apunkt/SurfSense
1
0
Fork 0
mirror of https://github.com/MODSetter/SurfSense.git synced 2026-06-08 20:25:19 +02:00
Code Issues Projects Releases Packages Wiki Activity

Renaming resources

Browse source
This commit is contained in:
samkul-swe 2025-11-22 19:19:00 -08:00
parent 6d19e0fad8
commit 121e2f0c0e
24 changed files with 117 additions and 273 deletions
Download patch file Download diff file Expand all files Collapse all files

47
surfsense_web/app/dashboard/[search_space_id]/connectors/[connector_id]/edit/page.tsx
View file

@ -18,7 +18,16 @@ import {
CardHeader,
CardTitle,
} from "@/components/ui/card";
import { Form } from "@/components/ui/form";
import {
Form,
FormControl,
FormDescription,
FormField,
FormItem,
FormLabel,
FormMessage,
} from "@/components/ui/form";
import { Textarea } from "@/components/ui/textarea";
import { getConnectorIcon } from "@/contracts/enums/connectorIcons";
import { useConnectorEditPage } from "@/hooks/use-connector-edit-page";
// Import Utils, Types, Hook, and Components
@ -285,13 +294,35 @@ export default function EditConnectorPage() {
{/* == Webcrawler == */}
{connector.connector_type === "WEBCRAWLER_CONNECTOR" && (
<EditSimpleTokenForm
control={editForm.control}
fieldName="FIRECRAWL_API_KEY"
fieldLabel="Firecrawl API Key (Optional)"
fieldDescription="Add a Firecrawl API key for enhanced crawling capabilities. If not provided, will use AsyncChromiumLoader as fallback."
placeholder="fc-xxxxxxxxxxxxx"
/>
<div className="space-y-4">
<EditSimpleTokenForm
control={editForm.control}
fieldName="FIRECRAWL_API_KEY"
fieldLabel="Firecrawl API Key (Optional)"
fieldDescription="Add a Firecrawl API key for enhanced crawling capabilities. If not provided, will use AsyncChromiumLoader as fallback."
placeholder="fc-xxxxxxxxxxxxx"
/>
<FormField
control={editForm.control}
name="INITIAL_URLS"
render={({ field }) => (
<FormItem>
<FormLabel>URLs to Crawl</FormLabel>
<FormControl>
<Textarea
placeholder="https://example.com&#10;https://docs.example.com&#10;https://blog.example.com"
className="min-h-[150px] font-mono text-sm"
{...field}
/>
</FormControl>
<FormDescription>
Enter URLs to crawl (one per line). These URLs will be indexed when you trigger indexing.
</FormDescription>
<FormMessage />
</FormItem>
)}
/>
</div>
)}
</CardContent>

2
surfsense_web/app/dashboard/[search_space_id]/connectors/[connector_id]/page.tsx
View file

@ -55,7 +55,7 @@ const getConnectorTypeDisplay = (type: string): string => {
AIRTABLE_CONNECTOR: "Airtable Connector",
LUMA_CONNECTOR: "Luma Connector",
ELASTICSEARCH_CONNECTOR: "Elasticsearch Connector",
WEBCRAWLER_CONNECTOR: "Web Crawler Connector",
WEBCRAWLER_CONNECTOR: "Web Page Connector",
// Add other connector types here as needed
};
return typeMap[type] || type;

10
surfsense_web/app/dashboard/[search_space_id]/connectors/add/webcrawler-connector/page.tsx
View file

