WordPress/wp-includes/rewrite.php
Pascal Birchler f079f57357 Rewrite: Add a remove_rewrite_tag() helper function.
It can be used to properly remove registered rewrite tags. Adds unit tests.

Fixes #35236.
Built from https://develop.svn.wordpress.org/trunk@36217


git-svn-id: http://core.svn.wordpress.org/trunk@36184 1a063a9b-81f0-0310-95a4-ce76da25c4cd
2016-01-07 09:39:27 +00:00

591 lines
16 KiB
PHP

<?php
/**
* WordPress Rewrite API
*
* @package WordPress
* @subpackage Rewrite
*/
/**
* Endpoint Mask for default, which is nothing.
*
* @since 2.1.0
*/
define('EP_NONE', 0);
/**
* Endpoint Mask for Permalink.
*
* @since 2.1.0
*/
define('EP_PERMALINK', 1);
/**
* Endpoint Mask for Attachment.
*
* @since 2.1.0
*/
define('EP_ATTACHMENT', 2);
/**
* Endpoint Mask for date.
*
* @since 2.1.0
*/
define('EP_DATE', 4);
/**
* Endpoint Mask for year
*
* @since 2.1.0
*/
define('EP_YEAR', 8);
/**
* Endpoint Mask for month.
*
* @since 2.1.0
*/
define('EP_MONTH', 16);
/**
* Endpoint Mask for day.
*
* @since 2.1.0
*/
define('EP_DAY', 32);
/**
* Endpoint Mask for root.
*
* @since 2.1.0
*/
define('EP_ROOT', 64);
/**
* Endpoint Mask for comments.
*
* @since 2.1.0
*/
define('EP_COMMENTS', 128);
/**
* Endpoint Mask for searches.
*
* @since 2.1.0
*/
define('EP_SEARCH', 256);
/**
* Endpoint Mask for categories.
*
* @since 2.1.0
*/
define('EP_CATEGORIES', 512);
/**
* Endpoint Mask for tags.
*
* @since 2.3.0
*/
define('EP_TAGS', 1024);
/**
* Endpoint Mask for authors.
*
* @since 2.1.0
*/
define('EP_AUTHORS', 2048);
/**
* Endpoint Mask for pages.
*
* @since 2.1.0
*/
define('EP_PAGES', 4096);
/**
* Endpoint Mask for all archive views.
*
* @since 3.7.0
*/
define( 'EP_ALL_ARCHIVES', EP_DATE | EP_YEAR | EP_MONTH | EP_DAY | EP_CATEGORIES | EP_TAGS | EP_AUTHORS );
/**
* Endpoint Mask for everything.
*
* @since 2.1.0
*/
define( 'EP_ALL', EP_PERMALINK | EP_ATTACHMENT | EP_ROOT | EP_COMMENTS | EP_SEARCH | EP_PAGES | EP_ALL_ARCHIVES );
/**
* Adds a rewrite rule that transforms a URL structure to a set of query vars.
*
* Any value in the $after parameter that isn't 'bottom' will result in the rule
* being placed at the top of the rewrite rules.
*
* @since 2.1.0
* @since 4.4.0 Array support was added to the `$query` parameter.
*
* @global WP_Rewrite $wp_rewrite WordPress Rewrite Component.
*
* @param string $regex Regular expression to match request against.
* @param string|array $query The corresponding query vars for this rewrite rule.
* @param string $after Optional. Priority of the new rule. Accepts 'top'
* or 'bottom'. Default 'bottom'.
*/
function add_rewrite_rule( $regex, $query, $after = 'bottom' ) {
global $wp_rewrite;
$wp_rewrite->add_rule( $regex, $query, $after );
}
/**
* Add a new rewrite tag (like %postname%).
*
* The $query parameter is optional. If it is omitted you must ensure that
* you call this on, or before, the 'init' hook. This is because $query defaults
* to "$tag=", and for this to work a new query var has to be added.
*
* @since 2.1.0
*
* @global WP_Rewrite $wp_rewrite
* @global WP $wp
*
* @param string $tag Name of the new rewrite tag.
* @param string $regex Regular expression to substitute the tag for in rewrite rules.
* @param string $query Optional. String to append to the rewritten query. Must end in '='. Default empty.
*/
function add_rewrite_tag( $tag, $regex, $query = '' ) {
// validate the tag's name
if ( strlen( $tag ) < 3 || $tag[0] != '%' || $tag[ strlen($tag) - 1 ] != '%' )
return;
global $wp_rewrite, $wp;
if ( empty( $query ) ) {
$qv = trim( $tag, '%' );
$wp->add_query_var( $qv );
$query = $qv . '=';
}
$wp_rewrite->add_rewrite_tag( $tag, $regex, $query );
}
/**
* Removes an existing rewrite tag (like %postname%).
