feat: Add PDF screenshot generation and display (#995)

* Updated pdf2json to 3.1.5

* Extract and store a screenshot from PDF files using pdf2pic

* Installing graphicsmagick and ghostscript

* Generate Missing PDF screenshot with tidyAssets worker for backward support

* Display PDF screenshot instead of the PDF in web if it exists.

* Display PDF screenshot in mobile app if exists.

* Updated pnpm-lock.yaml

* Removed console.log

* Revert the unnecessary changes in package.json

* Revert pnpm-lock changes

* Prevent rendering PDF files if the screenshot is not generated

* refactor: replace useEffect with useMemo for section initialization

* feat: show PDF file download button and handle large PDFs by defaulting to screenshot view

* feat: add file size to openapi spec

* feature: Add Assets preprocessing in fix mode to admin actions

* i18n: add reprocess_assets_fix_mode translation

* i18n: Add missing ar translations

* A bunch of fixes

* Fix openspec schema

---------

Co-authored-by: Mohamed Bassem <me@mbassem.com>

23 files changed, 389 insertions, 94 deletions

diff --git a/apps/mobile/components/bookmarks/BookmarkCard.tsx b/apps/mobile/components/bookmarks/BookmarkCard.tsx
index ce294a6f..3cbd064e 100644
--- a/apps/mobile/components/bookmarks/BookmarkCard.tsx
+++ b/apps/mobile/components/bookmarks/BookmarkCard.tsx
@@ -1,3 +1,4 @@
+import React from "react";
 import {
   ActivityIndicator,
   Alert,
@@ -300,11 +301,15 @@ function AssetCard({
   }
   const title = bookmark.title ?? bookmark.content.fileName;
 
+  const assetImage =
+    bookmark.assets.find((r) => r.assetType == "assetScreenshot")?.id ??
+    bookmark.content.assetId;
+
   return (
     <View className="flex gap-2">
       <Pressable onPress={onOpenBookmark}>
         <BookmarkAssetImage
-          assetId={bookmark.content.assetId}
+          assetId={assetImage}
           className="h-56 min-h-56 w-full object-cover"
         />
       </Pressable>
diff --git a/apps/web/components/admin/AdminActions.tsx b/apps/web/components/admin/AdminActions.tsx
index 34b3d63a..fb151ac8 100644
--- a/apps/web/components/admin/AdminActions.tsx
+++ b/apps/web/components/admin/AdminActions.tsx
@@ -37,6 +37,21 @@ export default function AdminActions() {
       },
     });
 
+  const { mutate: reprocessAssetsFixMode, isPending: isReprocessingPending } =
+    api.admin.reprocessAssetsFixMode.useMutation({
+      onSuccess: () => {
+        toast({
+          description: "Reprocessing enqueued",
+        });
+      },
+      onError: (e) => {
+        toast({
+          variant: "destructive",
+          description: e.message,
+        });
+      },
+    });
+
   const {
     mutate: reRunInferenceOnAllBookmarks,
     isPending: isInferencePending,
@@ -126,6 +141,13 @@ export default function AdminActions() {
         </ActionButton>
         <ActionButton
           variant="destructive"
+          loading={isReprocessingPending}
+          onClick={() => reprocessAssetsFixMode()}
+        >
+          {t("admin.actions.reprocess_assets_fix_mode")}
+        </ActionButton>
+        <ActionButton
+          variant="destructive"
           loading={isTidyAssetsPending}
           onClick={() => tidyAssets()}
         >
diff --git a/apps/web/components/dashboard/bookmarks/AssetCard.tsx b/apps/web/components/dashboard/bookmarks/AssetCard.tsx
index 61b3bc8d..0cb75b3f 100644
--- a/apps/web/components/dashboard/bookmarks/AssetCard.tsx
+++ b/apps/web/components/dashboard/bookmarks/AssetCard.tsx
@@ -2,6 +2,8 @@
 
 import Image from "next/image";
 import Link from "next/link";
+import { cn } from "@/lib/utils";
+import { FileText } from "lucide-react";
 
