MediaWiki
REL1_19
|
00001 <?php 00030 require_once 'UtfNormalDefines.php'; 00031 require_once 'UtfNormalUtil.php'; 00032 require_once 'UtfNormal.php'; 00033 mb_internal_encoding( "utf-8" ); 00034 00035 $verbose = false; 00036 #$verbose = true; 00037 if( php_sapi_name() != 'cli' ) { 00038 die( "Run me from the command line please.\n" ); 00039 } 00040 00041 $in = fopen( "UTF-8-test.txt", "rt" ); 00042 if( !$in ) { 00043 print "Couldn't open UTF-8-test.txt -- can't run tests.\n"; 00044 print "If necessary, manually download this file. It can be obtained at\n"; 00045 print "http://www.cl.cam.ac.uk/~mgk25/ucs/examples/UTF-8-test.txt\n"; 00046 exit(-1); 00047 } 00048 00049 $columns = 0; 00050 while( false !== ( $line = fgets( $in ) ) ) { 00051 $matches = array(); 00052 if( preg_match( '/^(Here come the tests:\s*)\|$/', $line, $matches ) ) { 00053 $columns = strpos( $line, '|' ); 00054 break; 00055 } 00056 } 00057 00058 if( !$columns ) { 00059 print "Something seems to be wrong; couldn't extract line length.\n"; 00060 print "Check that UTF-8-test.txt was downloaded correctly from\n"; 00061 print "http://www.cl.cam.ac.uk/~mgk25/ucs/examples/UTF-8-test.txt\n"; 00062 exit(-1); 00063 } 00064 00065 # print "$columns\n"; 00066 00067 $ignore = array( 00068 # These two lines actually seem to be corrupt 00069 '2.1.1', '2.2.1' ); 00070 00071 $exceptions = array( 00072 # Tests that should mark invalid characters due to using long 00073 # sequences beyond what is now considered legal. 00074 '2.1.5', '2.1.6', '2.2.4', '2.2.5', '2.2.6', '2.3.5', 00075 00076 # Literal 0xffff, which is illegal 00077 '2.2.3' ); 00078 00079 $longTests = array( 00080 # These tests span multiple lines 00081 '3.1.9', '3.2.1', '3.2.2', '3.2.3', '3.2.4', '3.2.5', 00082 '3.4' ); 00083 00084 # These tests are not in proper subsections 00085 $sectionTests = array( '3.4' ); 00086 00087 $section = null; 00088 $test = ''; 00089 $failed = 0; 00090 $success = 0; 00091 $total = 0; 00092 while( false !== ( $line = fgets( $in ) ) ) { 00093 $matches = array(); 00094 if( preg_match( '/^(\d+)\s+(.*?)\s*\|/', $line, $matches ) ) { 00095 $section = $matches[1]; 00096 print $line; 00097 continue; 00098 } 00099 if( preg_match( '/^(\d+\.\d+\.\d+)\s*/', $line, $matches ) ) { 00100 $test = $matches[1]; 00101 00102 if( in_array( $test, $ignore ) ) { 00103 continue; 00104 } 00105 if( in_array( $test, $longTests ) ) { 00106 $line = fgets( $in ); 00107 for( $line = fgets( $in ); !preg_match( '/^\s+\|/', $line ); $line = fgets( $in ) ) { 00108 testLine( $test, $line, $total, $success, $failed, $columns, $exceptions, $verbose ); 00109 } 00110 } else { 00111 testLine( $test, $line, $total, $success, $failed, $columns, $exceptions, $verbose ); 00112 } 00113 } 00114 } 00115 00116 if( $failed ) { 00117 echo "\nFailed $failed tests.\n"; 00118 echo "UTF-8 DECODER TEST FAILED\n"; 00119 exit (-1); 00120 } 00121 00122 echo "UTF-8 DECODER TEST SUCCESS!\n"; 00123 exit (0); 00124 00125 00126 function testLine( $test, $line, &$total, &$success, &$failed, $columns, $exceptions, $verbose ) { 00127 $stripped = $line; 00128 UtfNormal::quickisNFCVerify( $stripped ); 00129 00130 $same = ( $line == $stripped ); 00131 $len = mb_strlen( substr( $stripped, 0, strpos( $stripped, '|' ) ) ); 00132 if( $len == 0 ) { 00133 $len = strlen( substr( $stripped, 0, strpos( $stripped, '|' ) ) ); 00134 } 00135 00136 $ok = $same ^ ($test >= 3 ); 00137 00138 $ok ^= in_array( $test, $exceptions ); 00139 00140 $ok &= ($columns == $len); 00141 00142 $total++; 00143 if( $ok ) { 00144 $success++; 00145 } else { 00146 $failed++; 00147 } 00148 00149 if( $verbose || !$ok ) { 00150 print str_replace( "\n", "$len\n", $stripped ); 00151 } 00152 }