WordPress.org

Make WordPress Core


Ignore:
Timestamp:
07/22/2015 05:14:50 AM (5 years ago)
Author:
pento
Message:

Shortcodes: Improve the reliablity of shortcodes inside HTML tags.

Props miqrogroove.

See #15694.

File:
1 edited

Legend:

Unmodified
Added
Removed
  • trunk/src/wp-includes/kses.php

    r32860 r33359  
    530530
    531531/**
     532 * Filters one attribute only and ensures its value is allowed.
     533 *
     534 * This function has the advantage of being more secure than esc_attr() and can
     535 * escape data in some situations where wp_kses() must strip the whole attribute.
     536 *
     537 * @since 4.2.3
     538 *
     539 * @param string $string The 'whole' attribute, including name and value.
     540 * @param string $element The element name to which the attribute belongs.
     541 * @return string Filtered attribute.
     542 */
     543function wp_kses_one_attr( $string, $element ) {
     544    $uris = array('xmlns', 'profile', 'href', 'src', 'cite', 'classid', 'codebase', 'data', 'usemap', 'longdesc', 'action');
     545    $allowed_html = wp_kses_allowed_html( 'post' );
     546    $allowed_protocols = wp_allowed_protocols();
     547    $string = wp_kses_no_null( $string, array( 'slash_zero' => 'keep' ) );
     548    $string = wp_kses_js_entities( $string );
     549   
     550    // Preserve leading and trailing whitespace.
     551    $matches = array();
     552    preg_match('/^\s*/', $string, $matches);
     553    $lead = $matches[0];
     554    preg_match('/\s*$/', $string, $matches);
     555    $trail = $matches[0];
     556    if ( empty( $trail ) ) {
     557        $string = substr( $string, strlen( $lead ) );
     558    } else {
     559        $string = substr( $string, strlen( $lead ), -strlen( $trail ) );
     560    }
     561   
     562    // Parse attribute name and value from input.
     563    $split = preg_split( '/\s*=\s*/', $string, 2 );
     564    $name = $split[0];
     565    if ( count( $split ) == 2 ) {
     566        $value = $split[1];
     567
     568        // Remove quotes surrounding $value.
     569        // Also guarantee correct quoting in $string for this one attribute.
     570        if ( '' == $value ) {
     571            $quote = '';
     572        } else {
     573            $quote = $value[0];
     574        }
     575        if ( '"' == $quote || "'" == $quote ) {
     576            if ( substr( $value, -1 ) != $quote ) {
     577                return '';
     578            }
     579            $value = substr( $value, 1, -1 );
     580        } else {
     581            $quote = '"';
     582        }
     583
     584        // Sanitize quotes, angle braces, and entities.
     585        $value = esc_attr( $value );
     586
     587        // Sanitize URI values.
     588        if ( in_array( strtolower( $name ), $uris ) ) {
     589            $value = wp_kses_bad_protocol( $value, $allowed_protocols );
     590        }
     591
     592        $string = "$name=$quote$value$quote";
     593        $vless = 'n';
     594    } else {
     595        $value = '';
     596        $vless = 'y';
     597    }
     598   
     599    // Sanitize attribute by name.
     600    wp_kses_attr_check( $name, $value, $string, $vless, $element, $allowed_html );
     601
     602    // Restore whitespace.
     603    return $lead . $string . $trail;
     604}
     605
     606/**
    532607 * Return a list of allowed tags and attributes for a given context.
    533608 *
     
    742817 */
    743818function wp_kses_attr($element, $attr, $allowed_html, $allowed_protocols) {
    744     // Is there a closing XHTML slash at the end of the attributes?
    745 
    746819    if ( ! is_array( $allowed_html ) )
    747820        $allowed_html = wp_kses_allowed_html( $allowed_html );
    748821
     822    // Is there a closing XHTML slash at the end of the attributes?
    749823    $xhtml_slash = '';
    750824    if (preg_match('%\s*/\s*$%', $attr))
     
