includes/normal/CleanUpTest.php

   1 <?php
   2 # Copyright (C) 2004 Brion Vibber <brion@pobox.com>
   3 # http://www.mediawiki.org/
   4 #
   5 # This program is free software; you can redistribute it and/or modify
   6 # it under the terms of the GNU General Public License as published by
   7 # the Free Software Foundation; either version 2 of the License, or
   8 # (at your option) any later version.
   9 #
  10 # This program is distributed in the hope that it will be useful,
  11 # but WITHOUT ANY WARRANTY; without even the implied warranty of
  12 # MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
  13 # GNU General Public License for more details.
  14 #
  15 # You should have received a copy of the GNU General Public License along
  16 # with this program; if not, write to the Free Software Foundation, Inc.,
  17 # 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA.
  18 # http://www.gnu.org/copyleft/gpl.html
  19
  20
  21 if( php_sapi_name() != 'cli' ) {
  22         die( "Run me from the command line please.\n" );
  23 }
  24
  25 /** */
  26 if( isset( $_SERVER['argv'] ) && in_array( '--icu', $_SERVER['argv'] ) ) {
  27         dl( 'php_utfnormal.so' );
  28 }
  29
  30 #ini_set( 'memory_limit', '40M' );
  31
  32 require_once 'PHPUnit/Framework.php';
  33 require_once 'PHPUnit/TextUI/TestRunner.php';
  34
  35 require_once 'UtfNormal.php';
  36
  37 /**
  38  * Additional tests for UtfNormal::cleanUp() function, inclusion
  39  * regression checks for known problems.
  40  * Requires PHPUnit.
  41  *
  42  * @ingroup UtfNormal
  43  * @private
  44  */
  45 class CleanUpTest extends PHPUnit_Framework_TestCase {
  46         /** @todo document */
  47         function setUp() {
  48         }
  49
  50         /** @todo document */
  51         function tearDown() {
  52         }
  53
  54         /** @todo document */
  55         function testAscii() {
  56                 $text = 'This is plain ASCII text.';
  57                 $this->assertEquals( $text, UtfNormal::cleanUp( $text ) );
  58         }
  59
  60         /** @todo document */
  61         function testNull() {
  62                 $text = "a \x00 null";
  63                 $expect = "a \xef\xbf\xbd null";
  64                 $this->assertEquals(
  65                         bin2hex( $expect ),
  66                         bin2hex( UtfNormal::cleanUp( $text ) ) );
  67         }
  68
  69         /** @todo document */
  70         function testLatin() {
  71                 $text = "L'\xc3\xa9cole";
  72                 $this->assertEquals( $text, UtfNormal::cleanUp( $text ) );
  73         }
  74
  75         /** @todo document */
  76         function testLatinNormal() {
  77                 $text = "L'e\xcc\x81cole";
  78                 $expect = "L'\xc3\xa9cole";
  79                 $this->assertEquals( $expect, UtfNormal::cleanUp( $text ) );
  80         }
  81
  82         /**
  83          * This test is *very* expensive!
  84          * @todo document
  85          */
  86         function XtestAllChars() {
  87                 $rep = UTF8_REPLACEMENT;
  88                 for( $i = 0x0; $i < UNICODE_MAX; $i++ ) {
  89                         $char = codepointToUtf8( $i );
  90                         $clean = UtfNormal::cleanUp( $char );
  91                         $x = sprintf( "%04X", $i );
  92                         if( $i % 0x1000 == 0 ) echo "U+$x\n";
  93                         if( $i == 0x0009 ||
  94                             $i == 0x000a ||
  95                             $i == 0x000d ||
  96                             ($i > 0x001f && $i < UNICODE_SURROGATE_FIRST) ||
  97                             ($i > UNICODE_SURROGATE_LAST && $i < 0xfffe ) ||
  98                             ($i > 0xffff && $i <= UNICODE_MAX ) ) {
  99                                 if( isset( UtfNormal::$utfCanonicalComp[$char] ) || isset( UtfNormal::$utfCanonicalDecomp[$char] ) ) {
 100                                     $comp = UtfNormal::NFC( $char );
 101                                         $this->assertEquals(
 102                                                 bin2hex( $comp ),
 103                                                 bin2hex( $clean ),
 104                                                 "U+$x should be decomposed" );
 105                                 } else {
 106                                         $this->assertEquals(
 107                                                 bin2hex( $char ),
 108                                                 bin2hex( $clean ),
 109                                                 "U+$x should be intact" );
 110                                 }
 111                         } else {
 112                                 $this->assertEquals( bin2hex( $rep ), bin2hex( $clean ), $x );
 113                         }
 114                 }
 115         }
 116
 117         /** @todo document */
 118         function testAllBytes() {
 119                 $this->doTestBytes( '', '' );
 120                 $this->doTestBytes( 'x', '' );
 121                 $this->doTestBytes( '', 'x' );
 122                 $this->doTestBytes( 'x', 'x' );
 123         }
 124
 125         /** @todo document */
 126         function doTestBytes( $head, $tail ) {
 127                 for( $i = 0x0; $i < 256; $i++ ) {
 128                         $char = $head . chr( $i ) . $tail;
 129                         $clean = UtfNormal::cleanUp( $char );
