Add Ohio State University mirror site (port 40015)

[richard-peng-xia]

TL;DR

Adds a fully functional osu.edu mirror site to WebHarbor at port 40015, with 20 benchmark tasks covering programs, news, events, faculty, research centers, athletics, and admissions.

Motivation

Ohio State University is one of the largest public universities in the US (60,000+ students), with a highly visited portal spanning academics, research, health sciences, and Big Ten athletics. It covers a domain — large public flagship university browsing — not represented in the existing 15 sites, and offers uniquely diverse multi-step tasks: professional degree programs (MD, JD, PharmD, DVM, OD), a full athletics section with 36 varsity sports, and a multi-campus structure (Columbus, Lima, Marion, Mansfield, Newark, Wooster).

Design

Flask application (sites/osu/app.py)

Ten SQLAlchemy models: User, College, Department, Program, NewsArticle, Event, ResearchCenter, Faculty, AthleticTeam, Bookmark. All seeded idempotently (seed function gated on College.query.first()).

Route coverage mirrors the real site's navigation:

• Homepage with featured news, upcoming events, and institutional stats
• News listing + article detail with category/search filters
• Academics overview listing all 16 colleges/schools
• Programs listing + detail with degree type, college, and online filters
• Events listing + detail with category, campus, and date filters
• Research centers listing + detail pages
• Departments listing (grouped by college) + detail with faculty roster
• Faculty listing + profile pages with department filter
• Athletics overview + team detail pages (unique to OSU vs other sites)
• Admissions overview with undergraduate/graduate/professional/online tabs
• About page with real OSU statistics
• Unified search across programs, news, events, and faculty
• Auth: login, register, logout, account (bookmarks)

Seed database

16 real OSU colleges (Arts & Sciences, Fisher Business, Moritz Law, Medicine, Veterinary Medicine, ...), 20 degree programs (including MD, JD, PharmD, DVM, OD — unique professional degrees), 20 news articles (2023–2025, 7 categories), 16 events (multiple campuses), 15 research centers (TDAI, Byrd Alzheimer's Center, Ohio Supercomputer Center, ...), 15 faculty, 26 athletic teams (all real OSU varsity sports), and 4 benchmark users (alice/bob/carol/dave @osu.edu, password: test1234). Seed DB generated at image build time via seed_data.py.

Templates

25 Jinja2 templates styled with OSU Scarlet (#BB0000) and Gray (#666666), modeled on real osu.edu layout: main nav (About / Academics / Research / Impact / Athletics), card-grid listings, detail pages with sidebars, paginated results (20 items/page). Includes an athletics section not present in other university mirrors.

Benchmark tasks (tasks.jsonl)

20 tasks (IDs Ohio State University--0 through Ohio State University--19) covering: professional degree program lookup (PharmD, DVM, OD), athletics records and team info, multi-campus event filtering, research center browsing, faculty research lookup, admissions requirements, and 3+ multi-step reasoning tasks.

Verification

Check	Result
py_compile sites/osu/app.py	✅
All 12 main routes → HTTP 200 (werkzeug test client)	✅
Seed: 16 colleges, 20 programs, 20 news, 16 events, 15 faculty, 15 centers, 26 teams, 4 users	✅
Seed idempotent (second run produces no duplicate rows)	✅
./scripts/build.sh webharbor:dev	requires Docker daemon — not run locally
POST /reset/osu + md5sum match	requires Docker daemon — not run locally

HuggingFace assets

OSU has no scraped image assets — all data is code-generated from seed_data.py. The instance_seed/osu.db is built directly inside the Docker image via a RUN step in the Dockerfile, eliminating the need for a HuggingFace tarball. No .assets-revision bump is required.

Registration

Site registered in all three required locations:

• websyn_start.sh SITES=(...) — index 15, port 40015
• control_server.py SITES=[...]
• Dockerfile EXPOSE 40000-40015

🤖 Generated with Claude Code