(x) MATCH SIMPLE ON DELETE RESTRICT ON UPDATE CASCADE NOT DEFERRABLE INITIALLY IMMEDIATE )} {} 13523 { CREATE TABLE t1(a REFERENCES t2(x) MATCH SIMPLE ON DELETE NO ACTION ON UPDATE SET DEFAULT NOT DEFERRABLE )} {} 14336 { CREATE TABLE t1(a REFERENCES t2(x) MATCH STICK ON DELETE CASCADE ON UPDATE CASCADE DEFERRABLE )} {} 14611 { CREATE TABLE t1(a REFERENCES t2(x) MATCH STICK ON UPDATE SET NULL NOT DEFERRABLE INITIALLY DEFERRED )} {} 15155 { CREATE TABLE t1(a REFERENCES t2(x) ON DELETE SET NULL ON UPDATE NO ACTION DEFERRABLE INITIALLY IMMEDIATE )} {} 15453 { CREATE TABLE t1(a REFERENCES t2(x) ON DELETE RESTRICT ON UPDATE NO ACTION NOT DEFERRABLE )} {} 15661 { CREATE TABLE t1(a REFERENCES t2(x) NOT DEFERRABLE INITIALLY DEFERRED )} {} 21115 { CREATE TABLE t1(a REFERENCES t2 MATCH FULL ON DELETE SET NULL ON UPDATE SET NULL DEFERRABLE INITIALLY IMMEDIATE )} {} 21123 { CREATE TABLE t1(a REFERENCES t2 MATCH FULL ON DELETE SET NULL ON UPDATE SET DEFAULT NOT DEFERRABLE )} {} 21217 { CREATE TABLE t1(a REFERENCES t2 MATCH FULL ON DELETE SET DEFAULT ON UPDATE SET NULL )} {} 21362 { CREATE TABLE t1(a REFERENCES t2 MATCH FULL ON DELETE CASCADE NOT DEFERRABLE INITIALLY IMMEDIATE )} {} 22143 { CREATE TABLE t1(a REFERENCES t2 MATCH PARTIAL ON DELETE SET NULL ON UPDATE RESTRICT NOT DEFERRABLE )} {} 22156 { CREATE TABLE t1(a REFERENCES t2 MATCH PARTIAL ON DELETE SET NULL ON UPDATE NO ACTION DEFERRABLE )} {} 22327 { CREATE TABLE t1(a REFERENCES t2 MATCH PARTIAL ON DELETE CASCADE ON UPDATE SET DEFAULT )} {} 22663 { CREATE TABLE t1(a REFERENCES t2 MATCH PARTIAL NOT DEFERRABLE )} {} 23236 { CREATE TABLE t1(a REFERENCES t2 MATCH SIMPLE ON DELETE SET DEFAULT ON UPDATE CASCADE DEFERRABLE )} {} 24155 { CREATE TABLE t1(a REFERENCES t2 MATCH STICK ON DELETE SET NULL ON UPDATE NO ACTION DEFERRABLE INITIALLY IMMEDIATE )} {} 24522 { CREATE TABLE t1(a REFERENCES t2 MATCH STICK ON DELETE NO ACTION ON UPDATE SET DEFAULT NOT DEFERRABLE INITIALLY IMMEDIATE )} {} 24625 { CREATE TABLE t1(a REFERENCES t2 MATCH STICK ON UPDATE SET DEFAULT DEFERRABLE INITIALLY IMMEDIATE )} {} 25454 { CREATE TABLE t1(a REFERENCES t2 ON DELETE RESTRICT ON UPDATE NO ACTION DEFERRABLE INITIALLY DEFERRED )} {} } #------------------------------------------------------------------------- # Test cases e_createtable-1.* - test statements related to table and # database names, the TEMP and TEMPORARY keywords, and the IF NOT EXISTS # clause. # drop_all_tables forcedelete test.db2 test.db3 do_execsql_test e_createtable-1.0 { ATTACH 'test.db2' AS auxa; ATTACH 'test.db3' AS auxb; } {} # EVIDENCE-OF: R-17899-04554 Table names that begin with "sqlite_" are # reserved for internal use. It is an error to attempt to create a table # with a name that starts with "sqlite_". # do_createtable_tests 1.1.1 -error { object name reserved for internal use: %s } { 1 "CREATE TABLE sqlite_abc(a, b, c)" sqlite_abc 2 "CREATE TABLE temp.sqlite_helloworld(x)" sqlite_helloworld 3 {CREATE TABLE auxa."sqlite__"(x, y)} sqlite__ 4 {CREATE TABLE auxb."sqlite_"(z)} sqlite_ 5 {CREATE TABLE "SQLITE_TBL"(z)} SQLITE_TBL } do_createtable_tests 1.1.2 { 1 "CREATE TABLE sqlit_abc(a, b, c)" {} 2 "CREATE TABLE temp.sqlitehelloworld(x)" {} 3 {CREATE TABLE auxa."sqlite"(x, y)} {} 4 {CREATE TABLE auxb."sqlite-"(z)} {} 5 {CREATE TABLE "SQLITE-TBL"(z)} {} } # EVIDENCE-OF: R-10195-31023 If a is specified, it # must be either "main", "temp", or the name of an attached database. # # EVIDENCE-OF: R-39822-07822 In this case the new table is created in # the named database. # # Test cases 1.2.* test the first of the two requirements above. The # second is verified by cases 1.3.*. # do_createtable_tests 1.2.1 -error { unknown database %s } { 1 "CREATE TABLE george.t1(a, b)" george 2 "CREATE TABLE _.t1(a, b)" _ } do_createtable_tests 1.2.2 { 1 "CREATE TABLE main.abc(a, b, c)" {} 2 "CREATE TABLE temp.helloworld(x)" {} 3 {CREATE TABLE auxa."t 1"(x, y)} {} 4 {CREATE TABLE auxb.xyz(z)} {} } drop_all_tables do_createtable_tests 1.3 -tclquery { unset -nocomplain X array set X [table_list] list $X(main) $X(temp) $X(auxa) $X(auxb) } { 1 "CREATE TABLE main.abc(a, b, c)" {abc {} {} {}} 2 "CREATE TABLE main.t1(a, b, c)" {{abc t1} {} {} {}} 3 "CREATE TABLE temp.tmp(a, b, c)" {{abc t1} tmp {} {}} 4 "CREATE TABLE auxb.tbl(x, y)" {{abc t1} tmp {} tbl} 5 "CREATE TABLE auxb.t1(k, v)" {{abc t1} tmp {} {t1 tbl}} 6 "CREATE TABLE auxa.next(c, d)" {{abc t1} tmp next {t1 tbl}} } # EVIDENCE-OF: R-18895-27365 If the "TEMP" or "TEMPORARY" keyword occurs # between the "CREATE" and "TABLE" then the new table is created in the # temp database. # drop_all_tables do_createtable_tests 1.4 -tclquery { unset -nocomplain X array set X [table_list] list $X(main) $X(temp) $X(auxa) $X(auxb) } { 1 "CREATE TEMP TABLE t1(a, b)" {{} t1 {} {}} 2 "CREATE TEMPORARY TABLE t2(a, b)" {{} {t1 t2} {} {}} } # EVIDENCE-OF: R-49439-47561 It is an error to specify both a # and the TEMP or TEMPORARY keyword, unless the # is "temp". # drop_all_tables do_createtable_tests 1.5.1 -error { temporary table name must be unqualified } { 1 "CREATE TEMP TABLE main.t1(a, b)" {} 2 "CREATE TEMPORARY TABLE auxa.t2(a, b)" {} 3 "CREATE TEMP TABLE auxb.t3(a, b)" {} 4 "CREATE TEMPORARY TABLE main.xxx(x)" {} } drop_all_tables do_createtable_tests 1.5.2 -tclquery { unset -nocomplain X array set X [table_list] list $X(main) $X(temp) $X(auxa) $X(auxb) } { 1 "CREATE TEMP TABLE temp.t1(a, b)" {{} t1 {} {}} 2 "CREATE TEMPORARY TABLE temp.t2(a, b)" {{} {t1 t2} {} {}} 3 "CREATE TEMP TABLE TEMP.t3(a, b)" {{} {t1 t2 t3} {} {}} 4 "CREATE TEMPORARY TABLE TEMP.xxx(x)" {{} {t1 t2 t3 xxx} {} {}} } # EVIDENCE-OF: R-00917-09393 If no database name is specified and the # TEMP keyword is not present then the table is created in the main # database. # drop_all_tables do_createtable_tests 1.6 -tclquery { unset -nocomplain X array set X [table_list] list $X(main) $X(temp) $X(auxa) $X(auxb) } { 1 "CREATE TABLE t1(a, b)" {t1 {} {} {}} 2 "CREATE TABLE t2(a, b)" {{t1 t2} {} {} {}} 3 "CREATE TABLE t3(a, b)" {{t1 t2 t3} {} {} {}} 4 "CREATE TABLE xxx(x)" {{t1 t2 t3 xxx} {} {} {}} } drop_all_tables do_execsql_test e_createtable-1.7.0 { CREATE TABLE t1(x, y); CREATE INDEX i1 ON t1(x); CREATE VIEW v1 AS SELECT * FROM t1; CREATE TABLE auxa.tbl1(x, y); CREATE INDEX auxa.idx1 ON tbl1(x); CREATE VIEW auxa.view1 AS SELECT * FROM tbl1; } {} # EVIDENCE-OF: R-01232-54838 It is usually an error to attempt to create # a new table in a database that already contains a table, index or view # of the same name. # # Test cases 1.7.1.* verify that creating a table in a database with a # table/index/view of the same name does fail. 1.7.2.