2 # Copyright (C) 2009 Aryeh Gregor
3 # http://www.mediawiki.org/
5 # This program is free software; you can redistribute it and/or modify
6 # it under the terms of the GNU General Public License as published by
7 # the Free Software Foundation; either version 2 of the License, or
8 # (at your option) any later version.
10 # This program is distributed in the hope that it will be useful,
11 # but WITHOUT ANY WARRANTY; without even the implied warranty of
12 # MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
13 # GNU General Public License for more details.
15 # You should have received a copy of the GNU General Public License along
16 # with this program; if not, write to the Free Software Foundation, Inc.,
17 # 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA.
18 # http://www.gnu.org/copyleft/gpl.html
21 * This class is a collection of static functions that serve two purposes:
23 * 1) Implement any algorithms specified by HTML 5, or other HTML
24 * specifications, in a convenient and self-contained way.
26 * 2) Allow HTML elements to be conveniently and safely generated, like the
27 * current Xml class but a) less confused (Xml supports HTML-specific things,
28 * but only sometimes!) and b) not necessarily confined to XML-compatible
31 * There are two important configuration options this class uses:
33 * $wgHtml5: If this is set to false, then all output should be valid XHTML 1.0
35 * $wgWellFormedXml: If this is set to true, then all output should be
36 * well-formed XML (quotes on attributes, self-closing tags, etc.).
38 * This class is meant to be confined to utility functions that are called from
39 * trusted code paths. It does not do enforcement of policy like not allowing
43 # List of void elements from HTML 5, section 9.1.2 as of 2009-08-10
44 private static $voidElements = array(
61 # Boolean attributes, which may have the value omitted entirely. Manually
62 # collected from the HTML 5 spec as of 2009-08-10.
63 private static $boolAttribs = array(
87 * Returns an HTML element in a string. The major advantage here over
88 * manually typing out the HTML is that it will escape all attribute
89 * values. If you're hardcoding all the attributes, or there are none, you
90 * should probably type out the string yourself.
92 * This is quite similar to Xml::element(), but it implements some useful
93 * HTML-specific logic. For instance, there is no $allowShortTag
94 * parameter: the closing tag is magically omitted if $element has an empty
95 * content model. If $wgWellFormedXml is false, then a few bytes will be
96 * shaved off the HTML output as well. In the future, other HTML-specific
97 * features might be added, like allowing arrays for the values of
98 * attributes like class= and media=.
100 * One notable difference to Xml::element() is that $contents is *not*
101 * escaped. This means that Html::element() can be usefully nested, rather
102 * than using the rather clumsy Xml::openElement() and Xml::closeElement().
104 * @param $element string The element's name, e.g., 'a'
105 * @param $attribs array Associative array of attributes, e.g., array(
106 * 'href' => 'http://www.mediawiki.org/' ). Values will be HTML-escaped.
107 * @param $contents string The raw HTML contents of the element: *not*
109 * @return string Raw HTML
111 public static function rawElement( $element, $attribs = array(), $contents = '' ) {
112 global $wgWellFormedXml;
113 $element = strtolower( $element );
114 $start = "<$element" . self
::expandAttributes( $attribs );
115 if ( in_array( $element, self
::$voidElements ) ) {
116 if ( $wgWellFormedXml ) {
121 return "$start>$contents</$element>";
126 * Identical to rawElement(), but HTML-escapes $contents.
128 public static function element( $element, $attribs = array(), $contents = '' ) {
129 return self
::rawElement( $element, $attribs, htmlspecialchars( $contents ) );
133 * Given an associative array of element attributes, generate a string
134 * to stick after the element name in HTML output. Like array( 'href' =>
135 * 'http://www.mediawiki.org/' ) becomes something like
136 * ' href="http://www.mediawiki.org"'. Again, this is like
137 * Xml::expandAttributes(), but it implements some HTML-specific logic.
138 * For instance, it will omit quotation marks if $wgWellFormedXml is false,
139 * and will treat boolean attributes specially.
141 * @param $attribs array Associative array of attributes, e.g., array(
142 * 'href' => 'http://www.mediawiki.org/' ). Values will be HTML-escaped.
