content = Filter::stripXmlTag($content); // Encode everything in UTF-8 if ($xml_encoding == 'windows-1251' || $http_encoding == 'windows-1251') { $this->content = Encoding::cp1251ToUtf8($this->content); } else { $this->content = Encoding::toUTF8($this->content); } // Workarounds $this->content = $this->normalizeData($this->content); } public function filterHtml($item_content, $item_url) { $content = ''; if ($this->grabber && ! in_array($item_url, $this->grabber_ignore_urls)) { $grabber = new Grabber($item_url); $grabber->download($this->grabber_timeout, $this->grabber_user_agent); if ($grabber->parse()) $item_content = $grabber->content; } if ($item_content) { $filter = new Filter($item_content, $item_url); $content = $filter->execute(); } return $content; } public function getXmlErrors() { $errors = array(); foreach(\libxml_get_errors() as $error) { $errors[] = sprintf('XML error: %s (Line: %d - Column: %d - Code: %d)', $error->message, $error->line, $error->column, $error->code ); } return implode(', ', $errors); } // Dirty quickfix before XML parsing public function normalizeData($data) { $data = str_replace("\xc3\x20", '', $data); $data = $this->replaceEntityAttribute($data); return $data; } // For each href attribute, replace & by & // Useful for broken XML feeds public function replaceEntityAttribute($content) { $content = preg_replace_callback('/href="[^"]+"/', function(array $matches) { return htmlspecialchars($matches[0], ENT_NOQUOTES, 'UTF-8', false); }, $content); return $content; } // Trim whitespace from the begining, the end and inside a string and don't break utf-8 string public function stripWhiteSpace($value) { $value = str_replace("\r", "", $value); $value = str_replace("\t", "", $value); $value = str_replace("\n", "", $value); return trim($value); } public function generateId() { // crc32b seems to be faster and shorter than other hash algorithms return hash('crc32b', implode(func_get_args())); } public function parseDate($value) { // Format => truncate to this length if not null $formats = array( DATE_ATOM => null, DATE_RSS => null, DATE_COOKIE => null, DATE_ISO8601 => null, DATE_RFC822 => null, DATE_RFC850 => null, DATE_RFC1036 => null, DATE_RFC1123 => null, DATE_RFC2822 => null, DATE_RFC3339 => null, 'D, d M Y H:i:s' => 25, 'D, d M Y h:i:s' => 25, 'D M d Y H:i:s' => 24, 'Y-m-d H:i:s' => 19, 'Y-m-d\TH:i:s' => 19, 'd/m/Y H:i:s' => 19, 'D, d M Y' => 16, 'Y-m-d' => 10, 'd-m-Y' => 10, 'm-d-Y' => 10, 'd.m.Y' => 10, 'm.d.Y' => 10, 'd/m/Y' => 10, 'm/d/Y' => 10, ); $value = trim($value); foreach ($formats as $format => $length) { $truncated_value = $value; if ($length !== null) { $truncated_value = substr($truncated_value, 0, $length); } $timestamp = $this->getValidDate($format, $truncated_value); if ($timestamp > 0) { return $timestamp; } } return time(); } public function getValidDate($format, $value) { $date = \DateTime::createFromFormat($format, $value); if ($date !== false) { $errors = \DateTime::getLastErrors(); if ($errors['error_count'] === 0 && $errors['warning_count'] === 0) return $date->getTimestamp(); } return 0; } // Hardcoded list of hostname/token to exclude from id generation public function isExcludedFromId($url) { $exclude_list = array('ap.org', 'jacksonville.com'); foreach ($exclude_list as $token) { if (strpos($url, $token) !== false) return true; } return false; } }