Merge branch 'master' of github.com:khoj-ai/khoj into features/add-chat-controls

This commit is contained in:
sabaimran
2025-01-21 10:32:47 -08:00
13 changed files with 1379 additions and 497 deletions

View File

@@ -31,6 +31,7 @@ interface ChatMessageState {
rawResponse: string;
rawQuery: string;
isVoice: boolean;
turnId: string;
}
interface Location {
@@ -41,6 +42,17 @@ interface Location {
timezone: string;
}
interface RenderMessageOptions {
chatBodyEl: Element;
message: string;
sender: string;
turnId?: string;
dt?: Date;
raw?: boolean;
willReplace?: boolean;
isSystemMessage?: boolean;
}
export class KhojChatView extends KhojPaneView {
result: string;
setting: KhojSetting;
@@ -480,6 +492,7 @@ export class KhojChatView extends KhojPaneView {
chatEl: Element,
message: string,
sender: string,
turnId: string,
context?: string[],
onlineContext?: object,
dt?: Date,
@@ -500,9 +513,21 @@ export class KhojChatView extends KhojPaneView {
mermaidjsDiagram ||
excalidrawDiagram) {
let imageMarkdown = this.generateImageMarkdown(message, intentType ?? "", inferredQueries, conversationId, images, excalidrawDiagram, mermaidjsDiagram);
chatMessageEl = this.renderMessage(chatEl, imageMarkdown, sender, dt);
chatMessageEl = this.renderMessage({
chatBodyEl: chatEl,
message: imageMarkdown,
sender,
dt,
turnId
});
} else {
chatMessageEl = this.renderMessage(chatEl, message, sender, dt);
chatMessageEl = this.renderMessage({
chatBodyEl: chatEl,
message,
sender,
dt,
turnId
});
}
// If no document or online context is provided, skip rendering the reference section
@@ -547,7 +572,7 @@ export class KhojChatView extends KhojPaneView {
return imageMarkdown;
}
renderMessage(chatBodyEl: Element, message: string, sender: string, dt?: Date, raw: boolean = false, willReplace: boolean = true): Element {
renderMessage({ chatBodyEl, message, sender, dt, turnId, raw = false, willReplace = true, isSystemMessage = false }: RenderMessageOptions): Element {
let message_time = this.formatDate(dt ?? new Date());
// Append message to conversation history HTML element.
@@ -555,7 +580,8 @@ export class KhojChatView extends KhojPaneView {
let chatMessageEl = chatBodyEl.createDiv({
attr: {
"data-meta": message_time,
class: `khoj-chat-message ${sender}`
class: `khoj-chat-message ${sender}`,
...(turnId && { "data-turnId": turnId })
},
})
let chatMessageBodyEl = chatMessageEl.createDiv();
@@ -574,7 +600,7 @@ export class KhojChatView extends KhojPaneView {
// Add action buttons to each chat message element
if (willReplace === true) {
this.renderActionButtons(message, chatMessageBodyTextEl);
this.renderActionButtons(message, chatMessageBodyTextEl, isSystemMessage);
}
// Remove user-select: none property to make text selectable
@@ -618,7 +644,7 @@ export class KhojChatView extends KhojPaneView {
this.scrollChatToBottom();
}
renderActionButtons(message: string, chatMessageBodyTextEl: HTMLElement) {
renderActionButtons(message: string, chatMessageBodyTextEl: HTMLElement, isSystemMessage: boolean = false) {
let copyButton = this.contentEl.createEl('button');
copyButton.classList.add("chat-action-button");
copyButton.title = "Copy Message to Clipboard";
@@ -632,6 +658,25 @@ export class KhojChatView extends KhojPaneView {
setIcon(pasteToFile, "clipboard-paste");
pasteToFile.addEventListener('click', (event) => { pasteTextAtCursor(createCopyParentText(message, 'clipboard-paste')(event)); });
// Add delete button
let deleteButton = null;
if (!isSystemMessage) {
deleteButton = this.contentEl.createEl('button');
deleteButton.classList.add("chat-action-button");
deleteButton.title = "Delete Message";
setIcon(deleteButton, "trash-2");
deleteButton.addEventListener('click', () => {
const messageEl = chatMessageBodyTextEl.closest('.khoj-chat-message');
if (messageEl) {
// Ask for confirmation before deleting
if (confirm('Are you sure you want to delete this message?')) {
this.deleteMessage(messageEl as HTMLElement);
}
}
});
}
// Only enable the speech feature if the user is subscribed
let speechButton = null;
@@ -646,7 +691,9 @@ export class KhojChatView extends KhojPaneView {
// Append buttons to parent element
chatMessageBodyTextEl.append(copyButton, pasteToFile);
if (deleteButton) {
chatMessageBodyTextEl.append(deleteButton);
}
if (speechButton) {
chatMessageBodyTextEl.append(speechButton);
}
@@ -672,7 +719,7 @@ export class KhojChatView extends KhojPaneView {
if (chatInput) {
chatInput.placeholder = this.startingMessage;
}
this.renderMessage(chatBodyEl, "Hey 👋🏾, what's up?", "khoj");
this.renderMessage({chatBodyEl, message: "Hey 👋🏾, what's up?", sender: "khoj", isSystemMessage: true});
}
async toggleChatSessions(forceShow: boolean = false): Promise<boolean> {
@@ -883,7 +930,12 @@ export class KhojChatView extends KhojPaneView {
if (responseJson.detail) {
// If the server returns error details in response, render a setup hint.
let setupMsg = "Hi 👋🏾, to start chatting add available chat models options via [the Django Admin panel](/server/admin) on the Server";
this.renderMessage(chatBodyEl, setupMsg, "khoj", undefined);
this.renderMessage({
chatBodyEl,
