refactor: enhance model configuration and plugin execution

- Simplified the `createModel` function to directly accept the `ModelConfig` object, improving clarity. - Updated `createBaseModel` to include `extraModelConfig` for extended configuration options. - Introduced `executeConfigureContext` method in `PluginManager` to handle context configuration for plugins. - Adjusted type definitions in `types.ts` to ensure consistency with the new configuration structure. - Refactored plugin execution methods in `PluginEngine` to utilize the resolved model directly, enhancing the flow of data through the plugin system.
2025-12-28 05:11:24 +08:00 · 2025-07-07 18:33:47 +08:00 · 2025-07-07 18:33:47 +08:00 · 0a908a334b
commit 0a908a334b
parent c72156b2da
7 changed files with 85 additions and 42 deletions
--- a/packages/aiCore/src/core/models/ProviderCreator.ts
+++ b/packages/aiCore/src/core/models/ProviderCreator.ts
@ -27,24 +27,28 @@ export class ProviderCreationError extends Error {
 export async function createBaseModel<T extends ProviderId>({
  providerId,
  modelId,
-  providerSettings
+  providerSettings,
+  extraModelConfig
  // middlewares
 }: {
  providerId: T
  modelId: string
  providerSettings: ProviderSettingsMap[T]
+  extraModelConfig?: any
  // middlewares?: LanguageModelV1Middleware[]
 }): Promise<LanguageModelV2>

 export async function createBaseModel({
  providerId,
  modelId,
-  providerSettings
+  providerSettings,
+  extraModelConfig
  // middlewares
 }: {
  providerId: string
  modelId: string
  providerSettings: ProviderSettingsMap['openai-compatible']
+  extraModelConfig?: any
  // middlewares?: LanguageModelV1Middleware[]
 }): Promise<LanguageModelV2>

--- a/packages/aiCore/src/core/models/factory.ts
+++ b/packages/aiCore/src/core/models/factory.ts
@ -16,11 +16,7 @@ export async function createModel(config: ModelConfig): Promise<LanguageModelV2>
  validateModelConfig(config)

  // 1. 创建基础模型
-  const baseModel = await createBaseModel({
-    providerId: config.providerId,
-    modelId: config.modelId,
-    providerSettings: config.options
-  })
+  const baseModel = await createBaseModel(config)

  // 2. 应用中间件（如果有）
  return config.middlewares?.length ? wrapModelWithMiddlewares(baseModel, config.middlewares) : baseModel
@ -43,7 +39,7 @@ function validateModelConfig(config: ModelConfig): void {
  if (!config.modelId) {
    throw new Error('ModelConfig: modelId is required')
  }
-  if (!config.options) {
+  if (!config.providerSettings) {
    throw new Error('ModelConfig: providerSettings is required')
  }
 }
--- a/packages/aiCore/src/core/models/types.ts
+++ b/packages/aiCore/src/core/models/types.ts
@ -5,9 +5,10 @@ import { LanguageModelV2Middleware } from '@ai-sdk/provider'

 import type { ProviderId, ProviderSettingsMap } from '../../types'

-export interface ModelConfig {
-  providerId: ProviderId
+export interface ModelConfig<T extends ProviderId = ProviderId> {
+  providerId: T
  modelId: string
-  options: ProviderSettingsMap[ProviderId]
+  providerSettings: ProviderSettingsMap[T]
  middlewares?: LanguageModelV2Middleware[]
+  extraModelConfig?: Record<string, any>
 }
--- a/packages/aiCore/src/core/plugins/manager.ts
+++ b/packages/aiCore/src/core/plugins/manager.ts
@ -71,7 +71,7 @@ export class PluginManager {
   * 执行 Sequential 钩子 - 链式数据转换
   */
  async executeSequential<T>(
-    hookName: 'transformParams' | 'transformResult' | 'configureModel',
+    hookName: 'transformParams' | 'transformResult',
    initialValue: T,
    context: AiRequestContext
  ): Promise<T> {
@ -87,6 +87,18 @@ export class PluginManager {
    return result
  }

