gmakstutis/openrag
1
0
Fork 0
Code Issues Pull requests Projects Releases Packages Wiki Activity Actions

Merge branch 'main' into feat-google-drive-folder-select

Browse source
This commit is contained in:
Eric Hare 2025-10-06 12:07:54 -07:00 • committed by GitHub
commit 22dc5b9b65
No known key found for this signature in database
GPG key ID: B5690EEEBB952194
27 changed files with 2241 additions and 1478 deletions
Show stats Download patch file Download diff file Expand all files Collapse all files

2
.gitignore vendored
View file

@ -21,3 +21,5 @@ wheels/
.DS_Store .DS_Store
config/ config/
.docling.pid

2
docker-compose-cpu.yml
View file

@ -43,7 +43,7 @@ services:
# build: # build:
# context: . # context: .
# dockerfile: Dockerfile.backend # dockerfile: Dockerfile.backend
# container_name: openrag-backend container_name: openrag-backend
depends_on: depends_on:
- langflow - langflow
environment: environment:

2
docker-compose.yml
View file

@ -43,7 +43,7 @@ services:
# build: # build:
# context: . # context: .
# dockerfile: Dockerfile.backend # dockerfile: Dockerfile.backend
# container_name: openrag-backend container_name: openrag-backend
depends_on: depends_on:
- langflow - langflow
environment: environment:

66
frontend/components/duplicate-handling-dialog.tsx Normal file
View file

@ -0,0 +1,66 @@
"use client";
import { RotateCcw } from "lucide-react";
import type React from "react";
import { Button } from "./ui/button";
import {
Dialog,
DialogContent,
DialogDescription,
DialogFooter,
DialogHeader,
DialogTitle,
} from "./ui/dialog";
interface DuplicateHandlingDialogProps {
open: boolean;
onOpenChange: (open: boolean) => void;
onOverwrite: () => void | Promise<void>;
isLoading?: boolean;
}
export const DuplicateHandlingDialog: React.FC<
DuplicateHandlingDialogProps
> = ({ open, onOpenChange, onOverwrite, isLoading = false }) => {
const handleOverwrite = async () => {
await onOverwrite();
onOpenChange(false);
};
return (
<Dialog open={open} onOpenChange={onOpenChange}>
<DialogContent className="sm:max-w-[450px]">
<DialogHeader>
<DialogTitle>Overwrite document</DialogTitle>
<DialogDescription className="pt-2 text-muted-foreground">
Overwriting will replace the existing document with another version.
This can't be undone.
</DialogDescription>
</DialogHeader>
<DialogFooter className="flex-row gap-2 justify-end">
<Button
type="button"
variant="ghost"
onClick={() => onOpenChange(false)}
disabled={isLoading}
size="sm"
>
Cancel
</Button>
<Button
type="button"
variant="default"
size="sm"
onClick={handleOverwrite}
disabled={isLoading}
className="flex items-center gap-2 !bg-accent-amber-foreground hover:!bg-foreground text-primary-foreground"
>
<RotateCcw className="h-3.5 w-3.5" />
Overwrite
</Button>
</DialogFooter>
</DialogContent>
</Dialog>
);
};

1189
frontend/components/knowledge-dropdown.tsx
View file

File diff suppressed because it is too large Load diff

2
frontend/components/logo/ibm-logo.tsx
View file

@ -9,7 +9,7 @@ export default function IBMLogo(props: React.SVGProps<SVGSVGElement>) {
{...props} {...props}
> >
<title>IBM watsonx.ai Logo</title> <title>IBM watsonx.ai Logo</title>
<g clip-path="url(#clip0_2620_2081)"> <g clipPath="url(#clip0_2620_2081)">
<path <path
d="M13 12.0007C12.4477 12.0007 12 12.4484 12 13.0007C12 13.0389 12.0071 13.0751 12.0112 13.1122C10.8708 14.0103 9.47165 14.5007 8 14.5007C5.86915 14.5007 4 12.5146 4 10.2507C4 7.90722 5.9065 6.00072 8.25 6.00072H8.5V5.00072H8.25C5.3552 5.00072 3 7.35592 3 10.2507C3 11.1927 3.2652 12.0955 3.71855 12.879C2.3619 11.6868 1.5 9.94447 1.5 8.00072C1.5 6.94312 1.74585 5.93432 2.23095 5.00292L1.34375 4.54102C0.79175 5.60157 0.5 6.79787 0.5 8.00072C0.5 12.1362 3.8645 15.5007 8 15.5007C9.6872 15.5007 11.2909 14.9411 12.6024 13.9176C12.7244 13.9706 12.8586 14.0007 13 14.0007C13.5523 14.0007 14 13.553 14 13.0007C14 12.4484 13.5523 12.0007 13 12.0007Z" d="M13 12.0007C12.4477 12.0007 12 12.4484 12 13.0007C12 13.0389 12.0071 13.0751 12.0112 13.1122C10.8708 14.0103 9.47165 14.5007 8 14.5007C5.86915 14.5007 4 12.5146 4 10.2507C4 7.90722 5.9065 6.00072 8.25 6.00072H8.5V5.00072H8.25C5.3552 5.00072 3 7.35592 3 10.2507C3 11.1927 3.2652 12.0955 3.71855 12.879C2.3619 11.6868 1.5 9.94447 1.5 8.00072C1.5 6.94312 1.74585 5.93432 2.23095 5.00292L1.34375 4.54102C0.79175 5.60157 0.5 6.79787 0.5 8.00072C0.5 12.1362 3.8645 15.5007 8 15.5007C9.6872 15.5007 11.2909 14.9411 12.6024 13.9176C12.7244 13.9706 12.8586 14.0007 13 14.0007C13.5523 14.0007 14 13.553 14 13.0007C14 12.4484 13.5523 12.0007 13 12.0007Z"
fill="currentColor" fill="currentColor"

2
frontend/components/ui/input.tsx
View file

@ -44,7 +44,7 @@ const Input = React.forwardRef<HTMLInputElement, InputProps>(
placeholder={placeholder} placeholder={placeholder}
className={cn( className={cn(
"primary-input", "primary-input",
icon && "pl-9", icon && "!pl-9",
type === "password" && "!pr-8", type === "password" && "!pr-8",
icon ? inputClassName : className icon ? inputClassName : className
)} )}

47
frontend/src/app/api/mutations/useCancelTaskMutation.ts Normal file
View file

@ -0,0 +1,47 @@
import {
type UseMutationOptions,
useMutation,
useQueryClient,
} from "@tanstack/react-query";
export interface CancelTaskRequest {
taskId: string;
}
export interface CancelTaskResponse {
status: string;
task_id: string;
}
export const useCancelTaskMutation = (
options?: Omit<
UseMutationOptions<CancelTaskResponse, Error, CancelTaskRequest>,
"mutationFn"
>
) => {
const queryClient = useQueryClient();
async function cancelTask(
variables: CancelTaskRequest,
): Promise<CancelTaskResponse> {
const response = await fetch(`/api/tasks/${variables.taskId}/cancel`, {
method: "POST",
});
if (!response.ok) {
const errorData = await response.json().catch(() => ({}));
throw new Error(errorData.error || "Failed to cancel task");
}
return response.json();
}
return useMutation({
mutationFn: cancelTask,
onSuccess: () => {
// Invalidate tasks query to refresh the list
queryClient.invalidateQueries({ queryKey: ["tasks"] });
},
...options,
});
};

3
frontend/src/app/api/queries/useGetNudgesQuery.ts
View file

@ -7,9 +7,6 @@ import {
type Nudge = string; type Nudge = string;
const DEFAULT_NUDGES = [ const DEFAULT_NUDGES = [
"Show me this quarter's top 10 deals",
"Summarize recent client interactions",
"Search OpenSearch for mentions of our competitors",
]; ];
export const useGetNudgesQuery = ( export const useGetNudgesQuery = (

79
frontend/src/app/api/queries/useGetTasksQuery.ts Normal file
View file

@ -0,0 +1,79 @@
import {
type UseQueryOptions,
useQuery,
useQueryClient,
} from "@tanstack/react-query";
export interface Task {
task_id: string;
status:
| "pending"
| "running"
| "processing"
| "completed"
| "failed"
| "error";
total_files?: number;
processed_files?: number;
successful_files?: number;
failed_files?: number;
running_files?: number;
pending_files?: number;
created_at: string;
updated_at: string;
duration_seconds?: number;
result?: Record<string, unknown>;
error?: string;
files?: Record<string, Record<string, unknown>>;
}
export interface TasksResponse {
tasks: Task[];
}
export const useGetTasksQuery = (
options?: Omit<UseQueryOptions<Task[]>, "queryKey" | "queryFn">
) => {
const queryClient = useQueryClient();
async function getTasks(): Promise<Task[]> {
const response = await fetch("/api/tasks");
if (!response.ok) {
throw new Error("Failed to fetch tasks");
}
const data: TasksResponse = await response.json();
return data.tasks || [];
}
const queryResult = useQuery(
{
queryKey: ["tasks"],
queryFn: getTasks,
refetchInterval: (query) => {
// Only poll if there are tasks with pending or running status
const data = query.state.data;
if (!data || data.length === 0) {
return false; // Stop polling if no tasks
}
const hasActiveTasks = data.some(
(task: Task) =>
task.status === "pending" ||
task.status === "running" ||
task.status === "processing"
);
return hasActiveTasks ? 3000 : false; // Poll every 3 seconds if active tasks exist
},
refetchIntervalInBackground: true,
staleTime: 0, // Always consider data stale to ensure fresh updates
gcTime: 5 * 60 * 1000, // Keep in cache for 5 minutes
...options,
},
queryClient,
);
return queryResult;
};

