langchain 0.1.25 → 0.1.27

package/README.md

@@ -2,7 +2,7 @@
 
 ⚡ Building applications with LLMs through composability ⚡
 
-[![CI](https://github.com/langchain-ai/langchainjs/actions/workflows/ci.yml/badge.svg)](https://github.com/langchain-ai/langchainjs/actions/workflows/ci.yml) ![npm](https://img.shields.io/npm/dw/langchain) [![License: MIT](https://img.shields.io/badge/License-MIT-yellow.svg)](https://opensource.org/licenses/MIT) [![Twitter](https://img.shields.io/twitter/url/https/twitter.com/langchainai.svg?style=social&label=Follow%20%40LangChainAI)](https://twitter.com/langchainai) [![](https://dcbadge.vercel.app/api/server/6adMQxSpJS?compact=true&style=flat)](https://discord.gg/6adMQxSpJS) [![Open in Dev Containers](https://img.shields.io/static/v1?label=Dev%20Containers&message=Open&color=blue&logo=visualstudiocode)](https://vscode.dev/redirect?url=vscode://ms-vscode-remote.remote-containers/cloneInVolume?url=https://github.com/langchain-ai/langchainjs)
+[![CI](https://github.com/langchain-ai/langchainjs/actions/workflows/ci.yml/badge.svg)](https://github.com/langchain-ai/langchainjs/actions/workflows/ci.yml) ![npm](https://img.shields.io/npm/dm/langchain) [![License: MIT](https://img.shields.io/badge/License-MIT-yellow.svg)](https://opensource.org/licenses/MIT) [![Twitter](https://img.shields.io/twitter/url/https/twitter.com/langchainai.svg?style=social&label=Follow%20%40LangChainAI)](https://twitter.com/langchainai) [![](https://dcbadge.vercel.app/api/server/6adMQxSpJS?compact=true&style=flat)](https://discord.gg/6adMQxSpJS) [![Open in Dev Containers](https://img.shields.io/static/v1?label=Dev%20Containers&message=Open&color=blue&logo=visualstudiocode)](https://vscode.dev/redirect?url=vscode://ms-vscode-remote.remote-containers/cloneInVolume?url=https://github.com/langchain-ai/langchainjs)
 [<img src="https://github.com/codespaces/badge.svg" title="Open in Github Codespace" width="150" height="20">](https://codespaces.new/langchain-ai/langchainjs)
 
 Looking for the Python version? Check out [LangChain](https://github.com/langchain-ai/langchain).

package/dist/document_loaders/web/gitbook.cjs

@@ -23,6 +23,13 @@ class GitbookLoader extends cheerio_js_1.CheerioWebBaseLoader {
             writable: true,
             value: false
         });
+        Object.defineProperty(this, "baseUrl", {
+            enumerable: true,
+            configurable: true,
+            writable: true,
+            value: void 0
+        });
+        this.baseUrl = webPath;
         this.webPath = path;
         this.shouldLoadAllPaths =
             params.shouldLoadAllPaths ?? this.shouldLoadAllPaths;
@@ -84,9 +91,10 @@ class GitbookLoader extends cheerio_js_1.CheerioWebBaseLoader {
             .map((element) => $(element).text());
         const documents = [];
         for (const url of urls) {
-            console.log(`Fetching text from ${url}`);
-            const html = await GitbookLoader._scrape(url, this.caller, this.timeout);
-            documents.push(...this.loadPath(html, url));
+            const buildUrl = url.includes(this.baseUrl) ? url : this.baseUrl + url;
+            console.log(`Fetching text from ${buildUrl}`);
+            const html = await GitbookLoader._scrape(buildUrl, this.caller, this.timeout);
+            documents.push(...this.loadPath(html, buildUrl));
         }
         console.log(`Fetched ${documents.length} documents.`);
         return documents;