+  /**
+   * 执行 ConfigureContext 钩子 - 串行配置上下文
+   */
+  async executeConfigureContext(context: AiRequestContext): Promise<void> {
+    for (const plugin of this.plugins) {
+      const hook = plugin.configureContext
+      if (hook) {
+        await hook(context)
+      }
+    }
+  }
+
  /**
   * 执行 Parallel 钩子 - 并行副作用
   */
--- a/packages/aiCore/src/core/plugins/types.ts
+++ b/packages/aiCore/src/core/plugins/types.ts
@ -1,4 +1,4 @@
-import type { TextStreamPart, ToolSet } from 'ai'
+import type { LanguageModel, TextStreamPart, ToolSet } from 'ai'

 import { ProviderId } from '../providers/registry'

@ -32,13 +32,13 @@ export interface AiPlugin {
  enforce?: 'pre' | 'post'

  // 【First】首个钩子 - 只执行第一个返回值的插件
-  resolveModel?: (modelId: string, context: AiRequestContext) => string | null | Promise<string | null>
+  resolveModel?: (modelId: string, context: AiRequestContext) => Promise<LanguageModel | null> | LanguageModel | null
  loadTemplate?: (templateName: string, context: AiRequestContext) => any | null | Promise<any | null>

  // 【Sequential】串行钩子 - 链式执行，支持数据转换
+  configureContext?: (context: AiRequestContext) => void | Promise<void>
  transformParams?: (params: any, context: AiRequestContext) => any | Promise<any>
  transformResult?: (result: any, context: AiRequestContext) => any | Promise<any>
-  configureModel?: (model: any, context: AiRequestContext) => any | Promise<any>

  // 【Parallel】并行钩子 - 不依赖顺序，用于副作用
  onRequestStart?: (context: AiRequestContext) => void | Promise<void>
--- a/packages/aiCore/src/core/runtime/executor.ts
+++ b/packages/aiCore/src/core/runtime/executor.ts
@ -7,7 +7,7 @@ import { generateObject, generateText, LanguageModel, streamObject, streamText }

 import { type ProviderId, type ProviderSettingsMap } from '../../types'
 import { createModel, getProviderInfo } from '../models'
-import { type AiPlugin } from '../plugins'
+import { type AiPlugin, type AiRequestContext, definePlugin } from '../plugins'
 import { PluginEngine } from './pluginEngine'
 import { type RuntimeConfig } from './types'

@ -28,6 +28,19 @@ export class RuntimeExecutor<T extends ProviderId = ProviderId> {
    this.pluginEngine = new PluginEngine(config.providerId, config.plugins || [])
  }

+  createResolveModelPlugin(middlewares?: LanguageModelV2Middleware[]) {
+    return definePlugin({
+      name: '_internal_resolveModel',
+      enforce: 'post',
+
+      resolveModel: async (modelId: string, context: AiRequestContext) => {
+        // 从 context 中读取由用户插件注入的 extraModelConfig
+        const extraModelConfig = context.extraModelConfig || {}
+        return await this.resolveModel(modelId, middlewares, extraModelConfig)
+      }
+    })
+  }
+
  // === 高阶重载：直接使用模型 ===

  /**
@ -59,14 +72,14 @@ export class RuntimeExecutor<T extends ProviderId = ProviderId> {
      middlewares?: LanguageModelV2Middleware[]
    }
  ): Promise<ReturnType<typeof streamText>> {
-    const model = await this.resolveModel(modelOrId, options?.middlewares)
+    this.pluginEngine.use(this.createResolveModelPlugin(options?.middlewares))

