fix: complete exploration data flow

This commit is contained in:
2026-07-23 22:25:12 +08:00
parent 55ff320cec
commit 5a1b04c4ed
12 changed files with 952 additions and 189 deletions

View File

@@ -35,17 +35,17 @@ const visualExperts = [
let visualDatasetSources = [
{
id: '019b0000-0000-7000-8000-000000000030', name: '手动收集', kind: 'manual', url: '', iconUrl: '',
description: '手动收集的文章与线索', enabled: true, lastSyncedAt: null, itemCount: 1,
description: '手动收集的文章与线索', enabled: true, builtIn: false, lastSyncedAt: null, itemCount: 1,
createdAt: '2026-07-22T06:00:00Z', updatedAt: '2026-07-22T06:00:00Z',
},
{
id: '019b0000-0000-7000-8000-000000000031', name: '需求文档', kind: 'link', url: 'https://example.com/requirements', iconUrl: '',
description: '产品需求与业务文档', enabled: true, lastSyncedAt: null, itemCount: 1,
description: '产品需求与业务文档', enabled: true, builtIn: false, lastSyncedAt: null, itemCount: 1,
createdAt: '2026-07-22T06:00:00Z', updatedAt: '2026-07-22T06:00:00Z',
},
{
id: '019b0000-0000-7000-8000-000000000032', name: '架构讨论', kind: 'link', url: 'https://example.com/architecture', iconUrl: '',
description: '架构方案与系统设计讨论', enabled: true, lastSyncedAt: null, itemCount: 0,
description: '架构方案与系统设计讨论', enabled: true, builtIn: false, lastSyncedAt: null, itemCount: 0,
createdAt: '2026-07-22T06:00:00Z', updatedAt: '2026-07-22T06:00:00Z',
},
]
@@ -225,6 +225,7 @@ await page.route('http://localhost:9150/api/v1/**', async (route) => {
const created = {
id: `019b0000-0000-7000-8000-${String(50 + visualDatasetSources.length).padStart(12, '0')}`,
...input,
builtIn: false,
lastSyncedAt: null,
itemCount: 0,
createdAt: new Date().toISOString(),
@@ -251,7 +252,24 @@ await page.route('http://localhost:9150/api/v1/**', async (route) => {
return
}
if (url.pathname === '/api/v1/dataset-items' && route.request().method() === 'GET') {
await route.fulfill({ json: visualDatasetItems })
const sourceId = url.searchParams.get('sourceId')
const offset = Number(url.searchParams.get('offset') ?? 0)
const limit = Number(url.searchParams.get('limit') ?? 50)
const matchingItems = sourceId
? visualDatasetItems.filter((item) => item.sourceId === sourceId)
: visualDatasetItems
await route.fulfill({
json: {
items: matchingItems.slice(offset, offset + limit),
total: matchingItems.length,
offset,
limit,
},
})
return
}
if (url.pathname.endsWith('/deposit') && route.request().method() === 'POST') {
await route.fulfill({ status: 201, json: { noteId: '019b0000-0000-7000-8000-000000000099', projectId } })
return
}
if (url.pathname.startsWith('/api/v1/dataset-items/') && route.request().method() === 'PATCH') {
@@ -281,6 +299,10 @@ await page.route('http://localhost:9150/api/v1/**', async (route) => {
})
return
}
if (url.pathname === '/api/v1/dataset-crons' && route.request().method() === 'GET') {
await route.fulfill({ json: [] })
return
}
await route.fulfill({ status: 404, json: { error: { code: 'not_found', message: '视觉检查未配置该接口' } } })
})

View File

@@ -10,6 +10,7 @@ export type DatasetSourceDTO = {
iconUrl: string
description: string
enabled: boolean
builtIn: boolean
lastSyncedAt: string | null
itemCount: number
createdAt: string
@@ -33,16 +34,25 @@ export type DatasetItemDTO = {
export type DatasetCronDTO = {
id: string
sourceId: string
schedule: string
status: string
enabled: boolean
nextRunAt: string | null
lastRunAt: string | null
lastResult: string
createdAt: string
updatedAt: string
}
export type DatasetItemPageDTO = {
items: DatasetItemDTO[]
total: number
limit: number
offset: number
}
export type DatasetDepositDTO = {
noteId: string
projectId: string
}
export type DatasetSourceInput = {
name: string
kind: DatasetSourceKind
@@ -53,8 +63,8 @@ export type DatasetSourceInput = {
}
export type DatasetItemUpdate = {
status: DatasetItemDTO['status']
starred: boolean
status?: DatasetItemDTO['status']
starred?: boolean
}
export function listDatasetSources(session: ApiSession, signal?: AbortSignal) {
@@ -85,9 +95,16 @@ export function deleteDatasetSource(session: ApiSession, sourceId: string) {
})
}
export function listDatasetItems(session: ApiSession, sourceId?: string, signal?: AbortSignal) {
const query = sourceId ? `?sourceId=${encodeURIComponent(sourceId)}` : ''
return apiRequest<DatasetItemDTO[]>(`/api/v1/dataset-items${query}`, { token: session.token, signal })
export function listDatasetItems(
session: ApiSession,
sourceId?: string,
offset = 0,
limit = 50,
signal?: AbortSignal,
) {
const params = new URLSearchParams({ offset: String(offset), limit: String(limit) })
if (sourceId) params.set('sourceId', sourceId)
return apiRequest<DatasetItemPageDTO>(`/api/v1/dataset-items?${params}`, { token: session.token, signal })
}
export function updateDatasetItem(session: ApiSession, itemId: string, input: DatasetItemUpdate) {
@@ -105,3 +122,15 @@ export function queueDatasetSync(session: ApiSession) {
body: {},
})
}
export function listDatasetCrons(session: ApiSession, signal?: AbortSignal) {
return apiRequest<DatasetCronDTO[]>('/api/v1/dataset-crons', { token: session.token, signal })
}
export function depositDatasetItem(session: ApiSession, itemId: string, projectId: string) {
return apiRequest<DatasetDepositDTO>(`/api/v1/dataset-items/${itemId}/deposit`, {
method: 'POST',
token: session.token,
body: { projectId },
})
}

View File

@@ -7,6 +7,8 @@ import { createAISession, listAIExperts, listAISessions, type CreateAISessionInp
import {
createDatasetSource,
deleteDatasetSource,
depositDatasetItem,
listDatasetCrons,
listDatasetItems,
listDatasetSources,
queueDatasetSync,
@@ -75,9 +77,9 @@ function App() {
return listDatasetSources(session, signal)
}, [session])
const handleListDatasetItems = useCallback((signal?: AbortSignal) => {
const handleListDatasetItems = useCallback((sourceId?: string, offset = 0, limit = 50, signal?: AbortSignal) => {
if (!session) return Promise.reject(new Error('未登录'))
return listDatasetItems(session, undefined, signal)
return listDatasetItems(session, sourceId, offset, limit, signal)
}, [session])
const handleCreateDatasetSource = useCallback((input: DatasetSourceInput) => {
@@ -105,6 +107,27 @@ function App() {
return queueDatasetSync(session)
}, [session])
const handleListDatasetCrons = useCallback((signal?: AbortSignal) => {
if (!session) return Promise.reject(new Error('未登录'))
return listDatasetCrons(session, signal)
}, [session])
const handleDepositDatasetItem = useCallback(async (itemId: string, projectId: string) => {
if (!session) throw new Error('未登录')
const result = await depositDatasetItem(session, itemId, projectId)
const workspaceIndex = workspaces.findIndex((workspace) => workspace.project.id === projectId)
if (workspaceIndex >= 0) {
try {
const payload = await fetchProjectWorkspace(session, projectId)
const refreshed = mapWorkspace(payload, workspaceIndex)
setWorkspaces((current) => current.map((workspace) => workspace.project.id === projectId ? refreshed : workspace))
} catch {
return { ...result, refreshFailed: true }
}
}
return result
}, [session, workspaces])
const dark = theme === 'dark'
const activeWorkspace = workspaces.find((workspace) => workspace.project.id === activeProjectID) ?? workspaces[0]
const activeTagOptions = activeWorkspace?.tags.filter((tag) => tag !== 'all' && tag !== '全部') ?? []
@@ -418,6 +441,8 @@ function App() {
onDeleteDatasetSource={handleDeleteDatasetSource}
onUpdateDatasetItem={handleUpdateDatasetItem}
onQueueDatasetSync={handleQueueDatasetSync}
onListDatasetCrons={handleListDatasetCrons}
onDepositDatasetItem={handleDepositDatasetItem}
/>
) : (
<Spin loading />

