rss-bridge/lib/contents.php

<?php
/**
 * This file is part of RSS-Bridge, a PHP project capable of generating RSS and
 * Atom feeds for websites that don't have one.
 *
 * For the full license information, please view the UNLICENSE file distributed
 * with this source code.
 *
 * @package	Core
 * @license	http://unlicense.org/ UNLICENSE
 * @link	https://github.com/rss-bridge/rss-bridge
 */


/**
 * Exception class to handle all errors, when executing getContents
 */
class GetContentsException extends \Exception {
	public function __construct($details, $code = 0, Throwable $previous = null) {
		$message = trim($this->getMessageHeading() . "\n$details");

		$lastError = error_get_last();
		if($lastError !== null)
			$message .= "\nLast PHP Error: " . $lastError['message'];

		parent::__construct($message, $code, $previous);
	}

	protected function getMessageHeading() {
		return 'Could not get contents';
	}
}

/**
 * Exception class to handle HTTP responses with Cloudflare challenges
 **/
class CloudflareChallengeException extends \Exception {
	public function __construct($code = 0, Throwable $previous = null) {
		$message = <<<EOD
The server responded with a Cloudflare challenge, which is not supported by RSS-Bridge!
If this error persists longer than a week, please consider opening an issue on GitHub!
EOD;

		parent::__construct($message, $code, $previous);
	}
}

/**
 * Exception class to handle non-20x HTTP responses
 **/
class UnexpectedResponseException extends \GetContentsException {
	private $responseCode;
	private $responseHeaders;
	private $responseBody;

	protected function getMessageHeading() {
		return 'Unexpected response from upstream';
	}

	public function __construct($responseBody, $responseHeaders, $responseCode = 500, Throwable $previous = null) {
		$this->responseCode = $responseCode;
		$this->responseHeaders = $responseHeaders;
		$this->responseBody = $responseBody;

		parent::__construct('', $responseCode, $previous);
	}

	public function getResponseCode() {
		return $this->responseCode;
	}

	public function getResponseHeaders() {
		return $this->responseHeaders;
	}

	public function getResponseBody() {
		return $this->responseBody();
	}
}

/**
 * Gets contents from the Internet.
 *
 * **Content caching** (disabled in debug mode)
 *
 * A copy of the received content is stored in a local cache folder `server/` at
 * {@see PATH_CACHE}. The `If-Modified-Since` header is added to the request, if
 * the provided URL has been cached before.
 *
 * When the server responds with `304 Not Modified`, the cached data is returned.
 * This will improve response times and reduce bandwidth for servers that support
 * the `If-Modified-Since` header.
 *
 * Cached files are forcefully removed after 24 hours.
 *
 * @link https://developer.mozilla.org/en-US/docs/Web/HTTP/Headers/If-Modified-Since
 * If-Modified-Since
 *
 * @param string $url The URL.
 * @param array $header (optional) A list of cURL header.
 * For more information follow the links below.
 * * https://php.net/manual/en/function.curl-setopt.php
 * * https://curl.haxx.se/libcurl/c/CURLOPT_HTTPHEADER.html
 * @param array $opts (optional) A list of cURL options as associative array in
 * the format `$opts[$option] = $value;`, where `$option` is any `CURLOPT_XXX`
 * option and `$value` the corresponding value.
 * @param bool $returnHeader Returns an array of two elements 'header' and
 * 'content' if enabled.
 *
 * For more information see http://php.net/manual/en/function.curl-setopt.php
 * @return string|array The contents.
 */
function getContents($url, $header = array(), $opts = array(), $returnHeader = false){
	Debug::log('Reading contents from "' . $url . '"');

	// Initialize cache
	$cacheFac = new CacheFactory();
	$cacheFac->setWorkingDir(PATH_LIB_CACHES);
	$cache = $cacheFac->create(Configuration::getConfig('cache', 'type'));
	$cache->setScope('server');
	$cache->purgeCache(86400); // 24 hours (forced)

	$params = array($url);
	$cache->setKey($params);

	$retVal = array(
		'header' => '',
		'content' => '',
	);

