fix: Goodreads cover scraper fetches via system curl (Cloudflare bypass)

The v0.7.45-rc.1 Goodreads cover lookup used PHP's HTTP client (Guzzle), which
Cloudflare answers with a 202 anti-bot page on real servers — so it returned
NULL in production and only "worked" from a residential dev IP. Verified on a
live server: PHP's curl-extension and Guzzle both get 202, the system `curl`
binary gets 200.

Open Library plugin 1.0.3: getGoodreadsCover now shells out to the system curl
binary (found via a guarded lookup, skipped when shell_exec is disabled), which
passes Cloudflare's TLS fingerprint check. The ISBN is validated to [0-9Xx]{10,13}
and passed through escapeshellarg, so no shell injection is possible. Where exec
is unavailable it degrades to no cover, exactly as before. Verified end-to-end
from a real cPanel server AND a php:8.4-apache container (both return the og:image).

Bumps version to 0.7.45-rc.2.

Author: fabiodalez-dev

README.md

@@ -45,7 +45,7 @@ A catalogue-consistency and scraper-maintenance release.
 - **Catalogue availability facets now reflect real circulation state** — reserved, on-loan, available and non-circulating records are shown and counted separately; records without physical copies remain visible under the complete catalogue without being mislabelled as loans ([#303](https://github.com/fabiodalez-dev/Pinakes/issues/303)).
 - **Catalogue card rows stay aligned** — a card with a subtitle no longer pushes its author, publisher and "Details" out of line with neighbouring cards; the subtitle space is reserved per row only where a row actually needs it, and the layout re-aligns after AJAX filtering, on resize, and once web fonts settle ([#298](https://github.com/fabiodalez-dev/Pinakes/issues/298)).
 - **Language completion statistics are derived from the Italian source catalogue** — every locale (list and edit views alike) now uses the same live denominator, so translation percentages cannot drift from the shipped keys.
-- **Open Library plugin 1.0.2** — replaces the retired third-party Goodreads-cover service with a direct, timeout-bounded lookup of the public Goodreads ISBN page. Only HTTPS cover URLs from exact Goodreads/Amazon CDN domains or their subdomains are accepted; its manifest now correctly requires Pinakes 0.7.16+ and PHP 8.2+.
+- **Open Library plugin 1.0.3** — replaces the retired third-party Goodreads-cover service with a direct, timeout-bounded lookup of the public Goodreads ISBN page. Because Goodreads sits behind Cloudflare (which serves an anti-bot page to PHP's HTTP client but not to the system `curl`), the cover lookup shells out to the `curl` binary when the host allows process execution, and degrades gracefully to no cover where it doesn't. Only HTTPS cover URLs from exact Goodreads/Amazon CDN domains or their subdomains are accepted; its manifest requires Pinakes 0.7.16+ and PHP 8.2+.
 - **Email templates: `{{pickup_deadline}}` now resolves**, and every placeholder token carries a localised description tooltip in the template editor ([#304](https://github.com/fabiodalez-dev/Pinakes/issues/304)).
 
 ### Database Changes

storage/plugins/open-library/OpenLibraryPlugin.php

@@ -419,25 +419,84 @@ private function getGoodreadsCover(string $isbn): ?string
             return null;
         }
 
-        try {
-            $res = \App\Support\HttpClient::get(
-                'https://www.goodreads.com/book/isbn/' . urlencode($clean),
-                [
-                    'Accept' => 'text/html,application/xhtml+xml',
-                    'Accept-Language' => 'en-US,en;q=0.9',
-                ],
-                ['timeout' => 5, 'max_redirects' => 3, 'user_agent' => self::USER_AGENT]
-            );
-
-            if (!$res['ok'] || $res['status'] !== 200 || empty($res['body'])) {
-                return null;
-            }
+        $html = $this->fetchGoodreadsHtml($clean);
+        if ($html === null) {
+            return null;
+        }
+        return $this->extractGoodreadsCoverFromHtml($html);
+    }
 