message: setupMsg,
sender: "khoj",
isSystemMessage: true
});
return false;
} else if (responseJson.response) {
@@ -897,6 +949,7 @@ export class KhojChatView extends KhojPaneView {
chatBodyEl,
chatLog.message,
chatLog.by,
chatLog.turnId,
chatLog.context,
chatLog.onlineContext,
new Date(chatLog.created),
@@ -927,7 +980,12 @@ export class KhojChatView extends KhojPaneView {
}
} catch (err) {
let errorMsg = "Unable to get response from Khoj server ❤️‍🩹. Ensure server is running or contact developers for help at [team@khoj.dev](mailto:team@khoj.dev) or in [Discord](https://discord.gg/BDgyabRM6e)";
this.renderMessage(chatBodyEl, errorMsg, "khoj", undefined);
this.renderMessage({
chatBodyEl,
message: errorMsg,
sender: "khoj",
isSystemMessage: true
});
return false;
}
return true;
@@ -972,7 +1030,7 @@ export class KhojChatView extends KhojPaneView {
this.textToSpeech(this.chatMessageState.rawResponse);
// Append any references after all the data has been streamed
this.finalizeChatBodyResponse(this.chatMessageState.references, this.chatMessageState.newResponseTextEl);
this.finalizeChatBodyResponse(this.chatMessageState.references, this.chatMessageState.newResponseTextEl, this.chatMessageState.turnId);
const liveQuery = this.chatMessageState.rawQuery;
// Reset variables
@@ -985,6 +1043,7 @@ export class KhojChatView extends KhojPaneView {
rawQuery: liveQuery,
isVoice: false,
generatedAssets: "",
turnId: "",
};
} else if (chunk.type === "references") {
this.chatMessageState.references = { "notes": chunk.data.context, "online": chunk.data.onlineContext };
@@ -1006,6 +1065,12 @@ export class KhojChatView extends KhojPaneView {
this.chatMessageState.rawResponse += chunkData;
this.handleStreamResponse(this.chatMessageState.newResponseTextEl, this.chatMessageState.rawResponse + this.chatMessageState.generatedAssets, this.chatMessageState.loadingEllipsis);
}
} else if (chunk.type === "metadata") {
const { turnId } = chunk.data;
if (turnId) {
// Append turnId to chatMessageState
this.chatMessageState.turnId = turnId;
}
}
}
@@ -1065,7 +1130,7 @@ export class KhojChatView extends KhojPaneView {
// Render user query as chat message
let chatBodyEl = this.contentEl.getElementsByClassName("khoj-chat-body")[0] as HTMLElement;
this.renderMessage(chatBodyEl, query, "you");
this.renderMessage({chatBodyEl, message: query, sender: "you"});
let conversationId = chatBodyEl.dataset.conversationId;
if (!conversationId) {
@@ -1111,6 +1176,7 @@ export class KhojChatView extends KhojPaneView {
rawResponse: "",
isVoice: isVoice,
generatedAssets: "",
turnId: "",
};
let response = await fetch(chatUrl, {
@@ -1410,10 +1476,15 @@ export class KhojChatView extends KhojPaneView {
return rawResponse;
}
finalizeChatBodyResponse(references: object, newResponseElement: HTMLElement | null) {
finalizeChatBodyResponse(references: object, newResponseElement: HTMLElement | null, turnId: string) {
if (!!newResponseElement && references != null && Object.keys(references).length > 0) {
newResponseElement.appendChild(this.createReferenceSection(references));
}
if (!!newResponseElement && turnId) {
// Set the turnId for the new response and the previous user message
newResponseElement.parentElement?.setAttribute("data-turnId", turnId);
newResponseElement.parentElement?.previousElementSibling?.setAttribute("data-turnId", turnId);
}
this.scrollChatToBottom();
let chatInput = this.contentEl.getElementsByClassName("khoj-chat-input")[0];
if (chatInput) chatInput.removeAttribute("disabled");
@@ -1482,4 +1553,49 @@ export class KhojChatView extends KhojPaneView {
}
}
}
// Add this new method to handle message deletion
async deleteMessage(messageEl: HTMLElement) {
const chatBodyEl = this.contentEl.getElementsByClassName("khoj-chat-body")[0] as HTMLElement;
const conversationId = chatBodyEl.dataset.conversationId;
// Get the turnId from the message's data-turn attribute
const turnId = messageEl.getAttribute("data-turnId");
if (!turnId || !conversationId) return;
try {
const response = await fetch(`${this.setting.khojUrl}/api/chat/conversation/message`, {
method: "DELETE",
headers: {
"Content-Type": "application/json",
"Authorization": `Bearer ${this.setting.khojApiKey}`
},
body: JSON.stringify({
conversation_id: conversationId,
turn_id: turnId
})
});
if (response.ok) {
// Remove both the user message and Khoj response (the conversation turn)
const isKhojMessage = messageEl.classList.contains("khoj");
const messages = Array.from(chatBodyEl.getElementsByClassName("khoj-chat-message"));
const messageIndex = messages.indexOf(messageEl);
if (isKhojMessage && messageIndex > 0) {
// If it is a Khoj message, remove the previous user message too
messages[messageIndex - 1].remove();
} else if (!isKhojMessage && messageIndex < messages.length - 1) {
// If it is a user message, remove the next Khoj message too
messages[messageIndex + 1].remove();
}
messageEl.remove();
} else {
this.flashStatusInChatInput("Failed to delete message");
}
} catch (error) {
console.error("Error deleting message:", error);
this.flashStatusInChatInput("Error deleting message");
}
}
}

