refactor: 拆分 gateway 单体为独立 crate，新增 systemd 部署方案

将 gateway 内部的 model-fetch、provider-transport、scheduler-core、 usage-runtime、video-tasks-core 模块提取为独立 crate；重构 gateway 内部模块结构（state/router/cache/data/query 等）；移除大量遗留模块文件；新增 systemd 二进制部署骨架及相关文档；更新前端 usage 相关 API 和组件。
2026-09-02 09:20:22 +08:00 · 2026-04-05 20:23:16 +08:00
parent cbc811f6ce
commit 763ff03a7b
777 changed files with 42659 additions and 21469 deletions
--- a/frontend/src/features/providers/components/PriorityManagementDialog.vue
+++ b/frontend/src/features/providers/components/PriorityManagementDialog.vue
@@ -538,6 +538,7 @@ const loadingKeys = ref(false)
 const saving = ref(false)

 const SAVE_CONCURRENCY = 6
+const PRIORITY_REQUEST_TIMEOUT_MS = 5 * 60 * 1000

 let originalProviderPriorityById = new Map<string, number>()
 let originalPoolPriorityByProviderId = new Map<string, number | null>()
@@ -761,6 +762,11 @@ function snapshotKeyBaseline() {
  )
 }

+function snapshotCurrentPriorityBaseline() {
+  snapshotProviderBaseline(sortedProviders.value)
+  snapshotKeyBaseline()
+}
+
 function arePriorityMapsEqual(
  left: Record<string, number> | undefined,
  right: Record<string, number> | undefined,
@@ -1035,7 +1041,9 @@ async function loadKeysByFormat() {
  try {
    loadingKeys.value = true
    const { default: client } = await import('@/api/client')
-    const response = await client.get('/api/admin/endpoints/keys/grouped-by-format')
+    const response = await client.get('/api/admin/endpoints/keys/grouped-by-format', {
+      timeout: PRIORITY_REQUEST_TIMEOUT_MS,
+    })

    // 每个格式独立管理优先级，额外做一次前端归一化兜底，避免历史数据导致重复格式/脏键
    const data: Record<string, KeyWithMeta[]> = {}
@@ -1475,7 +1483,11 @@ async function save() {
      }

      if (Object.keys(payload).length > 0) {
-        providerTasks.push(() => updateProvider(provider.id, payload))
+        providerTasks.push(() => updateProvider(
+          provider.id,
+          payload,
+          { timeout: PRIORITY_REQUEST_TIMEOUT_MS },
+        ))
      }
    })

@@ -1487,7 +1499,11 @@ async function save() {
        priorityByFormat,
      ))
      .map(([keyId, priorityByFormat]) => () =>
-        updateProviderKey(keyId, { global_priority_by_format: priorityByFormat })
+        updateProviderKey(
+          keyId,
+          { global_priority_by_format: priorityByFormat },
+          { timeout: PRIORITY_REQUEST_TIMEOUT_MS },
+        )
      )

    await runTasksWithConcurrency([...providerTasks, ...keyTasks])
@@ -1497,16 +1513,16 @@ async function save() {
    await adminApi.updateSystemConfig(
      'provider_priority_mode',
      newMode,
-      'Provider/Key 优先级策略：provider(提供商优先模式) 或 global_key(全局Key优先模式)'
+      'Provider/Key 优先级策略：provider(提供商优先模式) 或 global_key(全局Key优先模式)',
+      { timeout: PRIORITY_REQUEST_TIMEOUT_MS },
    )
    await adminApi.updateSystemConfig(
      'scheduling_mode',
      schedulingMode.value,
-      '调度模式：cache_affinity(缓存亲和模式) 或 load_balance(负载均衡模式) 或 fixed_order(固定顺序模式)'
+      '调度模式：cache_affinity(缓存亲和模式) 或 load_balance(负载均衡模式) 或 fixed_order(固定顺序模式)',
+      { timeout: PRIORITY_REQUEST_TIMEOUT_MS },
    )
-
-    await loadAllProviders()
-    await loadKeysByFormat()
+    snapshotCurrentPriorityBaseline()

