vendor/wikimedia/css-sanitizer/src/Parser/StringDataSource.php

   1 <?php
   2 /**
   3  * @file
   4  * @license https://opensource.org/licenses/Apache-2.0 Apache-2.0
   5  */
   6
   7 namespace Wikimedia\CSS\Parser;
   8
   9 /**
  10  * Read data for the CSS parser
  11  */
  12 class StringDataSource implements DataSource {
  13
  14         /** @var string */
  15         protected $string;
  16
  17         /** @var int */
  18         protected $len = 0, $pos = 0;
  19
  20         /** @var string[] */
  21         protected $putBack = [];
  22
  23         /**
  24          * @param string $string Input string. Must be valid UTF-8 with no BOM.
  25          */
  26         public function __construct( $string ) {
  27                 static $newPHP;
  28
  29                 $this->string = (string)$string;
  30                 $this->len = strlen( $this->string );
  31
  32                 // HHVM 3.4 and older come with an outdated version of libmbfl that
  33                 // incorrectly allows values above U+10FFFF, so we have to check
  34                 // for them separately. (This issue also exists in PHP 5.3 and
  35                 // older, which are no longer supported.)
  36                 // @codeCoverageIgnoreStart
  37                 if ( $newPHP === null ) {
  38                         $newPHP = !mb_check_encoding( "\xf4\x90\x80\x80", 'UTF-8' );
  39                 }
  40                 // @codeCoverageIgnoreEnd
  41
  42                 if ( !mb_check_encoding( $this->string, 'UTF-8' ) ||
  43                         !$newPHP && preg_match( "/\xf4[\x90-\xbf]|[\xf5-\xff]/S", $this->string ) !== 0
  44                 ) {
  45                         throw new \InvalidArgumentException( '$string is not valid UTF-8' );
  46                 }
  47         }
  48
  49         public function readCharacter() {
  50                 if ( $this->putBack ) {
  51                         return array_pop( $this->putBack );
  52                 }
  53
  54                 if ( $this->pos >= $this->len ) {
  55                         return self::EOF;
  56                 }
  57
  58                 // We already checked that the string is valid UTF-8 in the
  59                 // constructor, so we can do a quick binary "get next character" here.
  60                 $p = $this->pos;
  61                 $c = $this->string[$p];
  62                 $cc = ord( $this->string[$p] );
  63                 if ( $cc <= 0x7f ) {
  64                         $this->pos += 1;
  65                         return $c;
  66                 } elseif ( ( $cc & 0xe0 ) === 0xc0 ) {
  67                         $this->pos += 2;
  68                         return substr( $this->string, $p, 2 );
  69                 } elseif ( ( $cc & 0xf0 ) === 0xe0 ) {
  70                         $this->pos += 3;
  71                         return substr( $this->string, $p, 3 );
  72                 } elseif ( ( $cc & 0xf8 ) === 0xf0 ) {
  73                         $this->pos += 4;
  74                         return substr( $this->string, $p, 4 );
  75                 } else {
  76                         // WTF? Should never get here because it should have failed
  77                         // validation in the constructor.
  78                         // @codeCoverageIgnoreStart
  79                         throw new \UnexpectedValueException(
  80                                 sprintf( 'Unexpected byte %02X in string at position %d.', $cc, $this->pos )
  81                         );
  82                         // @codeCoverageIgnoreEnd
  83                 }
  84         }
  85
  86         public function putBackCharacter( $char ) {
  87                 if ( $char !== self::EOF ) {
  88                         $this->putBack[] = $char;
  89                 }
  90         }
  91 }