I fixed some issues in the new code, and activated *all* of the commented out tests (some of which fail, referring to existing or newly created) bug reports.
Some extra tests were created as well.

I might solve some of the issues I raised, but not right now.

I will be abroad (Uhh actually I will *not* be abroad, as I will be in my country of birth for a change) until after New Year, so it might take a while before I respond if there are issues.

As they say in Germany: "Guten Rutsch"

Regards
Anthon

On 2011-12-27 01:31, lkcl luke wrote:
On Mon, Dec 26, 2011 at 11:32 PM, Anthon van der Neut<[email protected]>  wrote:
I can fix some and can have some refer to Issue on the issue tracker (some
of that is done but not committed yet).
There is however also code that has nothing to do with format, but which
throws an exception.
  ah deep joy.

This is e.g. the case with
     INF = float('inf')
     self.assertEqual('{0:f}'.format(INF), 'inf')


I have changed this into:
     try:
         INF = float('inf')
     except ValueError:
             pass
     else:
         self.assertEqual('{0:f}'.format(INF), 'inf')

so that when the proper handling of float('inf') gets implemented, the tests
will run.

Or is there a better way to solve that kind of dependencies?

Would it not be a bit harsh to leave format mentioned in BUILTINS.TODO
because float() is incomplete and not have float() in that list?
  no, it wouldn't... unless... unless what you're saying is that
float's not correctly implemented, in which case there should be a
unit test for that [specifically named as such, and with an associated
bugreport]...

  _and_ that's the _only_ thing that's stopping format from being complete.

  if however there are more like this then it's probably best to *sigh*
raise bugreports for them all, link them to the bugreport on
format.... that's what bugtrackers are for.

  l.

Reply via email to