langchain 0.0.196 → 0.0.197-rc.0

package/dist/chat_models/anthropic.js

@@ -1,346 +1 @@
-import { Anthropic, AI_PROMPT, HUMAN_PROMPT, } from "@anthropic-ai/sdk";
-import { AIMessage, AIMessageChunk, ChatGenerationChunk, ChatMessage, } from "../schema/index.js";
-import { getEnvironmentVariable } from "../util/env.js";
-import { BaseChatModel } from "./base.js";
-/**
- * Extracts the custom role of a generic chat message.
- * @param message The chat message from which to extract the custom role.
- * @returns The custom role of the chat message.
- */
-function extractGenericMessageCustomRole(message) {
-    if (message.role !== AI_PROMPT &&
-        message.role !== HUMAN_PROMPT &&
-        message.role !== "") {
-        console.warn(`Unknown message role: ${message.role}`);
-    }
-    return message.role;
-}
-/**
- * Gets the Anthropic prompt from a base message.
- * @param message The base message from which to get the Anthropic prompt.
- * @returns The Anthropic prompt from the base message.
- */
-function getAnthropicPromptFromMessage(message) {
-    const type = message._getType();
-    switch (type) {
-        case "ai":
-            return AI_PROMPT;
-        case "human":
-            return HUMAN_PROMPT;
-        case "system":
-            return "";
-        case "generic": {
-            if (!ChatMessage.isInstance(message))
-                throw new Error("Invalid generic chat message");
-            return extractGenericMessageCustomRole(message);
-        }
-        default:
-            throw new Error(`Unknown message type: ${type}`);
-    }
-}
-export const DEFAULT_STOP_SEQUENCES = [HUMAN_PROMPT];
-/**
- * Wrapper around Anthropic large language models.
- *
- * To use you should have the `@anthropic-ai/sdk` package installed, with the
- * `ANTHROPIC_API_KEY` environment variable set.
- *
- * @remarks
- * Any parameters that are valid to be passed to {@link
- * https://console.anthropic.com/docs/api/reference |
- * `anthropic.complete`} can be passed through {@link invocationKwargs},
- * even if not explicitly available on this class.
- * @example
- * ```typescript
- * const model = new ChatAnthropic({
- *   temperature: 0.9,
- *   anthropicApiKey: 'YOUR-API-KEY',
- * });
- * const res = await model.invoke({ input: 'Hello!' });
- * console.log(res);
- * ```
- */
-export class ChatAnthropic extends BaseChatModel {
-    static lc_name() {
-        return "ChatAnthropic";
-    }
-    get lc_secrets() {
-        return {
-            anthropicApiKey: "ANTHROPIC_API_KEY",
-        };
-    }
-    get lc_aliases() {
-        return {
-            modelName: "model",
-        };
-    }
-    constructor(fields) {
-        super(fields ?? {});
-        Object.defineProperty(this, "lc_serializable", {
-            enumerable: true,
-            configurable: true,
-            writable: true,
-            value: true
-        });
-        Object.defineProperty(this, "anthropicApiKey", {
-            enumerable: true,
-            configurable: true,
-            writable: true,
-            value: void 0
-        });
-        Object.defineProperty(this, "apiUrl", {
-            enumerable: true,
-            configurable: true,
-            writable: true,
-            value: void 0
-        });
-        Object.defineProperty(this, "temperature", {
-            enumerable: true,
-            configurable: true,
-            writable: true,
-            value: 1
-        });
-        Object.defineProperty(this, "topK", {
-            enumerable: true,
-            configurable: true,
-            writable: true,
-            value: -1
-        });
-        Object.defineProperty(this, "topP", {
-            enumerable: true,
-            configurable: true,
-            writable: true,
-            value: -1
-        });
-        Object.defineProperty(this, "maxTokensToSample", {
-            enumerable: true,
-            configurable: true,
-            writable: true,
-            value: 2048
-        });
-        Object.defineProperty(this, "modelName", {
-            enumerable: true,
-            configurable: true,
