feat: add USE_UNSTRUCTURED_FOR_PDF environment variable to knowledge config

Author: fzlzjerry

src/config/knowledge.ts

@@ -1,19 +1,17 @@
 import { createEnv } from '@t3-oss/env-nextjs';
 import { z } from 'zod';
 
-export const getKnowledgeConfig = () => {
-  return createEnv({
-    runtimeEnv: {
-      DEFAULT_FILES_CONFIG: process.env.DEFAULT_FILES_CONFIG,
-      UNSTRUCTURED_API_KEY: process.env.UNSTRUCTURED_API_KEY,
-      UNSTRUCTURED_SERVER_URL: process.env.UNSTRUCTURED_SERVER_URL,
-    },
-    server: {
-      DEFAULT_FILES_CONFIG: z.string().optional(),
-      UNSTRUCTURED_API_KEY: z.string().optional(),
-      UNSTRUCTURED_SERVER_URL: z.string().optional(),
-    },
-  });
-};
-
-export const knowledgeEnv = getKnowledgeConfig();
+export const knowledgeEnv = createEnv({
+  runtimeEnv: {
+    DEFAULT_FILES_CONFIG: process.env.DEFAULT_FILES_CONFIG,
+    UNSTRUCTURED_API_KEY: process.env.UNSTRUCTURED_API_KEY,
+    UNSTRUCTURED_SERVER_URL: process.env.UNSTRUCTURED_SERVER_URL,
+    USE_UNSTRUCTURED_FOR_PDF: process.env.USE_UNSTRUCTURED_FOR_PDF,
+  },
+  server: {
+    DEFAULT_FILES_CONFIG: z.string().optional(),
+    UNSTRUCTURED_API_KEY: z.string().optional(),
+    UNSTRUCTURED_SERVER_URL: z.string().optional(),
+    USE_UNSTRUCTURED_FOR_PDF: z.string().optional(),
+  },
+});

src/server/modules/ContentChunk/index.ts

@@ -3,12 +3,8 @@ import { Strategy } from 'unstructured-client/sdk/models/shared';
 
 import { NewChunkItem, NewUnstructuredChunkItem } from '@/database/schemas';
 import { ChunkingStrategy, Unstructured } from '@/libs/unstructured';
-import { getEnvironment } from '../../utils/env';
 import { knowledgeEnv } from '@/config/knowledge';
 
-// Get environment variable to control PDF processing method
-const USE_UNSTRUCTURED_FOR_PDF = getEnvironment('USE_UNSTRUCTURED_FOR_PDF') === 'true';
-
 export interface ChunkContentParams {
   content: Uint8Array;
   fileType: string;
@@ -32,7 +28,7 @@ export class ContentChunk {
 
   isUsingUnstructured(params: ChunkContentParams) {
     return params.fileType === 'application/pdf' &&
-           USE_UNSTRUCTURED_FOR_PDF &&
+           !!knowledgeEnv.USE_UNSTRUCTURED_FOR_PDF &&
            !!knowledgeEnv.UNSTRUCTURED_API_KEY &&
            !!knowledgeEnv.UNSTRUCTURED_SERVER_URL;
   }