feature(web): Add ability to manually trigger full page archives. Fixes

hoarder-app#398 (hoarder-app#418) * [Feature Request] Ability to select what to "crawl full page archive" hoarder-app#398 Added the ability to start a full page crawl for links and also in bulk operations added the ability to refresh links as a bulk operation as well * minor icon and wording changes --------- Co-authored-by: MohamedBassem <me@mbassem.com>
kamtschatka · Sep 30, 2024 · 8b69cdd · 8b69cdd
1 parent 5281531
commit 8b69cdd
Show file tree

Hide file tree

Showing 5 changed files with 89 additions and 6 deletions.
diff --git a/apps/web/components/dashboard/BulkBookmarksAction.tsx b/apps/web/components/dashboard/BulkBookmarksAction.tsx
@@ -8,10 +8,21 @@ import {
 import ActionConfirmingDialog from "@/components/ui/action-confirming-dialog";
 import { useToast } from "@/components/ui/use-toast";
 import useBulkActionsStore from "@/lib/bulkActions";
-import { CheckCheck, Hash, Link, List, Pencil, Trash2, X } from "lucide-react";
+import {
+  CheckCheck,
+  FileDown,
+  Hash,
+  Link,
+  List,
+  Pencil,
+  RotateCw,
+  Trash2,
+  X,
+} from "lucide-react";
 
 import {
   useDeleteBookmark,
+  useRecrawlBookmark,
   useUpdateBookmark,
 } from "@hoarder/shared-react/hooks/bookmarks";
 import { BookmarkTypes } from "@hoarder/shared/types/bookmarks";
@@ -63,11 +74,35 @@ export default function BulkBookmarksAction() {
     onError,
   });
 
+  const recrawlBookmarkMutator = useRecrawlBookmark({
+    onSuccess: () => {
+      setIsBulkEditEnabled(false);
+    },
+    onError,
+  });
+
   interface UpdateBookmarkProps {
     favourited?: boolean;
     archived?: boolean;
   }
 
+  const recrawlBookmarks = async (archiveFullPage: boolean) => {
+    const links = selectedBookmarks.filter(
+      (item) => item.content.type === BookmarkTypes.LINK,
+    );
+    await Promise.all(
+      links.map((item) =>
+        recrawlBookmarkMutator.mutateAsync({
+          bookmarkId: item.id,
+          archiveFullPage,
+        }),
+      ),
+    );
+    toast({
+      description: `${links.length} bookmarks will be ${archiveFullPage ? "re-crawled and archived!" : "refreshed!"}`,
+    });
+  };
+
   function isClipboardAvailable() {
     if (typeof window === "undefined") {
       return false;
@@ -172,6 +207,20 @@ export default function BulkBookmarksAction() {
       isPending: updateBookmarkMutator.isPending,
       hidden: !isBulkEditEnabled,
     },
+    {
+      name: "Download Full Page Archive",
+      icon: <FileDown size={18} />,
+      action: () => recrawlBookmarks(true),
+      isPending: recrawlBookmarkMutator.isPending,
+      hidden: !isBulkEditEnabled,
+    },
+    {
+      name: "Refresh",
+      icon: <RotateCw size={18} />,
+      action: () => recrawlBookmarks(false),
+      isPending: recrawlBookmarkMutator.isPending,
+      hidden: !isBulkEditEnabled,
+    },
     {
       name: "Delete",
       icon: <Trash2 size={18} color="red" />,

diff --git a/apps/web/components/dashboard/bookmarks/BookmarkOptions.tsx b/apps/web/components/dashboard/bookmarks/BookmarkOptions.tsx
@@ -11,6 +11,7 @@ import {
 import { useToast } from "@/components/ui/use-toast";
 import { useClientConfig } from "@/lib/clientConfig";
 import {
+  FileDown,
   Link,
   List,
   ListX,
@@ -88,6 +89,15 @@ export default function BookmarkOptions({ bookmark }: { bookmark: ZBookmark }) {
     onError,
   });
 
