]> git.immae.eu Git - github/wallabag/wallabag.git/blame_incremental - src/Wallabag/CoreBundle/Helper/ContentProxy.php
Update with new strings
[github/wallabag/wallabag.git] / src / Wallabag / CoreBundle / Helper / ContentProxy.php
... / ...
CommitLineData
1<?php
2
3namespace Wallabag\CoreBundle\Helper;
4
5use Graby\Graby;
6use Psr\Log\LoggerInterface;
7use Symfony\Component\HttpFoundation\File\MimeType\MimeTypeExtensionGuesser;
8use Symfony\Component\Validator\Constraints\Locale as LocaleConstraint;
9use Symfony\Component\Validator\Constraints\Url as UrlConstraint;
10use Symfony\Component\Validator\Validator\ValidatorInterface;
11use Wallabag\CoreBundle\Entity\Entry;
12use 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 */
18class ContentProxy
19{
20 protected $graby;
21 protected $tagger;
22 protected $validator;
23 protected $logger;
24 protected $mimeGuesser;
25 protected $fetchingErrorMessage;
26 protected $eventDispatcher;
27
28 public function __construct(Graby $graby, RuleBasedTagger $tagger, ValidatorInterface $validator, LoggerInterface $logger, $fetchingErrorMessage)
29 {
30 $this->graby = $graby;
31 $this->tagger = $tagger;
32 $this->validator = $validator;
33 $this->logger = $logger;
34 $this->mimeGuesser = new MimeTypeExtensionGuesser();
35 $this->fetchingErrorMessage = $fetchingErrorMessage;
36 }
37
38 /**
39 * Update entry using either fetched or provided content.
40 *
41 * @param Entry $entry Entry to update
42 * @param string $url Url of the content
43 * @param array $content Array with content provided for import with AT LEAST keys title, html, url to skip the fetchContent from the url
44 * @param bool $disableContentUpdate Whether to skip trying to fetch content using Graby
45 */
46 public function updateEntry(Entry $entry, $url, array $content = [], $disableContentUpdate = false)
47 {
48 if (!empty($content['html'])) {
49 $content['html'] = $this->graby->cleanupHtml($content['html'], $url);
50 }
51
52 if ((empty($content) || false === $this->validateContent($content)) && false === $disableContentUpdate) {
53 $fetchedContent = $this->graby->fetchContent($url);
54
55 // when content is imported, we have information in $content
56 // in case fetching content goes bad, we'll keep the imported information instead of overriding them
57 if (empty($content) || $fetchedContent['html'] !== $this->fetchingErrorMessage) {
58 $content = $fetchedContent;
59 }
60 }
61
62 // be sure to keep the url in case of error
63 // so we'll be able to refetch it in the future
64 $content['url'] = !empty($content['url']) ? $content['url'] : $url;
65
66 $this->stockEntry($entry, $content);
67 }
68
69 /**
70 * Use a Symfony validator to ensure the language is well formatted.
71 *
72 * @param Entry $entry
73 * @param string $value Language to validate and save
74 */
75 public function updateLanguage(Entry $entry, $value)
76 {
77 // some lang are defined as fr-FR, es-ES.
78 // replacing - by _ might increase language support
79 $value = str_replace('-', '_', $value);
80
81 $errors = $this->validator->validate(
82 $value,
83 (new LocaleConstraint())
84 );
85
86 if (0 === count($errors)) {
87 $entry->setLanguage($value);
88
89 return;
90 }
91
92 $this->logger->warning('Language validation failed. ' . (string) $errors);
93 }
94
95 /**
96 * Use a Symfony validator to ensure the preview picture is a real url.
97 *
98 * @param Entry $entry
99 * @param string $value URL to validate and save
100 */
101 public function updatePreviewPicture(Entry $entry, $value)
102 {
103 $errors = $this->validator->validate(
104 $value,
105 (new UrlConstraint())
106 );
107
108 if (0 === count($errors)) {
109 $entry->setPreviewPicture($value);
110
111 return;
112 }
113
114 $this->logger->warning('PreviewPicture validation failed. ' . (string) $errors);
115 }
116
117 /**
118 * Update date.