	// Use file_get_contents() if curl module is not installed
	if(! function_exists('curl_version')) {

		$httpHeaders = '';

		foreach ($header as $headerL) {
			$httpHeaders .= $headerL . "\r\n";
		}

		$ctx = stream_context_create(array(
			'http' => array(
				'header' => $httpHeaders
			)
		));

		$data = @file_get_contents($url, 0, $ctx);

		if($data === false) {
			$errorCode = 500;
		} else {
			$errorCode = 200;
			$retVal['header'] = implode("\r\n", $http_response_header);
		}

		$curlError = '';
		$curlErrno = '';
		$headerSize = 0;
		$finalHeader = array();
	} else {
		$ch = curl_init($url);
		curl_setopt($ch, CURLOPT_RETURNTRANSFER, true);
		curl_setopt($ch, CURLOPT_FOLLOWLOCATION, true);

		if(is_array($header) && count($header) !== 0) {

			Debug::log('Setting headers: ' . json_encode($header));
			curl_setopt($ch, CURLOPT_HTTPHEADER, $header);

		}

		curl_setopt($ch, CURLOPT_USERAGENT, Configuration::getConfig('http', 'useragent'));
		curl_setopt($ch, CURLOPT_TIMEOUT, Configuration::getConfig('http', 'timeout'));
		curl_setopt($ch, CURLOPT_ENCODING, '');
		curl_setopt($ch, CURLOPT_PROTOCOLS, CURLPROTO_HTTP | CURLPROTO_HTTPS);

		if(is_array($opts) && count($opts) !== 0) {

			Debug::log('Setting options: ' . json_encode($opts));

			foreach($opts as $key => $value) {
				curl_setopt($ch, $key, $value);
			}

		}

		if(defined('PROXY_URL') && !defined('NOPROXY')) {

			Debug::log('Setting proxy url: ' . PROXY_URL);
			curl_setopt($ch, CURLOPT_PROXY, PROXY_URL);

		}

		// We always want the response header as part of the data!
		curl_setopt($ch, CURLOPT_HEADER, true);

		// Build "If-Modified-Since" header
		if(!Debug::isEnabled() && $time = $cache->getTime()) { // Skip if cache file doesn't exist
			Debug::log('Adding If-Modified-Since');
			curl_setopt($ch, CURLOPT_TIMEVALUE, $time);
			curl_setopt($ch, CURLOPT_TIMECONDITION, CURL_TIMECOND_IFMODSINCE);
		}

		// Enables logging for the outgoing header
		curl_setopt($ch, CURLINFO_HEADER_OUT, true);

		$data = curl_exec($ch);
		$errorCode = curl_getinfo($ch, CURLINFO_HTTP_CODE);

		$curlError = curl_error($ch);
		$curlErrno = curl_errno($ch);
		$curlInfo = curl_getinfo($ch);

		Debug::log('Outgoing header: ' . json_encode($curlInfo));

		if($data === false)
			Debug::log('Cant\'t download ' . $url . ' cUrl error: ' . $curlError . ' (' . $curlErrno . ')');

		$headerSize = curl_getinfo($ch, CURLINFO_HEADER_SIZE);
		$header = substr($data, 0, $headerSize);
		$retVal['header'] = $header;

		Debug::log('Response header: ' . $header);

		$headers = parseResponseHeader($header);
		$finalHeader = end($headers);

		curl_close($ch);
	}

	$finalHeader = array_change_key_case($finalHeader, CASE_LOWER);

	switch($errorCode) {
		case 200: // Contents OK
		case 201: // Contents Created
		case 202: // Contents Accepted
			Debug::log('New contents received');
			$data = substr($data, $headerSize);
			// Disable caching if the server responds with "Cache-Control: no-cache"
			// or "Cache-Control: no-store"