    // 2. 执行插件处理
    return this.pluginEngine.executeStreamWithPlugins(
      'streamText',
-      typeof modelOrId === 'string' ? modelOrId : model.modelId,
+      typeof modelOrId === 'string' ? modelOrId : modelOrId.modelId,
      params,
-      async (finalModelId, transformedParams, streamTransforms) => {
+      async (model, transformedParams, streamTransforms) => {
        const experimental_transform =
          params?.experimental_transform ?? (streamTransforms.length > 0 ? streamTransforms : undefined)

@ -110,13 +123,13 @@ export class RuntimeExecutor<T extends ProviderId = ProviderId> {
      middlewares?: LanguageModelV2Middleware[]
    }
  ): Promise<ReturnType<typeof generateText>> {
-    const model = await this.resolveModel(modelOrId, options?.middlewares)
+    this.pluginEngine.use(this.createResolveModelPlugin(options?.middlewares))

    return this.pluginEngine.executeWithPlugins(
      'generateText',
-      typeof modelOrId === 'string' ? modelOrId : model.modelId,
+      typeof modelOrId === 'string' ? modelOrId : modelOrId.modelId,
      params,
-      async (finalModelId, transformedParams) => {
+      async (model, transformedParams) => {
        return await generateText({ model, ...transformedParams })
      }
    )
@ -151,13 +164,13 @@ export class RuntimeExecutor<T extends ProviderId = ProviderId> {
      middlewares?: LanguageModelV2Middleware[]
    }
  ): Promise<ReturnType<typeof generateObject>> {
-    const model = await this.resolveModel(modelOrId, options?.middlewares)
+    this.pluginEngine.use(this.createResolveModelPlugin(options?.middlewares))

    return this.pluginEngine.executeWithPlugins(
      'generateObject',
-      typeof modelOrId === 'string' ? modelOrId : model.modelId,
+      typeof modelOrId === 'string' ? modelOrId : modelOrId.modelId,
      params,
-      async (finalModelId, transformedParams) => {
+      async (model, transformedParams) => {
        return await generateObject({ model, ...transformedParams })
      }
    )
@ -192,13 +205,13 @@ export class RuntimeExecutor<T extends ProviderId = ProviderId> {
      middlewares?: LanguageModelV2Middleware[]
    }
  ): Promise<ReturnType<typeof streamObject>> {
-    const model = await this.resolveModel(modelOrId, options?.middlewares)
+    this.pluginEngine.use(this.createResolveModelPlugin(options?.middlewares))

    return this.pluginEngine.executeWithPlugins(
      'streamObject',
-      typeof modelOrId === 'string' ? modelOrId : model.modelId,
+      typeof modelOrId === 'string' ? modelOrId : modelOrId.modelId,
      params,
-      async (finalModelId, transformedParams) => {
+      async (model, transformedParams) => {
        return await streamObject({ model, ...transformedParams })
      }
    )
@ -211,15 +224,17 @@ export class RuntimeExecutor<T extends ProviderId = ProviderId> {
   */
  private async resolveModel(
    modelOrId: LanguageModel,
-    middlewares?: LanguageModelV2Middleware[]
+    middlewares?: LanguageModelV2Middleware[],
+    extraModelConfig?: Record<string, any>
  ): Promise<LanguageModelV2> {
    if (typeof modelOrId === 'string') {
      // 字符串modelId，需要创建模型
      return await createModel({
        providerId: this.config.providerId,
        modelId: modelOrId,
-        options: this.config.providerSettings,
-        middlewares
+        providerSettings: this.config.providerSettings,
+        middlewares,
+        extraModelConfig
      })
    } else {
      // 已经是模型，直接返回
--- a/packages/aiCore/src/core/runtime/pluginEngine.ts
+++ b/packages/aiCore/src/core/runtime/pluginEngine.ts
@ -1,3 +1,5 @@
+import { LanguageModel } from 'ai'
+
 import { type ProviderId, type ProviderSettingsMap } from '../../types'
 import { type AiPlugin, createContext, PluginManager } from '../plugins'
 import { isProviderSupported } from '../providers/registry'
@ -63,7 +65,7 @@ export class PluginEngine<T extends ProviderId = ProviderId> {
    methodName: string,
    modelId: string,
    params: TParams,
-    executor: (finalModelId: string, transformedParams: TParams) => Promise<TResult>,
+    executor: (model: LanguageModel, transformedParams: TParams) => Promise<TResult>,
    _context?: ReturnType<typeof createContext>
  ): Promise<TResult> {
    // 使用正确的createContext创建请求上下文
@ -79,18 +81,23 @@ export class PluginEngine<T extends ProviderId = ProviderId> {
    }

