first commit

packages/@n8n/nodes-langchain/nodes/document_loaders/DocumentBinaryInputLoader/DocumentBinaryInputLoader.node.ts

@@ -0,0 +1,200 @@
+import type { TextSplitter } from '@langchain/textsplitters';
+import {
+	NodeConnectionTypes,
+	type INodeType,
+	type INodeTypeDescription,
+	type ISupplyDataFunctions,
+	type SupplyData,
+} from 'n8n-workflow';
+
+import {
+	logWrapper,
+	N8nBinaryLoader,
+	getConnectionHintNoticeField,
+	metadataFilterField,
+} from '@n8n/ai-utilities';
+
+// Dependencies needed underneath the hood for the loaders. We add them
+// here only to track where what dependency is sued
+// import 'd3-dsv'; // for csv
+import 'mammoth'; // for docx
+import 'epub2'; // for epub
+import 'pdf-parse'; // for pdf
+
+export class DocumentBinaryInputLoader implements INodeType {
+	description: INodeTypeDescription = {
+		// This node is deprecated and will be removed in the future.
+		// The functionality was merged with the `DocumentJSONInputLoader` to `DocumentDefaultDataLoader`
+		hidden: true,
+		displayName: 'Binary Input Loader',
+		name: 'documentBinaryInputLoader',
+		icon: 'file:binary.svg',
+		group: ['transform'],
+		version: 1,
+		description: 'Use binary data from a previous step in the workflow',
+		defaults: {
+			name: 'Binary Input Loader',
+		},
+		codex: {
+			categories: ['AI'],
+			subcategories: {
+				AI: ['Document Loaders'],
+			},
+			resources: {
+				primaryDocumentation: [
+					{
+						url: 'https://docs.n8n.io/integrations/builtin/cluster-nodes/sub-nodes/n8n-nodes-langchain.documentdefaultdataloader/',
+					},
+				],
+			},
+		},
+
+		inputs: [
+			{
+				displayName: 'Text Splitter',
+				maxConnections: 1,
+				type: NodeConnectionTypes.AiTextSplitter,
+				required: true,
+			},
+		],
+
+		outputs: [NodeConnectionTypes.AiDocument],
+		outputNames: ['Document'],
+		builderHint: {
+			inputs: {
+				ai_textSplitter: { required: true },
+			},
+		},
+		properties: [
+			getConnectionHintNoticeField([NodeConnectionTypes.AiVectorStore]),
+			{
+				displayName: 'Loader Type',
+				name: 'loader',
+				type: 'options',
+				default: 'jsonLoader',
+				required: true,
+				options: [
+					{
+						name: 'CSV Loader',
+						value: 'csvLoader',
+						description: 'Load CSV files',
+					},
+					{
+						name: 'Docx Loader',
+						value: 'docxLoader',
+						description: 'Load Docx documents',
+					},
+					{
+						name: 'EPub Loader',
+						value: 'epubLoader',
+						description: 'Load EPub files',
+					},
+					{
+						name: 'JSON Loader',
+						value: 'jsonLoader',
+						description: 'Load JSON files',
+					},
+					{
+						name: 'PDF Loader',
+						value: 'pdfLoader',
+						description: 'Load PDF documents',
+					},
+					{
+						name: 'Text Loader',
+						value: 'textLoader',
+						description: 'Load plain text files',
+					},
+				],
+			},
+			{
+				displayName: 'Binary Data Key',
+				name: 'binaryDataKey',
+				type: 'string',
+				default: 'data',
+				required: true,
+				description: 'Name of the binary property from which to read the file buffer',
+			},
+			// PDF Only Fields
+			{
+				displayName: 'Split Pages',
+				name: 'splitPages',
+				type: 'boolean',
+				default: true,
+				displayOptions: {
+					show: {
+						loader: ['pdfLoader'],
+					},
+				},
+			},
+			// CSV Only Fields
+			{
+				displayName: 'Column',
+				name: 'column',
+				type: 'string',
+				default: '',
+				description: 'Column to extract from CSV',
+				displayOptions: {
+					show: {
+						loader: ['csvLoader'],
+					},
+				},
+			},
+			{
+				displayName: 'Separator',
+				name: 'separator',
+				type: 'string',
+				description: 'Separator to use for CSV',
+				default: ',',
+				displayOptions: {
+					show: {
+						loader: ['csvLoader'],
+					},
+				},
+			},
+			// JSON Only Fields
+			{
+				displayName: 'Pointers',
+				name: 'pointers',
+				type: 'string',
+				default: '',
+				description: 'Pointers to extract from JSON, e.g. "/text" or "/text, /meta/title"',
+				displayOptions: {
+					show: {
+						loader: ['jsonLoader'],
+					},
+				},
+			},
+			{
+				displayName: 'Options',
+				name: 'options',
+				type: 'collection',
+				placeholder: 'Add Option',
+				default: {},
+				options: [
+					{
+						...metadataFilterField,
+						displayName: 'Metadata',
+						description:
+							'Metadata to add to each document. Could be used for filtering during retrieval',
+						placeholder: 'Add property',
+					},
+				],
+			},
+		],
+	};
+
+	async supplyData(this: ISupplyDataFunctions): Promise<SupplyData> {
+		this.logger.debug('Supply Data for Binary Input Loader');
+		const textSplitter = (await this.getInputConnectionData(
+			NodeConnectionTypes.AiTextSplitter,
+			0,
+		)) as TextSplitter | undefined;
+
+		const binaryDataKey = this.getNodeParameter('binaryDataKey', 0) as string;
+		const processor = new N8nBinaryLoader(this, undefined, binaryDataKey, textSplitter);
+
+		return {
+			response: logWrapper(processor, this),
+		};
+	}
+}

packages/@n8n/nodes-langchain/nodes/document_loaders/DocumentBinaryInputLoader/binary.svg

@@ -0,0 +1 @@
+<svg xmlns="http://www.w3.org/2000/svg" width="768" height="1024"><path fill="#7D7D87" d="M0 960V64h576l192 192v704zm704-640L512 128H64v768h640zM320 512H128V256h192zm-64-192h-64v128h64zm0 448h64v64H128v-64h64V640h-64v-64h128zm256-320h64v64H384v-64h64V320h-64v-64h128zm64 384H384V576h192zm-64-192h-64v128h64z"/></svg>