My Personal GitHub Page

Source for my site at https://yarinamomo.github.io/. GitHub Pages builds it with Jekyll on every push to main.

Where things are

To change…	Edit
Home page text, links	index.html
Profile photo	replace images/profile.jpg (keep the name)
CV page, Publications page	the Overleaf project, then regenerate (see below)
Hobbies gallery	_data/hobbies.yml + images in images/hobbies/
Navigation bar	_data/nav.yml
Blog posts	_posts/YYYY-MM-DD-title.md (listed on /blogs.html)
Styles (all pages)	css/style.css
Page wrapper (<head>, nav)	_layouts/default.html, _includes/nav.html
Site title / description	_config.yml

Pages start with YAML front matter (layout, title, body_class). body_class: home is special: the layout then leaves out the nav, because index.html places it below the hero itself.

Add a publication

Publications have one source: own-bib.bib in the Overleaf CV. Add the entry there plus a \nocite{key} in publications.tex (the \nocite order is the display order), then regenerate as described under Updating the CV. This updates the PDF, cv.html and publications.html together.

Add a drawing

Put the image in images/hobbies/ and add to _data/hobbies.yml:

- image: /images/hobbies/my-drawing.jpg   # file names are case-sensitive on GitHub Pages
  alt: Short description for screen readers
  caption: Caption shown under the image

Add a blog post

Create _posts/YYYY-MM-DD-some-title.md with front matter title: (and optionally tags:). The post layout is applied automatically. Put its images in images/<post-name>/ and reference them as {{ '/images/<post-name>/file.png' | relative_url }}. The blog page is /blogs.html; add it to _data/nav.yml to show it in the menu.

Updating the CV

cv.html and publications.html are generated from the Overleaf LaTeX project, so the website always matches the PDF. Don't edit either file by hand; the next run overwrites them.

1. In Overleaf: Menu → Download → Source and save the zip as _data/CV.zip.

2. From the repository root, run (Python 3.8+, no packages needed):

   python scripts/generate_cv.py

3. Read any warning: lines, preview the site, and commit _data/CV.zip, cv.html and publications.html together.

What the generator reads:

• Header: the name line and each \makefield{icon}{content} inside \leftheader{...}.
• Sections: every \makerubric{file} / \input{file} between \begin{document} and \end{document}, in that order. Adding a new \makerubric{awards} with an awards.tex rubric adds an Awards section automatically.
• Rubric files: \begin{rubric}{Title} with \entry*[date] text items (\par starts a new paragraph) and optional \subrubric{Heading}.
• Publications (CV section and the Publications page): the file with \printbibliography; entries come from the \addbibresource file in \nocite order (\nocite{*} = file order). \printbibliography[type=article, title={Journal Articles}] style splits are supported.

LaTeX it doesn't know is reported as a warning (the text is kept, the command dropped). To use a different source: python scripts/generate_cv.py --source path/to/folder-or.zip --output cv.html --publications-output publications.html.

Preview locally with Docker

The Docker image uses the same Jekyll version as GitHub Pages. From the repository root in PowerShell:

# once (and after editing the Dockerfile)
docker build -t my-jekyll-site .

# every time; pages rebuild when you save
docker run --rm -it -p 4000:4000 -p 35729:35729 -v "${PWD}:/srv/jekyll" my-jekyll-site

Then open http://localhost:4000/. Changes to _config.yml need a restart (Ctrl+C, run again).