@ -64,7 +64,7 @@ export default function WebcrawlerConnectorPage() {
const form = useForm<WebcrawlerConnectorFormValues>({
resolver: zodResolver(webcrawlerConnectorFormSchema),
defaultValues: {
name: "Web Crawler",
name: "Web Pages",
api_key: "",
initial_urls: "",
},
@ -150,7 +150,7 @@ export default function WebcrawlerConnectorPage() {
{getConnectorIcon(EnumConnectorName.WEBCRAWLER_CONNECTOR, "h-6 w-6")}
</div>
<div>
<h1 className="text-3xl font-bold tracking-tight">Connect Web Crawler</h1>
<h1 className="text-3xl font-bold tracking-tight">Connect Web Pages</h1>
<p className="text-muted-foreground">Crawl and index web pages for search.</p>
</div>
</div>
@ -160,9 +160,9 @@ export default function WebcrawlerConnectorPage() {
{!doesConnectorExist ? (
<Card>
<CardHeader>
<CardTitle>Set Up Web Crawler</CardTitle>
<CardTitle>Set Up Web Page crawler</CardTitle>
<CardDescription>
Configure your web crawler to index web pages. Optionally add a Firecrawl API key
Configure your web page crawler to index web pages. Optionally add a Firecrawl API key
for enhanced crawling capabilities.
</CardDescription>
</CardHeader>
@ -277,7 +277,7 @@ export default function WebcrawlerConnectorPage() {
/* Success Card */
<Card>
<CardHeader>
<CardTitle>✅ Your web crawler is successfully set up!</CardTitle>
<CardTitle>✅ Your web page crawler is successfully set up!</CardTitle>
<CardDescription>
You can now add URLs to crawl from the connector management page.
</CardDescription>

201
surfsense_web/app/dashboard/[search_space_id]/documents/webpage/page.tsx
View file