View File

@@ -114,3 +114,33 @@ export function useDebounce<T>(value: T, delay: number): T {
return debouncedValue;
}
export const formatDateTime = (isoString: string): string => {
try {
const date = new Date(isoString);
const now = new Date();
const diffInMinutes = Math.floor((now.getTime() - date.getTime()) / 60000);
// Show relative time for recent dates
if (diffInMinutes < 1) return "just now";
if (diffInMinutes < 60) return `${diffInMinutes} minutes ago`;
if (diffInMinutes < 120) return "1 hour ago";
if (diffInMinutes < 1440) return `${Math.floor(diffInMinutes / 60)} hours ago`;
// For older dates, show full formatted date
const formatter = new Intl.DateTimeFormat("en-US", {
month: "long",
day: "numeric",
year: "numeric",
hour: "numeric",
minute: "2-digit",
hour12: true,
timeZoneName: "short",
});
return formatter.format(date);
} catch (error) {
console.error("Error formatting date:", error);
return isoString;
}
};

View File

@@ -2,6 +2,7 @@ import type { Metadata } from "next";
import "../globals.css";
import { ContentSecurityPolicy } from "../common/layoutHelper";
import { Toaster } from "@/components/ui/toaster";
export const metadata: Metadata = {
title: "Khoj AI - Search",
@@ -35,7 +36,10 @@ export default function RootLayout({
return (
<html>
<ContentSecurityPolicy />
<body>{children}</body>
<body>
{children}
<Toaster />
</body>
</html>
);
}

