Centralize HTTP limits

NEXT_STEPS.md

@@ -43,14 +43,6 @@
 - Optionales Host-Allowlist-Feature vorhanden.
 - Tests fuer geblockte und erlaubte Ziele vorhanden.
 
-- #TODO Centralize HTTP limits (timeout/redirect/size)
-- Aufwand: `S`
-- Labels: `robustness`, `network`
-- Akzeptanzkriterien:
-- Eine zentrale Konfiguration fuer HTTP-Limits.
-- `og.php` und `link-meta.php` nutzen dieselben Limits.
-- Default-Werte sind in README dokumentiert.
-
 - #TODO Improve SQL error handling + logging
 - Aufwand: `M`
 - Labels: `sql`, `robustness`

README.md

@@ -36,6 +36,7 @@ echo decade(12345);                         // "12.345 K" (je nach PHP-Konvertie
 - `numbers.php`: Zahlen-Helfer (`decade`, `onlyNumeric`)
 - `sql.php`: Klasse `SQL` fuer Datenbankzugriffe (`get`, `single`, `list`, `keyval`, `set`)
 - `mail.php`: Mailfunktionen (`send_mail`, `send_html_mail`, `send_php_mail`)
+- `http-limits.php`: Zentrale HTTP-Limits (`httpLimits`)
 - `link-meta.php`: URL-Validierung, Fetching, Meta-Parsing, Bilddownload, Tag-Sanitization
 - `og.php`: Einfacher OG-Scan (`scanOG`)
 - `troy-api.php`: API-Helfer fuer Troy/Gitea (`sendToTroy`, `sendToGitea`)
@@ -63,6 +64,21 @@ $giteaRepo = 'repo';
 $giteaToken = 'token';
 ```
 
+HTTP-Defaults fuer Netzwerkmodule (`link-meta.php`, `og.php`):
+
+- `LIB_HTTP_TIMEOUT = 8` (Sekunden)
+- `LIB_HTTP_MAX_REDIRECTS = 4`
+- `LIB_HTTP_MAX_BYTES = 5242880` (5 MiB)
+
+Optional vor dem Include ueberschreiben:
+
+```php
+<?php
+define('LIB_HTTP_TIMEOUT', 10);
+define('LIB_HTTP_MAX_REDIRECTS', 5);
+define('LIB_HTTP_MAX_BYTES', 8 * 1024 * 1024);
+```
+
 ## Runnable Examples
 
 ### `string.php`

http-limits.php

@@ -0,0 +1,24 @@
+<?php
+declare(strict_types=1);
+
+if (!defined('LIB_HTTP_TIMEOUT')) {
+    define('LIB_HTTP_TIMEOUT', 8);
+}
+
+if (!defined('LIB_HTTP_MAX_REDIRECTS')) {
+    define('LIB_HTTP_MAX_REDIRECTS', 4);
+}
+
+if (!defined('LIB_HTTP_MAX_BYTES')) {
+    define('LIB_HTTP_MAX_BYTES', 5 * 1024 * 1024);
+}
+
+function httpLimits(): array {
+    return [
+        'timeout' => max(1, (int) LIB_HTTP_TIMEOUT),
+        'max_redirects' => max(0, (int) LIB_HTTP_MAX_REDIRECTS),
+        'max_bytes' => max(1, (int) LIB_HTTP_MAX_BYTES),
+        'user_agent' => 'star-citizen.de-linkbot/1.0'
+    ];
+}
+

link-meta.php

@@ -1,13 +1,16 @@
 <?php
 declare(strict_types=1);
+require_once __DIR__ . '/http-limits.php';
 
-function httpContext(int $timeout = 8) {
+function httpContext(?int $timeout = null) {
+    $limits = httpLimits();
+    $resolvedTimeout = $timeout === null ? $limits['timeout'] : max(1, $timeout);
     return stream_context_create([
         'http' => [
-            'timeout' => $timeout,
+            'timeout' => $resolvedTimeout,
             'follow_location' => 1,
-            'max_redirects' => 4,
-            'user_agent' => 'star-citizen.de-linkbot/1.0',
+            'max_redirects' => $limits['max_redirects'],
+            'user_agent' => $limits['user_agent'],
             'ignore_errors' => true
         ],
         'ssl' => [
@@ -66,10 +69,14 @@ function resolveUrl(string $url, string $baseUrl): ?string {
     return $baseParts['scheme'] . '://' . $baseParts['host'] . $path . $url;
 }
 
-function safeFetch(string $url, int $timeout = 8): ?string {
+function safeFetch(string $url, ?int $timeout = null): ?string {
+    $limits = httpLimits();
     $ctx = httpContext($timeout);
     $content = @file_get_contents($url, false, $ctx);
-    return $content === false ? null : $content;
+    if ($content === false || strlen($content) > $limits['max_bytes']) {
+        return null;
+    }
+    return $content;
 }
 
 function downloadImageFromUrl(string $url, string $baseUrl, string $destinationFolder = 'upl/'): ?string {
@@ -82,8 +89,8 @@ function downloadImageFromUrl(string $url, string $baseUrl, string $destinationF
         return null;
     }
 
-    $imageContent = safeFetch($resolved, 10);
-    if ($imageContent === null || strlen($imageContent) === 0 || strlen($imageContent) > (5 * 1024 * 1024)) {
+    $imageContent = safeFetch($resolved);
+    if ($imageContent === null || strlen($imageContent) === 0) {
         return null;
     }
 
@@ -130,7 +137,7 @@ function getPageInfo(string $url): array {
         return $ret;
     }
 
-    $html = safeFetch($normalized, 10);
+    $html = safeFetch($normalized);
     if ($html === null) {
         $ret['error'] = 'seite_nicht_erreichbar';
         return $ret;

og.php

@@ -1,9 +1,27 @@
 <?php
 declare(strict_types=1);
+require_once __DIR__ . '/http-limits.php';
 
 function scanOG(string $url): array {
   $og = array();
-  $html = file_get_contents($url);
+  $limits = httpLimits();
+  $ctx = stream_context_create([
+    'http' => [
+      'timeout' => $limits['timeout'],
+      'follow_location' => 1,
+      'max_redirects' => $limits['max_redirects'],
+      'user_agent' => $limits['user_agent'],
+      'ignore_errors' => true
+    ],
+    'ssl' => [
+      'verify_peer' => true,
+      'verify_peer_name' => true
+    ]
+  ]);
+  $html = @file_get_contents($url, false, $ctx);
+  if ($html === false || strlen($html) > $limits['max_bytes']) {
+    return $og;
+  }
 
   $re = '/<meta (name|property)=("|\')(.*?)("|\').*?content=("|\')(.*?)("|\')/m';
   preg_match_all($re, $html, $matches, PREG_SET_ORDER, 0);