View File

@@ -1,6 +1,6 @@
import { useEffect, useMemo, useState } from 'react'
import type { ReactNode } from 'react'
import { Alert, Button, Card, Empty, Grid, Input, Message, Modal, Select, Space, Spin, Switch, Typography } from '@arco-design/web-react'
import { Alert, Button, Card, Empty, Grid, Input, Message, Modal, Space, Spin, Switch, Typography } from '@arco-design/web-react'
import {
IconBook,
IconCheckCircle,
@@ -15,7 +15,9 @@ import {
} from '@arco-design/web-react/icon'
import type {
DatasetCronDTO,
DatasetDepositDTO,
DatasetItemDTO,
DatasetItemPageDTO,
DatasetItemUpdate,
DatasetSourceDTO,
DatasetSourceInput,
@@ -25,6 +27,7 @@ import type {
const { Row, Col } = Grid
const { Title, Text, Paragraph } = Typography
const { TextArea } = Input
const ITEM_PAGE_SIZE = 50
type DataSourceCard = DatasetSourceDTO & {
icon: ReactNode
@@ -42,7 +45,7 @@ type SourceDraft = {
const EMPTY_SOURCE_DRAFT: SourceDraft = {
name: '',
kind: 'link',
kind: 'rss',
url: '',
iconUrl: '',
description: '',
@@ -50,7 +53,8 @@ const EMPTY_SOURCE_DRAFT: SourceDraft = {
}
export function WorkspaceExplorePage({
onSelectItem,
activeProjectID,
activeProjectName,
onListSources,
onListItems,
onCreateSource,
@@ -58,18 +62,24 @@ export function WorkspaceExplorePage({
onDeleteSource,
onUpdateItem,
onQueueSync,
onListCrons,
onDepositItem,
}: {
onSelectItem: (title: string) => void
activeProjectID: string
activeProjectName: string
onListSources: (signal?: AbortSignal) => Promise<DatasetSourceDTO[]>
onListItems: (signal?: AbortSignal) => Promise<DatasetItemDTO[]>
onListItems: (sourceId?: string, offset?: number, limit?: number, signal?: AbortSignal) => Promise<DatasetItemPageDTO>
onCreateSource: (input: DatasetSourceInput) => Promise<DatasetSourceDTO>
onUpdateSource: (sourceId: string, input: DatasetSourceInput) => Promise<DatasetSourceDTO>
onDeleteSource: (sourceId: string) => Promise<void>
onUpdateItem: (itemId: string, input: DatasetItemUpdate) => Promise<DatasetItemDTO>
onQueueSync: () => Promise<DatasetCronDTO[]>
onListCrons: (signal?: AbortSignal) => Promise<DatasetCronDTO[]>
onDepositItem: (itemId: string, projectId: string) => Promise<DatasetDepositDTO & { refreshFailed?: boolean }>
}) {
const [sourceRecords, setSourceRecords] = useState<DatasetSourceDTO[]>([])
const [items, setItems] = useState<DatasetItemDTO[]>([])
const [itemTotal, setItemTotal] = useState(0)
const [activeSourceID, setActiveSourceID] = useState('all')
const [activeItemID, setActiveItemID] = useState<string | null>(null)
const [sourceModalMode, setSourceModalMode] = useState<'add' | 'edit' | null>(null)
@@ -78,17 +88,18 @@ export function WorkspaceExplorePage({
const [sourceError, setSourceError] = useState('')
const [loadError, setLoadError] = useState('')
const [loading, setLoading] = useState(true)
const [itemsLoading, setItemsLoading] = useState(true)
const [saving, setSaving] = useState(false)
const [syncing, setSyncing] = useState(false)
const [depositingItemID, setDepositingItemID] = useState<string | null>(null)
useEffect(() => {
const controller = new AbortController()
setLoading(true)
setLoadError('')
Promise.all([onListSources(controller.signal), onListItems(controller.signal)])
.then(([nextSources, nextItems]) => {
onListSources(controller.signal)
.then((nextSources) => {
setSourceRecords(nextSources)
setItems(nextItems)
})
.catch((error: unknown) => {
if (!controller.signal.aborted) {
@@ -99,7 +110,31 @@ export function WorkspaceExplorePage({
if (!controller.signal.aborted) setLoading(false)
})
return () => controller.abort()
}, [onListItems, onListSources])
}, [onListSources])
useEffect(() => {
const controller = new AbortController()
setItemsLoading(true)
setLoadError('')
setItems([])
setItemTotal(0)
setActiveItemID(null)
const sourceID = activeSourceID === 'all' ? undefined : activeSourceID
onListItems(sourceID, 0, ITEM_PAGE_SIZE, controller.signal)
.then((page) => {
setItems(page.items)
setItemTotal(page.total)
})
.catch((error: unknown) => {
if (!controller.signal.aborted) {
setLoadError(error instanceof Error ? error.message : '探索内容加载失败')
}
})
.finally(() => {
if (!controller.signal.aborted) setItemsLoading(false)
})
return () => controller.abort()
}, [activeSourceID, onListItems])
const sources = useMemo<DataSourceCard[]>(() => {
const allSource: DataSourceCard = {
@@ -110,8 +145,9 @@ export function WorkspaceExplorePage({
iconUrl: '',
description: '全部探索内容',
enabled: true,
builtIn: true,
lastSyncedAt: null,
itemCount: items.length,
itemCount: sourceRecords.reduce((total, source) => total + source.itemCount, 0),
createdAt: '',
updatedAt: '',
icon: <IconStorage />,
@@ -127,12 +163,9 @@ export function WorkspaceExplorePage({
color: sourceColor(source.kind),
})),
]
}, [items.length, sourceRecords])
}, [sourceRecords])
const filteredItems = useMemo(
() => activeSourceID === 'all' ? items : items.filter((item) => item.sourceId === activeSourceID),
[activeSourceID, items],
)
const filteredItems = items
useEffect(() => {
if (filteredItems.length === 0) {
@@ -147,6 +180,9 @@ export function WorkspaceExplorePage({
const selected = filteredItems.find((item) => item.id === activeItemID) ?? filteredItems[0]
const activeSource = sourceByID(sources, activeSourceID)
const selectedSource = selected ? sourceByID(sources, selected.sourceId) : activeSource
const editingSource = editingSourceID
? sourceRecords.find((source) => source.id === editingSourceID)
: undefined
function openAddSource() {
setEditingSourceID(null)
@@ -174,7 +210,7 @@ export function WorkspaceExplorePage({
name: sourceDraft.name.trim(),
kind: sourceDraft.kind,
url: sourceDraft.url.trim(),
iconUrl: sourceDraft.iconUrl,
iconUrl: sourceDraft.iconUrl.trim(),
description: sourceDraft.description.trim(),
enabled: sourceDraft.enabled,
}
@@ -182,8 +218,8 @@ export function WorkspaceExplorePage({
setSourceError('请输入数据源名称')
return
}
if (input.kind !== 'manual' && !input.url) {
setSourceError('链接和 RSS 数据源必须填写地址')
if (input.kind === 'rss' && !input.url) {
setSourceError('RSS 数据源必须填写地址')
return
}
setSaving(true)
@@ -229,14 +265,21 @@ export function WorkspaceExplorePage({
async function syncSources() {
setSyncing(true)
try {
const crons = await onQueueSync()
if (crons.length === 0) {
const queued = await onQueueSync()
if (queued.length === 0) {
Message.info('暂无已启用的 RSS 数据源')
} else {
const failed = crons.filter((cron) => cron.status === 'failed').length
const [nextSources, nextItems] = await Promise.all([onListSources(), onListItems()])
Message.info(`已提交 ${queued.length} 个采集任务`)