623
frontend/src/app/knowledge/chunks/page.tsx
View file

@ -1,201 +1,204 @@
"use client"; "use client";
import { ArrowLeft, Check, Copy, Loader2, Search, X } from "lucide-react"; import { ArrowLeft, Check, Copy, Loader2, Search, X } from "lucide-react";
import { Suspense, useCallback, useEffect, useMemo, useState } from "react";
import { useRouter, useSearchParams } from "next/navigation"; import { useRouter, useSearchParams } from "next/navigation";
import { Suspense, useCallback, useEffect, useMemo, useState } from "react";
// import { Label } from "@/components/ui/label";
// import { Checkbox } from "@/components/ui/checkbox";
import { filterAccentClasses } from "@/components/knowledge-filter-panel";
import { ProtectedRoute } from "@/components/protected-route"; import { ProtectedRoute } from "@/components/protected-route";
import { Button } from "@/components/ui/button"; import { Button } from "@/components/ui/button";
import { Checkbox } from "@/components/ui/checkbox";
import { Input } from "@/components/ui/input";
import { Label } from "@/components/ui/label";
import { useKnowledgeFilter } from "@/contexts/knowledge-filter-context"; import { useKnowledgeFilter } from "@/contexts/knowledge-filter-context";
import { useLayout } from "@/contexts/layout-context"; import { useLayout } from "@/contexts/layout-context";
import { useTask } from "@/contexts/task-context"; import { useTask } from "@/contexts/task-context";
import { import {
type ChunkResult, type ChunkResult,
type File, type File,
useGetSearchQuery, useGetSearchQuery,
} from "../../api/queries/useGetSearchQuery"; } from "../../api/queries/useGetSearchQuery";
// import { Label } from "@/components/ui/label";
// import { Checkbox } from "@/components/ui/checkbox";
import { filterAccentClasses } from "@/components/knowledge-filter-panel";
const getFileTypeLabel = (mimetype: string) => { const getFileTypeLabel = (mimetype: string) => {
if (mimetype === "application/pdf") return "PDF"; if (mimetype === "application/pdf") return "PDF";
if (mimetype === "text/plain") return "Text"; if (mimetype === "text/plain") return "Text";
if (mimetype === "application/msword") return "Word Document"; if (mimetype === "application/msword") return "Word Document";
return "Unknown"; return "Unknown";
}; };
function ChunksPageContent() { function ChunksPageContent() {
const router = useRouter(); const router = useRouter();
const searchParams = useSearchParams(); const searchParams = useSearchParams();
const { selectedFilter, setSelectedFilter, parsedFilterData, isPanelOpen } = const { selectedFilter, setSelectedFilter, parsedFilterData, isPanelOpen } =
useKnowledgeFilter(); useKnowledgeFilter();
const { isMenuOpen } = useTask(); const { isMenuOpen } = useTask();
const { totalTopOffset } = useLayout(); const { totalTopOffset } = useLayout();
const filename = searchParams.get("filename"); const filename = searchParams.get("filename");
const [chunks, setChunks] = useState<ChunkResult[]>([]); const [chunks, setChunks] = useState<ChunkResult[]>([]);
const [chunksFilteredByQuery, setChunksFilteredByQuery] = useState< const [chunksFilteredByQuery, setChunksFilteredByQuery] = useState<
ChunkResult[] ChunkResult[]
>([]); >([]);
const [selectedChunks, setSelectedChunks] = useState<Set<number>>(new Set()); const [selectedChunks, setSelectedChunks] = useState<Set<number>>(new Set());
const [activeCopiedChunkIndex, setActiveCopiedChunkIndex] = useState< const [activeCopiedChunkIndex, setActiveCopiedChunkIndex] = useState<
number | null number | null
>(null); >(null);
// Calculate average chunk length // Calculate average chunk length
const averageChunkLength = useMemo( const averageChunkLength = useMemo(
() => () =>
chunks.reduce((acc, chunk) => acc + chunk.text.length, 0) / chunks.reduce((acc, chunk) => acc + chunk.text.length, 0) /
chunks.length || 0, chunks.length || 0,
[chunks] [chunks],
); );
const [selectAll, setSelectAll] = useState(false); const [selectAll, setSelectAll] = useState(false);
const [queryInputText, setQueryInputText] = useState( const [queryInputText, setQueryInputText] = useState(
parsedFilterData?.query ?? "" parsedFilterData?.query ?? "",
); );
// Use the same search query as the knowledge page, but we'll filter for the specific file // Use the same search query as the knowledge page, but we'll filter for the specific file
const { data = [], isFetching } = useGetSearchQuery("*", parsedFilterData); const { data = [], isFetching } = useGetSearchQuery("*", parsedFilterData);
useEffect(() => { useEffect(() => {
if (queryInputText === "") { if (queryInputText === "") {
setChunksFilteredByQuery(chunks); setChunksFilteredByQuery(chunks);
} else { } else {
setChunksFilteredByQuery( setChunksFilteredByQuery(
chunks.filter((chunk) => chunks.filter((chunk) =>
chunk.text.toLowerCase().includes(queryInputText.toLowerCase()) chunk.text.toLowerCase().includes(queryInputText.toLowerCase()),
) ),
); );
} }
}, [queryInputText, chunks]); }, [queryInputText, chunks]);
const handleCopy = useCallback((text: string, index: number) => { const handleCopy = useCallback((text: string, index: number) => {
// Trim whitespace and remove new lines/tabs for cleaner copy // Trim whitespace and remove new lines/tabs for cleaner copy
navigator.clipboard.writeText(text.trim().replace(/[\n\r\t]/gm, "")); navigator.clipboard.writeText(text.trim().replace(/[\n\r\t]/gm, ""));
setActiveCopiedChunkIndex(index); setActiveCopiedChunkIndex(index);
setTimeout(() => setActiveCopiedChunkIndex(null), 10 * 1000); // 10 seconds setTimeout(() => setActiveCopiedChunkIndex(null), 10 * 1000); // 10 seconds
}, []); }, []);
const fileData = (data as File[]).find( const fileData = (data as File[]).find(
(file: File) => file.filename === filename (file: File) => file.filename === filename,
); );
// Extract chunks for the specific file // Extract chunks for the specific file
useEffect(() => { useEffect(() => {
if (!filename || !(data as File[]).length) { if (!filename || !(data as File[]).length) {
setChunks([]); setChunks([]);
return; return;
} }
setChunks( setChunks(
fileData?.chunks?.map((chunk, i) => ({ ...chunk, index: i + 1 })) || [] fileData?.chunks?.map((chunk, i) => ({ ...chunk, index: i + 1 })) || [],
); );
}, [data, filename]); }, [data, filename]);
// Set selected state for all checkboxes when selectAll changes // Set selected state for all checkboxes when selectAll changes
useEffect(() => { useEffect(() => {
if (selectAll) { if (selectAll) {
setSelectedChunks(new Set(chunks.map((_, index) => index))); setSelectedChunks(new Set(chunks.map((_, index) => index)));
} else { } else {
setSelectedChunks(new Set()); setSelectedChunks(new Set());
} }
}, [selectAll, setSelectedChunks, chunks]); }, [selectAll, setSelectedChunks, chunks]);
const handleBack = useCallback(() => { const handleBack = useCallback(() => {
router.push("/knowledge"); router.push("/knowledge");
}, [router]); }, [router]);
// const handleChunkCardCheckboxChange = useCallback( // const handleChunkCardCheckboxChange = useCallback(
// (index: number) => { // (index: number) => {
// setSelectedChunks((prevSelected) => { // setSelectedChunks((prevSelected) => {
// const newSelected = new Set(prevSelected); // const newSelected = new Set(prevSelected);
// if (newSelected.has(index)) { // if (newSelected.has(index)) {
// newSelected.delete(index); // newSelected.delete(index);
// } else { // } else {
// newSelected.add(index); // newSelected.add(index);
// } // }
// return newSelected; // return newSelected;
// }); // });
// }, // },
// [setSelectedChunks] // [setSelectedChunks]
// ); // );
if (!filename) { if (!filename) {
return ( return (
<div className="flex items-center justify-center h-64"> <div className="flex items-center justify-center h-64">
<div className="text-center"> <div className="text-center">
<Search className="h-12 w-12 mx-auto mb-4 text-muted-foreground/50" /> <Search className="h-12 w-12 mx-auto mb-4 text-muted-foreground/50" />
<p className="text-lg text-muted-foreground">No file specified</p> <p className="text-lg text-muted-foreground">No file specified</p>
<p className="text-sm text-muted-foreground/70 mt-2"> <p className="text-sm text-muted-foreground/70 mt-2">
Please select a file from the knowledge page Please select a file from the knowledge page
</p> </p>
</div> </div>
</div> </div>
); );
} }
return ( return (
<div <div
className={`fixed inset-0 md:left-72 flex flex-row transition-all duration-300 ${ className={`fixed inset-0 md:left-72 flex flex-row transition-all duration-300 ${
isMenuOpen && isPanelOpen isMenuOpen && isPanelOpen
? "md:right-[704px]" ? "md:right-[704px]"
: // Both open: 384px (menu) + 320px (KF panel) : // Both open: 384px (menu) + 320px (KF panel)
isMenuOpen isMenuOpen
? "md:right-96" ? "md:right-96"
: // Only menu open: 384px : // Only menu open: 384px
isPanelOpen isPanelOpen
? "md:right-80" ? "md:right-80"
: // Only KF panel open: 320px : // Only KF panel open: 320px
"md:right-6" // Neither open: 24px "md:right-6" // Neither open: 24px
}`} }`}
style={{ top: `${totalTopOffset}px` }} style={{ top: `${totalTopOffset}px` }}
> >
<div className="flex-1 flex flex-col min-h-0 px-6 py-6"> <div className="flex-1 flex flex-col min-h-0 px-6 py-6">
{/* Header */} {/* Header */}
<div className="flex flex-col mb-6"> <div className="flex flex-col mb-6">
<div className="flex flex-row items-center gap-3 mb-6"> <div className="flex flex-row items-center gap-3 mb-6">
<Button variant="ghost" onClick={handleBack} size="sm"> <Button variant="ghost" onClick={handleBack} size="sm">
<ArrowLeft size={24} /> <ArrowLeft size={24} />
</Button> </Button>
<h1 className="text-lg font-semibold"> <h1 className="text-lg font-semibold">
{/* Removes file extension from filename */} {/* Removes file extension from filename */}
{filename.replace(/\.[^/.]+$/, "")} {filename.replace(/\.[^/.]+$/, "")}
</h1> </h1>
</div> </div>
<div className="flex flex-col items-start mt-2"> <div className="flex flex-col items-start mt-2">
<div className="flex-1 flex items-center gap-2 w-full max-w-[640px]"> <div className="flex-1 flex items-center gap-2 w-full max-w-[640px]">
<div className="primary-input min-h-10 !flex items-center flex-nowrap focus-within:border-foreground transition-colors !p-[0.3rem]"> <div className="primary-input min-h-10 !flex items-center flex-nowrap focus-within:border-foreground transition-colors !p-[0.3rem]">
{selectedFilter?.name && ( {selectedFilter?.name && (
<div <div
className={`flex items-center gap-1 h-full px-1.5 py-0.5 mr-1 rounded max-w-[25%] ${ className={`flex items-center gap-1 h-full px-1.5 py-0.5 mr-1 rounded max-w-[25%] ${
filterAccentClasses[parsedFilterData?.color || "zinc"] filterAccentClasses[parsedFilterData?.color || "zinc"]
}`} }`}
> >
<span className="truncate">{selectedFilter?.name}</span> <span className="truncate">{selectedFilter?.name}</span>
<X <X
aria-label="Remove filter" aria-label="Remove filter"
className="h-4 w-4 flex-shrink-0 cursor-pointer" className="h-4 w-4 flex-shrink-0 cursor-pointer"
onClick={() => setSelectedFilter(null)} onClick={() => setSelectedFilter(null)}
/> />
</div> </div>
)} )}
<Search <Search
className="h-4 w-4 ml-1 flex-shrink-0 text-placeholder-foreground" className="h-4 w-4 ml-1 flex-shrink-0 text-placeholder-foreground"
strokeWidth={1.5} strokeWidth={1.5}
/> />
<input <input
className="bg-transparent w-full h-full ml-2 focus:outline-none focus-visible:outline-none font-mono placeholder:font-mono" className="bg-transparent w-full h-full ml-2 focus:outline-none focus-visible:outline-none font-mono placeholder:font-mono"
name="search-query" name="search-query"
id="search-query" id="search-query"
type="text" type="text"
placeholder="Enter your search query..." placeholder="Enter your search query..."
onChange={(e) => setQueryInputText(e.target.value)} onChange={(e) => setQueryInputText(e.target.value)}
value={queryInputText} value={queryInputText}
/> />
</div> </div>
</div> </div>
{/* <div className="flex items-center pl-4 gap-2"> {/* <div className="flex items-center pl-4 gap-2">
<Checkbox <Checkbox
id="selectAllChunks" id="selectAllChunks"
checked={selectAll} checked={selectAll}
@ -210,39 +213,39 @@ function ChunksPageContent() {
Select all Select all
</Label> </Label>
</div> */} </div> */}
</div> </div>
</div> </div>
{/* Content Area - matches knowledge page structure */} {/* Content Area - matches knowledge page structure */}
<div className="flex-1 overflow-scroll pr-6"> <div className="flex-1 overflow-scroll pr-6">
{isFetching ? ( {isFetching ? (
<div className="flex items-center justify-center h-64"> <div className="flex items-center justify-center h-64">
<div className="text-center"> <div className="text-center">
<Loader2 className="h-12 w-12 mx-auto mb-4 text-muted-foreground/50 animate-spin" /> <Loader2 className="h-12 w-12 mx-auto mb-4 text-muted-foreground/50 animate-spin" />
<p className="text-lg text-muted-foreground"> <p className="text-lg text-muted-foreground">
Loading chunks... Loading chunks...
</p> </p>
</div> </div>
</div> </div>
) : chunks.length === 0 ? ( ) : chunks.length === 0 ? (
<div className="flex items-center justify-center h-64"> <div className="flex items-center justify-center h-64">
<div className="text-center"> <div className="text-center">
<p className="text-xl font-semibold mb-2">No knowledge</p> <p className="text-xl font-semibold mb-2">No knowledge</p>
<p className="text-sm text-secondary-foreground"> <p className="text-sm text-secondary-foreground">
Clear the knowledge filter or return to the knowledge page Clear the knowledge filter or return to the knowledge page
</p> </p>
</div> </div>
</div> </div>
) : ( ) : (
<div className="space-y-4 pb-6"> <div className="space-y-4 pb-6">
{chunksFilteredByQuery.map((chunk, index) => ( {chunksFilteredByQuery.map((chunk, index) => (
<div <div
key={chunk.filename + index} key={chunk.filename + index}
className="bg-muted rounded-lg p-4 border border-border/50" className="bg-muted rounded-lg p-4 border border-border/50"
> >
<div className="flex items-center justify-between mb-2"> <div className="flex items-center justify-between mb-2">
<div className="flex items-center gap-3"> <div className="flex items-center gap-3">
{/* <div> {/* <div>
<Checkbox <Checkbox
checked={selectedChunks.has(index)} checked={selectedChunks.has(index)}
onCheckedChange={() => onCheckedChange={() =>
@ -250,73 +253,73 @@ function ChunksPageContent() {
} }
/> />
</div> */} </div> */}
<span className="text-sm font-bold"> <span className="text-sm font-bold">
Chunk {chunk.index} Chunk {chunk.index}
</span> </span>
<span className="bg-background p-1 rounded text-xs text-muted-foreground/70"> <span className="bg-background p-1 rounded text-xs text-muted-foreground/70">
{chunk.text.length} chars {chunk.text.length} chars
</span> </span>
<div className="py-1"> <div className="py-1">
<Button <Button
onClick={() => handleCopy(chunk.text, index)} onClick={() => handleCopy(chunk.text, index)}
variant="ghost" variant="ghost"
size="sm" size="sm"
> >
{activeCopiedChunkIndex === index ? ( {activeCopiedChunkIndex === index ? (
<Check className="text-muted-foreground" /> <Check className="text-muted-foreground" />
) : ( ) : (
<Copy className="text-muted-foreground" /> <Copy className="text-muted-foreground" />
)} )}
</Button> </Button>
</div> </div>
</div> </div>
<span className="bg-background p-1 rounded text-xs text-muted-foreground/70"> <span className="bg-background p-1 rounded text-xs text-muted-foreground/70">
{chunk.score.toFixed(2)} score {chunk.score.toFixed(2)} score
</span> </span>
{/* TODO: Update to use active toggle */} {/* TODO: Update to use active toggle */}
{/* <span className="px-2 py-1 text-green-500"> {/* <span className="px-2 py-1 text-green-500">
<Switch <Switch
className="ml-2 bg-green-500" className="ml-2 bg-green-500"
checked={true} checked={true}
/> />
Active Active
</span> */} </span> */}
</div> </div>
<blockquote className="text-sm text-muted-foreground leading-relaxed ml-1.5"> <blockquote className="text-sm text-muted-foreground leading-relaxed ml-1.5">
{chunk.text} {chunk.text}
</blockquote> </blockquote>
</div> </div>
))} ))}
</div> </div>
)} )}
</div> </div>
</div> </div>
{/* Right panel - Summary (TODO), Technical details, */} {/* Right panel - Summary (TODO), Technical details, */}
{chunks.length > 0 && ( {chunks.length > 0 && (
<div className="w-[320px] py-20 px-2"> <div className="w-[320px] py-20 px-2">
<div className="mb-8"> <div className="mb-8">
<h2 className="text-xl font-semibold mt-3 mb-4"> <h2 className="text-xl font-semibold mt-3 mb-4">
Technical details Technical details
</h2> </h2>
<dl> <dl>
<div className="sm:grid sm:grid-cols-3 sm:gap-4 sm:px-0 mb-2.5"> <div className="sm:grid sm:grid-cols-3 sm:gap-4 sm:px-0 mb-2.5">
<dt className="text-sm/6 text-muted-foreground"> <dt className="text-sm/6 text-muted-foreground">
Total chunks Total chunks
</dt> </dt>
<dd className="mt-1 text-sm/6 text-gray-100 sm:col-span-2 sm:mt-0"> <dd className="mt-1 text-sm/6 text-gray-100 sm:col-span-2 sm:mt-0">
{chunks.length} {chunks.length}
</dd> </dd>
</div> </div>
<div className="sm:grid sm:grid-cols-3 sm:gap-4 sm:px-0 mb-2.5"> <div className="sm:grid sm:grid-cols-3 sm:gap-4 sm:px-0 mb-2.5">
<dt className="text-sm/6 text-muted-foreground">Avg length</dt> <dt className="text-sm/6 text-muted-foreground">Avg length</dt>
<dd className="mt-1 text-sm/6 text-gray-100 sm:col-span-2 sm:mt-0"> <dd className="mt-1 text-sm/6 text-gray-100 sm:col-span-2 sm:mt-0">
{averageChunkLength.toFixed(0)} chars {averageChunkLength.toFixed(0)} chars
</dd> </dd>
</div> </div>
{/* TODO: Uncomment after data is available */} {/* TODO: Uncomment after data is available */}
{/* <div className="sm:grid sm:grid-cols-3 sm:gap-4 sm:px-0 mb-2.5"> {/* <div className="sm:grid sm:grid-cols-3 sm:gap-4 sm:px-0 mb-2.5">
<dt className="text-sm/6 text-muted-foreground">Process time</dt> <dt className="text-sm/6 text-muted-foreground">Process time</dt>
<dd className="mt-1 text-sm/6 text-gray-100 sm:col-span-2 sm:mt-0"> <dd className="mt-1 text-sm/6 text-gray-100 sm:col-span-2 sm:mt-0">
</dd> </dd>
@ -326,79 +329,79 @@ function ChunksPageContent() {
<dd className="mt-1 text-sm/6 text-gray-100 sm:col-span-2 sm:mt-0"> <dd className="mt-1 text-sm/6 text-gray-100 sm:col-span-2 sm:mt-0">
</dd> </dd>
</div> */} </div> */}
</dl> </dl>
</div> </div>
<div className="mb-8"> <div className="mb-8">
<h2 className="text-xl font-semibold mt-2 mb-3"> <h2 className="text-xl font-semibold mt-2 mb-3">
Original document Original document
</h2> </h2>
<dl> <dl>
{/* <div className="sm:grid sm:grid-cols-3 sm:gap-4 sm:px-0 mb-2.5"> {/* <div className="sm:grid sm:grid-cols-3 sm:gap-4 sm:px-0 mb-2.5">
<dt className="text-sm/6 text-muted-foreground">Name</dt> <dt className="text-sm/6 text-muted-foreground">Name</dt>
<dd className="mt-1 text-sm/6 text-gray-100 sm:col-span-2 sm:mt-0"> <dd className="mt-1 text-sm/6 text-gray-100 sm:col-span-2 sm:mt-0">
{fileData?.filename} {fileData?.filename}
</dd> </dd>
</div> */} </div> */}
<div className="sm:grid sm:grid-cols-3 sm:gap-4 sm:px-0 mb-2.5"> <div className="sm:grid sm:grid-cols-3 sm:gap-4 sm:px-0 mb-2.5">
<dt className="text-sm/6 text-muted-foreground">Type</dt> <dt className="text-sm/6 text-muted-foreground">Type</dt>
<dd className="mt-1 text-sm/6 text-gray-100 sm:col-span-2 sm:mt-0"> <dd className="mt-1 text-sm/6 text-gray-100 sm:col-span-2 sm:mt-0">
{fileData ? getFileTypeLabel(fileData.mimetype) : "Unknown"} {fileData ? getFileTypeLabel(fileData.mimetype) : "Unknown"}
</dd> </dd>
</div> </div>
<div className="sm:grid sm:grid-cols-3 sm:gap-4 sm:px-0 mb-2.5"> <div className="sm:grid sm:grid-cols-3 sm:gap-4 sm:px-0 mb-2.5">
<dt className="text-sm/6 text-muted-foreground">Size</dt> <dt className="text-sm/6 text-muted-foreground">Size</dt>
<dd className="mt-1 text-sm/6 text-gray-100 sm:col-span-2 sm:mt-0"> <dd className="mt-1 text-sm/6 text-gray-100 sm:col-span-2 sm:mt-0">
{fileData?.size {fileData?.size
? `${Math.round(fileData.size / 1024)} KB` ? `${Math.round(fileData.size / 1024)} KB`
: "Unknown"} : "Unknown"}
</dd> </dd>
</div> </div>
{/* <div className="sm:grid sm:grid-cols-3 sm:gap-4 sm:px-0 mb-2.5"> {/* <div className="sm:grid sm:grid-cols-3 sm:gap-4 sm:px-0 mb-2.5">
<dt className="text-sm/6 text-muted-foreground">Uploaded</dt> <dt className="text-sm/6 text-muted-foreground">Uploaded</dt>
<dd className="mt-1 text-sm/6 text-gray-100 sm:col-span-2 sm:mt-0"> <dd className="mt-1 text-sm/6 text-gray-100 sm:col-span-2 sm:mt-0">
N/A N/A
</dd> </dd>
</div> */} </div> */}
{/* TODO: Uncomment after data is available */} {/* TODO: Uncomment after data is available */}
{/* <div className="sm:grid sm:grid-cols-3 sm:gap-4 sm:px-0 mb-2.5"> {/* <div className="sm:grid sm:grid-cols-3 sm:gap-4 sm:px-0 mb-2.5">
<dt className="text-sm/6 text-muted-foreground">Source</dt> <dt className="text-sm/6 text-muted-foreground">Source</dt>
<dd className="mt-1 text-sm/6 text-gray-100 sm:col-span-2 sm:mt-0"></dd> <dd className="mt-1 text-sm/6 text-gray-100 sm:col-span-2 sm:mt-0"></dd>
</div> */} </div> */}
{/* <div className="sm:grid sm:grid-cols-3 sm:gap-4 sm:px-0 mb-2.5"> {/* <div className="sm:grid sm:grid-cols-3 sm:gap-4 sm:px-0 mb-2.5">
<dt className="text-sm/6 text-muted-foreground">Updated</dt> <dt className="text-sm/6 text-muted-foreground">Updated</dt>
<dd className="mt-1 text-sm/6 text-gray-100 sm:col-span-2 sm:mt-0"> <dd className="mt-1 text-sm/6 text-gray-100 sm:col-span-2 sm:mt-0">
N/A N/A
</dd> </dd>
</div> */} </div> */}
</dl> </dl>
</div> </div>
</div> </div>
)} )}
</div> </div>
); );
} }
function ChunksPage() { function ChunksPage() {
return ( return (
<Suspense <Suspense
fallback={ fallback={
<div className="flex items-center justify-center h-64"> <div className="flex items-center justify-center h-64">
<div className="text-center"> <div className="text-center">
<Loader2 className="h-12 w-12 mx-auto mb-4 text-muted-foreground/50 animate-spin" /> <Loader2 className="h-12 w-12 mx-auto mb-4 text-muted-foreground/50 animate-spin" />
<p className="text-lg text-muted-foreground">Loading...</p> <p className="text-lg text-muted-foreground">Loading...</p>
</div> </div>
</div> </div>
} }
> >
<ChunksPageContent /> <ChunksPageContent />
</Suspense> </Suspense>
); );
} }
export default function ProtectedChunksPage() { export default function ProtectedChunksPage() {
return ( return (
<ProtectedRoute> <ProtectedRoute>
<ChunksPage /> <ChunksPage />
</ProtectedRoute> </ProtectedRoute>
); );
} }

