feat: #77

app/api/langchain-tools/rag_search.ts

@@ -7,6 +7,8 @@ import { RunnableSequence } from "@langchain/core/runnables";
 import { StringOutputParser } from "@langchain/core/output_parsers";
 import { Pinecone } from "@pinecone-database/pinecone";
 import { PineconeStore } from "@langchain/pinecone";
+import { getServerSideConfig } from "@/app/config/server";
+import { QdrantVectorStore } from "@langchain/community/vectorstores/qdrant";
 
 export class RAGSearch extends Tool {
   static lc_name() {
@@ -34,21 +36,32 @@ export class RAGSearch extends Tool {
 
   /** @ignore */
   async _call(inputs: string, runManager?: CallbackManagerForToolRun) {
-    const pinecone = new Pinecone();
-    const pineconeIndex = pinecone.Index(process.env.PINECONE_INDEX!);
-    const vectorStore = await PineconeStore.fromExistingIndex(this.embeddings, {
-      pineconeIndex,
-    });
+    const serverConfig = getServerSideConfig();
+    // const pinecone = new Pinecone();
+    // const pineconeIndex = pinecone.Index(serverConfig.pineconeIndex!);
+    // const vectorStore = await PineconeStore.fromExistingIndex(this.embeddings, {
+    //   pineconeIndex,
+    // });
+    const vectorStore = await QdrantVectorStore.fromExistingCollection(
+      this.embeddings,
+      {
+        url: process.env.QDRANT_URL,
+        apiKey: process.env.QDRANT_API_KEY,
+        collectionName: this.sessionId,
+      },
+    );
 
     let context;
-    const returnCunt = process.env.RAG_RETURN_COUNT
-      ? parseInt(process.env.RAG_RETURN_COUNT, 10)
+    const returnCunt = serverConfig.ragReturnCount
+      ? parseInt(serverConfig.ragReturnCount, 10)
       : 4;
-    const results = await vectorStore.similaritySearch(inputs, returnCunt, {
-      sessionId: this.sessionId,
-    });
+    console.log("[rag-search]", { inputs, returnCunt });
+    // const results = await vectorStore.similaritySearch(inputs, returnCunt, {
+    //   sessionId: this.sessionId,
+    // });
+    const results = await vectorStore.similaritySearch(inputs, returnCunt);
     context = formatDocumentsAsString(results);
-    console.log("[rag-search]", context);
+    console.log("[rag-search]", { context });
     return context;
     // const input = `Text:${context}\n\nQuestion:${inputs}\n\nI need you to answer the question based on the text.`;
 

app/api/langchain/rag/search/route.ts

@@ -4,6 +4,7 @@ import { ACCESS_CODE_PREFIX, ModelProvider } from "@/app/constant";
 import { OpenAIEmbeddings } from "@langchain/openai";
 import { Pinecone } from "@pinecone-database/pinecone";
 import { PineconeStore } from "@langchain/pinecone";
+import { QdrantVectorStore } from "@langchain/community/vectorstores/qdrant";
 import { getServerSideConfig } from "@/app/config/server";
 
 interface RequestBody {
@@ -27,26 +28,40 @@ async function handle(req: NextRequest) {
     const reqBody: RequestBody = await req.json();
     const authToken = req.headers.get("Authorization") ?? "";
     const token = authToken.trim().replaceAll("Bearer ", "").trim();
-
-    const pinecone = new Pinecone();
-    const pineconeIndex = pinecone.Index(process.env.PINECONE_INDEX!);
+    const serverConfig = getServerSideConfig();
+    // const pinecone = new Pinecone();
+    // const pineconeIndex = pinecone.Index(serverConfig.pineconeIndex!);
     const apiKey = getOpenAIApiKey(token);
     const baseUrl = getOpenAIBaseUrl(reqBody.baseUrl);
     const embeddings = new OpenAIEmbeddings(
       {
-        modelName: process.env.RAG_EMBEDDING_MODEL ?? "text-embedding-3-large",
+        modelName: serverConfig.ragEmbeddingModel ?? "text-embedding-3-large",
         openAIApiKey: apiKey,
       },
       { basePath: baseUrl },
     );
-    const vectorStore = await PineconeStore.fromExistingIndex(embeddings, {
-      pineconeIndex,
-    });
-    const results = await vectorStore.similaritySearch(reqBody.query, 1, {
-      sessionId: reqBody.sessionId,
-    });
-    console.log(results);
-    return NextResponse.json(results, {
+    // const vectorStore = await PineconeStore.fromExistingIndex(embeddings, {
+    //   pineconeIndex,
+    // });
+    // const results = await vectorStore.similaritySearch(reqBody.query, 4, {
+    //   sessionId: reqBody.sessionId,
+    // });
+    const vectorStore = await QdrantVectorStore.fromExistingCollection(
+      embeddings,
+      {
+        url: process.env.QDRANT_URL,
+        apiKey: process.env.QDRANT_API_KEY,
+        collectionName: reqBody.sessionId,
+      },
+    );
+    const returnCunt = serverConfig.ragReturnCount
+      ? parseInt(serverConfig.ragReturnCount, 10)
+      : 4;
+    const response = await vectorStore.similaritySearch(
+      reqBody.query,
+      returnCunt,
+    );
+    return NextResponse.json(response, {
       status: 200,
     });
   } catch (e) {