const crons = await waitForDatasetRuns(queued, onListCrons)
const sourceID = activeSourceID === 'all' ? undefined : activeSourceID
const [nextSources, nextPage] = await Promise.all([
onListSources(),
onListItems(sourceID, 0, ITEM_PAGE_SIZE),
])
setSourceRecords(nextSources)
setItems(nextItems)
setItems(nextPage.items)
setItemTotal(nextPage.total)
const failed = crons.filter((cron) => cron.status === 'failed').length
if (failed > 0) {
Message.warning(`${crons.length - failed} 个数据源采集完成,${failed} 个失败`)
} else {
@@ -250,6 +293,21 @@ export function WorkspaceExplorePage({
}
}
async function loadMoreItems() {
if (itemsLoading || items.length >= itemTotal) return
setItemsLoading(true)
try {
const sourceID = activeSourceID === 'all' ? undefined : activeSourceID
const page = await onListItems(sourceID, items.length, ITEM_PAGE_SIZE)
setItems((current) => [...current, ...page.items])
setItemTotal(page.total)
} catch (error) {
Message.error(error instanceof Error ? error.message : '更多探索内容加载失败')
} finally {
setItemsLoading(false)
}
}
async function updateSelected(input: DatasetItemUpdate) {
if (!selected) return
try {
@@ -260,6 +318,23 @@ export function WorkspaceExplorePage({
}
}
async function depositSelected() {
if (!selected || !activeProjectID) return
setDepositingItemID(selected.id)
try {
const result = await onDepositItem(selected.id, activeProjectID)
if (result.refreshFailed) {
Message.warning(`笔记已沉淀到项目“${activeProjectName}”,但项目数据刷新失败`)
} else {
Message.success(`已沉淀到项目“${activeProjectName}”的笔记资料`)
}
} catch (error) {
Message.error(error instanceof Error ? error.message : '沉淀到项目失败')
} finally {
setDepositingItemID(null)
}
}
return (
<div className="workspace-explore-page overview-page">
<div className="overview-head">
@@ -275,7 +350,7 @@ export function WorkspaceExplorePage({
{loadError ? <Alert type="error" content={loadError} /> : null}
<Spin loading={loading} style={{ width: '100%' }}>
<Spin loading={loading || (itemsLoading && items.length === 0)} style={{ width: '100%' }}>
<Row gutter={10} className="explore-source-row">
{sources.map((source) => (
<Col span={6} key={source.id}>
@@ -287,12 +362,14 @@ export function WorkspaceExplorePage({
<span className={`explore-source-icon ${source.color}`}>{source.icon}</span>
<span className="explore-source-copy">
<Text className="explore-source-name">{source.name}</Text>
<Text type="secondary">{source.itemCount} </Text>
<Text type="secondary">{sourceSubtitle(source)}</Text>
</span>
{source.id !== 'all' ? (
<span className="explore-source-actions" onClick={(event) => event.stopPropagation()}>
<Button aria-label={`编辑${source.name}`} type="text" size="mini" icon={<IconEdit />} onClick={() => openEditSource(source)} />
<Button aria-label={`删除${source.name}`} type="text" size="mini" status="danger" icon={<IconDelete />} onClick={() => deleteSource(source)} />
{!source.builtIn ? (
<Button aria-label={`删除${source.name}`} type="text" size="mini" status="danger" icon={<IconDelete />} onClick={() => deleteSource(source)} />
) : null}
</span>
) : null}
</Card>
@@ -319,10 +396,7 @@ export function WorkspaceExplorePage({
<button
key={item.id}
className={item.id === selected.id ? 'explore-article-item active' : 'explore-article-item'}
onClick={() => {
setActiveItemID(item.id)
onSelectItem(item.title)
}}
onClick={() => setActiveItemID(item.id)}
>
<span className={`explore-source-logo ${itemSource.color}`}>{sourceInitial(itemSource.name)}</span>
<span className="explore-article-copy">
@@ -335,6 +409,11 @@ export function WorkspaceExplorePage({
</button>
)
})}
{items.length < itemTotal ? (
<Button long type="text" loading={itemsLoading} onClick={() => void loadMoreItems()}>
{items.length}/{itemTotal}
</Button>
) : null}
</div>
</Card>
@@ -354,7 +433,13 @@ export function WorkspaceExplorePage({
icon={<IconStar />}
onClick={() => void updateSelected({ status: selected.status, starred: !selected.starred })}
/>
<Button type="text" icon={<IconBook />} onClick={() => onSelectItem(selected.title)} />
<Button
aria-label={`沉淀到项目${activeProjectName}`}
type="text"
icon={<IconBook />}
loading={depositingItemID === selected.id}
onClick={() => void depositSelected()}
/>
{selected.url ? <Button type="text" icon={<IconLaunch />} href={selected.url} target="_blank" /> : null}
</Space>
</div>
@@ -393,18 +478,25 @@ export function WorkspaceExplorePage({
</label>
<label>
<Text></Text>
<Select
value={sourceDraft.kind}
onChange={(kind) => setSourceDraft((current) => ({ ...current, kind: kind as DatasetSourceKind }))}
>
<Select.Option value="link"></Select.Option>
<Select.Option value="rss">RSS </Select.Option>
<Select.Option value="manual"></Select.Option>
</Select>
<Input value={sourceDraft.kind === 'rss' ? 'RSS 订阅' : '历史数据源(仅维护)'} disabled />
</label>
<label>
<Text></Text>
<Input value={sourceDraft.url} placeholder="https://example.com/feed" onChange={(url) => setSourceDraft((current) => ({ ...current, url }))} />
<Input
value={sourceDraft.url}
disabled={Boolean(editingSource?.builtIn || (editingSource && editingSource.kind !== 'rss'))}
placeholder="https://example.com/feed"
onChange={(url) => setSourceDraft((current) => ({ ...current, url }))}
/>
</label>
<label>
<Text></Text>
<Input
value={sourceDraft.iconUrl}
disabled={Boolean(editingSource?.builtIn || (editingSource && editingSource.kind !== 'rss'))}
placeholder="https://example.com/icon.png"
onChange={(iconUrl) => setSourceDraft((current) => ({ ...current, iconUrl }))}
/>
</label>
<label>
<Text></Text>
@@ -448,3 +540,26 @@ function itemTime(item: DatasetItemDTO) {
if (Number.isNaN(parsed.getTime())) return ''
return new Intl.DateTimeFormat('zh-CN', { dateStyle: 'medium' }).format(parsed)
}
function sourceSubtitle(source: DatasetSourceDTO) {
if (!source.lastSyncedAt) return `${source.itemCount}`
const parsed = new Date(source.lastSyncedAt)
if (Number.isNaN(parsed.getTime())) return `${source.itemCount}`
return `${source.itemCount} 条 · ${new Intl.DateTimeFormat('zh-CN', { dateStyle: 'short', timeStyle: 'short' }).format(parsed)}`
}
async function waitForDatasetRuns(
queued: DatasetCronDTO[],
listRuns: (signal?: AbortSignal) => Promise<DatasetCronDTO[]>,
) {
const queuedIDs = new Set(queued.map((run) => run.id))
const deadline = Date.now() + 10 * 60 * 1000
while (Date.now() < deadline) {
const runs = (await listRuns()).filter((run) => queuedIDs.has(run.id))
if (runs.length === queuedIDs.size && runs.every((run) => run.status === 'completed' || run.status === 'failed')) {
return runs
}
await new Promise((resolve) => setTimeout(resolve, 1500))
}
throw new Error('采集任务仍在后台运行,请稍后刷新')
}

