3 * Interfaces for preprocessors
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
24 use MediaWiki\Logger\LoggerFactory
;
29 abstract class Preprocessor
{
31 const CACHE_VERSION
= 1;
34 * @var array Brace matching rules.
48 'names' => [ 2 => null ],
54 'names' => [ 2 => null ],
61 * Store a document tree in the cache.
67 protected function cacheSetTree( $text, $flags, $tree ) {
68 $config = RequestContext
::getMain()->getConfig();
70 $length = strlen( $text );
71 $threshold = $config->get( 'PreprocessorCacheThreshold' );
72 if ( $threshold === false ||
$length < $threshold ||
$length > 1e6
) {
76 $cache = ObjectCache
::getLocalClusterInstance();
77 $key = $cache->makeKey(
78 defined( 'static::CACHE_PREFIX' ) ?
static::CACHE_PREFIX
: static::class,
79 md5( $text ), $flags );
80 $value = sprintf( "%08d", static::CACHE_VERSION
) . $tree;
82 $cache->set( $key, $value, 86400 );
84 LoggerFactory
::getInstance( 'Preprocessor' )
85 ->info( "Cached preprocessor output (key: $key)" );
89 * Attempt to load a precomputed document tree for some given wikitext
94 * @return PPNode_Hash_Tree|bool
96 protected function cacheGetTree( $text, $flags ) {
97 $config = RequestContext
::getMain()->getConfig();
99 $length = strlen( $text );
100 $threshold = $config->get( 'PreprocessorCacheThreshold' );
101 if ( $threshold === false ||
$length < $threshold ||
$length > 1e6
) {
105 $cache = ObjectCache
::getLocalClusterInstance();
107 $key = $cache->makeKey(
108 defined( 'static::CACHE_PREFIX' ) ?
static::CACHE_PREFIX
: static::class,
109 md5( $text ), $flags );
111 $value = $cache->get( $key );
116 $version = intval( substr( $value, 0, 8 ) );
117 if ( $version !== static::CACHE_VERSION
) {
121 LoggerFactory
::getInstance( 'Preprocessor' )
122 ->info( "Loaded preprocessor output from cache (key: $key)" );
124 return substr( $value, 8 );
128 * Create a new top-level frame for expansion of a page
132 abstract public function newFrame();
135 * Create a new custom frame for programmatic use of parameter replacement
136 * as used in some extensions.
142 abstract public function newCustomFrame( $args );
145 * Create a new custom node for programmatic use of parameter replacement
146 * as used in some extensions.
148 * @param array $values
150 abstract public function newPartNodeArray( $values );
153 * Preprocess text to a PPNode
155 * @param string $text
160 abstract public function preprocessToObj( $text, $flags = 0 );
168 const NO_TEMPLATES
= 2;
169 const STRIP_COMMENTS
= 4;
171 const RECOVER_COMMENTS
= 16;
174 const RECOVER_ORIG
= self
::NO_ARGS | self
::NO_TEMPLATES | self
::NO_IGNORE |
175 self
::RECOVER_COMMENTS | self
::NO_TAGS
;
177 /** This constant exists when $indexOffset is supported in newChild() */
178 const SUPPORTS_INDEX_OFFSET
= 1;
181 * Create a child frame
183 * @param array|bool $args
184 * @param bool|Title $title
185 * @param int $indexOffset A number subtracted from the index attributes of the arguments
189 public function newChild( $args = false, $title = false, $indexOffset = 0 );
192 * Expand a document tree node, caching the result on its parent with the given key
193 * @param string|int $key
194 * @param string|PPNode $root
198 public function cachedExpand( $key, $root, $flags = 0 );
201 * Expand a document tree node
202 * @param string|PPNode $root
206 public function expand( $root, $flags = 0 );
209 * Implode with flags for expand()
212 * @param string|PPNode $args,...
215 public function implodeWithFlags( $sep, $flags /*, ... */ );
218 * Implode with no flags specified
220 * @param string|PPNode $args,...