 import type { ZBookmarkTypeAsset } from "@hoarder/shared/types/bookmarks";
 import { getAssetUrl } from "@hoarder/shared-react/utils/assetUtils";
@@ -32,12 +34,28 @@ function AssetImage({
       );
     }
     case "pdf": {
+      const screenshotAssetId = bookmark.assets.find(
+        (r) => r.assetType === "assetScreenshot",
+      )?.id;
+      if (!screenshotAssetId) {
+        return (
+          <div
+            className={cn(className, "flex items-center justify-center")}
+            title="PDF screenshot not available. Run asset preprocessing job to generate one screenshot"
+          >
+            <FileText size={80} />
+          </div>
+        );
+      }
       return (
-        <iframe
-          title={bookmarkedAsset.assetId}
-          className={className}
-          src={getAssetUrl(bookmarkedAsset.assetId)}
-        />
+        <Link href={`/dashboard/preview/${bookmark.id}`}>
+          <Image
+            alt="asset"
+            src={getAssetUrl(screenshotAssetId)}
+            fill={true}
+            className={className}
+          />
+        </Link>
       );
     }
     default: {
diff --git a/apps/web/components/dashboard/preview/AssetContentSection.tsx b/apps/web/components/dashboard/preview/AssetContentSection.tsx
index 03ab8a43..8590d2ad 100644
--- a/apps/web/components/dashboard/preview/AssetContentSection.tsx
+++ b/apps/web/components/dashboard/preview/AssetContentSection.tsx
@@ -1,42 +1,117 @@
+import { useMemo, useState } from "react";
 import Image from "next/image";
 import Link from "next/link";
+import {
+  Select,
+  SelectContent,
+  SelectGroup,
+  SelectItem,
+  SelectTrigger,
+  SelectValue,
+} from "@/components/ui/select";
+import { useTranslation } from "@/lib/i18n/client";
 
+import { getAssetUrl } from "@hoarder/shared-react/utils/assetUtils";
 import { BookmarkTypes, ZBookmark } from "@hoarder/shared/types/bookmarks";
 
-export function AssetContentSection({ bookmark }: { bookmark: ZBookmark }) {
+// 20 MB
+const BIG_FILE_SIZE = 20 * 1024 * 1024;
+
+function PDFContentSection({ bookmark }: { bookmark: ZBookmark }) {
   if (bookmark.content.type != BookmarkTypes.ASSET) {
     throw new Error("Invalid content type");
   }
+  const { t } = useTranslation();
 
