Building SQL query

I wrote a messy function and I am wondering if you see any way I could clean it up. Essentially it takes in a list e_1=2&e_2=23&e_3=1 and makes a queryset out of it, maintaining the ordering.

from operator import itemgetter
def ids_to_students(items, prefix=0):

    if prefix == 0:
        # Make ordered QuerySets of Students based on there ids
        etuples   = sorted([(k,v) for k,v in items if k[:2] == 'e_'], key=itemgetter(0))
        ituples   = sorted([(k,v) for k,v in items if k[:2] == 'i_'], key=itemgetter(0))
        tuples = etuples+ituples
    else:
        tuples   = sorted([(k,v) for k,v in items if k[:2] == '%s_'%prefix], key=itemgetter(0))

    pk_list  = [v for (k,v) in tuples]
    clauses   = ' '.join(['WHEN id=%s THEN %s' % (pk, i) for i, pk in enumerate(pk_list)])
    ordering  = 'CASE %s END' % clauses
    students = Student.objects.filter(pk__in=pk_list).extra(
                select={'ordering': ordering}, order_by=('ordering',))
    return students

It's called like this:

students = ids_to_students(request.GET.items())
e_students = ids_to_students(request.GET.items(), 'e')

Tuples are sorted primarily by the first item of each. If they are all unique, specifying key=itemgetter(0) makes no difference.

This code

if prefix == 0:
    etuples   = sorted([(k,v) for k,v in items if k[:2] == 'e_'], key=itemgetter(0))
    ituples   = sorted([(k,v) for k,v in items if k[:2] == 'i_'], key=itemgetter(0))
    tuples = etuples+ituples
else:
    tuples   = sorted([(k,v) for k,v in items if k[:2] == '%s_'%prefix], key=itemgetter(0))

can be rearranged like this to avoid repetition:

prefixes = ['%s_'%prefix] if prefix else ['e_', 'i_']
tuples = sorted(item for item in items if item[0][:2] in prefixes)

StackExchange Code Review Q#45023, answer score: 6