Surface ipinfo.io location check failures (closes #1311)

[kunal-10-cloud]

Summary

Closes #1311.

The ipinfo.io lookup performed in AnophelesBase.__init__ swallowed every OSError silently, leaving _client_details = None. That cascaded into _get_gcp_region() returning None, which in turn caused the GCS URL optimisation (selecting a region-local bucket) to fall back to the default bucket — without any indication that the location check had failed.

The failure mode was indistinguishable from a successful lookup that returned no location data, which is exactly the situation #1311 describes.

Why it matters

• Performance: Researchers near a GCS region (e.g. us-central1) could unknowingly pull multi-GB datasets from a distant bucket.
• Debuggability: Slow-download reports had nothing in the application state or logs to correlate with a failed ipinfo.io lookup.
• Prevalence: The networks most likely to block external lookups (corporate firewalls, institutional VPNs, restricted regions) are exactly the environments where many researchers operate.

Change

malariagen_data/anoph/base.py, lines 100–105.

Before:

```python
self._client_details = None
if check_location:
try:
self._client_details = ipinfo.getHandler().getDetails()
except OSError:
pass
```

After:

```python
self._client_details = None
if check_location:
try:
self._client_details = ipinfo.getHandler().getDetails()
except OSError as exc:
self._log.debug(f"Location check failed: {exc}")
warnings.warn(
"Could not determine client location via ipinfo.io "
f"({exc}). Region-optimal GCS bucket selection is "
"unavailable. You can manually specify a region using "
"the `url` parameter.",
UserWarning,
stacklevel=2,
)
```

The fallback behaviour itself is unchanged; the fix is purely additive.

• self._log.debug(...) captures the underlying exception for users running with debug=True.
• warnings.warn(..., UserWarning, stacklevel=2) surfaces the failure to end users with an actionable hint about the url override, without being intrusive in the common success path.
• warnings is already imported at the top of the module; self._log is initialised on the line immediately preceding this block, so the warning is safe to emit in __init__.

Test plan

• pre-commit run --files malariagen_data/anoph/base.py — passed (ruff, ruff-format, trailing-whitespace, end-of-file)
• poetry run mypy malariagen_data/anoph/base.py — passed (no issues)
• poetry run pytest tests/anoph/test_base.py — 44 passed
• CI matrix on PR (linting, coverage, tests across Python 3.10/3.11/3.12 × numpy 2.0.2 / >=2.0.2,<2.1)

Backwards compatibility

No public API changes. No change to fallback behaviour. The only observable difference is a UserWarning (and a DEBUG log entry when debug=True) when the ipinfo.io lookup fails, which previously produced no signal at all.

[Copilot]

Pull request overview

This PR makes failures in the ipinfo.io-based client geolocation check visible to users, so region-optimal GCS bucket selection doesn’t silently degrade when the lookup fails (closes #1311).

Changes:

• Add DEBUG logging when the ipinfo.io lookup raises OSError.
• Emit a UserWarning with actionable guidance when the location lookup fails, while keeping the existing fallback behavior unchanged.

---

💡 Add Copilot custom instructions for smarter, more guided reviews. Learn how to get started.