-  switch (bookmark.content.assetType) {
-    case "image": {
-      return (
-        <div className="relative h-full min-w-full">
-          <Link
-            href={`/api/assets/${bookmark.content.assetId}`}
-            target="_blank"
-          >
-            <Image
-              alt="asset"
-              fill={true}
-              className="object-contain"
-              src={`/api/assets/${bookmark.content.assetId}`}
-            />
-          </Link>
-        </div>
-      );
+  const initialSection = useMemo(() => {
+    if (bookmark.content.type != BookmarkTypes.ASSET) {
+      throw new Error("Invalid content type");
     }
-    case "pdf": {
-      return (
-        <iframe
-          title={bookmark.content.assetId}
-          className="h-full w-full"
-          src={`/api/assets/${bookmark.content.assetId}`}
-        />
-      );
+
+    const screenshot = bookmark.assets.find(
+      (item) => item.assetType === "assetScreenshot",
+    );
+    const bigSize =
+      bookmark.content.size && bookmark.content.size > BIG_FILE_SIZE;
+    if (bigSize && screenshot) {
+      return "screenshot";
     }
-    default: {
+    return "pdf";
+  }, [bookmark]);
+  const [section, setSection] = useState(initialSection);
+
+  const screenshot = bookmark.assets.find(
+    (r) => r.assetType === "assetScreenshot",
+  )?.id;
+
+  const content =
+    section === "screenshot" && screenshot ? (
+      <div className="relative h-full min-w-full">
+        <Image
+          alt="screenshot"
+          src={getAssetUrl(screenshot)}
+          fill={true}
+          className="object-contain"
+        />
+      </div>
+    ) : (
+      <iframe
+        title={bookmark.content.assetId}
+        className="h-full w-full"
+        src={getAssetUrl(bookmark.content.assetId)}
+      />
+    );
+
+  return (
+    <div className="flex h-full flex-col items-center gap-2">
+      <div className="flex w-full items-center justify-center gap-4">
+        <Select onValueChange={setSection} value={section}>
+          <SelectTrigger className="w-fit">
+            <SelectValue />
+          </SelectTrigger>
+          <SelectContent>
+            <SelectGroup>
+              <SelectItem value="screenshot" disabled={!screenshot}>
+                {t("common.screenshot")}
+              </SelectItem>
+              <SelectItem value="pdf">PDF</SelectItem>
+            </SelectGroup>
+          </SelectContent>
+        </Select>
+      </div>
+      {content}
+    </div>
+  );
+}
+
+function ImageContentSection({ bookmark }: { bookmark: ZBookmark }) {
+  if (bookmark.content.type != BookmarkTypes.ASSET) {
+    throw new Error("Invalid content type");
+  }
+  return (
+    <div className="relative h-full min-w-full">
+      <Link href={getAssetUrl(bookmark.content.assetId)} target="_blank">
+        <Image
+          alt="asset"
+          fill={true}
+          className="object-contain"
+          src={getAssetUrl(bookmark.content.assetId)}
+        />
+      </Link>
+    </div>
+  );
+}
+
+export function AssetContentSection({ bookmark }: { bookmark: ZBookmark }) {
+  if (bookmark.content.type != BookmarkTypes.ASSET) {
+    throw new Error("Invalid content type");
+  }
+  switch (bookmark.content.assetType) {
+    case "image":
+      return <ImageContentSection bookmark={bookmark} />;
+    case "pdf":
+      return <PDFContentSection bookmark={bookmark} />;
+    default:
       return <div>Unsupported asset type</div>;
-    }
   }
 }
diff --git a/apps/web/components/dashboard/preview/AttachmentBox.tsx b/apps/web/components/dashboard/preview/AttachmentBox.tsx
index 6547ae51..32939cb0 100644
--- a/apps/web/components/dashboard/preview/AttachmentBox.tsx
+++ b/apps/web/components/dashboard/preview/AttachmentBox.tsx
@@ -45,6 +45,7 @@ export default function AttachmentBox({ bookmark }: { bookmark: ZBookmark }) {
   const { t } = useTranslation();
   const typeToIcon: Record<ZAssetType, React.ReactNode> = {
     screenshot: <Camera className="size-4" />,
+    assetScreenshot: <Camera className="size-4" />,
     fullPageArchive: <Archive className="size-4" />,
     precrawledArchive: <Archive className="size-4" />,
     bannerImage: <Image className="size-4" />,
diff --git a/apps/web/lib/i18n/locales/ar/translation.json b/apps/web/lib/i18n/locales/ar/translation.json
index 7bd0bcad..e9239e70 100644
--- a/apps/web/lib/i18n/locales/ar/translation.json
+++ b/apps/web/lib/i18n/locales/ar/translation.json
@@ -153,7 +153,7 @@
     }
   },
   "admin": {
-    "admin_settings": "إعدادات المدير",
+    "admin_settings": "إعدادات المشرف",
     "server_stats": {
       "server_stats": "إحصائيات الخادم",
       "total_users": "إجمالي المستخدمين",
@@ -161,15 +161,36 @@
       "server_version": "إصدار الخادم"
     },
     "background_jobs": {
-      "background_jobs": "المهام التلقائية",
+      "background_jobs": "المهام الخلفية",
       "crawler_jobs": "مهام الاستكشاف",
       "indexing_jobs": "مهام الفهرسة",
-      "inference_jobs": "مهام التحليل الذكي",
-      "tidy_assets_jobs": "مهام تنظيم الملفات",
+      "inference_jobs": "مهام الاستدلال",
+      "tidy_assets_jobs": "مهام تنظيم الوسائط",
       "job": "مهمة",
       "queued": "في قائمة الانتظار",
-      "pending": "معلق",
-      "failed": "فشل"
+      "pending": "قيد الانتظار",
+      "failed": "فشلت"
+    },
+    "actions": {
+      "recrawl_failed_links_only": "إعادة استكشاف الروابط الفاشلة فقط",
+      "recrawl_all_links": "إعادة استكشاف جميع الروابط",
+      "without_inference": "بدون استدلال",
+      "regenerate_ai_tags_for_failed_bookmarks_only": "إعادة إنشاء علامات الذكاء الاصطناعي للإشارات المرجعية الفاشلة فقط",
