WIP feat(Mistral OCR Node): New node

packages/nodes-base/credentials/MistralOcrApi.credentials.ts

@@ -0,0 +1,55 @@
+import type {
+	IAuthenticateGeneric,
+	ICredentialTestRequest,
+	ICredentialType,
+	INodeProperties,
+} from 'n8n-workflow';
+
+export class MistralOcrApi implements ICredentialType {
+	name = 'mistralOcrApi';
+
+	displayName = 'Mistral OCR';
+
+	documentationUrl = 'https://docs.mistral.ai/capabilities/document/';
+
+	properties: INodeProperties[] = [
+		{
+			displayName: 'API Key',
+			name: 'apiKey',
+			type: 'string',
+			typeOptions: { password: true },
+			required: true,
+			default: '',
+			description: 'Your Mistral OCR API key. Get it from your Mistral account.',
+		},
+		{
+			displayName: 'Base URL',
+			name: 'baseUrl',
+			type: 'string',
+			default: 'https://api.mistral.ai',
+			description: 'The base URL for the Mistral OCR API.',
+		},
+	];
+
+	authenticate: IAuthenticateGeneric = {
+		type: 'generic',
+		properties: {
+			headers: {
+				Authorization: '=Bearer {{$credentials.apiKey}}',
+			},
+		},
+	};
+
+	test: ICredentialTestRequest = {
+		request: {
+			baseURL: '={{$credentials.baseUrl}}',
+			url: '/v1/models',
+			method: 'GET',
+			headers: {
+				Authorization: '=Bearer {{$credentials.apiKey}}',
+				'Content-Type': 'application/json',
+			},
+			json: true,
+		},
+	};
+}

packages/nodes-base/nodes/MistralOcr/GenericFunctions.ts

@@ -0,0 +1,46 @@
+import type {
+	IExecuteSingleFunctions,
+	IHttpRequestOptions,
+	IN8nHttpFullResponse,
+	INodeExecutionData,
+	JsonObject,
+} from 'n8n-workflow';
+import { NodeApiError } from 'n8n-workflow';
+
+export async function sendErrorPostReceive(
+	this: IExecuteSingleFunctions,
+	_data: INodeExecutionData[],
+	response: IN8nHttpFullResponse,
+): Promise<INodeExecutionData[]> {
+	if (response.statusCode && response.statusCode >= 400) {
+		const errorBody = response.body as JsonObject;
+
+		if (response.statusCode === 422) {
+			const errors = (errorBody.detail as any[])?.map(
+				(err) => `${err.loc?.join('.') || 'field'}: ${err.msg || 'Invalid value'}`,
+			) || ['Invalid request parameters'];
+
+			throw new NodeApiError(this.getNode(), errorBody, {
+				message: 'Validation Error',
+				description: errors.join('\n'),
+			});
+		}
+
+		const message = (errorBody.message || response.statusMessage) as string;
+		throw new NodeApiError(this.getNode(), errorBody, {
+			message: message || 'Unknown error occurred',
+		});
+	}
+	return [];
+}
+
+// ToDo: Remove before completing the pull request
+export async function presendTest(
+	this: IExecuteSingleFunctions,
+
+	requestOptions: IHttpRequestOptions,
+): Promise<IHttpRequestOptions> {
+	console.log('requestOptions', requestOptions);
+
+	return requestOptions;
+}

packages/nodes-base/nodes/MistralOcr/MistralOcr.node.json

@@ -0,0 +1,18 @@
+{
+	"node": "n8n-nodes-base.mistralOcr",
+	"nodeVersion": "1.0",
+	"codexVersion": "1.0",
+	"categories": ["Utility"],
+	"resources": {
+		"credentialDocumentation": [
+			{
+				"url": "https://docs.n8n.io/integrations/builtin/credentials/mistralOcr/"
+			}
+		],
+		"primaryDocumentation": [
+			{
+				"url": "https://docs.n8n.io/integrations/builtin/app-nodes/n8n-nodes-langchain.mistralOcr/"
+			}
+		]
+	}
+}

packages/nodes-base/nodes/MistralOcr/MistralOcr.node.ts

@@ -0,0 +1,52 @@
+import type { INodeType, INodeTypeDescription } from 'n8n-workflow';
+import { NodeConnectionTypes } from 'n8n-workflow';
+
+import { mistralOcrOperations, mistralOcrFields } from './descriptions/MistralOcrDescription';
+
+export class MistralOcr implements INodeType {
+	description: INodeTypeDescription = {
+		displayName: 'Mistral OCR',
+		name: 'mistralOcr',
+		icon: {
+			light: 'file:mistralOcr.svg',
+			dark: 'file:mistralOcr.dark.svg',
+		},
+		group: ['transform'],
+		version: 1,
+		subtitle: '={{ $parameter["operation"] + ": " + $parameter["resource"] }}',
+		description: 'Extract text and layout information from documents using Mistral’s OCR API.',
+		defaults: {
+			name: 'Mistral OCR',
+		},
+		inputs: [NodeConnectionTypes.Main],
+		outputs: [NodeConnectionTypes.Main],
+		usableAsTool: true,
+		credentials: [
+			{
+				name: 'mistralOcrApi',
+				required: true,
+			},
+		],
+		requestDefaults: {
+			baseURL: '={{ $credentials.baseUrl }}',
+			ignoreHttpStatusErrors: true,
+		},
+		properties: [
+			{
+				displayName: 'Resource',
+				name: 'resource',
+				type: 'options',
+				noDataExpression: true,
+				options: [
+					{
+						name: 'Document',
+						value: 'document',
+					},
+				],
+				default: 'document',
+			},
+			...mistralOcrOperations,
+			...mistralOcrFields,
+		],
+	};
+}

