<?php
-# Copyright (C) 2004 Brion Vibber <brion@pobox.com>
-# http://www.mediawiki.org/
-#
-# This program is free software; you can redistribute it and/or modify
-# it under the terms of the GNU General Public License as published by
-# the Free Software Foundation; either version 2 of the License, or
-# (at your option) any later version.
-#
-# This program is distributed in the hope that it will be useful,
-# but WITHOUT ANY WARRANTY; without even the implied warranty of
-# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
-# GNU General Public License for more details.
-#
-# You should have received a copy of the GNU General Public License along
-# with this program; if not, write to the Free Software Foundation, Inc.,
-# 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA.
-# http://www.gnu.org/copyleft/gpl.html
-
-/** */
-require_once dirname(__FILE__).'/UtfNormalUtil.php';
-
-global $utfCombiningClass, $utfCanonicalComp, $utfCanonicalDecomp;
-$utfCombiningClass = NULL;
-$utfCanonicalComp = NULL;
-$utfCanonicalDecomp = NULL;
-
-# Load compatibility decompositions on demand if they are needed.
-global $utfCompatibilityDecomp;
-$utfCompatibilityDecomp = NULL;
-
-define( 'UNICODE_HANGUL_FIRST', 0xac00 );
-define( 'UNICODE_HANGUL_LAST', 0xd7a3 );
-
-define( 'UNICODE_HANGUL_LBASE', 0x1100 );
-define( 'UNICODE_HANGUL_VBASE', 0x1161 );
-define( 'UNICODE_HANGUL_TBASE', 0x11a7 );
-
-define( 'UNICODE_HANGUL_LCOUNT', 19 );
-define( 'UNICODE_HANGUL_VCOUNT', 21 );
-define( 'UNICODE_HANGUL_TCOUNT', 28 );
-define( 'UNICODE_HANGUL_NCOUNT', UNICODE_HANGUL_VCOUNT * UNICODE_HANGUL_TCOUNT );
-
-define( 'UNICODE_HANGUL_LEND', UNICODE_HANGUL_LBASE + UNICODE_HANGUL_LCOUNT - 1 );
-define( 'UNICODE_HANGUL_VEND', UNICODE_HANGUL_VBASE + UNICODE_HANGUL_VCOUNT - 1 );
-define( 'UNICODE_HANGUL_TEND', UNICODE_HANGUL_TBASE + UNICODE_HANGUL_TCOUNT - 1 );
-
-define( 'UNICODE_SURROGATE_FIRST', 0xd800 );
-define( 'UNICODE_SURROGATE_LAST', 0xdfff );
-define( 'UNICODE_MAX', 0x10ffff );
-define( 'UNICODE_REPLACEMENT', 0xfffd );
-
-
-define( 'UTF8_HANGUL_FIRST', "\xea\xb0\x80" /*codepointToUtf8( UNICODE_HANGUL_FIRST )*/ );
-define( 'UTF8_HANGUL_LAST', "\xed\x9e\xa3" /*codepointToUtf8( UNICODE_HANGUL_LAST )*/ );
-
-define( 'UTF8_HANGUL_LBASE', "\xe1\x84\x80" /*codepointToUtf8( UNICODE_HANGUL_LBASE )*/ );
-define( 'UTF8_HANGUL_VBASE', "\xe1\x85\xa1" /*codepointToUtf8( UNICODE_HANGUL_VBASE )*/ );
-define( 'UTF8_HANGUL_TBASE', "\xe1\x86\xa7" /*codepointToUtf8( UNICODE_HANGUL_TBASE )*/ );
-
-define( 'UTF8_HANGUL_LEND', "\xe1\x84\x92" /*codepointToUtf8( UNICODE_HANGUL_LEND )*/ );
-define( 'UTF8_HANGUL_VEND', "\xe1\x85\xb5" /*codepointToUtf8( UNICODE_HANGUL_VEND )*/ );
-define( 'UTF8_HANGUL_TEND', "\xe1\x87\x82" /*codepointToUtf8( UNICODE_HANGUL_TEND )*/ );
-
-define( 'UTF8_SURROGATE_FIRST', "\xed\xa0\x80" /*codepointToUtf8( UNICODE_SURROGATE_FIRST )*/ );
-define( 'UTF8_SURROGATE_LAST', "\xed\xbf\xbf" /*codepointToUtf8( UNICODE_SURROGATE_LAST )*/ );
-define( 'UTF8_MAX', "\xf4\x8f\xbf\xbf" /*codepointToUtf8( UNICODE_MAX )*/ );
-define( 'UTF8_REPLACEMENT', "\xef\xbf\xbd" /*codepointToUtf8( UNICODE_REPLACEMENT )*/ );
-#define( 'UTF8_REPLACEMENT', '!' );
-
-define( 'UTF8_OVERLONG_A', "\xc1\xbf" );
-define( 'UTF8_OVERLONG_B', "\xe0\x9f\xbf" );
-define( 'UTF8_OVERLONG_C', "\xf0\x8f\xbf\xbf" );
-
-# These two ranges are illegal
-define( 'UTF8_FDD0', "\xef\xb7\x90" /*codepointToUtf8( 0xfdd0 )*/ );
-define( 'UTF8_FDEF', "\xef\xb7\xaf" /*codepointToUtf8( 0xfdef )*/ );
-define( 'UTF8_FFFE', "\xef\xbf\xbe" /*codepointToUtf8( 0xfffe )*/ );
-define( 'UTF8_FFFF', "\xef\xbf\xbf" /*codepointToUtf8( 0xffff )*/ );
+/**
+ * Unicode normalization routines
+ *
+ * Copyright © 2004 Brion Vibber <brion@pobox.com>
+ * http://www.mediawiki.org/
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License as published by
+ * the Free Software Foundation; either version 2 of the License, or
+ * (at your option) any later version.
