Add language-aware filtering to random book feature

[KrishnaShuk]

Closes #10810

Feature: Add language-aware filtering to the random book feature (/random endpoint)

When users select a non-English language for the Open Library interface, the random book feature now prioritizes showing books with editions available in their selected language. This applies only to "populated languages" (those with >15k borrowable ebooks), with automatic fallback to all languages for unpopulated languages.

Technical

Implementation follows the existing QueryCarousel.html pattern:

• Modified get_random_borrowable_ebook_keys() to accept optional language parameter
• Added per-language caching with dynamic keys: home.random_book.{lang} (e.g., home.random_book.fre, home.random_book.eng, home.random_book.all)
• Updated random_book.GET() to:
  • Extract user language from web.ctx.lang
  • Convert ISO 639-1 to MARC21 format using convert_iso_to_marc()
  • Validate language is in get_populated_languages() set
  • Pass language to Solr query only for populated languages
• Solr query becomes: type:edition AND ebook_access:[borrowable TO *] AND language:{lang} when language is populated

Populated languages (10 total): eng, fre, ger, spa, chi, ita, lat, dut, rus, jpn

Fallback behavior: For unpopulated languages (e.g., Hindi), the feature falls back to showing books from all languages, ensuring users always get results.

(cci:7://file:///home/krishna/Contribution/openlibrary/openlibrary/plugins/openlibrary/home.py:0:0-0:0)

Stakeholders

@cdrini