 130                         $x = sprintf( "%02X", $i );
 131                         if( $i == 0x0009 ||
 132                             $i == 0x000a ||
 133                             $i == 0x000d ||
 134                             ($i > 0x001f && $i < 0x80) ) {
 135                                 $this->assertEquals(
 136                                         bin2hex( $char ),
 137                                         bin2hex( $clean ),
 138                                         "ASCII byte $x should be intact" );
 139                                 if( $char != $clean ) return;
 140                         } else {
 141                                 $norm = $head . UTF8_REPLACEMENT . $tail;
 142                                 $this->assertEquals(
 143                                         bin2hex( $norm ),
 144                                         bin2hex( $clean ),
 145                                         "Forbidden byte $x should be rejected" );
 146                                 if( $norm != $clean ) return;
 147                         }
 148                 }
 149         }
 150
 151         /** @todo document */
 152         function testDoubleBytes() {
 153                 $this->doTestDoubleBytes( '', '' );
 154                 $this->doTestDoubleBytes( 'x', '' );
 155                 $this->doTestDoubleBytes( '', 'x' );
 156                 $this->doTestDoubleBytes( 'x', 'x' );
 157         }
 158
 159         /**
 160          * @todo document
 161          */
 162         function doTestDoubleBytes( $head, $tail ) {
 163                 for( $first = 0xc0; $first < 0x100; $first++ ) {
 164                         for( $second = 0x80; $second < 0x100; $second++ ) {
 165                                 $char = $head . chr( $first ) . chr( $second ) . $tail;
 166                                 $clean = UtfNormal::cleanUp( $char );
 167                                 $x = sprintf( "%02X,%02X", $first, $second );
 168                                 if( $first > 0xc1 &&
 169                                     $first < 0xe0 &&
 170                                     $second < 0xc0 ) {
 171                                     $norm = UtfNormal::NFC( $char );
 172                                         $this->assertEquals(
 173                                                 bin2hex( $norm ),
 174                                                 bin2hex( $clean ),
 175                                                 "Pair $x should be intact" );
 176                                     if( $norm != $clean ) return;
 177                                 } elseif( $first > 0xfd || $second > 0xbf ) {
 178                                         # fe and ff are not legal head bytes -- expect two replacement chars
 179                                         $norm = $head . UTF8_REPLACEMENT . UTF8_REPLACEMENT . $tail;
 180                                         $this->assertEquals(
 181                                                 bin2hex( $norm ),
 182                                                 bin2hex( $clean ),
 183                                                 "Forbidden pair $x should be rejected" );
 184                                         if( $norm != $clean ) return;
 185                                 } else {
 186                                         $norm = $head . UTF8_REPLACEMENT . $tail;
 187                                         $this->assertEquals(
 188                                                 bin2hex( $norm ),
 189                                                 bin2hex( $clean ),
 190                                                 "Forbidden pair $x should be rejected" );
 191                                         if( $norm != $clean ) return;
 192                                 }
 193                         }
 194                 }
 195         }
 196
 197         /** @todo document */
 198         function testTripleBytes() {
 199                 $this->doTestTripleBytes( '', '' );
 200                 $this->doTestTripleBytes( 'x', '' );
 201                 $this->doTestTripleBytes( '', 'x' );
 202                 $this->doTestTripleBytes( 'x', 'x' );
 203         }
 204
 205         /** @todo document */
 206         function doTestTripleBytes( $head, $tail ) {
 207                 for( $first = 0xc0; $first < 0x100; $first++ ) {
 208                         for( $second = 0x80; $second < 0x100; $second++ ) {
 209                                 #for( $third = 0x80; $third < 0x100; $third++ ) {
 210                                 for( $third = 0x80; $third < 0x81; $third++ ) {
 211                                         $char = $head . chr( $first ) . chr( $second ) . chr( $third ) . $tail;
 212                                         $clean = UtfNormal::cleanUp( $char );
 213                                         $x = sprintf( "%02X,%02X,%02X", $first, $second, $third );
 214                                         if( $first >= 0xe0 &&
 215                                                 $first < 0xf0 &&
 216                                                 $second < 0xc0 &&
 217                                                 $third < 0xc0 ) {
 218                                                 if( $first == 0xe0 && $second < 0xa0 ) {
 219                                                         $this->assertEquals(
 220                                                                 bin2hex( $head . UTF8_REPLACEMENT . $tail ),
