1
|
|
|
<?php |
2
|
|
|
|
3
|
|
|
declare(strict_types=1); |
4
|
|
|
|
5
|
|
|
namespace App\Controller; |
6
|
|
|
|
7
|
|
|
use App\Model\Query; |
8
|
|
|
use App\Repository\CloudElasticRepository; |
9
|
|
|
use App\Repository\WikiDomainRepository; |
10
|
|
|
use GuzzleHttp\Client; |
11
|
|
|
use Psr\Cache\CacheItemPoolInterface; |
12
|
|
|
use Symfony\Bundle\FrameworkBundle\Controller\AbstractController; |
13
|
|
|
use Symfony\Component\HttpFoundation\Request; |
14
|
|
|
use Symfony\Component\HttpFoundation\Response; |
15
|
|
|
use Symfony\Component\Routing\Annotation\Route; |
16
|
|
|
|
17
|
|
|
/** |
18
|
|
|
* A DefaultController serves the main routes for the application and processing the submitted form. |
19
|
|
|
*/ |
20
|
|
|
class DefaultController extends AbstractController |
21
|
|
|
{ |
22
|
|
|
/** @var Client */ |
23
|
|
|
private $client; |
24
|
|
|
|
25
|
|
|
/** @var CacheItemPoolInterface */ |
26
|
|
|
private $cache; |
27
|
|
|
|
28
|
|
|
/** @var bool Whether the results were pulled from cache. */ |
29
|
|
|
private $fromCache = false; |
30
|
|
|
|
31
|
|
|
/** @var string Duration of cache for main results set, as accepted by DateInterval::createFromDateString() */ |
32
|
|
|
private const CACHE_TIME = '10 minutes'; |
33
|
|
|
|
34
|
|
|
/** @var string[]|null Map from wiki dbname to domain name */ |
35
|
|
|
private $domainLookup; |
36
|
|
|
|
37
|
|
|
/** |
38
|
|
|
* DefaultController constructor. |
39
|
|
|
* @param CacheItemPoolInterface $cache |
40
|
|
|
*/ |
41
|
|
|
public function __construct(CacheItemPoolInterface $cache) |
42
|
|
|
{ |
43
|
|
|
$this->client = new Client([ |
44
|
|
|
'verify' => $_ENV['ELASTIC_INSECURE'] ? false : true, |
45
|
|
|
]); |
46
|
|
|
$this->cache = $cache; |
47
|
|
|
} |
48
|
|
|
|
49
|
|
|
/** |
50
|
|
|
* Splash page, shown when user is logged out. |
51
|
|
|
* @Route("/splash") |
52
|
|
|
*/ |
53
|
|
|
public function splashAction(): Response |
54
|
|
|
{ |
55
|
|
|
return $this->render('jumbotron.html.twig'); |
56
|
|
|
} |
57
|
|
|
|
58
|
|
|
/** |
59
|
|
|
* The main route. |
60
|
|
|
* @Route("/", name="home") |
61
|
|
|
* @param Request $request |
62
|
|
|
* @return Response |
63
|
|
|
*/ |
64
|
|
|
public function indexAction(Request $request): Response |
65
|
|
|
{ |
66
|
|
|
if (!$this->get('session')->get('logged_in_user')) { |
67
|
|
|
return $this->render('jumbotron.html.twig'); |
68
|
|
|
} |
69
|
|
|
$query = $request->query->get('q'); |
70
|
|
|
$regex = (bool)$request->query->get('regex'); |
71
|
|
|
$ignoreCase = (bool)$request->query->get('ignorecase'); |
72
|
|
|
[$namespaces, $namespaceIds] = $this->parseNamespaces($request); |
73
|
|
|
$purgeCache = (bool)$request->query->get('purge'); |
74
|
|
|
|
75
|
|
|
$ret = [ |
76
|
|
|
'q' => $query, |
77
|
|
|
'regex' => $regex, |
78
|
|
|
'max_results' => Query::MAX_RESULTS, |
79
|
|
|
'namespaces' => $namespaces, |
80
|
|
|
'ignore_case' => $ignoreCase, |
81
|
|
|
]; |
82
|
|
|
|
83
|
|
|
if ($query) { |
84
|
|
|
$ret = array_merge($ret, $this->getResults($query, $regex, $ignoreCase, $namespaceIds, $purgeCache)); |
85
|
|
|
$ret['from_cache'] = $this->fromCache; |
86
|
|
|
return $this->render('default/result.html.twig', $ret); |
87
|
|
|
} |
88
|
|
|
|
89
|
|
|
return $this->render('default/index.html.twig', $ret); |
90
|
|
|
} |
91
|
|
|
|
92
|
|
|
/** |
93
|
|
|
* Get results based on given Request. |
94
|
|
|
* @param string $query |
95
|
|
|
* @param bool $regex |
96
|
|
|
* @param bool $ignoreCase |
97
|
|
|
* @param int[] $namespaceIds |
98
|
|
|
* @param bool $purgeCache |
99
|
|
|
* @return mixed[] |
100
|
|
|
*/ |
101
|
|
|
public function getResults( |
102
|
|
|
string $query, |
103
|
|
|
bool $regex, |
104
|
|
|
bool $ignoreCase, |
105
|
|
|
array $namespaceIds, |
106
|
|
|
bool $purgeCache = false |
107
|
|
|
): array { |
108
|
|
|
$cacheItem = md5($query.$regex.$ignoreCase.implode('|', $namespaceIds)); |
109
|
|
|
|
110
|
|
|
if (!$purgeCache && $this->cache->hasItem($cacheItem)) { |
