[Python-3000] str.format vs. string.Formatter exceptions

[Eric Smith]

Ron Adam points out some differences in which exceptions are thrown by 
str.format and string.Formatter.  For example, on a missing positional 
argument:

 >>> "{0}".format()
Traceback (most recent call last):
   File "<stdin>", line 1, in <module>
ValueError: Not enough positional arguments in format string

 >>> Formatter().format("{0}")
Traceback (most recent call last):
   File "<stdin>", line 1, in <module>
   File "/shared/src/python/py3k/Lib/string.py", line 201, in format
     return self.vformat(format_string, args, kwargs)
   File "/shared/src/python/py3k/Lib/string.py", line 220, in vformat
     obj, arg_used = self.get_field(field_name, args, kwargs)
   File "/shared/src/python/py3k/Lib/string.py", line 278, in get_field
     obj = self.get_value(first, args, kwargs)
   File "/shared/src/python/py3k/Lib/string.py", line 235, in get_value
     return args[key]
IndexError: tuple index out of range

The PEP says: In general, exceptions generated by the formatter code 
itself are of the "ValueError" variety -- there is an error in the 
actual "value" of the format string.

I can easily change string.Formatter to make this a ValueError, and I 
think that's probably the right thing to do.  For example, if the string 
comes from a translation module, then there might be an extra parameter 
added by mistake, in which case ValueError seems right to me.

But I'd like to hear if anyone else thinks this should be an IndexError, 
or maybe they both should be some other exception.

Similarly "{x}".format()' currently raises ValueError, but 
'Formatter().format("{x}")' raises KeyError.