    761835    // in $attr2
    762836    $attr2 = '';
    763 
    764     $allowed_attr = $allowed_html[strtolower($element)];
    765     foreach ($attrarr as $arreach) {
    766         if ( ! isset( $allowed_attr[strtolower($arreach['name'])] ) )
    767             continue; // the attribute is not allowed
    768 
    769         $current = $allowed_attr[strtolower($arreach['name'])];
    770         if ( $current == '' )
    771             continue; // the attribute is not allowed
    772 
    773         if ( strtolower( $arreach['name'] ) == 'style' ) {
    774             $orig_value = $arreach['value'];
    775             $value = safecss_filter_attr( $orig_value );
    776 
    777             if ( empty( $value ) )
    778                 continue;
    779 
    780             $arreach['value'] = $value;
    781             $arreach['whole'] = str_replace( $orig_value, $value, $arreach['whole'] );
     837    foreach ( $attrarr as $arreach ) {
     838        if ( wp_kses_attr_check( $arreach['name'], $arreach['value'], $arreach['whole'], $arreach['vless'], $element, $allowed_html ) ) {
     839            $attr2 .= ' '.$arreach['whole'];
    782840        }
    783 
    784         if ( ! is_array($current) ) {
    785             $attr2 .= ' '.$arreach['whole'];
    786         // there are no checks
    787 
    788         } else {
    789             // there are some checks
    790             $ok = true;
    791             foreach ($current as $currkey => $currval) {
    792                 if ( ! wp_kses_check_attr_val($arreach['value'], $arreach['vless'], $currkey, $currval) ) {
    793                     $ok = false;
    794                     break;
    795                 }
    796             }
    797 
    798             if ( $ok )
    799                 $attr2 .= ' '.$arreach['whole']; // it passed them
    800         } // if !is_array($current)
    801     } // foreach
     841    }
    802842
    803843    // Remove any "<" or ">" characters
     
    805845
    806846    return "<$element$attr2$xhtml_slash>";
     847}
     848
     849/**
     850 * Determine whether an attribute is allowed.
     851 *
     852 * @since 4.2.3
     853 *
     854 * @param string $name The attribute name. Returns empty string when not allowed.
     855 * @param string $value The attribute value. Returns a filtered value.
     856 * @param string $whole The name=value input. Returns filtered input.
     857 * @param string $vless 'y' when attribute like "enabled", otherwise 'n'.
     858 * @param string $element The name of the element to which this attribute belongs.
     859 * @param array $allowed_html The full list of allowed elements and attributes.
     860 * @return bool Is the attribute allowed?
     861 */
     862function wp_kses_attr_check( &$name, &$value, &$whole, $vless, $element, $allowed_html ) {
     863    $allowed_attr = $allowed_html[strtolower( $element )];
     864
     865    $name_low = strtolower( $name );
     866    if ( ! isset( $allowed_attr[$name_low] ) || '' == $allowed_attr[$name_low] ) {
     867        $name = $value = $whole = '';
     868        return false;
     869    }
     870
     871    if ( 'style' == $name_low ) {
     872        $new_value = safecss_filter_attr( $value );
     873
     874        if ( empty( $new_value ) ) {
     875            $name = $value = $whole = '';
     876            return false;
     877        }
     878
     879        $whole = str_replace( $value, $new_value, $whole );
     880        $value = $new_value;
     881    }
     882
     883    if ( is_array( $allowed_attr[$name_low] ) ) {
     884        // there are some checks
     885        foreach ( $allowed_attr[$name_low] as $currkey => $currval ) {
     886            if ( ! wp_kses_check_attr_val( $value, $vless, $currkey, $currval ) ) {
     887                $name = $value = $whole = '';
     888                return false;
     889            }
     890        }
     891    }
     892
     893    return true;
    807894}
    808895
     