+      "regenerate_ai_tags_for_all_bookmarks": "إعادة إنشاء علامات الذكاء الاصطناعي لجميع الإشارات المرجعية",
+      "reindex_all_bookmarks": "إعادة فهرسة جميع الإشارات المرجعية",
+      "compact_assets": "ضغط الوسائط",
+      "reprocess_assets_fix_mode": "إعادة معالجة الوسائط (وضع الإصلاح)"
+    },
+    "users_list": {
+      "users_list": "قائمة المستخدمين",
+      "create_user": "إنشاء مستخدم",
+      "change_role": "تغيير الدور",
+      "reset_password": "إعادة تعيين كلمة المرور",
+      "delete_user": "حذف المستخدم",
+      "num_bookmarks": "عدد الإشارات المرجعية",
+      "asset_sizes": "أحجام الوسائط",
+      "local_user": "مستخدم محلي",
+      "confirm_password": "تأكيد كلمة المرور"
     }
   },
   "options": {
diff --git a/apps/web/lib/i18n/locales/da/translation.json b/apps/web/lib/i18n/locales/da/translation.json
index 3822d5c6..4fe69650 100644
--- a/apps/web/lib/i18n/locales/da/translation.json
+++ b/apps/web/lib/i18n/locales/da/translation.json
@@ -94,7 +94,8 @@
       "recrawl_all_links": "Gennemsøg alle links",
       "without_inference": "Uden inferens",
       "regenerate_ai_tags_for_all_bookmarks": "Genopret AI-tags for alle bogmærker",
-      "reindex_all_bookmarks": "Genindeksér alle bogmærker"
+      "reindex_all_bookmarks": "Genindeksér alle bogmærker",
+      "reprocess_assets_fix_mode": "Genbehandling af aktiver (Fix Mode)"
     },
     "background_jobs": {
       "inference_jobs": "Inferensopgaver",
diff --git a/apps/web/lib/i18n/locales/de/translation.json b/apps/web/lib/i18n/locales/de/translation.json
index c20a2273..ccebf1f1 100644
--- a/apps/web/lib/i18n/locales/de/translation.json
+++ b/apps/web/lib/i18n/locales/de/translation.json
@@ -175,7 +175,8 @@
       "regenerate_ai_tags_for_failed_bookmarks_only": "KI-Tags nur für fehlgeschlagene Lesezeichen neu generieren",
       "regenerate_ai_tags_for_all_bookmarks": "KI-Tags für alle Lesezeichen neu generieren",
       "reindex_all_bookmarks": "Alle Lesezeichen neu indizieren",
-      "compact_assets": "Assets komprimieren"
+      "compact_assets": "Assets komprimieren",
+      "reprocess_assets_fix_mode": "Assets neu verarbeiten (Fix-Modus)"
     },
     "users_list": {
       "users_list": "Benutzerliste",
diff --git a/apps/web/lib/i18n/locales/en/translation.json b/apps/web/lib/i18n/locales/en/translation.json
index 2e80f2f4..81ef942f 100644
--- a/apps/web/lib/i18n/locales/en/translation.json
+++ b/apps/web/lib/i18n/locales/en/translation.json
@@ -178,7 +178,8 @@
       "regenerate_ai_tags_for_failed_bookmarks_only": "Regenerate AI Tags for Failed Bookmarks Only",
       "regenerate_ai_tags_for_all_bookmarks": "Regenerate AI Tags for All Bookmarks",
       "reindex_all_bookmarks": "Reindex All Bookmarks",
-      "compact_assets": "Compact Assets"
+      "compact_assets": "Compact Assets",
+      "reprocess_assets_fix_mode": "Reprocess Assets (Fix Mode)"
     },
     "users_list": {
       "users_list": "Users List",
diff --git a/apps/web/lib/i18n/locales/es/translation.json b/apps/web/lib/i18n/locales/es/translation.json
index 40c6cb01..3a1a7e3c 100644
--- a/apps/web/lib/i18n/locales/es/translation.json
+++ b/apps/web/lib/i18n/locales/es/translation.json
@@ -146,7 +146,8 @@
       "compact_assets": "Optimizar multimedia",
       "without_inference": "Sin inferencia",
       "recrawl_failed_links_only": "Recrawlear solo los enlaces fallidos",
-      "recrawl_all_links": "Recrawlear todos los enlaces"
+      "recrawl_all_links": "Recrawlear todos los enlaces",
+      "reprocess_assets_fix_mode": "Reprocesar assets (modo fijo)"
     },
     "users_list": {
       "users_list": "Lista de usuarios",
diff --git a/apps/web/lib/i18n/locales/fr/translation.json b/apps/web/lib/i18n/locales/fr/translation.json
index b7834a7b..1772c2ff 100644
--- a/apps/web/lib/i18n/locales/fr/translation.json
+++ b/apps/web/lib/i18n/locales/fr/translation.json
@@ -146,7 +146,8 @@
       "regenerate_ai_tags_for_failed_bookmarks_only": "Régénérer les tags AI uniquement pour les favoris échoués",
       "regenerate_ai_tags_for_all_bookmarks": "Régénérer les tags AI pour tous les favoris",
       "reindex_all_bookmarks": "Réindexer tous les favoris",
-      "compact_assets": "Compacter les assets"
+      "compact_assets": "Compacter les assets",
+      "reprocess_assets_fix_mode": "Reprocesser les assets (mode fix)"
     },
     "users_list": {
       "users_list": "Liste des utilisateurs",
diff --git a/apps/web/lib/i18n/locales/gl/translation.json b/apps/web/lib/i18n/locales/gl/translation.json
index eb65ca64..363ffac8 100644
--- a/apps/web/lib/i18n/locales/gl/translation.json
+++ b/apps/web/lib/i18n/locales/gl/translation.json
@@ -178,7 +178,8 @@
       "regenerate_ai_tags_for_failed_bookmarks_only": "Rexenerar etiquetas IA so en marcadores errados",
       "regenerate_ai_tags_for_all_bookmarks": "Rexenerar etiquetas IA para todos os marcadores",
       "reindex_all_bookmarks": "Reindexar marcadores",
-      "compact_assets": "Optimizar multimedia"
+      "compact_assets": "Optimizar multimedia",
+      "reprocess_assets_fix_mode": "Reprocesar assets (modo fixo)"
     },
     "users_list": {
       "users_list": "Listado de usuarios",
diff --git a/apps/web/lib/i18n/locales/hr/translation.json b/apps/web/lib/i18n/locales/hr/translation.json
index 6e250924..7a72d295 100644
--- a/apps/web/lib/i18n/locales/hr/translation.json
+++ b/apps/web/lib/i18n/locales/hr/translation.json
@@ -36,7 +36,8 @@
       "recrawl_all_links": "Ponovno pregledavanje svih veza",
       "regenerate_ai_tags_for_all_bookmarks": "Ponovno generiranje AI oznaka za sve oznake",
       "without_inference": "Bez zaključivanja",
-      "compact_assets": "Kompaktiranje resursa"
+      "compact_assets": "Kompaktiranje resursa",