*
* @since 4.5.0
*
* @global WP_Rewrite $wp_rewrite WordPress rewrite component.
*
* @param string $tag Name of the rewrite tag.
*/
function remove_rewrite_tag( $tag ) {
global $wp_rewrite;
$wp_rewrite->remove_rewrite_tag( $tag );
}
/**
* Add permalink structure.
*
* @since 3.0.0
*
* @see WP_Rewrite::add_permastruct()
* @global WP_Rewrite $wp_rewrite WordPress rewrite component.
*
* @param string $name Name for permalink structure.
* @param string $struct Permalink structure.
* @param array $args Optional. Arguments for building the rules from the permalink structure,
* see WP_Rewrite::add_permastruct() for full details. Default empty array.
*/
function add_permastruct( $name, $struct, $args = array() ) {
global $wp_rewrite;
// backwards compatibility for the old parameters: $with_front and $ep_mask
if ( ! is_array( $args ) )
$args = array( 'with_front' => $args );
if ( func_num_args() == 4 )
$args['ep_mask'] = func_get_arg( 3 );
$wp_rewrite->add_permastruct( $name, $struct, $args );
}
/**
* Remove permalink structure.
*
* Can only be used to remove permastructs that were added using add_permastruct().
* Built-in permastructs cannot be removed.
*
* @since 4.5.0
*
* @see WP_Rewrite::remove_permastruct()
* @global WP_Rewrite $wp_rewrite WordPress rewrite component.
*
* @param string $name Name for permalink structure.
*/
function remove_permastruct( $name ) {
global $wp_rewrite;
$wp_rewrite->remove_permastruct( $name );
}
/**
* Add a new feed type like /atom1/.
*
* @since 2.1.0
*
* @global WP_Rewrite $wp_rewrite
*
* @param string $feedname Feed name.
* @param callable $function Callback to run on feed display.
* @return string Feed action name.
*/
function add_feed( $feedname, $function ) {
global $wp_rewrite;
if ( ! in_array( $feedname, $wp_rewrite->feeds ) ) {
$wp_rewrite->feeds[] = $feedname;
}
$hook = 'do_feed_' . $feedname;
// Remove default function hook
remove_action( $hook, $hook );
add_action( $hook, $function, 10, 2 );
return $hook;
}
/**
* Remove rewrite rules and then recreate rewrite rules.
*
* @since 3.0.0
*
* @global WP_Rewrite $wp_rewrite
*
* @param bool $hard Whether to update .htaccess (hard flush) or just update
* rewrite_rules transient (soft flush). Default is true (hard).
*/
function flush_rewrite_rules( $hard = true ) {
global $wp_rewrite;
$wp_rewrite->flush_rules( $hard );
}
/**
* Add an endpoint, like /trackback/.
*
* Adding an endpoint creates extra rewrite rules for each of the matching
* places specified by the provided bitmask. For example:
*
* add_rewrite_endpoint( 'json', EP_PERMALINK | EP_PAGES );
*
* will add a new rewrite rule ending with "json(/(.*))?/?$" for every permastruct
* that describes a permalink (post) or page. This is rewritten to "json=$match"
* where $match is the part of the URL matched by the endpoint regex (e.g. "foo" in
* "[permalink]/json/foo/").
*
* A new query var with the same name as the endpoint will also be created.
*
* When specifying $places ensure that you are using the EP_* constants (or a
* combination of them using the bitwise OR operator) as their values are not
* guaranteed to remain static (especially `EP_ALL`).
*
* Be sure to flush the rewrite rules - see flush_rewrite_rules() - when your plugin gets
* activated and deactivated.
*
* @since 2.1.0
* @since 4.3.0 Added support for skipping query var registration by passing `false` to `$query_var`.
*
* @global WP_Rewrite $wp_rewrite
*
* @param string $name Name of the endpoint.
* @param int $places Endpoint mask describing the places the endpoint should be added.
* @param string|bool $query_var Name of the corresponding query variable. Pass `false` to skip registering a query_var
* for this endpoint. Defaults to the value of `$name`.
*/
function add_rewrite_endpoint( $name, $places, $query_var = true ) {
global $wp_rewrite;
$wp_rewrite->add_endpoint( $name, $places, $query_var );
}
/**
* Filter the URL base for taxonomies.
*
* To remove any manually prepended /index.php/.
*
* @access private
* @since 2.6.0
*
* @param string $base The taxonomy base that we're going to filter
* @return string
*/
function _wp_filter_taxonomy_base( $base ) {
if ( !empty( $base ) ) {
$base = preg_replace( '|^/index\.php/|', '', $base );
$base = trim( $base, '/' );
}
return $base;
}
/**
* Resolve numeric slugs that collide with date permalinks.