148
frontend/src/app/knowledge/page.tsx
View file

@ -1,10 +1,24 @@
"use client"; "use client";
import type { ColDef } from "ag-grid-community"; import type { ColDef, GetRowIdParams } from "ag-grid-community";
import { AgGridReact, type CustomCellRendererProps } from "ag-grid-react"; import { AgGridReact, type CustomCellRendererProps } from "ag-grid-react";
import { Building2, Cloud, HardDrive, Search, Trash2, X } from "lucide-react"; import {
Building2,
Cloud,
Globe,
HardDrive,
Search,
Trash2,
X,
} from "lucide-react";
import { useRouter } from "next/navigation"; import { useRouter } from "next/navigation";
import { type ChangeEvent, useCallback, useRef, useState } from "react"; import {
type ChangeEvent,
useCallback,
useEffect,
useRef,
useState,
} from "react";
import { SiGoogledrive } from "react-icons/si"; import { SiGoogledrive } from "react-icons/si";
import { TbBrandOnedrive } from "react-icons/tb"; import { TbBrandOnedrive } from "react-icons/tb";
import { KnowledgeDropdown } from "@/components/knowledge-dropdown"; import { KnowledgeDropdown } from "@/components/knowledge-dropdown";
@ -18,14 +32,16 @@ import "@/components/AgGrid/registerAgGridModules";
import "@/components/AgGrid/agGridStyles.css"; import "@/components/AgGrid/agGridStyles.css";
import { toast } from "sonner"; import { toast } from "sonner";
import { KnowledgeActionsDropdown } from "@/components/knowledge-actions-dropdown"; import { KnowledgeActionsDropdown } from "@/components/knowledge-actions-dropdown";
import { filterAccentClasses } from "@/components/knowledge-filter-panel";
import { StatusBadge } from "@/components/ui/status-badge"; import { StatusBadge } from "@/components/ui/status-badge";
import { DeleteConfirmationDialog } from "../../../components/confirmation-dialog"; import { DeleteConfirmationDialog } from "../../../components/confirmation-dialog";
import { useDeleteDocument } from "../api/mutations/useDeleteDocument"; import { useDeleteDocument } from "../api/mutations/useDeleteDocument";
import { filterAccentClasses } from "@/components/knowledge-filter-panel";
// Function to get the appropriate icon for a connector type // Function to get the appropriate icon for a connector type
function getSourceIcon(connectorType?: string) { function getSourceIcon(connectorType?: string) {
switch (connectorType) { switch (connectorType) {
case "url":
return <Globe className="h-4 w-4 text-muted-foreground flex-shrink-0" />;
case "google_drive": case "google_drive":
return ( return (
<SiGoogledrive className="h-4 w-4 text-foreground flex-shrink-0" /> <SiGoogledrive className="h-4 w-4 text-foreground flex-shrink-0" />
@ -47,7 +63,7 @@ function getSourceIcon(connectorType?: string) {
function SearchPage() { function SearchPage() {
const router = useRouter(); const router = useRouter();
const { isMenuOpen, files: taskFiles } = useTask(); const { isMenuOpen, files: taskFiles, refreshTasks } = useTask();
const { totalTopOffset } = useLayout(); const { totalTopOffset } = useLayout();
const { selectedFilter, setSelectedFilter, parsedFilterData, isPanelOpen } = const { selectedFilter, setSelectedFilter, parsedFilterData, isPanelOpen } =
useKnowledgeFilter(); useKnowledgeFilter();
@ -56,15 +72,14 @@ function SearchPage() {
const deleteDocumentMutation = useDeleteDocument(); const deleteDocumentMutation = useDeleteDocument();
const { data = [], isFetching } = useGetSearchQuery( useEffect(() => {
refreshTasks();
}, [refreshTasks]);
const { data: searchData = [], isFetching } = useGetSearchQuery(
parsedFilterData?.query || "*", parsedFilterData?.query || "*",
parsedFilterData parsedFilterData,
); );
const handleTableSearch = (e: ChangeEvent<HTMLInputElement>) => {
gridRef.current?.api.setGridOption("quickFilterText", e.target.value);
};
// Convert TaskFiles to File format and merge with backend results // Convert TaskFiles to File format and merge with backend results
const taskFilesAsFiles: File[] = taskFiles.map((taskFile) => { const taskFilesAsFiles: File[] = taskFiles.map((taskFile) => {
return { return {
@ -77,13 +92,32 @@ function SearchPage() {
}; };
}); });
const backendFiles = data as File[]; // Create a map of task files by filename for quick lookup
const taskFileMap = new Map(
taskFilesAsFiles.map((file) => [file.filename, file]),
);
// Override backend files with task file status if they exist
const backendFiles = (searchData as File[])
.map((file) => {
const taskFile = taskFileMap.get(file.filename);
if (taskFile) {
// Override backend file with task file data (includes status)
return { ...file, ...taskFile };
}
return file;
})
.filter((file) => {
// Only filter out files that are currently processing AND in taskFiles
const taskFile = taskFileMap.get(file.filename);
return !taskFile || taskFile.status !== "processing";
});
const filteredTaskFiles = taskFilesAsFiles.filter((taskFile) => { const filteredTaskFiles = taskFilesAsFiles.filter((taskFile) => {
return ( return (
taskFile.status !== "active" && taskFile.status !== "active" &&
!backendFiles.some( !backendFiles.some(
(backendFile) => backendFile.filename === taskFile.filename (backendFile) => backendFile.filename === taskFile.filename,
) )
); );
}); });
@ -91,41 +125,60 @@ function SearchPage() {
// Combine task files first, then backend files // Combine task files first, then backend files
const fileResults = [...backendFiles, ...filteredTaskFiles]; const fileResults = [...backendFiles, ...filteredTaskFiles];
const handleTableSearch = (e: ChangeEvent<HTMLInputElement>) => {
gridRef.current?.api.setGridOption("quickFilterText", e.target.value);
};
const gridRef = useRef<AgGridReact>(null); const gridRef = useRef<AgGridReact>(null);
const [columnDefs] = useState<ColDef<File>[]>([ const columnDefs = [
{ {
field: "filename", field: "filename",
headerName: "Source", headerName: "Source",
checkboxSelection: true, checkboxSelection: (params: CustomCellRendererProps<File>) =>
(params?.data?.status || "active") === "active",
headerCheckboxSelection: true, headerCheckboxSelection: true,
initialFlex: 2, initialFlex: 2,
minWidth: 220, minWidth: 220,
cellRenderer: ({ data, value }: CustomCellRendererProps<File>) => { cellRenderer: ({ data, value }: CustomCellRendererProps<File>) => {
// Read status directly from data on each render
const status = data?.status || "active";
const isActive = status === "active";
console.log(data?.filename, status, "a");
return ( return (
<button <div className="flex items-center overflow-hidden w-full">
type="button" <div
className="flex items-center gap-2 cursor-pointer hover:text-blue-600 transition-colors text-left w-full" className={`transition-opacity duration-200 ${
onClick={() => { isActive ? "w-0" : "w-7"
router.push( }`}
`/knowledge/chunks?filename=${encodeURIComponent( ></div>
data?.filename ?? "" <button
)}` type="button"
); className="flex items-center gap-2 cursor-pointer hover:text-blue-600 transition-colors text-left flex-1 overflow-hidden"
}} onClick={() => {
> if (!isActive) {
{getSourceIcon(data?.connector_type)} return;
<span className="font-medium text-foreground truncate"> }
{value} router.push(
</span> `/knowledge/chunks?filename=${encodeURIComponent(
</button> data?.filename ?? "",
)}`,
);
}}
>
{getSourceIcon(data?.connector_type)}
<span className="font-medium text-foreground truncate">
{value}
</span>
</button>
</div>
); );
}, },
}, },
{ {
field: "size", field: "size",
headerName: "Size", headerName: "Size",
valueFormatter: (params) => valueFormatter: (params: CustomCellRendererProps<File>) =>
params.value ? `${Math.round(params.value / 1024)} KB` : "-", params.value ? `${Math.round(params.value / 1024)} KB` : "-",
}, },
{ {
@ -135,13 +188,14 @@ function SearchPage() {
{ {
field: "owner", field: "owner",
headerName: "Owner", headerName: "Owner",
valueFormatter: (params) => valueFormatter: (params: CustomCellRendererProps<File>) =>
params.data?.owner_name || params.data?.owner_email || "—", params.data?.owner_name || params.data?.owner_email || "—",
}, },
{ {
field: "chunkCount", field: "chunkCount",
headerName: "Chunks", headerName: "Chunks",
valueFormatter: (params) => params.data?.chunkCount?.toString() || "-", valueFormatter: (params: CustomCellRendererProps<File>) =>
params.data?.chunkCount?.toString() || "-",
}, },
{ {
field: "avgScore", field: "avgScore",
@ -159,6 +213,7 @@ function SearchPage() {
field: "status", field: "status",
headerName: "Status", headerName: "Status",
cellRenderer: ({ data }: CustomCellRendererProps<File>) => { cellRenderer: ({ data }: CustomCellRendererProps<File>) => {
console.log(data?.filename, data?.status, "b");
// Default to 'active' status if no status is provided // Default to 'active' status if no status is provided
const status = data?.status || "active"; const status = data?.status || "active";
return <StatusBadge status={status} />; return <StatusBadge status={status} />;
@ -166,6 +221,10 @@ function SearchPage() {
}, },
{ {
cellRenderer: ({ data }: CustomCellRendererProps<File>) => { cellRenderer: ({ data }: CustomCellRendererProps<File>) => {
const status = data?.status || "active";
if (status !== "active") {
return null;
}
return <KnowledgeActionsDropdown filename={data?.filename || ""} />; return <KnowledgeActionsDropdown filename={data?.filename || ""} />;
}, },
cellStyle: { cellStyle: {
@ -182,7 +241,7 @@ function SearchPage() {
sortable: false, sortable: false,
initialFlex: 0, initialFlex: 0,
}, },
]); ];
const defaultColDef: ColDef<File> = { const defaultColDef: ColDef<File> = {
resizable: false, resizable: false,
@ -204,7 +263,7 @@ function SearchPage() {
try { try {
// Delete each file individually since the API expects one filename at a time // Delete each file individually since the API expects one filename at a time
const deletePromises = selectedRows.map((row) => const deletePromises = selectedRows.map((row) =>
deleteDocumentMutation.mutateAsync({ filename: row.filename }) deleteDocumentMutation.mutateAsync({ filename: row.filename }),
); );
await Promise.all(deletePromises); await Promise.all(deletePromises);
@ -212,7 +271,7 @@ function SearchPage() {
toast.success( toast.success(
`Successfully deleted ${selectedRows.length} document${ `Successfully deleted ${selectedRows.length} document${
selectedRows.length > 1 ? "s" : "" selectedRows.length > 1 ? "s" : ""
}` }`,
); );
setSelectedRows([]); setSelectedRows([]);
setShowBulkDeleteDialog(false); setShowBulkDeleteDialog(false);
@ -225,7 +284,7 @@ function SearchPage() {
toast.error( toast.error(
error instanceof Error error instanceof Error
? error.message ? error.message
: "Failed to delete some documents" : "Failed to delete some documents",
); );
} }
}; };
@ -270,16 +329,13 @@ function SearchPage() {
/> />
</div> </div>
)} )}
<Search <Search className="h-4 w-4 ml-1 flex-shrink-0 text-placeholder-foreground" />
className="h-4 w-4 ml-1 flex-shrink-0 text-placeholder-foreground"
strokeWidth={1.5}
/>
<input <input
className="bg-transparent w-full h-full ml-2 focus:outline-none focus-visible:outline-none font-mono placeholder:font-mono" className="bg-transparent w-full h-full ml-2 focus:outline-none focus-visible:outline-none font-mono placeholder:font-mono"
name="search-query" name="search-query"
id="search-query" id="search-query"
type="text" type="text"
placeholder="Search your documents..." placeholder="Enter your search query..."
onChange={handleTableSearch} onChange={handleTableSearch}
/> />
</div> </div>
@ -317,7 +373,7 @@ function SearchPage() {
</div> </div>
<AgGridReact <AgGridReact
className="w-full overflow-auto" className="w-full overflow-auto"
columnDefs={columnDefs} columnDefs={columnDefs as ColDef<File>[]}
defaultColDef={defaultColDef} defaultColDef={defaultColDef}
loading={isFetching} loading={isFetching}
ref={gridRef} ref={gridRef}
@ -325,7 +381,7 @@ function SearchPage() {
rowSelection="multiple" rowSelection="multiple"
rowMultiSelectWithClick={false} rowMultiSelectWithClick={false}
suppressRowClickSelection={true} suppressRowClickSelection={true}
getRowId={(params) => params.data.filename} getRowId={(params: GetRowIdParams<File>) => params.data?.filename}
domLayout="normal" domLayout="normal"
onSelectionChanged={onSelectionChanged} onSelectionChanged={onSelectionChanged}
noRowsOverlayComponent={() => ( noRowsOverlayComponent={() => (

