* @defgroup UtfNormal UtfNormal
*/
-require_once dirname(__FILE__).'/UtfNormalUtil.php';
-
-/**
- * For using the ICU wrapper
- */
-define( 'UNORM_NONE', 1 );
-define( 'UNORM_NFD', 2 );
-define( 'UNORM_NFKD', 3 );
-define( 'UNORM_NFC', 4 );
-define( 'UNORM_DEFAULT', UNORM_NFC );
-define( 'UNORM_NFKC', 5 );
-define( 'UNORM_FCD', 6 );
-
define( 'NORMALIZE_ICU', function_exists( 'utf8_normalize' ) );
define( 'NORMALIZE_INTL', function_exists( 'normalizer_normalize' ) );
* @ingroup UtfNormal
*/
class UtfNormal {
+ /**
+ * For using the ICU wrapper
+ */
+ const UNORM_NONE = 1;
+ const UNORM_NFD = 2;
+ const UNORM_NFKD = 3;
+ const UNORM_NFC = 4;
+ const UNORM_NFKC = 5;
+ const UNORM_FCD = 6;
+ const UNORM_DEFAULT = self::UNORM_NFC;
+
static $utfCombiningClass = null;
static $utfCanonicalComp = null;
static $utfCanonicalDecomp = null;
# UnicodeString constructor fails if the string ends with a
# head byte. Add a junk char at the end, we'll strip it off.
- return rtrim( utf8_normalize( $string . "\x01", UNORM_NFC ), "\x01" );
+ return rtrim( utf8_normalize( $string . "\x01", self::UNORM_NFC ), "\x01" );
} elseif( NORMALIZE_INTL ) {
$string = self::replaceForNativeNormalize( $string );
$norm = normalizer_normalize( $string, Normalizer::FORM_C );
if( NORMALIZE_INTL )
return normalizer_normalize( $string, Normalizer::FORM_C );
elseif( NORMALIZE_ICU )
- return utf8_normalize( $string, UNORM_NFC );
+ return utf8_normalize( $string, self::UNORM_NFC );
elseif( UtfNormal::quickIsNFC( $string ) )
return $string;
else
if( NORMALIZE_INTL )
return normalizer_normalize( $string, Normalizer::FORM_D );
elseif( NORMALIZE_ICU )
- return utf8_normalize( $string, UNORM_NFD );
+ return utf8_normalize( $string, self::UNORM_NFD );
elseif( preg_match( '/[\x80-\xff]/', $string ) )
return UtfNormal::NFD( $string );
else
if( NORMALIZE_INTL )
return normalizer_normalize( $string, Normalizer::FORM_KC );
elseif( NORMALIZE_ICU )
- return utf8_normalize( $string, UNORM_NFKC );
+ return utf8_normalize( $string, self::UNORM_NFKC );
elseif( preg_match( '/[\x80-\xff]/', $string ) )
return UtfNormal::NFKC( $string );
else
if( NORMALIZE_INTL )
return normalizer_normalize( $string, Normalizer::FORM_KD );
elseif( NORMALIZE_ICU )
- return utf8_normalize( $string, UNORM_NFKD );
+ return utf8_normalize( $string, self::UNORM_NFKD );
elseif( preg_match( '/[\x80-\xff]/', $string ) )
return UtfNormal::NFKD( $string );
else
UtfNormal::loadData();
$len = strlen( $string );
for( $i = 0; $i < $len; $i++ ) {
- $c = $string{$i};
+ $c = $string[$i];
$n = ord( $c );
if( $n < 0x80 ) {
continue;
* Returns true if the string is _definitely_ in NFC.
* Returns false if not or uncertain.
* @param $string String: a UTF-8 string, altered on output to be valid UTF-8 safe for XML.
