|
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\JsonResponse; |
|
14
|
|
|
use Symfony\Component\HttpFoundation\Request; |
|
15
|
|
|
use Symfony\Component\HttpFoundation\Response; |
|
16
|
|
|
use Symfony\Component\Routing\Annotation\Route; |
|
17
|
|
|
|
|
18
|
|
|
/** |
|
19
|
|
|
* A DefaultController serves the main routes for the application and processing the submitted form. |
|
20
|
|
|
*/ |
|
21
|
|
|
class DefaultController extends AbstractController |
|
22
|
|
|
{ |
|
23
|
|
|
/** @var Client */ |
|
24
|
|
|
private $client; |
|
25
|
|
|
|
|
26
|
|
|
/** @var CacheItemPoolInterface */ |
|
27
|
|
|
private $cache; |
|
28
|
|
|
|
|
29
|
|
|
/** @var bool Whether the results were pulled from cache. */ |
|
30
|
|
|
private $fromCache = false; |
|
31
|
|
|
|
|
32
|
|
|
/** @var string Duration of cache for main results set, as accepted by DateInterval::createFromDateString() */ |
|
33
|
|
|
private const CACHE_TIME = '10 minutes'; |
|
34
|
|
|
|
|
35
|
|
|
/** @var string[]|null Map from wiki dbname to domain name */ |
|
36
|
|
|
private $domainLookup; |
|
37
|
|
|
|
|
38
|
|
|
/** |
|
39
|
|
|
* DefaultController constructor. |
|
40
|
|
|
* @param CacheItemPoolInterface $cache |
|
41
|
|
|
*/ |
|
42
|
|
|
public function __construct(CacheItemPoolInterface $cache) |
|
43
|
|
|
{ |
|
44
|
|
|
$this->client = new Client([ |
|
45
|
|
|
'verify' => $_ENV['ELASTIC_INSECURE'] ? false : true, |
|
46
|
|
|
]); |
|
47
|
|
|
$this->cache = $cache; |
|
48
|
|
|
} |
|
49
|
|
|
|
|
50
|
|
|
/** |
|
51
|
|
|
* Splash page, shown when user is logged out. |
|
52
|
|
|
* @Route("/splash") |
|
53
|
|
|
*/ |
|
54
|
|
|
public function splashAction(): Response |
|
55
|
|
|
{ |
|
56
|
|
|
return $this->render('jumbotron.html.twig'); |
|
57
|
|
|
} |
|
58
|
|
|
|
|
59
|
|
|
/** |
|
60
|
|
|
* The main route. |
|
61
|
|
|
* @Route("/", name="home") |
|
62
|
|
|
* @param Request $request |
|
63
|
|
|
* @return Response |
|
64
|
|
|
*/ |
|
65
|
|
|
public function indexAction(Request $request): Response |
|
66
|
|
|
{ |
|
67
|
|
|
if (!$this->get('session')->get('logged_in_user')) { |
|
68
|
|
|
return $this->render('jumbotron.html.twig'); |
|
69
|
|
|
} |
|
70
|
|
|
$query = $request->query->get('q'); |
|
71
|
|
|
$regex = (bool)$request->query->get('regex'); |
|
72
|
|
|
$ignoreCase = (bool)$request->query->get('ignorecase'); |
|
73
|
|
|
$namespaceIds = $this->parseNamespaces($request); |
|
74
|
|
|
$titlePattern = $request->query->get('title'); |
|
75
|
|
|
$purgeCache = (bool)$request->query->get('purge'); |
|
76
|
|
|
|
|
77
|
|
|
$ret = [ |
|
78
|
|
|
'q' => $query, |
|
79
|
|
|
'regex' => $regex, |
|
80
|
|
|
'max_results' => Query::MAX_RESULTS, |
|
81
|
|
|
'namespaces' => $namespaceIds, |
|
82
|
|
|
'title' => $titlePattern, |
|
83
|
|
|
'ignore_case' => $ignoreCase, |
|
84
|
|
|
]; |
|
85
|
|
|
|
|
86
|
|
|
if ($query) { |
|
87
|
|
|
$ret = array_merge($ret, $this->getResults( |
|
88
|
|
|
$query, |
|
89
|
|
|
$regex, |
|
90
|
|
|
$ignoreCase, |
|
91
|
|
|
$namespaceIds, |
|
92
|
|
|
$titlePattern, |
|
93
|
|
|
$purgeCache |
|
94
|
|
|
)); |
|
95
|
|
|
$ret['from_cache'] = $this->fromCache; |
|
96
|
|
|
return $this->formatResponse($request, $query, $ret); |
|
97
|
|
|
} |
|
98
|
|
|
|
|
99
|
|
|
return $this->render('default/index.html.twig', $ret); |
|
100
|
|
|
} |
|
101
|
|
|
|
|
102
|
|
|
/** |
|
103
|
|
|
* Get the rendered template for the requested format. |
|
104
|
|
|
* @param Request $request |
|
105
|
|
|