72
frontend/src/app/settings/page.tsx
View file

@ -633,30 +633,54 @@ function KnowledgeSourcesPage() {
</div> </div>
{/* Conditional Sync Settings or No-Auth Message */} {/* Conditional Sync Settings or No-Auth Message */}
{ {
isNoAuthMode ? ( isNoAuthMode ? (
<Card className="border-yellow-500/50 bg-yellow-500/5"> <Card className="border-yellow-500">
<CardHeader> <CardHeader>
<CardTitle className="text-lg text-yellow-600"> <CardTitle className="text-lg">
Cloud connectors are only available with auth mode enabled Cloud connectors require authentication
</CardTitle> </CardTitle>
<CardDescription className="text-sm"> <CardDescription className="text-sm">
Please provide the following environment variables and Add the Google OAuth variables below to your <code>.env</code>{" "}
restart: then restart the OpenRAG containers.
</CardDescription> </CardDescription>
</CardHeader> </CardHeader>
<CardContent> <CardContent>
<div className="bg-muted rounded-md p-4 font-mono text-sm"> <div className="bg-muted rounded-md p-4 font-mono text-sm">
<div className="text-muted-foreground mb-2"> <div className="text-muted-foreground">
# make here <div>
https://console.cloud.google.com/apis/credentials <span className="mr-3 text-placeholder-foreground">
</div> 27
<div>GOOGLE_OAUTH_CLIENT_ID=</div> </span>
<div>GOOGLE_OAUTH_CLIENT_SECRET=</div> <span># Google OAuth</span>
</div> </div>
</CardContent> <div>
</Card> <span className="mr-3 text-placeholder-foreground">
) : null 28
</span>
<span># Create credentials here:</span>
</div>
<div>
<span className="mr-3 text-placeholder-foreground">
29
</span>
<span>
# https://console.cloud.google.com/apis/credentials
</span>
</div>
</div>
<div>
<span className="mr-3 text-placeholder-foreground">30</span>
<span>GOOGLE_OAUTH_CLIENT_ID=</span>
</div>
<div>
<span className="mr-3 text-placeholder-foreground">31</span>
<span>GOOGLE_OAUTH_CLIENT_SECRET=</span>
</div>
</div>
</CardContent>
</Card>
) : null
// <div className="flex items-center justify-between py-4"> // <div className="flex items-center justify-between py-4">
// <div> // <div>
// <h3 className="text-lg font-medium">Sync Settings</h3> // <h3 className="text-lg font-medium">Sync Settings</h3>

11
frontend/src/components/task-notification-menu.tsx
View file

@ -1,6 +1,6 @@
"use client" "use client"
import { useState } from 'react' import { useEffect, useState } from 'react'
import { Bell, CheckCircle, XCircle, Clock, Loader2, ChevronDown, ChevronUp, X } from 'lucide-react' import { Bell, CheckCircle, XCircle, Clock, Loader2, ChevronDown, ChevronUp, X } from 'lucide-react'
import { Button } from '@/components/ui/button' import { Button } from '@/components/ui/button'
import { Card, CardContent, CardDescription, CardHeader, CardTitle } from '@/components/ui/card' import { Card, CardContent, CardDescription, CardHeader, CardTitle } from '@/components/ui/card'
@ -8,9 +8,16 @@ import { Badge } from '@/components/ui/badge'
import { useTask, Task } from '@/contexts/task-context' import { useTask, Task } from '@/contexts/task-context'
export function TaskNotificationMenu() { export function TaskNotificationMenu() {
const { tasks, isFetching, isMenuOpen, cancelTask } = useTask() const { tasks, isFetching, isMenuOpen, isRecentTasksExpanded, cancelTask } = useTask()
const [isExpanded, setIsExpanded] = useState(false) const [isExpanded, setIsExpanded] = useState(false)
// Sync local state with context state
useEffect(() => {
if (isRecentTasksExpanded) {
setIsExpanded(true)
}
}, [isRecentTasksExpanded])
// Don't render if menu is closed // Don't render if menu is closed
if (!isMenuOpen) return null if (!isMenuOpen) return null

52
frontend/src/components/ui/animated-processing-icon.tsx
View file

@ -1,26 +1,16 @@
interface AnimatedProcessingIconProps { import type { SVGProps } from "react";
className?: string;
size?: number;
}
export const AnimatedProcessingIcon = ({ export const AnimatedProcessingIcon = (props: SVGProps<SVGSVGElement>) => {
className = "", return (
size = 10, <svg
}: AnimatedProcessingIconProps) => { viewBox="0 0 8 12"
const width = Math.round((size * 6) / 10); fill="none"
const height = size; xmlns="http://www.w3.org/2000/svg"
{...props}
return ( >
<svg <title>Processing</title>
width={width} <style>
height={height} {`
viewBox="0 0 6 10"
fill="none"
xmlns="http://www.w3.org/2000/svg"
className={className}
>
<style>
{`
.dot-1 { animation: pulse-wave 1.5s infinite; animation-delay: 0s; } .dot-1 { animation: pulse-wave 1.5s infinite; animation-delay: 0s; }
.dot-2 { animation: pulse-wave 1.5s infinite; animation-delay: 0.1s; } .dot-2 { animation: pulse-wave 1.5s infinite; animation-delay: 0.1s; }
.dot-3 { animation: pulse-wave 1.5s infinite; animation-delay: 0.2s; } .dot-3 { animation: pulse-wave 1.5s infinite; animation-delay: 0.2s; }
@ -30,20 +20,18 @@ export const AnimatedProcessingIcon = ({
@keyframes pulse-wave { @keyframes pulse-wave {
0%, 60%, 100% { 0%, 60%, 100% {
opacity: 0.25; opacity: 0.25;
transform: scale(1);
} }
30% { 30% {
opacity: 1; opacity: 1;
transform: scale(1.2);
} }
} }
`} `}
</style> </style>
<circle className="dot-1" cx="1" cy="5" r="1" fill="currentColor" /> <circle className="dot-1" cx="2" cy="6" r="1" fill="currentColor" />
<circle className="dot-2" cx="1" cy="9" r="1" fill="currentColor" /> <circle className="dot-2" cx="2" cy="10" r="1" fill="currentColor" />
<circle className="dot-3" cx="5" cy="1" r="1" fill="currentColor" /> <circle className="dot-3" cx="6" cy="2" r="1" fill="currentColor" />
<circle className="dot-4" cx="5" cy="5" r="1" fill="currentColor" /> <circle className="dot-4" cx="6" cy="6" r="1" fill="currentColor" />
<circle className="dot-5" cx="5" cy="9" r="1" fill="currentColor" /> <circle className="dot-5" cx="6" cy="10" r="1" fill="currentColor" />
</svg> </svg>
); );
}; };

2
frontend/src/components/ui/status-badge.tsx
View file

@ -50,7 +50,7 @@ export const StatusBadge = ({ status, className }: StatusBadgeProps) => {
}`} }`}
> >
{status === "processing" && ( {status === "processing" && (
<AnimatedProcessingIcon className="text-current mr-2" size={10} /> <AnimatedProcessingIcon className="text-current h-3 w-3 shrink-0" />
)} )}
{config.label} {config.label}
</div> </div>