+  const fullPageArchiveBookmarkMutator = useRecrawlBookmark({
+    onSuccess: () => {
+      toast({
+        description: "Full Page Archive creation has been triggered",
+      });
+    },
+    onError,
+  });
+
   const removeFromListMutator = useRemoveBookmarkFromList({
     onSuccess: () => {
       toast({
@@ -152,6 +162,21 @@ export default function BookmarkOptions({ bookmark }: { bookmark: ZBookmark }) {
             />
             <span>{bookmark.archived ? "Un-archive" : "Archive"}</span>
           </DropdownMenuItem>
+
+          {bookmark.content.type === BookmarkTypes.LINK && (
+            <DropdownMenuItem
+              onClick={() => {
+                fullPageArchiveBookmarkMutator.mutate({
+                  bookmarkId: bookmark.id,
+                  archiveFullPage: true,
+                });
+              }}
+            >
+              <FileDown className="mr-2 size-4" />
+              <span>Download Full Page Archive</span>
+            </DropdownMenuItem>
+          )}
+
           {bookmark.content.type === BookmarkTypes.LINK && (
             <DropdownMenuItem
               onClick={() => {

diff --git a/apps/workers/crawlerWorker.ts b/apps/workers/crawlerWorker.ts
@@ -214,7 +214,7 @@ async function getBookmarkDetails(bookmarkId: string) {
   });
 
   if (!bookmark || !bookmark.link) {
-    throw new Error("The bookmark either doesn't exist or not a link");
+    throw new Error("The bookmark either doesn't exist or is not a link");
   }
   return {
     url: bookmark.link.url,
@@ -519,6 +519,7 @@ async function crawlAndParseUrl(
   oldScreenshotAssetId: string | undefined,
   oldImageAssetId: string | undefined,
   oldFullPageArchiveAssetId: string | undefined,
+  archiveFullPage: boolean,
 ) {
   const {
     htmlContent,
@@ -578,7 +579,7 @@ async function crawlAndParseUrl(
   ]);
 
   return async () => {
-    if (serverConfig.crawler.fullPageArchive) {
+    if (serverConfig.crawler.fullPageArchive || archiveFullPage) {
       const fullPageArchiveAssetId = await archiveWebpage(
         htmlContent,
         browserUrl,
@@ -615,7 +616,7 @@ async function runCrawler(job: DequeuedJob<ZCrawlLinkRequest>) {
     return;
   }
 
-  const { bookmarkId } = request.data;
+  const { bookmarkId, archiveFullPage } = request.data;
   const {
     url,
     userId,
@@ -654,6 +655,7 @@ async function runCrawler(job: DequeuedJob<ZCrawlLinkRequest>) {
       oldScreenshotAssetId,
       oldImageAssetId,
       oldFullPageArchiveAssetId,
+      archiveFullPage,
     );
   }
 

diff --git a/packages/shared/queues.ts b/packages/shared/queues.ts
@@ -17,8 +17,9 @@ export function runQueueDBMigrations() {
 export const zCrawlLinkRequestSchema = z.object({
   bookmarkId: z.string(),
   runInference: z.boolean().optional(),
+  archiveFullPage: z.boolean().optional().default(false),
 });
-export type ZCrawlLinkRequest = z.infer<typeof zCrawlLinkRequestSchema>;
+export type ZCrawlLinkRequest = z.input<typeof zCrawlLinkRequestSchema>;
 
 export const LinkCrawlerQueue = new SqliteQueue<ZCrawlLinkRequest>(
   "link_crawler_queue",

diff --git a/packages/trpc/routers/bookmarks.ts b/packages/trpc/routers/bookmarks.ts
@@ -426,11 +426,17 @@ export const bookmarksAppRouter = router({
       }
     }),
   recrawlBookmark: authedProcedure
-    .input(z.object({ bookmarkId: z.string() }))
+    .input(
+      z.object({
+        bookmarkId: z.string(),
+        archiveFullPage: z.boolean().optional().default(false),
+      }),
+    )
     .use(ensureBookmarkOwnership)
     .mutation(async ({ input }) => {
       await LinkCrawlerQueue.enqueue({
         bookmarkId: input.bookmarkId,
+        archiveFullPage: input.archiveFullPage,
       });
     }),
   getBookmark: authedProcedure