Evalua Javiera is a deliberately lightweight, zero-database web application. The backend is a Python FastAPI application that serves server-rendered HTML via Jinja2 templates. All interactive logic — authentication, teacher selection, and rating submission — runs in the browser using plain vanilla JavaScript, withDocumentation Index
Fetch the complete documentation index at: https://mintlify.com/groupTwoisTheBest/evaJav/llms.txt
Use this file to discover all available pages before exploring further.
localStorage acting as the sole persistence layer between page navigations. The application is deployed to Vercel using the official @vercel/python runtime with no build step required.
FastAPI
Python web framework powering all four backend routes. FastAPI handles HTTP
GET requests, renders Jinja2 templates through TemplateResponse, mounts the static/ directory, and returns HTMLResponse objects. Every route handler is declared async for non-blocking request handling.Jinja2
Server-side template engine used to render HTML pages. Templates follow an inheritance model:
base.html defines the document skeleton and three overridable blocks (title, stylesheets, content). Child templates extend base.html and fill in only what they need, keeping markup DRY.Vanilla JavaScript
All client-side interactivity is handled with plain JavaScript — no frameworks or build tools. Three modules cover the three interactive steps:
index.js (login), sele_prof.js (teacher selection), and cal_plataform.js (rating submission). Each module is loaded per-page via the linkJs context variable.CSS
A single global stylesheet (
style.css) provides the dark-themed, mobile-first responsive design. Key design tokens (--fondo: #222222, --texto: #f2f2f2, --border: 3px solid #3B4B6A) are defined as CSS custom properties. The login page loads an additional index.css for page-specific styles.Vercel
Cloud hosting platform. Evalua Javiera is deployed using the
@vercel/python runtime, configured in vercel.json. All incoming requests are routed to main.py via a catch-all route rule. No separate server process or Dockerfile is required.localStorage
The browser’s built-in
localStorage API serves as the sole state store between page navigations. The selected teacher’s name is stored under the key maestroSeleccionado, and the submitted ratings are stored as a JSON string under the key calification. No data is sent to or persisted on the server.Python Dependencies
All Python dependencies are listed inrequirements.txt:
@vercel/python runtime.
For local development, you also need an ASGI server to run the application. Uvicorn is the standard choice for FastAPI, but it is not listed in requirements.txt because Vercel does not require it. Install it separately:
Vercel Deployment Configuration
Thevercel.json file at the project root configures the Vercel build and routing:
| Key | Value | Description |
|---|---|---|
builds[].src | "main.py" | The entry-point file Vercel compiles with the Python runtime. |
builds[].use | "@vercel/python" | The Vercel runtime builder for Python WSGI/ASGI applications. |
routes[].src | "/(.*)" | A catch-all regex that matches every incoming URL path. |
routes[].dest | "main.py" | Routes all matched requests to the FastAPI application in main.py. |
Runtime Versions
| Requirement | Version |
|---|---|
| Python | 3.8 or later |
| FastAPI | Latest (no version constraint in requirements.txt) |
| Jinja2 | Latest (no version constraint in requirements.txt) |
| Uvicorn | Latest (local dev only, not in requirements.txt) |
Evalua Javiera uses no database of any kind. Student credentials live in
static/js/index.js, teacher names live in templates/selectProfesor.html, and all runtime state (selected teacher, submitted ratings) is stored exclusively in the browser’s localStorage. This means ratings are not persisted beyond the browser session and are not accessible server-side.