Change allow_existing default to True so duplicate fingerprint matches
return the existing device (200) instead of 409 Conflict. Add model
validator to force allow_existing=False when allow_duplicate is set.
Add model_validator to SaveSchema that safely handles SQLAlchemy lazy
relationships by checking inspect(obj).unloaded before attribute access.
This allows direct use of model_validate(orm_obj) instead of manually
building a dict and excluding device_syncs.
Add tests for duplicate device registration scenarios (409 conflict,
allow_existing, allow_duplicate, reset_syncs). Fix compute_file_hash
function references after relocation to assets_handler.
Move compute_file_hash, compute_zip_hash, and compute_content_hash from
scan_handler.py to filesystem/assets_handler.py as standalone module-level
functions. This follows the existing pattern for utility functions in
filesystem handlers.
Add order_by and order_dir parameters to get_saves() for flexible
sorting. Supports "updated_at" and "created_at" fields with "asc" or
"desc" direction (default: desc). Enables ascending order for pruning
scenarios.
Replace individual Body() parameters with DeviceCreatePayload and
DeviceUpdatePayload Pydantic models. This simplifies the function
signatures and leverages model_dump(exclude_unset=True) for cleaner
update handling.
Add fingerprint-based detection for duplicate device registration with
configurable behavior via new body params:
- allow_existing: return existing device if fingerprint matches
- allow_duplicate: skip fingerprint check, always create new device
- reset_syncs: clear tracked saves when reclaiming existing device
Fingerprint matching uses mac_address (primary) or hostname+platform
(fallback). Returns 409 Conflict with device_id when duplicate detected
without flags, 200 OK for existing device, 201 Created for new.
- Add device registration and save synchronization
- Implement slot-based save organization with datetime tagging
- Add conflict detection for multi-device sync scenarios
- Add content hash computation for save deduplication
- Support ZIP inner-file hashing for consistent deduplication
- Add confirm_download endpoint for sync state management
- Add overwrite parameter to bypass conflict checks
Some archives have an empty "Attributes" field, i.e. `Attributes = `.
But because we the line is strip, it became `Attributes =` and the
`elif` branch was never taken, resulting with `largest_file` being
`None` after processing.