119 *
120 * @param Entry $entry
121 * @param string $value Date to validate and save
122 */
123 public function updatePublishedAt(Entry $entry, $value)
124 {
125 $date = $value;
126
127 // is it a timestamp?
128 if (filter_var($date, FILTER_VALIDATE_INT) !== false) {
129 $date = '@' . $date;
130 }
131
132 try {
133 // is it already a DateTime?
134 // (it's inside the try/catch in case of fail to be parse time string)
135 if (!$date instanceof \DateTime) {
136 $date = new \DateTime($date);
137 }
138
139 $entry->setPublishedAt($date);
140 } catch (\Exception $e) {
141 $this->logger->warning('Error while defining date', ['e' => $e, 'url' => $entry->getUrl(), 'date' => $value]);
142 }
143 }
144
145 /**
146 * Stock entry with fetched or imported content.
147 * Will fall back to OpenGraph data if available.
148 *
149 * @param Entry $entry Entry to stock
150 * @param array $content Array with at least title, url & html
151 */
152 private function stockEntry(Entry $entry, array $content)
153 {
154 $entry->setUrl($content['url']);
155
156 $domainName = parse_url($entry->getUrl(), PHP_URL_HOST);
157 if (false !== $domainName) {
158 $entry->setDomainName($domainName);
159 }
160
161 if (!empty($content['title'])) {
162 $entry->setTitle($content['title']);
163 } elseif (!empty($content['open_graph']['og_title'])) {
164 $entry->setTitle($content['open_graph']['og_title']);
165 }
166
167 $html = $content['html'];
168 if (false === $html) {
169 $html = $this->fetchingErrorMessage;
170
171 if (!empty($content['open_graph']['og_description'])) {
172 $html .= '<p><i>But we found a short description: </i></p>';
173 $html .= $content['open_graph']['og_description'];
174 }
175 }
176
177 $entry->setContent($html);
178 $entry->setReadingTime(Utils::getReadingTime($html));
179
180 if (!empty($content['status'])) {
181 $entry->setHttpStatus($content['status']);
182 }
183
184 if (!empty($content['authors']) && is_array($content['authors'])) {
185 $entry->setPublishedBy($content['authors']);
186 }
187
188 if (!empty($content['all_headers'])) {
189 $entry->setHeaders($content['all_headers']);
190 }
191
192 if (!empty($content['date'])) {
193 $this->updatePublishedAt($entry, $content['date']);
194 }
195
196 if (!empty($content['language'])) {
197 $this->updateLanguage($entry, $content['language']);
198 }
199
200 if (!empty($content['open_graph']['og_image'])) {
201 $this->updatePreviewPicture($entry, $content['open_graph']['og_image']);
202 }
203
204 // if content is an image, define it as a preview too
205 if (!empty($content['content_type']) && in_array($this->mimeGuesser->guess($content['content_type']), ['jpeg', 'jpg', 'gif', 'png'], true)) {
206 $this->updatePreviewPicture($entry, $content['url']);
207 }
208
209 if (!empty($content['content_type'])) {
210 $entry->setMimetype($content['content_type']);
211 }
212
213 try {
214 $this->tagger->tag($entry);
215 } catch (\Exception $e) {
216 $this->logger->error('Error while trying to automatically tag an entry.', [
217 'entry_url' => $content['url'],
218 'error_msg' => $e->getMessage(),
219 ]);
220 }
221 }
222
223 /**
224 * Validate that the given content has at least a title, an html and a url.
225 *
226 * @param array $content
227 *
228 * @return bool true if valid otherwise false
229 */
230 private function validateContent(array $content)
231 {
232 return !empty($content['title']) && !empty($content['html']) && !empty($content['url']);
233 }
234}