-            writable: true,
-            value: "claude-2"
-        });
-        Object.defineProperty(this, "invocationKwargs", {
-            enumerable: true,
-            configurable: true,
-            writable: true,
-            value: void 0
-        });
-        Object.defineProperty(this, "stopSequences", {
-            enumerable: true,
-            configurable: true,
-            writable: true,
-            value: void 0
-        });
-        Object.defineProperty(this, "streaming", {
-            enumerable: true,
-            configurable: true,
-            writable: true,
-            value: false
-        });
-        Object.defineProperty(this, "clientOptions", {
-            enumerable: true,
-            configurable: true,
-            writable: true,
-            value: void 0
-        });
-        // Used for non-streaming requests
-        Object.defineProperty(this, "batchClient", {
-            enumerable: true,
-            configurable: true,
-            writable: true,
-            value: void 0
-        });
-        // Used for streaming requests
-        Object.defineProperty(this, "streamingClient", {
-            enumerable: true,
-            configurable: true,
-            writable: true,
-            value: void 0
-        });
-        this.anthropicApiKey =
-            fields?.anthropicApiKey ?? getEnvironmentVariable("ANTHROPIC_API_KEY");
-        if (!this.anthropicApiKey) {
-            throw new Error("Anthropic API key not found");
-        }
-        // Support overriding the default API URL (i.e., https://api.anthropic.com)
-        this.apiUrl = fields?.anthropicApiUrl;
-        this.modelName = fields?.modelName ?? this.modelName;
-        this.invocationKwargs = fields?.invocationKwargs ?? {};
-        this.temperature = fields?.temperature ?? this.temperature;
-        this.topK = fields?.topK ?? this.topK;
-        this.topP = fields?.topP ?? this.topP;
-        this.maxTokensToSample =
-            fields?.maxTokensToSample ?? this.maxTokensToSample;
-        this.stopSequences = fields?.stopSequences ?? this.stopSequences;
-        this.streaming = fields?.streaming ?? false;
-        this.clientOptions = fields?.clientOptions ?? {};
-    }
-    /**
-     * Get the parameters used to invoke the model
-     */
-    invocationParams(options) {
-        return {
-            model: this.modelName,
-            temperature: this.temperature,
-            top_k: this.topK,
-            top_p: this.topP,
-            stop_sequences: options?.stop?.concat(DEFAULT_STOP_SEQUENCES) ??
-                this.stopSequences ??
-                DEFAULT_STOP_SEQUENCES,
-            max_tokens_to_sample: this.maxTokensToSample,
-            stream: this.streaming,
-            ...this.invocationKwargs,
-        };
-    }
-    /** @ignore */
-    _identifyingParams() {
-        return {
-            model_name: this.modelName,
-            ...this.invocationParams(),
-        };
-    }
-    /**
-     * Get the identifying parameters for the model
-     */
-    identifyingParams() {
-        return {
-            model_name: this.modelName,
-            ...this.invocationParams(),
-        };
-    }
-    async *_streamResponseChunks(messages, options, runManager) {
-        const params = this.invocationParams(options);
-        const stream = await this.createStreamWithRetry({
-            ...params,
-            prompt: this.formatMessagesAsPrompt(messages),
-        });
-        let modelSent = false;
-        let stopReasonSent = false;
-        for await (const data of stream) {
-            if (options.signal?.aborted) {
-                stream.controller.abort();
-                throw new Error("AbortError: User aborted the request.");
-            }
-            const additional_kwargs = {};
-            if (data.model && !modelSent) {
-                additional_kwargs.model = data.model;
-                modelSent = true;
-            }
-            else if (data.stop_reason && !stopReasonSent) {
-                additional_kwargs.stop_reason = data.stop_reason;
-                stopReasonSent = true;
