wp/wp-includes/shortcodes.php
changeset 7 cf61fcea0001
parent 5 5e2f62d02dcd
child 9 177826044cd9
--- a/wp/wp-includes/shortcodes.php	Tue Jun 09 11:14:17 2015 +0000
+++ b/wp/wp-includes/shortcodes.php	Mon Oct 14 17:39:30 2019 +0200
@@ -1,7 +1,7 @@
 <?php
 /**
- * WordPress API for creating bbcode like tags or what WordPress calls
- * "shortcodes." The tag and attribute parsing or regular expression code is
+ * WordPress API for creating bbcode-like tags or what WordPress calls
+ * "shortcodes". The tag and attribute parsing or regular expression code is
  * based on the Textpattern tag parser.
  *
  * A few examples are below:
@@ -42,55 +42,41 @@
 $shortcode_tags = array();
 
 /**
- * Add hook for shortcode tag.
- *
- * There can only be one hook for each shortcode. Which means that if another
- * plugin has a similar shortcode, it will override yours or yours will override
- * theirs depending on which order the plugins are included and/or ran.
- *
- * Simplest example of a shortcode tag using the API:
- *
- *     // [footag foo="bar"]
- *     function footag_func( $atts ) {
- *         return "foo = {
- *             $atts[foo]
- *         }";
- *     }
- *     add_shortcode( 'footag', 'footag_func' );
- *
- * Example with nice attribute defaults:
+ * Adds a new shortcode.
  *
- *     // [bartag foo="bar"]
- *     function bartag_func( $atts ) {
- *         $args = shortcode_atts( array(
- *             'foo' => 'no foo',
- *             'baz' => 'default baz',
- *         ), $atts );
- *
- *         return "foo = {$args['foo']}";
- *     }
- *     add_shortcode( 'bartag', 'bartag_func' );
- *
- * Example with enclosed content:
- *
- *     // [baztag]content[/baztag]
- *     function baztag_func( $atts, $content = '' ) {
- *         return "content = $content";
- *     }
- *     add_shortcode( 'baztag', 'baztag_func' );
+ * Care should be taken through prefixing or other means to ensure that the
+ * shortcode tag being added is unique and will not conflict with other,
+ * already-added shortcode tags. In the event of a duplicated tag, the tag
+ * loaded last will take precedence.
  *
  * @since 2.5.0
  *
- * @uses $shortcode_tags
+ * @global array $shortcode_tags
  *
- * @param string $tag Shortcode tag to be searched in post content.
- * @param callable $func Hook to run when shortcode is found.
+ * @param string   $tag      Shortcode tag to be searched in post content.
+ * @param callable $callback The callback function to run when the shortcode is found.
+ *                           Every shortcode callback is passed three parameters by default,
+ *                           including an array of attributes (`$atts`), the shortcode content
+ *                           or null if not set (`$content`), and finally the shortcode tag
+ *                           itself (`$shortcode_tag`), in that order.
  */
-function add_shortcode($tag, $func) {
+function add_shortcode( $tag, $callback ) {
 	global $shortcode_tags;
 
-	if ( is_callable($func) )
-		$shortcode_tags[$tag] = $func;
+	if ( '' == trim( $tag ) ) {
+		$message = __( 'Invalid shortcode name: Empty name given.' );
+		_doing_it_wrong( __FUNCTION__, $message, '4.4.0' );
+		return;
+	}
+
+	if ( 0 !== preg_match( '@[<>&/\[\]\x00-\x20=]@', $tag ) ) {
+		/* translators: 1: shortcode name, 2: space separated list of reserved characters */
+		$message = sprintf( __( 'Invalid shortcode name: %1$s. Do not use spaces or reserved characters: %2$s' ), $tag, '& / < > [ ] =' );
+		_doing_it_wrong( __FUNCTION__, $message, '4.4.0' );
+		return;
+	}
+
+	$shortcode_tags[ $tag ] = $callback;
 }
 
