Resume LaTeX Rendering Service

A small FastAPI service that renders structured resume data (JSON) into a PDF using a Jinja2 LaTeX template and the tectonic compiler.

This repository provides a lightweight HTTP API to convert a Resume JSON payload into a downloadable PDF (resume.pdf). The service:

• Validates input using Pydantic models defined in app/models.py.
• Renders a LaTeX source via Jinja2 template app/templates/resume_template.tex.j2.
• Compiles LaTeX to PDF with tectonic inside a temporary directory and returns the PDF bytes.
• Writes the last generated LaTeX to debug_resume.tex at the repository root for troubleshooting.

Contents

• app/
  • main.py — FastAPI application exposing /render-resume.
  • models.py — Pydantic models describing the resume schema.
  • latex_renderer.py — Template rendering and PDF compilation logic.
  • templates/resume_template.tex.j2 — Jinja2 LaTeX template.
• requirements.txt — Python dependencies.
• Dockerfile — (optional) project containerization.
• debug_resume.tex — generated LaTeX debug file (created at runtime).

Quickstart (macOS / Linux)

Prerequisites

• Python 3.8+ (3.10 recommended)
• pip or a virtual environment tool (venv/poetry)
• tectonic installed and available on PATH. On macOS you can install it with Homebrew:

brew install tectonic

1. Create and activate a virtual environment (recommended)

python3 -m venv .venv
source .venv/bin/activate

2. Install Python dependencies

pip install -r requirements.txt

3. Run the app with Uvicorn

uvicorn app.main:app --reload

The service will be available at http://127.0.0.1:8000.

API

POST /render-resume

• Description: Render a resume JSON into a compiled PDF.
• Content-Type: application/json
• Request body: RenderRequest (see app/models.py) — a short summary of the model below.
• Success response: 200 OK with body application/pdf and header Content-Disposition: attachment; filename="resume.pdf".
• Error responses:
  • 422 — JSON validation error (Pydantic / FastAPI)
  • 500 — Rendering or compilation error. The LaTeX source is written to debug_resume.tex for inspection.

Example request shape (fields shown for clarity — see app/models.py for full schema):

{
  "resume": {
    "header": {
      "name": "Jane Developer",
      "phone": "(555) 555-5555",
      "email": "jane@example.com",
      "location": "San Francisco, CA",
      "links": { "portfolio": "https://ex.com", "linkedin": "https://linkedin.com/in/jane" }
    },
    "summary": "Short professional summary...",
    "education": [ { "school": "University", "degree": "B.S.", "location": "City", "startDate": "2016", "endDate": "2020" } ],
    "experience": [ { "title": "Software Engineer", "company": "Acme", "location": "Remote", "startDate": "2021", "endDate": "Present", "bullets": ["Did X","Did Y"] } ],
    "projects": [ { "name": "Project A", "stack": "Python, FastAPI", "year": "2024", "bullets": ["Built feature A"] } ],
    "skills": { "languages": ["Python"], "frameworks": ["FastAPI"], "databases": ["Postgres"], "cloud": ["AWS"], "concepts": ["TDD"] }
  }
}

You can test quickly with curl (replace payload.json with your file):

curl -X POST "http://127.0.0.1:8000/render-resume" \
  -H "Content-Type: application/json" \
  --data-binary @payload.json \
  --output resume.pdf

If the LaTeX compilation fails, inspect debug_resume.tex for the generated source and check tectonic output in the server logs.

Development notes

• The LaTeX escaping filter latex_escape intentionally does not remove backslashes so that deliberate LaTeX commands in templates or user-provided content (if intended) can remain. Be cautious when allowing arbitrary input.
• render_resume_to_pdf writes the last generated LaTeX to debug_resume.tex in the repo root to help debugging compilation errors.
• The code calls tectonic via subprocess.run. Consider adding timeout= and logging to avoid long-running or silent failures.

Troubleshooting

• Error: "tectonic: command not found"

  • Install tectonic and ensure it's on PATH (see prerequisites).

• LaTeX compilation errors

  • Check debug_resume.tex for the generated LaTeX source.
  • Run tectonic debug_resume.tex --outdir output_dir to get detailed errors.

• FastAPI returns 422

  • The request JSON didn't validate against the Pydantic models. Check required fields and types.

Suggested next steps / enhancements

• Add structured logging (e.g., structlog or Python logging) to capture compile stdout/stderr and request metadata.
• Add a timeout to subprocess.run and return a 504 when compilation times out.
• Add unit tests that call render_resume_to_pdf with a minimal Resume and assert the returned bytes start with %PDF.
• Provide a Docker image that includes tectonic so the runtime is consistent across environments.