+ *
+ * This program is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
+ * GNU General Public License for more details.
+ *
+ * You should have received a copy of the GNU General Public License along
+ * with this program; if not, write to the Free Software Foundation, Inc.,
+ * 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA.
+ * http://www.gnu.org/copyleft/gpl.html
+ *
+ * @file
+ * @ingroup UtfNormal
+ */
-define( 'UTF8_HEAD', false );
-define( 'UTF8_TAIL', true );
+/**
+ * @defgroup UtfNormal UtfNormal
+ */
+require_once dirname(__FILE__).'/UtfNormalUtil.php';
/**
* For using the ICU wrapper
define( 'UNORM_FCD', 6 );
define( 'NORMALIZE_ICU', function_exists( 'utf8_normalize' ) );
+define( 'NORMALIZE_INTL', function_exists( 'normalizer_normalize' ) );
/**
* Unicode normalization routines for working with UTF-8 strings.
*
* See description of forms at http://www.unicode.org/reports/tr15/
*
- * @addtogroup UtfNormal
+ * @ingroup UtfNormal
*/
class UtfNormal {
+ static $utfCombiningClass = null;
+ static $utfCanonicalComp = null;
+ static $utfCanonicalDecomp = null;
+
+ # Load compatibility decompositions on demand if they are needed.
+ static $utfCompatibilityDecomp = null;
+
+ static $utfCheckNFC;
+
/**
* The ultimate convenience function! Clean up invalid UTF-8 sequences,
* and convert to normal form C, canonical composition.
* Fast return for pure ASCII strings; some lesser optimizations for
* strings containing only known-good characters. Not as fast as toNFC().
*
- * @param string $string a UTF-8 string
+ * @param $string String: a UTF-8 string
* @return string a clean, shiny, normalized UTF-8 string
- * @static
*/
static function cleanUp( $string ) {
if( NORMALIZE_ICU ) {
# 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" );
+ } elseif( NORMALIZE_INTL ) {
+ $norm = normalizer_normalize( $string, Normalizer::FORM_C );
+ if( $norm === null || $norm === false ) {
+ # normalizer_normalize will either return false or null
+ # (depending on which doc you read) if invalid utf8 string.
+ # quickIsNFCVerify cleans up invalid sequences.
+
+ if( UtfNormal::quickIsNFCVerify( $string ) ) {
+ # if that's true, the string is actually already normal.
+ return $string;
+ } else {
+ # Now we are valid but non-normal
+ return normalizer_normalize( $string, Normalizer::FORM_C );
+ }
+ } else {
+ return $norm;
+ }
} elseif( UtfNormal::quickIsNFCVerify( $string ) ) {
# Side effect -- $string has had UTF-8 errors cleaned up.
return $string;
* Fast return for pure ASCII strings; some lesser optimizations for
* strings containing only known-good characters.
*
- * @param string $string a valid UTF-8 string. Input is not validated.
+ * @param $string String: a valid UTF-8 string. Input is not validated.
* @return string a UTF-8 string in normal form C
- * @static
*/
static function toNFC( $string ) {
- if( NORMALIZE_ICU )
+ if( NORMALIZE_INTL )
+ return normalizer_normalize( $string, Normalizer::FORM_C );
+ elseif( NORMALIZE_ICU )
return utf8_normalize( $string, UNORM_NFC );
elseif( UtfNormal::quickIsNFC( $string ) )
return $string;
* Convert a UTF-8 string to normal form D, canonical decomposition.
* Fast return for pure ASCII strings.