-            }
-            const delta = data.completion ?? "";
-            yield new ChatGenerationChunk({
-                message: new AIMessageChunk({
-                    content: delta,
-                    additional_kwargs,
-                }),
-                text: delta,
-            });
-            await runManager?.handleLLMNewToken(delta);
-            if (data.stop_reason) {
-                break;
-            }
-        }
-    }
-    /**
-     * Formats messages as a prompt for the model.
-     * @param messages The base messages to format as a prompt.
-     * @returns The formatted prompt.
-     */
-    formatMessagesAsPrompt(messages) {
-        return (messages
-            .map((message) => {
-            const messagePrompt = getAnthropicPromptFromMessage(message);
-            return `${messagePrompt} ${message.content}`;
-        })
-            .join("") + AI_PROMPT);
-    }
-    /** @ignore */
-    async _generate(messages, options, runManager) {
-        if (this.stopSequences && options.stop) {
-            throw new Error(`"stopSequence" parameter found in input and default params`);
-        }
-        const params = this.invocationParams(options);
-        let response;
-        if (params.stream) {
-            response = {
-                completion: "",
-                model: "",
-                stop_reason: "",
-            };
-            const stream = await this._streamResponseChunks(messages, options, runManager);
-            for await (const chunk of stream) {
-                response.completion += chunk.message.content;
-                response.model =
-                    chunk.message.additional_kwargs.model ?? response.model;
-                response.stop_reason =
-                    chunk.message.additional_kwargs.stop_reason ??
-                        response.stop_reason;
-            }
-        }
-        else {
-            response = await this.completionWithRetry({
-                ...params,
-                prompt: this.formatMessagesAsPrompt(messages),
-            }, { signal: options.signal });
-        }
-        const generations = (response.completion ?? "")
-            .split(AI_PROMPT)
-            .map((message) => ({
-            text: message,
-            message: new AIMessage(message),
-        }));
-        return {
-            generations,
-        };
-    }
-    /**
-     * Creates a streaming request with retry.
-     * @param request The parameters for creating a completion.
-     * @returns A streaming request.
-     */
-    async createStreamWithRetry(request) {
-        if (!this.streamingClient) {
-            const options = this.apiUrl ? { baseURL: this.apiUrl } : undefined;
-            this.streamingClient = new Anthropic({
-                ...this.clientOptions,
-                ...options,
-                apiKey: this.anthropicApiKey,
-                maxRetries: 0,
-            });
-        }
-        const makeCompletionRequest = async () => this.streamingClient.completions.create({ ...request, stream: true }, { headers: request.headers });
-        return this.caller.call(makeCompletionRequest);
-    }
-    /** @ignore */
-    async completionWithRetry(request, options) {
-        if (!this.anthropicApiKey) {
-            throw new Error("Missing Anthropic API key.");
-        }
-        if (!this.batchClient) {
-            const options = this.apiUrl ? { baseURL: this.apiUrl } : undefined;
-            this.batchClient = new Anthropic({
-                ...this.clientOptions,
-                ...options,
-                apiKey: this.anthropicApiKey,
-                maxRetries: 0,
-            });
-        }
-        const makeCompletionRequest = async () => this.batchClient.completions.create({ ...request, stream: false }, { headers: request.headers });
-        return this.caller.callWithOptions({ signal: options.signal }, makeCompletionRequest);
-    }
-    _llmType() {
-        return "anthropic";
-    }
-    /** @ignore */
-    _combineLLMOutput() {
-        return [];
-    }
-}
+export * from "@langchain/anthropic";

