It is ok to add FieldGetDbPrepValueMixin to IExact lookup?

Django Internals ORM
1

Context

I’m dealing with a legacy Oracle database with a column of type CHAR, which adds leading whitespaces on every query and it causes matching issues with the ORM (i.e. SomeModel.objects.filter(somefield="text") can’t find some results because “text” != "text "). I got a a solution overriding from_db_value() and get_db_prep_value() method of Charfield (see below, on LegacyCharField) and it works pretty well in most of the cases except for iexact lookups.

Main Question

Some django builtin lookups (see django/db/models/lookups.py) like ‘exact’, ‘lt’, ‘gte’ has the FieldGetDbPrepValueMixin, which transforms the field value before sending to the database.

However, the lookup iexact doesn’t have FieldGetDbPrepValueMixin. That’s very odd and I can’t figure it why (maybe someone just forgot to add the mixin?). It is ok to add the mixin into the IExact lookup? Or should I take a different approach?

@Field.register_lookup
class IExact(BuiltinLookup):
    ...

My custom CharField:

class LegacyCharField(models.CharField):
    def db_type(self, connection):
        if connection.vendor == 'oracle':
            return 'CHAR'
        else:
            return super().db_type(connection)

    def from_db_value(self, value, expression, connection):
        if connection.vendor == 'oracle':
            if value is None:
                return value
            # Strip whitespaces
            return value.rstrip()
        else:
            return value

    def get_db_prep_value(self, value, connection, prepared=False):
        value = super().get_db_prep_value(value, connection, prepared)
        if connection.vendor == 'oracle':
            if value is None:
                return value
            # Add whitespaces
            return bytes(value, encoding='utf-8') \
                .ljust(self.max_length) \
                .decode(encoding='utf-8')
        else:
            return value