feat(knowledge): 知识源维护与 TXT/Markdown/PDF 导入

支持文本、Markdown 与非扫描 PDF 上传解析，展示提取文本与切块结果。

frontend/src/__tests__/components/knowledge/KnowledgeAddSourceModal.scope.spec.ts

@@ -0,0 +1,71 @@
+/*
+Copyright 2024-2026 ReplyPier
+
+Licensed under the Apache License, Version 2.0 (the "License");
+you may not use this file except in compliance with the License.
+You may obtain a copy of the License at
+
+    http://www.apache.org/licenses/LICENSE-2.0
+
+Unless required by applicable law or agreed to in writing, software
+distributed under the License is distributed on an "AS IS" BASIS,
+WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+See the License for the specific language governing permissions and
+limitations under the License.
+*/
+
+import { describe, it, expect } from 'vitest'
+import { mount } from '@vue/test-utils'
+import KnowledgeAddSourceModal from '../../../components/knowledge/KnowledgeAddSourceModal.vue'
+
+const mountModal = async (url: string) => {
+  const wrapper = mount(KnowledgeAddSourceModal)
+  await wrapper.find('#kb-add-url').setValue(url)
+  return wrapper
+}
+
+const scopeTitles = (wrapper: ReturnType<typeof mount>) =>
+  wrapper.findAll('.scope__title').map((node) => node.text())
+
+const scopeOption = (wrapper: ReturnType<typeof mount>, title: string) =>
+  wrapper.findAll('.scope__opt').find((node) => node.find('.scope__title').text() === title)
+
+describe('KnowledgeAddSourceModal crawl scope', () => {
+  it('offers no section option for a homepage URL', async () => {
+    // "Pages under / " is not a section — it is the whole site, which the next
+    // option already covers.
+    const wrapper = await mountModal('https://paywithatoa.co.uk')
+    expect(scopeTitles(wrapper)).toEqual(['仅此页面', '整个站点', '整个域名'])
+    expect(scopeOption(wrapper, '整个站点')!.text()).toContain('paywithatoa.co.uk 上的所有页面')
+  })
+
+  it('names the section when the URL has a path', async () => {
+    const wrapper = await mountModal('https://help.example.com/hc/en')
+    expect(scopeTitles(wrapper)).toContain('本栏目')
+    expect(scopeOption(wrapper, '本栏目')!.text()).toContain(
+      '/hc/en/ 下的页面',
+    )
+  })
+
+  it('falls back to this-site when the section option disappears', async () => {
+    const wrapper = await mountModal('https://help.example.com/hc/en')
+    await scopeOption(wrapper, '本栏目')!.trigger('click')
+
+    await wrapper.find('#kb-add-url').setValue('https://help.example.com')
+    await wrapper.find('button.btn--primary').trigger('click')
+
+    expect(wrapper.emitted('submit')![0][0]).toMatchObject({
+      type: 'website',
+      url: 'https://help.example.com',
+      scope: 'host',
+    })
+  })
+
+  it('submits the selected scope', async () => {
+    const wrapper = await mountModal('https://help.example.com/hc/en')
+    await scopeOption(wrapper, '本栏目')!.trigger('click')
+    await wrapper.find('button.btn--primary').trigger('click')
+
+    expect(wrapper.emitted('submit')![0][0]).toMatchObject({ scope: 'path' })
+  })
+})