			if(array_key_exists('cache-control', $finalHeader)) {
				Debug::log('Server responded with "Cache-Control" header');
				$directives = explode(',', $finalHeader['cache-control']);
				$directives = array_map('trim', $directives);
				if(in_array('no-cache', $directives)
				|| in_array('no-store', $directives)) { // Skip caching
					Debug::log('Skip server side caching');
					$retVal['content'] = $data;
					break;
				}
			}
			Debug::log('Store response to cache');
			$cache->saveData($data);
			$retVal['content'] = $data;
			break;
		case 304: // Not modified, use cached data
			Debug::log('Contents not modified on host, returning cached data');
			$retVal['content'] = $cache->loadData();
			break;
		default:
			if(array_key_exists('server', $finalHeader) && stripos($finalHeader['server'], 'cloudflare') !== false) {
				throw new CloudflareChallengeException($errorCode);
			}

			if ($curlError || $curlErrno) {
				throw new GetContentsException('cURL error: ' . $curlError . ' (' . $curlErrno . ')');
			}

			throw new UnexpectedResponseException($retVal['content'], $retVal['header'], $errorCode);
	}

	return ($returnHeader === true) ? $retVal : $retVal['content'];
}

/**
 * Gets contents from the Internet as simplhtmldom object.
 *
 * @param string $url The URL.
 * @param array $header (optional) A list of cURL header.
 * For more information follow the links below.
 * * https://php.net/manual/en/function.curl-setopt.php
 * * https://curl.haxx.se/libcurl/c/CURLOPT_HTTPHEADER.html
 * @param array $opts (optional) A list of cURL options as associative array in
 * the format `$opts[$option] = $value;`, where `$option` is any `CURLOPT_XXX`
 * option and `$value` the corresponding value.
 *
 * For more information see http://php.net/manual/en/function.curl-setopt.php
 * @param bool $lowercase Force all selectors to lowercase.
 * @param bool $forceTagsClosed Forcefully close tags in malformed HTML.
 *
 * _Remarks_: Forcefully closing tags is great for malformed HTML, but it can
 * lead to parsing errors.
 * @param string $target_charset Defines the target charset.
 * @param bool $stripRN Replace all occurrences of `"\r"` and `"\n"` by `" "`.
 * @param string $defaultBRText Specifies the replacement text for `<br>` tags
 * when returning plaintext.
 * @param string $defaultSpanText Specifies the replacement text for `<span />`
 * tags when returning plaintext.
 * @return false|simple_html_dom Contents as simplehtmldom object.
 */
function getSimpleHTMLDOM($url,
	$header = array(),
	$opts = array(),
	$lowercase = true,
	$forceTagsClosed = true,
	$target_charset = DEFAULT_TARGET_CHARSET,
	$stripRN = true,
	$defaultBRText = DEFAULT_BR_TEXT,
	$defaultSpanText = DEFAULT_SPAN_TEXT){

	$content = getContents($url, $header, $opts);
	return str_get_html($content,
	$lowercase,
	$forceTagsClosed,
	$target_charset,
	$stripRN,
	$defaultBRText,
	$defaultSpanText);
}

/**
 * Gets contents from the Internet as simplhtmldom object. Contents are cached
 * and re-used for subsequent calls until the cache duration elapsed.
 *
 * _Notice_: Cached contents are forcefully removed after 24 hours (86400 seconds).
 *
 * @param string $url The URL.
 * @param int $duration Cache duration in seconds.
 * @param array $header (optional) A list of cURL header.
 * For more information follow the links below.
 * * https://php.net/manual/en/function.curl-setopt.php
 * * https://curl.haxx.se/libcurl/c/CURLOPT_HTTPHEADER.html
 * @param array $opts (optional) A list of cURL options as associative array in
 * the format `$opts[$option] = $value;`, where `$option` is any `CURLOPT_XXX`
 * option and `$value` the corresponding value.
 *
 * For more information see http://php.net/manual/en/function.curl-setopt.php
 * @param bool $lowercase Force all selectors to lowercase.
 * @param bool $forceTagsClosed Forcefully close tags in malformed HTML.
 *
 * _Remarks_: Forcefully closing tags is great for malformed HTML, but it can
 * lead to parsing errors.
 * @param string $target_charset Defines the target charset.
 * @param bool $stripRN Replace all occurrences of `"\r"` and `"\n"` by `" "`.
 * @param string $defaultBRText Specifies the replacement text for `<br>` tags
 * when returning plaintext.
 * @param string $defaultSpanText Specifies the replacement text for `<span />`
 * tags when returning plaintext.
 * @return false|simple_html_dom Contents as simplehtmldom object.
 */
function getSimpleHTMLDOMCached($url,
	$duration = 86400,
	$header = array(),
	$opts = array(),
	$lowercase = true,
	$forceTagsClosed = true,
	$target_charset = DEFAULT_TARGET_CHARSET,
	$stripRN = true,
	$defaultBRText = DEFAULT_BR_TEXT,
	$defaultSpanText = DEFAULT_SPAN_TEXT){

