I may be guilty of thinking in too object-oriented a manner here, but I keep coming back to a certain pattern which doesn't look SQL-friendly at all.
Simplified description:-
I have an Attendance table, with various attributes including clockin, lunchout, lunchin, and clockout (all of type Time), an employee ID, and a date.
Different employees can have different shift hours, which depends on the day of the week and day of the month. The employee's categories are summarized in an Employees table, the shift hours etc. are summarized in an Hours table. A Leave table specifies half/full day leave taken.
I have raw check-in times from another source (third party, basically consider this a csv) which I sort and then need to insert to the Attendance table properly, by which I imagine I should be doing:-a) Take the earliest time as clockinb) Take the latest time as clockoutc) Compare remaining times with assigned lunch hours according to an employee's shift for the day
c) is the one giving me problems, as some employees check-in at odd times (perhaps they need to run out for an hour to meet a client), and some simply don't check-in for meals (eating in, etc). So I need to fuzzy-match the remaining times with assigned lunch hours, meaning I have to access the Hours table as well.
I'm trying to do all of this in a function within the Attendance class, so that it looks something like this:-
class Attendance(Base):
... my attributes ...
def calc_attendance(self, listOfTimes):
# Do the necessary
However it looks like I'd be required to do more queries within the calc_attendance function. Not even sure if that's possible, and certainly doesn't look like something I should be doing.
I could do multiple queries in the calling function, but that seems more messy. I've also considered multiple joins so the requisite information (for each employee per day) is available. Neither of them feel right, please do correct me though, if necessary.

View Solution