 /**
@@ -98,7 +84,7 @@
  *
  * @since 2.5.0
  *
- * @uses $shortcode_tags
+ * @global array $shortcode_tags
  *
  * @param string $tag Shortcode tag to remove hook for.
  */
@@ -117,7 +103,7 @@
  *
  * @since 2.5.0
  *
- * @uses $shortcode_tags
+ * @global array $shortcode_tags
  */
 function remove_all_shortcodes() {
 	global $shortcode_tags;
@@ -157,7 +143,7 @@
 	}
 
 	if ( shortcode_exists( $tag ) ) {
-		preg_match_all( '/' . get_shortcode_regex() . '/s', $content, $matches, PREG_SET_ORDER );
+		preg_match_all( '/' . get_shortcode_regex() . '/', $content, $matches, PREG_SET_ORDER );
 		if ( empty( $matches ) )
 			return false;
 
@@ -184,9 +170,10 @@
  * @global array $shortcode_tags List of shortcode tags and their callback hooks.
  *
  * @param string $content Content to search for shortcodes.
+ * @param bool $ignore_html When true, shortcodes inside HTML elements will be skipped.
  * @return string Content with shortcodes filtered out.
  */
-function do_shortcode($content) {
+function do_shortcode( $content, $ignore_html = false ) {
 	global $shortcode_tags;
 
 	if ( false === strpos( $content, '[' ) ) {
@@ -196,8 +183,23 @@
 	if (empty($shortcode_tags) || !is_array($shortcode_tags))
 		return $content;
 
-	$pattern = get_shortcode_regex();
-	return preg_replace_callback( "/$pattern/s", 'do_shortcode_tag', $content );
+	// Find all registered tag names in $content.
+	preg_match_all( '@\[([^<>&/\[\]\x00-\x20=]++)@', $content, $matches );
+	$tagnames = array_intersect( array_keys( $shortcode_tags ), $matches[1] );
+
+	if ( empty( $tagnames ) ) {
+		return $content;
+	}
+
+	$content = do_shortcodes_in_html_tags( $content, $ignore_html, $tagnames );
+
+	$pattern = get_shortcode_regex( $tagnames );
+	$content = preg_replace_callback( "/$pattern/", 'do_shortcode_tag', $content );
+
+	// Always restore square braces so we don't break things like <!--[if IE ]>
+	$content = unescape_invalid_shortcodes( $content );
+
+	return $content;
 }
 
 /**
@@ -216,14 +218,19 @@
  * 6 - An extra ] to allow for escaping shortcodes with double [[]]
  *
  * @since 2.5.0
+ * @since 4.4.0 Added the `$tagnames` parameter.
  *
- * @uses $shortcode_tags
+ * @global array $shortcode_tags
  *
+ * @param array $tagnames Optional. List of shortcodes to find. Defaults to all registered shortcodes.
  * @return string The shortcode search regular expression
  */
-function get_shortcode_regex() {
+function get_shortcode_regex( $tagnames = null ) {
 	global $shortcode_tags;
-	$tagnames = array_keys($shortcode_tags);
+
+	if ( empty( $tagnames ) ) {
+		$tagnames = array_keys( $shortcode_tags );
+	}
 	$tagregexp = join( '|', array_map('preg_quote', $tagnames) );
 
 	// WARNING! Do not change this regex without changing do_shortcode_tag() and strip_shortcode_tag()
@@ -265,10 +272,11 @@
  *
  * @since 2.5.0
  * @access private
- * @uses $shortcode_tags
+ *
+ * @global array $shortcode_tags
  *
  * @param array $m Regular expression match array
- * @return mixed False on failure.
+ * @return string|false False on failure.
  */
 function do_shortcode_tag( $m ) {
 	global $shortcode_tags;
@@ -281,13 +289,179 @@
 	$tag = $m[2];
 	$attr = shortcode_parse_atts( $m[3] );
 
