feat: add backend OCR manual proxy endpoint (refs #135)
Add POST /api/ocr/extract/manual endpoint that proxies to the Python OCR service's manual extraction pipeline. Includes Pro tier gating via document.scanMaintenanceSchedule, PDF-only validation, 200MB file size limit, and async 202 job response for polling via existing job status endpoint. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
This commit is contained in:
@@ -336,6 +336,112 @@ export class OcrController {
|
||||
}
|
||||
}
|
||||
|
||||
/**
|
||||
* POST /api/ocr/extract/manual
|
||||
* Submit an async manual extraction job for PDF owner's manuals.
|
||||
* Requires Pro tier (document.scanMaintenanceSchedule).
|
||||
*/
|
||||
async extractManual(
|
||||
request: FastifyRequest,
|
||||
reply: FastifyReply
|
||||
) {
|
||||
const userId = (request as any).user?.sub as string;
|
||||
|
||||
logger.info('Manual extract requested', {
|
||||
operation: 'ocr.controller.extractManual',
|
||||
userId,
|
||||
});
|
||||
|
||||
const file = await (request as any).file({ limits: { files: 1 } });
|
||||
if (!file) {
|
||||
logger.warn('No file provided for manual extraction', {
|
||||
operation: 'ocr.controller.extractManual.no_file',
|
||||
userId,
|
||||
});
|
||||
return reply.code(400).send({
|
||||
error: 'Bad Request',
|
||||
message: 'No file provided',
|
||||
});
|
||||
}
|
||||
|
||||
const contentType = file.mimetype as string;
|
||||
if (contentType !== 'application/pdf') {
|
||||
logger.warn('Non-PDF file provided for manual extraction', {
|
||||
operation: 'ocr.controller.extractManual.not_pdf',
|
||||
userId,
|
||||
contentType,
|
||||
fileName: file.filename,
|
||||
});
|
||||
return reply.code(400).send({
|
||||
error: 'Bad Request',
|
||||
message: `Manual extraction requires PDF files. Received: ${contentType}`,
|
||||
});
|
||||
}
|
||||
|
||||
const chunks: Buffer[] = [];
|
||||
for await (const chunk of file.file) {
|
||||
chunks.push(chunk);
|
||||
}
|
||||
const fileBuffer = Buffer.concat(chunks);
|
||||
|
||||
if (fileBuffer.length === 0) {
|
||||
logger.warn('Empty file provided for manual extraction', {
|
||||
operation: 'ocr.controller.extractManual.empty_file',
|
||||
userId,
|
||||
fileName: file.filename,
|
||||
});
|
||||
return reply.code(400).send({
|
||||
error: 'Bad Request',
|
||||
message: 'Empty file provided',
|
||||
});
|
||||
}
|
||||
|
||||
// Get optional vehicle_id from form fields
|
||||
const vehicleId = file.fields?.vehicle_id?.value as string | undefined;
|
||||
|
||||
try {
|
||||
const result = await ocrService.submitManualJob(userId, {
|
||||
fileBuffer,
|
||||
contentType,
|
||||
vehicleId,
|
||||
});
|
||||
|
||||
logger.info('Manual extract job submitted', {
|
||||
operation: 'ocr.controller.extractManual.success',
|
||||
userId,
|
||||
jobId: result.jobId,
|
||||
status: result.status,
|
||||
estimatedSeconds: result.estimatedSeconds,
|
||||
});
|
||||
|
||||
return reply.code(202).send(result);
|
||||
} catch (error: any) {
|
||||
if (error.statusCode === 413) {
|
||||
return reply.code(413).send({
|
||||
error: 'Payload Too Large',
|
||||
message: error.message,
|
||||
});
|
||||
}
|
||||
if (error.statusCode === 400) {
|
||||
return reply.code(400).send({
|
||||
error: 'Bad Request',
|
||||
message: error.message,
|
||||
});
|
||||
}
|
||||
|
||||
logger.error('Manual extract failed', {
|
||||
operation: 'ocr.controller.extractManual.error',
|
||||
userId,
|
||||
error: error.message,
|
||||
});
|
||||
|
||||
return reply.code(500).send({
|
||||
error: 'Internal Server Error',
|
||||
message: 'Manual extraction submission failed',
|
||||
});
|
||||
}
|
||||
}
|
||||
|
||||
/**
|
||||
* POST /api/ocr/jobs
|
||||
* Submit an async OCR job for large files.