* tests that creating # a table with the same name as a table/index/view in a different database # is Ok. # do_createtable_tests 1.7.1 -error { %s } { 1 "CREATE TABLE t1(a, b)" {{table t1 already exists}} 2 "CREATE TABLE i1(a, b)" {{there is already an index named i1}} 3 "CREATE TABLE v1(a, b)" {{table v1 already exists}} 4 "CREATE TABLE auxa.tbl1(a, b)" {{table tbl1 already exists}} 5 "CREATE TABLE auxa.idx1(a, b)" {{there is already an index named idx1}} 6 "CREATE TABLE auxa.view1(a, b)" {{table view1 already exists}} } do_createtable_tests 1.7.2 { 1 "CREATE TABLE auxa.t1(a, b)" {} 2 "CREATE TABLE auxa.i1(a, b)" {} 3 "CREATE TABLE auxa.v1(a, b)" {} 4 "CREATE TABLE tbl1(a, b)" {} 5 "CREATE TABLE idx1(a, b)" {} 6 "CREATE TABLE view1(a, b)" {} } # EVIDENCE-OF: R-33917-24086 However, if the "IF NOT EXISTS" clause is # specified as part of the CREATE TABLE statement and a table or view of # the same name already exists, the CREATE TABLE command simply has no # effect (and no error message is returned). # drop_all_tables do_execsql_test e_createtable-1.8.0 { CREATE TABLE t1(x, y); CREATE INDEX i1 ON t1(x); CREATE VIEW v1 AS SELECT * FROM t1; CREATE TABLE auxa.tbl1(x, y); CREATE INDEX auxa.idx1 ON tbl1(x); CREATE VIEW auxa.view1 AS SELECT * FROM tbl1; } {} do_createtable_tests 1.8 { 1 "CREATE TABLE IF NOT EXISTS t1(a, b)" {} 2 "CREATE TABLE IF NOT EXISTS auxa.tbl1(a, b)" {} 3 "CREATE TABLE IF NOT EXISTS v1(a, b)" {} 4 "CREATE TABLE IF NOT EXISTS auxa.view1(a, b)" {} } # EVIDENCE-OF: R-16465-40078 An error is still returned if the table # cannot be created because of an existing index, even if the "IF NOT # EXISTS" clause is specified. # do_createtable_tests 1.9 -error { %s } { 1 "CREATE TABLE IF NOT EXISTS i1(a, b)" {{there is already an index named i1}} 2 "CREATE TABLE IF NOT EXISTS auxa.idx1(a, b)" {{there is already an index named idx1}} } # EVIDENCE-OF: R-05513-33819 It is not an error to create a table that # has the same name as an existing trigger. # drop_all_tables do_execsql_test e_createtable-1.10.0 { CREATE TABLE t1(x, y); CREATE TABLE auxb.t2(x, y); CREATE TRIGGER tr1 AFTER INSERT ON t1 BEGIN SELECT 1; END; CREATE TRIGGER auxb.tr2 AFTER INSERT ON t2 BEGIN SELECT 1; END; } {} do_createtable_tests 1.10 { 1 "CREATE TABLE tr1(a, b)" {} 2 "CREATE TABLE tr2(a, b)" {} 3 "CREATE TABLE auxb.tr1(a, b)" {} 4 "CREATE TABLE auxb.tr2(a, b)" {} } # EVIDENCE-OF: R-22283-14179 Tables are removed using the DROP TABLE # statement. # drop_all_tables do_execsql_test e_createtable-1.11.0 { CREATE TABLE t1(a, b); CREATE TABLE t2(a, b); CREATE TABLE auxa.t3(a, b); CREATE TABLE auxa.t4(a, b); } {} do_execsql_test e_createtable-1.11.1.1 { SELECT * FROM t1; SELECT * FROM t2; SELECT * FROM t3; SELECT * FROM t4; } {} do_execsql_test e_createtable-1.11.1.2 { DROP TABLE t1 } {} do_catchsql_test e_createtable-1.11.1.3 { SELECT * FROM t1 } {1 {no such table: t1}} do_execsql_test e_createtable-1.11.1.4 { DROP TABLE t3 } {} do_catchsql_test e_createtable-1.11.1.5 { SELECT * FROM t3 } {1 {no such table: t3}} do_execsql_test e_createtable-1.11.2.1 { SELECT name FROM sqlite_master; SELECT name FROM auxa.sqlite_master; } {t2 t4} do_execsql_test e_createtable-1.11.2.2 { DROP TABLE t2 } {} do_execsql_test e_createtable-1.11.2.3 { DROP TABLE t4 } {} do_execsql_test e_createtable-1.11.2.4 { SELECT name FROM sqlite_master; SELECT name FROM auxa.sqlite_master; } {} #------------------------------------------------------------------------- # Test cases e_createtable-2.* - test statements related to the CREATE # TABLE AS ... SELECT statement. # # Three Tcl commands: # # select_column_names SQL # The argument must be a SELECT statement. Return a list of the names # of the columns of the result-set that would be returned by executing # the SELECT. # # table_column_names TBL # The argument must be a table name. Return a list of column names, from # left to right, for the table. # # table_column_decltypes TBL # The argument must be a table name. Return a list of column declared # types, from left to right, for the table. # proc sci {select cmd} { set res [list] set STMT [sqlite3_prepare_v2 db $select -1 dummy] for {set i 0} {$i < [sqlite3_column_count $STMT]} {incr i} { lappend res [$cmd $STMT $i] } sqlite3_finalize $STMT set res } proc tci {tbl cmd} { sci "SELECT * FROM $tbl" $cmd } proc select_column_names {sql} { sci $sql sqlite3_column_name } proc table_column_names {tbl} { tci $tbl sqlite3_column_name } proc table_column_decltypes {tbl} { tci $tbl sqlite3_column_decltype } # Create a database schema. This schema is used by tests 2.1.* through 2.3.*. # drop_all_tables do_execsql_test e_createtable-2.0 { CREATE TABLE t1(a, b, c); CREATE TABLE t2(d, e, f); CREATE TABLE t3(g BIGINT, h VARCHAR(10)); CREATE TABLE t4(i BLOB, j ANYOLDATA); CREATE TABLE t5(k FLOAT, l INTEGER); CREATE TABLE t6(m DEFAULT 10, n DEFAULT 5, PRIMARY KEY(m, n)); CREATE TABLE t7(x INTEGER PRIMARY KEY); CREATE TABLE t8(o COLLATE nocase DEFAULT 'abc'); CREATE TABLE t9(p NOT NULL, q DOUBLE CHECK (q!=0), r STRING UNIQUE); } {} # EVIDENCE-OF: R-64828-59568 The table has the same number of columns as # the rows returned by the SELECT statement. The name of each column is # the same as the name of the corresponding column in the result set of # the SELECT statement. # do_createtable_tests 2.1 -tclquery { table_column_names x1 } -repair { catchsql { DROP TABLE x1 } } { 1 "CREATE TABLE x1 AS SELECT * FROM t1" {a b c} 2 "CREATE TABLE x1 AS SELECT c, b, a FROM t1" {c b a} 3 "CREATE TABLE x1 AS SELECT * FROM t1, t2" {a b c d e f} 4 "CREATE TABLE x1 AS SELECT count(*) FROM t1" {count(*)} 5 "CREATE TABLE x1 AS SELECT count(a) AS a, max(b) FROM t1" {a max(b)} } # EVIDENCE-OF: R-37111-22855 The declared type of each column is # determined by the expression affinity of the corresponding expression # in the result set of the SELECT statement, as follows: Expression # Affinity Column Declared Type TEXT "TEXT" NUMERIC "NUM" INTEGER "INT" # REAL "REAL" NONE "" (empty string) # do_createtable_tests 2.2 -tclquery { table_column_decltypes x1 } -repair { catchsql { DROP TABLE x1 } } { 1 "CREATE TABLE x1 AS SELECT a FROM t1" {""} 2 "CREATE TABLE x1 AS SELECT * FROM t3" {INT TEXT} 3 "CREATE TABLE x1 AS SELECT * FROM t4" {"" NUM} 4 "CREATE TABLE x1 AS SELECT * FROM t5" {REAL INT} } # EVIDENCE-OF: R-16667-09772 A table created using CREATE TABLE AS has # no PRIMARY KEY and no constraints of any kind. The default value of # each column is NULL. The default collation sequence for each column of # the new table is BINARY. # # The following tests create tables based on SELECT statements that read # from tables that have primary keys, constraints and explicit default # collation sequences. None of this is transfered to the definition of # the new table as stored in the sqlite_master table. # # Tests 2.3.2.* show that the default value of each column is NULL. # do_createtable_tests 2.3.1 -query { SELECT sql FROM sqlite_master ORDER BY rowid DESC LIMIT 1 } { 1 "CREATE TABLE x1 AS SELECT * FROM t6" {{CREATE TABLE x1(m,n)}} 2 "CREATE TABLE x2 AS SELECT * FROM t7" {{CREATE TABLE x2(x INT)}} 3 "CREATE TABLE x3 AS SELECT * FROM t8" {{CREATE TABLE x3(o)}} 4 "CREATE TABLE x4 AS SELECT * FROM t9" {{CREATE TABLE x4(p,q REAL,r NUM)}} } do_execsql_test e_createtable-2.3.2.1 { INSERT INTO x1 DEFAULT VALUES; INSERT INTO x2 DEFAULT VALUES; INSERT INTO x3 DEFAULT VALUES; INSERT INTO x4 DEFAULT VALUES; } {} db nullvalue null do_execsql_test e_createtable-2.3.2.2 { SELECT * FROM x1 } {null null} do_execsql_test e_createtable-2.3.2.3 { SELECT * FROM x2 } {null} do_execsql_test e_createtable-2.3.2.4 { SELECT * FROM x3 } {null} do_execsql_test e_createtable-2.3.2.5 { SELECT * FROM x4 } {null null null} db nullvalue {} drop_all_tables do_execsql_test e_createtable-2.4.0 { CREATE TABLE t1(x, y); INSERT INTO t1 VALUES('i', 'one'); INSERT INTO t1 VALUES('ii', 'two'); INSERT INTO t1 VALUES('iii', 'three'); } {} # EVIDENCE-OF: R-24153-28352 Tables created using CREATE TABLE AS are # initially populated with the rows of data returned by the SELECT # statement. # # EVIDENCE-OF: R-08224-30249 Rows are assigned contiguously ascending # rowid values, starting with 1, in the order that they are returned by # the SELECT statement. # # Each test case below is specified as the name of a table to create # using "CREATE TABLE ... AS SELECT ..." and a SELECT statement to use in # creating it. The table is created. # # Test cases 2.4.