Question

I’ve made a few Badge classes in Django, each containing some sort of description in a string variable:

"You get this badge because you've runned %d meters in %d minutes"
"You get this badge because you've killed %d monsters of the type %s"

etc. And the classes also have a function get_description(badge_level_requirements), so in the templates it will be called together with a list to assemble the string for a specific user:

class RunnerBadge(Badge):
    des=ugettext_lazy("You get this badge because you've runned %d meters in %d minutes")
    def get_description(cls,badge_level_requirements):
        return cls.des%badge_level_requirements

And I’ve stored the requirements lists in the database without any argument names already 🙁 As shown in the examples, different classes have different numbers of values to fill in the string, the values mean different things as well. So I can’t really name the arguments.

However, if I want to internationalize these strings, there’ll be errors: 'msgid' format string with unnamed arguments cannot be properly localized And the language file cannot be generated for this matter.

Is there a way to bypass this error?

Update

I’ve come across this method for bypassing the error without changing database.
In database, the level requirements are stored in a text field in the format of dict:

#Requirment of Runner's badge
"{'gold':(100,10),'silver':(50,5),'bronze':(25,2)}"

And in class definition, mannually add argument names as ‘arg_0’,’arg_1’… to descriptions. The get_description method is changed to pre-process the data before used for filling description strings.

class RunnersBadge(Badge):
    requirements=#get the previous dict from database
    description="You get this badge because you've runned %(arg_0)d meters in %(arg_1)d minutes"

    @classmethod
    def get_description(cls,level):
        '''
        This is actually a method of parent class Badge
        level is either 'gold','silver' or 'bronze'
        '''
        dic={}
        try:
            for (num,val) in enumerate(cls.requirements[level]):
                dic['arg_'+str(num)]=val
        except TypeError:
            dic['arg_0']=cls.requirements[level]


        return cls.description%dic

This method keeps most of the current structure (logic and database). And the translator just need to take care of the words’ placement.

Answer 1

1. like in code, variable names should be meaningful within their context, ‘meter_count’ and ‘minute_count’ are explicit, compared to ‘arg_0’ and ‘arg_1’ which are meaningless

2. use standard translation in python code, it is less error prone and is recognized by the amazingly useful makemessages command

3. use use named-string interpolation (e.g., %(day)s) instead of positional interpolation (e.g., %s or %d) whenever you have more than a single parameter, because the order of parameters can change depending on the language. I.e. germanic and latin languages inverse the noun/adjective order, dates are displayed differently depending on the language, etc, etc …

4. use ran instead of runned, double check the syntactical validity of your english translation strings

This:

class RunnersBadge(Badge):
    requirements=#get the previous dict from database
    description="You get this badge because you've runned %(arg_0)d meters in %(arg_1)d minutes"

Becomes:

from django.utils.translation import ugettext as _

class RunnersBadge(Badge):
    requirements=#get the previous dict from database
    description=_("You get this badge because you've ran %(meter_count)d meters in %(minute_count)d minutes")