File diff suppressed because it is too large Load Diff

View File

@@ -3,7 +3,7 @@
import styles from "./settings.module.css";
import "intl-tel-input/styles";
import { Suspense, useEffect, useRef, useState } from "react";
import { Suspense, useEffect, useState } from "react";
import { useToast } from "@/components/ui/use-toast";
import { useUserConfig, ModelOptions, UserConfig, SubscriptionStates } from "../common/auth";
@@ -23,14 +23,6 @@ import {
DropdownMenuTrigger,
} from "@/components/ui/dropdown-menu";
import { Table, TableBody, TableCell, TableRow } from "@/components/ui/table";
import {
CommandInput,
CommandList,
CommandEmpty,
CommandGroup,
CommandItem,
CommandDialog,
} from "@/components/ui/command";
import {
ArrowRight,
@@ -56,9 +48,10 @@ import {
ArrowCircleUp,
ArrowCircleDown,
ArrowsClockwise,
Check,
CaretDown,
Waveform,
MagnifyingGlass,
Brain,
EyeSlash,
Eye,
} from "@phosphor-icons/react";
@@ -66,312 +59,11 @@ import {
import Loading from "../components/loading/loading";
import IntlTelInput from "intl-tel-input/react";
import { uploadDataForIndexing } from "../common/chatFunctions";
import {
AlertDialog,
AlertDialogAction,
AlertDialogContent,
AlertDialogDescription,
AlertDialogHeader,
AlertDialogTitle,
} from "@/components/ui/alert-dialog";
import { Progress } from "@/components/ui/progress";
import Link from "next/link";
import { SidebarInset, SidebarProvider, SidebarTrigger } from "@/components/ui/sidebar";
import { AppSidebar } from "../components/appSidebar/appSidebar";
import { Separator } from "@/components/ui/separator";
import { KhojLogoType } from "../components/logo/khojLogo";
const ManageFilesModal: React.FC<{ onClose: () => void }> = ({ onClose }) => {
const [syncedFiles, setSyncedFiles] = useState<string[]>([]);
const [selectedFiles, setSelectedFiles] = useState<string[]>([]);
const [searchQuery, setSearchQuery] = useState("");
const [isDragAndDropping, setIsDragAndDropping] = useState(false);
const [warning, setWarning] = useState<string | null>(null);
const [error, setError] = useState<string | null>(null);
const [uploading, setUploading] = useState(false);
const [progressValue, setProgressValue] = useState(0);
const [uploadedFiles, setUploadedFiles] = useState<string[]>([]);
const fileInputRef = useRef<HTMLInputElement>(null);
useEffect(() => {
if (!uploading) {
setProgressValue(0);
}
if (uploading) {
const interval = setInterval(() => {
setProgressValue((prev) => {
const increment = Math.floor(Math.random() * 5) + 1; // Generates a random number between 1 and 5
const nextValue = prev + increment;
return nextValue < 100 ? nextValue : 100; // Ensures progress does not exceed 100
});
}, 800);
return () => clearInterval(interval);
}
}, [uploading]);
useEffect(() => {
const fetchFiles = async () => {
try {
const response = await fetch("/api/content/computer");
if (!response.ok) throw new Error("Failed to fetch files");
// Extract resonse
const syncedFiles = await response.json();
// Validate response
if (Array.isArray(syncedFiles)) {
// Set synced files state
setSyncedFiles(syncedFiles.toSorted());
} else {
console.error("Unexpected data format from API");
}
} catch (error) {
console.error("Error fetching files:", error);
}
};
fetchFiles();
}, [uploadedFiles]);
const filteredFiles = syncedFiles.filter((file) =>
file.toLowerCase().includes(searchQuery.toLowerCase()),
);
const deleteSelected = async () => {
let filesToDelete = selectedFiles.length > 0 ? selectedFiles : filteredFiles;
if (filesToDelete.length === 0) {
return;
}
try {
const response = await fetch("/api/content/files", {
method: "DELETE",
headers: {
"Content-Type": "application/json",
},
body: JSON.stringify({ files: filesToDelete }),
});
if (!response.ok) throw new Error("Failed to delete files");
// Update the syncedFiles state
setSyncedFiles((prevFiles) =>
prevFiles.filter((file) => !filesToDelete.includes(file)),
);
// Reset selectedFiles
setSelectedFiles([]);
} catch (error) {
console.error("Error deleting files:", error);
}
};
const deleteFile = async (filename: string) => {
try {
const response = await fetch(
`/api/content/file?filename=${encodeURIComponent(filename)}`,
{
method: "DELETE",
headers: {
"Content-Type": "application/json",
},
},
);
if (!response.ok) throw new Error("Failed to delete file");
// Update the syncedFiles state
setSyncedFiles((prevFiles) => prevFiles.filter((file) => file !== filename));
// Remove the file from selectedFiles if it's there
setSelectedFiles((prevSelected) => prevSelected.filter((file) => file !== filename));
} catch (error) {
console.error("Error deleting file:", error);
}
};
function handleDragOver(event: React.DragEvent<HTMLDivElement>) {
event.preventDefault();
setIsDragAndDropping(true);
}
function handleDragLeave(event: React.DragEvent<HTMLDivElement>) {
event.preventDefault();
setIsDragAndDropping(false);
}
function handleDragAndDropFiles(event: React.DragEvent<HTMLDivElement>) {
event.preventDefault();
setIsDragAndDropping(false);
if (!event.dataTransfer.files) return;
uploadFiles(event.dataTransfer.files);
}
function openFileInput() {
if (fileInputRef && fileInputRef.current) {
fileInputRef.current.click();
}
}
function handleFileChange(event: React.ChangeEvent<HTMLInputElement>) {