467
frontend/src/contexts/task-context.tsx
View file

@ -7,33 +7,18 @@ import {
useCallback, useCallback,
useContext, useContext,
useEffect, useEffect,
useRef,
useState, useState,
} from "react"; } from "react";
import { toast } from "sonner"; import { toast } from "sonner";
import { useCancelTaskMutation } from "@/app/api/mutations/useCancelTaskMutation";
import {
type Task,
useGetTasksQuery,
} from "@/app/api/queries/useGetTasksQuery";
import { useAuth } from "@/contexts/auth-context"; import { useAuth } from "@/contexts/auth-context";
export interface Task { // Task interface is now imported from useGetTasksQuery
task_id: string;
status:
| "pending"
| "running"
| "processing"
| "completed"
| "failed"
| "error";
total_files?: number;
processed_files?: number;
successful_files?: number;
failed_files?: number;
running_files?: number;
pending_files?: number;
created_at: string;
updated_at: string;
duration_seconds?: number;
result?: Record<string, unknown>;
error?: string;
files?: Record<string, Record<string, unknown>>;
}
export interface TaskFile { export interface TaskFile {
filename: string; filename: string;
@ -51,27 +36,54 @@ interface TaskContextType {
files: TaskFile[]; files: TaskFile[];
addTask: (taskId: string) => void; addTask: (taskId: string) => void;
addFiles: (files: Partial<TaskFile>[], taskId: string) => void; addFiles: (files: Partial<TaskFile>[], taskId: string) => void;
removeTask: (taskId: string) => void;
refreshTasks: () => Promise<void>; refreshTasks: () => Promise<void>;
cancelTask: (taskId: string) => Promise<void>; cancelTask: (taskId: string) => Promise<void>;
isPolling: boolean; isPolling: boolean;
isFetching: boolean; isFetching: boolean;
isMenuOpen: boolean; isMenuOpen: boolean;
toggleMenu: () => void; toggleMenu: () => void;
isRecentTasksExpanded: boolean;
setRecentTasksExpanded: (expanded: boolean) => void;
// React Query states
isLoading: boolean;
error: Error | null;
} }
const TaskContext = createContext<TaskContextType | undefined>(undefined); const TaskContext = createContext<TaskContextType | undefined>(undefined);
export function TaskProvider({ children }: { children: React.ReactNode }) { export function TaskProvider({ children }: { children: React.ReactNode }) {
const [tasks, setTasks] = useState<Task[]>([]);
const [files, setFiles] = useState<TaskFile[]>([]); const [files, setFiles] = useState<TaskFile[]>([]);
const [isPolling, setIsPolling] = useState(false);
const [isFetching, setIsFetching] = useState(false);
const [isMenuOpen, setIsMenuOpen] = useState(false); const [isMenuOpen, setIsMenuOpen] = useState(false);
const [isRecentTasksExpanded, setIsRecentTasksExpanded] = useState(false);
const previousTasksRef = useRef<Task[]>([]);
const { isAuthenticated, isNoAuthMode } = useAuth(); const { isAuthenticated, isNoAuthMode } = useAuth();
const queryClient = useQueryClient(); const queryClient = useQueryClient();
// Use React Query hooks
const {
data: tasks = [],
isLoading,
error,
refetch: refetchTasks,
isFetching,
} = useGetTasksQuery({
enabled: isAuthenticated || isNoAuthMode,
});
const cancelTaskMutation = useCancelTaskMutation({
onSuccess: () => {
toast.success("Task cancelled", {
description: "Task has been cancelled successfully",
});
},
onError: (error) => {
toast.error("Failed to cancel task", {
description: error.message,
});
},
});
const refetchSearch = useCallback(() => { const refetchSearch = useCallback(() => {
queryClient.invalidateQueries({ queryClient.invalidateQueries({
queryKey: ["search"], queryKey: ["search"],
@ -99,265 +111,216 @@ export function TaskProvider({ children }: { children: React.ReactNode }) {
[], [],
); );
const fetchTasks = useCallback(async () => { // Handle task status changes and file updates
if (!isAuthenticated && !isNoAuthMode) return; useEffect(() => {
if (tasks.length === 0) {
setIsFetching(true); // Store current tasks as previous for next comparison
try { previousTasksRef.current = tasks;
const response = await fetch("/api/tasks"); return;
if (response.ok) {
const data = await response.json();
const newTasks = data.tasks || [];
// Update tasks and check for status changes in the same state update
setTasks((prevTasks) => {
// Check for newly completed tasks to show toasts
if (prevTasks.length > 0) {
newTasks.forEach((newTask: Task) => {
const oldTask = prevTasks.find(
(t) => t.task_id === newTask.task_id,
);
// Update or add files from task.files if available
if (newTask.files && typeof newTask.files === "object") {
const taskFileEntries = Object.entries(newTask.files);
const now = new Date().toISOString();
taskFileEntries.forEach(([filePath, fileInfo]) => {
if (typeof fileInfo === "object" && fileInfo) {
const fileName = filePath.split("/").pop() || filePath;
const fileStatus = fileInfo.status as string;
// Map backend file status to our TaskFile status
let mappedStatus: TaskFile["status"];
switch (fileStatus) {
case "pending":
case "running":
mappedStatus = "processing";
break;
case "completed":
mappedStatus = "active";
break;
case "failed":
mappedStatus = "failed";
break;
default:
mappedStatus = "processing";
}
setFiles((prevFiles) => {
const existingFileIndex = prevFiles.findIndex(
(f) =>
f.source_url === filePath &&
f.task_id === newTask.task_id,
);
// Detect connector type based on file path or other indicators
let connectorType = "local";
if (filePath.includes("/") && !filePath.startsWith("/")) {
// Likely S3 key format (bucket/path/file.ext)
connectorType = "s3";
}
const fileEntry: TaskFile = {
filename: fileName,
mimetype: "", // We don't have this info from the task
source_url: filePath,
size: 0, // We don't have this info from the task
connector_type: connectorType,
status: mappedStatus,
task_id: newTask.task_id,
created_at:
typeof fileInfo.created_at === "string"
? fileInfo.created_at
: now,
updated_at:
typeof fileInfo.updated_at === "string"
? fileInfo.updated_at
: now,
};
if (existingFileIndex >= 0) {
// Update existing file
const updatedFiles = [...prevFiles];
updatedFiles[existingFileIndex] = fileEntry;
return updatedFiles;
} else {
// Add new file
return [...prevFiles, fileEntry];
}
});
}
});
}
if (
oldTask &&
oldTask.status !== "completed" &&
newTask.status === "completed"
) {
// Task just completed - show success toast
toast.success("Task completed successfully", {
description: `Task ${newTask.task_id} has finished processing.`,
action: {
label: "View",
onClick: () => console.log("View task", newTask.task_id),
},
});
refetchSearch();
// Dispatch knowledge updated event for all knowledge-related pages
console.log(
"Task completed successfully, dispatching knowledgeUpdated event",
);
window.dispatchEvent(new CustomEvent("knowledgeUpdated"));
// Remove files for this completed task from the files list
setFiles((prevFiles) =>
prevFiles.filter((file) => file.task_id !== newTask.task_id),
);
} else if (
oldTask &&
oldTask.status !== "failed" &&
oldTask.status !== "error" &&
(newTask.status === "failed" || newTask.status === "error")
) {
// Task just failed - show error toast
toast.error("Task failed", {
description: `Task ${newTask.task_id} failed: ${
newTask.error || "Unknown error"
}`,
});
// Files will be updated to failed status by the file parsing logic above
}
});
}
return newTasks;
});
}
} catch (error) {
console.error("Failed to fetch tasks:", error);
} finally {
setIsFetching(false);
} }
}, [isAuthenticated, isNoAuthMode, refetchSearch]); // Removed 'tasks' from dependencies to prevent infinite loop!
const addTask = useCallback((taskId: string) => { // Check for task status changes by comparing with previous tasks
// Immediately start aggressive polling for the new task tasks.forEach((currentTask) => {
let pollAttempts = 0; const previousTask = previousTasksRef.current.find(
const maxPollAttempts = 30; // Poll for up to 30 seconds (prev) => prev.task_id === currentTask.task_id,
);
const aggressivePoll = async () => { // Only show toasts if we have previous data and status has changed
try { if (
const response = await fetch("/api/tasks"); (previousTask && previousTask.status !== currentTask.status) ||
if (response.ok) { (!previousTask && previousTasksRef.current.length !== 0)
const data = await response.json(); ) {
const newTasks = data.tasks || []; // Process files from failed task and add them to files list
const foundTask = newTasks.find( if (currentTask.files && typeof currentTask.files === "object") {
(task: Task) => task.task_id === taskId, const taskFileEntries = Object.entries(currentTask.files);
); const now = new Date().toISOString();
if (foundTask) { taskFileEntries.forEach(([filePath, fileInfo]) => {
// Task found! Update the tasks state if (typeof fileInfo === "object" && fileInfo) {
setTasks((prevTasks) => { // Use the filename from backend if available, otherwise extract from path
// Check if task is already in the list const fileName =
const exists = prevTasks.some((t) => t.task_id === taskId); (fileInfo as any).filename ||
if (!exists) { filePath.split("/").pop() ||
return [...prevTasks, foundTask]; filePath;
const fileStatus = fileInfo.status as string;
// Map backend file status to our TaskFile status
let mappedStatus: TaskFile["status"];
switch (fileStatus) {
case "pending":
case "running":
mappedStatus = "processing";
break;
case "completed":
mappedStatus = "active";
break;
case "failed":
mappedStatus = "failed";
break;
default:
mappedStatus = "processing";
} }
// Update existing task
return prevTasks.map((t) => setFiles((prevFiles) => {
t.task_id === taskId ? foundTask : t, const existingFileIndex = prevFiles.findIndex(
); (f) =>
}); f.source_url === filePath &&
return; // Stop polling, we found it f.task_id === currentTask.task_id,
} );
// Detect connector type based on file path or other indicators
let connectorType = "local";
if (filePath.includes("/") && !filePath.startsWith("/")) {
// Likely S3 key format (bucket/path/file.ext)
connectorType = "s3";
}
const fileEntry: TaskFile = {
filename: fileName,
mimetype: "", // We don't have this info from the task
source_url: filePath,
size: 0, // We don't have this info from the task
connector_type: connectorType,
status: mappedStatus,
task_id: currentTask.task_id,
created_at:
typeof fileInfo.created_at === "string"
? fileInfo.created_at
: now,
updated_at:
typeof fileInfo.updated_at === "string"
? fileInfo.updated_at
: now,
};
if (existingFileIndex >= 0) {
// Update existing file
const updatedFiles = [...prevFiles];
updatedFiles[existingFileIndex] = fileEntry;
return updatedFiles;
} else {
// Add new file
return [...prevFiles, fileEntry];
}
});
}
});
} }
} catch (error) { if (
console.error("Aggressive polling failed:", error); previousTask &&
} previousTask.status !== "completed" &&
currentTask.status === "completed"
) {
// Task just completed - show success toast with file counts
const successfulFiles = currentTask.successful_files || 0;
const failedFiles = currentTask.failed_files || 0;
pollAttempts++; let description = "";
if (pollAttempts < maxPollAttempts) { if (failedFiles > 0) {
// Continue polling every 1 second for new tasks description = `${successfulFiles} file${
setTimeout(aggressivePoll, 1000); successfulFiles !== 1 ? "s" : ""
} } uploaded successfully, ${failedFiles} file${
}; failedFiles !== 1 ? "s" : ""
} failed`;
} else {
description = `${successfulFiles} file${
successfulFiles !== 1 ? "s" : ""
} uploaded successfully`;
}
// Start aggressive polling after a short delay to allow backend to process toast.success("Task completed", {
setTimeout(aggressivePoll, 500); description,
}, []); action: {
label: "View",
onClick: () => {
setIsMenuOpen(true);
setIsRecentTasksExpanded(true);
},
},
});
setTimeout(() => {
setFiles((prevFiles) =>
prevFiles.filter(
(file) =>
file.task_id !== currentTask.task_id ||
file.status === "failed",
),
);
refetchSearch();
}, 500);
} else if (
previousTask &&
previousTask.status !== "failed" &&
previousTask.status !== "error" &&
(currentTask.status === "failed" || currentTask.status === "error")
) {
// Task just failed - show error toast
toast.error("Task failed", {
description: `Task ${currentTask.task_id} failed: ${
currentTask.error || "Unknown error"
}`,
});
}
}
});
// Store current tasks as previous for next comparison
previousTasksRef.current = tasks;
}, [tasks, refetchSearch]);
const addTask = useCallback(
(_taskId: string) => {
// React Query will automatically handle polling when tasks are active
// Just trigger a refetch to get the latest data
setTimeout(() => {
refetchTasks();
}, 500);
},
[refetchTasks],
);
const refreshTasks = useCallback(async () => { const refreshTasks = useCallback(async () => {
await fetchTasks(); setFiles([]);
}, [fetchTasks]); await refetchTasks();
}, [refetchTasks]);
const removeTask = useCallback((taskId: string) => {
setTasks((prev) => prev.filter((task) => task.task_id !== taskId));
}, []);
const cancelTask = useCallback( const cancelTask = useCallback(
async (taskId: string) => { async (taskId: string) => {
try { cancelTaskMutation.mutate({ taskId });
const response = await fetch(`/api/tasks/${taskId}/cancel`, {
method: "POST",
});
if (response.ok) {
// Immediately refresh tasks to show the updated status
await fetchTasks();
toast.success("Task cancelled", {
description: `Task ${taskId.substring(0, 8)}... has been cancelled`,
});
} else {
const errorData = await response.json().catch(() => ({}));
throw new Error(errorData.error || "Failed to cancel task");
}
} catch (error) {
console.error("Failed to cancel task:", error);
toast.error("Failed to cancel task", {
description: error instanceof Error ? error.message : "Unknown error",
});
}
}, },
[fetchTasks], [cancelTaskMutation],
); );
const toggleMenu = useCallback(() => { const toggleMenu = useCallback(() => {
setIsMenuOpen((prev) => !prev); setIsMenuOpen((prev) => !prev);
}, []); }, []);
// Periodic polling for task updates // Determine if we're polling based on React Query's refetch interval
useEffect(() => { const isPolling =
if (!isAuthenticated && !isNoAuthMode) return; isFetching &&
tasks.some(
setIsPolling(true); (task) =>
task.status === "pending" ||
// Initial fetch task.status === "running" ||
fetchTasks(); task.status === "processing",
);
// Set up polling interval - every 3 seconds (more responsive for active tasks)
const interval = setInterval(fetchTasks, 3000);
return () => {
clearInterval(interval);
setIsPolling(false);
};
}, [isAuthenticated, isNoAuthMode, fetchTasks]);
const value: TaskContextType = { const value: TaskContextType = {
tasks, tasks,
files, files,
addTask, addTask,
addFiles, addFiles,
removeTask,
refreshTasks, refreshTasks,
cancelTask, cancelTask,
isPolling, isPolling,
isFetching, isFetching,
isMenuOpen, isMenuOpen,
toggleMenu, toggleMenu,
isRecentTasksExpanded,
setRecentTasksExpanded: setIsRecentTasksExpanded,
isLoading,
error,
}; };
return <TaskContext.Provider value={value}>{children}</TaskContext.Provider>; return <TaskContext.Provider value={value}>{children}</TaskContext.Provider>;

62
src/api/documents.py
View file

@ -6,6 +6,54 @@ from config.settings import INDEX_NAME
logger = get_logger(__name__) logger = get_logger(__name__)
async def check_filename_exists(request: Request, document_service, session_manager):
"""Check if a document with a specific filename already exists"""
filename = request.query_params.get("filename")
if not filename:
return JSONResponse({"error": "filename parameter is required"}, status_code=400)
user = request.state.user
jwt_token = session_manager.get_effective_jwt_token(user.user_id, request.state.jwt_token)
try:
# Get user's OpenSearch client
opensearch_client = session_manager.get_user_opensearch_client(
user.user_id, jwt_token
)
# Search for any document with this exact filename
from utils.opensearch_queries import build_filename_search_body
search_body = build_filename_search_body(filename, size=1, source=["filename"])
logger.debug(f"Checking filename existence: {filename}")
response = await opensearch_client.search(
index=INDEX_NAME,
body=search_body
)
# Check if any hits were found
hits = response.get("hits", {}).get("hits", [])
exists = len(hits) > 0
logger.debug(f"Filename check result - exists: {exists}, hits: {len(hits)}")
return JSONResponse({
"exists": exists,
"filename": filename
}, status_code=200)
except Exception as e:
logger.error("Error checking filename existence", filename=filename, error=str(e))
error_str = str(e)
if "AuthenticationException" in error_str:
return JSONResponse({"error": "Access denied: insufficient permissions"}, status_code=403)
else:
return JSONResponse({"error": str(e)}, status_code=500)
async def delete_documents_by_filename(request: Request, document_service, session_manager): async def delete_documents_by_filename(request: Request, document_service, session_manager):
"""Delete all documents with a specific filename""" """Delete all documents with a specific filename"""
data = await request.json() data = await request.json()
@ -24,15 +72,11 @@ async def delete_documents_by_filename(request: Request, document_service, sessi
) )
# Delete by query to remove all chunks of this document # Delete by query to remove all chunks of this document
delete_query = { from utils.opensearch_queries import build_filename_delete_body
"query": {
"bool": { delete_query = build_filename_delete_body(filename)
"must": [
{"term": {"filename": filename}} logger.debug(f"Deleting documents with filename: {filename}")
]
}
}
}
result = await opensearch_client.delete_by_query( result = await opensearch_client.delete_by_query(
index=INDEX_NAME, index=INDEX_NAME,

11
src/api/langflow_files.py
View file

@ -193,15 +193,16 @@ async def upload_and_ingest_user_file(
# Read file content # Read file content
content = await upload_file.read() content = await upload_file.read()
# Create temporary file # Create temporary file with the actual filename (not a temp prefix)
# Store in temp directory but use the real filename
temp_dir = tempfile.gettempdir()
safe_filename = upload_file.filename.replace(" ", "_").replace("/", "_") safe_filename = upload_file.filename.replace(" ", "_").replace("/", "_")
temp_fd, temp_path = tempfile.mkstemp( temp_path = os.path.join(temp_dir, safe_filename)
suffix=f"_{safe_filename}"
)
try: try:
# Write content to temp file # Write content to temp file
with os.fdopen(temp_fd, 'wb') as temp_file: with open(temp_path, 'wb') as temp_file:
temp_file.write(content) temp_file.write(content)
logger.debug("Created temporary file for task processing", temp_path=temp_path) logger.debug("Created temporary file for task processing", temp_path=temp_path)