|
||||
|
||||
@@ -29,6 +29,12 @@ export const ocrRoutes: FastifyPluginAsync = async (
|
||||
handler: ctrl.extractReceipt.bind(ctrl),
|
||||
});
|
||||
|
||||
// POST /api/ocr/extract/manual - Manual extraction (Pro tier required)
|
||||
fastify.post('/ocr/extract/manual', {
|
||||
preHandler: [requireAuth, fastify.requireTier({ featureKey: 'document.scanMaintenanceSchedule' })],
|
||||
handler: ctrl.extractManual.bind(ctrl),
|
||||
});
|
||||
|
||||
// POST /api/ocr/jobs - Submit async OCR job
|
||||
fastify.post('/ocr/jobs', {
|
||||
preHandler: [requireAuth],
|
||||
|
||||
@@ -5,6 +5,8 @@ import { logger } from '../../../core/logging/logger';
|
||||
import { ocrClient, JobNotFoundError } from '../external/ocr-client';
|
||||
import type {
|
||||
JobResponse,
|
||||
ManualJobResponse,
|
||||
ManualJobSubmitRequest,
|
||||
OcrExtractRequest,
|
||||
OcrJobSubmitRequest,
|
||||
OcrResponse,
|
||||
@@ -278,6 +280,66 @@ export class OcrService {
|
||||
}
|
||||
}
|
||||
|
||||
/**
|
||||
* Submit an async manual extraction job for PDF owner's manuals.
|
||||
*
|
||||
* @param userId - User ID for logging
|
||||
* @param request - Manual job submission request
|
||||
* @returns Manual job response with job ID
|
||||
*/
|
||||
async submitManualJob(userId: string, request: ManualJobSubmitRequest): Promise<ManualJobResponse> {
|
||||
// Validate file size for async processing (200MB max)
|
||||
if (request.fileBuffer.length > MAX_ASYNC_SIZE) {
|
||||
const err: any = new Error(
|
||||
`File too large. Max: ${MAX_ASYNC_SIZE / (1024 * 1024)}MB.`
|
||||
);
|
||||
err.statusCode = 413;
|
||||
throw err;
|
||||
}
|
||||
|
||||
// Manual extraction only supports PDF
|
||||
if (request.contentType !== 'application/pdf') {
|
||||
const err: any = new Error(
|
||||
`Unsupported file type: ${request.contentType}. Manual extraction requires PDF files.`
|
||||
);
|
||||
err.statusCode = 400;
|
||||
throw err;
|
||||
}
|
||||
|
||||
logger.info('Manual job submit requested', {
|
||||
operation: 'ocr.service.submitManualJob',
|
||||
userId,
|
||||
contentType: request.contentType,
|
||||
fileSize: request.fileBuffer.length,
|
||||
hasVehicleId: !!request.vehicleId,
|
||||
});
|
||||
|
||||
try {
|
||||
const result = await ocrClient.submitManualJob(
|
||||
request.fileBuffer,
|
||||
request.contentType,
|
||||
request.vehicleId
|
||||
);
|
||||
|
||||
logger.info('Manual job submitted', {
|
||||
operation: 'ocr.service.submitManualJob.success',
|
||||
userId,
|
||||
jobId: result.jobId,
|
||||
status: result.status,
|
||||
estimatedSeconds: result.estimatedSeconds,
|
||||
});
|
||||
|
||||
return result;
|
||||
} catch (error) {
|
||||
logger.error('Manual job submit failed', {
|
||||
operation: 'ocr.service.submitManualJob.error',
|
||||
userId,
|
||||
error: error instanceof Error ? error.message : 'Unknown error',
|
||||
});
|
||||
throw error;
|
||||
}
|
||||
}
|
||||
|
||||
/**
|
||||
* Get the status of an async OCR job.