*
* Permalinks of posts with numeric slugs can sometimes look to WP_Query::parse_query()
* like a date archive, as when your permalink structure is `/%year%/%postname%/` and
* a post with post_name '05' has the URL `/2015/05/`.
*
* This function detects conflicts of this type and resolves them in favor of the
* post permalink.
*
* Note that, since 4.3.0, wp_unique_post_slug() prevents the creation of post slugs
* that would result in a date archive conflict. The resolution performed in this
* function is primarily for legacy content, as well as cases when the admin has changed
* the site's permalink structure in a way that introduces URL conflicts.
*
* @since 4.3.0
*
* @param array $query_vars Optional. Query variables for setting up the loop, as determined in
* WP::parse_request(). Default empty array.
* @return array Returns the original array of query vars, with date/post conflicts resolved.
*/
function wp_resolve_numeric_slug_conflicts( $query_vars = array() ) {
if ( ! isset( $query_vars['year'] ) && ! isset( $query_vars['monthnum'] ) && ! isset( $query_vars['day'] ) ) {
return $query_vars;
}
// Identify the 'postname' position in the permastruct array.
$permastructs = array_values( array_filter( explode( '/', get_option( 'permalink_structure' ) ) ) );
$postname_index = array_search( '%postname%', $permastructs );
if ( false === $postname_index ) {
return $query_vars;
}
/*
* A numeric slug could be confused with a year, month, or day, depending on position. To account for
* the possibility of post pagination (eg 2015/2 for the second page of a post called '2015'), our
* `is_*` checks are generous: check for year-slug clashes when `is_year` *or* `is_month`, and check
* for month-slug clashes when `is_month` *or* `is_day`.
*/
$compare = '';
if ( 0 === $postname_index && ( isset( $query_vars['year'] ) || isset( $query_vars['monthnum'] ) ) ) {
$compare = 'year';
} elseif ( '%year%' === $permastructs[ $postname_index - 1 ] && ( isset( $query_vars['monthnum'] ) || isset( $query_vars['day'] ) ) ) {
$compare = 'monthnum';
} elseif ( '%monthnum%' === $permastructs[ $postname_index - 1 ] && isset( $query_vars['day'] ) ) {
$compare = 'day';
}
if ( ! $compare ) {
return $query_vars;
}
// This is the potentially clashing slug.
$value = $query_vars[ $compare ];
$post = get_page_by_path( $value, OBJECT, 'post' );
if ( ! ( $post instanceof WP_Post ) ) {
return $query_vars;
}
// If the date of the post doesn't match the date specified in the URL, resolve to the date archive.
if ( preg_match( '/^([0-9]{4})\-([0-9]{2})/', $post->post_date, $matches ) && isset( $query_vars['year'] ) && ( 'monthnum' === $compare || 'day' === $compare ) ) {
// $matches[1] is the year the post was published.
if ( intval( $query_vars['year'] ) !== intval( $matches[1] ) ) {
return $query_vars;
}
// $matches[2] is the month the post was published.
if ( 'day' === $compare && isset( $query_vars['monthnum'] ) && intval( $query_vars['monthnum'] ) !== intval( $matches[2] ) ) {
return $query_vars;
}
}
/*
* If the located post contains nextpage pagination, then the URL chunk following postname may be
* intended as the page number. Verify that it's a valid page before resolving to it.
*/
$maybe_page = '';
if ( 'year' === $compare && isset( $query_vars['monthnum'] ) ) {
$maybe_page = $query_vars['monthnum'];
} elseif ( 'monthnum' === $compare && isset( $query_vars['day'] ) ) {
$maybe_page = $query_vars['day'];
}
// Bug found in #11694 - 'page' was returning '/4'
$maybe_page = (int) trim( $maybe_page, '/' );
$post_page_count = substr_count( $post->post_content, '<!--nextpage-->' ) + 1;
// If the post doesn't have multiple pages, but a 'page' candidate is found, resolve to the date archive.
if ( 1 === $post_page_count && $maybe_page ) {
return $query_vars;
}
// If the post has multiple pages and the 'page' number isn't valid, resolve to the date archive.
if ( $post_page_count > 1 && $maybe_page > $post_page_count ) {
return $query_vars;
}
// If we've gotten to this point, we have a slug/date clash. First, adjust for nextpage.
if ( '' !== $maybe_page ) {
$query_vars['page'] = intval( $maybe_page );
}
// Next, unset autodetected date-related query vars.
unset( $query_vars['year'] );
unset( $query_vars['monthnum'] );
unset( $query_vars['day'] );
// Then, set the identified post.
