*/
class Tweeper {
- private static $userAgent = "Mozilla/5.0";
+ private static $userAgent = "APIs-Google (+https://developers.google.com/webmasters/APIs-Google.html)";
private static $maxConnectionTimeout = 5;
private static $maxConnectionRetries = 5;
}
/**
- * Convert Twitter mobile date to the date format expected in a RSS document.
- */
- public static function twitterToRssDate($date) {
- // Twitter uses relative timestamps in minutes for recent tweets.
- if (preg_match('/^(\d+)m$/', $date, $matches)) {
- $timestamp = strtotime("+" . $matches[1] . " min", time());
- if (FALSE === $timestamp) {
- $timestamp = 0;
- }
- }
- else {
- /*
- * In case the time is specified put it after the date,
- * to make it recognized by strptime().
- */
- if (preg_match('/(.*) - (.*)/', $date, $matches)) {
- $date = $matches[2] . " " . $matches[1];
- }
-
- $timestamp = strtotime($date);
- if (FALSE === $timestamp) {
- $timestamp = 0;
- }
-
- /*
- * The twitter mobile UI usually only specifies the month and the day, so
- * strtotime($date) may interpret the date as being in the future.
- *
- * If the date is in the future it is probably in the same day but in the
- * previous year.
- */
- if ($timestamp > time()) {
- $timestamp = strtotime('-1 years', $timestamp);
- }
- }
-
- return Tweeper::epochToRssDate($timestamp);
- }
-
- /**
* Convert string to UpperCamelCase.
*/
public static function toUpperCamelCase($str, $delim = ' ') {
}
} while (curl_errno($ch) == CURLE_OPERATION_TIMEDOUT && ++$attempt < Tweeper::$maxConnectionRetries);
+ $response_code = curl_getinfo($ch, CURLINFO_RESPONSE_CODE);
+ if (FALSE === $response_code) {
+ trigger_error(curl_error($ch), E_USER_WARNING);
+ return FALSE;
+ }
+
+ if ($response_code >= 400) {
+ trigger_error("HTTP reponse code $response_code", E_USER_WARNING);
+ return FALSE;
+ }
+
return $ret;
}
}
/**
+ * Generate a data URL.
+ */
+ public static function generateDataURL($url) {
+ $supported_content_types = [
+ "image/gif",
+ "image/jpeg",
+ "image/png",
+ ];
+
+ $url_info = Tweeper::getUrlInfo($url);
+ if (FALSE === $url_info) {
+ trigger_error("Failed to retrieve info for URL: " . $url, E_USER_WARNING);
+ return '';
+ }
+
+ $supported = in_array($url_info['content_type'], $supported_content_types);
+ if (!$supported) {
+ trigger_error("Unsupported data URL type \"" . $url_info['content_type'] . "\" for URL: " . $url_info['url'], E_USER_WARNING);
+ return '';
+ }
+
+ $base64Data = base64_encode(file_get_contents($url));
+ $dataURL = 'data: ' . $url_info['content_type'] . ';base64,' . $base64Data;
+
+ return $dataURL;
+ }
+
+ /**
* Mimic the message from libxml.c::php_libxml_ctx_error_level()
*/
private static function logXmlError($error) {
$data = json_decode($matches[1], $assoc = TRUE);
- // The "qe" object contains elements which will result in invalid XML
- // element names, so remove it.
+ // Remove items that may contain elements which can result in invalid XML
+ // element names (for example names starting with a number).
unset($data["qe"]);
-
- // The "knobs" object contains elements with undefined namespaces, so
- // remove it to silence an error message.
- unset($data["knobs"]);
+ unset($data["knobx"]);
+ unset($data["to_cache"]);
+
+ // Stop here in case Instagram redirected to the login page, this can
+ // happen when too many consecutive requests have been made from the same
+ // IP.
+ if (array_key_exists("LoginAndSignupPage", $data["entry_data"])) {
+ trigger_error("Cannot open instagram page: redirected to Login page.\n", E_USER_WARNING);
+ return NULL;
+ }
$json = json_encode($data);
$get_xml_host_method = 'getXml' . Tweeper::toUpperCamelCase($host, '.');
if (method_exists($this, $get_xml_host_method)) {
$xml_data = call_user_func_array([$this, $get_xml_host_method], [$html]);
+ if (NULL === $xml_data) {
+ return NULL;
+ }
$xmlDoc->loadXML($xml_data);
}
else {
return NULL;
}
- // Override User-Agent for twitter.com to force it to serve the mobile UI.
- if ($host == "twitter.com") {
- $user_agent = "Mozilla/5.0 (Linux; U; Android 4.3; en-us; SM-N900T Build/JSS15J)";
- }
- else {
- $user_agent = NULL;
- }
-
- $html = Tweeper::getUrlContents($src_url, $user_agent);
+ $html = Tweeper::getUrlContents($src_url);
if (FALSE === $html) {
trigger_error("Failed to retrieve $src_url", E_USER_WARNING);
return NULL;