View File

@@ -12,7 +12,9 @@ import type { InboxSuggestionDTO } from '../api/inbox'
import type { AIExpertDTO, AISessionDTO, CreateAISessionInput } from '../api/ai'
import type {
DatasetCronDTO,
DatasetDepositDTO,
DatasetItemDTO,
DatasetItemPageDTO,
DatasetItemUpdate,
DatasetSourceDTO,
DatasetSourceInput,
@@ -62,6 +64,8 @@ export function ProjectPage({
onDeleteDatasetSource,
onUpdateDatasetItem,
onQueueDatasetSync,
onListDatasetCrons,
onDepositDatasetItem,
}: {
activeView: WorkbenchView
activeWorkspace: ProjectWorkspace
@@ -97,12 +101,14 @@ export function ProjectPage({
onListAIExperts: (signal?: AbortSignal) => Promise<AIExpertDTO[]>
onCreateAISession: (projectId: string, input: CreateAISessionInput, signal?: AbortSignal) => Promise<AISessionDTO>
onListDatasetSources: (signal?: AbortSignal) => Promise<DatasetSourceDTO[]>
onListDatasetItems: (signal?: AbortSignal) => Promise<DatasetItemDTO[]>
onListDatasetItems: (sourceId?: string, offset?: number, limit?: number, signal?: AbortSignal) => Promise<DatasetItemPageDTO>
onCreateDatasetSource: (input: DatasetSourceInput) => Promise<DatasetSourceDTO>
onUpdateDatasetSource: (sourceId: string, input: DatasetSourceInput) => Promise<DatasetSourceDTO>
onDeleteDatasetSource: (sourceId: string) => Promise<void>
onUpdateDatasetItem: (itemId: string, input: DatasetItemUpdate) => Promise<DatasetItemDTO>
onQueueDatasetSync: () => Promise<DatasetCronDTO[]>
onListDatasetCrons: (signal?: AbortSignal) => Promise<DatasetCronDTO[]>
onDepositDatasetItem: (itemId: string, projectId: string) => Promise<DatasetDepositDTO & { refreshFailed?: boolean }>
}) {
const isProject = activeView === 'project'
const projects = workspaces.map((workspace) => workspace.project)
@@ -159,7 +165,8 @@ export function ProjectPage({
<WorkspacePage workspaces={workspaces} onOpenTask={onOpenTask} onUpdateTask={onUpdateWorkspaceTask} />
) : activeView === 'workspace-explore' ? (
<WorkspaceExplorePage
onSelectItem={onSelectItem}
activeProjectID={activeWorkspace.project.id}
activeProjectName={activeWorkspace.project.name}
onListSources={onListDatasetSources}
onListItems={onListDatasetItems}
onCreateSource={onCreateDatasetSource}
@@ -167,6 +174,8 @@ export function ProjectPage({
onDeleteSource={onDeleteDatasetSource}
onUpdateItem={onUpdateDatasetItem}
onQueueSync={onQueueDatasetSync}
onListCrons={onListDatasetCrons}
onDepositItem={onDepositDatasetItem}
/>
) : (
<ProjectChannelPage

View File

@@ -8,6 +8,7 @@ import (
"senlinai-agent/backend/internal/httpx"
"senlinai-agent/backend/internal/logic/ai"
"senlinai-agent/backend/internal/logic/auth"
"senlinai-agent/backend/internal/logic/dataset"
"senlinai-agent/backend/internal/logic/files"
"senlinai-agent/backend/internal/logic/inbox"
"senlinai-agent/backend/internal/logic/projects"
@@ -28,6 +29,7 @@ func TestMainRegistrarsRegisterTogether(t *testing.T) {
cfg,
authService.VerifySession,
auth.NewHandler(authService),
dataset.NewHandler(dataset.NewService(nil)),
projects.NewHandler(projectService),
projects.NewTagHandler(projectService),
projects.NewCronHandler(projectService),