package/dist/document_loaders/web/gitbook.d.ts

@@ -15,6 +15,7 @@ interface GitbookLoaderParams {
 export declare class GitbookLoader extends CheerioWebBaseLoader {
     webPath: string;
     shouldLoadAllPaths: boolean;
+    private readonly baseUrl;
     constructor(webPath: string, params?: GitbookLoaderParams);
     /**
      * Method that scrapes the web document using Cheerio and loads the

package/dist/document_loaders/web/gitbook.js

@@ -20,6 +20,13 @@ export class GitbookLoader extends CheerioWebBaseLoader {
             writable: true,
             value: false
         });
+        Object.defineProperty(this, "baseUrl", {
+            enumerable: true,
+            configurable: true,
+            writable: true,
+            value: void 0
+        });
+        this.baseUrl = webPath;
         this.webPath = path;
         this.shouldLoadAllPaths =
             params.shouldLoadAllPaths ?? this.shouldLoadAllPaths;
@@ -81,9 +88,10 @@ export class GitbookLoader extends CheerioWebBaseLoader {
             .map((element) => $(element).text());
         const documents = [];
         for (const url of urls) {
-            console.log(`Fetching text from ${url}`);
-            const html = await GitbookLoader._scrape(url, this.caller, this.timeout);
-            documents.push(...this.loadPath(html, url));
+            const buildUrl = url.includes(this.baseUrl) ? url : this.baseUrl + url;
+            console.log(`Fetching text from ${buildUrl}`);
+            const html = await GitbookLoader._scrape(buildUrl, this.caller, this.timeout);
+            documents.push(...this.loadPath(html, buildUrl));
         }
         console.log(`Fetched ${documents.length} documents.`);
         return documents;

package/dist/experimental/chat_models/anthropic_functions.cjs

@@ -28,6 +28,7 @@ for the weather in SF you would respond:
 
 <tool>search</tool><tool_input><query>weather in SF</query></tool_input>
 <observation>64 degrees</observation>`);
+/** @deprecated Install and use in "@langchain/anthropic/experimental" instead */
 class AnthropicFunctions extends chat_models_1.BaseChatModel {
     static lc_name() {
         return "AnthropicFunctions";

package/dist/experimental/chat_models/anthropic_functions.d.ts

@@ -6,13 +6,16 @@ import { BaseFunctionCallOptions } from "@langchain/core/language_models/base";
 import { CallbackManagerForLLMRun } from "@langchain/core/callbacks/manager";
 import { BasePromptTemplate } from "@langchain/core/prompts";
 import { type AnthropicInput } from "../../chat_models/anthropic.js";
+/** @deprecated Install and use in "@langchain/anthropic/experimental" instead */
 export interface ChatAnthropicFunctionsCallOptions extends BaseFunctionCallOptions {
     tools?: StructuredToolInterface[];
 }
+/** @deprecated Install and use in "@langchain/anthropic/experimental" instead */
 export type AnthropicFunctionsInput = Partial<AnthropicInput> & BaseChatModelParams & {
     llm?: BaseChatModel;
     systemPromptTemplate?: BasePromptTemplate;
 };
+/** @deprecated Install and use in "@langchain/anthropic/experimental" instead */
 export declare class AnthropicFunctions extends BaseChatModel<ChatAnthropicFunctionsCallOptions> {
     llm: BaseChatModel;
     stopSequences?: string[];

package/dist/experimental/chat_models/anthropic_functions.js

@@ -25,6 +25,7 @@ for the weather in SF you would respond:
 
 <tool>search</tool><tool_input><query>weather in SF</query></tool_input>
 <observation>64 degrees</observation>`);
+/** @deprecated Install and use in "@langchain/anthropic/experimental" instead */
 export class AnthropicFunctions extends BaseChatModel {
     static lc_name() {
         return "AnthropicFunctions";

package/dist/llms/bedrock/index.cjs

@@ -11,6 +11,7 @@ class Bedrock extends web_js_1.Bedrock {
     static lc_name() {
         return "Bedrock";
     }
+    // eslint-disable-next-line @typescript-eslint/no-explicit-any
     constructor(fields) {
         super({
             ...fields,

