aboutsummaryrefslogtreecommitdiffstats
path: root/apps
diff options
context:
space:
mode:
authorMohamed Bassem <me@mbassem.com>2024-11-03 17:09:47 +0000
committerMohamed Bassem <me@mbassem.com>2024-11-03 17:09:47 +0000
commitcf1a25131fd45ab7c9a72b837be525c24457cd8b (patch)
treecb8b3d4a57a4ce06e500e4c7ceea43924d64a5d4 /apps
parent2efc7c8c01866fafad5322fb8783d94821e32ff1 (diff)
downloadkarakeep-cf1a25131fd45ab7c9a72b837be525c24457cd8b.tar.zst
feature: Add support for subscribing to RSS feeds. Fixes #202
Diffstat (limited to 'apps')
-rw-r--r--apps/web/app/settings/feeds/page.tsx5
-rw-r--r--apps/web/components/settings/FeedSettings.tsx398
-rw-r--r--apps/web/components/settings/sidebar/items.tsx14
-rw-r--r--apps/workers/feedWorker.ts149
-rw-r--r--apps/workers/index.ts8
-rw-r--r--apps/workers/package.json2
-rw-r--r--apps/workers/trpc.ts33
7 files changed, 606 insertions, 3 deletions
diff --git a/apps/web/app/settings/feeds/page.tsx b/apps/web/app/settings/feeds/page.tsx
new file mode 100644
index 00000000..d2d1f182
--- /dev/null
+++ b/apps/web/app/settings/feeds/page.tsx
@@ -0,0 +1,5 @@
+import FeedSettings from "@/components/settings/FeedSettings";
+
+export default function FeedSettingsPage() {
+ return <FeedSettings />;
+}
diff --git a/apps/web/components/settings/FeedSettings.tsx b/apps/web/components/settings/FeedSettings.tsx
new file mode 100644
index 00000000..9f2dbda9
--- /dev/null
+++ b/apps/web/components/settings/FeedSettings.tsx
@@ -0,0 +1,398 @@
+"use client";
+
+import React from "react";
+import { ActionButton } from "@/components/ui/action-button";
+import {
+ Form,
+ FormControl,
+ FormField,
+ FormItem,
+ FormLabel,
+ FormMessage,
+} from "@/components/ui/form";
+import { FullPageSpinner } from "@/components/ui/full-page-spinner";
+import { Input } from "@/components/ui/input";
+import { toast } from "@/components/ui/use-toast";
+import { api } from "@/lib/trpc";
+import { zodResolver } from "@hookform/resolvers/zod";
+import {
+ ArrowDownToLine,
+ CheckCircle,
+ CircleDashed,
+ Edit,
+ Plus,
+ Save,
+ Trash2,
+ XCircle,
+} from "lucide-react";
+import { useForm } from "react-hook-form";
+import { z } from "zod";
+
+import {
+ ZFeed,
+ zNewFeedSchema,
+ zUpdateFeedSchema,
+} from "@hoarder/shared/types/feeds";
+
+import ActionConfirmingDialog from "../ui/action-confirming-dialog";
+import { Button } from "../ui/button";
+import {
+ Dialog,
+ DialogClose,
+ DialogContent,
+ DialogFooter,
+ DialogHeader,
+ DialogTitle,
+ DialogTrigger,
+} from "../ui/dialog";
+import {
+ Table,
+ TableBody,
+ TableCell,
+ TableHead,
+ TableHeader,
+ TableRow,
+} from "../ui/table";
+
+export function FeedsEditorDialog() {
+ const [open, setOpen] = React.useState(false);
+ const apiUtils = api.useUtils();
+
+ const form = useForm<z.infer<typeof zNewFeedSchema>>({
+ resolver: zodResolver(zNewFeedSchema),
+ defaultValues: {
+ name: "",
+ url: "",
+ },
+ });
+
+ React.useEffect(() => {
+ if (open) {
+ form.reset();
+ }
+ }, [open]);
+
+ const { mutateAsync: createFeed, isPending: isCreating } =
+ api.feeds.create.useMutation({
+ onSuccess: () => {
+ toast({
+ description: "Feed has been created!",
+ });
+ apiUtils.feeds.list.invalidate();