View File

@@ -3,9 +3,11 @@ package dataset
import (
"context"
"errors"
"fmt"
"net/url"
"os"
"testing"
"time"
"github.com/stretchr/testify/require"
"senlinai-agent/backend/internal/models"
@@ -164,8 +166,68 @@ func TestSyncAllSourcesCollectsEnabledRSSSourcesForEveryUser(t *testing.T) {
require.Equal(t, int64(2), itemCount)
}
func TestSyncSourcesCollectsMultipleSourcesForOneOwner(t *testing.T) {
database := newDatasetTestDatabase(t)
user := createDatasetTestUser(t, database, "multi-feed-owner@example.com")
for index := 0; index < 5; index++ {
require.NoError(t, database.Create(&models.SaDatasetSource{
CreatedBy: user.ID, Name: fmt.Sprintf("Feed %d", index), Kind: "rss",
URL: fmt.Sprintf("https://example.com/feed-%d", index), Enabled: true,
}).Error)
}
runs, err := newServiceWithFetcher(database, staticFeedFetcher{}).SyncSources(context.Background(), user.ID)
require.NoError(t, err)
require.Len(t, runs, 5)
for _, run := range runs {
require.Equal(t, "completed", run.Status)
}
}
func TestQueueSourcesRunsInBackgroundAndRejectsOverlap(t *testing.T) {
database := newDatasetTestDatabase(t)
user := createDatasetTestUser(t, database, "queued-feed-owner@example.com")
source := models.SaDatasetSource{
CreatedBy: user.ID, Name: "Queued RSS", Kind: "rss",
URL: "https://example.com/feed", Enabled: true,
}
require.NoError(t, database.Create(&source).Error)
fetcher := &blockingFeedFetcher{
started: make(chan struct{}),
release: make(chan struct{}),
}
service := newServiceWithFetcher(database, fetcher)
queued, err := service.QueueSources(user.ID)
require.NoError(t, err)
require.Len(t, queued, 1)
require.Equal(t, "pending", queued[0].Status)
<-fetcher.started
_, err = service.QueueSources(user.ID)
require.ErrorIs(t, err, ErrSyncInProgress)
require.ErrorIs(t, service.DeleteSource(user.ID, source.Identity), ErrSyncInProgress)
close(fetcher.release)
require.Eventually(t, func() bool {
runs, listErr := service.ListCrons(user.ID)
return listErr == nil && len(runs) == 1 && runs[0].Status == "completed"
}, time.Second, 10*time.Millisecond)
}
type failingFeedFetcher struct{}
func (failingFeedFetcher) Fetch(context.Context, string) (ParsedFeed, error) {
return ParsedFeed{}, errors.New("feed unavailable")
}
type blockingFeedFetcher struct {
started chan struct{}
release chan struct{}
}
func (f *blockingFeedFetcher) Fetch(context.Context, string) (ParsedFeed, error) {
close(f.started)
<-f.release
return staticFeedFetcher{}.Fetch(context.Background(), "")
}

View File

@@ -3,6 +3,7 @@ package dataset
import (
"errors"
"net/http"
"strconv"
"strings"
"time"
@@ -25,6 +26,7 @@ type SourceDTO struct {
IconURL string `json:"iconUrl"`
Description string `json:"description"`
Enabled bool `json:"enabled"`
BuiltIn bool `json:"builtIn"`
LastSyncedAt *time.Time `json:"lastSyncedAt"`
ItemCount int64 `json:"itemCount"`
CreatedAt time.Time `json:"createdAt"`
@@ -45,19 +47,28 @@ type ItemDTO struct {
UpdatedAt time.Time `json:"updatedAt"`
}
type ItemPageDTO struct {
Items []ItemDTO `json:"items"`
Total int64 `json:"total"`
Limit int `json:"limit"`
Offset int `json:"offset"`
}
type CronDTO struct {
ID string `json:"id"`
SourceID string `json:"sourceId"`
Schedule string `json:"schedule"`
Status string `json:"status"`
Enabled bool `json:"enabled"`
NextRunAt *time.Time `json:"nextRunAt"`
LastRunAt *time.Time `json:"lastRunAt"`
LastResult string `json:"lastResult"`
CreatedAt time.Time `json:"createdAt"`
UpdatedAt time.Time `json:"updatedAt"`
}
type DepositDTO struct {
NoteID string `json:"noteId"`
ProjectID string `json:"projectId"`
}
type sourceRequest struct {
Name string `json:"name"`
Kind string `json:"kind"`
@@ -79,6 +90,7 @@ func (h *Handler) Register(router gin.IRouter) {
router.GET("/dataset-items", h.listItems)
router.POST("/dataset-items", h.createItem)
router.PATCH("/dataset-items/:id", h.updateItem)
router.POST("/dataset-items/:id/deposit", h.depositItem)
router.GET("/dataset-crons", h.listCrons)
router.POST("/dataset-crons/sync", h.queueSync)
}
@@ -134,18 +146,21 @@ func (h *Handler) updateSource(c *gin.Context) {
if !ok {
return
}
var input sourceRequest
var input struct {
Name *string `json:"name"`
Kind *string `json:"kind"`
URL *string `json:"url"`
IconURL *string `json:"iconUrl"`
Description *string `json:"description"`
Enabled *bool `json:"enabled"`
}
if err := c.ShouldBindJSON(&input); err != nil {
writeInvalidRequest(c)
return
}
enabled := true
if input.Enabled != nil {
enabled = *input.Enabled
}
source, err := h.service.UpdateSource(userID, identity, SourceInput{
source, err := h.service.PatchSource(userID, identity, SourcePatch{
Name: input.Name, Kind: input.Kind, URL: input.URL, IconURL: input.IconURL,
Description: input.Description, Enabled: enabled,
Description: input.Description, Enabled: input.Enabled,
})
if err != nil {
writeError(c, err)
@@ -180,16 +195,34 @@ func (h *Handler) listItems(c *gin.Context) {
if !ok {
return
}
items, err := h.service.ListItems(userID, c.Query("sourceId"))
limit, err := optionalNonNegativeInt(c.Query("limit"))
if err != nil {
writeInvalidRequest(c)
return
}
offset, err := optionalNonNegativeInt(c.Query("offset"))
if err != nil {
writeInvalidRequest(c)
return
}
page, err := h.service.ListItemsPage(userID, c.Query("sourceId"), limit, offset)
if err != nil {
writeError(c, err)
return
}
result := make([]ItemDTO, 0, len(items))
for _, item := range items {
result := make([]ItemDTO, 0, len(page.Items))
for _, item := range page.Items {
result = append(result, itemDTO(item))
}
c.JSON(http.StatusOK, result)
_, hasLimit := c.GetQuery("limit")
_, hasOffset := c.GetQuery("offset")
if !hasLimit && !hasOffset {
c.JSON(http.StatusOK, result)
return
}
c.JSON(http.StatusOK, ItemPageDTO{
Items: result, Total: page.Total, Limit: page.Limit, Offset: page.Offset,
})
}
func (h *Handler) createItem(c *gin.Context) {
@@ -235,14 +268,14 @@ func (h *Handler) updateItem(c *gin.Context) {
return
}
var input struct {
Status string `json:"status"`
Starred bool `json:"starred"`
Status *string `json:"status"`
Starred *bool `json:"starred"`
}
if err := c.ShouldBindJSON(&input); err != nil {
writeInvalidRequest(c)
return
}
item, err := h.service.UpdateItem(userID, identity, ItemUpdate{Status: input.Status, Starred: input.Starred})
item, err := h.service.PatchItem(userID, identity, ItemPatch{Status: input.Status, Starred: input.Starred})
if err != nil {
writeError(c, err)
return
@@ -250,6 +283,32 @@ func (h *Handler) updateItem(c *gin.Context) {
c.JSON(http.StatusOK, itemDTO(*item))
}
func (h *Handler) depositItem(c *gin.Context) {
userID, ok := currentUser(c)
if !ok {
return
}
identity, ok := httpx.IdentityParam(c, "id")
if !ok {
return
}
var input struct {
ProjectID string `json:"projectId"`
}
if err := c.ShouldBindJSON(&input); err != nil || strings.TrimSpace(input.ProjectID) == "" {
writeInvalidRequest(c)
return
}
result, err := h.service.DepositItem(userID, identity, input.ProjectID)
if err != nil {
writeError(c, err)
return
}
c.JSON(http.StatusCreated, DepositDTO{
NoteID: result.NoteIdentity, ProjectID: result.ProjectIdentity,
})
}
func (h *Handler) listCrons(c *gin.Context) {
userID, ok := currentUser(c)
if !ok {
@@ -272,7 +331,7 @@ func (h *Handler) queueSync(c *gin.Context) {
if !ok {
return
}
crons, err := h.service.SyncSources(c.Request.Context(), userID)
crons, err := h.service.QueueSources(userID)
if err != nil {
writeError(c, err)
return
@@ -281,7 +340,7 @@ func (h *Handler) queueSync(c *gin.Context) {
for _, cron := range crons {
result = append(result, cronDTO(cron))
}
c.JSON(http.StatusOK, result)
c.JSON(http.StatusAccepted, result)
}
func currentUser(c *gin.Context) (uint, bool) {
@@ -297,6 +356,10 @@ func writeError(c *gin.Context, err error) {
switch {
case errors.Is(err, gorm.ErrRecordNotFound):
httpx.Error(c, http.StatusNotFound, "not_found", "数据源或数据条目不存在")
case errors.Is(err, ErrBuiltInSource):
httpx.Error(c, http.StatusConflict, "built_in_source", "内置数据源不能删除,可以将其停用")
case errors.Is(err, ErrSyncInProgress):
httpx.Error(c, http.StatusConflict, "sync_in_progress", "已有数据源采集任务正在运行")
case errors.Is(err, ErrNameRequired), errors.Is(err, ErrKindInvalid),
errors.Is(err, ErrURLInvalid), errors.Is(err, ErrTitleRequired), errors.Is(err, ErrStatusInvalid):
writeInvalidRequest(c)
@@ -313,7 +376,7 @@ func sourceDTO(source models.SaDatasetSource, itemCount int64) SourceDTO {
return SourceDTO{
ID: source.Identity, Name: source.Name, Kind: source.Kind, URL: source.URL,
IconURL: source.IconURL, Description: source.Description, Enabled: source.Enabled,
LastSyncedAt: utcTime(source.LastSyncedAt), ItemCount: itemCount,
BuiltIn: source.SeedKey != nil, LastSyncedAt: utcTime(source.LastSyncedAt), ItemCount: itemCount,
CreatedAt: source.CreatedAt.UTC(), UpdatedAt: source.UpdatedAt.UTC(),
}
}
@@ -329,13 +392,23 @@ func itemDTO(item models.SaDatasetItem) ItemDTO {
func cronDTO(cron models.SaDatasetCron) CronDTO {
return CronDTO{
ID: cron.Identity, SourceID: cron.SourceIdentity, Schedule: cron.Schedule,
Status: cron.Status, Enabled: cron.Enabled, NextRunAt: utcTime(cron.NextRunAt),
ID: cron.Identity, SourceID: cron.SourceIdentity, Status: cron.Status,
LastRunAt: utcTime(cron.LastRunAt), LastResult: cron.LastResult,
CreatedAt: cron.CreatedAt.UTC(), UpdatedAt: cron.UpdatedAt.UTC(),
}
}
func optionalNonNegativeInt(value string) (int, error) {
if strings.TrimSpace(value) == "" {
return 0, nil
}
parsed, err := strconv.Atoi(value)
if err != nil || parsed < 0 {
return 0, errors.New("invalid non-negative integer")
}
return parsed, nil
}
func parseOptionalTime(value string) (*time.Time, error) {
if strings.TrimSpace(value) == "" {
return nil, nil