-	if ( isset( $m[5] ) ) {
-		// enclosing tag - extra parameter
-		return $m[1] . call_user_func( $shortcode_tags[$tag], $attr, $m[5], $tag ) . $m[6];
-	} else {
-		// self-closing tag
-		return $m[1] . call_user_func( $shortcode_tags[$tag], $attr, null,  $tag ) . $m[6];
+	if ( ! is_callable( $shortcode_tags[ $tag ] ) ) {
+		/* translators: %s: shortcode tag */
+		$message = sprintf( __( 'Attempting to parse a shortcode without a valid callback: %s' ), $tag );
+		_doing_it_wrong( __FUNCTION__, $message, '4.3.0' );
+		return $m[0];
+	}
+
+	/**
+	 * Filters whether to call a shortcode callback.
+	 *
+	 * Passing a truthy value to the filter will effectively short-circuit the
+	 * shortcode generation process, returning that value instead.
+	 *
+	 * @since 4.7.0
+	 *
+	 * @param bool|string $return      Short-circuit return value. Either false or the value to replace the shortcode with.
+	 * @param string       $tag         Shortcode name.
+	 * @param array|string $attr        Shortcode attributes array or empty string.
+	 * @param array        $m           Regular expression match array.
+	 */
+	$return = apply_filters( 'pre_do_shortcode_tag', false, $tag, $attr, $m );
+	if ( false !== $return ) {
+		return $return;
 	}
+
+	$content = isset( $m[5] ) ? $m[5] : null;
+
+	$output = $m[1] . call_user_func( $shortcode_tags[ $tag ], $attr, $content, $tag ) . $m[6];
+
+	/**
+	 * Filters the output created by a shortcode callback.
+	 *
+	 * @since 4.7.0
+	 *
+	 * @param string       $output Shortcode output.
+	 * @param string       $tag    Shortcode name.
+	 * @param array|string $attr   Shortcode attributes array or empty string.
+	 * @param array        $m      Regular expression match array.
+	 */
+	return apply_filters( 'do_shortcode_tag', $output, $tag, $attr, $m );
+}
+
+/**
+ * Search only inside HTML elements for shortcodes and process them.
+ *
+ * Any [ or ] characters remaining inside elements will be HTML encoded
+ * to prevent interference with shortcodes that are outside the elements.
+ * Assumes $content processed by KSES already.  Users with unfiltered_html
+ * capability may get unexpected output if angle braces are nested in tags.
+ *
+ * @since 4.2.3
+ *
+ * @param string $content Content to search for shortcodes
+ * @param bool $ignore_html When true, all square braces inside elements will be encoded.
+ * @param array $tagnames List of shortcodes to find.
+ * @return string Content with shortcodes filtered out.
+ */
+function do_shortcodes_in_html_tags( $content, $ignore_html, $tagnames ) {
+	// Normalize entities in unfiltered HTML before adding placeholders.
+	$trans = array( '&#91;' => '&#091;', '&#93;' => '&#093;' );
+	$content = strtr( $content, $trans );
+	$trans = array( '[' => '&#91;', ']' => '&#93;' );
+
+	$pattern = get_shortcode_regex( $tagnames );
+	$textarr = wp_html_split( $content );
+
+	foreach ( $textarr as &$element ) {
+		if ( '' == $element || '<' !== $element[0] ) {
+			continue;
+		}
+
+		$noopen = false === strpos( $element, '[' );
+		$noclose = false === strpos( $element, ']' );
+		if ( $noopen || $noclose ) {
+			// This element does not contain shortcodes.
+			if ( $noopen xor $noclose ) {
+				// Need to encode stray [ or ] chars.
+				$element = strtr( $element, $trans );
+			}
+			continue;
+		}
+
+		if ( $ignore_html || '<!--' === substr( $element, 0, 4 ) || '<![CDATA[' === substr( $element, 0, 9 ) ) {
+			// Encode all [ and ] chars.
+			$element = strtr( $element, $trans );
+			continue;
+		}
+
+		$attributes = wp_kses_attr_parse( $element );
+		if ( false === $attributes ) {
+			// Some plugins are doing things like [name] <[email]>.
+			if ( 1 === preg_match( '%^<\s*\[\[?[^\[\]]+\]%', $element ) ) {
+				$element = preg_replace_callback( "/$pattern/", 'do_shortcode_tag', $element );
+			}
+
+			// Looks like we found some crazy unfiltered HTML.  Skipping it for sanity.
+			$element = strtr( $element, $trans );
+			continue;
+		}
+
+		// Get element name
+		$front = array_shift( $attributes );
+		$back = array_pop( $attributes );
+		$matches = array();
+		preg_match('%[a-zA-Z0-9]+%', $front, $matches);
+		$elname = $matches[0];
+
+		// Look for shortcodes in each attribute separately.
+		foreach ( $attributes as &$attr ) {
+			$open = strpos( $attr, '[' );
+			$close = strpos( $attr, ']' );
+			if ( false === $open || false === $close ) {
+				continue; // Go to next attribute.  Square braces will be escaped at end of loop.
+			}
+			$double = strpos( $attr, '"' );
+			$single = strpos( $attr, "'" );
+			if ( ( false === $single || $open < $single ) && ( false === $double || $open < $double ) ) {
+				// $attr like '[shortcode]' or 'name = [shortcode]' implies unfiltered_html.
+				// In this specific situation we assume KSES did not run because the input
+				// was written by an administrator, so we should avoid changing the output
+				// and we do not need to run KSES here.
+				$attr = preg_replace_callback( "/$pattern/", 'do_shortcode_tag', $attr );
+			} else {
+				// $attr like 'name = "[shortcode]"' or "name = '[shortcode]'"
+				// We do not know if $content was unfiltered. Assume KSES ran before shortcodes.
+				$count = 0;
+				$new_attr = preg_replace_callback( "/$pattern/", 'do_shortcode_tag', $attr, -1, $count );
+				if ( $count > 0 ) {
+					// Sanitize the shortcode output using KSES.
+					$new_attr = wp_kses_one_attr( $new_attr, $elname );
+					if ( '' !== trim( $new_attr ) ) {
+						// The shortcode is safe to use now.
+						$attr = $new_attr;
+					}
+				}
+			}
+		}
+		$element = $front . implode( '', $attributes ) . $back;
+
+		// Now encode any remaining [ or ] chars.
+		$element = strtr( $element, $trans );
+	}
+
+	$content = implode( '', $textarr );
+
+	return $content;
+}
+
+/**
+ * Remove placeholders added by do_shortcodes_in_html_tags().
+ *
+ * @since 4.2.3
+ *
+ * @param string $content Content to search for placeholders.
+ * @return string Content with placeholders removed.
+ */
+function unescape_invalid_shortcodes( $content ) {
+        // Clean up entire string, avoids re-parsing HTML.
+        $trans = array( '&#91;' => '[', '&#93;' => ']' );
+        $content = strtr( $content, $trans );
+
+        return $content;
+}
+
+/**
+ * Retrieve the shortcode attributes regex.
+ *
+ * @since 4.4.0
+ *
+ * @return string The shortcode attribute regular expression
+ */
+function get_shortcode_atts_regex() {
+	return '/([\w-]+)\s*=\s*"([^"]*)"(?:\s|$)|([\w-]+)\s*=\s*\'([^\']*)\'(?:\s|$)|([\w-]+)\s*=\s*([^\s\'"]+)(?:\s|$)|"([^"]*)"(?:\s|$)|\'([^\']*)\'(?:\s|$)|(\S+)(?:\s|$)/';
 }
 