+      "reprocess_assets_fix_mode": "Ponovno postupanje s resursima (fiksni mod)"
     }
   },
   "layouts": {
diff --git a/apps/web/lib/i18n/locales/hu/translation.json b/apps/web/lib/i18n/locales/hu/translation.json
index 38ef96b4..439212f4 100644
--- a/apps/web/lib/i18n/locales/hu/translation.json
+++ b/apps/web/lib/i18n/locales/hu/translation.json
@@ -258,7 +258,8 @@
       "regenerate_ai_tags_for_all_bookmarks": "Minden könyvjelző MI címkéjének lecserélése",
       "regenerate_ai_tags_for_failed_bookmarks_only": "Hibás könyvjelzők MI címkéjének lecserélése",
       "reindex_all_bookmarks": "Minden könyvjelző újraindexelése",
-      "compact_assets": "Kompakt tulajdonok"
+      "compact_assets": "Kompakt tulajdonok",
+      "reprocess_assets_fix_mode": "Tulajdonok függvényezése (Fix Mod)"
     },
     "users_list": {
       "asset_sizes": "Tulajdon méretek",
diff --git a/apps/web/lib/i18n/locales/it/translation.json b/apps/web/lib/i18n/locales/it/translation.json
index e24b6b7f..4b093b72 100644
--- a/apps/web/lib/i18n/locales/it/translation.json
+++ b/apps/web/lib/i18n/locales/it/translation.json
@@ -201,7 +201,8 @@
       "regenerate_ai_tags_for_failed_bookmarks_only": "Rigenera tag AI solo per i segnalibri falliti",
       "regenerate_ai_tags_for_all_bookmarks": "Rigenera tag AI per tutti i segnalibri",
       "compact_assets": "Compatta asset",
-      "reindex_all_bookmarks": "Reindicizza tutti i segnalibri"
+      "reindex_all_bookmarks": "Reindicizza tutti i segnalibri",
+      "reprocess_assets_fix_mode": "Riprocessa asset (modalità fissa)"
     },
     "users_list": {
       "users_list": "Lista utenti",
diff --git a/apps/web/lib/i18n/locales/pl/translation.json b/apps/web/lib/i18n/locales/pl/translation.json
index 0d026542..66921560 100644
--- a/apps/web/lib/i18n/locales/pl/translation.json
+++ b/apps/web/lib/i18n/locales/pl/translation.json
@@ -148,7 +148,8 @@
       "regenerate_ai_tags_for_failed_bookmarks_only": "Regeneruj tagi AI tylko dla nieudanych zakładek",
       "regenerate_ai_tags_for_all_bookmarks": "Regeneruj tagi AI dla wszystkich zakładek",
       "reindex_all_bookmarks": "Ponowne indeksowanie wszystkich zakładek",
-      "compact_assets": "Kompaktuj zasoby"
+      "compact_assets": "Kompaktuj zasoby",
+      "reprocess_assets_fix_mode": "Ponowne przetwarzanie zasobów (tryb fiksny)"
     }
   },
   "tags": {
diff --git a/apps/web/lib/i18n/locales/ru/translation.json b/apps/web/lib/i18n/locales/ru/translation.json
index 4a8cdd52..1d4c50bd 100644
--- a/apps/web/lib/i18n/locales/ru/translation.json
+++ b/apps/web/lib/i18n/locales/ru/translation.json
@@ -211,7 +211,8 @@
       "compact_assets": "Сжать ресурсы",
       "regenerate_ai_tags_for_failed_bookmarks_only": "Перегенерировать ИИ метки только для неудачных закладок",
       "reindex_all_bookmarks": "Переиндексировать все закладки",
-      "recrawl_all_links": "Пересканировать все ссылки"
+      "recrawl_all_links": "Пересканировать все ссылки",
+      "reprocess_assets_fix_mode": "Перепроцессировать ресурсы (фиксный режим)"
     },
     "admin_settings": "Настройки администратора"
   },
diff --git a/apps/web/lib/i18n/locales/tr/translation.json b/apps/web/lib/i18n/locales/tr/translation.json
index 9840c6f0..227f6dac 100644
--- a/apps/web/lib/i18n/locales/tr/translation.json
+++ b/apps/web/lib/i18n/locales/tr/translation.json
@@ -148,7 +148,8 @@
       "regenerate_ai_tags_for_failed_bookmarks_only": "Yalnızca Başarısız Yer İşaretleri için Yapay Zeka Etiketlerini Yeniden Oluştur",
       "regenerate_ai_tags_for_all_bookmarks": "Tüm Yer İşaretleri için Yapay Zeka Etiketlerini Yeniden Oluştur",
       "reindex_all_bookmarks": "Tüm Yer İşaretlerini Yeniden Dizine Al",
-      "compact_assets": "Varlıkları Sıkıştır"
+      "compact_assets": "Varlıkları Sıkıştır",
+      "reprocess_assets_fix_mode": "Varlıkları Yeniden İşle (Fix Mod)"
     },
     "users_list": {
       "users_list": "Kullanıcı Listesi",
diff --git a/apps/web/lib/i18n/locales/zh/translation.json b/apps/web/lib/i18n/locales/zh/translation.json
index 84a9e17a..d798b716 100644
--- a/apps/web/lib/i18n/locales/zh/translation.json
+++ b/apps/web/lib/i18n/locales/zh/translation.json
@@ -175,7 +175,8 @@
       "regenerate_ai_tags_for_failed_bookmarks_only": "仅为失败书签重新生成AI标签",
       "regenerate_ai_tags_for_all_bookmarks": "为所有书签重新生成AI标签",
       "reindex_all_bookmarks": "重新索引所有书签",
-      "compact_assets": "压缩资产"
+      "compact_assets": "压缩资产",
+      "reprocess_assets_fix_mode": "重新处理资产(固定模式)"
     },
     "users_list": {
       "users_list": "用户列表",
diff --git a/apps/web/lib/i18n/locales/zhtw/translation.json b/apps/web/lib/i18n/locales/zhtw/translation.json
index aada5492..284b5de2 100644
--- a/apps/web/lib/i18n/locales/zhtw/translation.json
+++ b/apps/web/lib/i18n/locales/zhtw/translation.json
@@ -156,7 +156,8 @@
       "regenerate_ai_tags_for_failed_bookmarks_only": "僅重新產生失敗書籤的 AI 標籤",
       "regenerate_ai_tags_for_all_bookmarks": "重新產生所有書籤的 AI 標籤",
       "reindex_all_bookmarks": "重新索引所有書籤",
-      "compact_assets": "壓縮資源"
+      "compact_assets": "壓縮資源",
+      "reprocess_assets_fix_mode": "重新處理資源(固定模式)"
     },
     "users_list": {
       "users_list": "使用者清單",
diff --git a/apps/workers/assetPreprocessingWorker.ts b/apps/workers/assetPreprocessingWorker.ts
index 5c4937e5..f94eeb9e 100644
--- a/apps/workers/assetPreprocessingWorker.ts
+++ b/apps/workers/assetPreprocessingWorker.ts
@@ -2,12 +2,18 @@ import os from "os";
 import { eq } from "drizzle-orm";
 import { DequeuedJob, Runner } from "liteque";
 import PDFParser from "pdf2json";
+import { fromBuffer } from "pdf2pic";
 import { createWorker } from "tesseract.js";
 