View File

@@ -8,6 +8,7 @@ import (
"net/http"
"net/http/httptest"
"testing"
"time"
"github.com/glebarez/sqlite"
"github.com/stretchr/testify/require"
@@ -21,12 +22,18 @@ func TestDatasetAPIConnectsSourcesItemsAndSyncTasks(t *testing.T) {
database := newDatasetTestDatabase(t)
owner := createDatasetTestUser(t, database, "owner@example.com")
other := createDatasetTestUser(t, database, "other@example.com")
project := models.SaProject{OwnerID: owner.ID, Name: "Research"}
require.NoError(t, database.Create(&project).Error)
ownerRouter := datasetTestRouter(database, owner.ID)
invalid := performDatasetRequest(t, ownerRouter, http.MethodPost, "/api/v1/dataset-sources", map[string]any{
"name": "Invalid RSS", "kind": "rss", "url": "",
})
require.Equal(t, http.StatusBadRequest, invalid.Code)
unsupported := performDatasetRequest(t, ownerRouter, http.MethodPost, "/api/v1/dataset-sources", map[string]any{
"name": "Manual source", "kind": "manual", "enabled": true,
})
require.Equal(t, http.StatusBadRequest, unsupported.Code)
createdSource := performDatasetRequest(t, ownerRouter, http.MethodPost, "/api/v1/dataset-sources", map[string]any{
"name": "Industry feed", "kind": "rss", "url": "https://example.com/feed",
@@ -53,6 +60,11 @@ func TestDatasetAPIConnectsSourcesItemsAndSyncTasks(t *testing.T) {
var item ItemDTO
require.NoError(t, json.Unmarshal(createdItem.Body.Bytes(), &item))
require.Equal(t, source.ID, item.SourceID)
legacyItems := performDatasetRequest(t, ownerRouter, http.MethodGet, "/api/v1/dataset-items", nil)
require.Equal(t, http.StatusOK, legacyItems.Code)
var legacyItemList []ItemDTO
require.NoError(t, json.Unmarshal(legacyItems.Body.Bytes(), &legacyItemList))
require.Len(t, legacyItemList, 1)
listSources := performDatasetRequest(t, ownerRouter, http.MethodGet, "/api/v1/dataset-sources", nil)
require.Equal(t, http.StatusOK, listSources.Code)
@@ -68,20 +80,42 @@ func TestDatasetAPIConnectsSourcesItemsAndSyncTasks(t *testing.T) {
require.NoError(t, json.Unmarshal(updatedItem.Body.Bytes(), &item))
require.Equal(t, "read", item.Status)
require.True(t, item.Starred)
partialItemUpdate := performDatasetRequest(t, ownerRouter, http.MethodPatch, "/api/v1/dataset-items/"+item.ID, map[string]any{
"starred": false,
})
require.Equal(t, http.StatusOK, partialItemUpdate.Code)
require.NoError(t, json.Unmarshal(partialItemUpdate.Body.Bytes(), &item))
require.Equal(t, "read", item.Status)
require.False(t, item.Starred)
deposited := performDatasetRequest(t, ownerRouter, http.MethodPost, "/api/v1/dataset-items/"+item.ID+"/deposit", map[string]any{
"projectId": project.Identity,
})
require.Equal(t, http.StatusCreated, deposited.Code)
var depositedNote models.SaNote
require.NoError(t, database.Where("project_id = ?", project.ID).First(&depositedNote).Error)
require.Equal(t, item.Title, depositedNote.Title)
require.Contains(t, depositedNote.Markdown, "原文链接")
firstSync := performDatasetRequest(t, ownerRouter, http.MethodPost, "/api/v1/dataset-crons/sync", map[string]any{})
require.Equal(t, http.StatusOK, firstSync.Code)
require.Equal(t, http.StatusAccepted, firstSync.Code)
var firstCrons []CronDTO
require.NoError(t, json.Unmarshal(firstSync.Body.Bytes(), &firstCrons))
require.Len(t, firstCrons, 1)
require.Equal(t, "completed", firstCrons[0].Status)
require.Contains(t, firstCrons[0].LastResult, "format=rss")
require.Equal(t, "pending", firstCrons[0].Status)
firstCompleted := waitForDatasetRun(t, ownerRouter, firstCrons[0].ID)
require.Equal(t, "completed", firstCompleted.Status)
require.Contains(t, firstCompleted.LastResult, "format=rss")
secondSync := performDatasetRequest(t, ownerRouter, http.MethodPost, "/api/v1/dataset-crons/sync", map[string]any{})
require.Equal(t, http.StatusOK, secondSync.Code)
var secondSync *httptest.ResponseRecorder
require.Eventually(t, func() bool {
secondSync = performDatasetRequest(t, ownerRouter, http.MethodPost, "/api/v1/dataset-crons/sync", map[string]any{})
return secondSync.Code == http.StatusAccepted
}, time.Second, 10*time.Millisecond)
var secondCrons []CronDTO
require.NoError(t, json.Unmarshal(secondSync.Body.Bytes(), &secondCrons))
require.NotEqual(t, firstCrons[0].ID, secondCrons[0].ID)
waitForDatasetRun(t, ownerRouter, secondCrons[0].ID)
var collectedItems []models.SaDatasetItem
require.NoError(t, database.Where("source_id = ?", 1).Find(&collectedItems).Error)
require.Len(t, collectedItems, 2)
@@ -121,6 +155,84 @@ func TestDatasetSourceAuthorizationUsesOwner(t *testing.T) {
creatorSources, err := NewService(database).ListSources(creator.ID)
require.NoError(t, err)
require.Empty(t, creatorSources)
item := models.SaDatasetItem{
SourceID: source.ID, CreatedBy: creator.ID, Title: "Creator-authored item", Status: "unread",
}
require.NoError(t, database.Create(&item).Error)
page, err := NewService(database).ListItemsPage(owner.ID, source.Identity, 50, 0)
require.NoError(t, err)
require.Len(t, page.Items, 1)
creatorPage, err := NewService(database).ListItemsPage(creator.ID, "", 50, 0)
require.NoError(t, err)
require.Empty(t, creatorPage.Items)
creatorProject := models.SaProject{OwnerID: creator.ID, Name: "Creator project"}
require.NoError(t, database.Create(&creatorProject).Error)
_, err = NewService(database).DepositItem(owner.ID, item.Identity, creatorProject.Identity)
require.ErrorIs(t, err, gorm.ErrRecordNotFound)
require.NoError(t, NewService(database).DeleteSource(owner.ID, source.Identity))
var itemCount int64
require.NoError(t, database.Model(&models.SaDatasetItem{}).Count(&itemCount).Error)
require.Zero(t, itemCount)
}
func TestDatasetItemsArePaginatedByOwnedSource(t *testing.T) {
database := newDatasetTestDatabase(t)
owner := createDatasetTestUser(t, database, "owner@example.com")
source := models.SaDatasetSource{
OwnerID: owner.ID, CreatedBy: owner.ID,
Name: "Feed", Kind: "rss", URL: "https://example.com/feed", Enabled: true,
}
require.NoError(t, database.Create(&source).Error)
for index := 0; index < 3; index++ {
require.NoError(t, database.Create(&models.SaDatasetItem{
SourceID: source.ID, CreatedBy: owner.ID,
Title: fmt.Sprintf("Item %d", index), Status: "unread",
}).Error)
}
router := datasetTestRouter(database, owner.ID)
response := performDatasetRequest(t, router, http.MethodGet,
"/api/v1/dataset-items?sourceId="+source.Identity+"&limit=2&offset=1", nil)
require.Equal(t, http.StatusOK, response.Code)
var page ItemPageDTO
require.NoError(t, json.Unmarshal(response.Body.Bytes(), &page))
require.Equal(t, int64(3), page.Total)
require.Equal(t, 2, page.Limit)
require.Equal(t, 1, page.Offset)
require.Len(t, page.Items, 2)
}
func TestBuiltInDatasetSourceKeepsManagedFieldsAndCannotBeDeleted(t *testing.T) {
database := newDatasetTestDatabase(t)
owner := createDatasetTestUser(t, database, "owner@example.com")
seedKey := "built-in"
source := models.SaDatasetSource{
OwnerID: owner.ID, CreatedBy: owner.ID, SeedKey: &seedKey,
Name: "Built-in", Kind: "rss", URL: "https://example.com/feed",
IconURL: "https://example.com/icon.png", Enabled: true,
}
require.NoError(t, database.Create(&source).Error)
service := NewService(database)
updated, err := service.UpdateSource(owner.ID, source.Identity, SourceInput{
Name: "Renamed", Kind: "rss", URL: "https://attacker.example/feed",
IconURL: "https://attacker.example/icon.png", Enabled: false,
})
require.NoError(t, err)
require.Equal(t, "Renamed", updated.Name)
require.Equal(t, "https://example.com/feed", updated.URL)
require.Equal(t, "https://example.com/icon.png", updated.IconURL)
require.False(t, updated.Enabled)
renamed := "Renamed again"
updated, err = service.PatchSource(owner.ID, source.Identity, SourcePatch{Name: &renamed})
require.NoError(t, err)
require.Equal(t, renamed, updated.Name)
require.False(t, updated.Enabled)
require.Equal(t, "https://example.com/feed", updated.URL)
require.ErrorIs(t, service.DeleteSource(owner.ID, source.Identity), ErrBuiltInSource)
}
func newDatasetTestDatabase(t *testing.T) *gorm.DB {
@@ -132,14 +244,39 @@ func newDatasetTestDatabase(t *testing.T) *gorm.DB {
&models.SaDatasetSource{},
&models.SaDatasetItem{},
&models.SaDatasetCron{},
&models.SaProject{},
&models.SaNote{},
))
require.NoError(t, database.Exec(`
CREATE UNIQUE INDEX idx_sa_dataset_item_source_external
CREATE UNIQUE INDEX IF NOT EXISTS idx_sa_dataset_item_source_external
ON sa_dataset_items (source_id, external_id)
`).Error)
return database
}
func waitForDatasetRun(t *testing.T, router http.Handler, identity string) CronDTO {
t.Helper()
var completed CronDTO
require.Eventually(t, func() bool {
response := performDatasetRequest(t, router, http.MethodGet, "/api/v1/dataset-crons", nil)
if response.Code != http.StatusOK {
return false
}
var runs []CronDTO
if json.Unmarshal(response.Body.Bytes(), &runs) != nil {
return false
}
for _, run := range runs {
if run.ID == identity && (run.Status == "completed" || run.Status == "failed") {
completed = run
return true
}
}
return false
}, time.Second, 10*time.Millisecond)
return completed
}
func createDatasetTestUser(t *testing.T, database *gorm.DB, email string) models.SaUser {
t.Helper()
user := models.SaUser{Email: email, DisplayName: email, PasswordHash: "hash", Role: "user"}