 /**
@@ -300,11 +474,14 @@
  * @since 2.5.0
  *
  * @param string $text
- * @return array List of attributes and their value.
+ * @return array|string List of attribute values.
+ *                      Returns empty array if trim( $text ) == '""'.
+ *                      Returns empty string if trim( $text ) == ''.
+ *                      All other matches are checked for not empty().
  */
 function shortcode_parse_atts($text) {
 	$atts = array();
-	$pattern = '/(\w+)\s*=\s*"([^"]*)"(?:\s|$)|(\w+)\s*=\s*\'([^\']*)\'(?:\s|$)|(\w+)\s*=\s*([^\s\'"]+)(?:\s|$)|"([^"]*)"(?:\s|$)|(\S+)(?:\s|$)/';
+	$pattern = get_shortcode_atts_regex();
 	$text = preg_replace("/[\x{00a0}\x{200b}]+/u", " ", $text);
 	if ( preg_match_all($pattern, $text, $match, PREG_SET_ORDER) ) {
 		foreach ($match as $m) {
@@ -316,8 +493,19 @@
 				$atts[strtolower($m[5])] = stripcslashes($m[6]);
 			elseif (isset($m[7]) && strlen($m[7]))
 				$atts[] = stripcslashes($m[7]);
-			elseif (isset($m[8]))
+			elseif (isset($m[8]) && strlen($m[8]))
 				$atts[] = stripcslashes($m[8]);
+			elseif (isset($m[9]))
+				$atts[] = stripcslashes($m[9]);
+		}
+
+		// Reject any unclosed HTML elements
+		foreach( $atts as &$value ) {
+			if ( false !== strpos( $value, '<' ) ) {
+				if ( 1 !== preg_match( '/^[^<]*+(?:<[^>]*+>[^<]*+)*+$/', $value ) ) {
+					$value = '';
+				}
+			}
 		}
 	} else {
 		$atts = ltrim($text);
@@ -337,34 +525,37 @@
  *
  * @since 2.5.0
  *
- * @param array $pairs Entire list of supported attributes and their defaults.
- * @param array $atts User defined attributes in shortcode tag.
+ * @param array  $pairs     Entire list of supported attributes and their defaults.
+ * @param array  $atts      User defined attributes in shortcode tag.
  * @param string $shortcode Optional. The name of the shortcode, provided for context to enable filtering
  * @return array Combined and filtered attribute list.
  */
 function shortcode_atts( $pairs, $atts, $shortcode = '' ) {
 	$atts = (array)$atts;
 	$out = array();
-	foreach($pairs as $name => $default) {
+	foreach ($pairs as $name => $default) {
 		if ( array_key_exists($name, $atts) )
 			$out[$name] = $atts[$name];
 		else
 			$out[$name] = $default;
 	}
 	/**
-	 * Filter a shortcode's default attributes.
+	 * Filters a shortcode's default attributes.
 	 *
 	 * If the third parameter of the shortcode_atts() function is present then this filter is available.
 	 * The third parameter, $shortcode, is the name of the shortcode.
 	 *
 	 * @since 3.6.0
+	 * @since 4.4.0 Added the `$shortcode` parameter.
 	 *
-	 * @param array $out The output array of shortcode attributes.
-	 * @param array $pairs The supported attributes and their defaults.
-	 * @param array $atts The user defined shortcode attributes.
+	 * @param array  $out       The output array of shortcode attributes.
+	 * @param array  $pairs     The supported attributes and their defaults.
+	 * @param array  $atts      The user defined shortcode attributes.
+	 * @param string $shortcode The shortcode name.
 	 */
-	if ( $shortcode )
-		$out = apply_filters( "shortcode_atts_{$shortcode}", $out, $pairs, $atts );
+	if ( $shortcode ) {
+		$out = apply_filters( "shortcode_atts_{$shortcode}", $out, $pairs, $atts, $shortcode );
+	}
 