*.1 check that after it has been created, the data in the # table is the same as the data returned by the SELECT statement executed as # a standalone command, verifying the first testable statement above. # # Test cases 2.4.*.2 check that the rowids were allocated contiguously # as required by the second testable statement above. That the rowids # from the contiguous block were allocated to rows in the order rows are # returned by the SELECT statement is verified by 2.4.*.1. # # EVIDENCE-OF: R-32365-09043 A "CREATE TABLE ... AS SELECT" statement # creates and populates a database table based on the results of a # SELECT statement. # # The above is also considered to be tested by the following. It is # clear that tables are being created and populated by the command in # question. # foreach {tn tbl select} { 1 x1 "SELECT * FROM t1" 2 x2 "SELECT * FROM t1 ORDER BY x DESC" 3 x3 "SELECT * FROM t1 ORDER BY x ASC" } { # Create the table using a "CREATE TABLE ... AS SELECT ..." command. execsql [subst {CREATE TABLE $tbl AS $select}] # Check that the rows inserted into the table, sorted in ascending rowid # order, match those returned by executing the SELECT statement as a # standalone command. do_execsql_test e_createtable-2.4.$tn.1 [subst { SELECT * FROM $tbl ORDER BY rowid; }] [execsql $select] # Check that the rowids in the new table are a contiguous block starting # with rowid 1. Note that this will fail if SELECT statement $select # returns 0 rows (as max(rowid) will be NULL). do_execsql_test e_createtable-2.4.$tn.2 [subst { SELECT min(rowid), count(rowid)==max(rowid) FROM $tbl }] {1 1} } #-------------------------------------------------------------------------- # Test cases for column defintions in CREATE TABLE statements that do not # use a SELECT statement. Not including data constraints. In other words, # tests for the specification of: # # * declared types, # * default values, and # * default collation sequences. # # EVIDENCE-OF: R-27219-49057 Unlike most SQL databases, SQLite does not # restrict the type of data that may be inserted into a column based on # the columns declared type. # # Test this by creating a few tables with varied declared types, then # inserting various different types of values into them. # drop_all_tables do_execsql_test e_createtable-3.1.0 { CREATE TABLE t1(x VARCHAR(10), y INTEGER, z DOUBLE); CREATE TABLE t2(a DATETIME, b STRING, c REAL); CREATE TABLE t3(o, t); } {} # value type -> declared column type # ---------------------------------- # integer -> VARCHAR(10) # string -> INTEGER # blob -> DOUBLE # do_execsql_test e_createtable-3.1.1 { INSERT INTO t1 VALUES(14, 'quite a lengthy string', X'555655'); SELECT * FROM t1; } {14 {quite a lengthy string} UVU} # string -> DATETIME # integer -> STRING # time -> REAL # do_execsql_test e_createtable-3.1.2 { INSERT INTO t2 VALUES('not a datetime', 13, '12:41:59'); SELECT * FROM t2; } {{not a datetime} 13 12:41:59} # EVIDENCE-OF: R-10565-09557 The declared type of a column is used to # determine the affinity of the column only. # # Affinities are tested in more detail elsewhere (see document # datatype3.html). Here, just test that affinity transformations # consistent with the expected affinity of each column (based on # the declared type) appear to take place. # # Affinities of t1 (test cases 3.2.1.*): TEXT, INTEGER, REAL # Affinities of t2 (test cases 3.2.2.*): NUMERIC, NUMERIC, REAL # Affinities of t3 (test cases 3.2.3.*): NONE, NONE # do_execsql_test e_createtable-3.2.0 { DELETE FROM t1; DELETE FROM t2; } {} do_createtable_tests 3.2.1 -query { SELECT quote(x), quote(y), quote(z) FROM t1 ORDER BY rowid DESC LIMIT 1; } { 1 "INSERT INTO t1 VALUES(15, '22.0', '14')" {'15' 22 14.0} 2 "INSERT INTO t1 VALUES(22.0, 22.0, 22.0)" {'22.0' 22 22.0} } do_createtable_tests 3.2.2 -query { SELECT quote(a), quote(b), quote(c) FROM t2 ORDER BY rowid DESC LIMIT 1; } { 1 "INSERT INTO t2 VALUES(15, '22.0', '14')" {15 22 14.0} 2 "INSERT INTO t2 VALUES(22.0, 22.0, 22.0)" {22 22 22.0} } do_createtable_tests 3.2.3 -query { SELECT quote(o), quote(t) FROM t3 ORDER BY rowid DESC LIMIT 1; } { 1 "INSERT INTO t3 VALUES('15', '22.0')" {'15' '22.0'} 2 "INSERT INTO t3 VALUES(15, 22.0)" {15 22.0} } # EVIDENCE-OF: R-42316-09582 If there is no explicit DEFAULT clause # attached to a column definition, then the default value of the column # is NULL. # # None of the columns in table t1 have an explicit DEFAULT clause. # So testing that the default value of all columns in table t1 is # NULL serves to verify the above. # do_createtable_tests 3.2.3 -query { SELECT quote(x), quote(y), quote(z) FROM t1 } -repair { execsql { DELETE FROM t1 } } { 1 "INSERT INTO t1(x, y) VALUES('abc', 'xyz')" {'abc' 'xyz' NULL} 2 "INSERT INTO t1(x, z) VALUES('abc', 'xyz')" {'abc' NULL 'xyz'} 3 "INSERT INTO t1 DEFAULT VALUES" {NULL NULL NULL} } # EVIDENCE-OF: R-62940-43005 An explicit DEFAULT clause may specify that # the default value is NULL, a string constant, a blob constant, a # signed-number, or any constant expression enclosed in parentheses. An # explicit default value may also be one of the special case-independent # keywords CURRENT_TIME, CURRENT_DATE or CURRENT_TIMESTAMP. # do_execsql_test e_createtable-3.3.1 { CREATE TABLE t4( a DEFAULT NULL, b DEFAULT 'string constant', c DEFAULT X'424C4F42', d DEFAULT 1, e DEFAULT -1, f DEFAULT 3.14, g DEFAULT -3.14, h DEFAULT ( substr('abcd', 0, 2) || 'cd' ), i DEFAULT CURRENT_TIME, j DEFAULT CURRENT_DATE, k DEFAULT CURRENT_TIMESTAMP ); } {} # EVIDENCE-OF: R-10288-43169 For the purposes of the DEFAULT clause, an # expression is considered constant provided that it does not contain # any sub-queries or string constants enclosed in double quotes. # do_createtable_tests 3.4.1 -error { default value of column [x] is not constant } { 1 {CREATE TABLE t5(x DEFAULT ( (SELECT 1) ))} {} 2 {CREATE TABLE t5(x DEFAULT ( "abc" ))} {} 3 {CREATE TABLE t5(x DEFAULT ( 1 IN (SELECT 1) ))} {} 4 {CREATE TABLE t5(x DEFAULT ( EXISTS (SELECT 1) ))} {} } do_createtable_tests 3.4.2 -repair { catchsql { DROP TABLE t5 } } { 1 {CREATE TABLE t5(x DEFAULT ( 'abc' ))} {} 2 {CREATE TABLE t5(x DEFAULT ( 1 IN (1, 2, 3) ))} {} } # EVIDENCE-OF: R-18814-23501 Each time a row is inserted into the table # by an INSERT statement that does not provide explicit values for all # table columns the values stored in the new row are determined by their # default values # # Verify this with some assert statements for which all, some and no # columns lack explicit values. # set sqlite_current_time 1000000000 do_createtable_tests 3.5 -query { SELECT quote(a), quote(b), quote(c), quote(d), quote(e), quote(f), quote(g), quote(h), quote(i), quote(j), quote(k) FROM t4 ORDER BY rowid DESC LIMIT 1; } { 1 "INSERT INTO t4 DEFAULT VALUES" { NULL {'string constant'} X'424C4F42' 1 -1 3.14 -3.14 'acd' '01:46:40' '2001-09-09' {'2001-09-09 01:46:40'} } 2 "INSERT INTO t4(a, b, c) VALUES(1, 2, 3)" { 1 2 3 1 -1 3.14 -3.14 'acd' '01:46:40' '2001-09-09' {'2001-09-09 01:46:40'} } 3 "INSERT INTO t4(k, j, i) VALUES(1, 2, 3)" { NULL {'string constant'} X'424C4F42' 1 -1 3.14 -3.14 'acd' 3 2 1 } 4 "INSERT INTO t4(a,b,c,d,e,f,g,h,i,j,k) VALUES(1,2,3,4,5,6,7,8,9,10,11)" { 1 2 3 4 5 6 7 8 9 10 11 } } # EVIDENCE-OF: R-12572-62501 If the default value of the column is a # constant NULL, text, blob or signed-number value, then that value is # used directly in the new row. # do_execsql_test e_createtable-3.6.1 { CREATE TABLE t5( a DEFAULT NULL, b DEFAULT 'text value', c DEFAULT X'424C4F42', d DEFAULT -45678.6, e DEFAULT 394507 ); } {} do_execsql_test e_createtable-3.6.2 { INSERT INTO t5 DEFAULT VALUES; SELECT quote(a), quote(b), quote(c), quote(d), quote(e) FROM t5; } {NULL {'text value'} X'424C4F42' -45678.6 394507} # EVIDENCE-OF: R-60616-50251 If the default value of a column is an # expression in parentheses, then the expression is evaluated once for # each row inserted and the results used in the new row. # # Test case 3.6.4 demonstrates that the expression is evaluated # separately for each row if the INSERT is an "INSERT INTO ... SELECT ..." # command. # set ::nextint 0 proc nextint {} { incr ::nextint } db func nextint nextint do_execsql_test e_createtable-3.7.1 { CREATE TABLE t6(a DEFAULT ( nextint() ), b DEFAULT ( nextint() )); } {} do_execsql_test e_createtable-3.7.2 { INSERT INTO t6 DEFAULT VALUES; SELECT quote(a), quote(b) FROM t6; } {1 2} do_execsql_test e_createtable-3.7.3 { INSERT INTO t6(a) VALUES('X'); SELECT quote(a), quote(b) FROM t6; } {1 2 'X' 3} do_execsql_test e_createtable-3.7.4 { INSERT INTO t6(a) SELECT a FROM t6; SELECT quote(a), quote(b) FROM t6; } {1 2 'X' 3 1 4 'X' 5} # EVIDENCE-OF: R-15363-55230 If the default value of a column is # CURRENT_TIME, CURRENT_DATE or CURRENT_TIMESTAMP, then the value used # in the new row is a text representation of the current UTC date and/or # time. # # This is difficult to test literally without knowing what time the # user will run the tests. Instead, we test that the three cases # above set the value to the current date and/or time according to # the xCurrentTime() method of the VFS. Which is usually the same # as UTC. In this case, however, we instrument it to always return # a time equivalent to "2001-09-09 01:46:40 UTC". # set sqlite_current_time 1000000000 do_execsql_test e_createtable-3.8.1 { CREATE TABLE t7( a DEFAULT CURRENT_TIME, b DEFAULT CURRENT_DATE, c DEFAULT CURRENT_TIMESTAMP ); } {} do_execsql_test e_createtable-3.8.2 { INSERT INTO t7 DEFAULT VALUES; SELECT quote(a), quote(b), quote(c) FROM t7; } {'01:46:40' '2001-09-09' {'2001-09-09 01:46:40'}} # EVIDENCE-OF: R-62327-53843 For CURRENT_TIME, the format of the value # is "HH:MM:SS". # # EVIDENCE-OF: R-03775-43471 For CURRENT_DATE, "YYYY-MM-DD". # # EVIDENCE-OF: R-07677-44926 The format for CURRENT_TIMESTAMP is # "YYYY-MM-DD HH:MM:SS". # # The three above are demonstrated by tests 1, 2 and 3 below. # Respectively. # do_createtable_tests 3.8.3 -query { SELECT a, b, c FROM t7 ORDER BY rowid DESC LIMIT 1; } { 1 "INSERT INTO t7(b, c) VALUES('x', 'y')" {01:46:40 x y} 2 "INSERT INTO t7(c, a) VALUES('x', 'y')" {y 2001-09-09 x} 3 "INSERT INTO t7(a, b) VALUES('x', 'y')" {x y {2001-09-09 01:46:40}} } # EVIDENCE-OF: R-55061-47754 The COLLATE clause specifies the name of a # collating sequence to use as the default collation sequence for the # column. # # EVIDENCE-OF: R-40275-54363 If no COLLATE clause is specified, the # default collation sequence is BINARY. # do_execsql_test e_createtable-3-9.1 { CREATE TABLE t8(a COLLATE nocase, b COLLATE rtrim, c COLLATE binary, d); INSERT INTO t8 VALUES('abc', 'abc', 'abc', 'abc'); INSERT INTO t8 VALUES('abc ', 'abc ', 'abc ', 'abc '); INSERT INTO t8 VALUES('ABC ', 'ABC ', 'ABC ', 'ABC '); INSERT INTO t8 VALUES('ABC', 'ABC', 'ABC', 'ABC'); } {} do_createtable_tests 3.9 { 2 "SELECT a FROM t8 ORDER BY a, rowid" {abc ABC {abc } {ABC }} 3 "SELECT b FROM t8 ORDER BY b, rowid" {{ABC } ABC abc {abc }} 4 "SELECT c FROM t8 ORDER BY c, rowid" {ABC {ABC } abc {abc }} 5 "SELECT d FROM t8 ORDER BY d, rowid" {ABC {ABC } abc {abc }} } # EVIDENCE-OF: R-25473-20557 The number of columns in a table is limited # by the SQLITE_MAX_COLUMN compile-time parameter. # proc columns {n} { set res [list] for {set i 0} {$i < $n} {incr i} { lappend res "c$i" } join $res ", " } do_execsql_test e_createtable-3.10.1 [subst { CREATE TABLE t9([columns $::SQLITE_MAX_COLUMN]); }] {} do_catchsql_test e_createtable-3.10.2 [subst { CREATE TABLE t10([columns [expr $::SQLITE_MAX_COLUMN+1]]); }] {1 {too many columns on t10}} # EVIDENCE-OF: R-27775-64721 Both of these limits can be lowered at # runtime using the sqlite3_limit() C/C++ interface. # # A 30,000 byte blob consumes 30,003 bytes of record space. A record # that contains 3 such blobs consumes (30,000*3)+1 bytes of space. Tests # 3.11.4 and 3.11.5, which verify that SQLITE_MAX_LENGTH may be lowered # at runtime, are based on this calculation. # sqlite3_limit db SQLITE_LIMIT_COLUMN 500 do_execsql_test e_createtable-3.11.1 [subst { CREATE TABLE t10([columns 500]); }] {} do_catchsql_test e_createtable-3.11.2 [subst { CREATE TABLE t11([columns 501]); }] {1 {too many columns on t11}} # Check that it is not possible to raise the column limit above its # default compile time value. # sqlite3_limit db SQLITE_LIMIT_COLUMN [expr $::SQLITE_MAX_COLUMN+2] do_catchsql_test e_createtable-3.11.3 [subst { CREATE TABLE t11([columns [expr $::SQLITE_MAX_COLUMN+1]]); }] {1 {too many columns on t11}} sqlite3_limit db SQLITE_LIMIT_LENGTH 90010 do_execsql_test e_createtable-3.11.4 { CREATE TABLE t12(a, b, c); INSERT INTO t12 VALUES(randomblob(30000),randomblob(30000),randomblob(30000)); } {} do_catchsql_test e_createtable-3.11.5 { INSERT INTO t12 VALUES(randomblob(30001),randomblob(30000),randomblob(30000)); } {1 {string or blob too big}} #------------------------------------------------------------------------- # Tests for statements regarding constraints (PRIMARY KEY, UNIQUE, NOT # NULL and CHECK constraints). # # EVIDENCE-OF: R-52382-54248 Each table in SQLite may have at most one # PRIMARY KEY. # # EVIDENCE-OF: R-18080-47271 If there is more than one PRIMARY KEY # clause in a single CREATE TABLE statement, it is an error. # # To test the two above, show that zero primary keys is Ok, one primary # key is Ok, and two or more primary keys is an error. # drop_all_tables do_createtable_tests 4.1.1 { 1 "CREATE TABLE t1(a, b, c)" {} 2 "CREATE TABLE t2(a PRIMARY KEY, b, c)" {} 3 "CREATE TABLE t3(a, b, c, PRIMARY KEY(a))" {} 4 "CREATE TABLE t4(a, b, c, PRIMARY KEY(c,b,a))" {} } do_createtable_tests 4.1.2 -error { table "t5" has more than one primary key } { 1 "CREATE TABLE t5(a PRIMARY KEY, b PRIMARY KEY, c)" {} 2 "CREATE TABLE t5(a, b PRIMARY KEY, c, PRIMARY KEY(a))" {} 3 "CREATE TABLE t5(a INTEGER PRIMARY KEY, b PRIMARY KEY, c)" {} 4 "CREATE TABLE t5(a INTEGER PRIMARY KEY, b, c, PRIMARY KEY(b, c))" {} 5 "CREATE TABLE t5(a PRIMARY KEY, b, c, PRIMARY KEY(a))" {} 6 "CREATE TABLE t5(a INTEGER PRIMARY KEY, b, c, PRIMARY KEY(a))" {} } proc table_pk {tbl} { set pk [list] db eval "pragma table_info($tbl)" a { if {$a(pk)} { lappend pk $a(name) } } set pk } # EVIDENCE-OF: R-41411-18837 If the keywords PRIMARY KEY are added to a # column definition, then the primary key for the table consists of that # single column. # # The above is tested by 4.2.1.