Make GCS optional for transfer elevation cache and cleanup

[jeremyzilar]

Summary

Makes Google Cloud Storage (GCS) optional for the data transfer pipeline. When GCS_BUCKET_NAME is not set, the transfer runs without GCS instead of failing with ValueError: Cannot determine path without bucket name.

Problem

Running python -m transfers.transfer locally (without GCS credentials) failed during well transfer initialization. WellTransferer calls get_cached_elevations() to load a precomputed elevation cache from GCS. If GCS_BUCKET_NAME is empty or unset, the GCS client creates a bucket object with no name, and blob.exists() raises:

ValueError: Cannot determine path without bucket name.

This blocked local development and testing of the transfer without GCS.

Solution

Guard all GCS usage in well_transfer_util.py with a check for GCS_BUCKET_NAME:

• get_cached_elevations() – Returns an empty dict when GCS is not configured. The transfer proceeds without a cache; elevation lookups may be slower or use defaults.
• dump_cached_elevations() – No-op when GCS is not configured. The in-memory cache is not persisted.
• cleanup_locations() – Skips GCS read/write when GCS is not configured. Location updates (state, county, quad_name) still run using in-memory lookups; only the GCS-backed cache is skipped.

Behavior

Scenario	Before	After
GCS_BUCKET_NAME set	Uses GCS for elevation cache and location cleanup cache	Same behavior
GCS_BUCKET_NAME unset or empty	Transfer crashes on well transfer init	Transfer runs; GCS operations are skipped

Testing

• Run python -m transfers.transfer with GCS_BUCKET_NAME unset in .env; the transfer should complete successfully.
• With GCS_BUCKET_NAME set and valid credentials, the transfer should still use GCS for elevation and location cleanup caches.

[jirhiker]

I'd like to minimize any modification or continued use of the transfer functionality. For this purpose I implemented PR #597.

[jirhiker]

see inline comment