fix(person): normalize URL trailing slash to prevent urljoin path loss

[dtateks]

Summary

• urljoin() drops the username path segment when the input URL lacks a trailing slash, causing 404 errors on all detail page navigations
• Fix: normalize linkedin_url to always end with / at the start of PersonScraper.scrape()

Problem

from urllib.parse import urljoin

# WITHOUT trailing slash — BROKEN
urljoin("https://linkedin.com/in/username", "details/patents/")
# → "https://linkedin.com/in/details/patents/"  ← username lost, 404!

# WITH trailing slash — CORRECT
urljoin("https://linkedin.com/in/username/", "details/patents/")
# → "https://linkedin.com/in/username/details/patents/"

This affects all fallback navigations in PersonScraper:

• details/experience
• details/education
• details/interests/
• details/{certifications,honors,patents,...}/ (8 accomplishment sections)
• overlay/contact-info/

Fix

4-line change at the top of scrape():

if not linkedin_url.endswith("/"):
    linkedin_url = linkedin_url + "/"

All downstream urljoin(base_url, ...) calls then produce correct URLs.

Testing

• All 11 unit tests pass
• Manually verified URL generation for all affected paths