Hüseyin Uğur Yıldız

Academic Website

Publications, research, teaching, supervision, service, and a web + PDF curriculum vitae.

---

Source code and scholarly content for huguryildiz.com, the academic website of Hüseyin Uğur Yıldız. The site presents publications, research projects, teaching, graduate supervision, professional service, and a web and PDF curriculum vitae. It also preserves research notes and announcements and publishes an aggregated, privacy-conscious site-reach report.

The website is generated with Jekyll and deployed to GitHub Pages from the master branch. It uses no theme: every page renders through a single custom academic interface.

Site sections

Section	Content
Home	Academic profile, research focus, news, and an interactive ocean digital twin
Publications	Publication catalogue with type, year, and quartile filters; DOI and local full-text links
Research	Research program, interactive topic map, software, and learning resources
Writing	Notes, announcements, and longer pieces retained as part of the scholarly record
CV	Web curriculum vitae and the downloadable PDF version
Teaching	Undergraduate and graduate courses
Students	Graduate supervision and information for prospective students
Service	Reviewing, committee, chairing, institutional, and invited-talk activities
Site Reach	Aggregated GoatCounter page-view, content, geographic, referral, and reading-environment summaries, selectable by date range, with explicit data limitations

Rendering architecture

The repository has one presentation path. Every page uses it.

Path	Pages	Rendering chain
Academic interface	All of them — home, publications, research, writing, posts, CV, teaching, students, service, site reach, terms, and 404	Page → _layouts/academic.html → _layouts/compress.html → assets/css/redesign.css; posts first pass through _layouts/post.html

The primary navigation is defined directly in _layouts/academic.html; there is no navigation data file. The Minimal Mistakes remote theme and its local overrides were removed — no _sass/, no main.scss/custom.css, no default.html/single.html. Extend assets/css/redesign.css rather than reintroducing a theme.

Repository map

Path	Responsibility
index.md	Home-page content
_pages/	Main and utility pages
_posts/	Dated writing entries published under /writing/:title/
_layouts/academic.html	Primary navigation, page shell, footer, theme control, and shared interactions
_layouts/post.html	Writing-entry metadata, reading time, tags, and optional source link
_includes/_shared/	Shared fragments: favicon markup and animation, direct GA4, GoatCounter
_includes/hero-uwsn.html	Interactive ocean digital twin and underwater acoustic-network scenario
_includes/research-map.html	Interactive research map rendered from _data/research_map.yml
assets/css/redesign.css	All site styles
_data/cv.yml	CV data — the source for _pages/cv.md, _pages/service.md, _pages/teaching.md, and the downloadable PDF
_data/publications.yml	Publication records — the source for _pages/publications.md and the downloadable PDF
_data/scholar_metrics.json	Google Scholar metrics snapshot; rendered by index.md and _pages/cv.md
_data/openalex_metrics.json	OpenAlex metrics snapshot; dormant fallback, not rendered
_data/site_stats.json	Sanitized GoatCounter snapshot rendered by _pages/stats.md
files/	CV, theses, papers, and other downloadable scholarly documents
assets/images/	Portraits, research graphics, course images, project images, and favicons
assets/video/courses/, assets/video/topics/	Course and research-topic media
scripts/	Google Scholar, OpenAlex, and GoatCounter retrieval plus optional CV-conversion utilities; excluded from the build (see scripts/README.md)
_config.yml	Jekyll, metadata, SEO, and plugin configuration
PRODUCT.md	Product, data-honesty, design, and accessibility contract for the ocean digital twin

Generated output under _site/ is not a source and should not be edited.

What reaches the public site is decided by exclude in _config.yml, not by .gitignore. Jekyll skips dot- and underscore-prefixed paths automatically; any other tracked path is copied into _site/ and served. Before adding a file at the repository root, decide whether it should become a public URL.

Local development

Docker

Docker is the most reproducible option because it does not depend on the host's Ruby installation:

docker compose up --build

The container combines _config.yml with _config_docker.yml and serves the site at http://localhost:4000. Source changes are watched automatically.

Native Ruby

With a compatible Ruby installation and Bundler:

bundle install
bundle exec jekyll serve

For a production-style build:

JEKYLL_ENV=production bundle exec jekyll build

A change to _config.yml requires restarting the development server.

Content maintenance

The web CV, the downloadable PDF, and the publication catalogue share one source of truth: _data/cv.yml and _data/publications.yml.