* @param string $query Query string, used for filenames. |
|
106
|
|
|
* @param mixed[] $data Data that should be passed to the view. |
|
107
|
|
|
* @return Response |
|
108
|
|
|
*/ |
|
109
|
|
|
private function formatResponse(Request $request, string $query, array $data): Response |
|
110
|
|
|
{ |
|
111
|
|
|
$format = $request->query->get('format', 'html'); |
|
112
|
|
|
if ('' == $format) { |
|
113
|
|
|
// The default above doesn't work when the 'format' parameter is blank. |
|
114
|
|
|
$format = 'html'; |
|
115
|
|
|
} |
|
116
|
|
|
|
|
117
|
|
|
$formatMap = [ |
|
118
|
|
|
'html' => 'text/html', |
|
119
|
|
|
'wikitext' => 'text/plain', |
|
120
|
|
|
'csv' => 'text/csv', |
|
121
|
|
|
'tsv' => 'text/tab-separated-values', |
|
122
|
|
|
'json' => 'application/json', |
|
123
|
|
|
]; |
|
124
|
|
|
|
|
125
|
|
|
// Use HTML if unknown format requested. |
|
126
|
|
|
$format = isset($formatMap[$format]) ? $format : 'html'; |
|
127
|
|
|
|
|
128
|
|
|
// Nothing more needed if requesting JSON. |
|
129
|
|
|
if ('json' === $format) { |
|
130
|
|
|
return new JsonResponse($data); |
|
131
|
|
|
} |
|
132
|
|
|
|
|
133
|
|
|
$response = new Response(); |
|
134
|
|
|
|
|
135
|
|
|
$response->headers->set('Content-Type', $formatMap[$format]); |
|
136
|
|
|
if (in_array($format, ['csv', 'tsv'])) { |
|
137
|
|
|
$filename = $this->getFilenameForRequest($query); |
|
138
|
|
|
$response->headers->set( |
|
139
|
|
|
'Content-Disposition', |
|
140
|
|
|
"attachment; filename=\"{$filename}.$format\"" |
|
141
|
|
|
); |
|
142
|
|
|
} |
|
143
|
|
|
|
|
144
|
|
|
return $this->render("default/result.$format.twig", $data, $response); |
|
145
|
|
|
} |
|
146
|
|
|
|
|
147
|
|
|
/** |
|
148
|
|
|
* Returns pretty filename from the given query, with problematic characters filtered out. |
|
149
|
|
|
* @param string $query |
|
150
|
|
|
* @return string |
|
151
|
|
|
*/ |
|
152
|
|
|
private function getFilenameForRequest(string $query): string |
|
153
|
|
|
{ |
|
154
|
|
|
$filename = trim($query, '/'); |
|
155
|
|
|
return trim(preg_replace('/[-\/\\:;*?|<>%#"]+/', '-', $filename)); |
|
156
|
|
|
} |
|
157
|
|
|
|
|
158
|
|
|
/** |
|
159
|
|
|
* Get results based on given Request. |
|
160
|
|
|
* @param string $query |
|
161
|
|
|
* @param bool $regex |
|
162
|
|
|
* @param bool $ignoreCase |
|
163
|
|
|
* @param int[] $namespaceIds |
|
164
|
|
|
* @param string|null $titlePattern |
|
165
|
|
|
* @param bool $purgeCache |
|
166
|
|
|
* @return mixed[] |
|
167
|
|
|
*/ |
|
168
|
|
|
public function getResults( |
|
169
|
|
|
string $query, |
|
170
|
|
|
bool $regex, |
|
171
|
|
|
bool $ignoreCase, |
|
172
|
|
|
array $namespaceIds, |
|
173
|
|
|
?string $titlePattern = null, |
|
174
|
|
|
bool $purgeCache = false |
|
175
|
|
|
): array { |
|
176
|
|
|
$cacheItem = md5($query.$regex.$ignoreCase.$titlePattern.implode('|', $namespaceIds)); |
|
177
|
|
|
|
|
178
|
|
|
if (!$purgeCache && $this->cache->hasItem($cacheItem)) { |
|
179
|
|
|
$this->fromCache = true; |
|
180
|
|
|
return $this->cache->getItem($cacheItem)->get(); |
|
181
|
|
|
} |
|
182
|
|
|
|
|
183
|
|
|
$query = new Query($query, $namespaceIds, $regex, $ignoreCase, $titlePattern); |
|
184
|
|
|
$params = $query->getParams(); |
|
185
|
|
|
$res = (new CloudElasticRepository($this->client, $params))->makeRequest(); |
|
186
|
|
|
$data = [ |
|
187
|
|
|
'regex' => $regex, |
|
188
|
|
|
'ignore_case' => $ignoreCase, |
|
189
|
|
|
'title' => $titlePattern, |
|
190
|
|
|
'total' => $res['hits']['total'], |
|
191
|
|
|