+ setOpen(false);
+ },
+ });
+
+ return (
+ <Dialog open={open} onOpenChange={setOpen}>
+ <DialogTrigger asChild>
+ <Button>
+ <Plus className="mr-2 size-4" />
+ Add a Subscription
+ </Button>
+ </DialogTrigger>
+ <DialogContent>
+ <DialogHeader>
+ <DialogTitle>Subscribe to a new Feed</DialogTitle>
+ </DialogHeader>
+ <Form {...form}>
+ <form
+ className="flex flex-col gap-3"
+ onSubmit={form.handleSubmit(async (value) => {
+ await createFeed(value);
+ form.resetField("name");
+ form.resetField("url");
+ })}
+ >
+ <FormField
+ control={form.control}
+ name="name"
+ render={({ field }) => {
+ return (
+ <FormItem className="flex-1">
+ <FormLabel>Name</FormLabel>
+ <FormControl>
+ <Input placeholder="Feed Name" type="text" {...field} />
+ </FormControl>
+ <FormMessage />
+ </FormItem>
+ );
+ }}
+ />
+ <FormField
+ control={form.control}
+ name="url"
+ render={({ field }) => {
+ return (
+ <FormItem className="flex-1">
+ <FormLabel>URL</FormLabel>
+ <FormControl>
+ <Input placeholder="Feed URL" type="text" {...field} />
+ </FormControl>
+ <FormMessage />
+ </FormItem>
+ );
+ }}
+ />
+ </form>
+ </Form>
+ <DialogFooter>
+ <DialogClose asChild>
+ <Button type="button" variant="secondary">
+ Close
+ </Button>
+ </DialogClose>
+ <ActionButton
+ onClick={form.handleSubmit(async (value) => {
+ await createFeed(value);
+ })}
+ loading={isCreating}
+ variant="default"
+ className="items-center"
+ >
+ <Plus className="mr-2 size-4" />
+ Add
+ </ActionButton>
+ </DialogFooter>
+ </DialogContent>
+ </Dialog>
+ );
+}
+
+export function EditFeedDialog({ feed }: { feed: ZFeed }) {
+ const apiUtils = api.useUtils();
+ const [open, setOpen] = React.useState(false);
+ React.useEffect(() => {
+ if (open) {
+ form.reset({
+ feedId: feed.id,
+ name: feed.name,
+ url: feed.url,
+ });
+ }
+ }, [open]);
+ const { mutateAsync: updateFeed, isPending: isUpdating } =
+ api.feeds.update.useMutation({
+ onSuccess: () => {
+ toast({
+ description: "Feed has been updated!",
+ });
+ setOpen(false);
+ apiUtils.feeds.list.invalidate();
+ },
+ });
+ const form = useForm<z.infer<typeof zUpdateFeedSchema>>({
+ resolver: zodResolver(zUpdateFeedSchema),
+ defaultValues: {
+ feedId: feed.id,
+ name: feed.name,
+ url: feed.url,
+ },
+ });
+ return (
+ <Dialog open={open} onOpenChange={setOpen}>
+ <DialogTrigger asChild>
+ <Button variant="secondary">
+ <Edit className="mr-2 size-4" />
+ Edit
+ </Button>
+ </DialogTrigger>
+ <DialogContent>
+ <DialogHeader>
+ <DialogTitle>Edit Feed</DialogTitle>
+ </DialogHeader>
+
+ <Form {...form}>
+ <form
+ className="flex flex-col gap-3"
+ onSubmit={form.handleSubmit(async (value) => {
+ await updateFeed(value);
+ })}
+ >
+ <FormField
+ control={form.control}
+ name="feedId"
+ render={({ field }) => {
+ return (
+ <FormItem className="hidden">
+ <FormControl>
+ <Input type="hidden" {...field} />
+ </FormControl>
+ <FormMessage />
+ </FormItem>
+ );
+ }}
+ />
+ <FormField
+ control={form.control}
+ name="name"
+ render={({ field }) => {
+ return (
+ <FormItem className="flex-1">
+ <FormLabel>Name</FormLabel>
+ <FormControl>
+ <Input placeholder="Feed name" type="text" {...field} />
+ </FormControl>
+ <FormMessage />
+ </FormItem>