 import type { AssetPreprocessingRequest } from "@hoarder/shared/queues";
 import { db } from "@hoarder/db";
-import { bookmarkAssets, bookmarks } from "@hoarder/db/schema";
-import { readAsset } from "@hoarder/shared/assetdb";
+import {
+  assets,
+  AssetTypes,
+  bookmarkAssets,
+  bookmarks,
+} from "@hoarder/db/schema";
+import { newAssetId, readAsset, saveAsset } from "@hoarder/shared/assetdb";
 import serverConfig from "@hoarder/shared/config";
 import logger from "@hoarder/shared/logger";
 import {
@@ -67,17 +73,14 @@ async function readImageText(buffer: Buffer) {
 
 async function readPDFText(buffer: Buffer): Promise<{
   text: string;
-  metadata: Record<string, string>;
+  metadata: Record<string, object>;
 }> {
   return new Promise((resolve, reject) => {
-    // Need raw text flag represents as number (1), reference : https://github.com/modesty/pdf2json/issues/76#issuecomment-236569265
-    const pdfParser = new PDFParser(null, 1);
+    const pdfParser = new PDFParser(null, true);
     pdfParser.on("pdfParser_dataError", reject);
     pdfParser.on("pdfParser_dataReady", (pdfData) => {
       resolve({
-        // The type isn't set correctly, reference : https://github.com/modesty/pdf2json/issues/327
-        // eslint-disable-next-line
-        text: (pdfParser as any).getRawTextContent(),
+        text: pdfParser.getRawTextContent(),
         metadata: pdfData.Meta,
       });
     });
@@ -85,11 +88,102 @@ async function readPDFText(buffer: Buffer): Promise<{
   });
 }
 
