afrobeard <[EMAIL PROTECTED]> writes: > Arnaud's code wont work if self.opt1 is None, an empty list, an empty > tuple, False, etc, because all these evaluate to false. They wont > print the internal state of these variables. [Just an informational > notice, this may be the behavior you expect]
??? My suggestion is to get rid of attributes altogether and does not test any truth values. > Secondly, I'm not sure if you know the variable names from before hand > in which case Casey's approach will work, or you need to know them via > introspection. http://www.ibm.com/developerworks/library/l-pyint.html > [Scroll down to attributes]. > > On May 16, 1:44 am, Arnaud Delobelle <[EMAIL PROTECTED]> wrote: >> Casey <[EMAIL PROTECTED]> writes: >> > Hi, >> >> > I have some classes that print variable outputs depending on their >> > internal state, like so: >> >> > def __str__(self): >> > out = [] >> > if self.opt1: out += ['option 1 is %s' % self.opt1'] >> > if self.opt2: out += ['option 2 is %s' % self.opt2'] >> > .... >> > return '\n'.join(out) >> >> > Is there any way to make this cleaner? >> >> Maybe. >> >> Have a dictionary of options rather than individual attributes; >> options not in the dictionary are not set. E.g. >> >> mask = { >> 'opt1': 'option 1 is %s', >> 'opt2': 'option 2 is %s', >> ... >> } >> >> def __str__(self): >> return '\n'.join(mask[o] % v for o,v in self.options.iteritems()) >> >> -- >> Arnaud -- http://mail.python.org/mailman/listinfo/python-list