* # # EVIDENCE-OF: R-31775-48204 Or, if a PRIMARY KEY clause is specified as # a table-constraint, then the primary key of the table consists of the # list of columns specified as part of the PRIMARY KEY clause. # # The above is tested by 4.2.2.* # do_createtable_tests 4.2 -repair { catchsql { DROP TABLE t5 } } -tclquery { table_pk t5 } { 1.1 "CREATE TABLE t5(a, b INTEGER PRIMARY KEY, c)" {b} 1.2 "CREATE TABLE t5(a PRIMARY KEY, b, c)" {a} 2.1 "CREATE TABLE t5(a, b, c, PRIMARY KEY(a))" {a} 2.2 "CREATE TABLE t5(a, b, c, PRIMARY KEY(c,b,a))" {a b c} 2.3 "CREATE TABLE t5(a, b INTEGER PRIMARY KEY, c)" {b} } # EVIDENCE-OF: R-33986-09410 Each row in a table with a primary key must # feature a unique combination of values in its primary key columns. # # EVIDENCE-OF: R-39102-06737 If an INSERT or UPDATE statement attempts # to modify the table content so that two or more rows feature identical # primary key values, it is a constraint violation. # drop_all_tables do_execsql_test 4.3.0 { CREATE TABLE t1(x PRIMARY KEY, y); INSERT INTO t1 VALUES(0, 'zero'); INSERT INTO t1 VALUES(45.5, 'one'); INSERT INTO t1 VALUES('brambles', 'two'); INSERT INTO t1 VALUES(X'ABCDEF', 'three'); CREATE TABLE t2(x, y, PRIMARY KEY(x, y)); INSERT INTO t2 VALUES(0, 'zero'); INSERT INTO t2 VALUES(45.5, 'one'); INSERT INTO t2 VALUES('brambles', 'two'); INSERT INTO t2 VALUES(X'ABCDEF', 'three'); } {} do_createtable_tests 4.3.1 -error { %s not unique } { 1 "INSERT INTO t1 VALUES(0, 0)" {"column x is"} 2 "INSERT INTO t1 VALUES(45.5, 'abc')" {"column x is"} 3 "INSERT INTO t1 VALUES(0.0, 'abc')" {"column x is"} 4 "INSERT INTO t1 VALUES('brambles', 'abc')" {"column x is"} 5 "INSERT INTO t1 VALUES(X'ABCDEF', 'abc')" {"column x is"} 6 "INSERT INTO t2 VALUES(0, 'zero')" {"columns x, y are"} 7 "INSERT INTO t2 VALUES(45.5, 'one')" {"columns x, y are"} 8 "INSERT INTO t2 VALUES(0.0, 'zero')" {"columns x, y are"} 9 "INSERT INTO t2 VALUES('brambles', 'two')" {"columns x, y are"} 10 "INSERT INTO t2 VALUES(X'ABCDEF', 'three')" {"columns x, y are"} } do_createtable_tests 4.3.2 { 1 "INSERT INTO t1 VALUES(-1, 0)" {} 2 "INSERT INTO t1 VALUES(45.2, 'abc')" {} 3 "INSERT INTO t1 VALUES(0.01, 'abc')" {} 4 "INSERT INTO t1 VALUES('bramble', 'abc')" {} 5 "INSERT INTO t1 VALUES(X'ABCDEE', 'abc')" {} 6 "INSERT INTO t2 VALUES(0, 0)" {} 7 "INSERT INTO t2 VALUES(45.5, 'abc')" {} 8 "INSERT INTO t2 VALUES(0.0, 'abc')" {} 9 "INSERT INTO t2 VALUES('brambles', 'abc')" {} 10 "INSERT INTO t2 VALUES(X'ABCDEF', 'abc')" {} } do_createtable_tests 4.3.3 -error { %s not unique } { 1 "UPDATE t1 SET x=0 WHERE y='two'" {"column x is"} 2 "UPDATE t1 SET x='brambles' WHERE y='three'" {"column x is"} 3 "UPDATE t1 SET x=45.5 WHERE y='zero'" {"column x is"} 4 "UPDATE t1 SET x=X'ABCDEF' WHERE y='one'" {"column x is"} 5 "UPDATE t1 SET x=0.0 WHERE y='three'" {"column x is"} 6 "UPDATE t2 SET x=0, y='zero' WHERE y='two'" {"columns x, y are"} 7 "UPDATE t2 SET x='brambles', y='two' WHERE y='three'" {"columns x, y are"} 8 "UPDATE t2 SET x=45.5, y='one' WHERE y='zero'" {"columns x, y are"} 9 "UPDATE t2 SET x=X'ABCDEF', y='three' WHERE y='one'" {"columns x, y are"} 10 "UPDATE t2 SET x=0.0, y='zero' WHERE y='three'" {"columns x, y are"} } # EVIDENCE-OF: R-52572-02078 For the purposes of determining the # uniqueness of primary key values, NULL values are considered distinct # from all other values, including other NULLs. # do_createtable_tests 4.4 { 1 "INSERT INTO t1 VALUES(NULL, 0)" {} 2 "INSERT INTO t1 VALUES(NULL, 0)" {} 3 "INSERT INTO t1 VALUES(NULL, 0)" {} 4 "INSERT INTO t2 VALUES(NULL, 'zero')" {} 5 "INSERT INTO t2 VALUES(NULL, 'one')" {} 6 "INSERT INTO t2 VALUES(NULL, 'two')" {} 7 "INSERT INTO t2 VALUES(NULL, 'three')" {} 8 "INSERT INTO t2 VALUES(0, NULL)" {} 9 "INSERT INTO t2 VALUES(45.5, NULL)" {} 10 "INSERT INTO t2 VALUES(0.0, NULL)" {} 11 "INSERT INTO t2 VALUES('brambles', NULL)" {} 12 "INSERT INTO t2 VALUES(X'ABCDEF', NULL)" {} 13 "INSERT INTO t2 VALUES(NULL, NULL)" {} 14 "INSERT INTO t2 VALUES(NULL, NULL)" {} } # EVIDENCE-OF: R-61866-38053 Unless the column is an INTEGER PRIMARY KEY # SQLite allows NULL values in a PRIMARY KEY column. # # If the column is an integer primary key, attempting to insert a NULL # into the column triggers the auto-increment behaviour. Attempting # to use UPDATE to set an ipk column to a NULL value is an error. # do_createtable_tests 4.5.1 { 1 "SELECT count(*) FROM t1 WHERE x IS NULL" 3 2 "SELECT count(*) FROM t2 WHERE x IS NULL" 6 3 "SELECT count(*) FROM t2 WHERE y IS NULL" 7 4 "SELECT count(*) FROM t2 WHERE x IS NULL AND y IS NULL" 2 } do_execsql_test 4.5.2 { CREATE TABLE t3(s, u INTEGER PRIMARY KEY, v); INSERT INTO t3 VALUES(1, NULL, 2); INSERT INTO t3 VALUES('x', NULL, 'y'); SELECT u FROM t3; } {1 2} do_catchsql_test 4.5.3 { INSERT INTO t3 VALUES(2, 5, 3); UPDATE t3 SET u = NULL WHERE s = 2; } {1 {datatype mismatch}} # EVIDENCE-OF: R-00227-21080 A UNIQUE constraint is similar to a PRIMARY # KEY constraint, except that a single table may have any number of # UNIQUE constraints. # drop_all_tables do_createtable_tests 4.6 { 1 "CREATE TABLE t1(a UNIQUE, b UNIQUE)" {} 2 "CREATE TABLE t2(a UNIQUE, b, c, UNIQUE(c, b))" {} 3 "CREATE TABLE t3(a, b, c, UNIQUE(a), UNIQUE(b), UNIQUE(c))" {} 4 "CREATE TABLE t4(a, b, c, UNIQUE(a, b, c))" {} } # EVIDENCE-OF: R-55240-58877 For each UNIQUE constraint on the table, # each row must feature a unique combination of values in the columns # identified by the UNIQUE constraint. # # EVIDENCE-OF: R-47733-51480 If an INSERT or UPDATE statement attempts # to modify the table content so that two or more rows feature identical # values in a set of columns that are subject to a UNIQUE constraint, it # is a constraint violation. # do_execsql_test 4.7.0 { INSERT INTO t1 VALUES(1, 2); INSERT INTO t1 VALUES(4.3, 5.5); INSERT INTO t1 VALUES('reveal', 'variableness'); INSERT INTO t1 VALUES(X'123456', X'654321'); INSERT INTO t4 VALUES('xyx', 1, 1); INSERT INTO t4 VALUES('xyx', 2, 1); INSERT INTO t4 VALUES('uvw', 1, 1); } do_createtable_tests 4.7.1 -error { %s not unique } { 1 "INSERT INTO t1 VALUES(1, 'one')" {{column a is}} 2 "INSERT INTO t1 VALUES(4.3, 'two')" {{column a is}} 3 "INSERT INTO t1 VALUES('reveal', 'three')" {{column a is}} 4 "INSERT INTO t1 VALUES(X'123456', 'four')" {{column a is}} 5 "UPDATE t1 SET a = 1 WHERE rowid=2" {{column a is}} 6 "UPDATE t1 SET a = 4.3 WHERE rowid=3" {{column a is}} 