Support split Internet Archive items - #20

Support archives split across a base Internet Archive item and consecutive _extN items.

• parse and lint every IA item linked in a README row
• require archive URLs to be space-separated with no intervening text
• count zip files across all parts when checking ctlogsize
• require consecutive _extN names
• require 000.zip in base items while reading the first available zip from _extN items
• document the split-item convention with concrete identifiers

This is a companion to #19.

Tests:

• uv run --script .github/scripts/lint_ia.py
• table-cell canonicalization checks
• base and extension metadata zip-selection checks