    success('优先级已保存')
    emit('saved')
--- a/frontend/src/features/usage/components/UsageRecordsTable.vue
+++ b/frontend/src/features/usage/components/UsageRecordsTable.vue
@@ -264,9 +264,9 @@
              :response-time-ms="record.response_time_ms ?? null"
            /></span>
            <span
-              v-else-if="record.response_time_ms != null"
+              v-else-if="record.response_time_ms != null || record.first_byte_time_ms != null"
              class="tabular-nums"
-            >{{ record.first_byte_time_ms != null ? (record.first_byte_time_ms / 1000).toFixed(1) + '/' : '' }}{{ (record.response_time_ms / 1000).toFixed(1) }}s</span>
+            >{{ record.first_byte_time_ms != null ? (record.first_byte_time_ms / 1000).toFixed(1) + '/' : '' }}{{ record.response_time_ms != null ? (record.response_time_ms / 1000).toFixed(1) : '-' }}{{ record.response_time_ms != null ? 's' : '' }}</span>
            <span
              v-else
              class="tabular-nums"
@@ -419,11 +419,11 @@
              <div class="flex flex-col text-xs gap-0.5">
                <span>{{ record.provider }}</span>
                <span
-                  v-if="record.api_key_name"
+                  v-if="record.provider_key_name || record.api_key_name"
                  class="text-muted-foreground truncate"
-                  :title="record.api_key_name"
+                  :title="record.provider_key_name || record.api_key_name"
                >
-                  {{ record.api_key_name }}
+                  {{ record.provider_key_name || record.api_key_name }}
                  <span
                    v-if="record.rate_multiplier && record.rate_multiplier !== 1.0"
                    class="text-foreground/60"
@@ -608,7 +608,7 @@
            </div>
            <!-- 已完成状态：首字 + 总耗时 -->
            <div
-              v-else-if="record.response_time_ms != null"
+              v-else-if="record.response_time_ms != null || record.first_byte_time_ms != null"
              class="flex flex-col items-end text-xs gap-0.5"
            >
              <span
@@ -619,7 +619,14 @@
                v-else
                class="text-muted-foreground"
              >-</span>
-              <span class="text-muted-foreground tabular-nums">{{ (record.response_time_ms / 1000).toFixed(2) }}s</span>
+              <span
+                v-if="record.response_time_ms != null"
+                class="text-muted-foreground tabular-nums"
+              >{{ (record.response_time_ms / 1000).toFixed(2) }}s</span>
+              <span
+                v-else
+                class="text-muted-foreground"
+              >-</span>
            </div>
            <span
              v-else
--- a/frontend/src/features/usage/composables/useUsageData.ts
+++ b/frontend/src/features/usage/composables/useUsageData.ts
@@ -395,6 +395,7 @@ export function useUsageData(options: UseUsageDataOptions) {
          endpoint_api_format: existing.endpoint_api_format || record.endpoint_api_format,
          has_format_conversion: existing.has_format_conversion ?? record.has_format_conversion,
          api_key_name: existing.api_key_name || record.api_key_name,
+          provider_key_name: existing.provider_key_name || record.provider_key_name,