if (!event.target.files) return;
uploadFiles(event.target.files);
}
function uploadFiles(files: FileList) {
uploadDataForIndexing(files, setWarning, setUploading, setError, setUploadedFiles);
}
return (
<CommandDialog open={true} onOpenChange={onClose}>
<AlertDialog open={warning !== null || error != null}>
<AlertDialogContent>
<AlertDialogHeader>
<AlertDialogTitle>Alert</AlertDialogTitle>
</AlertDialogHeader>
<AlertDialogDescription>{warning || error}</AlertDialogDescription>
<AlertDialogAction
className="bg-slate-400 hover:bg-slate-500"
onClick={() => {
setWarning(null);
setError(null);
setUploading(false);
}}
>
Close
</AlertDialogAction>
</AlertDialogContent>
</AlertDialog>
<div
className={`flex flex-col h-full`}
onDragOver={handleDragOver}
onDragLeave={handleDragLeave}
onDrop={handleDragAndDropFiles}
onClick={openFileInput}
>
<input
type="file"
multiple
ref={fileInputRef}
style={{ display: "none" }}
onChange={handleFileChange}
/>
<div className="flex-none p-4">
Upload files
{uploading && (
<Progress
indicatorColor="bg-slate-500"
className="w-full h-2 rounded-full"
value={progressValue}
/>
)}
</div>
<div
className={`flex-none p-4 bg-secondary border-b ${isDragAndDropping ? "animate-pulse" : ""} rounded-lg`}
>
<div className="flex items-center justify-center w-full h-32 border-2 border-dashed border-gray-300 rounded-lg">
{isDragAndDropping ? (
<div className="flex items-center justify-center w-full h-full">
<Waveform className="h-6 w-6 mr-2" />
<span>Drop files to upload</span>
</div>
) : (
<div className="flex items-center justify-center w-full h-full">
<Plus className="h-6 w-6 mr-2" />
<span>Drag and drop files here</span>
</div>
)}
</div>
</div>
</div>
<div className="flex flex-col h-full">
<div className="flex-none p-4 bg-background border-b">
<CommandInput
placeholder="Find synced files"
value={searchQuery}
onValueChange={setSearchQuery}
/>
</div>
<div className="flex-grow overflow-auto">
<CommandList>
<CommandEmpty>
{syncedFiles.length === 0 ? (
<div className="flex items-center justify-center">
<ExclamationMark className="h-4 w-4 mr-2" weight="bold" />
No files synced
</div>
) : (
<div>
Could not find a good match.
<Link href="/search" className="block">
Need advanced search? Click here.
</Link>
</div>
)}
</CommandEmpty>
<CommandGroup heading="Synced files">
{filteredFiles.map((filename: string) => (
<CommandItem
key={filename}
value={filename}
onSelect={(value) => {
setSelectedFiles((prev) =>
prev.includes(value)
? prev.filter((f) => f !== value)
: [...prev, value],
);
}}
>
<div className="flex items-center justify-between w-full">
<div
className={`flex items-center ${selectedFiles.includes(filename) ? "font-semibold" : ""}`}
>
{selectedFiles.includes(filename) && (
<Check className="h-4 w-4 mr-2" />
)}
<span className="break-all">{filename}</span>
</div>
<Button
variant="outline"
size="sm"
onClick={() => deleteFile(filename)}
className="ml-auto"
>
<Trash className="h-4 w-4" />
</Button>
</div>
</CommandItem>
))}
</CommandGroup>
</CommandList>
</div>
<div className="flex-none p-4 bg-background border-t">
<div className="flex justify-between">
<Button
variant="outline"
size="sm"
onClick={deleteSelected}
className="mr-2"
>
<Trash className="h-4 w-4 mr-2" />
{selectedFiles.length > 0
? `Delete Selected (${selectedFiles.length})`
: "Delete All"}
</Button>
</div>
</div>
</div>
</CommandDialog>
);
};
interface DropdownComponentProps {
items: ModelOptions[];
selected: number;
@@ -604,7 +296,6 @@ export default function SettingsView() {
const [numberValidationState, setNumberValidationState] = useState<PhoneNumberValidationState>(
PhoneNumberValidationState.Verified,
);
const [isManageFilesModalOpen, setIsManageFilesModalOpen] = useState(false);
const { toast } = useToast();
const isMobileWidth = useIsMobileWidth();
@@ -1162,18 +853,13 @@ export default function SettingsView() {
</Card>
</div>
</div>
{isManageFilesModalOpen && (
<ManageFilesModal
onClose={() => setIsManageFilesModalOpen(false)}
/>
)}
<div className="section grid gap-8">
<div className="text-2xl">Content</div>
<div className="cards flex flex-wrap gap-16">
<Card id="computer" className={cardClassName}>
<CardHeader className="flex flex-row text-2xl">
<Laptop className="h-8 w-8 mr-2" />
Files
<CardHeader className="flex flex-row text-xl">
<Brain className="h-8 w-8 mr-2" />
Knowledge Base
{userConfig.enabled_content_source.computer && (
<CheckCircle
className="h-6 w-6 ml-auto text-green-500"
@@ -1182,20 +868,19 @@ export default function SettingsView() {
)}
</CardHeader>
<CardContent className="overflow-hidden pb-12 text-gray-400">
Manage your synced files
Manage and search through your digital brain.
</CardContent>
<CardFooter className="flex flex-wrap gap-4">
<Button
variant="outline"
size="sm"
title="Search thorugh files"
onClick={() =>
setIsManageFilesModalOpen(true)
(window.location.href = "/search")
}
>
<>
<Files className="h-5 w-5 inline mr-1" />
Manage
</>
<MagnifyingGlass className="h-5 w-5 inline mr-1" />
Search
</Button>
<Button
variant="outline"
@@ -1206,7 +891,7 @@ export default function SettingsView() {
}
>
<CloudSlash className="h-5 w-5 inline mr-1" />
Disable
Clear All
</Button>
</CardFooter>
</Card>