    9341021
    9351022    return $attrarr;
     1023}
     1024
     1025/**
     1026 * Finds all attributes of an HTML element.
     1027 *
     1028 * Does not modify input.  May return "evil" output.
     1029 *
     1030 * Based on wp_kses_split2() and wp_kses_attr()
     1031 *
     1032 * @since 4.2.3
     1033 *
     1034 * @param string $element HTML element/tag
     1035 * @return array|bool List of attributes found in $element. Returns false on failure.
     1036 */
     1037function wp_kses_attr_parse( $element ) {
     1038    $valid = preg_match('%^(<\s*)(/\s*)?([a-zA-Z0-9]+\s*)([^>]*)(>?)$%', $element, $matches);
     1039    if ( 1 !== $valid ) {
     1040        return false;
     1041    }
     1042
     1043    $begin =  $matches[1];
     1044    $slash =  $matches[2];
     1045    $elname = $matches[3];
     1046    $attr =   $matches[4];
     1047    $end =    $matches[5];
     1048
     1049    if ( '' !== $slash ) {
     1050        // Closing elements do not get parsed.
     1051        return false;
     1052    }
     1053
     1054    // Is there a closing XHTML slash at the end of the attributes?
     1055    if ( 1 === preg_match( '%\s*/\s*$%', $attr, $matches ) ) {
     1056        $xhtml_slash = $matches[0];
     1057        $attr = substr( $attr, 0, -strlen( $xhtml_slash ) );
     1058    } else {
     1059        $xhtml_slash = '';
     1060    }
     1061   
     1062    // Split it
     1063    $attrarr = wp_kses_hair_parse( $attr );
     1064    if ( false === $attrarr ) {
     1065        return false;
     1066    }
     1067
     1068    // Make sure all input is returned by adding front and back matter.
     1069    array_unshift( $attrarr, $begin . $slash . $elname );
     1070    array_push( $attrarr, $xhtml_slash . $end );
     1071   
     1072    return $attrarr;
     1073}
     1074
     1075/**
     1076 * Builds an attribute list from string containing attributes.
     1077 *
     1078 * Does not modify input.  May return "evil" output.
     1079 * In case of unexpected input, returns false instead of stripping things.
     1080 *
     1081 * Based on wp_kses_hair() but does not return a multi-dimensional array.
     1082 *
     1083 * @since 4.2.3
     1084 *
     1085 * @param string $attr Attribute list from HTML element to closing HTML element tag
     1086 * @return array|bool List of attributes found in $attr. Returns false on failure.
     1087 */
     1088function wp_kses_hair_parse( $attr ) {
     1089    if ( '' === $attr ) {
     1090        return array();
     1091    }
     1092
     1093    $regex =
     1094      '(?:'
     1095    .     '[-a-zA-Z:]+'   // Attribute name.
     1096    . '|'
     1097    .     '\[\[?[^\[\]]+\]\]?' // Shortcode in the name position implies unfiltered_html.
     1098    . ')'
     1099    . '(?:'               // Attribute value.
     1100    .     '\s*=\s*'       // All values begin with '='
     1101    .     '(?:'
     1102    .         '"[^"]*"'   // Double-quoted
     1103    .     '|'
     1104    .         "'[^']*'"   // Single-quoted
     1105    .     '|'
     1106    .         '[^\s"\']+' // Non-quoted
     1107    .         '(?:\s|$)'  // Must have a space
     1108    .     ')'
     1109    . '|'
     1110    .     '(?:\s|$)'      // If attribute has no value, space is required.
     1111    . ')'
     1112    . '\s*';              // Trailing space is optional except as mentioned above.
     1113
     1114    // Although it is possible to reduce this procedure to a single regexp,
     1115    // we must run that regexp twice to get exactly the expected result.
     1116
     1117    $validation = "%^($regex)+$%";
     1118    $extraction = "%$regex%";
     1119
     1120    if ( 1 === preg_match( $validation, $attr ) ) {
     1121        preg_match_all( $extraction, $attr, $attrarr );
     1122        return $attrarr[0];
     1123    } else {
     1124        return false;
     1125    }
    9361126}
    9371127
Note: See TracChangeset for help on using the changeset viewer.