fix rss: support rel=self and XML content-type feeds

[carlotran4]

Fixes DiscoverFeedURL to handle two cases:

1. XML content-type: When a URL already returns RSS/Atom/XML (e.g. techcrunch.com/tag/AI/feed/), use it directly instead of trying to parse HTML.

2. rel="self" links: Some feeds use rel="self" instead of rel="alternate" in their HTML. Now checks both.

This fixes TechCrunch tag/category feeds which return XML directly and use rel="self" links.

Before: DiscoverFeedURL would fall back to /feed for TechCrunch tag feeds.
After: Correctly returns the tag feed URL.

// Content-type check
if strings.Contains(contentType, "rss") || strings.Contains(contentType, "atom") || strings.Contains(contentType, "xml") {
    return blogURL, nil
}

// rel="self" fallback
selection = doc.Find(fmt.Sprintf("link[rel='self'][type='%s']", feedType)).First()