111
|
|
|
$this->fromCache = true; |
112
|
|
|
return $this->cache->getItem($cacheItem)->get(); |
113
|
|
|
} |
114
|
|
|
|
115
|
|
|
$query = new Query($query, $namespaceIds, $regex, $ignoreCase); |
116
|
|
|
$params = $query->getParams(); |
117
|
|
|
$res = (new CloudElasticRepository($this->client, $params))->makeRequest(); |
118
|
|
|
$data = [ |
119
|
|
|
'query' => $query, |
120
|
|
|
'regex' => $regex, |
121
|
|
|
'ignore_case' => $ignoreCase, |
122
|
|
|
'total' => $res['hits']['total'], |
123
|
|
|
'hits' => $this->formatHits($res), |
124
|
|
|
]; |
125
|
|
|
|
126
|
|
|
$cacheItem = $this->cache->getItem($cacheItem) |
127
|
|
|
->set($data) |
128
|
|
|
->expiresAfter(\DateInterval::createFromDateString(self::CACHE_TIME)); |
129
|
|
|
$this->cache->save($cacheItem); |
130
|
|
|
return $data; |
131
|
|
|
} |
132
|
|
|
|
133
|
|
|
/** |
134
|
|
|
* Build the data structure for each hit, giving the view what it needs. |
135
|
|
|
* @param mixed[] $data |
136
|
|
|
* @return mixed[] |
137
|
|
|
*/ |
138
|
|
|
private function formatHits(array $data): array |
139
|
|
|
{ |
140
|
|
|
$hits = $data['hits']['hits']; |
141
|
|
|
$newData = []; |
142
|
|
|
|
143
|
|
|
foreach ($hits as $hit) { |
144
|
|
|
$result = $hit['_source']; |
145
|
|
|
$title = ($result['namespace_text'] ? $result['namespace_text'].':' : '').$result['title']; |
146
|
|
|
$domain = $this->getWikiDomainFromDbName($result['wiki']); |
147
|
|
|
$newData[] = [ |
148
|
|
|
'wiki' => rtrim($domain, '.org'), |
149
|
|
|
'title' => $title, |
150
|
|
|
'url' => $this->getUrlForTitle($domain, $title), |
151
|
|
|
'source_text' => $this->highlightQuery( |
152
|
|
|
$hit['highlight']['source_text.plain'][0] ?? '' |
153
|
|
|
), |
154
|
|
|
]; |
155
|
|
|
} |
156
|
|
|
|
157
|
|
|
return $newData; |
158
|
|
|
} |
159
|
|
|
|
160
|
|
|
/** |
161
|
|
|
* Get the URL to the page with the given title on the given wiki. |
162
|
|
|
* @param string $domain |
163
|
|
|
* @param string $title |
164
|
|
|
* @return string |
165
|
|
|
*/ |
166
|
|
|
private function getUrlForTitle(string $domain, string $title): string |
167
|
|
|
{ |
168
|
|
|
return 'https://'.$domain.'/wiki/'.$title; |
169
|
|
|
} |
170
|
|
|
|
171
|
|
|
/** |
172
|
|
|
* Query Siteinfo API to get the domain of the wiki with the given database name. |
173
|
|
|
* @param string $wiki |
174
|
|
|
* @return string |
175
|
|
|
*/ |
176
|
|
|
private function getWikiDomainFromDbName(string $wiki): string |
177
|
|
|
{ |
178
|
|
|
if (null === $this->domainLookup) { |
179
|
|
|
$this->domainLookup = (new WikiDomainRepository($this->client, $this->cache))->load(); |
180
|
|
|
} |
181
|
|
|
return $this->domainLookup[$wiki] ?? 'WIKINOTFOUND'; |
182
|
|
|
} |
183
|
|
|
|
184
|
|
|
/** |
185
|
|
|
* Make the highlight text safe and wrap the search term in a span so that we can style it. |
186
|
|
|
* @param string $text |
187
|
|
|
* @return string |
188
|
|
|
*/ |
189
|
|
|
private function highlightQuery(string $text): string |
190
|
|
|
{ |
191
|
|
|
$text = htmlspecialchars($text); |
192
|
|
|
return strtr($text, [ |
193
|
|
|
Query::PRE_TAG => "<span class='highlight'>", |
194
|
|
|
Query::POST_TAG => "</span>", |
195
|
|
|
]); |
196
|
|
|
} |
197
|
|
|
|
198
|
|
|
/** |
199
|
|
|
* Parse the namespaces parameter of the query string. |
200
|
|
|
* @param Request $request |
201
|
|
|
* @return mixed[] [normalized comma-separated list as a string, array of ids as ints] |
202
|
|
|
*/ |
203
|
|
|
private function parseNamespaces(Request $request): array |
204
|
|
|
{ |
205
|
|
|
$param = $request->query->get('namespaces', ''); |
206
|
|
|
|
207
|
|
|
if ('' === $param) { |
208
|
|
|
$ids = []; |
209
|
|
|
} else { |
210
|
|
|
$ids = array_map( |
211
|
|
|
'intval', |
212
|
|
|
explode(',', $param) |
213
|
|
|
); |
214
|
|
|
} |
215
|
|
|
|
216
|
|
|
return [ |
217
|
|
|
implode(',', $ids), |
218
|
|
|
$ids, |
219
|
|
|
]; |
220
|
|
|
} |
221
|
|
|
} |
222
|
|
|
|