package/dist/chat_models/base.cjs

@@ -14,4 +14,4 @@ var __exportStar = (this && this.__exportStar) || function(m, exports) {
     for (var p in m) if (p !== "default" && !Object.prototype.hasOwnProperty.call(exports, p)) __createBinding(exports, m, p);
 };
 Object.defineProperty(exports, "__esModule", { value: true });
-__exportStar(require("langchain-core/chat_model"), exports);
+__exportStar(require("langchain-core/language_models/chat_models"), exports);

package/dist/chat_models/base.d.ts

@@ -1 +1 @@
-export * from "langchain-core/chat_model";
+export * from "langchain-core/language_models/chat_models";

package/dist/chat_models/base.js

@@ -1 +1 @@
-export * from "langchain-core/chat_model";
+export * from "langchain-core/language_models/chat_models";

package/dist/chat_models/bedrock/web.cjs

@@ -321,11 +321,31 @@ class BedrockChat extends base_js_1.SimpleChatModel {
     }
     // eslint-disable-next-line @typescript-eslint/no-explicit-any
     _readChunks(reader) {
+        function _concatChunks(a, b) {
+            const newBuffer = new Uint8Array(a.length + b.length);
+            newBuffer.set(a);
+            newBuffer.set(b, a.length);
+            return newBuffer;
+        }
+        function getMessageLength(buffer) {
+            if (buffer.byteLength === 0)
+                return 0;
+            const view = new DataView(buffer.buffer, buffer.byteOffset, buffer.byteLength);
+            return view.getUint32(0, false);
+        }
         return {
             async *[Symbol.asyncIterator]() {
                 let readResult = await reader.read();
+                let buffer = new Uint8Array(0);
                 while (!readResult.done) {
-                    yield readResult.value;
+                    const chunk = readResult.value;
+                    buffer = _concatChunks(buffer, chunk);
+                    let messageLength = getMessageLength(buffer);
+                    while (buffer.byteLength > 0 && buffer.byteLength >= messageLength) {
+                        yield buffer.slice(0, messageLength);
+                        buffer = buffer.slice(messageLength);
+                        messageLength = getMessageLength(buffer);
+                    }
                     readResult = await reader.read();
                 }
             },

package/dist/chat_models/bedrock/web.d.ts

@@ -72,7 +72,7 @@ export declare class BedrockChat extends SimpleChatModel implements BaseBedrockI
     }): Promise<Response>;
     _streamResponseChunks(messages: BaseMessage[], options: this["ParsedCallOptions"], runManager?: CallbackManagerForLLMRun): AsyncGenerator<ChatGenerationChunk>;
     _readChunks(reader: any): {
-        [Symbol.asyncIterator](): AsyncGenerator<any, void, unknown>;
+        [Symbol.asyncIterator](): AsyncGenerator<Uint8Array, void, unknown>;
     };
     _combineLLMOutput(): {};
 }

package/dist/chat_models/bedrock/web.js

@@ -316,11 +316,31 @@ export class BedrockChat extends SimpleChatModel {
     }
     // eslint-disable-next-line @typescript-eslint/no-explicit-any
     _readChunks(reader) {
+        function _concatChunks(a, b) {
+            const newBuffer = new Uint8Array(a.length + b.length);
+            newBuffer.set(a);
+            newBuffer.set(b, a.length);
+            return newBuffer;
+        }
+        function getMessageLength(buffer) {
+            if (buffer.byteLength === 0)
+                return 0;
+            const view = new DataView(buffer.buffer, buffer.byteOffset, buffer.byteLength);
+            return view.getUint32(0, false);
+        }
         return {
             async *[Symbol.asyncIterator]() {
                 let readResult = await reader.read();
+                let buffer = new Uint8Array(0);
                 while (!readResult.done) {
-                    yield readResult.value;
+                    const chunk = readResult.value;
+                    buffer = _concatChunks(buffer, chunk);
+                    let messageLength = getMessageLength(buffer);
+                    while (buffer.byteLength > 0 && buffer.byteLength >= messageLength) {
+                        yield buffer.slice(0, messageLength);
+                        buffer = buffer.slice(messageLength);
+                        messageLength = getMessageLength(buffer);
+                    }
                     readResult = await reader.read();
                 }
             },

package/dist/document.cjs

@@ -1,5 +1,5 @@
 "use strict";
 Object.defineProperty(exports, "__esModule", { value: true });
 exports.Document = void 0;
-var document_1 = require("langchain-core/schema/document");
-Object.defineProperty(exports, "Document", { enumerable: true, get: function () { return document_1.Document; } });
+var documents_1 = require("langchain-core/documents");
+Object.defineProperty(exports, "Document", { enumerable: true, get: function () { return documents_1.Document; } });

package/dist/document.d.ts

@@ -1 +1 @@
-export { type DocumentInput, Document } from "langchain-core/schema/document";
+export { type DocumentInput, Document } from "langchain-core/documents";

package/dist/document.js

@@ -1 +1 @@
-export { Document } from "langchain-core/schema/document";
+export { Document } from "langchain-core/documents";