View File

@@ -0,0 +1,118 @@
import * as React from "react"
import { ChevronLeft, ChevronRight, MoreHorizontal } from "lucide-react"
import { cn } from "@/lib/utils"
import { ButtonProps, buttonVariants } from "@/components/ui/button"
const Pagination = ({ className, ...props }: React.ComponentProps<"nav">) => (
<nav
role="navigation"
aria-label="pagination"
className={cn("mx-auto flex w-full justify-center", className)}
{...props}
/>
)
Pagination.displayName = "Pagination"
const PaginationContent = React.forwardRef<
HTMLUListElement,
React.ComponentProps<"ul">
>(({ className, ...props }, ref) => (
<ul
ref={ref}
className={cn("flex flex-row items-center gap-1", className)}
{...props}
/>
))
PaginationContent.displayName = "PaginationContent"
const PaginationItem = React.forwardRef<
HTMLLIElement,
React.ComponentProps<"li">
>(({ className, ...props }, ref) => (
<li ref={ref} className={cn("", className)} {...props} />
))
PaginationItem.displayName = "PaginationItem"
type PaginationLinkProps = {
isActive?: boolean
} & Pick<ButtonProps, "size"> &
React.ComponentProps<"a">
const PaginationLink = ({
className,
isActive,
size = "icon",
...props
}: PaginationLinkProps) => (
<a
aria-current={isActive ? "page" : undefined}
className={cn(
buttonVariants({
variant: isActive ? "outline" : "ghost",
size,
}),
"no-underline",
className
)}
{...props}
/>
)
PaginationLink.displayName = "PaginationLink"
const PaginationPrevious = ({
className,
...props
}: React.ComponentProps<typeof PaginationLink>) => (
<PaginationLink
aria-label="Go to previous page"
size="default"
className={cn("gap-1 pl-2.5", className)}
{...props}
>
<ChevronLeft className="h-4 w-4" />
<span>Previous</span>
</PaginationLink>
)
PaginationPrevious.displayName = "PaginationPrevious"
const PaginationNext = ({
className,
...props
}: React.ComponentProps<typeof PaginationLink>) => (
<PaginationLink
aria-label="Go to next page"
size="default"
className={cn("gap-1 pr-2.5", className)}
{...props}
>
<span>Next</span>
<ChevronRight className="h-4 w-4" />
</PaginationLink>
)
PaginationNext.displayName = "PaginationNext"
const PaginationEllipsis = ({
className,
...props
}: React.ComponentProps<"span">) => (
<span
aria-hidden
className={cn("flex h-9 w-9 items-center justify-center", className)}
{...props}
>
<MoreHorizontal className="h-4 w-4" />
<span className="sr-only">More pages</span>
</span>
)
PaginationEllipsis.displayName = "PaginationEllipsis"
export {
Pagination,
PaginationContent,
PaginationEllipsis,
PaginationItem,
PaginationLink,
PaginationNext,
PaginationPrevious,
}

View File

@@ -1523,9 +1523,16 @@ class FileObjectAdapters:
return await sync_to_async(list)(FileObject.objects.filter(user=user, file_name__in=file_names))
@staticmethod
@arequire_valid_user
async def aget_all_file_objects(user: KhojUser):
return await sync_to_async(list)(FileObject.objects.filter(user=user))
@require_valid_user
async def aget_all_file_objects(user: KhojUser, start: int = 0, limit: int = 10):
query = FileObject.objects.filter(user=user).order_by("-updated_at")[start : start + limit]
return await sync_to_async(list)(query)
@staticmethod
@require_valid_user
async def aget_number_of_pages(user: KhojUser, limit: int = 10):
count = await FileObject.objects.filter(user=user).acount()
return math.ceil(count / limit)
@staticmethod
@arequire_valid_user

View File

@@ -0,0 +1,56 @@
from django.core.management.base import BaseCommand
from django.db.models import Exists, OuterRef
from khoj.database.models import Entry, FileObject
class Command(BaseCommand):
help = "Deletes FileObjects that have no associated Entries"
def add_arguments(self, parser):
parser.add_argument(
"--apply",
action="store_true",
help="Actually perform the deletion. Without this flag, only shows what would be deleted.",
)
def handle(self, *args, **options):
# Find FileObjects with no related entries using subquery
orphaned_files = FileObject.objects.annotate(
has_entries=Exists(Entry.objects.filter(file_object=OuterRef("pk")))
).filter(has_entries=False)
total_orphaned = orphaned_files.count()
mode = "DELETE" if options["apply"] else "DRY RUN"
self.stdout.write(f"[{mode}] Found {total_orphaned} orphaned FileObjects")
if total_orphaned == 0:
self.stdout.write("No orphaned FileObjects to process")
return
# Process in batches of 1000
batch_size = 1000
processed = 0
while processed < total_orphaned:
# Get batch of IDs to process
batch_ids = list(orphaned_files.values_list("id", flat=True)[:batch_size])
if not batch_ids:
break
if options["apply"]:
# Delete by ID to avoid slice/limit issues
count = FileObject.objects.filter(id__in=batch_ids).delete()[0]
processed += count
self.stdout.write(f"Deleted {processed}/{total_orphaned} orphaned FileObjects")
else:
processed += len(batch_ids)
self.stdout.write(f"Would delete {processed}/{total_orphaned} orphaned FileObjects")
# Re-query to get fresh state
orphaned_files = FileObject.objects.annotate(
has_entries=Exists(Entry.objects.filter(file_object=OuterRef("pk")))
).filter(has_entries=False)
action = "Deleted" if options["apply"] else "Would delete"
self.stdout.write(self.style.SUCCESS(f"{action} {processed} orphaned FileObjects"))