 221                                                                 bin2hex( $clean ),
 222                                                                 "Overlong triplet $x should be rejected" );
 223                                                 } elseif( $first == 0xed &&
 224                                                         ( chr( $first ) . chr( $second ) . chr( $third ))  >= UTF8_SURROGATE_FIRST ) {
 225                                                         $this->assertEquals(
 226                                                                 bin2hex( $head . UTF8_REPLACEMENT . $tail ),
 227                                                                 bin2hex( $clean ),
 228                                                                 "Surrogate triplet $x should be rejected" );
 229                                                 } else {
 230                                                         $this->assertEquals(
 231                                                                 bin2hex( UtfNormal::NFC( $char ) ),
 232                                                                 bin2hex( $clean ),
 233                                                                 "Triplet $x should be intact" );
 234                                                 }
 235                                         } elseif( $first > 0xc1 && $first < 0xe0 && $second < 0xc0 ) {
 236                                                 $this->assertEquals(
 237                                                         bin2hex( UtfNormal::NFC( $head . chr( $first ) . chr( $second ) ) . UTF8_REPLACEMENT . $tail ),
 238                                                         bin2hex( $clean ),
 239                                                         "Valid 2-byte $x + broken tail" );
 240                                         } elseif( $second > 0xc1 && $second < 0xe0 && $third < 0xc0 ) {
 241                                                 $this->assertEquals(
 242                                                         bin2hex( $head . UTF8_REPLACEMENT . UtfNormal::NFC( chr( $second ) . chr( $third ) . $tail ) ),
 243                                                         bin2hex( $clean ),
 244                                                         "Broken head + valid 2-byte $x" );
 245                                         } elseif( ( $first > 0xfd || $second > 0xfd ) &&
 246                                                     ( ( $second > 0xbf && $third > 0xbf ) ||
 247                                                       ( $second < 0xc0 && $third < 0xc0 ) ||
 248                                                       ( $second > 0xfd ) ||
 249                                                       ( $third > 0xfd ) ) ) {
 250                                                 # fe and ff are not legal head bytes -- expect three replacement chars
 251                                                 $this->assertEquals(
 252                                                         bin2hex( $head . UTF8_REPLACEMENT . UTF8_REPLACEMENT . UTF8_REPLACEMENT . $tail ),
 253                                                         bin2hex( $clean ),
 254                                                         "Forbidden triplet $x should be rejected" );
 255                                         } elseif( $first > 0xc2 && $second < 0xc0 && $third < 0xc0 ) {
 256                                                 $this->assertEquals(
 257                                                         bin2hex( $head . UTF8_REPLACEMENT . $tail ),
 258                                                         bin2hex( $clean ),
 259                                                         "Forbidden triplet $x should be rejected" );
 260                                         } else {
 261                                                 $this->assertEquals(
 262                                                         bin2hex( $head . UTF8_REPLACEMENT . UTF8_REPLACEMENT . $tail ),
 263                                                         bin2hex( $clean ),
 264                                                         "Forbidden triplet $x should be rejected" );
 265                                         }
 266                                 }
 267                         }
 268                 }
 269         }
 270
 271         /** @todo document */
 272         function testChunkRegression() {
 273                 # Check for regression against a chunking bug
 274                 $text   = "\x46\x55\xb8" .
 275                           "\xdc\x96" .
 276                           "\xee" .
 277                           "\xe7" .
 278                           "\x44" .
 279                           "\xaa" .
 280                           "\x2f\x25";
 281                 $expect = "\x46\x55\xef\xbf\xbd" .
 282                           "\xdc\x96" .
 283                           "\xef\xbf\xbd" .
 284                           "\xef\xbf\xbd" .
 285                           "\x44" .
 286                           "\xef\xbf\xbd" .
 287                           "\x2f\x25";
 288
 289                 $this->assertEquals(
 290                         bin2hex( $expect ),
 291                         bin2hex( UtfNormal::cleanUp( $text ) ) );
 292         }
 293
 294         /** @todo document */
 295         function testInterposeRegression() {
 296                 $text   = "\x4e\x30" .