143 * @return string HTML fragment that goes between element name and '>'
144 * (starting with a space if at least one attribute is output)
146 public static function expandAttributes( $attribs ) {
147 global $wgHtml5, $wgWellFormedXml;
150 foreach ( $attribs as $key => $value ) {
151 # See the "Attributes" section in the HTML syntax part of HTML 5,
152 # 9.1.2.3 as of 2009-08-10. Most attributes can have quotation
153 # marks omitted, but not all. (Although a literal " is not
154 # permitted, we don't check for that, since it will be escaped
156 if ( $wgWellFormedXml ||
$value == ''
157 ||
preg_match( "/[ '=<>]/", $value ) ) {
163 if ( in_array( $key, self
::$boolAttribs ) ) {
164 # In XHTML 1.0 Transitional, the value needs to be equal to the
165 # key. In HTML 5, we can leave the value empty instead. If we
166 # don't need well-formed XML, we can omit the = entirely.
167 if ( !$wgWellFormedXml ) {
169 } elseif ( $wgHtml5 ) {
170 $ret .= " $key=\"\"";
172 $ret .= " $key=\"$key\"";
175 # Apparently we need to entity-encode \n, \r, \t, although the
176 # spec doesn't mention that. Since we're doing strtr() anyway,
177 # and we don't need <> escaped here, we may as well not call
178 # htmlspecialchars(). FIXME: verify that we actually need to
179 # escape \n\r\t here, and explain why, exactly.
180 $ret .= " $key=$quote" . strtr( $value, array(
193 * Output a <script> tag with the given contents. TODO: do some useful
194 * escaping as well, like if $contents contains literal '</script>' or (for
195 * XML) literal "]]>".
197 * Note that $contents will not be escaped, since JS may legitimately
198 * contain unescaped characters like "<". Make sure you don't output
199 * untrusted user input here!
201 * @param $contents string JavaScript
202 * @return string Raw HTML
204 public static function inlineScript( $contents ) {
205 global $wgHtml5, $wgJsMimeType;
209 $attrs['type'] = $wgJsMimeType;
210 $contents = "/*<![CDATA[*/$contents/*]]>*/";
212 return self
::rawElement( 'script', $attrs, $contents );
216 * Output a <script> tag linking to the given URL, e.g.,
217 * <script src=foo.js></script>.
220 * @return string Raw HTML
222 public static function linkedScript( $url ) {
223 global $wgHtml5, $wgJsMimeType;
225 $attrs = array( 'src' => $url );
227 $attrs['type'] = $wgJsMimeType;
229 return self
::element( 'script', $attrs );
233 * Output a <style> tag with the given contents for the given media type
234 * (if any). TODO: do some useful escaping as well, like if $contents
235 * contains literal '</style>' (admittedly unlikely).
237 * Note that $contents will not be escaped, since CSS may legitimately
238 * contain unescaped characters like "<". Make sure you don't output
239 * untrusted user input here!
241 * @param $contents string CSS
242 * @param $media mixed A media type string, like 'screen', or null for all
244 * @return string Raw HTML
246 public static function inlineStyle( $contents, $media = null ) {
251 # Technically we should probably add CDATA stuff here like with
252 # scripts, but in practice, stylesheets tend not to have
253 # problematic characters anyway.
254 $attrs['type'] = 'text/css';
256 if ( $media !== null ) {
257 $attrs['media'] = $media;
259 return self
::rawElement( 'style', $attrs, $contents );
263 * Output a <link rel=stylesheet> linking to the given URL for the given
264 * media type (if any).
267 * @param $media mixed A media type string, like 'screen', or null for all
269 * @return string Raw HTML
271 public static function linkedStyle( $url, $media = null ) {
274 $attrs = array( 'rel' => 'stylesheet', 'href' => $url );
276 $attrs['type'] = 'text/css';
278 if ( $media !== null ) {
279 $attrs['media'] = $media;
281 return self
::element( 'link', $attrs );
285 * Convenience function to produce an <input> element. This supports the
286 * new HTML 5 input types and attributes, and will silently strip them if
289 * @param $name string name attribute
290 * @param $value mixed value attribute (null = omit)
291 * @param $type string type attribute
292 * @param $attribs array Assocative array of miscellaneous extra attributes,
293 * passed to Html::element()
294 * @return string Raw HTML
296 public static function input( $name, $value = null, $type = 'text', $attribs = array() ) {
300 // With $wgHtml5 off we want to validate as XHTML 1, so we
301 // strip out any fancy HTML 5-only input types for now.
303 // Whitelist of valid types:
316 if ( !in_array( $type, $validTypes ) ) {
319 // Here we're blacklisting some HTML5-only attributes...
320 $html5attribs = array(
331 foreach ( $html5attribs as $badAttr ) {
332 unset( $attribs[$badAttr] );
335 if ( $type != 'text' ) {
336 $attribs['type'] = $type;
338 if ( $value !== null ) {
339 $attribs['value'] = $value;
341 $attribs['name'] = $name;
343 return self
::element( 'input', $attribs );