View File

@@ -15,11 +15,20 @@ import (
)
var (
ErrNameRequired = errors.New("dataset source name is required")
ErrKindInvalid = errors.New("dataset source kind is invalid")
ErrURLInvalid = errors.New("dataset source url is invalid")
ErrTitleRequired = errors.New("dataset item title is required")
ErrStatusInvalid = errors.New("dataset item status is invalid")
ErrNameRequired = errors.New("dataset source name is required")
ErrKindInvalid = errors.New("dataset source kind is invalid")
ErrURLInvalid = errors.New("dataset source url is invalid")
ErrTitleRequired = errors.New("dataset item title is required")
ErrStatusInvalid = errors.New("dataset item status is invalid")
ErrBuiltInSource = errors.New("built-in dataset source cannot be deleted")
ErrSyncInProgress = errors.New("dataset sync is already in progress")
)
const (
defaultItemPageSize = 50
maxItemPageSize = 200
datasetRunRetention = 30 * 24 * time.Hour
maxConcurrentFetches = 4
)
type Service struct {
@@ -37,6 +46,15 @@ type SourceInput struct {
Enabled bool
}
type SourcePatch struct {
Name *string
Kind *string
URL *string
IconURL *string
Description *string
Enabled *bool
}
type ItemInput struct {
SourceIdentity string
Title string
@@ -51,11 +69,33 @@ type ItemUpdate struct {
Starred bool
}
type ItemPatch struct {
Status *string
Starred *bool
}
type DepositResult struct {
NoteIdentity string
ProjectIdentity string
}
type SourceRecord struct {
Source models.SaDatasetSource
ItemCount int64
}
type ItemPage struct {
Items []models.SaDatasetItem
Total int64
Limit int
Offset int
}
type queuedSource struct {
Source models.SaDatasetSource
Run models.SaDatasetCron
}
func NewService(database *gorm.DB) *Service {
return &Service{db: database, fetcher: NewHTTPFeedFetcher()}
}
@@ -81,7 +121,7 @@ func (s *Service) ListSources(userID uint) ([]SourceRecord, error) {
}
if err := s.db.Model(&models.SaDatasetItem{}).
Select("source_id, COUNT(*) AS count").
Where("created_by = ? AND source_id IN ?", userID, sourceIDs).
Where("source_id IN ?", sourceIDs).
Group("source_id").
Scan(&rows).Error; err != nil {
return nil, err
@@ -111,14 +151,60 @@ func (s *Service) CreateSource(userID uint, input SourceInput) (*models.SaDatase
}
func (s *Service) UpdateSource(userID uint, identity string, input SourceInput) (*models.SaDatasetSource, error) {
normalized, err := normalizeSourceInput(input)
if err != nil {
return nil, err
return s.PatchSource(userID, identity, SourcePatch{
Name: &input.Name, Kind: &input.Kind, URL: &input.URL, IconURL: &input.IconURL,
Description: &input.Description, Enabled: &input.Enabled,
})
}
func (s *Service) PatchSource(userID uint, identity string, patch SourcePatch) (*models.SaDatasetSource, error) {
if !s.syncMu.TryLock() {
return nil, ErrSyncInProgress
}
defer s.syncMu.Unlock()
var source models.SaDatasetSource
if err := s.db.Where("identity = ? AND owner_id = ?", identity, userID).First(&source).Error; err != nil {
return nil, err
}
input := SourceInput{
Name: source.Name, Kind: source.Kind, URL: source.URL, IconURL: source.IconURL,
Description: source.Description, Enabled: source.Enabled,
}
if patch.Name != nil {
input.Name = *patch.Name
}
if patch.Kind != nil {
input.Kind = *patch.Kind
}
if patch.URL != nil {
input.URL = *patch.URL
}
if patch.IconURL != nil {
input.IconURL = *patch.IconURL
}
if patch.Description != nil {
input.Description = *patch.Description
}
if patch.Enabled != nil {
input.Enabled = *patch.Enabled
}
managedFields := source.SeedKey != nil || source.Kind != "rss"
if managedFields {
input.Kind = source.Kind
input.URL = source.URL
input.IconURL = source.IconURL
}
var normalized SourceInput
var err error
if source.Kind == "rss" {
normalized, err = normalizeSourceInput(input)
} else {
normalized, err = normalizeLegacySourceInput(input)
}
if err != nil {
return nil, err
}
if err := s.db.Model(&source).Updates(map[string]any{
"name": normalized.Name, "kind": normalized.Kind, "url": normalized.URL,
"icon_url": normalized.IconURL, "description": normalized.Description, "enabled": normalized.Enabled,
@@ -129,15 +215,23 @@ func (s *Service) UpdateSource(userID uint, identity string, input SourceInput)
}
func (s *Service) DeleteSource(userID uint, identity string) error {
if !s.syncMu.TryLock() {
return ErrSyncInProgress
}
defer s.syncMu.Unlock()
return s.db.Transaction(func(tx *gorm.DB) error {
var source models.SaDatasetSource
if err := tx.Where("identity = ? AND owner_id = ?", identity, userID).First(&source).Error; err != nil {
return err
}
if err := tx.Where("source_id = ? AND created_by = ?", source.ID, userID).Delete(&models.SaDatasetCron{}).Error; err != nil {
if source.SeedKey != nil {
return ErrBuiltInSource
}
if err := tx.Where("source_id = ?", source.ID).Delete(&models.SaDatasetCron{}).Error; err != nil {
return err
}
if err := tx.Where("source_id = ? AND created_by = ?", source.ID, userID).Delete(&models.SaDatasetItem{}).Error; err != nil {
if err := tx.Where("source_id = ?", source.ID).Delete(&models.SaDatasetItem{}).Error; err != nil {
return err
}
return tx.Delete(&source).Error
@@ -145,28 +239,46 @@ func (s *Service) DeleteSource(userID uint, identity string) error {
}
func (s *Service) ListItems(userID uint, sourceIdentity string) ([]models.SaDatasetItem, error) {
query := s.db.Where("created_by = ?", userID)
page, err := s.ListItemsPage(userID, sourceIdentity, maxItemPageSize, 0)
if err != nil {
return nil, err
}
return page.Items, nil
}
func (s *Service) ListItemsPage(userID uint, sourceIdentity string, limit, offset int) (ItemPage, error) {
limit, offset = normalizePage(limit, offset)
query := s.db.Model(&models.SaDatasetItem{}).
Joins("JOIN sa_dataset_sources ON sa_dataset_sources.id = sa_dataset_items.source_id").
Where("sa_dataset_sources.owner_id = ?", userID)
if sourceIdentity = strings.TrimSpace(sourceIdentity); sourceIdentity != "" {
source, err := s.findOwnedSource(userID, sourceIdentity)
if err != nil {
return nil, err
return ItemPage{}, err
}
query = query.Where("source_id = ?", source.ID)
query = query.Where("sa_dataset_items.source_id = ?", source.ID)
}
var total int64
if err := query.Session(&gorm.Session{}).Count(&total).Error; err != nil {
return ItemPage{}, err
}
var items []models.SaDatasetItem
if err := query.Order("COALESCE(published_at, created_at) desc, id desc").Limit(200).Find(&items).Error; err != nil {
return nil, err
if err := query.Select("sa_dataset_items.*").
Order("COALESCE(sa_dataset_items.published_at, sa_dataset_items.created_at) desc, sa_dataset_items.id desc").
Limit(limit).Offset(offset).Find(&items).Error; err != nil {
return ItemPage{}, err
}
if items == nil {
items = []models.SaDatasetItem{}
}
return items, nil
return ItemPage{Items: items, Total: total, Limit: limit, Offset: offset}, nil
}
func (s *Service) CountItems(userID uint, sourceID uint) (int64, error) {
var count int64
err := s.db.Model(&models.SaDatasetItem{}).
Where("created_by = ? AND source_id = ?", userID, sourceID).
Joins("JOIN sa_dataset_sources ON sa_dataset_sources.id = sa_dataset_items.source_id").
Where("sa_dataset_sources.owner_id = ? AND sa_dataset_items.source_id = ?", userID, sourceID).
Count(&count).Error
return count, err
}
@@ -193,23 +305,72 @@ func (s *Service) CreateItem(userID uint, input ItemInput) (*models.SaDatasetIte
}
func (s *Service) UpdateItem(userID uint, identity string, input ItemUpdate) (*models.SaDatasetItem, error) {
status := strings.TrimSpace(input.Status)
if status != "unread" && status != "read" && status != "archived" {
return nil, ErrStatusInvalid
}
var item models.SaDatasetItem
if err := s.db.Where("identity = ? AND created_by = ?", identity, userID).First(&item).Error; err != nil {
return s.PatchItem(userID, identity, ItemPatch{Status: &input.Status, Starred: &input.Starred})
}
func (s *Service) PatchItem(userID uint, identity string, patch ItemPatch) (*models.SaDatasetItem, error) {
item, err := s.findOwnedItem(userID, identity)
if err != nil {
return nil, err
}
if err := s.db.Model(&item).Updates(map[string]any{"status": status, "starred": input.Starred}).Error; err != nil {
updates := make(map[string]any, 2)
if patch.Status != nil {
status := strings.TrimSpace(*patch.Status)
if status != "unread" && status != "read" && status != "archived" {
return nil, ErrStatusInvalid
}
updates["status"] = status
}
if patch.Starred != nil {
updates["starred"] = *patch.Starred
}
if len(updates) == 0 {
return item, nil
}
if err := s.db.Model(item).Updates(updates).Error; err != nil {
return nil, err
}
return &item, nil
return item, nil
}
func (s *Service) DepositItem(userID uint, itemIdentity, projectIdentity string) (DepositResult, error) {
item, err := s.findOwnedItem(userID, itemIdentity)
if err != nil {
return DepositResult{}, err
}
var project models.SaProject
if err := s.db.Where("identity = ? AND owner_id = ?", strings.TrimSpace(projectIdentity), userID).
First(&project).Error; err != nil {
return DepositResult{}, err
}
body := strings.TrimSpace(item.Content)
if body == "" {
body = strings.TrimSpace(item.Summary)
}
if item.URL != "" {
if body != "" {
body += "\n\n"
}
body += "[原文链接](" + item.URL + ")"
}
note := models.SaNote{
ProjectID: project.ID, CreatedBy: userID,
Title: item.Title, Markdown: body,
}
if err := s.db.Create(&note).Error; err != nil {
return DepositResult{}, err
}
return DepositResult{NoteIdentity: note.Identity, ProjectIdentity: project.Identity}, nil
}
func (s *Service) ListCrons(userID uint) ([]models.SaDatasetCron, error) {
var crons []models.SaDatasetCron
if err := s.db.Where("created_by = ?", userID).Order("created_at desc, id desc").Limit(100).Find(&crons).Error; err != nil {
if err := s.db.Model(&models.SaDatasetCron{}).
Select("sa_dataset_crons.*").
Joins("JOIN sa_dataset_sources ON sa_dataset_sources.id = sa_dataset_crons.source_id").
Where("sa_dataset_sources.owner_id = ?", userID).
Order("sa_dataset_crons.created_at desc, sa_dataset_crons.id desc").
Limit(100).Find(&crons).Error; err != nil {
return nil, err
}
if crons == nil {
@@ -222,7 +383,36 @@ func (s *Service) SyncSources(ctx context.Context, userID uint) ([]models.SaData
s.syncMu.Lock()
defer s.syncMu.Unlock()
return s.syncSources(ctx, userID)
return s.syncSources(ctx, userID, "scheduled")
}
func (s *Service) QueueSources(userID uint) ([]models.SaDatasetCron, error) {
if !s.syncMu.TryLock() {
return nil, ErrSyncInProgress
}
sources, err := s.enabledRSSSources(userID)
if err != nil {
s.syncMu.Unlock()
return nil, err
}
queued, err := s.createDatasetRuns(userID, sources, "manual")
if err != nil {