*
- * @param string $string a valid UTF-8 string. Input is not validated.
+ * @param $string String: a valid UTF-8 string. Input is not validated.
* @return string a UTF-8 string in normal form D
- * @static
*/
static function toNFD( $string ) {
- if( NORMALIZE_ICU )
+ if( NORMALIZE_INTL )
+ return normalizer_normalize( $string, Normalizer::FORM_D );
+ elseif( NORMALIZE_ICU )
return utf8_normalize( $string, UNORM_NFD );
elseif( preg_match( '/[\x80-\xff]/', $string ) )
return UtfNormal::NFD( $string );
* This may cause irreversible information loss, use judiciously.
* Fast return for pure ASCII strings.
*
- * @param string $string a valid UTF-8 string. Input is not validated.
+ * @param $string String: a valid UTF-8 string. Input is not validated.
* @return string a UTF-8 string in normal form KC
- * @static
*/
static function toNFKC( $string ) {
- if( NORMALIZE_ICU )
+ if( NORMALIZE_INTL )
+ return normalizer_normalize( $string, Normalizer::FORM_KC );
+ elseif( NORMALIZE_ICU )
return utf8_normalize( $string, UNORM_NFKC );
elseif( preg_match( '/[\x80-\xff]/', $string ) )
return UtfNormal::NFKC( $string );
* This may cause irreversible information loss, use judiciously.
* Fast return for pure ASCII strings.
*
- * @param string $string a valid UTF-8 string. Input is not validated.
+ * @param $string String: a valid UTF-8 string. Input is not validated.
* @return string a UTF-8 string in normal form KD
- * @static
*/
static function toNFKD( $string ) {
- if( NORMALIZE_ICU )
+ if( NORMALIZE_INTL )
+ return normalizer_normalize( $string, Normalizer::FORM_KD );
+ elseif( NORMALIZE_ICU )
return utf8_normalize( $string, UNORM_NFKD );
elseif( preg_match( '/[\x80-\xff]/', $string ) )
return UtfNormal::NFKD( $string );
/**
* Load the basic composition data if necessary
* @private
- * @static
*/
static function loadData() {
- global $utfCombiningClass;
- if( !isset( $utfCombiningClass ) ) {
+ if( !isset( self::$utfCombiningClass ) ) {
require_once( dirname(__FILE__) . '/UtfNormalData.inc' );
}
}
/**
* Returns true if the string is _definitely_ in NFC.
* Returns false if not or uncertain.
- * @param string $string a valid UTF-8 string. Input is not validated.
+ * @param $string String: a valid UTF-8 string. Input is not validated.
* @return bool
- * @static
*/
static function quickIsNFC( $string ) {
# ASCII is always valid NFC!
if( !preg_match( '/[\x80-\xff]/', $string ) ) return true;
UtfNormal::loadData();
- global $utfCheckNFC, $utfCombiningClass;
$len = strlen( $string );
for( $i = 0; $i < $len; $i++ ) {
$c = $string{$i};
$c = substr( $string, $i, 2 );
$i++;
}
- if( isset( $utfCheckNFC[$c] ) ) {
+ if( isset( self::$utfCheckNFC[$c] ) ) {
# If it's NO or MAYBE, bail and do the slow check.
return false;
}
- if( isset( $utfCombiningClass[$c] ) ) {
+ if( isset( self::$utfCombiningClass[$c] ) ) {
# Combining character? We might have to do sorting, at least.
return false;
}
/**
* 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.
- * @static
+ * @param $string String: a UTF-8 string, altered on output to be valid UTF-8 safe for XML.
*/
static function quickIsNFCVerify( &$string ) {
# Screen out some characters that eg won't be allowed in XML
if( !isset( $checkit ) ) {
# Load/build some scary lookup tables...
UtfNormal::loadData();
- global $utfCheckNFC, $utfCombiningClass;
- $utfCheckOrCombining = array_merge( $utfCheckNFC, $utfCombiningClass );
+ $utfCheckOrCombining = array_merge( self::$utfCheckNFC, self::$utfCombiningClass );
# Head bytes for sequences which we should do further validity checks
$checkit = array_flip( array_map( 'chr',
$len = $chunk + 1; # Counting down is faster. I'm *so* sorry.