-            return $this->extractGoodreadsCoverFromHtml((string) $res['body']);
+    /**
+     * Fetch the Goodreads book page HTML.
+     *
+     * Goodreads sits behind Cloudflare, which serves a 202 anti-bot interstitial
+     * to PHP's HTTP stack (Guzzle/libcurl present a TLS/JA3 fingerprint Cloudflare
+     * challenges) — so an in-process HTTP client never gets the real page. The
+     * system `curl` binary's fingerprint passes, so we shell out to it when the
+     * host allows process execution; otherwise we give up gracefully (the caller
+     * simply gets no cover). $isbn is already validated to [0-9Xx]{10,13} by the
+     * caller and is passed through escapeshellarg, so no shell injection is
+     * possible.
+     *
+     * @param string $isbn Validated ISBN-10/13 (digits and X only)
+     * @return string|null Raw HTML, or null when unreachable / exec unavailable
+     */
+    private function fetchGoodreadsHtml(string $isbn): ?string
+    {
+        $curl = $this->findCurlBinary();
+        if ($curl === null) {
+            return null;
+        }
+        $url = 'https://www.goodreads.com/book/isbn/' . $isbn;
+        $cmd = sprintf(
+            '%s -sL --max-time 8 -A %s %s 2>/dev/null',
+            escapeshellarg($curl),
+            escapeshellarg(self::USER_AGENT),
+            escapeshellarg($url)
+        );
+        try {
+            $out = shell_exec($cmd);
         } catch (\Throwable $e) {
-            \App\Support\SecureLogger::warning('[OpenLibrary] Goodreads cover scrape error: ' . $e->getMessage());
+            \App\Support\SecureLogger::warning('[OpenLibrary] Goodreads cover fetch error: ' . $e->getMessage());
             return null;
         }
+        return (is_string($out) && $out !== '') ? $out : null;
+    }
+
+    /**
+     * Locate a usable system `curl` binary, or null when process execution is
+     * unavailable (shell_exec disabled — common on hardened shared hosting) or
+     * no curl is installed. Cached per instance.
+     *
+     * @return string|null Absolute path to curl, or null
+     */
+    private function findCurlBinary(): ?string
+    {
+        static $resolved = false;
+        static $path = null;
+        if ($resolved) {
+            return $path;
+        }
+        $resolved = true;
+
+        if (!function_exists('shell_exec')) {
+            return $path = null;
+        }
+        $disabled = array_map('trim', explode(',', (string) ini_get('disable_functions')));
+        if (in_array('shell_exec', $disabled, true)) {
+            return $path = null;
+        }
+        foreach (['/usr/bin/curl', '/bin/curl', '/usr/local/bin/curl'] as $candidate) {
+            if (@is_executable($candidate)) {
+                return $path = $candidate;
+            }
+        }
+        $found = @shell_exec('command -v curl 2>/dev/null');
+        $found = is_string($found) ? trim($found) : '';
+        if ($found !== '' && @is_executable($found)) {
+            return $path = $found;
+        }
+        return $path = null;
     }
 
     /**

storage/plugins/open-library/plugin.json

@@ -2,7 +2,7 @@
   "name": "open-library",
   "display_name": "Open Library Scraper",
   "description": "Integrazione con le API di Open Library (openlibrary.org) per lo scraping di metadati dei libri. Fornisce dati completi su edizioni, opere, autori e copertine ad alta risoluzione, con copertine di fallback recuperate dalla pagina pubblica di Goodreads.",
-  "version": "1.0.2",
+  "version": "1.0.3",
   "author": "Fabiodalez",
   "author_url": "",
   "plugin_url": "https://openlibrary.org",

tests/plugin-goodreads-cover.unit.php

@@ -5,10 +5,13 @@
  * Unit test for the Goodreads cover scraper in OpenLibraryPlugin
  * (replaces the retired bookcover.longitood.com service).
  *
- * The parsing (og:image extraction + trusted-CDN host guard) is tested
- * deterministically against HTML fixtures — no network. A final best-effort
- * live check hits Goodreads only if reachable, and never fails the suite when
- * the network is unavailable.
+ * The scraper fetches the Goodreads book page by shelling out to the system
+ * `curl` binary — Cloudflare's anti-bot serves a 202 to PHP's HTTP client but
+ * 200 to curl's TLS fingerprint — and parses the og:image with a trusted-CDN
+ * host guard. The parsing is tested deterministically against HTML fixtures
+ * (no network). A final best-effort live check hits Goodreads only when a curl
+ * binary is available and the page is reachable, and never fails the suite when
+ * exec is disabled or the network is unavailable.
  *
  * Run:  php tests/plugin-goodreads-cover.unit.php   (exit 0 iff all pass)
  */

version.json

@@ -1,5 +1,5 @@
 {
     "name": "Pinakes",
-    "version": "0.7.45-rc.1",
+    "version": "0.7.45-rc.2",
     "description": "Library Management System - Sistema di Gestione Bibliotecaria"
 }