View File

@@ -0,0 +1,75 @@
# Generated by Django 5.0.10 on 2025-01-10 18:28
import django.db.models.deletion
from django.db import migrations, models
def migrate_entry_objects(apps, schema_editor):
Entry = apps.get_model("database", "Entry")
FileObject = apps.get_model("database", "FileObject")
db_alias = schema_editor.connection.alias
# Create lookup dictionary of all file objects
file_objects_map = {(fo.user_id, fo.file_name): fo for fo in FileObject.objects.using(db_alias).all()}
# Process entries in chunks of 1000
chunk_size = 1000
processed = 0
processed_entry_ids = set()
while True:
entries = list(
Entry.objects.using(db_alias)
.select_related("user")
.filter(file_object__isnull=True)
.exclude(id__in=processed_entry_ids)
.only("id", "user", "file_path")[:chunk_size]
)
if not entries:
break
processed_entry_ids.update([entry.id for entry in entries])
entries_to_update = []
for entry in entries:
try:
file_object = file_objects_map.get((entry.user_id, entry.file_path))
if file_object:
entry.file_object = file_object
entries_to_update.append(entry)
except Exception as e:
print(f"Error processing entry {entry.id}: {str(e)}")
continue
if entries_to_update:
Entry.objects.using(db_alias).bulk_update(entries_to_update, ["file_object"], batch_size=chunk_size)
processed += len(entries)
print(f"Processed {processed} entries")
def reverse_migration(apps, schema_editor):
pass
class Migration(migrations.Migration):
dependencies = [
("database", "0078_khojuser_email_verification_code_expiry"),
]
operations = [
migrations.AddField(
model_name="entry",
name="file_object",
field=models.ForeignKey(
blank=True,
default=None,
null=True,
on_delete=django.db.models.deletion.CASCADE,
to="database.fileobject",
),
),
migrations.RunPython(migrate_entry_objects, reverse_migration),
]

View File

@@ -0,0 +1,14 @@
# Generated by Django 5.0.10 on 2025-01-20 16:33
from typing import List
from django.db import migrations
class Migration(migrations.Migration):
dependencies = [
("database", "0079_entry_file_object"),
("database", "0080_speechtotextmodeloptions_ai_model_api"),
]
operations: List[str] = []

View File

@@ -330,6 +330,7 @@ class ProcessLock(DbBaseModel):
INDEX_CONTENT = "index_content"
SCHEDULED_JOB = "scheduled_job"
SCHEDULE_LEADER = "schedule_leader"
APPLY_MIGRATIONS = "apply_migrations"
# We need to make sure that some operations are thread-safe. To do so, add locks for potentially shared operations.
# For example, we need to make sure that only one process is updating the embeddings at a time.
@@ -672,6 +673,14 @@ class ReflectiveQuestion(DbBaseModel):
user = models.ForeignKey(KhojUser, on_delete=models.CASCADE, default=None, null=True, blank=True)
class FileObject(DbBaseModel):
# Contains the full text of a file that has associated Entry objects
file_name = models.CharField(max_length=400, default=None, null=True, blank=True)
raw_text = models.TextField()
user = models.ForeignKey(KhojUser, on_delete=models.CASCADE, default=None, null=True, blank=True)
agent = models.ForeignKey(Agent, on_delete=models.CASCADE, default=None, null=True, blank=True)
class Entry(DbBaseModel):
class EntryType(models.TextChoices):
IMAGE = "image"
@@ -703,20 +712,13 @@ class Entry(DbBaseModel):
hashed_value = models.CharField(max_length=100)
corpus_id = models.UUIDField(default=uuid.uuid4, editable=False)
search_model = models.ForeignKey(SearchModelConfig, on_delete=models.SET_NULL, default=None, null=True, blank=True)
file_object = models.ForeignKey(FileObject, on_delete=models.CASCADE, default=None, null=True, blank=True)
def save(self, *args, **kwargs):
if self.user and self.agent:
raise ValidationError("An Entry cannot be associated with both a user and an agent.")
class FileObject(DbBaseModel):
# Same as Entry but raw will be a much larger string
file_name = models.CharField(max_length=400, default=None, null=True, blank=True)
raw_text = models.TextField()
user = models.ForeignKey(KhojUser, on_delete=models.CASCADE, default=None, null=True, blank=True)
agent = models.ForeignKey(Agent, on_delete=models.CASCADE, default=None, null=True, blank=True)
class EntryDates(DbBaseModel):
date = models.DateField()
entry = models.ForeignKey(Entry, on_delete=models.CASCADE, related_name="embeddings_dates")