@ -1,201 +0,0 @@
"use client";
import { type Tag, TagInput } from "emblor";
import { Globe, Loader2 } from "lucide-react";
import { useParams, useRouter } from "next/navigation";
import { useTranslations } from "next-intl";
import { useState } from "react";
import { toast } from "sonner";
import { Button } from "@/components/ui/button";
import {
Card,
CardContent,
CardDescription,
CardFooter,
CardHeader,
CardTitle,
} from "@/components/ui/card";
import { Label } from "@/components/ui/label";
// URL validation regex
const urlRegex = /^(https?:\/\/)?([\da-z.-]+)\.([a-z.]{2,6})([/\w .-]*)*\/?$/;
export default function WebpageCrawler() {
const t = useTranslations("add_webpage");
const params = useParams();
const router = useRouter();
const search_space_id = params.search_space_id as string;
const [urlTags, setUrlTags] = useState<Tag[]>([]);
const [activeTagIndex, setActiveTagIndex] = useState<number | null>(null);
const [isSubmitting, setIsSubmitting] = useState(false);
const [error, setError] = useState<string | null>(null);
// Function to validate a URL
const isValidUrl = (url: string): boolean => {
return urlRegex.test(url);
};
// Function to handle URL submission
const handleSubmit = async () => {
// Validate that we have at least one URL
if (urlTags.length === 0) {
setError(t("error_no_url"));
return;
}
// Validate all URLs
const invalidUrls = urlTags.filter((tag) => !isValidUrl(tag.text));
if (invalidUrls.length > 0) {
setError(t("error_invalid_urls", { urls: invalidUrls.map((tag) => tag.text).join(", ") }));
return;
}
setError(null);
setIsSubmitting(true);
try {
toast(t("crawling_toast"), {
description: t("crawling_toast_desc"),
});
// Extract URLs from tags
const urls = urlTags.map((tag) => tag.text);
// Make API call to backend
const response = await fetch(
`${process.env.NEXT_PUBLIC_FASTAPI_BACKEND_URL}/api/v1/documents`,
{
method: "POST",
headers: {
"Content-Type": "application/json",
Authorization: `Bearer ${localStorage.getItem("surfsense_bearer_token")}`,
},
body: JSON.stringify({
document_type: "CRAWLED_URL",
content: urls,
search_space_id: parseInt(search_space_id),
}),
}
);
if (!response.ok) {
throw new Error("Failed to crawl URLs");
}
await response.json();
toast(t("success_toast"), {
description: t("success_toast_desc"),
});
// Redirect to documents page
router.push(`/dashboard/${search_space_id}/documents`);
} catch (error: any) {
setError(error.message || t("error_generic"));
toast(t("error_toast"), {
description: `${t("error_toast_desc")}: ${error.message}`,
});
} finally {
setIsSubmitting(false);
}
};
// Function to add a new URL tag
const handleAddTag = (text: string) => {
// Basic URL validation
if (!isValidUrl(text)) {
toast(t("invalid_url_toast"), {
description: t("invalid_url_toast_desc"),
});
return;
}
// Check for duplicates
if (urlTags.some((tag) => tag.text === text)) {
toast(t("duplicate_url_toast"), {
description: t("duplicate_url_toast_desc"),
});
return;
}
// Add the new tag
const newTag: Tag = {
id: Date.now().toString(),
text: text,
};
setUrlTags([...urlTags, newTag]);
};
return (
<div className="container mx-auto py-8">
<Card className="max-w-2xl mx-auto">
<CardHeader>
<CardTitle className="flex items-center gap-2">
<Globe className="h-5 w-5" />
{t("title")}
</CardTitle>
<CardDescription>{t("subtitle")}</CardDescription>
</CardHeader>
<CardContent>
<div className="space-y-4">
<div className="space-y-2">
<Label htmlFor="url-input">{t("label")}</Label>
<TagInput
id="url-input"
tags={urlTags}
setTags={setUrlTags}
placeholder={t("placeholder")}
onAddTag={handleAddTag}
styleClasses={{
inlineTagsContainer:
"border-input rounded-lg bg-background shadow-sm shadow-black/5 transition-shadow focus-within:border-ring focus-within:outline-none focus-within:ring-[3px] focus-within:ring-ring/20 p-1 gap-1",
input: "w-full min-w-[80px] focus-visible:outline-none shadow-none px-2 h-7",
tag: {
body: "h-7 relative bg-background border border-input hover:bg-background rounded-md font-medium text-xs ps-2 pe-7 flex",
closeButton:
"absolute -inset-y-px -end-px p-0 rounded-e-lg flex size-7 transition-colors outline-0 focus-visible:outline focus-visible:outline-2 focus-visible:outline-ring/70 text-muted-foreground/80 hover:text-foreground",
},
}}
activeTagIndex={activeTagIndex}
setActiveTagIndex={setActiveTagIndex}
/>
<p className="text-xs text-muted-foreground mt-1">{t("hint")}</p>
</div>
{error && <div className="text-sm text-red-500 mt-2">{error}</div>}
<div className="bg-muted/50 rounded-lg p-4 text-sm">
<h4 className="font-medium mb-2">{t("tips_title")}</h4>
<ul className="list-disc pl-5 space-y-1 text-muted-foreground">
<li>{t("tip_1")}</li>
<li>{t("tip_2")}</li>
<li>{t("tip_3")}</li>
<li>{t("tip_4")}</li>
</ul>
</div>
</div>
</CardContent>
<CardFooter className="flex justify-between">
<Button
variant="outline"
onClick={() => router.push(`/dashboard/${search_space_id}/documents`)}
>
{t("cancel")}
</Button>
<Button onClick={handleSubmit} disabled={isSubmitting || urlTags.length === 0}>
{isSubmitting ? (
<>
<Loader2 className="mr-2 h-4 w-4 animate-spin" />
{t("submitting")}
</>
) : (
t("submit")
)}
</Button>
</CardFooter>
</Card>
</div>
);
}

2
surfsense_web/components/dashboard-breadcrumb.tsx
View file

@ -138,7 +138,7 @@ export function DashboardBreadcrumb() {
"linkup-api": "LinkUp API",
"luma-connector": "Luma",
"elasticsearch-connector": "Elasticsearch",
"webcrawler-connector": "WebCrawler",
"webcrawler-connector": "Web Pages",
};
const connectorLabel = connectorLabels[connectorType] || connectorType;

