fix: Improve GTFS metrics reliability and test accuracy

[Copilot]

Addresses code review feedback from PR #49 to fix metrics accuracy and error handling gaps in GTFS processing tasks.

Changes

• Error handling: Wrapped table import loop in try-except to capture CSV parsing and database insertion failures, preventing silent errors
• Metrics accuracy: Moved GTFS_SCHEDULE_UPDATES_TOTAL increment from before import to after successful completion - counter no longer increments on failed imports
• Test correctness: Fixed test_schedule_updates_counter to use provider label instead of incorrect feed_id label

Example

Before:

GTFS_SCHEDULE_UPDATES_TOTAL.labels(provider=company).inc()  # Line 82 - increments before import
for table_name in tables.keys():
    # ... import logic with no error handling

After:

try:
    for table_name in tables.keys():
        # ... import logic
    GTFS_SCHEDULE_UPDATES_TOTAL.labels(provider=company).inc()  # Line 122 - after success
except Exception as e:
    GTFS_PROCESSING_ERRORS_TOTAL.labels(task="get_schedule", provider=company).inc()
    logging.error(f"Error importing schedule data: {e}")
    return "Error importing schedule data"

---

✨ Let Copilot coding agent set things up for you — coding agent works faster and does higher quality work when set up for your repo.