💻 Job Research Application

Job Research Application is a Spring Boot backend for managing job opportunity records, including position, company, location, skills, work mode, status, posted date, description, application URL, salary, and source/origin tracking. It provides REST APIs for the React frontend, supports JWT-based authentication, role-based access control, CRUD operations, paginated job retrieval, backend sorting, and field-based dynamic filtering using Spring Data JPA Specification. Job records are populated both through manual entry from the frontend and through an automated n8n scraping pipeline. The backend is packaged as a Docker container, deployed on Render, and integrates with PostgreSQL.

---

🧱 Tech Stack

• Java 17
• Spring Boot
• Spring Security + JWT
• PostgreSQL (hosted on Supabase)
• Spring Data JPA for pagination, sorting, and Specification-based filtering
• Maven
• CORS Configuration (for frontend integration)

---

🔐 Authentication & Roles

• JWT-based authentication
• Two roles:
  • USER: can view jobs
  • ADMIN: can create, update, and delete jobs

---

📦 REST API Overview

Method	Endpoint	Access	Description
POST	/login	Public	Authenticates user and returns JWT token
GET	/health	Public	Lightweight health check, used to wake the backend from idle on Render's free tier
GET	/api/jobs/filter	USER/ADMIN	Retrieve paginated job list with sorting and field-based filtering
POST	/api/jobs	ADMIN	Create new job entry
PUT	/api/jobs/{id}	ADMIN	Update existing job
DELETE	/api/jobs/{id}	ADMIN	Delete job entry

---

🤖 Automated Job Ingestion (n8n)

In addition to manual entry through the frontend, job records are written to this backend by an external n8n workflow that scrapes multiple job boards (Jooble, WeWorkRemotely, RemoteOK, Remotive, Himalayas) on a schedule.

• The workflow authenticates against POST /login like any other client, then calls POST /api/jobs for each newly discovered listing.

• The n8n integration uses the following fields:

  • scrapedFrom`: the platform from which the job listing was collected (e.g. Jooble, We Work Remotely, RemoteOK), populated by the n8n pipeline.
  • externalJobId: the identifier of the listing on its original platform, used by the workflow to avoid creating duplicate records across runs.

• The GET /health endpoint exists specifically to support this integration: since the backend can spin down when idle on Render's free tier, the workflow pings this endpoint with retries before authenticating, so a cold backend doesn't cause a scheduled run to fail.

---

Job List Query Parameters

The /api/jobs/filter endpoint supports pagination, sorting, and field-based filtering.

Example:

GET /api/jobs/filter?page=0&size=10&sortBy=postedDate&sortDir=desc&position=frontend&company=supabase&location=remote&mode=Remote&status=APPLIED

Parameter	Description
page	Page number, starting from 0
size	Number of records per page
sortBy	Field used for sorting
sortDir	Sort direction: asc or desc
position	Filter jobs by position/title
company	Filter jobs by company name
location	Filter jobs by location
mode	Filter jobs by work mode
status	Filter jobs by job status

---

Additional job fields — url, salary, externalJobId, and scrapedFrom — are stored and returned on every job record, but are not currently exposed as filter parameters on this endpoint.

---

🧩 Testing

This project includes tests for JobService, JwtService, JWTAuthorizationFilter, and JobController, covering job sorting, JWT generation, authorization filter behavior, and API responses.

Run tests locally

Windows:

.\mvnw.cmd test -DskipTests=false

macOS / Linux:

./mvnw test -DskipTests=false

Test Coverage

Test Class	Type	What it verifies
JobServiceTest	Unit test (Mockito)	Fallback to postedDate for invalid sort fields, ascending sorting by company, and an exception when updating a non-existent job
JwtServiceTest	Unit test	Generated JWT subject, roles claim, and expiration approximately 24 hours after token creation
JWTAuthorizationFilterTest	Unit test (Mockito)	Authentication is set for a valid JWT, while requests without a token remain unauthenticated and continue through the filter chain
JobControllerTest	Web layer test (@WebMvcTest)	GET /api/jobs/{id} returns 200 OK with job data when found and 404 Not Found when the job does not exist

Note: JobresearchApplicationTests.contextLoads() is currently disabled because @SpringBootTest loads the full application context and requires database configuration that is not available in the local test environment. Tests are run locally before changes are committed, while Render builds skip them to reduce deployment time.

🚀 Deployment

The backend is deployed on Render, with PostgreSQL hosted on Supabase.

• Deployed: [https://jobresearch-backend.onrender.com]

---