package/dist/llms/bedrock/index.d.ts

@@ -1,7 +1,6 @@
 import { BaseLLMParams } from "@langchain/core/language_models/llms";
-import { BaseBedrockInput } from "../../util/bedrock.js";
 import { Bedrock as BaseBedrock } from "./web.js";
 export declare class Bedrock extends BaseBedrock {
     static lc_name(): string;
-    constructor(fields?: Partial<BaseBedrockInput> & BaseLLMParams);
+    constructor(fields?: Partial<any> & BaseLLMParams);
 }

package/dist/llms/bedrock/index.js

@@ -8,6 +8,7 @@ export class Bedrock extends BaseBedrock {
     static lc_name() {
         return "Bedrock";
     }
+    // eslint-disable-next-line @typescript-eslint/no-explicit-any
     constructor(fields) {
         super({
             ...fields,

package/dist/retrievers/parent_document.cjs

@@ -153,9 +153,10 @@ class ParentDocumentRetriever extends multi_vector_js_1.MultiVectorRetriever {
      * This can be false if and only if `ids` are provided. You may want
      *   to set this to False if the documents are already in the docstore
      *   and you don't want to re-add them.
+     * @param config.chunkHeaderOptions Object with options for adding Contextual chunk headers
      */
     async addDocuments(docs, config) {
-        const { ids, addToDocstore = true } = config ?? {};
+        const { ids, addToDocstore = true, childDocChunkHeaderOptions = {}, } = config ?? {};
         const parentDocs = this.parentSplitter
             ? await this.parentSplitter.splitDocuments(docs)
             : docs;
@@ -177,7 +178,7 @@ class ParentDocumentRetriever extends multi_vector_js_1.MultiVectorRetriever {
         for (let i = 0; i < parentDocs.length; i += 1) {
             const parentDoc = parentDocs[i];
             const parentDocId = parentDocIds[i];
-            const subDocs = await this.childSplitter.splitDocuments([parentDoc]);
+            const subDocs = await this.childSplitter.splitDocuments([parentDoc], childDocChunkHeaderOptions);
             const taggedSubDocs = subDocs.map((subDoc) => new documents_1.Document({
                 pageContent: subDoc.pageContent,
                 metadata: { ...subDoc.metadata, [this.idKey]: parentDocId },

package/dist/retrievers/parent_document.d.ts

@@ -1,6 +1,6 @@
 import { type VectorStoreInterface, type VectorStoreRetrieverInterface } from "@langchain/core/vectorstores";
 import { Document } from "@langchain/core/documents";
-import { TextSplitter } from "../text_splitter.js";
+import { TextSplitter, TextSplitterChunkHeaderOptions } from "../text_splitter.js";
 import { MultiVectorRetriever, type MultiVectorRetrieverInput } from "./multi_vector.js";
 /**
  * Interface for the fields required to initialize a
@@ -69,9 +69,11 @@ export declare class ParentDocumentRetriever extends MultiVectorRetriever {
      * This can be false if and only if `ids` are provided. You may want
      *   to set this to False if the documents are already in the docstore
      *   and you don't want to re-add them.
+     * @param config.chunkHeaderOptions Object with options for adding Contextual chunk headers
      */
     addDocuments(docs: Document[], config?: {
         ids?: string[];
         addToDocstore?: boolean;
+        childDocChunkHeaderOptions?: TextSplitterChunkHeaderOptions;
     }): Promise<void>;
 }