1
surfsense_web/components/editConnector/types.ts
View file

@ -53,5 +53,6 @@ export const editConnectorSchema = z.object({
LUMA_API_KEY: z.string().optional(),
ELASTICSEARCH_API_KEY: z.string().optional(),
FIRECRAWL_API_KEY: z.string().optional(),
INITIAL_URLS: z.string().optional()
});
export type EditConnectorFormValues = z.infer<typeof editConnectorSchema>;

2
surfsense_web/components/homepage/integrations.tsx
View file

@ -29,7 +29,7 @@ const INTEGRATIONS: Integration[] = [
// Documentation & Knowledge
{ name: "Confluence", icon: "https://cdn.simpleicons.org/confluence/172B4D" },
{ name: "Notion", icon: "https://cdn.simpleicons.org/notion/000000/ffffff" },
{ name: "Web Crawler", icon: "https://cdn.jsdelivr.net/npm/lucide-static@0.294.0/icons/globe.svg"},
{ name: "Web Pages", icon: "https://cdn.jsdelivr.net/npm/lucide-static@0.294.0/icons/globe.svg"},
// Cloud Storage
{ name: "Google Drive", icon: "https://cdn.simpleicons.org/googledrive/4285F4" },

2
surfsense_web/components/sources/connector-data.tsx
View file

@ -140,7 +140,7 @@ export const connectorCategories: ConnectorCategory[] = [
},
{
id: "webcrawler-connector",
title: "Web Crawler",
title: "Web Pages",
description: "webcrawler_desc",
icon: getConnectorIcon(EnumConnectorName.WEBCRAWLER_CONNECTOR, "h-6 w-6"),
status: "available",

3
surfsense_web/content/docs/docker-installation.mdx
View file

@ -96,8 +96,7 @@ Before you begin, ensure you have:
| TTS_SERVICE_API_BASE | (Optional) Custom API base URL for the Text-to-Speech service |
| STT_SERVICE | Speech-to-Text API provider for Audio Files (e.g., `local/base`, `openai/whisper-1`). See [supported providers](https://docs.litellm.ai/docs/audio_transcription#supported-providers) |
| STT_SERVICE_API_KEY | (Optional if local) API key for the Speech-to-Text service |
| STT_SERVICE_API_BASE | (Optional) Custom API base URL for the Speech-to-Text service |
| FIRECRAWL_API_KEY | API key for Firecrawl service for web crawling |
| STT_SERVICE_API_BASE | (Optional) Custom API base URL for the Speech-to-Text service | |
| ETL_SERVICE | Document parsing service: `UNSTRUCTURED` (supports 34+ formats), `LLAMACLOUD` (supports 50+ formats including legacy document types), or `DOCLING` (local processing, supports PDF, Office docs, images, HTML, CSV) |
| UNSTRUCTURED_API_KEY | API key for Unstructured.io service for document parsing (required if ETL_SERVICE=UNSTRUCTURED) |
| LLAMA_CLOUD_API_KEY | API key for LlamaCloud service for document parsing (required if ETL_SERVICE=LLAMACLOUD) |

2
surfsense_web/contracts/enums/connectorIcons.tsx
View file

@ -62,6 +62,8 @@ export const getConnectorIcon = (connectorType: EnumConnectorName | string, clas
case EnumConnectorName.WEBCRAWLER_CONNECTOR:
return <Globe {...iconProps} />;
// Additional cases for non-enum connector types
case "CRAWLED_URL":
return <Globe {...iconProps} />;
case "YOUTUBE_VIDEO":
return <IconBrandYoutube {...iconProps} />;
case "FILE":