|
||||
*
|
||||
|
||||
@@ -79,3 +79,49 @@ export interface OcrJobSubmitRequest {
|
||||
contentType: string;
|
||||
callbackUrl?: string;
|
||||
}
|
||||
|
||||
/** Request to submit a manual extraction job */
|
||||
export interface ManualJobSubmitRequest {
|
||||
fileBuffer: Buffer;
|
||||
contentType: string;
|
||||
vehicleId?: string;
|
||||
}
|
||||
|
||||
/** Vehicle info extracted from a manual */
|
||||
export interface ManualVehicleInfo {
|
||||
make: string | null;
|
||||
model: string | null;
|
||||
year: number | null;
|
||||
}
|
||||
|
||||
/** A single maintenance schedule item extracted from a manual */
|
||||
export interface MaintenanceScheduleItem {
|
||||
service: string;
|
||||
intervalMiles: number | null;
|
||||
intervalMonths: number | null;
|
||||
details: string | null;
|
||||
confidence: number;
|
||||
subtypes: string[];
|
||||
}
|
||||
|
||||
/** Result of manual extraction (nested in ManualJobResponse.result) */
|
||||
export interface ManualExtractionResult {
|
||||
success: boolean;
|
||||
vehicleInfo: ManualVehicleInfo;
|
||||
maintenanceSchedules: MaintenanceScheduleItem[];
|
||||
rawTables: unknown[];
|
||||
processingTimeMs: number;
|
||||
totalPages: number;
|
||||
pagesProcessed: number;
|
||||
error: string | null;
|
||||
}
|
||||
|
||||
/** Response for async manual extraction job */
|
||||
export interface ManualJobResponse {
|
||||
jobId: string;
|
||||
status: JobStatus;
|
||||
progress?: number;
|
||||
estimatedSeconds?: number;
|
||||
result?: ManualExtractionResult;
|
||||
error?: string;
|
||||
}
|
||||
|
||||
57
backend/src/features/ocr/external/ocr-client.ts
vendored
57
backend/src/features/ocr/external/ocr-client.ts
vendored
@@ -2,7 +2,7 @@
|
||||
* @ai-summary HTTP client for OCR service communication
|
||||
*/
|
||||
import { logger } from '../../../core/logging/logger';
|
||||
import type { JobResponse, OcrResponse, ReceiptExtractionResponse, VinExtractionResponse } from '../domain/ocr.types';
|
||||
import type { JobResponse, ManualJobResponse, OcrResponse, ReceiptExtractionResponse, VinExtractionResponse } from '../domain/ocr.types';
|
||||
|
||||
/** OCR service configuration */
|
||||
const OCR_SERVICE_URL = process.env.OCR_SERVICE_URL || 'http://mvp-ocr:8000';
|
||||
@@ -265,6 +265,61 @@ export class OcrClient {
|
||||
return (await response.json()) as JobResponse;
|
||||
}
|
||||
|
||||
/**
|
||||
* Submit an async manual extraction job for PDF owner's manuals.
|
||||
*
|
||||
* @param fileBuffer - PDF file buffer
|
||||
* @param contentType - MIME type of the file (must be application/pdf)
|
||||
* @param vehicleId - Optional vehicle ID for context
|
||||
* @returns Manual job submission response
|
||||
*/
|
||||
async submitManualJob(
|
||||
fileBuffer: Buffer,
|
||||
contentType: string,
|
||||
vehicleId?: string
|
||||
): Promise<ManualJobResponse> {
|
||||
const formData = this.buildFormData(fileBuffer, contentType);
|
||||
if (vehicleId) {
|
||||
formData.append('vehicle_id', vehicleId);
|
||||
}
|
||||
|
||||
const url = `${this.baseUrl}/extract/manual`;
|
||||
|
||||
logger.info('OCR manual job submit request', {
|
||||
operation: 'ocr.client.submitManualJob',
|
||||
url,
|
||||
contentType,
|
||||
fileSize: fileBuffer.length,
|
||||
hasVehicleId: !!vehicleId,
|
||||
});
|
||||
|
||||
const response = await this.fetchWithTimeout(url, {
|
||||
method: 'POST',
|
||||
body: formData,
|
||||
});
|
||||
|
||||
if (!response.ok) {
|
||||
const errorText = await response.text();
|
||||
logger.error('OCR manual job submit failed', {
|
||||
operation: 'ocr.client.submitManualJob.error',
|
||||
status: response.status,
|
||||
error: errorText,
|
||||
});
|
||||
throw new Error(`OCR service error: ${response.status} - ${errorText}`);
|
||||
}
|
||||
|
||||
const result = (await response.json()) as ManualJobResponse;
|
||||
|
||||
logger.info('OCR manual job submitted', {
|
||||
operation: 'ocr.client.submitManualJob.success',
|
||||
jobId: result.jobId,
|
||||
status: result.status,
|
||||
estimatedSeconds: result.estimatedSeconds,
|
||||
});
|
||||
|
||||
return result;
|
||||
}
|
||||
|
||||
/**
|
||||
* Check if the OCR service is healthy.