package/dist/retrievers/parent_document.js

@@ -127,9 +127,10 @@ export class ParentDocumentRetriever extends MultiVectorRetriever {
      * This can be false if and only if `ids` are provided. You may want
      *   to set this to False if the documents are already in the docstore
      *   and you don't want to re-add them.
+     * @param config.chunkHeaderOptions Object with options for adding Contextual chunk headers
      */
     async addDocuments(docs, config) {
-        const { ids, addToDocstore = true } = config ?? {};
+        const { ids, addToDocstore = true, childDocChunkHeaderOptions = {}, } = config ?? {};
         const parentDocs = this.parentSplitter
             ? await this.parentSplitter.splitDocuments(docs)
             : docs;
@@ -151,7 +152,7 @@ export class ParentDocumentRetriever extends MultiVectorRetriever {
         for (let i = 0; i < parentDocs.length; i += 1) {
             const parentDoc = parentDocs[i];
             const parentDocId = parentDocIds[i];
-            const subDocs = await this.childSplitter.splitDocuments([parentDoc]);
+            const subDocs = await this.childSplitter.splitDocuments([parentDoc], childDocChunkHeaderOptions);
             const taggedSubDocs = subDocs.map((subDoc) => new Document({
                 pageContent: subDoc.pageContent,
                 metadata: { ...subDoc.metadata, [this.idKey]: parentDocId },

package/dist/util/sql_utils.cjs

@@ -83,7 +83,8 @@ const getTableAndColumnsName = async (appDataSource) => {
         const rep = await appDataSource.query(sql);
         return formatToSqlTable(rep);
     }
-    if (appDataSource.options.type === "mysql") {
+    if (appDataSource.options.type === "mysql" ||
+        appDataSource.options.type === "aurora-mysql") {
         sql =
             "SELECT " +
                 "TABLE_NAME AS table_name, " +

package/dist/util/sql_utils.js

@@ -77,7 +77,8 @@ export const getTableAndColumnsName = async (appDataSource) => {
         const rep = await appDataSource.query(sql);
         return formatToSqlTable(rep);
     }
-    if (appDataSource.options.type === "mysql") {
+    if (appDataSource.options.type === "mysql" ||
+        appDataSource.options.type === "aurora-mysql") {
         sql =
             "SELECT " +
                 "TABLE_NAME AS table_name, " +

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "langchain",
-  "version": "0.1.25",
+  "version": "0.1.27",
   "description": "Typescript bindings for langchain",
   "type": "module",
   "engines": {
@@ -1328,7 +1328,7 @@
     "couchbase": "^4.2.10",
     "d3-dsv": "^2.0.0",
     "epub2": "^3.0.1",
-    "fast-xml-parser": "^4.2.7",
+    "fast-xml-parser": "*",
     "google-auth-library": "^8.9.0",
     "handlebars": "^4.7.8",
     "html-to-text": "^9.0.5",
@@ -1512,9 +1512,9 @@
   },
   "dependencies": {
     "@anthropic-ai/sdk": "^0.9.1",
-    "@langchain/community": "~0.0.33",
-    "@langchain/core": "~0.1.36",
-    "@langchain/openai": "~0.0.14",
+    "@langchain/community": "~0.0.36",
+    "@langchain/core": "~0.1.44",
+    "@langchain/openai": "~0.0.19",
     "binary-extensions": "^2.2.0",
     "expr-eval": "^2.0.2",
     "js-tiktoken": "^1.0.7",