'hits' => $this->formatHits($res), |
|
192
|
|
|
]; |
|
193
|
|
|
|
|
194
|
|
|
$cacheItem = $this->cache->getItem($cacheItem) |
|
195
|
|
|
->set($data) |
|
196
|
|
|
->expiresAfter(\DateInterval::createFromDateString(self::CACHE_TIME)); |
|
197
|
|
|
$this->cache->save($cacheItem); |
|
198
|
|
|
return $data; |
|
199
|
|
|
} |
|
200
|
|
|
|
|
201
|
|
|
/** |
|
202
|
|
|
* Build the data structure for each hit, giving the view what it needs. |
|
203
|
|
|
* @param mixed[] $data |
|
204
|
|
|
* @return mixed[] |
|
205
|
|
|
*/ |
|
206
|
|
|
private function formatHits(array $data): array |
|
207
|
|
|
{ |
|
208
|
|
|
$hits = $data['hits']['hits']; |
|
209
|
|
|
$newData = []; |
|
210
|
|
|
|
|
211
|
|
|
foreach ($hits as $hit) { |
|
212
|
|
|
$result = $hit['_source']; |
|
213
|
|
|
$title = ($result['namespace_text'] ? $result['namespace_text'].':' : '').$result['title']; |
|
214
|
|
|
$domain = $this->getWikiDomainFromDbName($result['wiki']); |
|
215
|
|
|
$newData[] = [ |
|
216
|
|
|
'wiki' => rtrim($domain, '.org'), |
|
217
|
|
|
'title' => $title, |
|
218
|
|
|
'url' => $this->getUrlForTitle($domain, $title), |
|
219
|
|
|
'source_text' => $this->highlightQuery( |
|
220
|
|
|
$hit['highlight']['source_text.plain'][0] ?? '' |
|
221
|
|
|
), |
|
222
|
|
|
]; |
|
223
|
|
|
} |
|
224
|
|
|
|
|
225
|
|
|
return $newData; |
|
226
|
|
|
} |
|
227
|
|
|
|
|
228
|
|
|
/** |
|
229
|
|
|
* Get the URL to the page with the given title on the given wiki. |
|
230
|
|
|
* @param string $domain |
|
231
|
|
|
* @param string $title |
|
232
|
|
|
* @return string |
|
233
|
|
|
*/ |
|
234
|
|
|
private function getUrlForTitle(string $domain, string $title): string |
|
235
|
|
|
{ |
|
236
|
|
|
return 'https://'.$domain.'/wiki/'.$title; |
|
237
|
|
|
} |
|
238
|
|
|
|
|
239
|
|
|
/** |
|
240
|
|
|
* Query Siteinfo API to get the domain of the wiki with the given database name. |
|
241
|
|
|
* @param string $wiki |
|
242
|
|
|
* @return string |
|
243
|
|
|
*/ |
|
244
|
|
|
private function getWikiDomainFromDbName(string $wiki): string |
|
245
|
|
|
{ |
|
246
|
|
|
if (null === $this->domainLookup) { |
|
247
|
|
|
$this->domainLookup = (new WikiDomainRepository($this->client, $this->cache))->load(); |
|
248
|
|
|
} |
|
249
|
|
|
return $this->domainLookup[$wiki] ?? 'WIKINOTFOUND'; |
|
250
|
|
|
} |
|
251
|
|
|
|
|
252
|
|
|
/** |
|
253
|
|
|
* Make the highlight text safe and wrap the search term in a span so that we can style it. |
|
254
|
|
|
* @param string $text |
|
255
|
|
|
* @return string |
|
256
|
|
|
*/ |
|
257
|
|
|
private function highlightQuery(string $text): string |
|
258
|
|
|
{ |
|
259
|
|
|
$text = htmlspecialchars($text); |
|
260
|
|
|
return strtr($text, [ |
|
261
|
|
|
Query::PRE_TAG => "<span class='highlight'>", |
|
262
|
|
|
Query::POST_TAG => "</span>", |
|
263
|
|
|
]); |
|
264
|
|
|
} |
|
265
|
|
|
|
|
266
|
|
|
/** |
|
267
|
|
|
* Parse the namespaces parameter of the query string. |
|
268
|
|
|
* @param Request $request |
|
269
|
|
|
* @return int[] Namespace IDs. |
|
270
|
|
|
*/ |
|
271
|
|
|
private function parseNamespaces(Request $request): array |
|
272
|
|
|
{ |
|
273
|
|
|
$param = $request->query->get('namespaces', ''); |
|
274
|
|
|
|
|
275
|
|
|
if ('' === $param) { |
|
276
|
|
|
$ids = []; |
|
277
|
|
|
} else { |
|
278
|
|
|
$ids = array_map( |
|
279
|
|
|
'intval', |
|
280
|
|
|
explode(',', $param) |
|
281
|
|
|
); |
|
282
|
|
|
} |
|
283
|
|
|
|
|
284
|
|
|
return $ids; |
|
285
|
|
|
} |
|
286
|
|
|
} |
|
287
|
|
|
|