sqlite/test/enc.test

# 2002 May 24
#
# 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 file is testing the SQLite routines used for converting between the
# various suported unicode encodings (UTF-8, UTF-16, UTF-16le and
# UTF-16be).
#
# $Id: enc.test,v 1.7 2007/05/23 16:23:09 danielk1977 Exp $

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

# Skip this test if the build does not support multiple encodings.
#
ifcapable {!utf16} {
  finish_test
  return
}

proc do_bincmp_test {testname got expect} {
  binary scan $expect \c* expectvals
  binary scan $got \c* gotvals
  do_test $testname [list set dummy $gotvals] $expectvals
}

# $utf16 is a UTF-16 encoded string. Swap each pair of bytes around
# to change the byte-order of the string.
proc swap_byte_order {utf16} {
  binary scan $utf16 \c* ints

  foreach {a b} $ints {
    lappend ints2 $b
    lappend ints2 $a
  }

  return [binary format \c* $ints2]
}

#
# Test that the SQLite routines for converting between UTF encodings
# produce the same results as their TCL counterparts.
#
# $testname is the prefix to be used for the test names.
# $str is a string to use for testing (encoded in UTF-8, as normal for TCL).
#
# The test procedure is:
# 1. Convert the string from UTF-8 to UTF-16le and check that the TCL and
#    SQLite routines produce the same results.
#
# 2. Convert the string from UTF-8 to UTF-16be and check that the TCL and
#    SQLite routines produce the same results.
#
# 3. Use the SQLite routines to convert the native machine order UTF-16
#    representation back to the original UTF-8. Check that the result
#    matches the original representation.
#
# 4. Add a byte-order mark to each of the UTF-16 representations and
#    check that the SQLite routines can convert them back to UTF-8.  For
#    byte-order mark info, refer to section 3.10 of the unicode standard.
#
# 5. Take the byte-order marked UTF-16 strings from step 4 and ensure
#    that SQLite can convert them both to native byte order UTF-16 
#    strings, sans BOM.
#
# Coverage:
#
# sqlite_utf8to16be (step 2)
# sqlite_utf8to16le (step 1)
# sqlite_utf16to8 (steps 3, 4)
# sqlite_utf16to16le (step 5)
# sqlite_utf16to16be (step 5)
#
proc test_conversion {testname str} {
 
  # Step 1.
  set utf16le_sqlite3 [test_translate $str UTF8 UTF16LE]
  set utf16le_tcl [encoding convertto unicode $str]
  append utf16le_tcl "\x00\x00"
  if { $::tcl_platform(byteOrder)!="littleEndian" } {
    set utf16le_tcl [swap_byte_order $utf16le_tcl]
  }
  do_bincmp_test $testname.1 $utf16le_sqlite3 $utf16le_tcl
  set utf16le $utf16le_tcl

  # Step 2.
  set utf16be_sqlite3 [test_translate $str UTF8 UTF16BE]
  set utf16be_tcl [encoding convertto unicode $str]
  append utf16be_tcl "\x00\x00"
  if { $::tcl_platform(byteOrder)=="littleEndian" } {
    set utf16be_tcl [swap_byte_order $utf16be_tcl]
  }
  do_bincmp_test $testname.2 $utf16be_sqlite3 $utf16be_tcl
  set utf16be $utf16be_tcl
 
  # Step 3.
  if { $::tcl_platform(byteOrder)=="littleEndian" } {
    set utf16 $utf16le
  } else {
    set utf16 $utf16be
  }
  set utf8_sqlite3 [test_translate $utf16 UTF16 UTF8]
  do_bincmp_test $testname.3 $utf8_sqlite3 [binarize $str]

  # Step 4 (little endian).
  append utf16le_bom "\xFF\xFE" $utf16le
  set utf8_sqlite3 [test_translate $utf16le_bom UTF16 UTF8 1]
  do_bincmp_test $testname.4.le $utf8_sqlite3 [binarize $str]

  # Step 4 (big endian).
  append utf16be_bom "\xFE\xFF" $utf16be
  set utf8_sqlite3 [test_translate $utf16be_bom UTF16 UTF8]
  do_bincmp_test $testname.4.be $utf8_sqlite3 [binarize $str]

  # Step 5 (little endian to little endian).
  set utf16_sqlite3 [test_translate $utf16le_bom UTF16LE UTF16LE]
  do_bincmp_test $testname.5.le.le $utf16_sqlite3 $utf16le

  # Step 5 (big endian to big endian).
  set utf16_sqlite3 [test_translate $utf16be_bom UTF16 UTF16BE]
  do_bincmp_test $testname.5.be.be $utf16_sqlite3 $utf16be