• Updating the CV means editing _data/cv.yml. _pages/cv.md, _pages/service.md, and _pages/teaching.md render it directly, and the PDF regenerates from the same file on every deploy via scripts/render_cv_tex.py — never edit files/Yildiz_HuseyinUgur_CV.pdf or main.tex by hand. Run scripts/build_cv_pdf.sh to refresh the local PDF from the same data and template used by production.

• Publication records live in _data/publications.yml; _pages/publications.md renders them and also drives the PDF's publication list. Verify metadata, DOI links, and local PDF paths when adding a record.

• Research-map content belongs in _data/research_map.yml; its labels and counts must remain consistent with _includes/research-map.html.

• The writing index is _pages/writing.md; dated entries live in _posts/. Post URLs follow /writing/:title/, as configured in _config.yml.

• Do not infer or hand-edit bibliometric values without a source. The citation counts, h-index, and work count shown on the home page and the CV come from _data/scholar_metrics.json. To refresh that snapshot locally, run:

  SERPAPI_KEY=... SCHOLAR_AUTHOR_ID=nQwHS1gAAAAJ python3 scripts/fetch_scholar.py

  Google Scholar has no public API, so the script reads the profile through SerpApi's Google Scholar Author endpoint. The key is a secret: never print it, embed it in browser code, or commit it. To refresh the OpenAlex snapshot, run:

  OPENALEX_AUTHOR_ID=A5085505896 python3 scripts/fetch_openalex.py

  Both scripts require Python's requests package. OPENALEX_MAILTO is optional.

• _pages/stats.md renders the sanitized snapshot in _data/site_stats.json. The daily GitHub Actions workflow runs scripts/fetch_goatcounter.py with the GOATCOUNTER_API_TOKEN repository secret. The token must never be placed in client-side code or committed data. The public report does not infer visitors, sessions, bounce rate, or other measures unavailable from the source API.

The CV conversion scripts are optional maintenance utilities, not an active site-generation pipeline. Use scripts/update_cv_json.sh only when JSON Resume output is explicitly needed, and review the generated data before retaining it.

Validation

There is no automated test suite. Before publishing a change:

1. Run git diff --check and inspect the focused diff.
2. Complete a Jekyll production build.
3. Inspect the affected page at desktop and narrow mobile widths.
4. Check light and dark themes.
5. Exercise changed controls with both keyboard and pointer input.
6. Check the browser console when JavaScript changes.
7. Confirm that affected internal links and local assets resolve.

Changes to shared layouts, navigation, CSS, or theme behavior should be checked on every academic page — which is now every page on the site.

Deployment and metrics

A push to master, a daily schedule, or manual dispatch starts .github/workflows/jekyll.yml, which builds the site with Ruby 3.1, renders and compiles the CV PDF from _data/cv.yml + _data/publications.yml with scripts/render_cv_tex.py and pdflatex, places it at _site/files/Yildiz_HuseyinUgur_CV.pdf, and deploys to GitHub Pages. A failed or empty PDF fails the build so the previously deployed site stays live. The custom domain is declared in CNAME.

.github/workflows/update-scholar.yml runs weekly and can also be triggered manually. It reads the Google Scholar profile through SerpApi and commits an updated _data/scholar_metrics.json only when the snapshot changes. It needs the SERPAPI_KEY and SCHOLAR_AUTHOR_ID repository secrets.

.github/workflows/update-openalex.yml is a dormant fallback: manual dispatch only, no schedule. _data/openalex_metrics.json is retained as a second bibliometric source but is not rendered by any page and is not refreshed automatically.

.github/workflows/update-goatcounter.yml runs daily and can also be triggered manually. It retrieves aggregated analytics with a repository secret and commits _data/site_stats.json only when the snapshot changes.

License

Code and content are licensed separately; see LICENSE for the exact paths each covers.

• Software — the layouts, includes, stylesheet, scripts, LaTeX template, and workflows are MIT licensed. Take any part of them, including the _data/cv.yml → web CV + LaTeX PDF pipeline.
• Content and media — the prose, biography, publication and teaching records, and images are CC BY-NC 4.0. This is a personal scholarly record, so reuse means credit and non-commercial use, not adoption as your own CV.
• Publication PDFs under files/ — excluded from both. Copyright in most of them belongs to their publishers and they are posted here only under the self-archiving terms each publisher permits.

This repository is a working site rather than a template: content is written directly into the pages, and _includes/hero-uwsn.html is specific to one research programme. It is published to be read and borrowed from, not forked wholesale.