langchain 0.1.26 → 0.1.27

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/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/package.json

@@ -1,6 +1,6 @@
 {
   "name": "langchain",
-  "version": "0.1.26",
+  "version": "0.1.27",
   "description": "Typescript bindings for langchain",
   "type": "module",
   "engines": {