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

@@ -47,6 +47,6 @@ export declare class AzureBlobStorageFileLoader extends BaseDocumentLoader {
      * are returned, and the temporary directory is deleted.
      * @returns An array of documents loaded from the file in Azure Blob Storage.
      */
-    load(): Promise<import("langchain-core/schema/document").Document<Record<string, any>>[]>;
+    load(): Promise<import("langchain-core/documents").Document<Record<string, any>>[]>;
 }
 export {};

package/dist/document_loaders/web/github.cjs

@@ -216,6 +216,22 @@ class GithubRepoLoader extends base_js_1.BaseDocumentLoader {
         }
         return documents;
     }
+    /**
+     * Asynchronously streams documents from the entire GitHub repository.
+     * It is suitable for situations where processing large repositories in a memory-efficient manner is required.
+     * @yields Yields a Promise that resolves to a Document object for each file or submodule content found in the repository.
+     */
+    async *loadAsStream() {
+        this.log(`Loading documents from ${this.baseUrl}/${this.owner}/${this.repo}/${this.initialPath}...`);
+        yield* await this.processRepoAsStream(this.initialPath);
+        if (!this.processSubmodules) {
+            return;
+        }
+        await this.getSubmoduleInfo();
+        for (const submoduleInfo of this.submoduleInfos) {
+            yield* await this.loadSubmoduleAsStream(submoduleInfo);
+        }
+    }
     /**
      * Loads the information about Git submodules from the repository, if available.
      */
@@ -323,6 +339,37 @@ class GithubRepoLoader extends base_js_1.BaseDocumentLoader {
             }).load();
         }
     }
+    /**
+     * Asynchronously processes and streams the contents of a specified submodule in the GitHub repository.
+     * @param submoduleInfo the info about the submodule to be loaded
+     * @yields Yields a Promise that resolves to a Document object for each file found in the submodule.
+     */
+    async *loadSubmoduleAsStream(submoduleInfo) {
+        if (!submoduleInfo.url.startsWith(this.baseUrl)) {
+            this.log(`Ignoring external submodule ${submoduleInfo.url}.`);
+            yield* [];
+        }
+        if (!submoduleInfo.path.startsWith(this.initialPath)) {
+            this.log(`Ignoring submodule ${submoduleInfo.url}, as it is not on initial path.`);
+            yield* [];
+        }
+        this.log(`Accessing submodule ${submoduleInfo.name} (${submoduleInfo.url})...`);
+        const submoduleLoader = new GithubRepoLoader(submoduleInfo.url, {
+            accessToken: this.accessToken,
+            baseUrl: this.baseUrl,
+            apiUrl: this.apiUrl,
+            branch: submoduleInfo.ref,
+            recursive: this.recursive,
+            processSubmodules: this.processSubmodules,
+            unknown: this.unknown,
+            ignoreFiles: this.ignoreFiles,
+            ignorePaths: this.ignorePaths,
+            verbose: this.verbose,
+            maxConcurrency: this.maxConcurrency,
+            maxRetries: this.maxRetries,
+        });
+        yield* await submoduleLoader.processRepoAsStream(submoduleInfo.path);
+    }
     /**
      * Determines whether a file or directory should be ignored based on its
      * path and type.
@@ -414,6 +461,35 @@ class GithubRepoLoader extends base_js_1.BaseDocumentLoader {
             return Promise.reject(error);
         }
     }
+    /**
+     * Asynchronously processes the contents of the entire GitHub repository,
+     * streaming each file as a Document object.
+     * @param path The path of the directory to process.
+     * @yields Yields a Promise that resolves to a Document object for each file found in the repository.
+     */
+    async *processRepoAsStream(path) {
+        const files = await this.fetchRepoFiles(path);
+        for (const file of files) {
+            if (this.shouldIgnore(file.path, file.type)) {
+                continue;
+            }
+            if (file.type === "file") {
+                try {
+                    const fileResponse = await this.fetchFileContentWrapper(file);
+                    yield new document_js_1.Document({
+                        pageContent: fileResponse.contents,
+                        metadata: fileResponse.metadata,
+                    });
+                }
+                catch (error) {
+                    this.handleError(`Failed to fetch file content: ${file.path}, ${error}`);
+                }
+            }
+            else if (this.recursive) {
+                yield* await this.processDirectoryAsStream(file.path);
+            }
+        }
+    }
     /**
      * Fetches the contents of a directory and maps the file / directory paths
      * to promises that will fetch the file / directory contents.
@@ -430,6 +506,35 @@ class GithubRepoLoader extends base_js_1.BaseDocumentLoader {
             return Promise.reject(error);
         }
     }
+    /**
+     * Asynchronously processes the contents of a given directory in the GitHub repository,
+     * streaming each file as a Document object.
+     * @param path The path of the directory to process.
+     * @yields Yields a Promise that resolves to a Document object for each file in the directory.
+     */
+    async *processDirectoryAsStream(path) {
+        const files = await this.fetchRepoFiles(path);
+        for (const file of files) {
+            if (this.shouldIgnore(file.path, file.type)) {
+                continue;
+            }
+            if (file.type === "file") {
+                try {
+                    const fileResponse = await this.fetchFileContentWrapper(file);
+                    yield new document_js_1.Document({
+                        pageContent: fileResponse.contents,
+                        metadata: fileResponse.metadata,
+                    });
+                }
+                catch {
+                    this.handleError(`Failed to fetch file content: ${file.path}`);
+                }
+            }
+            else if (this.recursive) {
+                yield* await this.processDirectoryAsStream(file.path);
+            }
+        }
+    }
     /**
      * Fetches the files from a GitHub repository.
      * If the path denotes a single file, the resulting array contains only one element.