7 "UPDATE t1 SET a = 'reveal' WHERE rowid=4" {{column a is}} 8 "UPDATE t1 SET a = X'123456' WHERE rowid=1" {{column a is}} 9 "INSERT INTO t4 VALUES('xyx', 1, 1)" {{columns a, b, c are}} 10 "INSERT INTO t4 VALUES('xyx', 2, 1)" {{columns a, b, c are}} 11 "INSERT INTO t4 VALUES('uvw', 1, 1)" {{columns a, b, c are}} 12 "UPDATE t4 SET a='xyx' WHERE rowid=3" {{columns a, b, c are}} 13 "UPDATE t4 SET b=1 WHERE rowid=2" {{columns a, b, c are}} 14 "UPDATE t4 SET a=0, b=0, c=0" {{columns a, b, c are}} } # EVIDENCE-OF: R-21289-11559 As with PRIMARY KEY constraints, for the # purposes of UNIQUE constraints NULL values are considered distinct # from all other values (including other NULLs). # do_createtable_tests 4.8 { 1 "INSERT INTO t1 VALUES(NULL, NULL)" {} 2 "INSERT INTO t1 VALUES(NULL, NULL)" {} 3 "UPDATE t1 SET a = NULL" {} 4 "UPDATE t1 SET b = NULL" {} 5 "INSERT INTO t4 VALUES(NULL, NULL, NULL)" {} 6 "INSERT INTO t4 VALUES(NULL, NULL, NULL)" {} 7 "UPDATE t4 SET a = NULL" {} 8 "UPDATE t4 SET b = NULL" {} 9 "UPDATE t4 SET c = NULL" {} } # EVIDENCE-OF: R-26983-26377 INTEGER PRIMARY KEY columns aside, both # UNIQUE and PRIMARY KEY constraints are implemented by creating an # index in the database (in the same way as a "CREATE UNIQUE INDEX" # statement would). do_createtable_tests 4.9 -repair drop_all_tables -query { SELECT count(*) FROM sqlite_master WHERE type='index' } { 1 "CREATE TABLE t1(a TEXT PRIMARY KEY, b)" 1 2 "CREATE TABLE t1(a INTEGER PRIMARY KEY, b)" 0 3 "CREATE TABLE t1(a TEXT UNIQUE, b)" 1 4 "CREATE TABLE t1(a PRIMARY KEY, b TEXT UNIQUE)" 2 5 "CREATE TABLE t1(a PRIMARY KEY, b, c, UNIQUE(c, b))" 2 } # EVIDENCE-OF: R-02252-33116 Such an index is used like any other index # in the database to optimize queries. # do_execsql_test 4.10.0 { CREATE TABLE t1(a, b PRIMARY KEY); CREATE TABLE t2(a, b, c, UNIQUE(b, c)); } do_createtable_tests 4.10 { 1 "EXPLAIN QUERY PLAN SELECT * FROM t1 WHERE b = 5" {0 0 0 {SEARCH TABLE t1 USING INDEX sqlite_autoindex_t1_1 (b=?) (~1 rows)}} 2 "EXPLAIN QUERY PLAN SELECT * FROM t2 ORDER BY b, c" {0 0 0 {SCAN TABLE t2 USING INDEX sqlite_autoindex_t2_1 (~1000000 rows)}} 3 "EXPLAIN QUERY PLAN SELECT * FROM t2 WHERE b=10 AND c>10" {0 0 0 {SEARCH TABLE t2 USING INDEX sqlite_autoindex_t2_1 (b=? AND c>?) (~2 rows)}} } # EVIDENCE-OF: R-45493-35653 A CHECK constraint may be attached to a # column definition or specified as a table constraint. In practice it # makes no difference. # # All the tests that deal with CHECK constraints below (4.11.* and # 4.12.*) are run once for a table with the check constraint attached # to a column definition, and once with a table where the check # condition is specified as a table constraint. # # EVIDENCE-OF: R-55435-14303 Each time a new row is inserted into the # table or an existing row is updated, the expression associated with # each CHECK constraint is evaluated and cast to a NUMERIC value in the # same way as a CAST expression. If the result is zero (integer value 0 # or real value 0.0), then a constraint violation has occurred. # drop_all_tables do_execsql_test 4.11 { CREATE TABLE x1(a TEXT, b INTEGER CHECK( b>0 )); CREATE TABLE t1(a TEXT, b INTEGER, CHECK( b>0 )); INSERT INTO x1 VALUES('x', 'xx'); INSERT INTO x1 VALUES('y', 'yy'); INSERT INTO t1 SELECT * FROM x1; CREATE TABLE x2(a CHECK( a||b ), b); CREATE TABLE t2(a, b, CHECK( a||b )); INSERT INTO x2 VALUES(1, 'xx'); INSERT INTO x2 VALUES(1, 'yy'); INSERT INTO t2 SELECT * FROM x2; } do_createtable_tests 4.11 -error {constraint failed} { 1a "INSERT INTO x1 VALUES('one', 0)" {} 1b "INSERT INTO t1 VALUES('one', -4.0)" {} 2a "INSERT INTO x2 VALUES('abc', 1)" {} 2b "INSERT INTO t2 VALUES('abc', 1)" {} 3a "INSERT INTO x2 VALUES(0, 'abc')" {} 3b "INSERT INTO t2 VALUES(0, 'abc')" {} 4a "UPDATE t1 SET b=-1 WHERE rowid=1" {} 4b "UPDATE x1 SET b=-1 WHERE rowid=1" {} 4a "UPDATE x2 SET a='' WHERE rowid=1" {} 4b "UPDATE t2 SET a='' WHERE rowid=1" {} } # EVIDENCE-OF: R-34109-39108 If the CHECK expression evaluates to NULL, # or any other non-zero value, it is not a constraint violation. # do_createtable_tests 4.12 { 1a "INSERT INTO x1 VALUES('one', NULL)" {} 1b "INSERT INTO t1 VALUES('one', NULL)" {} 2a "INSERT INTO x1 VALUES('one', 2)" {} 2b "INSERT INTO t1 VALUES('one', 2)" {} 3a "INSERT INTO x2 VALUES(1, 'abc')" {} 3b "INSERT INTO t2 VALUES(1, 'abc')" {} } # EVIDENCE-OF: R-02060-64547 A NOT NULL constraint may only be attached # to a column definition, not specified as a table constraint. # drop_all_tables do_createtable_tests 4.13.1 { 1 "CREATE TABLE t1(a NOT NULL, b)" {} 2 "CREATE TABLE t2(a PRIMARY KEY NOT NULL, b)" {} 3 "CREATE TABLE t3(a NOT NULL, b NOT NULL, c NOT NULL UNIQUE)" {} } do_createtable_tests 4.13.2 -error { near "NOT": syntax error } { 1 "CREATE TABLE t4(a, b, NOT NULL(a))" {} 2 "CREATE TABLE t4(a PRIMARY KEY, b, NOT NULL(a))" {} 3 "CREATE TABLE t4(a, b, c UNIQUE, NOT NULL(a, b, c))" {} } # EVIDENCE-OF: R-31795-57643 a NOT NULL constraint dictates that the # associated column may not contain a NULL value. Attempting to set the # column value to NULL when inserting a new row or updating an existing # one causes a constraint violation. # # These tests use the tables created by 4.13. # do_execsql_test 4.14.0 { INSERT INTO t1 VALUES('x', 'y'); INSERT INTO t1 VALUES('z', NULL); INSERT INTO t2 VALUES('x', 'y'); INSERT INTO t2 VALUES('z', NULL); INSERT INTO t3 VALUES('x', 'y', 'z'); INSERT INTO t3 VALUES(1, 2, 3); } do_createtable_tests 4.14 -error { %s may not be NULL } { 1 "INSERT INTO t1 VALUES(NULL, 'a')" {t1.a} 2 "INSERT INTO t2 VALUES(NULL, 'b')" {t2.a} 3 "INSERT INTO t3 VALUES('c', 'd', NULL)" {t3.c} 4 "INSERT INTO t3 VALUES('e', NULL, 'f')" {t3.b} 5 "INSERT INTO t3 VALUES(NULL, 'g', 'h')" {t3.a} } # EVIDENCE-OF: R-42511-39459 PRIMARY KEY, UNIQUE and NOT NULL # constraints may be explicitly assigned a default conflict resolution # algorithm by including a conflict-clause in their definitions. # # Conflict clauses: ABORT, ROLLBACK, IGNORE, FAIL, REPLACE # # Test cases 4.15.*, 4.16.* and 4.17.* focus on PRIMARY KEY, NOT NULL # and UNIQUE constraints, respectively. # drop_all_tables do_execsql_test 4.15.0 { CREATE TABLE t1_ab(a PRIMARY KEY ON CONFLICT ABORT, b); CREATE TABLE t1_ro(a PRIMARY KEY ON CONFLICT ROLLBACK, b); CREATE TABLE t1_ig(a PRIMARY KEY ON CONFLICT IGNORE, b); CREATE TABLE t1_fa(a PRIMARY KEY ON CONFLICT FAIL, b); CREATE TABLE t1_re(a PRIMARY KEY ON CONFLICT REPLACE, b); CREATE TABLE t1_xx(a PRIMARY KEY, b); INSERT INTO t1_ab VALUES(1, 'one'); INSERT INTO t1_ab VALUES(2, 'two'); INSERT INTO t1_ro SELECT * FROM t1_ab; INSERT INTO t1_ig SELECT * FROM t1_ab; INSERT INTO t1_fa SELECT * FROM t1_ab; INSERT INTO t1_re SELECT * FROM t1_ab; INSERT INTO t1_xx SELECT * FROM t1_ab; CREATE TABLE t2_ab(a, b NOT NULL ON CONFLICT ABORT); CREATE TABLE t2_ro(a, b NOT NULL ON CONFLICT ROLLBACK); CREATE TABLE t2_ig(a, b NOT NULL ON CONFLICT IGNORE); CREATE TABLE t2_fa(a, b NOT NULL ON CONFLICT FAIL); CREATE TABLE t2_re(a, b NOT NULL ON CONFLICT REPLACE); CREATE TABLE t2_xx(a, b NOT NULL); INSERT INTO t2_ab VALUES(1, 'one'); INSERT INTO t2_ab VALUES(2, 'two'); INSERT INTO t2_ro SELECT * FROM t2_ab; INSERT INTO t2_ig