-async function preprocessImage(
+export async function extractAndSavePDFScreenshot(
   jobId: string,
   asset: Buffer,
-): Promise<{ content: string; metadata: string | null } | undefined> {
+  bookmark: NonNullable<Awaited<ReturnType<typeof getBookmark>>>,
+  isFixMode: boolean,
+): Promise<boolean> {
+  {
+    const alreadyHasScreenshot =
+      bookmark.assets.find(
+        (r) => r.assetType === AssetTypes.ASSET_SCREENSHOT,
+      ) !== undefined;
+    if (alreadyHasScreenshot && isFixMode) {
+      logger.info(
+        `[assetPreprocessing][${jobId}] Skipping PDF screenshot generation as it's already been generated.`,
+      );
+      return false;
+    }
+  }
+  logger.info(
+    `[assetPreprocessing][${jobId}] Attempting to generate PDF screenshot for bookmarkId: ${bookmark.id}`,
+  );
+  try {
+    /**
+     * If you encountered any issues with this library, make sure you have ghostscript and graphicsmagick installed following this URL
+     * https://github.com/yakovmeister/pdf2image/blob/HEAD/docs/gm-installation.md
+     */
+    const screenshot = await fromBuffer(asset, {
+      density: 100,
+      quality: 100,
+      format: "png",
+      preserveAspectRatio: true,
+    })(1, { responseType: "buffer" });
+
+    if (!screenshot.buffer) {
+      logger.error(
+        `[assetPreprocessing][${jobId}] Failed to generate PDF screenshot`,
+      );
+      return false;
+    }
+
+    // Store the screenshot
+    const assetId = newAssetId();
+    const fileName = "screenshot.png";
+    const contentType = "image/png";
+    await saveAsset({
+      userId: bookmark.userId,
+      assetId,
+      asset: screenshot.buffer,
+      metadata: {
+        contentType,
+        fileName,
+      },
+    });
+
+    // Insert into database
+    await db.insert(assets).values({
+      id: assetId,
+      bookmarkId: bookmark.id,
+      userId: bookmark.userId,
+      assetType: AssetTypes.ASSET_SCREENSHOT,
+      contentType,
+      size: screenshot.buffer.byteLength,
+      fileName,
+    });
+
+    logger.info(
+      `[assetPreprocessing][${jobId}] Successfully saved PDF screenshot to database`,
+    );
+    return true;
+  } catch (error) {
+    logger.error(
+      `[assetPreprocessing][${jobId}] Failed to process PDF screenshot: ${error}`,
+    );
+    return false;
+  }
+}
+
+async function extractAndSaveImageText(
+  jobId: string,
+  asset: Buffer,
+  bookmark: NonNullable<Awaited<ReturnType<typeof getBookmark>>>,
+  isFixMode: boolean,
+): Promise<boolean> {
+  {
+    const alreadyHasText = !!bookmark.asset.content;
+    if (alreadyHasText && isFixMode) {
+      logger.info(
+        `[assetPreprocessing][${jobId}] Skipping image text extraction as it's already been extracted.`,
+      );
+      return false;
+    }
+  }
   let imageText = null;
+  logger.info(
+    `[assetPreprocessing][${jobId}] Attempting to extract text from image.`,
+  );
   try {
     imageText = await readImageText(asset);
   } catch (e) {
@@ -98,19 +192,40 @@ async function preprocessImage(
     );
   }
   if (!imageText) {
-    return undefined;
+    return false;
   }
 