app/api/langchain/rag/store/route.ts

@@ -20,6 +20,7 @@ import { FileInfo } from "@/app/client/platforms/utils";
 import mime from "mime";
 import LocalFileStorage from "@/app/utils/local_file_storage";
 import S3FileStorage from "@/app/utils/s3_file_storage";
+import { QdrantVectorStore } from "@langchain/community/vectorstores/qdrant";
 
 interface RequestBody {
   sessionId: string;
@@ -80,16 +81,17 @@ async function handle(req: NextRequest) {
     const apiKey = getOpenAIApiKey(token);
     const baseUrl = getOpenAIBaseUrl(reqBody.baseUrl);
     const serverConfig = getServerSideConfig();
-    const pinecone = new Pinecone();
-    const pineconeIndex = pinecone.Index(process.env.PINECONE_INDEX!);
+    // const pinecone = new Pinecone();
+    // const pineconeIndex = pinecone.Index(serverConfig.pineconeIndex!);
     const embeddings = new OpenAIEmbeddings(
       {
-        modelName: process.env.RAG_EMBEDDING_MODEL ?? "text-embedding-3-large",
+        modelName: serverConfig.ragEmbeddingModel,
         openAIApiKey: apiKey,
       },
       { basePath: baseUrl },
     );
-    //https://js.langchain.com/docs/integrations/vectorstores/pinecone
+    // https://js.langchain.com/docs/integrations/vectorstores/pinecone
+    // https://js.langchain.com/docs/integrations/vectorstores/qdrant
     // process files
     for (let i = 0; i < reqBody.fileInfos.length; i++) {
       const fileInfo = reqBody.fileInfos[i];
@@ -121,22 +123,33 @@ async function handle(req: NextRequest) {
         };
       });
       // split
-      const chunkSize = process.env.RAG_CHUNK_SIZE
-        ? parseInt(process.env.RAG_CHUNK_SIZE, 10)
+      const chunkSize = serverConfig.ragChunkSize
+        ? parseInt(serverConfig.ragChunkSize, 10)
         : 2000;
-      const chunkOverlap = process.env.RAG_CHUNK_OVERLAP
-        ? parseInt(process.env.RAG_CHUNK_OVERLAP, 10)
+      const chunkOverlap = serverConfig.ragChunkOverlap
+        ? parseInt(serverConfig.ragChunkOverlap, 10)
         : 200;
       const textSplitter = new RecursiveCharacterTextSplitter({
         chunkSize: chunkSize,
         chunkOverlap: chunkOverlap,
       });
       const splits = await textSplitter.splitDocuments(docs);
-      // remove history
-      await PineconeStore.fromDocuments(splits, embeddings, {
-        pineconeIndex,
-        maxConcurrency: 5,
-      });
+      const vectorStore = await QdrantVectorStore.fromDocuments(
+        splits,
+        embeddings,
+        {
+          url: process.env.QDRANT_URL,
+          apiKey: process.env.QDRANT_API_KEY,
+          collectionName: reqBody.sessionId,
+        },
+      );
+      // await PineconeStore.fromDocuments(splits, embeddings, {
+      //   pineconeIndex,
+      //   maxConcurrency: 5,
+      // });
+      // const vectorStore = await PineconeStore.fromExistingIndex(embeddings, {
+      //   pineconeIndex,
+      // });
     }
     return NextResponse.json(
       {