geonames: add non-ASCII/international query terms

`geonames/terms.txt` contains 45 ASCII-only place names. The Geonames dataset is global and includes names in Arabic, Chinese, Japanese, Cyrillic, and accented Latin scripts. The current terms do not exercise Solr's Unicode text analysis pipeline.

## Tasks
- Extend `terms.txt` with a representative set of non-ASCII place names drawn from the corpus
- Verify the existing analyzer chain handles them correctly (tokenisation, lowercasing, stop words)

**File:** `geonames/terms.txt`

Start with geonames/terms.txt and the referenced Geonames corpus; inspect the current 45 entries and analyzer behavior for representative Arabic, Chinese, Japanese, Cyrillic, and accented Latin terms. Done means terms.txt includes representative corpus names and verification covers tokenisation, lowercasing, and stop words.

Written by the indexing model from the issue text.