223 public function implode( $sep /*, ... */ );
226 * Makes an object that, when expand()ed, will be the same as one obtained
229 * @param string|PPNode $args,...
232 public function virtualImplode( $sep /*, ... */ );
235 * Virtual implode with brackets
236 * @param string $start
239 * @param string|PPNode $args,...
242 public function virtualBracketedImplode( $start, $sep, $end /*, ... */ );
245 * Returns true if there are no arguments in this frame
249 public function isEmpty();
252 * Returns all arguments of this frame
255 public function getArguments();
258 * Returns all numbered arguments of this frame
261 public function getNumberedArguments();
264 * Returns all named arguments of this frame
267 public function getNamedArguments();
270 * Get an argument to this frame by name
271 * @param int|string $name
272 * @return string|bool
274 public function getArgument( $name );
277 * Returns true if the infinite loop check is OK, false if a loop is detected
279 * @param Title $title
282 public function loopCheck( $title );
285 * Return true if the frame is a template frame
288 public function isTemplate();
291 * Set the "volatile" flag.
293 * Note that this is somewhat of a "hack" in order to make extensions
294 * with side effects (such as Cite) work with the PHP parser. New
295 * extensions should be written in a way that they do not need this
296 * function, because other parsers (such as Parsoid) are not guaranteed
297 * to respect it, and it may be removed in the future.
301 public function setVolatile( $flag = true );
304 * Get the "volatile" flag.
306 * Callers should avoid caching the result of an expansion if it has the
309 * @see self::setVolatile()
312 public function isVolatile();
315 * Get the TTL of the frame's output.
317 * This is the maximum amount of time, in seconds, that this frame's
318 * output should be cached for. A value of null indicates that no
319 * maximum has been specified.
321 * Note that this TTL only applies to caching frames as parts of pages.
322 * It is not relevant to caching the entire rendered output of a page.
326 public function getTTL();
329 * Set the TTL of the output of this frame and all of its ancestors.
330 * Has no effect if the new TTL is greater than the one already set.
331 * Note that it is the caller's responsibility to change the cache
332 * expiry of the page as a whole, if such behavior is desired.
334 * @see self::getTTL()
337 public function setTTL( $ttl );
340 * Get a title of frame
344 public function getTitle();
348 * There are three types of nodes:
349 * * Tree nodes, which have a name and contain other nodes as children
350 * * Array nodes, which also contain other nodes but aren't considered part of a tree
351 * * Leaf nodes, which contain the actual data
353 * This interface provides access to the tree structure and to the contents of array nodes,
354 * but it does not provide access to the internal structure of leaf nodes. Access to leaf
355 * data is provided via two means:
356 * * PPFrame::expand(), which provides expanded text
357 * * The PPNode::split*() functions, which provide metadata about certain types of tree node
362 * Get an array-type node containing the children of this node.
363 * Returns false if this is not a tree node.
366 public function getChildren();
369 * Get the first child of a tree node. False if there isn't one.
373 public function getFirstChild();
376 * Get the next sibling of any node. False if there isn't one
379 public function getNextSibling();
382 * Get all children of this tree node which have a given name.
383 * Returns an array-type node, or false if this is not a tree node.
384 * @param string $type
385 * @return bool|PPNode
387 public function getChildrenOfType( $type );
390 * Returns the length of the array, or false if this is not an array-type node
392 public function getLength();
395 * Returns an item of an array-type node
397 * @return bool|PPNode
399 public function item( $i );
402 * Get the name of this node. The following names are defined here:
405 * template A double-brace node.
406 * tplarg A triple-brace node.
407 * title The first argument to a template or tplarg node.
408 * part Subsequent arguments to a template or tplarg node.
409 * #nodelist An array-type node
411 * The subclass may define various other names for tree and leaf nodes.
414 public function getName();
417 * Split a "<part>" node into an associative array containing:
423 public function splitArg();
426 * Split an "<ext>" node into an associative array containing name, attr, inner and close
427 * All values in the resulting array are PPNodes. Inner and close are optional.
430 public function splitExt();
433 * Split an "<h>" node
436 public function splitHeading();