   logger.info(
     `[assetPreprocessing][${jobId}] Extracted ${imageText.length} characters from image.`,
   );
-  return { content: imageText, metadata: null };
+  await db
+    .update(bookmarkAssets)
+    .set({
+      content: imageText,
+      metadata: null,
+    })
+    .where(eq(bookmarkAssets.id, bookmark.id));
+  return true;
 }
 
-async function preProcessPDF(
+async function extractAndSavePDFText(
   jobId: string,
   asset: Buffer,
-): Promise<{ content: string; metadata: string | null } | undefined> {
+  bookmark: NonNullable<Awaited<ReturnType<typeof getBookmark>>>,
+  isFixMode: boolean,
+): Promise<boolean> {
+  {
+    const alreadyHasText = !!bookmark.asset.content;
+    if (alreadyHasText && isFixMode) {
+      logger.info(
+        `[assetPreprocessing][${jobId}] Skipping PDF text extraction as it's already been extracted.`,
+      );
+      return false;
+    }
+  }
+  logger.info(
+    `[assetPreprocessing][${jobId}] Attempting to extract text from pdf.`,
+  );
   const pdfParse = await readPDFText(asset);
   if (!pdfParse?.text) {
     throw new Error(
@@ -120,13 +235,28 @@ async function preProcessPDF(
   logger.info(
     `[assetPreprocessing][${jobId}] Extracted ${pdfParse.text.length} characters from pdf.`,
   );
-  return {
-    content: pdfParse.text,
-    metadata: pdfParse.metadata ? JSON.stringify(pdfParse.metadata) : null,
-  };
+  await db
+    .update(bookmarkAssets)
+    .set({
+      content: pdfParse.text,
+      metadata: pdfParse.metadata ? JSON.stringify(pdfParse.metadata) : null,
+    })
+    .where(eq(bookmarkAssets.id, bookmark.id));
+  return true;
+}
+
+async function getBookmark(bookmarkId: string) {
+  return db.query.bookmarks.findFirst({
+    where: eq(bookmarks.id, bookmarkId),
+    with: {
+      asset: true,
+      assets: true,
+    },
+  });
 }
 
 async function run(req: DequeuedJob<AssetPreprocessingRequest>) {
+  const isFixMode = req.data.fixMode;
   const jobId = req.id;
   const bookmarkId = req.data.bookmarkId;
 
@@ -134,6 +264,7 @@ async function run(req: DequeuedJob<AssetPreprocessingRequest>) {
     where: eq(bookmarks.id, bookmarkId),
     with: {
       asset: true,
+      assets: true,
     },
   });
 
@@ -162,15 +293,29 @@ async function run(req: DequeuedJob<AssetPreprocessingRequest>) {
     );
   }
 
-  let result: { content: string; metadata: string | null } | undefined =
-    undefined;
-
+  let anythingChanged = false;
   switch (bookmark.asset.assetType) {
     case "image":
-      result = await preprocessImage(jobId, asset);
+      anythingChanged ||= await extractAndSaveImageText(
+        jobId,
+        asset,
+        bookmark,
+        isFixMode,
+      );
       break;
     case "pdf":
-      result = await preProcessPDF(jobId, asset);
+      anythingChanged ||= await extractAndSavePDFText(
+        jobId,
+        asset,
+        bookmark,
+        isFixMode,
+      );
+      anythingChanged ||= await extractAndSavePDFScreenshot(
+        jobId,
+        asset,
+        bookmark,
+        isFixMode,
+      );
       break;
     default:
       throw new Error(
@@ -178,20 +323,12 @@ async function run(req: DequeuedJob<AssetPreprocessingRequest>) {
       );
   }
 
-  if (result) {
-    await db
-      .update(bookmarkAssets)
-      .set({
-        content: result.content,
-        metadata: result.metadata,
-      })
-      .where(eq(bookmarkAssets.id, bookmarkId));
-  }
-
-  await OpenAIQueue.enqueue({
-    bookmarkId,
-  });
+  if (anythingChanged) {
+    await OpenAIQueue.enqueue({
+      bookmarkId,
+    });
 
-  // Update the search index
-  await triggerSearchReindex(bookmarkId);
+    // Update the search index
+    await triggerSearchReindex(bookmarkId);
+  }
 }
diff --git a/apps/workers/crawlerWorker.ts b/apps/workers/crawlerWorker.ts
index 7611494e..17dba443 100644
--- a/apps/workers/crawlerWorker.ts
+++ b/apps/workers/crawlerWorker.ts
@@ -592,6 +592,7 @@ async function handleAsAssetBookmark(
   });
   await AssetPreprocessingQueue.enqueue({
     bookmarkId,
+    fixMode: false,
   });
 }
 
diff --git a/apps/workers/package.json b/apps/workers/package.json
index ebcae757..122c7cb1 100644
--- a/apps/workers/package.json
+++ b/apps/workers/package.json
@@ -30,7 +30,8 @@
     "metascraper-url": "^5.45.22",
     "node-cron": "^3.0.3",
     "node-fetch": "^3.3.2",
-    "pdf2json": "^3.0.5",
+    "pdf2json": "^3.1.5",
+    "pdf2pic": "^3.1.3",
     "pdfjs-dist": "^4.0.379",
     "puppeteer": "^22.0.0",
     "puppeteer-extra": "^3.3.6",
@@ -65,4 +66,4 @@
     ]
   },
   "prettier": "@hoarder/prettier-config"
-}
+}
\ No newline at end of file