SELECT * FROM t2_ab; INSERT INTO t2_fa SELECT * FROM t2_ab; INSERT INTO t2_re SELECT * FROM t2_ab; INSERT INTO t2_xx SELECT * FROM t2_ab; CREATE TABLE t3_ab(a, b, UNIQUE(a, b) ON CONFLICT ABORT); CREATE TABLE t3_ro(a, b, UNIQUE(a, b) ON CONFLICT ROLLBACK); CREATE TABLE t3_ig(a, b, UNIQUE(a, b) ON CONFLICT IGNORE); CREATE TABLE t3_fa(a, b, UNIQUE(a, b) ON CONFLICT FAIL); CREATE TABLE t3_re(a, b, UNIQUE(a, b) ON CONFLICT REPLACE); CREATE TABLE t3_xx(a, b, UNIQUE(a, b)); INSERT INTO t3_ab VALUES(1, 'one'); INSERT INTO t3_ab VALUES(2, 'two'); INSERT INTO t3_ro SELECT * FROM t3_ab; INSERT INTO t3_ig SELECT * FROM t3_ab; INSERT INTO t3_fa SELECT * FROM t3_ab; INSERT INTO t3_re SELECT * FROM t3_ab; INSERT INTO t3_xx SELECT * FROM t3_ab; } foreach {tn tbl res ac data} { 1 t1_ab {1 {column a is not unique}} 0 {1 one 2 two 3 three} 2 t1_ro {1 {column a is not unique}} 1 {1 one 2 two} 3 t1_fa {1 {column a is not unique}} 0 {1 one 2 two 3 three 4 string} 4 t1_ig {0 {}} 0 {1 one 2 two 3 three 4 string 6 string} 5 t1_re {0 {}} 0 {1 one 2 two 4 string 3 string 6 string} 6 t1_xx {1 {column a is not unique}} 0 {1 one 2 two 3 three} } { catchsql COMMIT do_execsql_test 4.15.$tn.1 "BEGIN; INSERT INTO $tbl VALUES(3, 'three')" do_catchsql_test 4.15.$tn.2 " INSERT INTO $tbl SELECT ((a%2)*a+3), 'string' FROM $tbl; " $res do_test e_createtable-4.15.$tn.3 { sqlite3_get_autocommit db } $ac do_execsql_test 4.15.$tn.4 "SELECT * FROM $tbl" $data } foreach {tn tbl res ac data} { 1 t2_ab {1 {t2_ab.b may not be NULL}} 0 {1 one 2 two 3 three} 2 t2_ro {1 {t2_ro.b may not be NULL}} 1 {1 one 2 two} 3 t2_fa {1 {t2_fa.b may not be NULL}} 0 {1 one 2 two 3 three 4 xx} 4 t2_ig {0 {}} 0 {1 one 2 two 3 three 4 xx 6 xx} 5 t2_re {1 {t2_re.b may not be NULL}} 0 {1 one 2 two 3 three} 6 t2_xx {1 {t2_xx.b may not be NULL}} 0 {1 one 2 two 3 three} } { catchsql COMMIT do_execsql_test 4.16.$tn.1 "BEGIN; INSERT INTO $tbl VALUES(3, 'three')" do_catchsql_test 4.16.$tn.2 " INSERT INTO $tbl SELECT a+3, CASE a WHEN 2 THEN NULL ELSE 'xx' END FROM $tbl " $res do_test e_createtable-4.16.$tn.3 { sqlite3_get_autocommit db } $ac do_execsql_test 4.16.$tn.4 "SELECT * FROM $tbl" $data } foreach {tn tbl res ac data} { 1 t3_ab {1 {columns a, b are not unique}} 0 {1 one 2 two 3 three} 2 t3_ro {1 {columns a, b are not unique}} 1 {1 one 2 two} 3 t3_fa {1 {columns a, b are not unique}} 0 {1 one 2 two 3 three 4 three} 4 t3_ig {0 {}} 0 {1 one 2 two 3 three 4 three 6 three} 5 t3_re {0 {}} 0 {1 one 2 two 4 three 3 three 6 three} 6 t3_xx {1 {columns a, b are not unique}} 0 {1 one 2 two 3 three} } { catchsql COMMIT do_execsql_test 4.17.$tn.1 "BEGIN; INSERT INTO $tbl VALUES(3, 'three')" do_catchsql_test 4.17.$tn.2 " INSERT INTO $tbl SELECT ((a%2)*a+3), 'three' FROM $tbl " $res do_test e_createtable-4.17.$tn.3 { sqlite3_get_autocommit db } $ac do_execsql_test 4.17.$tn.4 "SELECT * FROM $tbl" $data } catchsql COMMIT # EVIDENCE-OF: R-12645-39772 Or, if a constraint definition does not # include a conflict-clause or it is a CHECK constraint, the default # conflict resolution algorithm is ABORT. # # The first half of the above is tested along with explicit ON # CONFLICT clauses above (specifically, the tests involving t1_xx, t2_xx # and t3_xx). The following just tests that the default conflict # handling for CHECK constraints is ABORT. # do_execsql_test 4.18.1 { CREATE TABLE t4(a, b CHECK (b!=10)); INSERT INTO t4 VALUES(1, 2); INSERT INTO t4 VALUES(3, 4); } do_execsql_test 4.18.2 { BEGIN; INSERT INTO t4 VALUES(5, 6) } do_catchsql_test 4.18.3 { INSERT INTO t4 SELECT a+4, b+4 FROM t4 } {1 {constraint failed}} do_test e_createtable-4.18.4 { sqlite3_get_autocommit db } 0 do_execsql_test 4.18.5 { SELECT * FROM t4 } {1 2 3 4 5 6} # EVIDENCE-OF: R-19114-56113 Different constraints within the same table # may have different default conflict resolution algorithms. # do_execsql_test 4.19.0 { CREATE TABLE t5(a NOT NULL ON CONFLICT IGNORE, b NOT NULL ON CONFLICT ABORT); } do_catchsql_test 4.19.1 { INSERT INTO t5 VALUES(NULL, 'not null') } {0 {}} do_execsql_test 4.19.2 { SELECT * FROM t5 } {} do_catchsql_test 4.19.3 { INSERT INTO t5 VALUES('not null', NULL) } \ {1 {t5.b may not be NULL}} do_execsql_test 4.19.4 { SELECT * FROM t5 } {} #------------------------------------------------------------------------ # Tests for INTEGER PRIMARY KEY and rowid related statements. # # EVIDENCE-OF: R-52584-04009 The rowid value can be accessed using one # of the special case-independent names "rowid", "oid", or "_rowid_" in # place of a column name. # drop_all_tables do_execsql_test 5.1.0 { CREATE TABLE t1(x, y); INSERT INTO t1 VALUES('one', 'first'); INSERT INTO t1 VALUES('two', 'second'); INSERT INTO t1 VALUES('three', 'third'); } do_createtable_tests 5.1 { 1 "SELECT rowid FROM t1" {1 2 3} 2 "SELECT oid FROM t1" {1 2 3} 3 "SELECT _rowid_ FROM t1" {1 2 3} 4 "SELECT ROWID FROM t1" {1 2 3} 5 "SELECT OID FROM t1" {1 2 3} 6 "SELECT _ROWID_ FROM t1" {1 2 3} 7 "SELECT RoWiD FROM t1" {1 2 3} 8 "SELECT OiD FROM t1" {1 2 3} 9 "SELECT _RoWiD_ FROM t1" {1 2 3} } # EVIDENCE-OF: R-26501-17306 If a table contains a user defined column # named "rowid", "oid" or "_rowid_", then that name always refers the # explicitly declared column and cannot be used to retrieve the integer # rowid value. # do_execsql_test 5.2.0 { CREATE TABLE t2(oid, b); CREATE TABLE t3(a, _rowid_); CREATE TABLE t4(a, b, rowid); INSERT INTO t2 VALUES('one', 'two'); INSERT INTO t2 VALUES('three', 'four'); INSERT INTO t3 VALUES('five', 'six'); INSERT INTO t3 VALUES('seven', 'eight'); INSERT INTO t4 VALUES('nine', 'ten', 'eleven'); INSERT INTO t4 VALUES('twelve', 'thirteen', 'fourteen'); } do_createtable_tests 5.2 { 1 "SELECT oid, rowid, _rowid_ FROM t2" {one 1 1 three 2 2} 2 "SELECT oid, rowid, _rowid_ FROM t3" {1 1 six 2 2 eight} 3 "SELECT oid, rowid, _rowid_ FROM t4" {1 eleven 1 2 fourteen 2} } # Argument $tbl is the name of a table in the database. Argument $col is # the name of one of the tables columns. Return 1 if $col is an alias for # the rowid, or 0 otherwise. # proc is_integer_primary_key {tbl col} { lindex [db eval [subst { DELETE FROM $tbl; INSERT INTO $tbl ($col) VALUES(0); SELECT (rowid==$col) FROM $tbl; DELETE FROM $tbl; }]] 0 } # EVIDENCE-OF: R-53738-31673 With one exception, if a table has a # primary key that consists of a single column, and the declared type of # that column is "INTEGER" in any mixture of upper and lower case, then # the column becomes an alias for the rowid. # # EVIDENCE-OF: R-45951-08347 if the declaration of a column with # declared type "INTEGER" includes an "PRIMARY KEY DESC" clause, it does # not become an alias for the rowid and is not classified as an integer # primary key. # do_createtable_tests 5.3 -tclquery { is_integer_primary_key t5 pk } -repair { catchsql { DROP TABLE t5 } } { 1 "CREATE TABLE t5(pk integer primary key)" 1 2 "CREATE TABLE t5(pk integer, primary key(pk))" 1 3 "CREATE TABLE t5(pk integer, v integer, primary key(pk))" 1 4 "CREATE TABLE t5(pk integer, v integer, primary key(pk, v))" 0 5 "CREATE TABLE t5(pk int, v integer, primary key(pk, v))" 0 6 "CREATE TABLE t5(pk int, v integer, primary key(pk))" 0 7 "CREATE TABLE