    try {
+      // 0. 配置上下文
+      await this.pluginManager.executeConfigureContext(context)
+
      // 1. 触发请求开始事件
      await this.pluginManager.executeParallel('onRequestStart', context)

-      // 2. 解析模型别名
-      const resolvedModelId = await this.pluginManager.executeFirst<string>('resolveModel', modelId, context)
-      const finalModelId = resolvedModelId || modelId
+      // 2. 解析模型
+      const model = await this.pluginManager.executeFirst<LanguageModel>('resolveModel', modelId, context)
+      if (!model) {
+        throw new Error(`Failed to resolve model: ${modelId}`)
+      }

      // 3. 转换请求参数
      const transformedParams = await this.pluginManager.executeSequential('transformParams', params, context)

      // 4. 执行具体的 API 调用
-      const result = await executor(finalModelId, transformedParams)
+      const result = await executor(model, transformedParams)

      // 5. 转换结果（对于非流式调用）
      const transformedResult = await this.pluginManager.executeSequential('transformResult', result, context)
@ -114,7 +121,7 @@ export class PluginEngine<T extends ProviderId = ProviderId> {
    methodName: string,
    modelId: string,
    params: TParams,
-    executor: (finalModelId: string, transformedParams: TParams, streamTransforms: any[]) => Promise<TResult>,
+    executor: (model: LanguageModel, transformedParams: TParams, streamTransforms: any[]) => Promise<TResult>,
    _context?: ReturnType<typeof createContext>
  ): Promise<TResult> {
    // 创建请求上下文
@ -130,12 +137,18 @@ export class PluginEngine<T extends ProviderId = ProviderId> {
    }

    try {
+      // 0. 配置上下文
+      await this.pluginManager.executeConfigureContext(context)
+
      // 1. 触发请求开始事件
      await this.pluginManager.executeParallel('onRequestStart', context)

-      // 2. 解析模型别名
-      const resolvedModelId = await this.pluginManager.executeFirst<string>('resolveModel', modelId, context)
-      const finalModelId = resolvedModelId || modelId
+      // 2. 解析模型
+      const model = await this.pluginManager.executeFirst<LanguageModel>('resolveModel', modelId, context)
+
+      if (!model) {
+        throw new Error(`Failed to resolve model: ${modelId}`)
+      }

      // 3. 转换请求参数
      const transformedParams = await this.pluginManager.executeSequential('transformParams', params, context)
@ -144,12 +157,14 @@ export class PluginEngine<T extends ProviderId = ProviderId> {
      const streamTransforms = this.pluginManager.collectStreamTransforms(transformedParams, context)

      // 5. 执行流式 API 调用
-      const result = await executor(finalModelId, transformedParams, streamTransforms)
+      const result = await executor(model, transformedParams, streamTransforms)
+
+      const transformedResult = await this.pluginManager.executeSequential('transformResult', result, context)

      // 6. 触发完成事件（注意：对于流式调用，这里触发的是开始流式响应的事件）
      await this.pluginManager.executeParallel('onRequestEnd', context, { stream: true })

-      return result
+      return transformedResult
    } catch (error) {
      // 7. 触发错误事件
      await this.pluginManager.executeParallel('onError', context, undefined, error as Error)