Enumerate a BACnet device's object-list and each object's property-list, handling segmentation limits and very long object lists

Documented steps

1. Locate the device: i_ams = await app.who_is(device_instance, device_instance), then take i_am.pduSource as the address and i_am.iAmDeviceIdentifier as the device ObjectIdentifier.
2. Try the fast path first: object_list = await app.read_property(device_address, device_identifier, 'object-list') to pull the whole array in one ReadProperty.
3. Catch AbortPDU and inspect err.apduAbortRejectReason. If it equals AbortReason.segmentationNotSupported, the object-list is too large for an unsegmented response and you must page it.
4. Fallback: read the length with array_index=0, then loop for i in range(length): await app.read_property(device_address, device_identifier, 'object-list', array_index=i+1) to pull each ObjectIdentifier one at a time.
5. Resolve each object's class through vendor info: vendor_info = get_vendor_info(i_am.vendorID); object_class = vendor_info.get_object_class(object_identifier[0]). Log and skip proprietary types that do not resolve rather than aborting.
6. Read each object's 'property-list' to learn which optional properties it actually implements before attempting to read them, instead of guessing from the object type.
7. Only read properties that appear in the property-list or in object_class._required, and wrap each read in try/except ErrorRejectAbortNack with log-and-continue semantics.
8. Cache the result and use the device object's 'database-revision' property to detect when a re-enumeration is actually needed.
9. Access requirement: none at the protocol level; enumeration is unauthenticated.
10. Official documentation verified 2026-07-30: https://raw.githubusercontent.com/JoelBender/BACpypes3/main/samples/discover-objects.py | https://raw.githubusercontent.com/JoelBender/BACpypes3/main/bacpypes3/apdu.py | https://raw.githubusercontent.com/JoelBender/BACpypes3/main/bacpypes3/service/object.py