+ );
+ }}
+ />
+ <FormField
+ control={form.control}
+ name="url"
+ render={({ field }) => {
+ return (
+ <FormItem className="flex-1">
+ <FormLabel>URL</FormLabel>
+ <FormControl>
+ <Input placeholder="Feed url" type="text" {...field} />
+ </FormControl>
+ <FormMessage />
+ </FormItem>
+ );
+ }}
+ />
+ </form>
+ </Form>
+ <DialogFooter>
+ <DialogClose asChild>
+ <Button type="button" variant="secondary">
+ Close
+ </Button>
+ </DialogClose>
+ <ActionButton
+ loading={isUpdating}
+ onClick={form.handleSubmit(async (value) => {
+ await updateFeed(value);
+ })}
+ type="submit"
+ className="items-center"
+ >
+ <Save className="mr-2 size-4" />
+ Save
+ </ActionButton>
+ </DialogFooter>
+ </DialogContent>
+ </Dialog>
+ );
+}
+
+export function FeedRow({ feed }: { feed: ZFeed }) {
+ const apiUtils = api.useUtils();
+ const { mutate: deleteFeed, isPending: isDeleting } =
+ api.feeds.delete.useMutation({
+ onSuccess: () => {
+ toast({
+ description: "Feed has been deleted!",
+ });
+ apiUtils.feeds.list.invalidate();
+ },
+ });
+
+ const { mutate: fetchNow, isPending: isFetching } =
+ api.feeds.fetchNow.useMutation({
+ onSuccess: () => {
+ toast({
+ description: "Feed fetch has been enqueued!",
+ });
+ apiUtils.feeds.list.invalidate();
+ },
+ });
+
+ return (
+ <TableRow>
+ <TableCell>{feed.name}</TableCell>
+ <TableCell>{feed.url}</TableCell>
+ <TableCell>{feed.lastFetchedAt?.toLocaleString()}</TableCell>
+ <TableCell>
+ {feed.lastFetchedStatus === "success" ? (
+ <span title="Successful">
+ <CheckCircle />
+ </span>
+ ) : feed.lastFetchedStatus === "failure" ? (
+ <span title="Failed">
+ <XCircle />
+ </span>
+ ) : (
+ <span title="Pending">
+ <CircleDashed name="Pending" />
+ </span>
+ )}
+ </TableCell>
+ <TableCell className="flex items-center gap-2">
+ <EditFeedDialog feed={feed} />
+ <ActionButton
+ loading={isFetching}
+ variant="secondary"
+ className="items-center"
+ onClick={() => fetchNow({ feedId: feed.id })}
+ >
+ <ArrowDownToLine className="mr-2 size-4" />
+ Fetch Now
+ </ActionButton>
+ <ActionConfirmingDialog
+ title={`Delete Feed "${feed.name}"?`}
+ description={`Are you sure you want to delete the feed "${feed.name}"?`}
+ actionButton={() => (
+ <ActionButton
+ loading={isDeleting}
+ variant="destructive"
+ onClick={() => deleteFeed({ feedId: feed.id })}
+ className="items-center"
+ type="button"
+ >
+ <Trash2 className="mr-2 size-4" />
+ Delete
+ </ActionButton>
+ )}
+ >
+ <Button variant="destructive" disabled={isDeleting}>
+ <Trash2 className="mr-2 size-4" />
+ Delete
+ </Button>
+ </ActionConfirmingDialog>
+ </TableCell>
+ </TableRow>
+ );
+}
+
+export default function FeedSettings() {
+ const { data: feeds, isLoading } = api.feeds.list.useQuery();
+ return (
+ <>
+ <div className="rounded-md border bg-background p-4">
+ <div className="flex flex-col gap-2">
+ <div className="flex items-center justify-between">
+ <div className="mb-2 text-lg font-medium">RSS Subscriptions</div>
+ <FeedsEditorDialog />
+ </div>
+ {isLoading && <FullPageSpinner />}
+ {feeds && feeds.feeds.length == 0 && (
+ <p className="rounded-md bg-muted p-2 text-sm text-muted-foreground">
+ You don&apos;t have any RSS subscriptions yet.