package/dist/util/bedrock.cjs

@@ -1,75 +0,0 @@
-"use strict";
-Object.defineProperty(exports, "__esModule", { value: true });
-exports.BedrockLLMInputOutputAdapter = void 0;
-/**
- * A helper class used within the `Bedrock` class. It is responsible for
- * preparing the input and output for the Bedrock service. It formats the
- * input prompt based on the provider (e.g., "anthropic", "ai21",
- * "amazon") and extracts the generated text from the service response.
- */
-class BedrockLLMInputOutputAdapter {
-    /** Adapter class to prepare the inputs from Langchain to a format
-    that LLM model expects. Also, provides a helper function to extract
-    the generated text from the model response. */
-    static prepareInput(provider, prompt, maxTokens = 50, temperature = 0, stopSequences = undefined, modelKwargs = {}, bedrockMethod = "invoke") {
-        const inputBody = {};
-        if (provider === "anthropic") {
-            inputBody.prompt = prompt;
-            inputBody.max_tokens_to_sample = maxTokens;
-            inputBody.temperature = temperature;
-            inputBody.stop_sequences = stopSequences;
-        }
-        else if (provider === "ai21") {
-            inputBody.prompt = prompt;
-            inputBody.maxTokens = maxTokens;
-            inputBody.temperature = temperature;
-            inputBody.stopSequences = stopSequences;
-        }
-        else if (provider === "meta") {
-            inputBody.prompt = prompt;
-            inputBody.max_gen_len = maxTokens;
-            inputBody.temperature = temperature;
-        }
-        else if (provider === "amazon") {
-            inputBody.inputText = prompt;
-            inputBody.textGenerationConfig = {
-                maxTokenCount: maxTokens,
-                temperature,
-            };
-        }
-        else if (provider === "cohere") {
-            inputBody.prompt = prompt;
-            inputBody.max_tokens = maxTokens;
-            inputBody.temperature = temperature;
-            inputBody.stop_sequences = stopSequences;
-            if (bedrockMethod === "invoke-with-response-stream") {
-                inputBody.stream = true;
-            }
-        }
-        return { ...inputBody, ...modelKwargs };
-    }
-    /**
-     * Extracts the generated text from the service response.
-     * @param provider The provider name.
-     * @param responseBody The response body from the service.
-     * @returns The generated text.
-     */
-    // eslint-disable-next-line @typescript-eslint/no-explicit-any
-    static prepareOutput(provider, responseBody) {
-        if (provider === "anthropic") {
-            return responseBody.completion;
-        }
-        else if (provider === "ai21") {
-            return responseBody?.completions?.[0]?.data?.text ?? "";
-        }
-        else if (provider === "cohere") {
-            return responseBody?.generations?.[0]?.text ?? responseBody?.text ?? "";
-        }
-        else if (provider === "meta") {
-            return responseBody.generation;
-        }
-        // I haven't been able to get a response with more than one result in it.
-        return responseBody.results?.[0]?.outputText;
-    }
-}
-exports.BedrockLLMInputOutputAdapter = BedrockLLMInputOutputAdapter;

package/dist/util/bedrock.d.ts

@@ -1,64 +0,0 @@
-import type { AwsCredentialIdentity, Provider } from "@aws-sdk/types";
-export type CredentialType = AwsCredentialIdentity | Provider<AwsCredentialIdentity>;
-/** Bedrock models.
-    To authenticate, the AWS client uses the following methods to automatically load credentials:
-    https://boto3.amazonaws.com/v1/documentation/api/latest/guide/credentials.html
-    If a specific credential profile should be used, you must pass the name of the profile from the ~/.aws/credentials file that is to be used.
-    Make sure the credentials / roles used have the required policies to access the Bedrock service.
-*/
-export interface BaseBedrockInput {
-    /** Model to use.
-        For example, "amazon.titan-tg1-large", this is equivalent to the modelId property in the list-foundation-models api.
-    */
-    model: string;
-    /** The AWS region e.g. `us-west-2`.
-        Fallback to AWS_DEFAULT_REGION env variable or region specified in ~/.aws/config in case it is not provided here.
-    */
-    region?: string;
-    /** AWS Credentials.
-        If no credentials are provided, the default credentials from `@aws-sdk/credential-provider-node` will be used.
-     */
-    credentials?: CredentialType;
-    /** Temperature. */
-    temperature?: number;
-    /** Max tokens. */
-    maxTokens?: number;
-    /** A custom fetch function for low-level access to AWS API. Defaults to fetch(). */
-    fetchFn?: typeof fetch;
-    /** @deprecated Use endpointHost instead Override the default endpoint url. */
-    endpointUrl?: string;
-    /** Override the default endpoint hostname. */
-    endpointHost?: string;
-    /**
-     * Optional additional stop sequences to pass to the model. Currently only supported for Anthropic and AI21.
-     * @deprecated Use .bind({ "stop": [...] }) instead
-     * */
-    stopSequences?: string[];
-    /** Additional kwargs to pass to the model. */
-    modelKwargs?: Record<string, unknown>;
-    /** Whether or not to stream responses */
-    streaming: boolean;
-}
-type Dict = {
-    [key: string]: unknown;
-};
-/**
- * A helper class used within the `Bedrock` class. It is responsible for
- * preparing the input and output for the Bedrock service. It formats the
- * input prompt based on the provider (e.g., "anthropic", "ai21",
- * "amazon") and extracts the generated text from the service response.
- */
-export declare class BedrockLLMInputOutputAdapter {
-    /** Adapter class to prepare the inputs from Langchain to a format
-    that LLM model expects. Also, provides a helper function to extract
-    the generated text from the model response. */
-    static prepareInput(provider: string, prompt: string, maxTokens?: number, temperature?: number, stopSequences?: string[] | undefined, modelKwargs?: Record<string, unknown>, bedrockMethod?: "invoke" | "invoke-with-response-stream"): Dict;
-    /**
-     * Extracts the generated text from the service response.
-     * @param provider The provider name.
-     * @param responseBody The response body from the service.
-     * @returns The generated text.
-     */
-    static prepareOutput(provider: string, responseBody: any): string;
-}
-export {};