|
||||
*
|
||||
|
||||
213
backend/src/features/ocr/tests/unit/ocr-manual.test.ts
Normal file
213
backend/src/features/ocr/tests/unit/ocr-manual.test.ts
Normal file
@@ -0,0 +1,213 @@
|
||||
/**
|
||||
* @ai-summary Unit tests for OCR manual extraction endpoint
|
||||
*/
|
||||
|
||||
import { OcrService } from '../../domain/ocr.service';
|
||||
import { ocrClient } from '../../external/ocr-client';
|
||||
import type { ManualJobResponse } from '../../domain/ocr.types';
|
||||
|
||||
jest.mock('../../external/ocr-client');
|
||||
jest.mock('../../../../core/logging/logger');
|
||||
|
||||
const mockSubmitManualJob = ocrClient.submitManualJob as jest.MockedFunction<
|
||||
typeof ocrClient.submitManualJob
|
||||
>;
|
||||
|
||||
describe('OcrService.submitManualJob', () => {
|
||||
let service: OcrService;
|
||||
|
||||
const userId = 'test-user-id';
|
||||
|
||||
const mockManualJobResponse: ManualJobResponse = {
|
||||
jobId: 'manual-job-123',
|
||||
status: 'pending',
|
||||
progress: 0,
|
||||
estimatedSeconds: 45,
|
||||
result: undefined,
|
||||
error: undefined,
|
||||
};
|
||||
|
||||
const mockCompletedJobResponse: ManualJobResponse = {
|
||||
jobId: 'manual-job-123',
|
||||
status: 'completed',
|
||||
progress: 100,
|
||||
result: {
|
||||
success: true,
|
||||
vehicleInfo: {
|
||||
make: 'Honda',
|
||||
model: 'Civic',
|
||||
year: 2023,
|
||||
},
|
||||
maintenanceSchedules: [
|
||||
{
|
||||
service: 'Engine Oil Change',
|
||||
intervalMiles: 5000,
|
||||
intervalMonths: 6,
|
||||
details: 'Use 0W-20 full synthetic oil',
|
||||
confidence: 0.95,
|
||||
subtypes: ['oil_change'],
|
||||
},
|
||||
{
|
||||
service: 'Tire Rotation',
|
||||
intervalMiles: 7500,
|
||||
intervalMonths: 6,
|
||||
details: null,
|
||||
confidence: 0.90,
|
||||
subtypes: ['tire_rotation'],
|
||||
},
|
||||
],
|
||||
rawTables: [],
|
||||
processingTimeMs: 45000,
|
||||
totalPages: 120,
|
||||
pagesProcessed: 120,
|
||||
error: null,
|
||||
},
|
||||
error: undefined,
|
||||
};
|
||||
|
||||
beforeEach(() => {
|
||||
jest.clearAllMocks();
|
||||
service = new OcrService();
|
||||
});
|
||||
|
||||
describe('valid manual job submission', () => {
|
||||
it('should return 202-style response with jobId for PDF submission', async () => {
|
||||
mockSubmitManualJob.mockResolvedValue(mockManualJobResponse);
|
||||
|
||||
const result = await service.submitManualJob(userId, {
|
||||
fileBuffer: Buffer.from('fake-pdf-data'),
|
||||
contentType: 'application/pdf',
|
||||
});
|
||||
|
||||
expect(result.jobId).toBe('manual-job-123');
|
||||
expect(result.status).toBe('pending');
|
||||
expect(result.progress).toBe(0);
|
||||
expect(result.estimatedSeconds).toBe(45);
|
||||
expect(result.result).toBeUndefined();
|
||||
});
|
||||
|
||||
it('should pass vehicleId to client when provided', async () => {
|
||||
mockSubmitManualJob.mockResolvedValue(mockManualJobResponse);
|
||||
|
||||
await service.submitManualJob(userId, {
|
||||
fileBuffer: Buffer.from('fake-pdf-data'),
|
||||
contentType: 'application/pdf',
|
||||
vehicleId: 'vehicle-abc',
|
||||
});
|
||||
|
||||
expect(mockSubmitManualJob).toHaveBeenCalledWith(
|
||||
expect.any(Buffer),
|
||||
'application/pdf',
|
||||
'vehicle-abc'
|
||||
);
|
||||
});
|
||||
|
||||
it('should call client without vehicleId when not provided', async () => {
|
||||
mockSubmitManualJob.mockResolvedValue(mockManualJobResponse);
|
||||
|
||||
await service.submitManualJob(userId, {
|
||||
