郭子扬

郭子扬

Southeast University

CLAUDE.md

This file provides guidance to Claude Code (claude.ai/code) when working with code in this repository.

Repository Overview

This is a Jekyll-based academic personal homepage for Ziyang Guo (郭子扬), forked from RayeRen/acad-homepage.github.io (the “AcadHomepage” template) and migrated to the RerankerGuo GitHub account. The site is hosted at https://RerankerGuo.github.io.

The site has two distinguishing features beyond a vanilla Jekyll template:

  1. Auto-refreshing Google Scholar citations via a Python crawler + GitHub Action.
  2. Per-paper citation counts rendered client-side from a JSON file on a separate branch.

Local Development

bash run_server.sh        # runs `bundle exec jekyll liveserve` → http://127.0.0.1:4000

Requires Ruby + Bundler. First run: bundle install. The _config.yml is not auto-reloaded by jekyll serve — restart the server after editing it.

_site/ is the build output and is gitignored.

Architecture

Content vs. layout separation

Layer Location Edit when…
Site-wide config _config.yml Author info, repo URL, analytics IDs, plugins, SEO verifications
Single-page content _pages/about.md Anything a visitor reads (bio, news, projects, awards)
Sidebar nav links _data/navigation.yml Adding/removing top-nav anchors
Styling _sass/*.scss (compiled into _sass/_base.scss etc.) Visual changes — _sidebar.scss is the most recently customized file
Page chrome _layouts/default.html + _includes/*.html Header, sidebar, scripts, SEO meta

_config.yml sets defaults so every page under _pages/ gets layout: default and author_profile: true automatically.

Google Scholar citation pipeline (the non-Jekyll half)

Two independent pieces run on different schedules and live in different branches:

  1. google_scholar_crawler/main.py — Python script using scholarly + jsonpickle. Reads GOOGLE_SCHOLAR_ID env var, fetches the author profile, and writes two JSON files to results/:
    • gs_data.json — full profile incl. publications keyed by author_pub_id
    • gs_data_shieldsio.json — { "schemaVersion": 1, "label": "citations", "message": "<citedby>" } for shield-style badges
  2. .github/workflows/google_scholar_crawler.yaml — runs on page_build and daily at 08:00 UTC. Executes main.py, then force-pushes results/*.json to the google-scholar-stats branch (note: this is a branch, not a folder in main). The script exits 0 silently when GOOGLE_SCHOLAR_ID is unset, so missing-secret builds don’t fail.

Citation rendering on the page

The frontend never blocks on the citation JSON:

  • _config.yml toggles google_scholar_stats_use_cdn (default true) — when on, the page reads from cdn.jsdelivr.net/gh/{repository}@google-scholar-stats/... instead of raw.githubusercontent.com.
  • _includes/fetch_google_scholar_stats.html (loaded via _layouts/default.html) does the $.getJSON and writes the total into #total_cit.
  • To show a per-paper citation count, drop <span class='show_paper_citations' data='AUTHOR_PUB_ID'></span> into _pages/about.md. The AUTHOR_PUB_ID is the value of citation_for_view=... from the paper’s Google Scholar URL.

Fork-specific deviations from upstream AcadHomepage

  • site.repository in _config.yml is RerankerGuo/RerankerGuo.github.io — must stay in sync with the actual GitHub repo name or the CDN citation URLs 404.
  • _config.yml adds the hawkins gem alongside the standard Jekyll plugins.
  • _sass/_sidebar.scss was restyled locally (see commit 5afbafd).
  • google_scholar_crawler/main.py was patched (commit 042bece) to skip cleanly when the secret is absent rather than crashing.
  • Sidebar profile (_includes/author-profile.html) and SEO include (_includes/seo.html) have minor local tweaks.

Common Tasks

Update author info or social links → _config.yml under author:. Edit page content → _pages/about.md (single page; permalink: /). Add a nav anchor → _data/navigation.yml, then put a matching <span class='anchor' id='anchor-id'></span> in _pages/about.md. Change a color/layout detail → _sass/_variables.scss first (defines palette), then specific partials. Refresh citations manually → push a commit to main (triggers page_build); or wait for the 08:00 UTC cron. Verify the JSON landed on the google-scholar-stats branch. Set up citations on a fresh fork → add the GOOGLE_SCHOLAR_ID repo secret and enable Actions.

Things to avoid

  • Don’t commit _site/ — gitignored.
  • Don’t edit Gemfile.lock by hand; bundle update <gem> to modify.
  • The google-scholar-stats branch is overwritten by the workflow on every run — don’t base feature branches on it.
  • compress_html is configured to skip compression in development env only, so production HTML is minified.