]>
Commit | Line | Data |
---|---|---|
1 | <?php | |
2 | ||
3 | namespace Wallabag\CoreBundle\Helper; | |
4 | ||
5 | use Graby\Graby; | |
6 | use Psr\Log\LoggerInterface; | |
7 | use Symfony\Component\HttpFoundation\File\MimeType\MimeTypeExtensionGuesser; | |
8 | use Symfony\Component\Validator\Constraints\Locale as LocaleConstraint; | |
9 | use Symfony\Component\Validator\Constraints\Url as UrlConstraint; | |
10 | use Symfony\Component\Validator\Validator\ValidatorInterface; | |
11 | use Wallabag\CoreBundle\Entity\Entry; | |
12 | use Wallabag\CoreBundle\Tools\Utils; | |
13 | ||
14 | /** | |
15 | * This kind of proxy class take care of getting the content from an url | |
16 | * and update the entry with what it found. | |
17 | */ | |
18 | class ContentProxy | |
19 | { | |
20 | protected $graby; | |
21 | protected $tagger; | |
22 | protected $validator; | |
23 | protected $logger; | |
24 | protected $mimeGuesser; | |
25 | protected $fetchingErrorMessage; | |
26 | protected $eventDispatcher; | |
27 | protected $storeArticleHeaders; | |
28 | ||
29 | public function __construct(Graby $graby, RuleBasedTagger $tagger, ValidatorInterface $validator, LoggerInterface $logger, $fetchingErrorMessage, $storeArticleHeaders = false) | |
30 | { | |
31 | $this->graby = $graby; | |
32 | $this->tagger = $tagger; | |
33 | $this->validator = $validator; | |
34 | $this->logger = $logger; | |
35 | $this->mimeGuesser = new MimeTypeExtensionGuesser(); | |
36 | $this->fetchingErrorMessage = $fetchingErrorMessage; | |
37 | $this->storeArticleHeaders = $storeArticleHeaders; | |
38 | } | |
39 | ||
40 | /** | |
41 | * Update entry using either fetched or provided content. | |
42 | * | |
43 | * @param Entry $entry Entry to update | |
44 | * @param string $url Url of the content | |
45 | * @param array $content Array with content provided for import with AT LEAST keys title, html, url to skip the fetchContent from the url | |
46 | * @param bool $disableContentUpdate Whether to skip trying to fetch content using Graby | |
47 | */ | |
48 | public function updateEntry(Entry $entry, $url, array $content = [], $disableContentUpdate = false) | |
49 | { | |
50 | if (!empty($content['html'])) { | |
51 | $content['html'] = $this->graby->cleanupHtml($content['html'], $url); | |
52 | } | |
53 | ||
54 | if ((empty($content) || false === $this->validateContent($content)) && false === $disableContentUpdate) { | |
55 | $fetchedContent = $this->graby->fetchContent($url); | |
56 | ||
57 | // when content is imported, we have information in $content | |
58 | // in case fetching content goes bad, we'll keep the imported information instead of overriding them | |
59 | if (empty($content) || $fetchedContent['html'] !== $this->fetchingErrorMessage) { | |
60 | $content = $fetchedContent; | |
61 | } | |
62 | } | |
63 | ||
64 | // be sure to keep the url in case of error | |
65 | // so we'll be able to refetch it in the future | |
66 | $content['url'] = !empty($content['url']) ? $content['url'] : $url; | |
67 | ||
68 | // In one case (at least in tests), url is empty here | |
69 | // so we set it using $url provided in the updateEntry call. | |
70 | // Not sure what are the other possible cases where this property is empty | |
71 | if (empty($entry->getUrl()) && !empty($url)) { | |
72 | $entry->setUrl($url); | |
73 | } | |
74 | ||
75 | $this->stockEntry($entry, $content); | |
76 | } | |
77 | ||
78 | /** | |
79 | * Use a Symfony validator to ensure the language is well formatted. | |
80 | * | |
81 | * @param Entry $entry | |
82 | * @param string $value Language to validate and save | |
83 | */ | |
84 | public function updateLanguage(Entry $entry, $value) | |
85 | { | |
86 | // some lang are defined as fr-FR, es-ES. | |
87 | // replacing - by _ might increase language support | |
88 | $value = str_replace('-', '_', $value); | |
89 | ||
90 | $errors = $this->validator->validate( | |
91 | $value, | |
92 | (new LocaleConstraint()) | |
93 | ); | |
94 | ||
95 | if (0 === count($errors)) { | |
96 | $entry->setLanguage($value); | |
97 | ||
98 | return; | |
99 | } | |
100 | ||
101 | $this->logger->warning('Language validation failed. ' . (string) $errors); | |
102 | } | |
103 | ||
104 | /** | |
105 | * Use a Symfony validator to ensure the preview picture is a real url. | |
106 | * | |
107 | * @param Entry $entry | |
108 | * @param string $value URL to validate and save | |
109 | */ | |
110 | public function updatePreviewPicture(Entry $entry, $value) | |
111 | { | |
112 | $errors = $this->validator->validate( | |
113 | $value, | |
114 | (new UrlConstraint()) | |
115 | ); | |
116 | ||
117 | if (0 === count($errors)) { | |
118 | $entry->setPreviewPicture($value); | |
119 | ||
120 | return; | |
121 | } | |
122 | ||
123 | $this->logger->warning('PreviewPicture validation failed. ' . (string) $errors); | |
124 | } | |
125 | ||
126 | /** | |
127 | * Update date. | |
128 | * | |
129 | * @param Entry $entry | |
130 | * @param string $value Date to validate and save | |
131 | */ | |
132 | public function updatePublishedAt(Entry $entry, $value) | |
133 | { | |
134 | $date = $value; | |
