The Open Event Server provides backend support to Open Event Organizer Android App which is used to check-in attendees in an event. When checking in attendees, it is important for any event organizer to keep track of the device that was used to check someone in. For this, we provide an option in the Organizer App settings to set the device name. But this device name should have support in the server as well.
The problem is to be able to add device name data corresponding to each check-in time. Currently attendees model has an attribute called `checkin-times`, which is a csv of time strings. For each value in the csv, there has to be a corresponding device name value. This could be achieved by providing a similar csv key-value pair for “device-name-checkin”.
The constraints that we need to check for while handling device names are as follows:
- If there’s `device_name_checkin` in the request, there must be `is_checked_in` and `checkin_times` in the data as well.
- Number of items in checkin_times csv in data should be equal to the length of the device_name_checkin csv.
- If there’s checkin_times in data, and device-name-checkin is absent, it must be set to `-` indicating no set device name.
| if ‘device_name_checkin’ in data and data[‘device_name_checkin’] is not None: if ‘is_checked_in’ not in data or not data[‘is_checked_in’]: raise UnprocessableEntity( {‘pointer’: ‘/data/attributes/device_name_checkin’}, “Attendee needs to be checked in first” ) elif ‘checkin_times’ not in data or data[‘checkin_times’] is None: raise UnprocessableEntity( {‘pointer’: ‘/data/attributes/device_name_checkin’}, “Check in Times missing” ) elif len(data[‘checkin_times’].split(“,”)) != len(data[‘device_name_checkin’].split(“,”)): raise UnprocessableEntity( {‘pointer’: ‘/data/attributes/device_name_checkin’}, “Check in Times missing for the corresponding device name” ) if ‘checkin_times’ in data: if ‘device_name_checkin’ not in data or data[‘device_name_checkin’] is None: data[‘device_name_checkin’] = ‘-‘ |
The case is a little different for a PATCH request since we need to check for the number of items differently like this:
| if ‘device_name_checkin’ in data and data[‘device_name_checkin’] is not None: if obj.device_name_checkin is not None: data[‘device_name_checkin’] = ‘{},{}’.format(obj.device_name_checkin, data[‘device_name_checkin’]) if len(data[‘checkin_times’].split(“,”)) != len(data[‘device_name_checkin’].split(“,”)): raise UnprocessableEntity( {‘pointer’: ‘/data/attributes/device_name_checkin’}, “Check in Time missing for the corresponding device name”) |
Since we expect only the latest value to be present in a PATCH request, we first add it to the object by formatting using:
'{},{}'.format(obj.device_name_checkin, data['device_name_checkin'])
and then compare the length of the obtained CSVs for check in times and device names, so that corresponding to each check in time, we have either a device name or the default fill in value ‘-’.
That’s all. Read the full code here.
Requests and Responses:
Resources
- SQLAlchemy Docs
https://docs.sqlalchemy.org/en/latest/ - Alembic Docs
http://alembic.zzzcomputing.com/en/latest/ - Flask REST JSON API Classical CRUD operation
https://flask-rest-jsonapi.readthedocs.io/en/latest/quickstart.html#classical-crud-operations