packages/nodes-base/nodes/MistralOcr/descriptions/MistralOcrDescription.ts

@@ -0,0 +1,156 @@
+import type { INodeProperties } from 'n8n-workflow';
+
+import { presendTest, sendErrorPostReceive } from '../GenericFunctions';
+
+export const mistralOcrOperations: INodeProperties[] = [
+	{
+		displayName: 'Operation',
+		name: 'operation',
+		type: 'options',
+		noDataExpression: true,
+		displayOptions: {
+			show: {
+				resource: ['document'],
+			},
+		},
+		options: [
+			{
+				name: 'Extract Text',
+				value: 'extractText',
+				description: 'Extract text from documents using OCR',
+				action: 'Extract text',
+				routing: {
+					request: {
+						method: 'POST',
+						url: '/v1/ocr',
+						headers: {
+							'Content-Type': 'application/json',
+						},
+					},
+					output: {
+						postReceive: [sendErrorPostReceive],
+					},
+				},
+			},
+		],
+		default: 'extractText',
+	},
+];
+
+export const mistralOcrFields: INodeProperties[] = [
+	{
+		displayName: 'Model',
+		name: 'model',
+		type: 'options',
+		options: [{ name: 'mistral-ocr-latest', value: 'mistral-ocr-latest' }],
+		description: 'The OCR model to use',
+		required: true,
+		default: 'mistral-ocr-latest',
+		routing: {
+			send: {
+				type: 'body',
+				property: 'model',
+				preSend: [presendTest],
+			},
+		},
+	},
+	{
+		displayName: 'Input Type',
+		name: 'inputType',
+		type: 'options',
+		options: [
+			{ name: 'Binary Data', value: 'binary' },
+			{ name: 'URL', value: 'url' },
+		],
+		description: 'How the document will be provided',
+		required: true,
+		default: 'binary',
+		routing: {
+			send: {
+				type: 'body',
+				property: 'document.type',
+				value: '={{ $value === "binary" ? "image_data" : "document_url" }}',
+			},
+		},
+	},
+	{
+		displayName: 'Binary Property',
+		name: 'binaryProperty',
+		type: 'string',
+		description: 'Name of the binary property that contains the file to process',
+		placeholder: 'e.g. data',
+		hint: 'Uploaded document files must not exceed 50 MB in size and should be no longer than 1,000 pages.',
+		required: true,
+		default: 'data',
+		displayOptions: {
+			show: {
+				inputType: ['binary'],
+			},
+		},
+		routing: {
+			send: {
+				type: 'body',
+				property: 'document.data',
+				value:
+					'={{ { data: $input.binary[binaryProperty].data, mime_type: $input.binary[binaryProperty].mimeType } }}',
+			},
+		},
+	},
+	{
+		displayName: 'Document URL',
+		name: 'documentUrl',
+		type: 'string',
+		description: 'URL of the document to process',
+		placeholder: 'e.g. https://example.com/document.pdf',
+		required: true,
+		default: '',
+		displayOptions: {
+			show: {
+				inputType: ['url'],
+			},
+		},
+		routing: {
+			send: {
+				type: 'body',
+				property: 'document.document_url',
+			},
+		},
+	},
+	{
+		displayName: 'Enable Batch Processing',
+		name: 'enableBatchProcessing',
+		type: 'boolean',
+		description: 'Whether to process multiple documents in a single API call (more cost-efficient)',
+		default: false,
+		routing: {
+			send: {
+				type: 'body',
+				property: 'batch',
+				value: '={{ $value ? { size: $parameter.batchSize } : undefined }}',
+			},
+		},
+	},
+	{
+		displayName: 'Batch Size',
+		name: 'batchSize',
+		type: 'number',
+		description: 'Maximum number of documents to process in a single batch',
+		default: 25,
+		typeOptions: {
+			minValue: 1,
+			maxValue: 100,
+		},
+		required: true,
+		displayOptions: {
+			show: {
+				enableBatchProcessing: [true],
+			},
+		},
+		routing: {
+			send: {
+				type: 'body',
+				property: 'batch.size',
+			},
+		},
+	},
+];

packages/nodes-base/package.json

@@ -240,6 +240,7 @@
       "dist/credentials/MiroOAuth2Api.credentials.js",
       "dist/credentials/MispApi.credentials.js",
       "dist/credentials/MistApi.credentials.js",
+      "dist/credentials/MistralOcrApi.credentials.js",
       "dist/credentials/MoceanApi.credentials.js",
       "dist/credentials/MondayComApi.credentials.js",
       "dist/credentials/MondayComOAuth2Api.credentials.js",
@@ -649,6 +650,7 @@
       "dist/nodes/Microsoft/ToDo/MicrosoftToDo.node.js",
       "dist/nodes/Mindee/Mindee.node.js",
       "dist/nodes/Misp/Misp.node.js",
+      "dist/nodes/MistralOcr/MistralOcr.node.js",
       "dist/nodes/Mocean/Mocean.node.js",
       "dist/nodes/MondayCom/MondayCom.node.js",
       "dist/nodes/MongoDb/MongoDb.node.js",