 297                           "\xb1" .              # bad tail
 298                           "\x3a" .
 299                           "\x92" .              # bad tail
 300                           "\x62\x3a" .
 301                           "\x84" .              # bad tail
 302                           "\x43" .
 303                           "\xc6" .              # bad head
 304                           "\x3f" .
 305                           "\x92" .              # bad tail
 306                           "\xad" .              # bad tail
 307                           "\x7d" .
 308                           "\xd9\x95";
 309
 310                 $expect = "\x4e\x30" .
 311                           "\xef\xbf\xbd" .
 312                           "\x3a" .
 313                           "\xef\xbf\xbd" .
 314                           "\x62\x3a" .
 315                           "\xef\xbf\xbd" .
 316                           "\x43" .
 317                           "\xef\xbf\xbd" .
 318                           "\x3f" .
 319                           "\xef\xbf\xbd" .
 320                           "\xef\xbf\xbd" .
 321                           "\x7d" .
 322                           "\xd9\x95";
 323
 324                 $this->assertEquals(
 325                         bin2hex( $expect ),
 326                         bin2hex( UtfNormal::cleanUp( $text ) ) );
 327         }
 328
 329         /** @todo document */
 330         function testOverlongRegression() {
 331                 $text   = "\x67" .
 332                           "\x1a" . # forbidden ascii
 333                           "\xea" . # bad head
 334                           "\xc1\xa6" . # overlong sequence
 335                           "\xad" . # bad tail
 336                           "\x1c" . # forbidden ascii
 337                           "\xb0" . # bad tail
 338                           "\x3c" .
 339                           "\x9e";  # bad tail
 340                 $expect = "\x67" .
 341                           "\xef\xbf\xbd" .
 342                           "\xef\xbf\xbd" .
 343                           "\xef\xbf\xbd" .
 344                           "\xef\xbf\xbd" .
 345                           "\xef\xbf\xbd" .
 346                           "\xef\xbf\xbd" .
 347                           "\x3c" .
 348                           "\xef\xbf\xbd";
 349                 $this->assertEquals(
 350                         bin2hex( $expect ),
 351                         bin2hex( UtfNormal::cleanUp( $text ) ) );
 352         }
 353
 354         /** @todo document */
 355         function testSurrogateRegression() {
 356                 $text   = "\xed\xb4\x96" . # surrogate 0xDD16
 357                           "\x83" . # bad tail
 358                           "\xb4" . # bad tail
 359                           "\xac";  # bad head
 360                 $expect = "\xef\xbf\xbd" .
 361                           "\xef\xbf\xbd" .
 362                           "\xef\xbf\xbd" .
 363                           "\xef\xbf\xbd";
 364                 $this->assertEquals(
 365                         bin2hex( $expect ),
 366                         bin2hex( UtfNormal::cleanUp( $text ) ) );
 367         }
 368
 369         /** @todo document */
 370         function testBomRegression() {
 371                 $text   = "\xef\xbf\xbe" . # U+FFFE, illegal char
 372                           "\xb2" . # bad tail
 373                           "\xef" . # bad head
 374                           "\x59";
 375                 $expect = "\xef\xbf\xbd" .
 376                           "\xef\xbf\xbd" .
 377                           "\xef\xbf\xbd" .
 378                           "\x59";
 379                 $this->assertEquals(
 380                         bin2hex( $expect ),
 381                         bin2hex( UtfNormal::cleanUp( $text ) ) );
 382         }
 383
 384         /** @todo document */
 385         function testForbiddenRegression() {
 386                 $text   = "\xef\xbf\xbf"; # U+FFFF, illegal char
 387                 $expect = "\xef\xbf\xbd";
 388                 $this->assertEquals(
 389                         bin2hex( $expect ),
 390                         bin2hex( UtfNormal::cleanUp( $text ) ) );
 391         }
 392
 393         /** @todo document */
 394         function testHangulRegression() {
 395                 $text = "\xed\x9c\xaf" . # Hangul char
 396                                 "\xe1\x87\x81";  # followed by another final jamo
 397                 $expect = $text;         # Should *not* change.
 398                 $this->assertEquals(
 399                         bin2hex( $expect ),
 400                         bin2hex( UtfNormal::cleanUp( $text ) ) );
 401         }
 402 }
 403
 404
 405 $suite = new PHPUnit_Framework_TestSuite( 'CleanUpTest' );
 406 $result = PHPUnit_TextUI_TestRunner::run( $suite );
 407
 408 if( !$result->wasSuccessful() ) {
 409         exit( -1 );
 410 }
 411 exit( 0 );