View File

@@ -152,8 +152,22 @@ class TextToEntries(ABC):
with timer("Generated embeddings for entries to add to database in", logger):
entries_to_process = [hash_to_current_entries[hashed_val] for hashed_val in hashes_to_process]
data_to_embed = [getattr(entry, key) for entry in entries_to_process]
modified_files = {entry.file for entry in entries_to_process}
embeddings += self.embeddings_model[model.name].embed_documents(data_to_embed)
file_to_file_object_map = {}
if file_to_text_map and modified_files:
with timer("Indexed text of modified file in", logger):
# create or update text of each updated file indexed on DB
for modified_file in modified_files:
raw_text = file_to_text_map[modified_file]
file_object = FileObjectAdapters.get_file_object_by_name(user, modified_file)
if file_object:
FileObjectAdapters.update_raw_text(file_object, raw_text)
else:
file_object = FileObjectAdapters.create_file_object(user, modified_file, raw_text)
file_to_file_object_map[modified_file] = file_object
added_entries: list[DbEntry] = []
with timer("Added entries to database in", logger):
num_items = len(hashes_to_process)
@@ -165,6 +179,7 @@ class TextToEntries(ABC):
batch_embeddings_to_create: List[DbEntry] = []
for entry_hash, new_entry in entry_batch:
entry = hash_to_current_entries[entry_hash]
file_object = file_to_file_object_map.get(entry.file, None)
batch_embeddings_to_create.append(
DbEntry(
user=user,
@@ -178,6 +193,7 @@ class TextToEntries(ABC):
hashed_value=entry_hash,
corpus_id=entry.corpus_id,
search_model=model,
file_object=file_object,
)
)
try:
@@ -190,19 +206,6 @@ class TextToEntries(ABC):
logger.error(f"Error adding entries to database:\n{batch_indexing_error}\n---\n{e}", exc_info=True)
logger.debug(f"Added {len(added_entries)} {file_type} entries to database")
if file_to_text_map:
with timer("Indexed text of modified file in", logger):
# get the set of modified files from added_entries
modified_files = {entry.file_path for entry in added_entries}
# create or update text of each updated file indexed on DB
for modified_file in modified_files:
raw_text = file_to_text_map[modified_file]
file_object = FileObjectAdapters.get_file_object_by_name(user, modified_file)
if file_object:
FileObjectAdapters.update_raw_text(file_object, raw_text)
else:
FileObjectAdapters.create_file_object(user, modified_file, raw_text)
new_dates = []
with timer("Indexed dates from added entries in", logger):
for added_entry in added_entries:

View File

@@ -22,6 +22,7 @@ from starlette.authentication import requires
from khoj.database import adapters
from khoj.database.adapters import (
EntryAdapters,
FileObjectAdapters,
get_user_github_config,
get_user_notion_config,
)
@@ -270,6 +271,8 @@ async def delete_content_files(
await EntryAdapters.adelete_entry_by_file(user, filename)
await FileObjectAdapters.adelete_file_object_by_name(user, filename)
return {"status": "ok"}
@@ -294,6 +297,8 @@ async def delete_content_file(
)
deleted_count = await EntryAdapters.adelete_entries_by_filenames(user, files.files)
for file in files.files:
await FileObjectAdapters.adelete_file_object_by_name(user, file)
return {"status": "ok", "deleted_count": deleted_count}
@@ -325,6 +330,77 @@ def get_content_types(request: Request, client: Optional[str] = None):
return list(configured_content_types & all_content_types)
@api_content.get("/files", response_model=Dict[str, str])
@requires(["authenticated"])
async def get_all_files(
request: Request, client: Optional[str] = None, truncated: Optional[bool] = True, page: int = 0
):
user = request.user.object
update_telemetry_state(
request=request,
telemetry_type="api",
api="get_all_filenames",
client=client,
)
files_data = []
page_size = 10
file_objects = await FileObjectAdapters.aget_all_file_objects(user, start=page * page_size, limit=page_size)
num_pages = await FileObjectAdapters.aget_number_of_pages(user, page_size)
for file_object in file_objects:
files_data.append(
{
"file_name": file_object.file_name,
"raw_text": file_object.raw_text[:1000] if truncated else file_object.raw_text,
"updated_at": str(file_object.updated_at),
}
)
data_packet = {
"files": files_data,
"num_pages": num_pages,
}
return Response(content=json.dumps(data_packet), media_type="application/json", status_code=200)
@api_content.get("/file", response_model=Dict[str, str])
@requires(["authenticated"])
async def get_file_object(
request: Request,
file_name: str,
client: Optional[str] = None,
):
user = request.user.object
file_object = (await FileObjectAdapters.aget_file_objects_by_name(user, file_name))[0]
if not file_object:
return Response(
content=json.dumps({"error": "File not found"}),
media_type="application/json",
status_code=404,
)
update_telemetry_state(
request=request,
telemetry_type="api",
api="get_file",
client=client,
)
return Response(
content=json.dumps(
{"id": file_object.id, "file_name": file_object.file_name, "raw_text": file_object.raw_text}
),
media_type="application/json",
status_code=200,
)
@api_content.get("/{content_source}", response_model=List[str])
@requires(["authenticated"])
async def get_content_source(