          rate_multiplier: existing.rate_multiplier ?? record.rate_multiplier,
          target_model: existing.target_model || record.target_model
        }
--- a/frontend/src/features/usage/conversation/tests/stream.spec.ts
+++ b/frontend/src/features/usage/conversation/tests/stream.spec.ts
@@ -0,0 +1,73 @@
+import { describe, expect, it } from 'vitest'
+import { parseResponse, renderResponse } from '../registry'
+
+describe('Conversation stream compatibility', () => {
+  it('parses raw OpenAI chat SSE text from stored usage records', () => {
+    const requestBody = {
+      model: 'gpt-5.4',
+      stream: true,
+      messages: [
+        { role: 'user', content: 'Hello' },
+      ],
+    }
+    const rawSse = [
+      'data: {"id":"chatcmpl_123","object":"chat.completion.chunk","model":"gpt-5.4","choices":[{"index":0,"delta":{"role":"assistant"},"finish_reason":null}]}',
+      '',
+      'data: {"id":"chatcmpl_123","object":"chat.completion.chunk","model":"gpt-5.4","choices":[{"index":0,"delta":{"content":"Hello from stream"},"finish_reason":null}]}',
+      '',
+      'data: {"id":"chatcmpl_123","object":"chat.completion.chunk","model":"gpt-5.4","choices":[{"index":0,"delta":{},"finish_reason":"stop"}]}',
+      '',
+      'data: [DONE]',
+      '',
+    ].join('\n')
+
+    const parsed = parseResponse(rawSse, requestBody, 'openai:chat')
+    expect(parsed.apiFormat).toBe('openai')
+    expect(parsed.isStream).toBe(true)
+    expect(parsed.messages).toHaveLength(1)
+    expect(parsed.messages[0]?.content[0]).toMatchObject({
+      type: 'text',
+      text: 'Hello from stream',
+    })
+  })
+
+  it('renders raw OpenAI CLI SSE text from stored usage records', () => {
+    const requestBody = {
+      model: 'gpt-5.4',
+      stream: true,
+      input: 'Hello',
+    }
+    const rawSse = [
+      'event: response.created',
+      'data: {"type":"response.created","response":{"id":"resp_123","object":"response","model":"gpt-5.4","status":"in_progress"}}',
+      '',
+      'event: response.output_text.delta',
+      'data: {"type":"response.output_text.delta","delta":"Hello from CLI stream"}',
+      '',
+      'event: response.completed',
+      'data: {"type":"response.completed","response":{"id":"resp_123","object":"response","model":"gpt-5.4","status":"completed","output":[{"type":"message","role":"assistant","content":[{"type":"output_text","text":"Hello from CLI stream"}]}]}}',
+      '',
+      'data: [DONE]',
+      '',
+    ].join('\n')
+
+    const rendered = renderResponse(rawSse, requestBody, 'openai:cli')
+    expect(rendered.error).toBeUndefined()
+    expect(rendered.isStream).toBe(true)
+    expect(rendered.blocks).toHaveLength(1)
+    expect(rendered.blocks[0]).toMatchObject({
+      type: 'message',
+      role: 'assistant',
+    })
+
+    const firstBlock = rendered.blocks[0]
+    if (!firstBlock || firstBlock.type !== 'message') {
+      throw new Error('expected first render block to be message')
+    }
+
+    expect(firstBlock.content[0]).toMatchObject({
+      type: 'text',
+      content: 'Hello from CLI stream',
+    })
+  })
+})
--- a/frontend/src/features/usage/conversation/registry.ts
+++ b/frontend/src/features/usage/conversation/registry.ts
@@ -2,7 +2,7 @@
 * 解析器注册表和统一入口
 */