23
surfsense_web/hooks/use-connector-edit-page.ts
View file

@ -98,6 +98,7 @@ export function useConnectorEditPage(connectorId: number, searchSpaceId: string)
LUMA_API_KEY: "",
ELASTICSEARCH_API_KEY: "",
FIRECRAWL_API_KEY: "",
INITIAL_URLS: ""
},
});
@ -144,6 +145,7 @@ export function useConnectorEditPage(connectorId: number, searchSpaceId: string)
LUMA_API_KEY: config.LUMA_API_KEY || "",
ELASTICSEARCH_API_KEY: config.ELASTICSEARCH_API_KEY || "",
FIRECRAWL_API_KEY: config.FIRECRAWL_API_KEY || "",
INTIAL_URLS: config.INITIAL_URLS || ""
});
if (currentConnector.connector_type === "GITHUB_CONNECTOR") {
const savedRepos = config.repo_full_names || [];
@ -472,16 +474,28 @@ export function useConnectorEditPage(connectorId: number, searchSpaceId: string)
}
break;
case "WEBCRAWLER_CONNECTOR":
if (formData.FIRECRAWL_API_KEY !== originalConfig.FIRECRAWL_API_KEY) {
if (
formData.FIRECRAWL_API_KEY !== originalConfig.FIRECRAWL_API_KEY ||
formData.INITIAL_URLS !== originalConfig.INITIAL_URLS
) {
newConfig = {};
if (formData.FIRECRAWL_API_KEY && formData.FIRECRAWL_API_KEY.trim()) {
if (!formData.FIRECRAWL_API_KEY.startsWith("fc-")) {
toast.warning("Firecrawl API keys typically start with 'fc-'. Please verify your key.");
}
newConfig = { FIRECRAWL_API_KEY: formData.FIRECRAWL_API_KEY };
} else {
newConfig = {};
newConfig.FIRECRAWL_API_KEY = formData.FIRECRAWL_API_KEY.trim();
} else if (originalConfig.FIRECRAWL_API_KEY) {
toast.info("Firecrawl API key removed. Web crawler will use AsyncChromiumLoader as fallback.");
}
if (formData.INITIAL_URLS !== undefined) {
if (formData.INITIAL_URLS && formData.INITIAL_URLS.trim()) {
newConfig.INITIAL_URLS = formData.INITIAL_URLS.trim();
} else if (originalConfig.INITIAL_URLS) {
toast.info("URLs removed from crawler configuration.");
}
}
}
break;
}
@ -579,6 +593,7 @@ export function useConnectorEditPage(connectorId: number, searchSpaceId: string)
);
} else if (connector.connector_type == "WEBCRAWLER_CONNECTOR") {
editForm.setValue("FIRECRAWL_API_KEY",newlySavedConfig.FIRECRAWL_API_KEY || "");
editForm.setValue("INITIAL_URLS", newlySavedConfig.INITIAL_URLS || "");
}
}
if (connector.connector_type === "GITHUB_CONNECTOR") {

2
surfsense_web/lib/connectors/utils.ts
View file

@ -18,7 +18,7 @@ export const getConnectorTypeDisplay = (type: string): string => {
AIRTABLE_CONNECTOR: "Airtable",
LUMA_CONNECTOR: "Luma",
ELASTICSEARCH_CONNECTOR: "Elasticsearch",
WEBCRAWLER_CONNECTOR: "Web Crawler",
WEBCRAWLER_CONNECTOR: "Web Pages",
};
return typeMap[type] || type;
};

2
surfsense_web/messages/en.json
View file

@ -332,7 +332,7 @@
"calendar_desc": "Connect to Google Calendar to search events, meetings and schedules.",
"gmail_desc": "Connect to your Gmail account to search through your emails.",
"zoom_desc": "Connect to Zoom to access meeting recordings and transcripts.",
"webcrawler_desc": "Scrape web pages using FireCrawl."
"webcrawler_desc": "Crawl web pages"
},
"upload_documents": {
"title": "Upload Documents",