+ * @return bool
*/
static function quickIsNFCVerify( &$string ) {
# Screen out some characters that eg won't be allowed in XML
foreach( $matches[1] as $str ) {
$chunk = strlen( $str );
- if( $str{0} < "\x80" ) {
+ if( $str[0] < "\x80" ) {
# ASCII chunk: guaranteed to be valid UTF-8
# and in normal form C, so skip over it.
$base += $chunk;
$len = $chunk + 1; # Counting down is faster. I'm *so* sorry.
for( $i = -1; --$len; ) {
- $remaining = $tailBytes[$c = $str{++$i}];
+ $remaining = $tailBytes[$c = $str[++$i]];
if( $remaining ) {
# UTF-8 head byte!
$sequence = $head = $c;
do {
# Look for the defined number of tail bytes...
- if( --$len && ( $c = $str{++$i} ) >= "\x80" && $c < "\xc0" ) {
+ if( --$len && ( $c = $str[++$i] ) >= "\x80" && $c < "\xc0" ) {
# Legal tail bytes are nice.
$sequence .= $c;
} else {
$len = strlen( $string );
$out = '';
for( $i = 0; $i < $len; $i++ ) {
- $c = $string{$i};
+ $c = $string[$i];
$n = ord( $c );
if( $n < 0x80 ) {
# ASCII chars never decompose
# A lookup table would be slightly faster,
# but adds a lot of memory & disk needs.
#
- $index = ( (ord( $c{0} ) & 0x0f) << 12
- | (ord( $c{1} ) & 0x3f) << 6
- | (ord( $c{2} ) & 0x3f) )
+ $index = ( (ord( $c[0] ) & 0x0f) << 12
+ | (ord( $c[1] ) & 0x3f) << 6
+ | (ord( $c[2] ) & 0x3f) )
- UNICODE_HANGUL_FIRST;
$l = intval( $index / UNICODE_HANGUL_NCOUNT );
$v = intval( ($index % UNICODE_HANGUL_NCOUNT) / UNICODE_HANGUL_TCOUNT);
$combiners = array();
$lastClass = -1;
for( $i = 0; $i < $len; $i++ ) {
- $c = $string{$i};
+ $c = $string[$i];
$n = ord( $c );
if( $n >= 0x80 ) {
if( $n >= 0xf0 ) {
$x1 = ord(substr(UTF8_HANGUL_VBASE,0,1));
$x2 = ord(substr(UTF8_HANGUL_TEND,0,1));
for( $i = 0; $i < $len; $i++ ) {
- $c = $string{$i};
+ $c = $string[$i];
$n = ord( $c );
if( $n < 0x80 ) {
# No combining characters here...
#
#$lIndex = utf8ToCodepoint( $startChar ) - UNICODE_HANGUL_LBASE;
#$vIndex = utf8ToCodepoint( $c ) - UNICODE_HANGUL_VBASE;
- $lIndex = ord( $startChar{2} ) - 0x80;
- $vIndex = ord( $c{2} ) - 0xa1;
+ $lIndex = ord( $startChar[2] ) - 0x80;
+ $vIndex = ord( $c[2] ) - 0xa1;
$hangulPoint = UNICODE_HANGUL_FIRST +
UNICODE_HANGUL_TCOUNT *
$startChar <= UTF8_HANGUL_LAST &&
!$lastHangul ) {
# $tIndex = utf8ToCodepoint( $c ) - UNICODE_HANGUL_TBASE;
- $tIndex = ord( $c{2} ) - 0xa7;
- if( $tIndex < 0 ) $tIndex = ord( $c{2} ) - 0x80 + (0x11c0 - 0x11a7);
+ $tIndex = ord( $c[2] ) - 0xa7;
+ if( $tIndex < 0 ) $tIndex = ord( $c[2] ) - 0x80 + (0x11c0 - 0x11a7);
# Increment the code point by $tIndex, without
# the function overhead of decoding and recoding UTF-8
#
- $tail = ord( $startChar{2} ) + $tIndex;
+ $tail = ord( $startChar[2] ) + $tIndex;
if( $tail > 0xbf ) {
$tail -= 0x40;
- $mid = ord( $startChar{1} ) + 1;
+ $mid = ord( $startChar[1] ) + 1;
if( $mid > 0xbf ) {
- $startChar{0} = chr( ord( $startChar{0} ) + 1 );
+ $startChar[0] = chr( ord( $startChar[0] ) + 1 );
$mid -= 0x40;
}
- $startChar{1} = chr( $mid );
+ $startChar[1] = chr( $mid );
}
- $startChar{2} = chr( $tail );
+ $startChar[2] = chr( $tail );
# If there's another jamo char after this, *don't* try to merge it.
$lastHangul = 1;
$len = strlen( $string );
$out = '';
for( $i = 0; $i < $len; $i++ ) {
- $out .= $string{$i};
+ $out .= $string[$i];
}
return $out;
}
dépôts / lhc / web / wiklou.git / blobdiff