	Debug::log('Caching url ' . $url . ', duration ' . $duration);

	// Initialize cache
	$cacheFac = new CacheFactory();
	$cacheFac->setWorkingDir(PATH_LIB_CACHES);
	$cache = $cacheFac->create(Configuration::getConfig('cache', 'type'));
	$cache->setScope('pages');
	$cache->purgeCache(86400); // 24 hours (forced)

	$params = array($url);
	$cache->setKey($params);

	// Determine if cached file is within duration
	$time = $cache->getTime();
	if($time !== false
	&& (time() - $duration < $time)
	&& !Debug::isEnabled()) { // Contents within duration
		$content = $cache->loadData();
	} else { // Content not within duration
		$content = getContents($url, $header, $opts);
		if($content !== false) {
			$cache->saveData($content);
		}
	}

	return str_get_html($content,
	$lowercase,
	$forceTagsClosed,
	$target_charset,
	$stripRN,
	$defaultBRText,
	$defaultSpanText);
}

/**
 * Parses the cURL response header into an associative array
 *
 * Based on https://stackoverflow.com/a/18682872
 *
 * @param string $header The cURL response header.
 * @return array An associative array of response headers.
 */
function parseResponseHeader($header) {

	$headers = array();
	$requests = explode("\r\n\r\n", trim($header));

	foreach ($requests as $request) {

		$header = array();

		foreach (explode("\r\n", $request) as $i => $line) {

			if($i === 0) {
				$header['http_code'] = $line;
			} else {

				list ($key, $value) = explode(':', $line);
				$header[$key] = trim($value);

			}

		}

		$headers[] = $header;

	}

	return $headers;

}

/**
 * Determines the MIME type from a URL/Path file extension.
 *
 * _Remarks_:
 *
 * * The built-in functions `mime_content_type` and `fileinfo` require fetching
 * remote contents.
 * * A caller can hint for a MIME type by appending `#.ext` to the URL (i.e. `#.image`).
 *
 * Based on https://stackoverflow.com/a/1147952
 *
 * @param string $url The URL or path to the file.
 * @return string The MIME type of the file.
 */
function getMimeType($url) {
	static $mime = null;

	if (is_null($mime)) {
		// Default values, overriden by /etc/mime.types when present
		$mime = array(
			'jpg' => 'image/jpeg',
			'gif' => 'image/gif',
			'png' => 'image/png',
			'image' => 'image/*'
		);
		// '@' is used to mute open_basedir warning, see issue #818
		if (@is_readable('/etc/mime.types')) {
			$file = fopen('/etc/mime.types', 'r');
			while(($line = fgets($file)) !== false) {
				$line = trim(preg_replace('/#.*/', '', $line));
				if(!$line)
					continue;
				$parts = preg_split('/\s+/', $line);
				if(count($parts) == 1)
					continue;
				$type = array_shift($parts);
				foreach($parts as $part)
					$mime[$part] = $type;
			}
			fclose($file);
		}
	}

	if (strpos($url, '?') !== false) {
		$url_temp = substr($url, 0, strpos($url, '?'));
		if (strpos($url, '#') !== false) {
			$anchor = substr($url, strpos($url, '#'));
			$url_temp .= $anchor;
		}
		$url = $url_temp;
	}

	$ext = strtolower(pathinfo($url, PATHINFO_EXTENSION));
	if (!empty($mime[$ext])) {
		return $mime[$ext];
	}

	return 'application/octet-stream';
}