On Tue, 8 Feb 2005 10:10:49 +0100, Fredrik Lundh <[EMAIL PROTECTED]> wrote:
>
> >> @@ -399,9 +393,8 @@
> >> del self[name] # Won't fail if it doesn't exist
> >> self.dict[name.lower()] = value
> >> text = name + ": " + value
> >> - lines = text.split("\n")
> >> - for line in lines:
> >> - self.headers.append(line + "\n")
> >> + self.headers.extend(text.splitlines(True))
> >> + self.headers.append('\n')
> >
> > and you're 100% sure that the change in how things are stored
> > in headers won't affect any existing code?
> >
> > (the docstring says that headers contain a list of lines, which is no
> > longer true)
>
> and the module documentation says:
>
> Each line contains a trailing newline. The blank line terminating
> the headers is not contained in the list.
>
> which is no longer true (unless I'm missing something here)
This would have been caught if there was a unit test validating what
the documentation says. Why aren't there unit tests for this code? I
think we need to raise the bar for "wholistic" improvements to a
module: first write a unit test if there isn't already one (and if
there is one, make sure that it tests all documented behavior), *then*
refactor. Yes, this would be less fun. It's not supposed to be fun.
It's supposed to avoid breaking code.
Raymond, please roll back that change until this is taken care of.
--
--Guido van Rossum (home page: http://www.python.org/~guido/)
_______________________________________________
Python-Dev mailing list
Python-Dev@python.org
http://mail.python.org/mailman/listinfo/python-dev
Unsubscribe:
http://mail.python.org/mailman/options/python-dev/archive%40mail-archive.com