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

@@ -99,6 +99,12 @@ export declare class GithubRepoLoader extends BaseDocumentLoader implements Gith
      * @returns A promise that resolves to an array of Document instances.
      */
     load(): Promise<Document[]>;
+    /**
+     * Asynchronously streams documents from the entire GitHub repository.
+     * It is suitable for situations where processing large repositories in a memory-efficient manner is required.
+     * @yields Yields a Promise that resolves to a Document object for each file or submodule content found in the repository.
+     */
+    loadAsStream(): AsyncGenerator<Document, void, undefined>;
     /**
      * Loads the information about Git submodules from the repository, if available.
      */
@@ -115,6 +121,12 @@ export declare class GithubRepoLoader extends BaseDocumentLoader implements Gith
      * @param submoduleInfo the info about the submodule to be loaded
      */
     private loadSubmodule;
+    /**
+     * Asynchronously processes and streams the contents of a specified submodule in the GitHub repository.
+     * @param submoduleInfo the info about the submodule to be loaded
+     * @yields Yields a Promise that resolves to a Document object for each file found in the submodule.
+     */
+    private loadSubmoduleAsStream;
     /**
      * Determines whether a file or directory should be ignored based on its
      * path and type.
@@ -137,6 +149,13 @@ export declare class GithubRepoLoader extends BaseDocumentLoader implements Gith
      * Begins the process of fetching the contents of the repository
      */
     private processRepo;
+    /**
+     * Asynchronously processes the contents of the entire GitHub repository,
+     * streaming each file as a Document object.
+     * @param path The path of the directory to process.
+     * @yields Yields a Promise that resolves to a Document object for each file found in the repository.
+     */
+    private processRepoAsStream;
     /**
      * Fetches the contents of a directory and maps the file / directory paths
      * to promises that will fetch the file / directory contents.
@@ -144,6 +163,13 @@ export declare class GithubRepoLoader extends BaseDocumentLoader implements Gith
      * @returns A promise that resolves to an array of promises that will fetch the file / directory contents.
      */
     private processDirectory;
+    /**
+     * Asynchronously processes the contents of a given directory in the GitHub repository,
+     * streaming each file as a Document object.
+     * @param path The path of the directory to process.
+     * @yields Yields a Promise that resolves to a Document object for each file in the directory.
+     */
+    private processDirectoryAsStream;
     /**
      * Fetches the files from a GitHub repository.
      * If the path denotes a single file, the resulting array contains only one element.