s.syncMu.Unlock()
return nil, err
}
runs := make([]models.SaDatasetCron, 0, len(queued))
for _, entry := range queued {
runs = append(runs, entry.Run)
}
if len(queued) == 0 {
s.syncMu.Unlock()
return runs, nil
}
go func() {
defer s.syncMu.Unlock()
_, _ = s.executeDatasetRuns(context.Background(), queued)
}()
return runs, nil
}
func (s *Service) SyncAllSources(ctx context.Context) ([]models.SaDatasetCron, error) {
@@ -250,103 +440,166 @@ func (s *Service) SyncAllSources(ctx context.Context) ([]models.SaDatasetCron, e
return result, errors.Join(syncErrors...)
}
func (s *Service) syncSources(ctx context.Context, userID uint) ([]models.SaDatasetCron, error) {
result := make([]models.SaDatasetCron, 0)
func (s *Service) syncSources(ctx context.Context, userID uint, trigger string) ([]models.SaDatasetCron, error) {
sources, err := s.enabledRSSSources(userID)
if err != nil {
return nil, err
}
queued, err := s.createDatasetRuns(userID, sources, trigger)
if err != nil {
return nil, err
}
return s.executeDatasetRuns(ctx, queued)
}
func (s *Service) enabledRSSSources(userID uint) ([]models.SaDatasetSource, error) {
var sources []models.SaDatasetSource
if err := s.db.Where("owner_id = ? AND enabled = ? AND kind = ?", userID, true, "rss").
Order("id asc").Find(&sources).Error; err != nil {
return nil, err
}
for _, source := range sources {
if err := ctx.Err(); err != nil {
return result, err
}
now := time.Now().UTC()
cron := models.SaDatasetCron{
SourceID: source.ID, CreatedBy: userID, Schedule: "@once",
Status: "running", Enabled: true, NextRunAt: &now, LastRunAt: &now,
}
if err := s.db.Create(&cron).Error; err != nil {
return nil, err
}
return sources, nil
}
feed, fetchErr := s.fetcher.Fetch(ctx, source.URL)
if fetchErr != nil {
cron.Status = "failed"
cron.Enabled = false
cron.NextRunAt = nil
cron.LastResult = truncateResult(fetchErr.Error())
if err := s.db.Model(&cron).Updates(map[string]any{
"status": cron.Status, "last_result": cron.LastResult, "enabled": false, "next_run_at": nil,
}).Error; err != nil {
return nil, err
}
result = append(result, cron)
if err := ctx.Err(); err != nil {
return result, err
}
continue
func (s *Service) createDatasetRuns(userID uint, sources []models.SaDatasetSource, trigger string) ([]queuedSource, error) {
queued := make([]queuedSource, 0, len(sources))
now := time.Now().UTC()
err := s.db.Transaction(func(tx *gorm.DB) error {
if err := tx.Where("created_at < ?", now.Add(-datasetRunRetention)).
Delete(&models.SaDatasetCron{}).Error; err != nil {
return err
}
for _, source := range sources {
run := models.SaDatasetCron{
SourceID: source.ID, CreatedBy: userID, Schedule: trigger,
Status: "pending", Enabled: true, NextRunAt: &now,
}
if err := tx.Create(&run).Error; err != nil {
return err
}
queued = append(queued, queuedSource{Source: source, Run: run})
}
return nil
})
return queued, err
}
inserted, err := s.storeFeedItems(userID, source, feed.Items)
if err != nil {
cron.Status = "failed"
cron.Enabled = false
cron.NextRunAt = nil
cron.LastResult = truncateResult("store feed items: " + err.Error())
if updateErr := s.db.Model(&cron).Updates(map[string]any{
"status": cron.Status, "last_result": cron.LastResult, "enabled": false, "next_run_at": nil,
}).Error; updateErr != nil {
return nil, updateErr
}
result = append(result, cron)
continue
func (s *Service) executeDatasetRuns(ctx context.Context, queued []queuedSource) ([]models.SaDatasetCron, error) {
result := make([]models.SaDatasetCron, len(queued))
runErrors := make([]error, len(queued))
concurrency := maxConcurrentFetches
if s.db.Dialector.Name() == "sqlite" {
concurrency = 1
}
semaphore := make(chan struct{}, concurrency)
var workers sync.WaitGroup
for index, entry := range queued {
workers.Add(1)
go func() {
defer workers.Done()
semaphore <- struct{}{}
defer func() { <-semaphore }()
result[index], runErrors[index] = s.executeDatasetRun(ctx, entry)
}()
}
workers.Wait()
return result, errors.Join(runErrors...)
}
func (s *Service) executeDatasetRun(ctx context.Context, entry queuedSource) (models.SaDatasetCron, error) {
source := entry.Source
cron := entry.Run
if err := ctx.Err(); err != nil {
return cron, err
}
now := time.Now().UTC()
cron.Status = "running"
cron.LastRunAt = &now
if err := s.db.Model(&cron).Updates(map[string]any{
"status": "running", "last_run_at": now,
}).Error; err != nil {
return cron, err
}
feed, fetchErr := s.fetcher.Fetch(ctx, source.URL)
if fetchErr != nil {
cron.Status = "failed"
cron.Enabled = false
cron.NextRunAt = nil
cron.LastResult = truncateResult(fetchErr.Error())
if err := s.db.Model(&cron).Updates(map[string]any{
"status": cron.Status, "last_result": cron.LastResult, "enabled": false, "next_run_at": nil,
}).Error; err != nil {
return cron, err
}
return cron, ctx.Err()
}
inserted := 0
completedAt := time.Now().UTC()
err := s.db.Transaction(func(tx *gorm.DB) error {
var storeErr error
inserted, storeErr = storeFeedItems(tx, cron.CreatedBy, source, feed.Items)
if storeErr != nil {
return storeErr
}
completedAt := time.Now().UTC()
cron.Status = "completed"
cron.Enabled = false
cron.NextRunAt = nil
cron.LastResult = fmt.Sprintf("format=%s fetched=%d inserted=%d", feed.Format, len(feed.Items), inserted)
if err := s.db.Transaction(func(tx *gorm.DB) error {
if err := tx.Model(&cron).Updates(map[string]any{
"status": cron.Status, "last_result": cron.LastResult, "enabled": false, "next_run_at": nil,
}).Error; err != nil {
return err
}
return tx.Model(&source).Update("last_synced_at", completedAt).Error
}); err != nil {
return nil, err
if err := tx.Model(&cron).Updates(map[string]any{
"status": cron.Status, "last_result": cron.LastResult, "enabled": false, "next_run_at": nil,
}).Error; err != nil {
return err
}
source.LastSyncedAt = &completedAt
result = append(result, cron)
return tx.Model(&source).Update("last_synced_at", completedAt).Error
})
if err == nil {
return cron, nil
}
return result, nil
cron.Status = "failed"
cron.Enabled = false
cron.NextRunAt = nil
cron.LastResult = truncateResult("store feed items: " + err.Error())
if updateErr := s.db.Model(&cron).Updates(map[string]any{
"status": cron.Status, "last_result": cron.LastResult, "enabled": false, "next_run_at": nil,
}).Error; updateErr != nil {
return cron, updateErr
}
return cron, nil
}
func (s *Service) storeFeedItems(userID uint, source models.SaDatasetSource, items []FeedItem) (int, error) {
inserted := 0
err := s.db.Transaction(func(tx *gorm.DB) error {
for _, input := range items {
externalID := input.ExternalID
item := models.SaDatasetItem{
SourceID: source.ID, CreatedBy: userID, ExternalID: &externalID,
Title: input.Title, Summary: input.Summary, Content: input.Content,
URL: input.URL, Status: "unread", PublishedAt: input.PublishedAt,
}
result := tx.Clauses(clause.OnConflict{
Columns: []clause.Column{{Name: "source_id"}, {Name: "external_id"}},
DoNothing: true,
}).Create(&item)
if result.Error != nil {
return result.Error
}
inserted += int(result.RowsAffected)
}
return nil
var err error
inserted, err = storeFeedItems(tx, userID, source, items)
return err
})
return inserted, err
}
func storeFeedItems(tx *gorm.DB, userID uint, source models.SaDatasetSource, items []FeedItem) (int, error) {
inserted := 0
for _, input := range items {
externalID := input.ExternalID
item := models.SaDatasetItem{
SourceID: source.ID, CreatedBy: userID, ExternalID: &externalID,
Title: input.Title, Summary: input.Summary, Content: input.Content,
URL: input.URL, Status: "unread", PublishedAt: input.PublishedAt,
}
result := tx.Clauses(clause.OnConflict{
Columns: []clause.Column{{Name: "source_id"}, {Name: "external_id"}},
DoNothing: true,
}).Create(&item)
if result.Error != nil {
return inserted, result.Error
}
inserted += int(result.RowsAffected)
}
return inserted, nil
}
func truncateResult(value string) string {
return truncateRunes(value, 2000)
}
@@ -365,6 +618,16 @@ func (s *Service) findOwnedSource(userID uint, identity string) (*models.SaDatas
return &source, err
}
func (s *Service) findOwnedItem(userID uint, identity string) (*models.SaDatasetItem, error) {
var item models.SaDatasetItem
err := s.db.Model(&models.SaDatasetItem{}).
Select("sa_dataset_items.*").
Joins("JOIN sa_dataset_sources ON sa_dataset_sources.id = sa_dataset_items.source_id").
Where("sa_dataset_items.identity = ? AND sa_dataset_sources.owner_id = ?", strings.TrimSpace(identity), userID).
First(&item).Error
return &item, err
}
func normalizeSourceInput(input SourceInput) (SourceInput, error) {
input.Name = strings.TrimSpace(input.Name)
input.Kind = strings.ToLower(strings.TrimSpace(input.Kind))
@@ -374,7 +637,7 @@ func normalizeSourceInput(input SourceInput) (SourceInput, error) {
if input.Name == "" {
return input, ErrNameRequired
}
if input.Kind != "manual" && input.Kind != "link" && input.Kind != "rss" {
if input.Kind != "rss" {
return input, ErrKindInvalid
}
if input.URL != "" && !validHTTPURL(input.URL) {
@@ -383,12 +646,34 @@ func normalizeSourceInput(input SourceInput) (SourceInput, error) {
if input.IconURL != "" && !validHTTPURL(input.IconURL) {
return input, ErrURLInvalid
}
if input.Kind != "manual" && input.URL == "" {
if input.URL == "" {
return input, ErrURLInvalid
}
return input, nil
}
func normalizeLegacySourceInput(input SourceInput) (SourceInput, error) {
input.Name = strings.TrimSpace(input.Name)
input.Description = strings.TrimSpace(input.Description)
if input.Name == "" {
return input, ErrNameRequired
}
return input, nil
}
func normalizePage(limit, offset int) (int, int) {
if limit <= 0 {
limit = defaultItemPageSize
}
if limit > maxItemPageSize {
limit = maxItemPageSize
}
if offset < 0 {
offset = 0
}
return limit, offset
}
func validHTTPURL(value string) bool {
parsed, err := url.ParseRequestURI(value)
return err == nil && (parsed.Scheme == "http" || parsed.Scheme == "https") &&

View File

@@ -14,8 +14,8 @@ type SaDatasetCron struct {
Enabled bool `gorm:"not null;default:true;index"`
NextRunAt *time.Time
LastRunAt *time.Time
LastResult string `gorm:"type:text"`
CreatedAt time.Time
LastResult string `gorm:"type:text"`
CreatedAt time.Time `gorm:"index"`
UpdatedAt time.Time
}

View File

@@ -2,6 +2,7 @@ package models
import (
"database/sql"
"errors"
"github.com/google/uuid"
"gorm.io/gorm"
@@ -28,6 +29,9 @@ func (m *SaDatasetSource) BeforeCreate(tx *gorm.DB) error {
if m.CreatedBy == 0 {
m.CreatedBy = m.OwnerID
}
if m.OwnerID == 0 {
return errors.New("dataset source owner is required")
}
if err := resolveIdentity(tx, &SaUser{}, m.OwnerID, &m.OwnerIdentity); err != nil {
return err
}