  # Step 5 (big endian to little endian).
  set utf16_sqlite3 [test_translate $utf16be_bom UTF16 UTF16LE]
  do_bincmp_test $testname.5.be.le $utf16_sqlite3 $utf16le

  # Step 5 (little endian to big endian).
  set utf16_sqlite3 [test_translate $utf16le_bom UTF16 UTF16BE]
  do_bincmp_test $testname.5.le.be $utf16_sqlite3 $utf16be
}

translate_selftest

test_conversion enc-1 "hello world"
test_conversion enc-2 "sqlite"
test_conversion enc-3 ""
test_conversion enc-X "\u0100"
test_conversion enc-4 "\u1234"
test_conversion enc-5 "\u4321abc"
test_conversion enc-6 "\u4321\u1234"
test_conversion enc-7 [string repeat "abcde\u00EF\u00EE\uFFFCabc" 100]
test_conversion enc-8 [string repeat "\u007E\u007F\u0080\u0081" 100]
test_conversion enc-9 [string repeat "\u07FE\u07FF\u0800\u0801\uFFF0" 100]
test_conversion enc-10 [string repeat "\uE000" 100]

proc test_collate {enc zLeft zRight} {
  return [string compare $zLeft $zRight]
}
add_test_collate $::DB 0 0 1
do_test enc-11.1 {
  execsql {
    CREATE TABLE ab(a COLLATE test_collate, b);
    INSERT INTO ab VALUES(CAST (X'C388' AS TEXT), X'888800');
    INSERT INTO ab VALUES(CAST (X'C0808080808080808080808080808080808080808080808080808080808080808080808080808080808080808080808080808080808388' AS TEXT), X'888800');
    CREATE INDEX ab_i ON ab(a, b);
  }
} {}
do_test enc-11.2 {
  set cp200 "\u00C8"
  execsql {
    SELECT count(*) FROM ab WHERE a = $::cp200;
  }
} {2}

#-------------------------------------------------------------------------
reset_db
forcedelete test.db2
forcedelete test.db3

do_execsql_test enc-12.0 {
  PRAGMA encoding = 'utf-8';
  CREATE TABLE t1(a, b, c);
  INSERT INTO t1 VALUES('a', 'b', 'c');
  ATTACH 'test.db3' AS aux;
  CREATE TABLE aux.t3(x, y, z);
  INSERT INTO t3 VALUES('xxx', 'yyy', 'zzz');
  PRAGMA encoding;
} {UTF-8}

do_test enc-12.1 {
  sqlite3 db2 test.db2
  db2 eval {
    PRAGMA encoding = 'UTF-16le';
    CREATE TABLE t2(d, e, f);
    INSERT INTO t2 VALUES('d', 'e', 'f');
    PRAGMA encoding;
  }
} {UTF-16le}

do_test enc-12.2 {
  db2 backup test.db
  db2 close
} {}

do_catchsql_test enc-12.3 {
  SELECT * FROM t2;
} {1 {attached databases must use the same text encoding as main database}}

db close
sqlite3 db test.db3
do_execsql_test enc-12.4 {
  SELECT * FROM t3;
  PRAGMA encoding = 'UTF-16le';
  SELECT * FROM t3;
} {xxx yyy zzz xxx yyy zzz}

db close
sqlite3 db test.db3
breakpoint
do_execsql_test enc-12.5 {
  PRAGMA encoding = 'UTF-16le';
  PRAGMA encoding;
} {UTF-8}

reset_db
do_execsql_test enc-12.6 {
  PRAGMA encoding = 'UTF-8';
  CREATE TEMP TABLE t1(a, b, c);
  INSERT INTO t1 VALUES('xxx', 'yyy', 'zzz');
}
do_test enc-12.7 {
  sqlite3 db2 test.db2
  db2 backup test.db
  db2 close
  db eval {
    SELECT * FROM t1;
  }
} {xxx yyy zzz}
do_catchsql_test enc-12.8 {
  SELECT * FROM t2;
  SELECT * FROM t1;
} {1 {attached databases must use the same text encoding as main database}}

db close
sqlite3 db test.db
do_execsql_test enc-12.9 {
  CREATE TEMP TABLE t1(a, b, c);
  INSERT INTO t1 VALUES('xxx', 'yyy', 'zzz');
}
do_execsql_test enc-12.10 {
  SELECT * FROM t2;
  SELECT * FROM t1;
} {d e f xxx yyy zzz}

finish_test