frontend/src/__tests__/composables/useKnowledgeManagement.url.spec.ts

@@ -0,0 +1,116 @@
+import { describe, it, expect, vi, beforeEach } from 'vitest'
+import { setActivePinia, createPinia } from 'pinia'
+
+vi.mock('@/services/knowledge', () => ({
+  knowledgeService: {
+    getKnowledgeByAgent: vi.fn().mockResolvedValue({ knowledge: [], pagination: { total_pages: 0 } }),
+    getKnowledgeByOrganization: vi
+      .fn()
+      .mockResolvedValue({ knowledge: [], pagination: { total_pages: 0 } }),
+    getAgentQueueItems: vi.fn().mockResolvedValue({ queue_items: [] })
+  }
+}))
+
+import { useKnowledgeManagement } from '@/composables/useKnowledgeManagement'
+
+const setup = () => useKnowledgeManagement('agent-1', 'org-1')
+
+describe('useKnowledgeManagement - URL handling', () => {
+  beforeEach(() => {
+    setActivePinia(createPinia())
+    vi.clearAllMocks()
+  })
+
+  describe('withScheme', () => {
+    const { withScheme } = setup()
+
+    it.each([
+      ['docs.company.com', 'https://docs.company.com'],
+      ['company.co.uk', 'https://company.co.uk'],
+      ['  spaced.com  ', 'https://spaced.com'],
+      ['//protocol-relative.com', 'https://protocol-relative.com']
+    ])('defaults the scheme: %s -> %s', (input, expected) => {
+      expect(withScheme(input)).toBe(expected)
+    })
+
+    it.each([
+      'https://already.com',
+      'http://plain.com',
+      'HTTPS://shouty.com'
+    ])('leaves an existing scheme alone: %s', (input) => {
+      expect(withScheme(input)).toBe(input)
+    })
+
+    it('leaves an empty value empty', () => {
+      expect(withScheme('   ')).toBe('')
+    })
+  })
+
+  describe('isValidUrl', () => {
+    const { isValidUrl } = setup()
+
+    it.each(['https://docs.company.com', 'http://a.co', 'https://x.com/path?q=1'])(
+      'accepts %s',
+      (url) => expect(isValidUrl(url)).toBe(true)
+    )
+
+    it.each(['mailto:someone@x.com', 'foo:bar', 'not a url', 'https://'])(
+      'rejects %s',
+      (url) => expect(isValidUrl(url)).toBe(false)
+    )
+
+    it.each(['https://localhost', 'https://wiki', 'http://intranet:8080'])(
+      'accepts the intranet hostnames a self-hosted install indexes: %s',
+      (url) => expect(isValidUrl(url)).toBe(true)
+    )
+  })
+
+  describe('handleUrlAdd', () => {
+    it('stages a scheme-less URL instead of rejecting it', () => {
+      const { newUrl, urls, urlFormError, handleUrlAdd } = setup()
+      newUrl.value = 'docs.company.com'
+
+      handleUrlAdd()
+
+      expect(urls.value).toEqual(['https://docs.company.com'])
+      expect(urlFormError.value).toBeNull()
+      expect(newUrl.value).toBe('')
+    })
+
+    it('still rejects input that is not a URL at all', () => {
+      const { newUrl, urls, urlFormError, handleUrlAdd } = setup()
+      newUrl.value = 'not a url'
+
+      handleUrlAdd()
+
+      expect(urls.value).toEqual([])
+      expect(urlFormError.value).toBe('请输入有效的 URL')
+    })
+
+    it('clears a previous error when a later attempt succeeds', () => {
+      const { newUrl, urls, urlFormError, handleUrlAdd } = setup()
+
+      newUrl.value = 'not a url'
+      handleUrlAdd()
+      expect(urlFormError.value).toBe('请输入有效的 URL')
+
+      newUrl.value = 'docs.company.com'
+      handleUrlAdd()
+
+      expect(urlFormError.value).toBeNull()
+      expect(urls.value).toEqual(['https://docs.company.com'])
+    })
+
+    it('does not stage the same URL twice', () => {
+      const { newUrl, urls, urlFormError, handleUrlAdd } = setup()
+
+      newUrl.value = 'docs.company.com'
+      handleUrlAdd()
+      newUrl.value = 'https://docs.company.com'
+      handleUrlAdd()
+
+      expect(urls.value).toEqual(['https://docs.company.com'])
+      expect(urlFormError.value).toBe('该 URL 已添加到当前批次')
+    })
+  })
+})

frontend/src/__tests__/composables/useStagedSourceCap.spec.ts

@@ -0,0 +1,55 @@
+import { describe, it, expect, vi, beforeEach } from 'vitest'
+import { ref } from 'vue'
+
+const planLimits = vi.fn()
+vi.mock('@/utils/storage', () => ({
+  subscriptionStorage: { getPlanLimits: () => planLimits() }
+}))
+
+import { useStagedSourceCap } from '@/composables/useStagedSourceCap'
+
+const file = (name: string) => new File(['x'], name, { type: 'application/pdf' })
+
+const setup = (cap: number, indexed: number, urls: string[], files: File[]) => {
+  planLimits.mockReturnValue({ maxKnowledgeSources: cap })
+  const u = ref(urls)
+  const f = ref(files)
+  return { u, f, ...useStagedSourceCap(ref(indexed), u, f) }
+}
+
+describe('useStagedSourceCap', () => {
+  beforeEach(() => planLimits.mockReset())
+
+  it('does nothing when the plan sets no cap', () => {
+    const { u, f, fitToCap } = setup(0, 3, ['a', 'b'], [file('x.pdf')])
+    expect(fitToCap('urls')).toBeNull()
+    expect(u.value).toEqual(['a', 'b'])
+    expect(f.value).toHaveLength(1)
+  })
+
+  it('drops staged files before older URLs, oldest first', () => {
+    const { u, f, fitToCap } = setup(2, 0, ['a', 'b'], [file('x.pdf')])
+    expect(fitToCap('urls')).toBeNull()
+    expect(f.value).toEqual([])
+    expect(u.value).toEqual(['a', 'b'])
+  })
+
+  it('drops the oldest URL when only URLs are staged', () => {
+    const { u, fitToCap } = setup(2, 0, ['a', 'b', 'c'], [])
+    expect(fitToCap('urls')).toBeNull()
+    expect(u.value).toEqual(['b', 'c'])
+  })
+
+  it('unstages new files that cannot fit and says how many can', () => {
+    const { u, f, fitToCap } = setup(3, 1, ['a'], [file('x.pdf'), file('y.pdf'), file('z.pdf')])
+    expect(fitToCap('files', 3)).toBe('你的套餐包含 3 个知识库来源，最多可选 2 个。')
+    expect(f.value).toEqual([])
+    expect(u.value).toEqual(['a'])
+  })
+
+  it('reports a cap already filled by indexed sources', () => {
+    const { u, fitToCap } = setup(1, 1, ['a'], [])
+    expect(fitToCap('urls')).toBe('你的套餐包含 1 个知识库来源，且已在用。')
+    expect(u.value).toEqual([])
+  })
+})