+ </p>
+ )}
+ {feeds && feeds.feeds.length > 0 && (
+ <Table>
+ <TableHeader>
+ <TableRow>
+ <TableHead>Name</TableHead>
+ <TableHead>URL</TableHead>
+ <TableHead>Last Fetch</TableHead>
+ <TableHead>Last Status</TableHead>
+ <TableHead>Actions</TableHead>
+ </TableRow>
+ </TableHeader>
+ <TableBody>
+ {feeds.feeds.map((feed) => (
+ <FeedRow key={feed.id} feed={feed} />
+ ))}
+ </TableBody>
+ </Table>
+ )}
+ </div>
+ </div>
+ </>
+ );
+}
diff --git a/apps/web/components/settings/sidebar/items.tsx b/apps/web/components/settings/sidebar/items.tsx
index 999825db..047ee233 100644
--- a/apps/web/components/settings/sidebar/items.tsx
+++ b/apps/web/components/settings/sidebar/items.tsx
@@ -1,5 +1,12 @@
import React from "react";
-import { ArrowLeft, Download, KeyRound, Sparkles, User } from "lucide-react";
+import {
+ ArrowLeft,
+ Download,
+ KeyRound,
+ Rss,
+ Sparkles,
+ User,
+} from "lucide-react";
export const settingsSidebarItems: {
name: string;
@@ -22,6 +29,11 @@ export const settingsSidebarItems: {
path: "/settings/ai",
},
{
+ name: "RSS Subscriptions",
+ icon: <Rss size={18} />,
+ path: "/settings/feeds",
+ },
+ {
name: "Import / Export",
icon: <Download size={18} />,
path: "/settings/import",
diff --git a/apps/workers/feedWorker.ts b/apps/workers/feedWorker.ts
new file mode 100644
index 00000000..1bd24641
--- /dev/null
+++ b/apps/workers/feedWorker.ts
@@ -0,0 +1,149 @@
+import { and, eq, inArray } from "drizzle-orm";
+import { DequeuedJob, Runner } from "liteque";
+import Parser from "rss-parser";
+import { buildImpersonatingTRPCClient } from "trpc";
+
+import type { ZFeedRequestSchema } from "@hoarder/shared/queues";
+import { db } from "@hoarder/db";
+import { rssFeedImportsTable, rssFeedsTable } from "@hoarder/db/schema";
+import logger from "@hoarder/shared/logger";
+import { FeedQueue } from "@hoarder/shared/queues";
+import { BookmarkTypes } from "@hoarder/shared/types/bookmarks";
+
+export class FeedWorker {
+ static build() {
+ logger.info("Starting feed worker ...");
+ const worker = new Runner<ZFeedRequestSchema>(
+ FeedQueue,
+ {
+ run: run,
+ onComplete: async (job) => {
+ const jobId = job.id;
+ logger.info(`[feed][${jobId}] Completed successfully`);
+ await db
+ .update(rssFeedsTable)
+ .set({ lastFetchedStatus: "success", lastFetchedAt: new Date() })
+ .where(eq(rssFeedsTable.id, job.data?.feedId));
+ },
+ onError: async (job) => {
+ const jobId = job.id;
+ logger.error(
+ `[feed][${jobId}] Feed fetch job failed: ${job.error}\n${job.error.stack}`,
+ );
+ if (job.data) {
+ await db
+ .update(rssFeedsTable)
+ .set({ lastFetchedStatus: "failure", lastFetchedAt: new Date() })
+ .where(eq(rssFeedsTable.id, job.data?.feedId));
+ }
+ },
+ },
+ {
+ concurrency: 1,
+ pollIntervalMs: 1000,
+ timeoutSecs: 30,
+ },
+ );
+
+ return worker;
+ }
+}
+
+async function run(req: DequeuedJob<ZFeedRequestSchema>) {
+ const jobId = req.id;
+ const feed = await db.query.rssFeedsTable.findFirst({
+ where: eq(rssFeedsTable.id, req.data.feedId),
+ });
+ if (!feed) {
+ throw new Error(
+ `[feed][${jobId}] Feed with id ${req.data.feedId} not found`,
+ );
+ }
+
+ const response = await fetch(feed.url, {
+ signal: AbortSignal.timeout(5000),
+ });
+ const contentType = response.headers.get("content-type");
+ if (!contentType || !contentType.includes("application/xml")) {
+ throw new Error(
+ `[feed][${jobId}] Feed with id ${req.data.feedId} is not a valid RSS feed`,
+ );
+ }
+ const xmlData = await response.text();
+
+ logger.info(
+ `[feed][${jobId}] Successfully fetched feed "${feed.name}" (${feed.id}) ...`,
+ );
+
+ const parser = new Parser();
+ const feedData = await parser.parseString(xmlData);
+
+ logger.info(
+ `[feed][${jobId}] Found ${feedData.items.length} entries in feed "${feed.name}" (${feed.id}) ...`,
+ );
+
+ if (feedData.items.length === 0) {
+ logger.info(`[feed][${jobId}] No entries found.`);
+ return;
+ }
+
+ const exitingEntries = await db.query.rssFeedImportsTable.findMany({
+ where: and(
+ eq(rssFeedImportsTable.rssFeedId, feed.id),
+ inArray(
+ rssFeedImportsTable.entryId,
+ feedData.items
+ .map((item) => item.guid)
+ .filter((id): id is string => !!id),
+ ),
+ ),
+ });
+
+ const newEntries = feedData.items.filter(
+ (item) =>
+ !exitingEntries.some((entry) => entry.entryId === item.guid) && item.link,
+ );
+
+ if (newEntries.length === 0) {
+ logger.info(
+ `[feed][${jobId}] No new entries found in feed "${feed.name}" (${feed.id}).`,
+ );
+ return;
+ }
+
+ logger.info(
+ `[feed][${jobId}] Found ${newEntries.length} new entries in feed "${feed.name}" (${feed.id}) ...`,
+ );
+
+ const trpcClient = await buildImpersonatingTRPCClient(feed.userId);
+
+ const createdBookmarks = await Promise.allSettled(
+ newEntries.map((item) =>
+ trpcClient.bookmarks.createBookmark({
+ type: BookmarkTypes.LINK,
+ url: item.link!,
+ }),
+ ),
+ );
+
+ // It's ok if this is not transactional as the bookmarks will get linked in the next iteration.
