protected $mimeGuesser;
protected $fetchingErrorMessage;
protected $eventDispatcher;
+ protected $storeArticleHeaders;
- public function __construct(Graby $graby, RuleBasedTagger $tagger, ValidatorInterface $validator, LoggerInterface $logger, $fetchingErrorMessage)
+ public function __construct(Graby $graby, RuleBasedTagger $tagger, ValidatorInterface $validator, LoggerInterface $logger, $fetchingErrorMessage, $storeArticleHeaders = false)
{
$this->graby = $graby;
$this->tagger = $tagger;
$this->logger = $logger;
$this->mimeGuesser = new MimeTypeExtensionGuesser();
$this->fetchingErrorMessage = $fetchingErrorMessage;
+ $this->storeArticleHeaders = $storeArticleHeaders;
}
/**
if ((empty($content) || false === $this->validateContent($content)) && false === $disableContentUpdate) {
$fetchedContent = $this->graby->fetchContent($url);
+ $fetchedContent['title'] = $this->sanitizeUTF8Text($fetchedContent['title']);
// when content is imported, we have information in $content
// in case fetching content goes bad, we'll keep the imported information instead of overriding them
$this->stockEntry($entry, $content);
}
+ /**
+ * Remove invalid UTF-8 characters from the given string in following steps:
+ * - try to interpret the given string as ISO-8859-1, convert it to UTF-8 and return it (if its valid)
+ * - simply remove every invalid UTF-8 character and return the result (https://stackoverflow.com/a/1433665)
+ * @param String $rawText
+ * @return string
+ */
+ private function sanitizeUTF8Text(String $rawText) {
+ if (mb_check_encoding($rawText, 'utf-8')) {
+ return $rawText; // return because its valid utf-8 text
+ }
+
+ // we assume that $text is encoded in ISO-8859-1 (and not the similar Windows-1252 or other encoding)
+ $convertedText = utf8_encode($rawText);
+ if (mb_check_encoding($convertedText, 'utf-8')) {
+ return $convertedText;
+ }
+
+ // last resort: simply remove invalid UTF-8 character because $rawText can have some every exotic encoding
+ return iconv("UTF-8", "UTF-8//IGNORE", $rawText);
+ }
+
/**
* Use a Symfony validator to ensure the language is well formatted.
*
(new LocaleConstraint())
);
- if (0 === count($errors)) {
+ if (0 === \count($errors)) {
$entry->setLanguage($value);
return;
(new UrlConstraint())
);
- if (0 === count($errors)) {
+ if (0 === \count($errors)) {
$entry->setPreviewPicture($value);
return;
$date = $value;
// is it a timestamp?
- if (filter_var($date, FILTER_VALIDATE_INT) !== false) {
- $date = '@' . $value;
+ if (false !== filter_var($date, FILTER_VALIDATE_INT)) {
+ $date = '@' . $date;
}
try {
- $entry->setPublishedAt(new \DateTime($date));
+ // is it already a DateTime?
+ // (it's inside the try/catch in case of fail to be parse time string)
+ if (!$date instanceof \DateTime) {
+ $date = new \DateTime($date);
+ }
+
+ $entry->setPublishedAt($date);
} catch (\Exception $e) {
$this->logger->warning('Error while defining date', ['e' => $e, 'url' => $entry->getUrl(), 'date' => $value]);
}
}
+ /**
+ * Helper to extract and save host from entry url.
+ *
+ * @param Entry $entry
+ */
+ public function setEntryDomainName(Entry $entry)
+ {
+ $domainName = parse_url($entry->getUrl(), PHP_URL_HOST);
+ if (false !== $domainName) {
+ $entry->setDomainName($domainName);
+ }
+ }
+
+ /**
+ * Helper to set a default title using:
+ * - url basename, if applicable
+ * - hostname.
+ *
+ * @param Entry $entry
+ */
+ public function setDefaultEntryTitle(Entry $entry)
+ {
+ $url = parse_url($entry->getUrl());
+ $path = pathinfo($url['path'], PATHINFO_BASENAME);
+
+ if (empty($path)) {
+ $path = $url['host'];
+ }
+
+ $entry->setTitle($path);
+ }
+
/**
* Stock entry with fetched or imported content.
* Will fall back to OpenGraph data if available.
{
$entry->setUrl($content['url']);
- $domainName = parse_url($entry->getUrl(), PHP_URL_HOST);
- if (false !== $domainName) {
- $entry->setDomainName($domainName);
- }
+ $this->setEntryDomainName($entry);
if (!empty($content['title'])) {
$entry->setTitle($content['title']);
$entry->setHttpStatus($content['status']);
}
- if (!empty($content['authors']) && is_array($content['authors'])) {
+ if (!empty($content['authors']) && \is_array($content['authors'])) {
$entry->setPublishedBy($content['authors']);
}
- if (!empty($content['all_headers'])) {
+ if (!empty($content['all_headers']) && $this->storeArticleHeaders) {
$entry->setHeaders($content['all_headers']);
}
}
// if content is an image, define it as a preview too
- if (!empty($content['content_type']) && in_array($this->mimeGuesser->guess($content['content_type']), ['jpeg', 'jpg', 'gif', 'png'], true)) {
+ if (!empty($content['content_type']) && \in_array($this->mimeGuesser->guess($content['content_type']), ['jpeg', 'jpg', 'gif', 'png'], true)) {
$this->updatePreviewPicture($entry, $content['url']);
}