rss-bridge/lib/contents.php

Ignoring revisions in .git-blame-ignore-revs. Click here to bypass and see the normal blame view.

240 lines
8.9 KiB
PHP
Raw Normal View History

<?php
2022-05-11 23:34:18 +03:00
/**
* Fetch data from an http url
*
* @param array $httpHeaders E.g. ['Content-type: text/plain']
* @param array $curlOptions Associative array e.g. [CURLOPT_MAXREDIRS => 3]
* @param bool $returnFull Whether to return an array: ['code' => int, 'headers' => array, 'content' => string]
2022-05-11 23:34:18 +03:00
* @return string|array
*/
function getContents(
string $url,
array $httpHeaders = [],
array $curlOptions = [],
2022-05-18 00:46:37 +03:00
bool $returnFull = false
) {
$httpClient = RssBridge::getHttpClient();
$cache = RssBridge::getCache();
2023-09-20 03:45:48 +03:00
$httpHeadersNormalized = [];
foreach ($httpHeaders as $httpHeader) {
$parts = explode(':', $httpHeader);
$headerName = trim($parts[0]);
$headerValue = trim(implode(':', array_slice($parts, 1)));
$httpHeadersNormalized[$headerName] = $headerValue;
}
// Snagged from https://github.com/lwthiker/curl-impersonate/blob/main/firefox/curl_ff102
$defaultHttpHeaders = [
'Accept' => 'text/html,application/xhtml+xml,application/xml;q=0.9,image/avif,image/webp,*/*;q=0.8',
'Accept-Language' => 'en-US,en;q=0.5',
'Upgrade-Insecure-Requests' => '1',
'Sec-Fetch-Dest' => 'document',
'Sec-Fetch-Mode' => 'navigate',
'Sec-Fetch-Site' => 'none',
'Sec-Fetch-User' => '?1',
2023-04-21 03:50:48 +03:00
'TE' => 'trailers',
];
$config = [
'useragent' => Configuration::getConfig('http', 'useragent'),
'timeout' => Configuration::getConfig('http', 'timeout'),
'headers' => array_merge($defaultHttpHeaders, $httpHeadersNormalized),
'curl_options' => $curlOptions,
];
$maxFileSize = Configuration::getConfig('http', 'max_filesize');
if ($maxFileSize) {
2023-09-20 03:45:48 +03:00
// Convert from MB to B by multiplying with 2^20 (1M)
$config['max_filesize'] = $maxFileSize * 2 ** 20;
}
if (Configuration::getConfig('proxy', 'url') && !defined('NOPROXY')) {
$config['proxy'] = Configuration::getConfig('proxy', 'url');
}
$requestBodyHash = null;
if (isset($curlOptions[CURLOPT_POSTFIELDS])) {
$requestBodyHash = md5(Json::encode($curlOptions[CURLOPT_POSTFIELDS], false));
}
$cacheKey = implode('_', ['server', $url, $requestBodyHash]);
/** @var Response $cachedResponse */
$cachedResponse = $cache->get($cacheKey);
if ($cachedResponse) {
$cachedLastModified = $cachedResponse->getHeader('last-modified');
if ($cachedLastModified) {
try {
// Some servers send Unix timestamp instead of RFC7231 date. Prepend it with @ to allow parsing as DateTime
$cachedLastModified = new \DateTimeImmutable((is_numeric($cachedLastModified) ? '@' : '') . $cachedLastModified);
$config['if_not_modified_since'] = $cachedLastModified->getTimestamp();
} catch (Exception $dateTimeParseFailue) {
// Ignore invalid 'Last-Modified' HTTP header value
}
}
}
$response = $httpClient->request($url, $config);
switch ($response->getCode()) {
case 200:
case 201:
case 202:
$cacheControl = $response->getHeader('cache-control');
if ($cacheControl) {
$directives = explode(',', $cacheControl);
$directives = array_map('trim', $directives);
if (in_array('no-cache', $directives) || in_array('no-store', $directives)) {
2022-05-11 23:34:18 +03:00
// Don't cache as instructed by the server
break;
}
}
$cache->set($cacheKey, $response, 86400 * 10);
break;
case 301:
case 302:
case 303:
// todo: cache
break;
case 304:
// Not Modified
$response = $response->withBody($cachedResponse->getBody());
break;
default:
$exceptionMessage = sprintf(
'%s resulted in %s %s %s',
$url,
$response->getCode(),
$response->getStatusLine(),
// If debug, include a part of the response body in the exception message
Debug::isEnabled() ? mb_substr($response->getBody(), 0, 500) : '',
);
2023-09-20 03:45:48 +03:00
if (CloudFlareException::isCloudFlareResponse($response)) {
throw new CloudFlareException($exceptionMessage, $response->getCode());
}
throw new HttpException(trim($exceptionMessage), $response->getCode());
}
2022-05-18 00:46:37 +03:00
if ($returnFull === true) {
2023-09-20 03:45:48 +03:00
// todo: return the actual response object
return [
'code' => $response->getCode(),
'headers' => $response->getHeaders(),
// For legacy reasons, use 'content' instead of 'body'
'content' => $response->getBody(),
];
}
return $response->getBody();
}
2018-11-16 23:48:59 +03:00
/**
* 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.
2018-11-16 23:48:59 +03:00
*/
2017-02-14 19:28:07 +03:00
function getSimpleHTMLDOM(
$url,
$header = [],
$opts = [],
$lowercase = true,
$forceTagsClosed = true,
$target_charset = DEFAULT_TARGET_CHARSET,
$stripRN = true,
$defaultBRText = DEFAULT_BR_TEXT,
$defaultSpanText = DEFAULT_SPAN_TEXT
) {
2023-10-12 20:49:04 +03:00
$html = getContents($url, $header ?? [], $opts ?? []);
if ($html === '') {
throw new \Exception('Unable to parse dom because the http response was the empty string');
}
2017-02-14 19:28:07 +03:00
return str_get_html(
2023-10-12 20:49:04 +03:00
$html,
2017-02-14 19:28:07 +03:00
$lowercase,
$forceTagsClosed,
$target_charset,
$stripRN,
$defaultBRText,
$defaultSpanText
);
}
/**
2018-11-16 23:48:59 +03:00
* 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 $ttl Cache duration in seconds.
2018-11-16 23:48:59 +03:00
* @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.
*/
2017-02-14 19:28:07 +03:00
function getSimpleHTMLDOMCached(
$url,
$ttl = 86400,
$header = [],
$opts = [],
$lowercase = true,
$forceTagsClosed = true,
$target_charset = DEFAULT_TARGET_CHARSET,
$stripRN = true,
$defaultBRText = DEFAULT_BR_TEXT,
$defaultSpanText = DEFAULT_SPAN_TEXT
) {
$cache = RssBridge::getCache();
$cacheKey = 'pages_' . $url;
$content = $cache->get($cacheKey);
if (!$content) {
$content = getContents($url, $header ?? [], $opts ?? []);
$cache->set($cacheKey, $content, $ttl);
}
2017-02-14 19:28:07 +03:00
return str_get_html(
$content,
$lowercase,
$forceTagsClosed,
$target_charset,
$stripRN,
$defaultBRText,
$defaultSpanText
);
}