fileBuffer: Buffer.from('fake-pdf-data'),
|
||||
contentType: 'application/pdf',
|
||||
});
|
||||
|
||||
expect(mockSubmitManualJob).toHaveBeenCalledWith(
|
||||
expect.any(Buffer),
|
||||
'application/pdf',
|
||||
undefined
|
||||
);
|
||||
});
|
||||
});
|
||||
|
||||
describe('completed job result', () => {
|
||||
it('should return completed result with maintenanceSchedules', async () => {
|
||||
mockSubmitManualJob.mockResolvedValue(mockCompletedJobResponse);
|
||||
|
||||
const result = await service.submitManualJob(userId, {
|
||||
fileBuffer: Buffer.from('fake-pdf-data'),
|
||||
contentType: 'application/pdf',
|
||||
});
|
||||
|
||||
expect(result.status).toBe('completed');
|
||||
expect(result.result).toBeDefined();
|
||||
expect(result.result!.success).toBe(true);
|
||||
expect(result.result!.maintenanceSchedules).toHaveLength(2);
|
||||
expect(result.result!.maintenanceSchedules[0].service).toBe('Engine Oil Change');
|
||||
expect(result.result!.maintenanceSchedules[0].intervalMiles).toBe(5000);
|
||||
expect(result.result!.maintenanceSchedules[0].subtypes).toEqual(['oil_change']);
|
||||
expect(result.result!.vehicleInfo.make).toBe('Honda');
|
||||
});
|
||||
});
|
||||
|
||||
describe('error handling', () => {
|
||||
it('should throw 400 for non-PDF file (JPEG)', async () => {
|
||||
await expect(
|
||||
service.submitManualJob(userId, {
|
||||
fileBuffer: Buffer.from('fake-image-data'),
|
||||
contentType: 'image/jpeg',
|
||||
})
|
||||
).rejects.toMatchObject({
|
||||
statusCode: 400,
|
||||
});
|
||||
});
|
||||
|
||||
it('should throw 400 for non-PDF file (PNG)', async () => {
|
||||
await expect(
|
||||
service.submitManualJob(userId, {
|
||||
fileBuffer: Buffer.from('fake-image-data'),
|
||||
contentType: 'image/png',
|
||||
})
|
||||
).rejects.toMatchObject({
|
||||
statusCode: 400,
|
||||
});
|
||||
});
|
||||
|
||||
it('should throw 400 for text/plain', async () => {
|
||||
await expect(
|
||||
service.submitManualJob(userId, {
|
||||
fileBuffer: Buffer.from('not a pdf'),
|
||||
contentType: 'text/plain',
|
||||
})
|
||||
).rejects.toMatchObject({
|
||||
statusCode: 400,
|
||||
});
|
||||
});
|
||||
|
||||
it('should throw 413 for oversized file', async () => {
|
||||
const largeBuffer = Buffer.alloc(201 * 1024 * 1024); // 201MB
|
||||
|
||||
await expect(
|
||||
service.submitManualJob(userId, {
|
||||
fileBuffer: largeBuffer,
|
||||
contentType: 'application/pdf',
|
||||
})
|
||||
).rejects.toMatchObject({
|
||||
statusCode: 413,
|
||||
});
|
||||
});
|
||||
|
||||
it('should accept file at 200MB boundary', async () => {
|
||||
mockSubmitManualJob.mockResolvedValue(mockManualJobResponse);
|
||||
const exactBuffer = Buffer.alloc(200 * 1024 * 1024); // exactly 200MB
|
||||
|
||||
const result = await service.submitManualJob(userId, {
|
||||
fileBuffer: exactBuffer,
|
||||
contentType: 'application/pdf',
|
||||
});
|
||||
|
||||
expect(result.jobId).toBe('manual-job-123');
|
||||
});
|
||||
|
||||
it('should propagate OCR service errors', async () => {
|
||||
mockSubmitManualJob.mockRejectedValue(
|
||||
new Error('OCR service error: 500 - Internal error')
|
||||
);
|
||||
|
||||
await expect(
|
||||
service.submitManualJob(userId, {
|
||||
fileBuffer: Buffer.from('fake-pdf-data'),
|
||||
contentType: 'application/pdf',
|
||||
})
|
||||
).rejects.toThrow('OCR service error: 500 - Internal error');
|
||||
});
|
||||
});
|
||||
});
|
||||
Reference in New Issue
Block a user