[ContentUnderstanding] Precise page number assertions in content range tests

[Copilot]

Content range assertions in test_sample_analyze_binary.py were too loose — checking > 0 or >= 2 instead of exact counts, and not verifying which pages were actually extracted.

Changes

• Exact page count assertions: Replace loose bounds with assert == for all content range scenarios ("3-" → full_page_count - 2, "1-3,5,9-" → full_page_count - 4, "1,3-4" → 3)
• Individual page number verification: Extract and sort page_number from each returned page, compare against expected list (e.g., "1-3" → [1, 2, 3], "1,3-4" → [1, 3, 4])
• Endpoint assertions: Verify start_page_number/end_page_number on DocumentContent for all range scenarios

# Before: loose assertion
assert combine2_page_count >= 2

# After: exact count + specific page numbers
assert combine2_page_count == 3
actual_combine2_pages = sorted([p.page_number for p in combine2_doc.pages])
assert actual_combine2_pages == [1, 3, 4]

---

🔒 GitHub Advanced Security automatically protects Copilot coding agent pull requests. You can protect all pull requests by enabling Advanced Security for your repositories. Learn more about Advanced Security.