72
src/api/router.py
View file

@ -17,7 +17,7 @@ async def upload_ingest_router(
document_service=None, document_service=None,
langflow_file_service=None, langflow_file_service=None,
session_manager=None, session_manager=None,
task_service=None task_service=None,
): ):
""" """
Router endpoint that automatically routes upload requests based on configuration. Router endpoint that automatically routes upload requests based on configuration.
@ -31,7 +31,7 @@ async def upload_ingest_router(
try: try:
logger.debug( logger.debug(
"Router upload_ingest endpoint called", "Router upload_ingest endpoint called",
disable_langflow_ingest=DISABLE_INGEST_WITH_LANGFLOW disable_langflow_ingest=DISABLE_INGEST_WITH_LANGFLOW,
) )
# Route based on configuration # Route based on configuration
@ -42,7 +42,9 @@ async def upload_ingest_router(
else: else:
# Route to Langflow upload and ingest using task service # Route to Langflow upload and ingest using task service
logger.debug("Routing to Langflow upload-ingest pipeline via task service") logger.debug("Routing to Langflow upload-ingest pipeline via task service")
return await langflow_upload_ingest_task(request, langflow_file_service, session_manager, task_service) return await langflow_upload_ingest_task(
request, langflow_file_service, session_manager, task_service
)
except Exception as e: except Exception as e:
logger.error("Error in upload_ingest_router", error=str(e)) logger.error("Error in upload_ingest_router", error=str(e))
@ -57,10 +59,7 @@ async def upload_ingest_router(
async def langflow_upload_ingest_task( async def langflow_upload_ingest_task(
request: Request, request: Request, langflow_file_service, session_manager, task_service
langflow_file_service,
session_manager,
task_service
): ):
"""Task-based langflow upload and ingest for single/multiple files""" """Task-based langflow upload and ingest for single/multiple files"""
try: try:
@ -77,6 +76,7 @@ async def langflow_upload_ingest_task(
settings_json = form.get("settings") settings_json = form.get("settings")
tweaks_json = form.get("tweaks") tweaks_json = form.get("tweaks")
delete_after_ingest = form.get("delete_after_ingest", "true").lower() == "true" delete_after_ingest = form.get("delete_after_ingest", "true").lower() == "true"
replace_duplicates = form.get("replace_duplicates", "false").lower() == "true"
# Parse JSON fields if provided # Parse JSON fields if provided
settings = None settings = None
@ -85,6 +85,7 @@ async def langflow_upload_ingest_task(
if settings_json: if settings_json:
try: try:
import json import json
settings = json.loads(settings_json) settings = json.loads(settings_json)
except json.JSONDecodeError as e: except json.JSONDecodeError as e:
logger.error("Invalid settings JSON", error=str(e)) logger.error("Invalid settings JSON", error=str(e))
@ -93,6 +94,7 @@ async def langflow_upload_ingest_task(
if tweaks_json: if tweaks_json:
try: try:
import json import json
tweaks = json.loads(tweaks_json) tweaks = json.loads(tweaks_json)
except json.JSONDecodeError as e: except json.JSONDecodeError as e:
logger.error("Invalid tweaks JSON", error=str(e)) logger.error("Invalid tweaks JSON", error=str(e))
@ -106,26 +108,35 @@ async def langflow_upload_ingest_task(
jwt_token = getattr(request.state, "jwt_token", None) jwt_token = getattr(request.state, "jwt_token", None)
if not user_id: if not user_id:
return JSONResponse({"error": "User authentication required"}, status_code=401) return JSONResponse(
{"error": "User authentication required"}, status_code=401
)
# Create temporary files for task processing # Create temporary files for task processing
import tempfile import tempfile
import os import os
temp_file_paths = [] temp_file_paths = []
original_filenames = []
try: try:
# Create temp directory reference once
temp_dir = tempfile.gettempdir()
for upload_file in upload_files: for upload_file in upload_files:
# Read file content # Read file content
content = await upload_file.read() content = await upload_file.read()
# Create temporary file # Store ORIGINAL filename (not transformed)
original_filenames.append(upload_file.filename)
# Create temporary file with TRANSFORMED filename for filesystem safety
# Transform: spaces and / to underscore
safe_filename = upload_file.filename.replace(" ", "_").replace("/", "_") safe_filename = upload_file.filename.replace(" ", "_").replace("/", "_")
temp_fd, temp_path = tempfile.mkstemp( temp_path = os.path.join(temp_dir, safe_filename)
suffix=f"_{safe_filename}"
)
# Write content to temp file # Write content to temp file
with os.fdopen(temp_fd, 'wb') as temp_file: with open(temp_path, "wb") as temp_file:
temp_file.write(content) temp_file.write(content)
temp_file_paths.append(temp_path) temp_file_paths.append(temp_path)
@ -136,21 +147,22 @@ async def langflow_upload_ingest_task(
user_id=user_id, user_id=user_id,
has_settings=bool(settings), has_settings=bool(settings),
has_tweaks=bool(tweaks), has_tweaks=bool(tweaks),
delete_after_ingest=delete_after_ingest delete_after_ingest=delete_after_ingest,
) )
# Create langflow upload task # Create langflow upload task
print(f"tweaks: {tweaks}") logger.debug(
print(f"settings: {settings}") f"Preparing to create langflow upload task: tweaks={tweaks}, settings={settings}, jwt_token={jwt_token}, user_name={user_name}, user_email={user_email}, session_id={session_id}, delete_after_ingest={delete_after_ingest}, temp_file_paths={temp_file_paths}",
print(f"jwt_token: {jwt_token}") )
print(f"user_name: {user_name}") # Create a map between temp_file_paths and original_filenames
print(f"user_email: {user_email}") file_path_to_original_filename = dict(zip(temp_file_paths, original_filenames))
print(f"session_id: {session_id}") logger.debug(
print(f"delete_after_ingest: {delete_after_ingest}") f"File path to original filename map: {file_path_to_original_filename}",
print(f"temp_file_paths: {temp_file_paths}") )
task_id = await task_service.create_langflow_upload_task( task_id = await task_service.create_langflow_upload_task(
user_id=user_id, user_id=user_id,
file_paths=temp_file_paths, file_paths=temp_file_paths,
original_filenames=file_path_to_original_filename,
langflow_file_service=langflow_file_service, langflow_file_service=langflow_file_service,
session_manager=session_manager, session_manager=session_manager,
jwt_token=jwt_token, jwt_token=jwt_token,
@ -160,19 +172,24 @@ async def langflow_upload_ingest_task(
tweaks=tweaks, tweaks=tweaks,
settings=settings, settings=settings,
delete_after_ingest=delete_after_ingest, delete_after_ingest=delete_after_ingest,
replace_duplicates=replace_duplicates,
) )
logger.debug("Langflow upload task created successfully", task_id=task_id) logger.debug("Langflow upload task created successfully", task_id=task_id)
return JSONResponse({ return JSONResponse(
"task_id": task_id, {
"message": f"Langflow upload task created for {len(upload_files)} file(s)", "task_id": task_id,
"file_count": len(upload_files) "message": f"Langflow upload task created for {len(upload_files)} file(s)",
}, status_code=202) # 202 Accepted for async processing "file_count": len(upload_files),
},
status_code=202,
) # 202 Accepted for async processing
except Exception: except Exception:
# Clean up temp files on error # Clean up temp files on error
from utils.file_utils import safe_unlink from utils.file_utils import safe_unlink
for temp_path in temp_file_paths: for temp_path in temp_file_paths:
safe_unlink(temp_path) safe_unlink(temp_path)
raise raise
@ -184,5 +201,6 @@ async def langflow_upload_ingest_task(
error=str(e), error=str(e),
) )
import traceback import traceback
logger.error("Full traceback", traceback=traceback.format_exc()) logger.error("Full traceback", traceback=traceback.format_exc())
return JSONResponse({"error": str(e)}, status_code=500) return JSONResponse({"error": str(e)}, status_code=500)

11
src/main.py
View file

@ -953,6 +953,17 @@ async def create_app():
methods=["POST", "GET"], methods=["POST", "GET"],
), ),
# Document endpoints # Document endpoints
Route(
"/documents/check-filename",
require_auth(services["session_manager"])(
partial(
documents.check_filename_exists,
document_service=services["document_service"],
session_manager=services["session_manager"],
)
),
methods=["GET"],
),
Route( Route(
"/documents/delete-by-filename", "/documents/delete-by-filename",
require_auth(services["session_manager"])( require_auth(services["session_manager"])(

129
src/models/processors.py
View file

@ -55,6 +55,96 @@ class TaskProcessor:
await asyncio.sleep(retry_delay) await asyncio.sleep(retry_delay)
retry_delay *= 2 # Exponential backoff retry_delay *= 2 # Exponential backoff
async def check_filename_exists(
self,
filename: str,
opensearch_client,
) -> bool:
"""
Check if a document with the given filename already exists in OpenSearch.
Returns True if any chunks with this filename exist.
"""
from config.settings import INDEX_NAME
from utils.opensearch_queries import build_filename_search_body
import asyncio
max_retries = 3
retry_delay = 1.0
for attempt in range(max_retries):
try:
# Search for any document with this exact filename
search_body = build_filename_search_body(filename, size=1, source=False)
response = await opensearch_client.search(
index=INDEX_NAME,
body=search_body
)
# Check if any hits were found
hits = response.get("hits", {}).get("hits", [])
return len(hits) > 0
except (asyncio.TimeoutError, Exception) as e:
if attempt == max_retries - 1:
logger.error(
"OpenSearch filename check failed after retries",
filename=filename,
error=str(e),
attempt=attempt + 1
)
# On final failure, assume document doesn't exist (safer to reprocess than skip)
logger.warning(
"Assuming filename doesn't exist due to connection issues",
filename=filename
)
return False
else:
logger.warning(
"OpenSearch filename check failed, retrying",
filename=filename,
error=str(e),
attempt=attempt + 1,
retry_in=retry_delay
)
await asyncio.sleep(retry_delay)
retry_delay *= 2 # Exponential backoff
async def delete_document_by_filename(
self,
filename: str,
opensearch_client,
) -> None:
"""
Delete all chunks of a document with the given filename from OpenSearch.
"""
from config.settings import INDEX_NAME
from utils.opensearch_queries import build_filename_delete_body
try:
# Delete all documents with this filename
delete_body = build_filename_delete_body(filename)
response = await opensearch_client.delete_by_query(
index=INDEX_NAME,
body=delete_body
)
deleted_count = response.get("deleted", 0)
logger.info(
"Deleted existing document chunks",
filename=filename,
deleted_count=deleted_count
)
except Exception as e:
logger.error(
"Failed to delete existing document",
filename=filename,
error=str(e)
)
raise
async def process_document_standard( async def process_document_standard(
self, self,
file_path: str, file_path: str,
@ -527,6 +617,7 @@ class LangflowFileProcessor(TaskProcessor):
tweaks: dict = None, tweaks: dict = None,
settings: dict = None, settings: dict = None,
delete_after_ingest: bool = True, delete_after_ingest: bool = True,
replace_duplicates: bool = False,
): ):
super().__init__() super().__init__()
self.langflow_file_service = langflow_file_service self.langflow_file_service = langflow_file_service
@ -539,6 +630,7 @@ class LangflowFileProcessor(TaskProcessor):
self.tweaks = tweaks or {} self.tweaks = tweaks or {}
self.settings = settings self.settings = settings
self.delete_after_ingest = delete_after_ingest self.delete_after_ingest = delete_after_ingest
self.replace_duplicates = replace_duplicates
async def process_item( async def process_item(
self, upload_task: UploadTask, item: str, file_task: FileTask self, upload_task: UploadTask, item: str, file_task: FileTask
@ -554,37 +646,40 @@ class LangflowFileProcessor(TaskProcessor):
file_task.updated_at = time.time() file_task.updated_at = time.time()
try: try:
# Compute hash and check if already exists # Use the ORIGINAL filename stored in file_task (not the transformed temp path)
from utils.hash_utils import hash_id # This ensures we check/store the original filename with spaces, etc.
file_hash = hash_id(item) original_filename = file_task.filename or os.path.basename(item)
# Check if document already exists # Check if document with same filename already exists
opensearch_client = self.session_manager.get_user_opensearch_client( opensearch_client = self.session_manager.get_user_opensearch_client(
self.owner_user_id, self.jwt_token self.owner_user_id, self.jwt_token
) )
if await self.check_document_exists(file_hash, opensearch_client):
file_task.status = TaskStatus.COMPLETED filename_exists = await self.check_filename_exists(original_filename, opensearch_client)
file_task.result = {"status": "unchanged", "id": file_hash}
if filename_exists and not self.replace_duplicates:
# Duplicate exists and user hasn't confirmed replacement
file_task.status = TaskStatus.FAILED
file_task.error = f"File with name '{original_filename}' already exists"
file_task.updated_at = time.time() file_task.updated_at = time.time()
upload_task.successful_files += 1 upload_task.failed_files += 1
return return
elif filename_exists and self.replace_duplicates:
# Delete existing document before uploading new one
logger.info(f"Replacing existing document: {original_filename}")
await self.delete_document_by_filename(original_filename, opensearch_client)
# Read file content for processing # Read file content for processing
with open(item, 'rb') as f: with open(item, 'rb') as f:
content = f.read() content = f.read()
# Create file tuple for upload # Create file tuple for upload using ORIGINAL filename
temp_filename = os.path.basename(item) # This ensures the document is indexed with the original name
# Extract original filename from temp file suffix (remove tmp prefix) content_type, _ = mimetypes.guess_type(original_filename)
if "_" in temp_filename:
filename = temp_filename.split("_", 1)[1] # Get everything after first _
else:
filename = temp_filename
content_type, _ = mimetypes.guess_type(filename)
if not content_type: if not content_type:
content_type = 'application/octet-stream' content_type = 'application/octet-stream'
file_tuple = (filename, content, content_type) file_tuple = (original_filename, content, content_type)
# Get JWT token using same logic as DocumentFileProcessor # Get JWT token using same logic as DocumentFileProcessor
# This will handle anonymous JWT creation if needed # This will handle anonymous JWT creation if needed

1
src/models/tasks.py
View file

@ -20,6 +20,7 @@ class FileTask:
retry_count: int = 0 retry_count: int = 0
created_at: float = field(default_factory=time.time) created_at: float = field(default_factory=time.time)
updated_at: float = field(default_factory=time.time) updated_at: float = field(default_factory=time.time)
filename: Optional[str] = None # Original filename for display
@property @property
def duration_seconds(self) -> float: def duration_seconds(self) -> float:

28
src/services/task_service.py
View file

@ -1,6 +1,5 @@
import asyncio import asyncio
import random import random
from typing import Dict, Optional
import time import time
import uuid import uuid
@ -59,6 +58,7 @@ class TaskService:
file_paths: list, file_paths: list,
langflow_file_service, langflow_file_service,
session_manager, session_manager,
original_filenames: dict | None = None,
jwt_token: str = None, jwt_token: str = None,
owner_name: str = None, owner_name: str = None,
owner_email: str = None, owner_email: str = None,
@ -66,6 +66,7 @@ class TaskService:
tweaks: dict = None, tweaks: dict = None,
settings: dict = None, settings: dict = None,
delete_after_ingest: bool = True, delete_after_ingest: bool = True,
replace_duplicates: bool = False,
) -> str: ) -> str:
"""Create a new upload task for Langflow file processing with upload and ingest""" """Create a new upload task for Langflow file processing with upload and ingest"""
# Use LangflowFileProcessor with user context # Use LangflowFileProcessor with user context
@ -82,18 +83,35 @@ class TaskService:
tweaks=tweaks, tweaks=tweaks,
settings=settings, settings=settings,
delete_after_ingest=delete_after_ingest, delete_after_ingest=delete_after_ingest,
replace_duplicates=replace_duplicates,
) )
return await self.create_custom_task(user_id, file_paths, processor) return await self.create_custom_task(user_id, file_paths, processor, original_filenames)
async def create_custom_task(self, user_id: str, items: list, processor) -> str: async def create_custom_task(self, user_id: str, items: list, processor, original_filenames: dict | None = None) -> str:
"""Create a new task with custom processor for any type of items""" """Create a new task with custom processor for any type of items"""
import os
# Store anonymous tasks under a stable key so they can be retrieved later # Store anonymous tasks under a stable key so they can be retrieved later
store_user_id = user_id or AnonymousUser().user_id store_user_id = user_id or AnonymousUser().user_id
task_id = str(uuid.uuid4()) task_id = str(uuid.uuid4())
# Create file tasks with original filenames if provided
normalized_originals = (
{str(k): v for k, v in original_filenames.items()} if original_filenames else {}
)
file_tasks = {
str(item): FileTask(
file_path=str(item),
filename=normalized_originals.get(
str(item), os.path.basename(str(item))
),
)
for item in items
}
upload_task = UploadTask( upload_task = UploadTask(
task_id=task_id, task_id=task_id,
total_files=len(items), total_files=len(items),
file_tasks={str(item): FileTask(file_path=str(item)) for item in items}, file_tasks=file_tasks,
) )
# Attach the custom processor to the task # Attach the custom processor to the task
@ -268,6 +286,7 @@ class TaskService:
"created_at": file_task.created_at, "created_at": file_task.created_at,
"updated_at": file_task.updated_at, "updated_at": file_task.updated_at,
"duration_seconds": file_task.duration_seconds, "duration_seconds": file_task.duration_seconds,
"filename": file_task.filename,
} }
# Count running and pending files # Count running and pending files
@ -322,6 +341,7 @@ class TaskService:
"created_at": file_task.created_at, "created_at": file_task.created_at,
"updated_at": file_task.updated_at, "updated_at": file_task.updated_at,
"duration_seconds": file_task.duration_seconds, "duration_seconds": file_task.duration_seconds,
"filename": file_task.filename,
} }
if file_task.status.value == "running": if file_task.status.value == "running":

331
src/tui/managers/docling_manager.py
View file

@ -31,10 +31,14 @@ class DoclingManager:
self._process: Optional[subprocess.Popen] = None self._process: Optional[subprocess.Popen] = None
self._port = 5001 self._port = 5001
self._host = "127.0.0.1" self._host = self._get_host_for_containers() # Get appropriate host IP based on runtime
self._running = False self._running = False
self._external_process = False self._external_process = False
# PID file to track docling-serve across sessions (in current working directory)
from pathlib import Path
self._pid_file = Path.cwd() / ".docling.pid"
# Log storage - simplified, no queue # Log storage - simplified, no queue
self._log_buffer: List[str] = [] self._log_buffer: List[str] = []
self._max_log_lines = 1000 self._max_log_lines = 1000
@ -42,21 +46,197 @@ class DoclingManager:
self._initialized = True self._initialized = True
def cleanup(self): # Try to recover existing process from PID file
"""Cleanup resources and stop any running processes.""" self._recover_from_pid_file()
if self._process and self._process.poll() is None:
self._add_log_entry("Cleaning up docling-serve process on exit")
try:
self._process.terminate()
self._process.wait(timeout=5)
except subprocess.TimeoutExpired:
self._process.kill()
self._process.wait()
except Exception as e:
self._add_log_entry(f"Error during cleanup: {e}")
self._running = False def _get_host_for_containers(self) -> str:
self._process = None """
Return a host IP that containers can reach (a bridge/CNI gateway).
Prefers Docker/Podman network gateways; falls back to bridge interfaces.
"""
import subprocess, json, shutil, re, logging
logger = logging.getLogger(__name__)
def run(cmd, timeout=2, text=True):
return subprocess.run(cmd, capture_output=True, text=text, timeout=timeout)
gateways = []
compose_gateways = [] # Highest priority - compose project networks
active_gateways = [] # Medium priority - networks with containers
# ---- Docker: enumerate networks and collect gateways
if shutil.which("docker"):
try:
ls = run(["docker", "network", "ls", "--format", "{{.Name}}"])
if ls.returncode == 0:
for name in filter(None, ls.stdout.splitlines()):
try:
insp = run(["docker", "network", "inspect", name, "--format", "{{json .}}"])
if insp.returncode == 0 and insp.stdout.strip():
nw = json.loads(insp.stdout)[0] if insp.stdout.strip().startswith("[") else json.loads(insp.stdout)
ipam = nw.get("IPAM", {})
containers = nw.get("Containers", {})
for cfg in ipam.get("Config", []) or []:
gw = cfg.get("Gateway")
if gw:
# Highest priority: compose networks (ending in _default)
if name.endswith("_default"):
compose_gateways.append(gw)
# Medium priority: networks with active containers
elif len(containers) > 0:
active_gateways.append(gw)
# Low priority: empty networks
else:
gateways.append(gw)
except Exception:
pass
except Exception:
pass
# ---- Podman: enumerate networks and collect gateways (netavark)
if shutil.which("podman"):
try:
# modern podman supports JSON format
ls = run(["podman", "network", "ls", "--format", "json"])
if ls.returncode == 0 and ls.stdout.strip():
for net in json.loads(ls.stdout):
name = net.get("name") or net.get("Name")
if not name:
continue
try:
insp = run(["podman", "network", "inspect", name, "--format", "json"])
if insp.returncode == 0 and insp.stdout.strip():
arr = json.loads(insp.stdout)
for item in (arr if isinstance(arr, list) else [arr]):
for sn in item.get("subnets", []) or []:
gw = sn.get("gateway")
if gw:
# Prioritize compose/project networks
if name.endswith("_default") or "_" in name:
compose_gateways.append(gw)
else:
gateways.append(gw)
except Exception:
pass
except Exception:
pass
# ---- Fallback: parse host interfaces for common bridges
if not gateways:
try:
if shutil.which("ip"):
show = run(["ip", "-o", "-4", "addr", "show"])
if show.returncode == 0:
for line in show.stdout.splitlines():
# e.g. "12: br-3f0f... inet 172.18.0.1/16 ..."
m = re.search(r"^\d+:\s+([a-zA-Z0-9_.:-]+)\s+.*\binet\s+(\d+\.\d+\.\d+\.\d+)/", line)
if not m:
continue
ifname, ip = m.group(1), m.group(2)
if ifname == "docker0" or ifname.startswith(("br-", "cni")):
gateways.append(ip)
else:
# As a last resort, try net-tools ifconfig output
if shutil.which("ifconfig"):
show = run(["ifconfig"])
for block in show.stdout.split("\n\n"):
if any(block.strip().startswith(n) for n in ("docker0", "cni", "br-")):
m = re.search(r"inet (?:addr:)?(\d+\.\d+\.\d+\.\d+)", block)
if m:
gateways.append(m.group(1))
except Exception:
pass
# Dedup, prioritizing: 1) compose networks, 2) active networks, 3) all others
seen, uniq = set(), []
# First: compose project networks (_default suffix)
for ip in compose_gateways:
if ip not in seen:
uniq.append(ip)
seen.add(ip)
# Second: networks with active containers
for ip in active_gateways:
if ip not in seen:
uniq.append(ip)
seen.add(ip)
# Third: all other gateways
for ip in gateways:
if ip not in seen:
uniq.append(ip)
seen.add(ip)
if uniq:
if len(uniq) > 1:
logger.info("Container-reachable host IP candidates: %s", ", ".join(uniq))
else:
logger.info("Container-reachable host IP: %s", uniq[0])
return uniq[0]
# Nothing found: warn clearly
logger.warning(
"No container bridge IP found. If using rootless Podman (slirp4netns), there is no host bridge; publish ports or use 10.0.2.2 from the container."
)
# Returning localhost is honest only for same-namespace; keep it explicit:
return "127.0.0.1"
def cleanup(self):
"""Cleanup resources but keep docling-serve running across sessions."""
# Don't stop the process on exit - let it persist
# Just clean up our references
self._add_log_entry("TUI exiting - docling-serve will continue running")
# Note: We keep the PID file so we can reconnect in future sessions
def _save_pid(self, pid: int) -> None:
"""Save the process PID to a file for persistence across sessions."""
try:
self._pid_file.write_text(str(pid))
self._add_log_entry(f"Saved PID {pid} to {self._pid_file}")
except Exception as e:
self._add_log_entry(f"Failed to save PID file: {e}")
def _load_pid(self) -> Optional[int]:
"""Load the process PID from file."""
try:
if self._pid_file.exists():
pid_str = self._pid_file.read_text().strip()
if pid_str.isdigit():
return int(pid_str)
except Exception as e:
self._add_log_entry(f"Failed to load PID file: {e}")
return None
def _clear_pid_file(self) -> None:
"""Remove the PID file."""
try:
if self._pid_file.exists():
self._pid_file.unlink()
self._add_log_entry("Cleared PID file")
except Exception as e:
self._add_log_entry(f"Failed to clear PID file: {e}")
def _is_process_running(self, pid: int) -> bool:
"""Check if a process with the given PID is running."""
try:
# Send signal 0 to check if process exists (doesn't actually send a signal)
os.kill(pid, 0)
return True
except OSError:
return False
def _recover_from_pid_file(self) -> None:
"""Try to recover connection to existing docling-serve process from PID file."""
pid = self._load_pid()
if pid is not None:
if self._is_process_running(pid):
self._add_log_entry(f"Recovered existing docling-serve process (PID: {pid})")
# Mark as external process since we didn't start it in this session
self._external_process = True
self._running = True
# Note: We don't have a Popen object for this process, but that's OK
# We'll detect it's running via port check
else:
self._add_log_entry(f"Stale PID file found (PID: {pid} not running)")
self._clear_pid_file()
def _add_log_entry(self, message: str) -> None: def _add_log_entry(self, message: str) -> None:
"""Add a log entry to the buffer (thread-safe).""" """Add a log entry to the buffer (thread-safe)."""
@ -70,43 +250,35 @@ class DoclingManager:
self._log_buffer = self._log_buffer[-self._max_log_lines:] self._log_buffer = self._log_buffer[-self._max_log_lines:]
def is_running(self) -> bool: def is_running(self) -> bool:
"""Check if docling serve is running.""" """Check if docling serve is running (by PID only)."""
# First check our internal state # Check if we have a direct process handle
internal_running = self._running and self._process is not None and self._process.poll() is None if self._process is not None and self._process.poll() is None:
self._running = True
# If we think it's not running, check if something is listening on the port
# This handles cases where docling-serve was started outside the TUI
if not internal_running:
try:
import socket
s = socket.socket(socket.AF_INET, socket.SOCK_STREAM)
s.settimeout(0.5)
result = s.connect_ex((self._host, self._port))
s.close()
# If port is in use, something is running there
if result == 0:
# Only log this once when we first detect external process
if not self._external_process:
self._add_log_entry(f"Detected external docling-serve running on {self._host}:{self._port}")
# Set a flag to indicate this is an external process
self._external_process = True
return True
except Exception as e:
# Only log errors occasionally to avoid spam
if not hasattr(self, '_last_port_error') or self._last_port_error != str(e):
self._add_log_entry(f"Error checking port: {e}")
self._last_port_error = str(e)
else:
# If we started it, it's not external
self._external_process = False self._external_process = False
return True
return internal_running # Check if we have a PID from file
pid = self._load_pid()
if pid is not None and self._is_process_running(pid):
self._running = True
self._external_process = True
return True
# No running process found
self._running = False
self._external_process = False
return False
def get_status(self) -> Dict[str, Any]: def get_status(self) -> Dict[str, Any]:
"""Get current status of docling serve.""" """Get current status of docling serve."""
if self.is_running(): if self.is_running():
pid = self._process.pid if self._process else None # Try to get PID from process handle first, then from PID file
pid = None
if self._process:
pid = self._process.pid
else:
pid = self._load_pid()
return { return {
"status": "running", "status": "running",
"port": self._port, "port": self._port,
@ -127,13 +299,28 @@ class DoclingManager:
"pid": None "pid": None
} }
async def start(self, port: int = 5001, host: str = "127.0.0.1", enable_ui: bool = False) -> Tuple[bool, str]: async def start(self, port: int = 5001, host: Optional[str] = None, enable_ui: bool = False) -> Tuple[bool, str]:
"""Start docling serve as external process.""" """Start docling serve as external process."""
if self.is_running(): if self.is_running():
return False, "Docling serve is already running" return False, "Docling serve is already running"
self._port = port self._port = port
self._host = host # Use provided host or the bridge IP we detected in __init__
if host is not None:
self._host = host
# else: keep self._host as already set in __init__
# Check if port is already in use before trying to start
import socket
try:
s = socket.socket(socket.AF_INET, socket.SOCK_STREAM)
s.settimeout(0.5)
result = s.connect_ex((self._host, self._port))
s.close()
if result == 0:
return False, f"Port {self._port} on {self._host} is already in use by another process. Please stop it first."
except Exception as e:
self._add_log_entry(f"Error checking port availability: {e}")
# Clear log buffer when starting # Clear log buffer when starting
self._log_buffer = [] self._log_buffer = []
@ -146,14 +333,14 @@ class DoclingManager:
if shutil.which("uv") and (os.path.exists("pyproject.toml") or os.getenv("VIRTUAL_ENV")): if shutil.which("uv") and (os.path.exists("pyproject.toml") or os.getenv("VIRTUAL_ENV")):
cmd = [ cmd = [
"uv", "run", "python", "-m", "docling_serve", "run", "uv", "run", "python", "-m", "docling_serve", "run",
"--host", host, "--host", self._host,
"--port", str(port), "--port", str(self._port),
] ]
else: else:
cmd = [ cmd = [
sys.executable, "-m", "docling_serve", "run", sys.executable, "-m", "docling_serve", "run",
"--host", host, "--host", self._host,
"--port", str(port), "--port", str(self._port),
] ]
if enable_ui: if enable_ui:
@ -173,6 +360,9 @@ class DoclingManager:
self._running = True self._running = True
self._add_log_entry("External process started") self._add_log_entry("External process started")
# Save the PID to file for persistence
self._save_pid(self._process.pid)
# Start a thread to capture output # Start a thread to capture output
self._start_output_capture() self._start_output_capture()
@ -192,11 +382,11 @@ class DoclingManager:
import socket import socket
s = socket.socket(socket.AF_INET, socket.SOCK_STREAM) s = socket.socket(socket.AF_INET, socket.SOCK_STREAM)
s.settimeout(0.5) s.settimeout(0.5)
result = s.connect_ex((host, port)) result = s.connect_ex((self._host, self._port))
s.close() s.close()
if result == 0: if result == 0:
self._add_log_entry(f"Docling-serve is now listening on {host}:{port}") self._add_log_entry(f"Docling-serve is now listening on {self._host}:{self._port}")
break break
except: except:
pass pass
@ -298,9 +488,12 @@ class DoclingManager:
try: try:
self._add_log_entry("Stopping docling-serve process") self._add_log_entry("Stopping docling-serve process")
pid_to_stop = None
if self._process: if self._process:
# We started this process, so we can stop it directly # We have a direct process handle
self._add_log_entry(f"Terminating our process (PID: {self._process.pid})") pid_to_stop = self._process.pid
self._add_log_entry(f"Terminating our process (PID: {pid_to_stop})")
self._process.terminate() self._process.terminate()
# Wait for it to stop # Wait for it to stop
@ -315,16 +508,32 @@ class DoclingManager:
self._add_log_entry("Process force killed") self._add_log_entry("Process force killed")
elif self._external_process: elif self._external_process:
# This is an external process, we can't stop it directly # This is a process we recovered from PID file
self._add_log_entry("Cannot stop external docling-serve process - it was started outside the TUI") pid_to_stop = self._load_pid()
self._running = False if pid_to_stop and self._is_process_running(pid_to_stop):
self._external_process = False self._add_log_entry(f"Stopping process from PID file (PID: {pid_to_stop})")
return False, "Cannot stop external docling-serve process. Please stop it manually." try:
os.kill(pid_to_stop, 15) # SIGTERM
# Wait a bit for graceful shutdown
await asyncio.sleep(2)
if self._is_process_running(pid_to_stop):
# Still running, force kill
self._add_log_entry(f"Force killing process (PID: {pid_to_stop})")
os.kill(pid_to_stop, 9) # SIGKILL
except Exception as e:
self._add_log_entry(f"Error stopping external process: {e}")
return False, f"Error stopping external process: {str(e)}"
else:
self._add_log_entry("External process not found")
return False, "Process not found"
self._running = False self._running = False
self._process = None self._process = None
self._external_process = False self._external_process = False
# Clear the PID file since we intentionally stopped the service
self._clear_pid_file()
self._add_log_entry("Docling serve stopped successfully") self._add_log_entry("Docling serve stopped successfully")
return True, "Docling serve stopped successfully" return True, "Docling serve stopped successfully"