for( $i = -1; --$len; ) {
- if( $remaining = $tailBytes[$c = $str{++$i}] ) {
+ $remaining = $tailBytes[$c = $str{++$i}];
+ if( $remaining ) {
# UTF-8 head byte!
$sequence = $head = $c;
do {
# checking for validity or any optimization etc. Input must be
# VALID UTF-8!
/**
- * @param string $string
+ * @param $string string
* @return string
* @private
- * @static
*/
static function NFC( $string ) {
return UtfNormal::fastCompose( UtfNormal::NFD( $string ) );
}
/**
- * @param string $string
+ * @param $string string
* @return string
* @private
- * @static
*/
static function NFD( $string ) {
UtfNormal::loadData();
- global $utfCanonicalDecomp;
+
return UtfNormal::fastCombiningSort(
- UtfNormal::fastDecompose( $string, $utfCanonicalDecomp ) );
+ UtfNormal::fastDecompose( $string, self::$utfCanonicalDecomp ) );
}
/**
- * @param string $string
+ * @param $string string
* @return string
* @private
- * @static
*/
static function NFKC( $string ) {
return UtfNormal::fastCompose( UtfNormal::NFKD( $string ) );
}
/**
- * @param string $string
+ * @param $string string
* @return string
* @private
- * @static
*/
static function NFKD( $string ) {
- global $utfCompatibilityDecomp;
- if( !isset( $utfCompatibilityDecomp ) ) {
+ if( !isset( self::$utfCompatibilityDecomp ) ) {
require_once( 'UtfNormalDataK.inc' );
}
- return UtfNormal::fastCombiningSort(
- UtfNormal::fastDecompose( $string, $utfCompatibilityDecomp ) );
+ return self::fastCombiningSort(
+ self::fastDecompose( $string, self::$utfCompatibilityDecomp ) );
}
* (depending on which decomposition map is passed to us).
* Input is assumed to be *valid* UTF-8. Invalid code will break.
* @private
- * @param string $string Valid UTF-8 string
- * @param array $map hash of expanded decomposition map
+ * @param $string String: valid UTF-8 string
+ * @param $map Array: hash of expanded decomposition map
* @return string a UTF-8 string decomposed, not yet normalized (needs sorting)
- * @static
*/
static function fastDecompose( $string, $map ) {
UtfNormal::loadData();
* Sorts combining characters into canonical order. This is the
* final step in creating decomposed normal forms D and KD.
* @private
- * @param string $string a valid, decomposed UTF-8 string. Input is not validated.
+ * @param $string String: a valid, decomposed UTF-8 string. Input is not validated.
* @return string a UTF-8 string with combining characters sorted in canonical order
- * @static
*/
static function fastCombiningSort( $string ) {
UtfNormal::loadData();
- global $utfCombiningClass;
$len = strlen( $string );
$out = '';
$combiners = array();
$c = substr( $string, $i, 2 );
$i++;
}
- if( isset( $utfCombiningClass[$c] ) ) {
- $lastClass = $utfCombiningClass[$c];
+ if( isset( self::$utfCombiningClass[$c] ) ) {
+ $lastClass = self::$utfCombiningClass[$c];
if( isset( $combiners[$lastClass] ) ) {
$combiners[$lastClass] .= $c;
} else {
* Produces canonically composed sequences, i.e. normal form C or KC.
*
* @private
- * @param string $string a valid UTF-8 string in sorted normal form D or KD. Input is not validated.
+ * @param $string String: a valid UTF-8 string in sorted normal form D or KD. Input is not validated.
* @return string a UTF-8 string with canonical precomposed characters used where possible
- * @static
*/
static function fastCompose( $string ) {
UtfNormal::loadData();
- global $utfCanonicalComp, $utfCombiningClass;
$len = strlen( $string );
$out = '';
$lastClass = -1;
}
$pair = $startChar . $c;
if( $n > 0x80 ) {
- if( isset( $utfCombiningClass[$c] ) ) {
+ if( isset( self::$utfCombiningClass[$c] ) ) {
# A combining char; see what we can do with it
- $class = $utfCombiningClass[$c];
+ $class = self::$utfCombiningClass[$c];
if( !empty( $startChar ) &&
$lastClass < $class &&
$class > 0 &&
- isset( $utfCanonicalComp[$pair] ) ) {
- $startChar = $utfCanonicalComp[$pair];
+ isset( self::$utfCanonicalComp[$pair] ) ) {
+ $startChar = self::$utfCanonicalComp[$pair];
$class = 0;
} else {
$combining .= $c;
}
# New start char
if( $lastClass == 0 ) {
- if( isset( $utfCanonicalComp[$pair] ) ) {
- $startChar = $utfCanonicalComp[$pair];
+ if( isset( self::$utfCanonicalComp[$pair] ) ) {
+ $startChar = self::$utfCanonicalComp[$pair];
$lastHangul = 0;
continue;
}
/**
* This is just used for the benchmark, comparing how long it takes to
* interate through a string without really doing anything of substance.
- * @param string $string
+ * @param $string string
* @return string
- * @static
*/
static function placebo( $string ) {
$len = strlen( $string );
return $out;
}
}
-
-?>