135 | ||
136 | // is it a timestamp? | |
137 | if (false !== filter_var($date, FILTER_VALIDATE_INT)) { | |
138 | $date = '@' . $date; | |
139 | } | |
140 | ||
141 | try { | |
142 | // is it already a DateTime? | |
143 | // (it's inside the try/catch in case of fail to be parse time string) | |
144 | if (!$date instanceof \DateTime) { | |
145 | $date = new \DateTime($date); | |
146 | } | |
147 | ||
148 | $entry->setPublishedAt($date); | |
149 | } catch (\Exception $e) { | |
150 | $this->logger->warning('Error while defining date', ['e' => $e, 'url' => $entry->getUrl(), 'date' => $value]); | |
151 | } | |
152 | } | |
153 | ||
154 | /** | |
155 | * Helper to extract and save host from entry url. | |
156 | * | |
157 | * @param Entry $entry | |
158 | */ | |
159 | public function setEntryDomainName(Entry $entry) | |
160 | { | |
161 | $domainName = parse_url($entry->getUrl(), PHP_URL_HOST); | |
162 | if (false !== $domainName) { | |
163 | $entry->setDomainName($domainName); | |
164 | } | |
165 | } | |
166 | ||
167 | /** | |
168 | * Helper to set a default title using: | |
169 | * - url basename, if applicable | |
170 | * - hostname. | |
171 | * | |
172 | * @param Entry $entry | |
173 | */ | |
174 | public function setDefaultEntryTitle(Entry $entry) | |
175 | { | |
176 | $url = parse_url($entry->getUrl()); | |
177 | $path = pathinfo($url['path'], PATHINFO_BASENAME); | |
178 | ||
179 | if (empty($path)) { | |
180 | $path = $url['host']; | |
181 | } | |
182 | ||
183 | $entry->setTitle($path); | |
184 | } | |
185 | ||
186 | /** | |
187 | * Stock entry with fetched or imported content. | |
188 | * Will fall back to OpenGraph data if available. | |
189 | * | |
190 | * @param Entry $entry Entry to stock | |
191 | * @param array $content Array with at least title, url & html | |
192 | */ | |
193 | private function stockEntry(Entry $entry, array $content) | |
194 | { | |
195 | // When a redirection occurs while fetching an entry | |
196 | // we move the original url in origin_url property if empty | |
197 | // and set the entry url with the final value | |
198 | if (!empty($content['url']) && $entry->getUrl() !== $content['url']) { | |
199 | if (empty($entry->getOriginUrl())) { | |
200 | $entry->setOriginUrl($entry->getUrl()); | |
201 | } | |
202 | $entry->setUrl($content['url']); | |
203 | } | |
204 | ||
205 | $this->setEntryDomainName($entry); | |
206 | ||
207 | if (!empty($content['title'])) { | |
208 | $entry->setTitle($content['title']); | |
209 | } elseif (!empty($content['open_graph']['og_title'])) { | |
210 | $entry->setTitle($content['open_graph']['og_title']); | |
211 | } | |
212 | ||
213 | $html = $content['html']; | |
214 | if (false === $html) { | |
215 | $html = $this->fetchingErrorMessage; | |
216 | ||
217 | if (!empty($content['open_graph']['og_description'])) { | |
218 | $html .= '<p><i>But we found a short description: </i></p>'; | |
219 | $html .= $content['open_graph']['og_description']; | |
220 | } | |
221 | } | |
222 | ||
223 | $entry->setContent($html); | |
224 | $entry->setReadingTime(Utils::getReadingTime($html)); | |
225 | ||
226 | if (!empty($content['status'])) { | |
227 | $entry->setHttpStatus($content['status']); | |
228 | } | |
229 | ||
230 | if (!empty($content['authors']) && is_array($content['authors'])) { | |
231 | $entry->setPublishedBy($content['authors']); | |
232 | } | |
233 | ||
234 | if (!empty($content['all_headers']) && $this->storeArticleHeaders) { | |
235 | $entry->setHeaders($content['all_headers']); | |
236 | } | |
237 | ||
238 | if (!empty($content['date'])) { | |
239 | $this->updatePublishedAt($entry, $content['date']); | |
240 | } | |
241 | ||
242 | if (!empty($content['language'])) { | |
243 | $this->updateLanguage($entry, $content['language']); | |
244 | } | |
245 | ||
246 | if (!empty($content['open_graph']['og_image'])) { | |
247 | $this->updatePreviewPicture($entry, $content['open_graph']['og_image']); | |
248 | } | |
249 | ||
250 | // if content is an image, define it as a preview too | |
251 | if (!empty($content['content_type']) && in_array($this->mimeGuesser->guess($content['content_type']), ['jpeg', 'jpg', 'gif', 'png'], true)) { | |
252 | $this->updatePreviewPicture($entry, $content['url']); | |
253 | } | |
254 | ||
255 | if (!empty($content['content_type'])) { | |
256 | $entry->setMimetype($content['content_type']); | |
257 | } | |
258 | ||
259 | try { | |
260 | $this->tagger->tag($entry); | |
261 | } catch (\Exception $e) { | |
262 | $this->logger->error('Error while trying to automatically tag an entry.', [ | |
263 | 'entry_url' => $content['url'], | |
264 | 'error_msg' => $e->getMessage(), | |
265 | ]); | |
266 | } | |
267 | } | |
268 | ||
269 | /** | |
270 | * Validate that the given content has at least a title, an html and a url. | |
271 | * | |
272 | * @param array $content | |
273 | * | |
274 | * @return bool true if valid otherwise false | |
275 | */ | |
276 | private function validateContent(array $content) | |
277 | { | |
278 | return !empty($content['title']) && !empty($content['html']) && !empty($content['url']); | |
279 | } | |
280 | } |