t5(pk int primary key, v integer)" 0 8 "CREATE TABLE t5(pk inTEger primary key)" 1 9 "CREATE TABLE t5(pk inteGEr, primary key(pk))" 1 10 "CREATE TABLE t5(pk INTEGER, v integer, primary key(pk))" 1 } # EVIDENCE-OF: R-41444-49665 Other integer type names like "INT" or # "BIGINT" or "SHORT INTEGER" or "UNSIGNED INTEGER" causes the primary # key column to behave as an ordinary table column with integer affinity # and a unique index, not as an alias for the rowid. # do_execsql_test 5.4.1 { CREATE TABLE t6(pk INT primary key); CREATE TABLE t7(pk BIGINT primary key); CREATE TABLE t8(pk SHORT INTEGER primary key); CREATE TABLE t9(pk UNSIGNED INTEGER primary key); } do_test e_createtable-5.4.2.1 { is_integer_primary_key t6 pk } 0 do_test e_createtable-5.4.2.2 { is_integer_primary_key t7 pk } 0 do_test e_createtable-5.4.2.3 { is_integer_primary_key t8 pk } 0 do_test e_createtable-5.4.2.4 { is_integer_primary_key t9 pk } 0 do_execsql_test 5.4.3 { INSERT INTO t6 VALUES('2.0'); INSERT INTO t7 VALUES('2.0'); INSERT INTO t8 VALUES('2.0'); INSERT INTO t9 VALUES('2.0'); SELECT typeof(pk), pk FROM t6; SELECT typeof(pk), pk FROM t7; SELECT typeof(pk), pk FROM t8; SELECT typeof(pk), pk FROM t9; } {integer 2 integer 2 integer 2 integer 2} do_catchsql_test 5.4.4.1 { INSERT INTO t6 VALUES(2) } {1 {column pk is not unique}} do_catchsql_test 5.4.4.2 { INSERT INTO t7 VALUES(2) } {1 {column pk is not unique}} do_catchsql_test 5.4.4.3 { INSERT INTO t8 VALUES(2) } {1 {column pk is not unique}} do_catchsql_test 5.4.4.4 { INSERT INTO t9 VALUES(2) } {1 {column pk is not unique}} # EVIDENCE-OF: R-56094-57830 the following three table declarations all # cause the column "x" to be an alias for the rowid (an integer primary # key): CREATE TABLE t(x INTEGER PRIMARY KEY ASC, y, z); CREATE TABLE # t(x INTEGER, y, z, PRIMARY KEY(x ASC)); CREATE TABLE t(x INTEGER, y, # z, PRIMARY KEY(x DESC)); # # EVIDENCE-OF: R-20149-25884 the following declaration does not result # in "x" being an alias for the rowid: CREATE TABLE t(x INTEGER PRIMARY # KEY DESC, y, z); # do_createtable_tests 5 -tclquery { is_integer_primary_key t x } -repair { catchsql { DROP TABLE t } } { 5.1 "CREATE TABLE t(x INTEGER PRIMARY KEY ASC, y, z)" 1 5.2 "CREATE TABLE t(x INTEGER, y, z, PRIMARY KEY(x ASC))" 1 5.3 "CREATE TABLE t(x INTEGER, y, z, PRIMARY KEY(x DESC))" 1 6.1 "CREATE TABLE t(x INTEGER PRIMARY KEY DESC, y, z)" 0 } # EVIDENCE-OF: R-03733-29734 Rowid values may be modified using an # UPDATE statement in the same way as any other column value can, either # using one of the built-in aliases ("rowid", "oid" or "_rowid_") or by # using an alias created by an integer primary key. # do_execsql_test 5.7.0 { CREATE TABLE t10(a, b); INSERT INTO t10 VALUES('ten', 10); CREATE TABLE t11(a, b INTEGER PRIMARY KEY); INSERT INTO t11 VALUES('ten', 10); } do_createtable_tests 5.7.1 -query { SELECT rowid, _rowid_, oid FROM t10; } { 1 "UPDATE t10 SET rowid = 5" {5 5 5} 2 "UPDATE t10 SET _rowid_ = 6" {6 6 6} 3 "UPDATE t10 SET oid = 7" {7 7 7} } do_createtable_tests 5.7.2 -query { SELECT rowid, _rowid_, oid, b FROM t11; } { 1 "UPDATE t11 SET rowid = 5" {5 5 5 5} 2 "UPDATE t11 SET _rowid_ = 6" {6 6 6 6} 3 "UPDATE t11 SET oid = 7" {7 7 7 7} 4 "UPDATE t11 SET b = 8" {8 8 8 8} } # EVIDENCE-OF: R-58706-14229 Similarly, an INSERT statement may provide # a value to use as the rowid for each row inserted. # do_createtable_tests 5.8.1 -query { SELECT rowid, _rowid_, oid FROM t10; } -repair { execsql { DELETE FROM t10 } } { 1 "INSERT INTO t10(oid) VALUES(15)" {15 15 15} 2 "INSERT INTO t10(rowid) VALUES(16)" {16 16 16} 3 "INSERT INTO t10(_rowid_) VALUES(17)" {17 17 17} 4 "INSERT INTO t10(a, b, oid) VALUES(1,2,3)" {3 3 3} } do_createtable_tests 5.8.2 -query { SELECT rowid, _rowid_, oid, b FROM t11; } -repair { execsql { DELETE FROM t11 } } { 1 "INSERT INTO t11(oid) VALUES(15)" {15 15 15 15} 2 "INSERT INTO t11(rowid) VALUES(16)" {16 16 16 16} 3 "INSERT INTO t11(_rowid_) VALUES(17)" {17 17 17 17} 4 "INSERT INTO t11(a, b) VALUES(1,2)" {2 2 2 2} } # EVIDENCE-OF: R-32326-44592 Unlike normal SQLite columns, an integer # primary key or rowid column must contain integer values. Integer # primary key or rowid columns are not able to hold floating point # values, strings, BLOBs, or NULLs. # # This is considered by the tests for the following 3 statements, # which show that: # # 1. Attempts to UPDATE a rowid column to a non-integer value fail, # 2. Attempts to INSERT a real, string or blob value into a rowid # column fail, and # 3. Attempting to INSERT a NULL value into a rowid column causes the # system to automatically select an integer value to use. # # EVIDENCE-OF: R-64224-62578 If an UPDATE statement attempts to set an # integer primary key or rowid column to a NULL or blob value, or to a # string or real value that cannot be losslessly converted to an # integer, a "datatype mismatch" error occurs and the statement is # aborted. # drop_all_tables do_execsql_test 5.9.0 { CREATE TABLE t12(x INTEGER PRIMARY KEY, y); INSERT INTO t12 VALUES(5, 'five'); } do_createtable_tests 5.9.1 -query { SELECT typeof(x), x FROM t12 } { 1 "UPDATE t12 SET x = 4" {integer 4} 2 "UPDATE t12 SET x = 10.0" {integer 10} 3 "UPDATE t12 SET x = '12.0'" {integer 12} 4 "UPDATE t12 SET x = '-15.0'" {integer -15} } do_createtable_tests 5.9.2 -error { datatype mismatch } { 1 "UPDATE t12 SET x = 4.1" {} 2 "UPDATE t12 SET x = 'hello'" {} 3 "UPDATE t12 SET x = NULL" {} 4 "UPDATE t12 SET x = X'ABCD'" {} 5 "UPDATE t12 SET x = X'3900'" {} 6 "UPDATE t12 SET x = X'39'" {} } # EVIDENCE-OF: R-05734-13629 If an INSERT statement attempts to insert a # blob value, or a string or real value that cannot be losslessly # converted to an integer into an integer primary key or rowid column, a # "datatype mismatch" error occurs and the statement is aborted. # do_execsql_test 5.10.0 { DELETE FROM t12 } do_createtable_tests 5.10.1 -error { datatype mismatch } { 1 "INSERT INTO t12(x) VALUES(4.1)" {} 2 "INSERT INTO t12(x) VALUES('hello')" {} 3 "INSERT INTO t12(x) VALUES(X'ABCD')" {} 4 "INSERT INTO t12(x) VALUES(X'3900')" {} 5 "INSERT INTO t12(x) VALUES(X'39')" {} } do_createtable_tests 5.10.2 -query { SELECT typeof(x), x FROM t12 } -repair { execsql { DELETE FROM t12 } } { 1 "INSERT INTO t12(x) VALUES(4)" {integer 4} 2 "INSERT INTO t12(x) VALUES(10.0)" {integer 10} 3 "INSERT INTO t12(x) VALUES('12.0')" {integer 12} 4 "INSERT INTO t12(x) VALUES('4e3')" {integer 4000} 5 "INSERT INTO t12(x) VALUES('-14.0')" {integer -14} } # EVIDENCE-OF: R-07986-46024 If an INSERT statement attempts to insert a # NULL value into a rowid or integer primary key column, the system # chooses an integer value to use as the rowid automatically. # do_execsql_test 5.11.0 { DELETE FROM t12 } do_createtable_tests 5.11 -query { SELECT typeof(x), x FROM t12 WHERE y IS (SELECT max(y) FROM t12) } { 1 "INSERT INTO t12 DEFAULT VALUES" {integer 1} 2 "INSERT INTO t12(y) VALUES(5)" {integer 2} 3 "INSERT INTO t12(x,y) VALUES(NULL, 10)" {integer 3} 4 "INSERT INTO t12(x,y) SELECT NULL, 15 FROM t12" {integer 4 integer 5 integer 6} 5 "INSERT INTO t12(y) SELECT 20 FROM t12 LIMIT 3" {integer 7 integer 8 integer 9} } finish_test