1<?php
2
3/**
4 * Sitemap handling functions
5 *
6 * @license    GPL 2 (http://www.gnu.org/licenses/gpl.html)
7 * @author     Michael Hamann <michael@content-space.de>
8 */
9
10namespace dokuwiki\Sitemap;
11
12use dokuwiki\Extension\Event;
13use dokuwiki\HTTP\DokuHTTPClient;
14use dokuwiki\Logger;
15
16/**
17 * A class for building sitemaps and pinging search engines with the sitemap URL.
18 *
19 * @author Michael Hamann
20 */
21class Mapper
22{
23    /**
24     * Builds a Google Sitemap of all public pages known to the indexer
25     *
26     * The map is placed in the cache directory named sitemap.xml.gz - This
27     * file needs to be writable!
28     *
29     * @author Michael Hamann
30     * @author Andreas Gohr
31     * @link   https://www.google.com/webmasters/sitemaps/docs/en/about.html
32     * @link   http://www.sitemaps.org/
33     *
34     * @return bool
35     */
36    public static function generate()
37    {
38        global $conf;
39        if ($conf['sitemap'] < 1 || !is_numeric($conf['sitemap'])) return false;
40
41        $sitemap = Mapper::getFilePath();
42
43        if (file_exists($sitemap)) {
44            if (!is_writable($sitemap)) return false;
45        } elseif (!is_writable(dirname($sitemap))) {
46            return false;
47        }
48
49        if (
50            @filesize($sitemap) &&
51            @filemtime($sitemap) > (time() - ($conf['sitemap'] * 86400))
52        ) { // 60*60*24=86400
53            Logger::debug('Sitemapper::generate(): Sitemap up to date');
54            return false;
55        }
56
57        Logger::debug("Sitemapper::generate(): using $sitemap");
58
59        $pages = idx_get_indexer()->getPages();
60        Logger::debug('Sitemapper::generate(): creating sitemap using ' . count($pages) . ' pages');
61        $items = [];
62
63        // build the sitemap items
64        foreach ($pages as $id) {
65            //skip hidden, non existing and restricted files
66            if (isHiddenPage($id)) continue;
67            if (auth_aclcheck($id, '', []) < AUTH_READ) continue;
68            $item = Item::createFromID($id);
69            if ($item instanceof Item)
70                $items[] = $item;
71        }
72
73        $eventData = ['items' => &$items, 'sitemap' => &$sitemap];
74        $event = new Event('SITEMAP_GENERATE', $eventData);
75        if ($event->advise_before(true)) {
76            //save the new sitemap
77            $event->result = io_saveFile($sitemap, (new Mapper())->getXML($items));
78        }
79        $event->advise_after();
80
81        return $event->result;
82    }
83
84    /**
85     * Builds the sitemap XML string from the given array auf SitemapItems.
86     *
87     * @param $items array The SitemapItems that shall be included in the sitemap.
88     * @return string The sitemap XML.
89     *
90     * @author Michael Hamann
91     */
92    private function getXML($items)
93    {
94        ob_start();
95        echo '<?xml version="1.0" encoding="UTF-8"?>' . NL;
96        echo '<urlset xmlns="http://www.sitemaps.org/schemas/sitemap/0.9">' . NL;
97        foreach ($items as $item) {
98            /** @var Item $item */
99            echo $item->toXML();
100        }
101        echo '</urlset>' . NL;
102        $result = ob_get_contents();
103        ob_end_clean();
104        return $result;
105    }
106
107    /**
108     * Helper function for getting the path to the sitemap file.
109     *
110     * @return string The path to the sitemap file.
111     *
112     * @author Michael Hamann
113     */
114    public static function getFilePath()
115    {
116        global $conf;
117
118        $sitemap = $conf['cachedir'] . '/sitemap.xml';
119        if (self::sitemapIsCompressed()) {
120            $sitemap .= '.gz';
121        }
122
123        return $sitemap;
124    }
125
126    /**
127     * Helper function for checking if the sitemap is compressed
128     *
129     * @return bool If the sitemap file is compressed
130     */
131    public static function sitemapIsCompressed()
132    {
133        global $conf;
134        return $conf['compression'] === 'bz2' || $conf['compression'] === 'gz';
135    }
136
137    /**
138     * Pings search engines with the sitemap url. Plugins can add or remove
139     * urls to ping using the SITEMAP_PING event.
140     *
141     * @author Michael Hamann
142     *
143     * @return bool
144     */
145    public static function pingSearchEngines()
146    {
147        //ping search engines...
148        $http = new DokuHTTPClient();
149        $http->timeout = 8;
150
151        $encoded_sitemap_url = urlencode(wl('', ['do' => 'sitemap'], true, '&'));
152        $ping_urls = [
153            'google'    => 'https://www.google.com/ping?sitemap=' . $encoded_sitemap_url,
154            'yandex'    => 'https://webmaster.yandex.com/ping?sitemap=' . $encoded_sitemap_url
155        ];
156
157        $data = [
158            'ping_urls' => $ping_urls,
159            'encoded_sitemap_url' => $encoded_sitemap_url
160        ];
161        $event = new Event('SITEMAP_PING', $data);
162        if ($event->advise_before(true)) {
163            foreach ($data['ping_urls'] as $name => $url) {
164                Logger::debug("Sitemapper::PingSearchEngines(): pinging $name");
165                $resp = $http->get($url);
166                if ($http->error) {
167                    Logger::debug("Sitemapper:pingSearchengines(): $http->error", $resp);
168                }
169            }
170        }
171        $event->advise_after();
172
173        return true;
174    }
175}
176