package/dist/util/bedrock.js

@@ -1,71 +0,0 @@
-/**
- * A helper class used within the `Bedrock` class. It is responsible for
- * preparing the input and output for the Bedrock service. It formats the
- * input prompt based on the provider (e.g., "anthropic", "ai21",
- * "amazon") and extracts the generated text from the service response.
- */
-export class BedrockLLMInputOutputAdapter {
-    /** Adapter class to prepare the inputs from Langchain to a format
-    that LLM model expects. Also, provides a helper function to extract
-    the generated text from the model response. */
-    static prepareInput(provider, prompt, maxTokens = 50, temperature = 0, stopSequences = undefined, modelKwargs = {}, bedrockMethod = "invoke") {
-        const inputBody = {};
-        if (provider === "anthropic") {
-            inputBody.prompt = prompt;
-            inputBody.max_tokens_to_sample = maxTokens;
-            inputBody.temperature = temperature;
-            inputBody.stop_sequences = stopSequences;
-        }
-        else if (provider === "ai21") {
-            inputBody.prompt = prompt;
-            inputBody.maxTokens = maxTokens;
-            inputBody.temperature = temperature;
-            inputBody.stopSequences = stopSequences;
-        }
-        else if (provider === "meta") {
-            inputBody.prompt = prompt;
-            inputBody.max_gen_len = maxTokens;
-            inputBody.temperature = temperature;
-        }
-        else if (provider === "amazon") {
-            inputBody.inputText = prompt;
-            inputBody.textGenerationConfig = {
-                maxTokenCount: maxTokens,
-                temperature,
-            };
-        }
-        else if (provider === "cohere") {
-            inputBody.prompt = prompt;
-            inputBody.max_tokens = maxTokens;
-            inputBody.temperature = temperature;
-            inputBody.stop_sequences = stopSequences;
-            if (bedrockMethod === "invoke-with-response-stream") {
-                inputBody.stream = true;
-            }
-        }
-        return { ...inputBody, ...modelKwargs };
-    }
-    /**
-     * Extracts the generated text from the service response.
-     * @param provider The provider name.
-     * @param responseBody The response body from the service.
-     * @returns The generated text.
-     */
-    // eslint-disable-next-line @typescript-eslint/no-explicit-any
-    static prepareOutput(provider, responseBody) {
-        if (provider === "anthropic") {
-            return responseBody.completion;
-        }
-        else if (provider === "ai21") {
-            return responseBody?.completions?.[0]?.data?.text ?? "";
-        }
-        else if (provider === "cohere") {
-            return responseBody?.generations?.[0]?.text ?? responseBody?.text ?? "";
-        }
-        else if (provider === "meta") {
-            return responseBody.generation;
-        }
-        // I haven't been able to get a response with more than one result in it.
-        return responseBody.results?.[0]?.outputText;
-    }
-}