Jan Miksa — research website
Minimalist academic website built with Hugo on top of the
hugo-website template (PaperMod
theme). The publication list is synchronized automatically with
Google Scholar and
the CV PDF is generated from structured data, so the website, the CV, and
LinkedIn all tell the same story.
Structure
config/_default/hugo.yaml site configuration (menu, homepage, social links)
config/development/hugo.yaml local dev override (serves at http://localhost:1313/)
content/ homepage copy lives in config; plus:
publications/_index.md publications page (rendered from data/)
projects.md gmumcl, ActivationScope
data/
cv.yaml CV source of truth (experience, education, skills)
publications_manual.yaml manual overrides + manual-only publications
publications.yaml GENERATED: merged list used by site + CV
scholar_cache.yaml GENERATED: raw Scholar snapshot (committed)
layouts/ custom homepage, publications page, partials
assets/css/extended/ site CSS (homepage spacing, publication styles)
scripts/
sync_publications.py Scholar sync + manual merge
generate_cv.py cv/cv.tex + static/cv.pdf from data
cv/cv.tex GENERATED: LaTeX source of the CV
static/cv.pdf GENERATED: compiled CV served by the site
.github/workflows/
hugo.yml build + deploy on every push
sync.yml weekly Scholar sync, CV rebuild, commit + deploy
Common tasks
Install the Python dependencies once:
python3 -m venv .venv
.venv/bin/pip install -r scripts/requirements.txt
Then:
make serve # local preview at http://localhost:1313/ (hugo server)
make site # build into public/
make sync # merge cached Scholar data + manual overrides (offline)
make sync-online # fetch from Google Scholar, then merge
make cv # regenerate cv/cv.tex and compile static/cv.pdf
To edit a publication (fix venue, add code link, change status, reorder), edit
data/publications_manual.yaml and run make sync. To add a paper that is not
on Scholar yet, add it to the entries: list in the same file. Never edit
data/publications.yaml by hand — it is regenerated.
To edit the CV, change data/cv.yaml (or data/publications_manual.yaml for
publication lines) and run make cv.
Automatic updates
The workflow .github/workflows/sync.yml runs every Monday:
- fetches the publication list from Google Scholar,
- merges it with the manual overrides,
- regenerates the CV PDF,
- commits
data/, cv/, and static/cv.pdf if anything changed,
- rebuilds and deploys the site.
Google Scholar blocks anonymous scraping intermittently. If a scheduled fetch
fails, the workflow keeps the last committed data (no breakage), and the next
run tries again. For a more reliable sync, create a SerpAPI
account and add the key as the repository secret SERPAPI_API_KEY; the script
detects it automatically and uses the API instead of scraping.
Deployment
The site is deployed by GitHub Actions to GitHub Pages. Set the final URL in
config/_default/hugo.yaml (baseURL) once the repository and domain are
known. The
canonical URL of the live site should match it.
License
Template code: MIT (see LICENSE.md). Site content © Jan Miksa.