+ await db
+ .insert(rssFeedImportsTable)
+ .values(
+ newEntries.map((item, idx) => {
+ const b = createdBookmarks[idx];
+ return {
+ entryId: item.guid!,
+ bookmarkId: b.status === "fulfilled" ? b.value.id : null,
+ rssFeedId: feed.id,
+ };
+ }),
+ )
+ .onConflictDoNothing();
+
+ logger.info(
+ `[feed][${jobId}] Successfully imported ${newEntries.length} new enteries from feed "${feed.name}" (${feed.id}).`,
+ );
+
+ return Promise.resolve();
+}
diff --git a/apps/workers/index.ts b/apps/workers/index.ts
index 3b5896e4..c8978adc 100644
--- a/apps/workers/index.ts
+++ b/apps/workers/index.ts
@@ -1,5 +1,6 @@
import "dotenv/config";
+import { FeedWorker } from "feedWorker";
import { TidyAssetsWorker } from "tidyAssetsWorker";
import serverConfig from "@hoarder/shared/config";
@@ -16,12 +17,13 @@ async function main() {
logger.info(`Workers version: ${serverConfig.serverVersion ?? "not set"}`);
runQueueDBMigrations();
- const [crawler, openai, search, tidyAssets, video] = [
+ const [crawler, openai, search, tidyAssets, video, feed] = [
await CrawlerWorker.build(),
OpenAiWorker.build(),
SearchIndexingWorker.build(),
TidyAssetsWorker.build(),
VideoWorker.build(),
+ FeedWorker.build(),
];
await Promise.any([
@@ -31,11 +33,12 @@ async function main() {
search.run(),
tidyAssets.run(),
video.run(),
+ feed.run(),
]),
shutdownPromise,
]);
logger.info(
- "Shutting down crawler, openai, tidyAssets, video and search workers ...",
+ "Shutting down crawler, openai, tidyAssets, video, feed and search workers ...",
);
crawler.stop();
@@ -43,6 +46,7 @@ async function main() {
search.stop();
tidyAssets.stop();
video.stop();
+ feed.stop();
}
main();
diff --git a/apps/workers/package.json b/apps/workers/package.json
index 7f64e715..a7579319 100644
--- a/apps/workers/package.json
+++ b/apps/workers/package.json
@@ -6,6 +6,7 @@
"dependencies": {
"@hoarder/db": "workspace:^0.1.0",
"@hoarder/shared": "workspace:^0.1.0",
+ "@hoarder/trpc": "workspace:^0.1.0",
"@hoarder/tsconfig": "workspace:^0.1.0",
"@mozilla/readability": "^0.5.0",
"@tsconfig/node21": "^21.0.1",
@@ -32,6 +33,7 @@
"puppeteer-extra": "^3.3.6",
"puppeteer-extra-plugin-adblocker": "^2.13.6",
"puppeteer-extra-plugin-stealth": "^2.11.2",
+ "rss-parser": "^3.13.0",
"tesseract.js": "^5.1.1",
"tsx": "^4.7.1",
"typescript": "^5.3.3",
diff --git a/apps/workers/trpc.ts b/apps/workers/trpc.ts
new file mode 100644
index 00000000..cd2e4c99
--- /dev/null
+++ b/apps/workers/trpc.ts
@@ -0,0 +1,33 @@
+import { eq } from "drizzle-orm";
+
+import { db } from "@hoarder/db";
+import { users } from "@hoarder/db/schema";
+import { createCallerFactory } from "@hoarder/trpc";
+import { appRouter } from "@hoarder/trpc/routers/_app";
+
+/**
+ * This is only safe to use in the context of a worker.
+ */
+export async function buildImpersonatingTRPCClient(userId: string) {
+ const createCaller = createCallerFactory(appRouter);
+
+ const user = await db.query.users.findFirst({
+ where: eq(users.id, userId),
+ });
+ if (!user) {
+ throw new Error("User not found");
+ }
+
+ return createCaller({
+ user: {
+ id: user.id,
+ name: user.name,
+ email: user.email,
+ role: user.role,
+ },
+ db,
+ req: {
+ ip: null,
+ },
+ });
+}