184
src/tui/screens/welcome.py
View file

@ -118,9 +118,16 @@ class WelcomeScreen(Screen):
welcome_text.append(ascii_art, style="bold white") welcome_text.append(ascii_art, style="bold white")
welcome_text.append("Terminal User Interface for OpenRAG\n\n", style="dim") welcome_text.append("Terminal User Interface for OpenRAG\n\n", style="dim")
if self.services_running: # Check if all services are running
all_services_running = self.services_running and self.docling_running
if all_services_running:
welcome_text.append( welcome_text.append(
"✓ Services are currently running\n\n", style="bold green" "✓ All services are running\n\n", style="bold green"
)
elif self.services_running or self.docling_running:
welcome_text.append(
"⚠ Some services are running\n\n", style="bold yellow"
) )
elif self.has_oauth_config: elif self.has_oauth_config:
welcome_text.append( welcome_text.append(
@ -140,16 +147,19 @@ class WelcomeScreen(Screen):
buttons = [] buttons = []
if self.services_running: # Check if all services (native + container) are running
# Services running - show app link first, then stop services all_services_running = self.services_running and self.docling_running
if all_services_running:
# All services running - show app link first, then stop all
buttons.append( buttons.append(
Button("Launch OpenRAG", variant="success", id="open-app-btn") Button("Launch OpenRAG", variant="success", id="open-app-btn")
) )
buttons.append( buttons.append(
Button("Stop Container Services", variant="error", id="stop-services-btn") Button("Stop All Services", variant="error", id="stop-all-services-btn")
) )
else: else:
# Services not running - show setup options and start services # Some or no services running - show setup options and start all
if has_oauth: if has_oauth:
# If OAuth is configured, only show advanced setup # If OAuth is configured, only show advanced setup
buttons.append( buttons.append(
@ -165,25 +175,7 @@ class WelcomeScreen(Screen):
) )
buttons.append( buttons.append(
Button("Start Container Services", variant="primary", id="start-services-btn") Button("Start All Services", variant="primary", id="start-all-services-btn")
)
# Native services controls
if self.docling_running:
buttons.append(
Button(
"Stop Native Services",
variant="warning",
id="stop-native-services-btn",
)
)
else:
buttons.append(
Button(
"Start Native Services",
variant="primary",
id="start-native-services-btn",
)
) )
# Always show status option # Always show status option
@ -213,7 +205,7 @@ class WelcomeScreen(Screen):
) )
# Set default button focus # Set default button focus
if self.services_running: if self.services_running and self.docling_running:
self.default_button_id = "open-app-btn" self.default_button_id = "open-app-btn"
elif self.has_oauth_config: elif self.has_oauth_config:
self.default_button_id = "advanced-setup-btn" self.default_button_id = "advanced-setup-btn"
@ -234,7 +226,7 @@ class WelcomeScreen(Screen):
def _focus_appropriate_button(self) -> None: def _focus_appropriate_button(self) -> None:
"""Focus the appropriate button based on current state.""" """Focus the appropriate button based on current state."""
try: try:
if self.services_running: if self.services_running and self.docling_running:
self.query_one("#open-app-btn").focus() self.query_one("#open-app-btn").focus()
elif self.has_oauth_config: elif self.has_oauth_config:
self.query_one("#advanced-setup-btn").focus() self.query_one("#advanced-setup-btn").focus()
@ -253,20 +245,16 @@ class WelcomeScreen(Screen):
self.action_monitor() self.action_monitor()
elif event.button.id == "diagnostics-btn": elif event.button.id == "diagnostics-btn":
self.action_diagnostics() self.action_diagnostics()
elif event.button.id == "start-services-btn": elif event.button.id == "start-all-services-btn":
self.action_start_stop_services() self.action_start_all_services()
elif event.button.id == "stop-services-btn": elif event.button.id == "stop-all-services-btn":
self.action_start_stop_services() self.action_stop_all_services()
elif event.button.id == "start-native-services-btn":
self.action_start_native_services()
elif event.button.id == "stop-native-services-btn":
self.action_stop_native_services()
elif event.button.id == "open-app-btn": elif event.button.id == "open-app-btn":
self.action_open_app() self.action_open_app()
def action_default_action(self) -> None: def action_default_action(self) -> None:
"""Handle Enter key - go to default action based on state.""" """Handle Enter key - go to default action based on state."""
if self.services_running: if self.services_running and self.docling_running:
self.action_open_app() self.action_open_app()
elif self.has_oauth_config: elif self.has_oauth_config:
self.action_full_setup() self.action_full_setup()
@ -297,28 +285,13 @@ class WelcomeScreen(Screen):
self.app.push_screen(DiagnosticsScreen()) self.app.push_screen(DiagnosticsScreen())
def action_start_stop_services(self) -> None: def action_start_all_services(self) -> None:
"""Start or stop all services (containers + docling).""" """Start all services (native first, then containers)."""
if self.services_running: self.run_worker(self._start_all_services())
# Stop services - show modal with progress
if self.container_manager.is_available(): def action_stop_all_services(self) -> None:
command_generator = self.container_manager.stop_services() """Stop all services (containers first, then native)."""
modal = CommandOutputModal( self.run_worker(self._stop_all_services())
"Stopping Services",
command_generator,
on_complete=self._on_services_operation_complete,
)
self.app.push_screen(modal)
else:
# Start services - show modal with progress
if self.container_manager.is_available():
command_generator = self.container_manager.start_services()
modal = CommandOutputModal(
"Starting Services",
command_generator,
on_complete=self._on_services_operation_complete,
)
self.app.push_screen(modal)
async def _on_services_operation_complete(self) -> None: async def _on_services_operation_complete(self) -> None:
"""Handle completion of services start/stop operation.""" """Handle completion of services start/stop operation."""
@ -334,7 +307,7 @@ class WelcomeScreen(Screen):
def _update_default_button(self) -> None: def _update_default_button(self) -> None:
"""Update the default button target based on state.""" """Update the default button target based on state."""
if self.services_running: if self.services_running and self.docling_running:
self.default_button_id = "open-app-btn" self.default_button_id = "open-app-btn"
elif self.has_oauth_config: elif self.has_oauth_config:
self.default_button_id = "advanced-setup-btn" self.default_button_id = "advanced-setup-btn"
@ -362,51 +335,84 @@ class WelcomeScreen(Screen):
self.call_after_refresh(self._focus_appropriate_button) self.call_after_refresh(self._focus_appropriate_button)
def action_start_native_services(self) -> None: async def _start_all_services(self) -> None:
"""Start native services (docling).""" """Start all services: containers first, then native services."""
if self.docling_running: # Step 1: Start container services first (to create the network)
self.notify("Native services are already running.", severity="warning") if self.container_manager.is_available():
return command_generator = self.container_manager.start_services()
modal = CommandOutputModal(
"Starting Container Services",
command_generator,
on_complete=self._on_containers_started_start_native,
)
self.app.push_screen(modal)
else:
self.notify("No container runtime available", severity="warning")
# Still try to start native services
await self._start_native_services_after_containers()
self.run_worker(self._start_native_services()) async def _on_containers_started_start_native(self) -> None:
"""Called after containers start successfully, now start native services."""
# Update container state
self._detect_services_sync()
async def _start_native_services(self) -> None: # Now start native services (docling-serve can now detect the compose network)
"""Worker task to start native services.""" await self._start_native_services_after_containers()
try:
async def _start_native_services_after_containers(self) -> None:
"""Start native services after containers have been started."""
if not self.docling_manager.is_running():
self.notify("Starting native services...", severity="information")
success, message = await self.docling_manager.start() success, message = await self.docling_manager.start()
if success: if success:
self.docling_running = True
self.notify(message, severity="information") self.notify(message, severity="information")
else: else:
self.notify(f"Failed to start native services: {message}", severity="error") self.notify(f"Failed to start native services: {message}", severity="error")
except Exception as exc: else:
self.notify(f"Error starting native services: {exc}", severity="error") self.notify("Native services already running", severity="information")
finally:
self.docling_running = self.docling_manager.is_running()
await self._refresh_welcome_content()
def action_stop_native_services(self) -> None: # Update state
"""Stop native services (docling).""" self.docling_running = self.docling_manager.is_running()
if not self.docling_running and not self.docling_manager.is_running(): await self._refresh_welcome_content()
self.notify("Native services are not running.", severity="warning")
return
self.run_worker(self._stop_native_services()) async def _stop_all_services(self) -> None:
"""Stop all services: containers first, then native."""
# Step 1: Stop container services
if self.container_manager.is_available() and self.services_running:
command_generator = self.container_manager.stop_services()
modal = CommandOutputModal(
"Stopping Container Services",
command_generator,
on_complete=self._on_stop_containers_complete,
)
self.app.push_screen(modal)
else:
# No containers to stop, go directly to stopping native services
await self._stop_native_services_after_containers()
async def _stop_native_services(self) -> None: async def _on_stop_containers_complete(self) -> None:
"""Worker task to stop native services.""" """Called after containers are stopped, now stop native services."""
try: # Update container state
self._detect_services_sync()
# Now stop native services
await self._stop_native_services_after_containers()
async def _stop_native_services_after_containers(self) -> None:
"""Stop native services after containers have been stopped."""
if self.docling_manager.is_running():
self.notify("Stopping native services...", severity="information")
success, message = await self.docling_manager.stop() success, message = await self.docling_manager.stop()
if success: if success:
self.docling_running = False
self.notify(message, severity="information") self.notify(message, severity="information")
else: else:
self.notify(f"Failed to stop native services: {message}", severity="error") self.notify(f"Failed to stop native services: {message}", severity="error")
except Exception as exc: else:
self.notify(f"Error stopping native services: {exc}", severity="error") self.notify("Native services already stopped", severity="information")
finally:
self.docling_running = self.docling_manager.is_running() # Update state
await self._refresh_welcome_content() self.docling_running = self.docling_manager.is_running()
await self._refresh_welcome_content()
def action_open_app(self) -> None: def action_open_app(self) -> None:
"""Open the OpenRAG app in the default browser.""" """Open the OpenRAG app in the default browser."""

55
src/utils/opensearch_queries.py Normal file
View file

@ -0,0 +1,55 @@
"""
Utility functions for constructing OpenSearch queries consistently.
"""
from typing import Union, List
def build_filename_query(filename: str) -> dict:
"""
Build a standardized query for finding documents by filename.
Args:
filename: The exact filename to search for
Returns:
A dict containing the OpenSearch query body
"""
return {
"term": {
"filename": filename
}
}
def build_filename_search_body(filename: str, size: int = 1, source: Union[bool, List[str]] = False) -> dict:
"""
Build a complete search body for checking if a filename exists.
Args:
filename: The exact filename to search for
size: Number of results to return (default: 1)
source: Whether to include source fields, or list of specific fields to include (default: False)
Returns:
A dict containing the complete OpenSearch search body
"""
return {
"query": build_filename_query(filename),
"size": size,
"_source": source
}
def build_filename_delete_body(filename: str) -> dict:
"""
Build a delete-by-query body for removing all documents with a filename.
Args:
filename: The exact filename to delete
Returns:
A dict containing the OpenSearch delete-by-query body
"""
return {
"query": build_filename_query(filename)
}