[Python-Dev] Changes to PEP 327: Decimal data type

[Michael Chermside]

Greg writes:
> I just did an experiment which involved running the Python
> script appended below over the Library Reference index.
> The results were:
> 
> Total names: 1908
> lower_names: 1684
> Percent lower: 88.2599580713

Unfortunately, there is a flaw in your methodology, and I
can't think of a reasonable way to correct for it. Consider
this code:

  class AllMixedCase:
      def thisIsMixedCase(): pass
      def thisIsAlsoMixed(): pass
      def mixed(): pass

  class AllLowerCase:
      def thisisalllowercase(): pass
      def thisisalsolowercase(): pass
      def lower(): pass

I'm guessing that your code counts 4 lowercase and 2 mixed
case. The problem is that a one-word mixedCase looks just
like a one-word lowercase. And when it's possible, using
a simple one-word name is a GOOD thing... so hopefully our
standard library is chock full of one-word functions.

I could also point out that in those cases where it ISN'T
possible to use one word, the alllowercaseformat is a real
pain to read, especially for non-native english speakers.
But that would start a coding-style war, and a major purpose
of having standards is to AVOID coding-style wars, so I'd
better not bring it up. ;-)

-- Michael Chermside