Searchengine.php 8.31 KB
Newer Older
1
2
<?php

3
namespace App\Models;
4
5

use App\Jobs\Search;
6
use App\MetaGer;
7
8
use Cache;
use Illuminate\Foundation\Bus\DispatchesJobs;
9
10
use Log;
use Redis;
11

12
abstract class Searchengine
13
{
14
    use DispatchesJobs;
15

16
17
18
19
20
    protected $ch; # Curl Handle zum erhalten der Ergebnisse
    public $fp;
    protected $getString = "";
    protected $engine;
    protected $counter      = 0;
21
    protected $socketNumber = null;
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
    public $enabled         = true;
    public $results         = [];
    public $ads             = [];
    public $write_time      = 0;
    public $connection_time = 0;
    public $loaded          = false;
    public $cached          = false;

    public function __construct(\SimpleXMLElement $engine, MetaGer $metager)
    {
        foreach ($engine->attributes() as $key => $value) {
            $this->$key = $value->__toString();
        }
        if (!isset($this->homepage)) {
            $this->homepage = "https://metager.de";
        }

39
        $this->engine = $engine->asXML();
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70

        if (!isset($this->cacheDuration)) {
            $this->cacheDuration = 60;
        }

        # Eine Suchmaschine kann automatisch temporär deaktiviert werden, wenn es Verbindungsprobleme gab:
        if (isset($this->disabled) && strtotime($this->disabled) <= time()) {
            # In diesem Fall ist der Timeout der Suchmaschine abgelaufen.
            $this->enable($metager->getSumaFile(), "Die Suchmaschine " . $this->name . " wurde wieder eingeschaltet.");
        } elseif (isset($this->disabled) && strtotime($this->disabled) > time()) {
            $this->enabled = false;
            return;
        }

        # User-Agent definieren:
        $this->useragent = "Mozilla/5.0 (Windows NT 6.1; WOW64; rv:40.0) Gecko/20100101 Firefox/40.1";
        $this->ip        = $metager->getIp();
        $this->gefVon    = "<a href=\"" . $this->homepage . "\" target=\"_blank\">" . $this->displayName . "</a>";
        $this->startTime = microtime();

        $q = "";
        if (isset($this->hasSiteSearch) && $this->hasSiteSearch === "1") {
            if (strlen($metager->getSite()) === 0) {
                $q = $metager->getQ();
            } else {
                $q = $metager->getQ() . " site:" . $metager->getSite();
            }

        } else {
            $q = $metager->getQ();
        }
Dominik Hebeler's avatar
Dominik Hebeler committed
71
        $this->getString  = $this->generateGetString($q, $metager->getUrl(), $metager->getLanguage(), $metager->getCategory());
72
73
        $this->hash       = md5($this->host . $this->getString . $this->port . $this->name);
        $this->resultHash = $metager->getHashCode();
74
        $this->canCache   = $metager->canCache();
75
76
77
78
    }

    abstract public function loadResults($result);

79
80
    public function getNext(MetaGer $metager, $result)
    {
81
82
83
84
85

    }

    public function startSearch(\App\MetaGer $metager)
    {
86
        if ($this->canCache && Cache::has($this->hash)) {
87
            $this->cached = true;
88
            $this->retrieveResults($metager);
89
90
91
92
93
94
        } else {
            # Die Anfragen an die Suchmaschinen werden nun von der Laravel-Queue bearbeitet:
            # Hinweis: solange in der .env der QUEUE_DRIVER auf "sync" gestellt ist, werden die Abfragen
            # nacheinander abgeschickt.
            # Sollen diese Parallel verarbeitet werden, muss ein anderer QUEUE_DRIVER verwendet werden.
            # siehe auch: https://laravel.com/docs/5.2/queues
95
            $this->dispatch(new Search($this->resultHash, $this->host, $this->port, $this->name, $this->getString, $this->useragent));
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
        }
    }

    public function rank(\App\MetaGer $metager)
    {
        foreach ($this->results as $result) {
            $result->rank($metager);
        }
    }

    private function setStatistic($key, $val)
    {

        $oldVal = floatval(Redis::hget($this->name, $key)) * $this->uses;
        $newVal = ($oldVal + max($val, 0)) / $this->uses;
        Redis::hset($this->name, $key, $newVal);
        $this->$key = $newVal;
    }

    public function enable($sumaFile, $message)
    {
        Log::info($message);
        $xml = simplexml_load_file($sumaFile);
        unset($xml->xpath("//sumas/suma[@name='" . $this->name . "']")['0']['disabled']);
        $xml->saveXML($sumaFile);
    }

    public function closeFp()
    {
        fclose($this->fp);
    }

    public function getSocket()
    {
        $number = Redis::hget('search.' . $this->hash, $this->name);
        if ($number === null) {
            die("test");
            return null;
        } else {
            return pfsockopen($this->getHost() . ":" . $this->port . "/$number", $this->port, $errstr, $errno, 1);
        }
    }