 	return $out;
 }
@@ -374,7 +565,7 @@
  *
  * @since 2.5.0
  *
- * @uses $shortcode_tags
+ * @global array $shortcode_tags
  *
  * @param string $content Content to remove shortcode tags.
  * @return string Content without shortcode tags.
@@ -389,11 +580,46 @@
 	if (empty($shortcode_tags) || !is_array($shortcode_tags))
 		return $content;
 
-	$pattern = get_shortcode_regex();
+	// Find all registered tag names in $content.
+	preg_match_all( '@\[([^<>&/\[\]\x00-\x20=]++)@', $content, $matches );
+
+	$tags_to_remove = array_keys( $shortcode_tags );
+
+	/**
+	 * Filters the list of shortcode tags to remove from the content.
+	 *
+	 * @since 4.7.0
+	 *
+	 * @param array  $tag_array Array of shortcode tags to remove.
+	 * @param string $content   Content shortcodes are being removed from.
+	 */
+	$tags_to_remove = apply_filters( 'strip_shortcodes_tagnames', $tags_to_remove, $content );
 
-	return preg_replace_callback( "/$pattern/s", 'strip_shortcode_tag', $content );
+	$tagnames = array_intersect( $tags_to_remove, $matches[1] );
+
+	if ( empty( $tagnames ) ) {
+		return $content;
+	}
+
+	$content = do_shortcodes_in_html_tags( $content, true, $tagnames );
+
+	$pattern = get_shortcode_regex( $tagnames );
+	$content = preg_replace_callback( "/$pattern/", 'strip_shortcode_tag', $content );
+
+	// Always restore square braces so we don't break things like <!--[if IE ]>
+	$content = unescape_invalid_shortcodes( $content );
+
+	return $content;
 }
 
+/**
+ * Strips a shortcode tag based on RegEx matches against post content.
+ *
+ * @since 3.3.0
+ *
+ * @param array $m RegEx matches against post content.
+ * @return string|false The content stripped of the tag, otherwise false.
+ */
 function strip_shortcode_tag( $m ) {
 	// allow [[foo]] syntax for escaping a tag
 	if ( $m[1] == '[' && $m[6] == ']' ) {