$query_vars['name'] = $post->post_name;
// Finally, return the modified query vars.
return $query_vars;
}
/**
* Examine a url and try to determine the post ID it represents.
*
* Checks are supposedly from the hosted site blog.
*
* @since 1.0.0
*
* @global WP_Rewrite $wp_rewrite
* @global WP $wp
*
* @param string $url Permalink to check.
* @return int Post ID, or 0 on failure.
*/
function url_to_postid( $url ) {
global $wp_rewrite;
/**
* Filter the URL to derive the post ID from.
*
* @since 2.2.0
*
* @param string $url The URL to derive the post ID from.
*/
$url = apply_filters( 'url_to_postid', $url );
// First, check to see if there is a 'p=N' or 'page_id=N' to match against
if ( preg_match('#[?&](p|page_id|attachment_id)=(\d+)#', $url, $values) ) {
$id = absint($values[2]);
if ( $id )
return $id;
}
// Check to see if we are using rewrite rules
$rewrite = $wp_rewrite->wp_rewrite_rules();
// Not using rewrite rules, and 'p=N' and 'page_id=N' methods failed, so we're out of options
if ( empty($rewrite) )
return 0;
// Get rid of the #anchor
$url_split = explode('#', $url);
$url = $url_split[0];
// Get rid of URL ?query=string
$url_split = explode('?', $url);
$url = $url_split[0];
// Set the correct URL scheme.
$url = set_url_scheme( $url );
// Add 'www.' if it is absent and should be there
if ( false !== strpos(home_url(), '://www.') && false === strpos($url, '://www.') )
$url = str_replace('://', '://www.', $url);
// Strip 'www.' if it is present and shouldn't be
if ( false === strpos(home_url(), '://www.') )
$url = str_replace('://www.', '://', $url);
// Strip 'index.php/' if we're not using path info permalinks
if ( !$wp_rewrite->using_index_permalinks() )
$url = str_replace( $wp_rewrite->index . '/', '', $url );
if ( false !== strpos( trailingslashit( $url ), home_url( '/' ) ) ) {
// Chop off http://domain.com/[path]
$url = str_replace(home_url(), '', $url);
} else {
// Chop off /path/to/blog
$home_path = parse_url( home_url( '/' ) );
$home_path = isset( $home_path['path'] ) ? $home_path['path'] : '' ;
$url = preg_replace( sprintf( '#^%s#', preg_quote( $home_path ) ), '', trailingslashit( $url ) );
}
// Trim leading and lagging slashes
$url = trim($url, '/');
$request = $url;
$post_type_query_vars = array();
foreach ( get_post_types( array() , 'objects' ) as $post_type => $t ) {
if ( ! empty( $t->query_var ) )
$post_type_query_vars[ $t->query_var ] = $post_type;
}
// Look for matches.
$request_match = $request;
foreach ( (array)$rewrite as $match => $query) {
// If the requesting file is the anchor of the match, prepend it
// to the path info.
if ( !empty($url) && ($url != $request) && (strpos($match, $url) === 0) )
$request_match = $url . '/' . $request;
if ( preg_match("#^$match#", $request_match, $matches) ) {
if ( $wp_rewrite->use_verbose_page_rules && preg_match( '/pagename=\$matches\[([0-9]+)\]/', $query, $varmatch ) ) {
// This is a verbose page match, let's check to be sure about it.
$page = get_page_by_path( $matches[ $varmatch[1] ] );
if ( ! $page ) {
continue;
}
$post_status_obj = get_post_status_object( $page->post_status );
if ( ! $post_status_obj->public && ! $post_status_obj->protected
&& ! $post_status_obj->private && $post_status_obj->exclude_from_search ) {
continue;
}
}
// Got a match.
// Trim the query of everything up to the '?'.
$query = preg_replace("!^.+\?!", '', $query);
// Substitute the substring matches into the query.
$query = addslashes(WP_MatchesMapRegex::apply($query, $matches));
// Filter out non-public query vars
global $wp;
parse_str( $query, $query_vars );
$query = array();
foreach ( (array) $query_vars as $key => $value ) {
if ( in_array( $key, $wp->public_query_vars ) ){
$query[$key] = $value;
if ( isset( $post_type_query_vars[$key] ) ) {
$query['post_type'] = $post_type_query_vars[$key];
$query['name'] = $value;
}
}
}
// Resolve conflicts between posts with numeric slugs and date archive queries.
$query = wp_resolve_numeric_slug_conflicts( $query );
// Do the query
$query = new WP_Query( $query );
if ( ! empty( $query->posts ) && $query->is_singular )
return $query->post->ID;
else
return 0;
}
}
return 0;
}