139
    public function retrieveResults(MetaGer $metager)
140
141
142
    {
        if ($this->loaded) {
            return true;
143
144
        }

145
        $body = "";
146
        if ($this->canCache && $this->cacheDuration > 0 && Cache::has($this->hash)) {
147
148
149
            $body = Cache::get($this->hash);
        } elseif (Redis::hexists('search.' . $this->resultHash, $this->name)) {
            $body = Redis::hget('search.' . $this->resultHash, $this->name);
150
            if ($this->canCache && $this->cacheDuration > 0) {
151
152
153
154
155
156
157
                Cache::put($this->hash, $body, $this->cacheDuration);
            }

        }

        if ($body !== "") {
            $this->loadResults($body);
158
            $this->getNext($metager, $body);
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
            $this->loaded = true;
            Redis::hdel('search.' . $this->hash, $this->name);
            return true;
        } else {
            return false;
        }
    }

    public function shutdown()
    {
        Redis::del($this->host . "." . $this->socketNumber);
    }

    protected function getHost()
    {
        $return = "";
        if ($this->port === "443") {
            $return .= "tls://";
        } else {
            $return .= "tcp://";
        }
        $return .= $this->host;
        return $return;
    }

    public function getCurlInfo()
    {
        return curl_getinfo($this->ch);
    }

    public function getCurlErrors()
    {
        return curl_errno($this->ch);
    }

    public function addCurlHandle($mh)
    {
        curl_multi_add_handle($mh, $this->ch);
    }

    public function removeCurlHandle($mh)
    {
        curl_multi_remove_handle($mh, $this->ch);
    }

    private function generateGetString($query, $url, $language, $category)
    {
        $getString = "";

        # Skript:
        if (strlen($this->skript) > 0) {
            $getString .= $this->skript;
        } else {
            $getString .= "/";
        }

        # FormData:
        if (strlen($this->formData) > 0) {
            $getString .= "?" . $this->formData;
        }

        # Wir müssen noch einige Platzhalter in dem GET-String ersetzen:
        if (strpos($getString, "<<USERAGENT>>")) {
            $getString = str_replace("<<USERAGENT>>", $this->urlEncode($this->useragent), $getString);
        }

        if (strpos($getString, "<<QUERY>>")) {
            $getString = str_replace("<<QUERY>>", $this->urlEncode($query), $getString);
        }

        if (strpos($getString, "<<IP>>")) {
            $getString = str_replace("<<IP>>", $this->urlEncode($this->ip), $getString);
        }

        if (strpos($getString, "<<LANGUAGE>>")) {
            $getString = str_replace("<<LANGUAGE>>", $this->urlEncode($language), $getString);
        }

        if (strpos($getString, "<<CATEGORY>>")) {
            $getString = str_replace("<<CATEGORY>>", $this->urlEncode($category), $getString);
        }

        if (strpos($getString, "<<AFFILDATA>>")) {
            $getString = str_replace("<<AFFILDATA>>", $this->getOvertureAffilData($url), $getString);
        }
        return $getString;
    }

    protected function urlEncode($string)
    {
        if (isset($this->inputEncoding)) {
            return urlencode(mb_convert_encoding($string, $this->inputEncoding));
        } else {
            return urlencode($string);
        }
    }

    private function getOvertureAffilData($url)
    {
        $affil_data = 'ip=' . $this->ip;
        $affil_data .= '&ua=' . $this->useragent;
        if (isset($_SERVER['HTTP_X_FORWARDED_FOR'])) {
            $affil_data .= '&xfip=' . $_SERVER['HTTP_X_FORWARDED_FOR'];
        }
        $affilDataValue = $this->urlEncode($affil_data);
        # Wir benötigen die ServeUrl:
        $serveUrl = $this->urlEncode($url);

        return "&affilData=" . $affilDataValue . "&serveUrl=" . $serveUrl;
    }

    public function isEnabled()
    {
        return $this->enabled;
    }
}