-import type { ApiFormat, ApiFormatParser, ParsedConversation } from './types'
+import type { ApiFormat, ApiFormatParser, ParsedConversation, StreamResponseBody } from './types'
 import { createEmptyConversation, isStreamResponse } from './types'
 import type { RenderResult } from './render'
 import { createEmptyRenderResult } from './render'
@@ -10,6 +10,67 @@ import { claudeParser } from './claude'
 import { openaiParser } from './openai'
 import { geminiParser } from './gemini'

+function parseRawSseResponse(responseBody: unknown): StreamResponseBody | null {
+  if (typeof responseBody !== 'string') {
+    return null
+  }
+
+  const text = responseBody.trim()
+  if (!text.includes('data:')) {
+    return null
+  }
+
+  const chunks: unknown[] = []
+  const eventBlocks = text.split(/\r?\n\r?\n/)
+
+  for (const block of eventBlocks) {
+    if (!block.trim()) {
+      continue
+    }
+
+    const dataLines = block
+      .split(/\r?\n/)
+      .map(line => line.trim())
+      .filter(line => line.startsWith('data:'))
+      .map(line => line.slice(5).trim())
+      .filter(Boolean)
+
+    if (dataLines.length === 0) {
+      continue
+    }
+
+    const payload = dataLines.join('\n')
+    if (payload === '[DONE]') {
+      continue
+    }
+
+    try {
+      chunks.push(JSON.parse(payload))
+    } catch {
+      // 保持原始数据不变，交给现有解析器兜底
+    }
+  }
+
+  if (chunks.length === 0) {
+    return null
+  }
+
+  return {
+    metadata: {
+      stream: true,
+    },
+    chunks,
+  }
+}
+
+function normalizeResponseBody(responseBody: unknown): unknown {
+  if (isStreamResponse(responseBody)) {
+    return responseBody
+  }
+
+  return parseRawSseResponse(responseBody) ?? responseBody
+}
+
 /**
 * 解析器注册表
 */
@@ -84,7 +145,11 @@ export function parseRequest(
    return createEmptyConversation('unknown', '无请求体')
  }

-  const parser = parserRegistry.detectParser(requestBody, responseBody, formatHint)
+  const parser = parserRegistry.detectParser(
+    requestBody,
+    normalizeResponseBody(responseBody),
+    formatHint
+  )
  if (!parser) {
    return createEmptyConversation('unknown', '无法识别的 API 格式')
  }
@@ -104,17 +169,18 @@ export function parseResponse(
    return createEmptyConversation('unknown', '无响应体')
  }

-  const parser = parserRegistry.detectParser(requestBody, responseBody, formatHint)
+  const normalizedResponseBody = normalizeResponseBody(responseBody)
+  const parser = parserRegistry.detectParser(requestBody, normalizedResponseBody, formatHint)
  if (!parser) {
    return createEmptyConversation('unknown', '无法识别的 API 格式')
  }

  // 判断是否为流式响应
-  if (isStreamResponse(responseBody)) {
-    return parser.parseStreamResponse(responseBody.chunks || [])
+  if (isStreamResponse(normalizedResponseBody)) {
+    return parser.parseStreamResponse(normalizedResponseBody.chunks || [])
  }

-  return parser.parseResponse(responseBody)
+  return parser.parseResponse(normalizedResponseBody)
 }

 /**
@@ -125,7 +191,7 @@ export function detectApiFormat(
  responseBody: unknown,
  hint?: string
 ): ApiFormat {
-  return parserRegistry.detectFormat(requestBody, responseBody, hint)
+  return parserRegistry.detectFormat(requestBody, normalizeResponseBody(responseBody), hint)
 }

 /**
@@ -140,7 +206,11 @@ export function renderRequest(
    return createEmptyRenderResult('无请求体')
  }

-  const parser = parserRegistry.detectParser(requestBody, responseBody, formatHint)
+  const parser = parserRegistry.detectParser(
+    requestBody,
+    normalizeResponseBody(responseBody),
+    formatHint
+  )
  if (!parser) {
    return createEmptyRenderResult('无法识别的 API 格式')
  }
@@ -160,10 +230,11 @@ export function renderResponse(
    return createEmptyRenderResult('无响应体')
  }

-  const parser = parserRegistry.detectParser(requestBody, responseBody, formatHint)
+  const normalizedResponseBody = normalizeResponseBody(responseBody)
+  const parser = parserRegistry.detectParser(requestBody, normalizedResponseBody, formatHint)
  if (!parser) {
    return createEmptyRenderResult('无法识别的 API 格式')
  }

-  return parser.renderResponse(responseBody)
+  return parser.renderResponse(normalizedResponseBody)
 }
--- a/frontend/src/features/usage/types.ts
+++ b/frontend/src/features/usage/types.ts
@@ -83,6 +83,7 @@ export interface UsageRecord {
  } | null
  provider?: string  // 仅管理员可见
  api_key_name?: string
+  provider_key_name?: string | null
  rate_multiplier?: number
  model: string
  target_model?: string | null  // 映射后的目标模型名（若无映射则为空）