feat: add etag support to getContents (#3893)

README.md

@@ -163,7 +163,7 @@ PHP ini config:
 ```ini
 ; /etc/php/8.2/fpm/conf.d/30-rss-bridge.ini
 
-max_execution_time = 20
+max_execution_time = 15
 memory_limit = 64M
 ```
 

config.default.ini.php

@@ -48,7 +48,7 @@ enable_maintenance_mode = false
 
 [http]
 ; Operation timeout in seconds
-timeout = 30
+timeout = 15
 
 ; Operation retry count in case of curl error
 retries = 2

lib/BridgeCard.php

@@ -16,7 +16,7 @@ final class BridgeCard
 
         $bridge = $bridgeFactory->create($bridgeClassName);
 
-        $isHttps = strpos($bridge->getURI(), 'https') === 0;
+        $isHttps = str_starts_with($bridge->getURI(), 'https');
 
         $uri = $bridge->getURI();
         $name = $bridge->getName();
@@ -113,8 +113,7 @@ EOD;
         }
 
         if (!$isHttps) {
-            $form .= '<div class="secure-warning">Warning :
+            $form .= '<div class="secure-warning">Warning: This bridge is not fetching its content through a secure connection</div>';
-This bridge is not fetching its content through a secure connection</div>';
         }
 
         return $form;

lib/FeedExpander.php

@@ -41,7 +41,7 @@ abstract class FeedExpander extends BridgeAbstract
     }
 
     /**
-     * This method is overidden by bridges
+     * This method is overridden by bridges
      *
      * @return array
      */

lib/FeedParser.php

@@ -7,9 +7,9 @@ declare(strict_types=1);
  *
  * Scrapes out rss 0.91, 1.0, 2.0 and atom 1.0.
  *
- * Produce arrays meant to be used inside rss-bridge.
+ * Produces array meant to be used inside rss-bridge.
  *
- * The item structure is tweaked so that works with FeedItem
+ * The item structure is tweaked so that it works with FeedItem
  */
 final class FeedParser
 {

lib/XPathAbstract.php

@@ -518,7 +518,10 @@ abstract class XPathAbstract extends BridgeAbstract
         if (strlen($value) === 0) {
             return '';
         }
-        if (strpos($value, 'http://') === 0 || strpos($value, 'https://') === 0) {
+        if (
+            strpos($value, 'http://') === 0
+            || strpos($value, 'https://') === 0
+        ) {
             return $value;
         }
 

lib/contents.php

@@ -24,6 +24,32 @@ function getContents(
         $headerValue = trim(implode(':', array_slice($parts, 1)));
         $httpHeadersNormalized[$headerName] = $headerValue;
     }
+
+    $requestBodyHash = null;
+    if (isset($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) {
+        $lastModified = $cachedResponse->getHeader('last-modified');
+        if ($lastModified) {
+            try {
+                // Some servers send Unix timestamp instead of RFC7231 date. Prepend it with @ to allow parsing as DateTime
+                $lastModified = new \DateTimeImmutable((is_numeric($lastModified) ? '@' : '') . $lastModified);
+                $config['if_not_modified_since'] = $lastModified->getTimestamp();
+            } catch (Exception $e) {
+                // Failed to parse last-modified
+            }
+        }
+        $etag = $cachedResponse->getHeader('etag');
+        if ($etag) {
+            $httpHeadersNormalized['if-none-match'] = $etag;
+        }
+    }
+
     // 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',
@@ -35,6 +61,7 @@ function getContents(
         'Sec-Fetch-User' => '?1',
         'TE' => 'trailers',
     ];
+
     $config = [
         'useragent' => Configuration::getConfig('http', 'useragent'),
         'timeout' => Configuration::getConfig('http', 'timeout'),
@@ -53,28 +80,6 @@ function getContents(
         $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
-            }
-        }
-        // todo: We should also check for Etag
-    }
-
     $response = $httpClient->request($url, $config);
 
     switch ($response->getCode()) {

lib/http.php

@@ -258,6 +258,10 @@ final class Response
     }
 
     /**
+     * HTTP response may have multiple headers with the same name.
+     *
+     * This method by default, returns only the last header.
+     *
      * @return string[]|string|null
      */
     public function getHeader(string $name, bool $all = false)