[Python-Dev] The fate of transform() and untransform() methods (original) (raw)

Alexander Belopolsky alexander.belopolsky at gmail.com
Thu Dec 9 15:51:36 CET 2010

On Fri, Dec 3, 2010 at 1:05 PM, Guido van Rossum <guido at python.org> wrote:

On Fri, Dec 3, 2010 at 9:58 AM, R. David Murray <rdmurray at bitdance.com> wrote: ..

I believe MAL's thought was that the addition of these methods had been approved pre-moratorium, but I don't know if that is a sufficient argument or not. It is not. The moratorium is intended to freeze the state of the language as implemented, not whatever was discussed and approved but didn't get implemented (that'd be a hole big enough to drive a truck through, as the saying goes :-). Regardless of what I or others may have said before, I am not currently a fan of adding transform() to either str or bytes.

I would like to restart the discussion under a separate subject because the original thread [1] went off the specific topic of the six new methods (2 methods x 3 types) added to builtins shortly before 3.2 beta was released. [2] The ticket that introduced the change is currently closed [3] even though the last message suggests that at least part of the change needs to be reverted.

Note that reverting part of the patch is not entirely trivial because new codecs' documentation refers to bytes.[un]transform() both in the docstrings and the library reference.

I think it will be the best to revert r86934 and resume the discussion of adding this functionality to 3.3 when we won't be constrained by the language moratorium. I will write a separate message with my thoughts about adding bytes codecs in 3.3. Let's keep this thread focused on what has to be done for 3.2.

[1] http://mail.python.org/pipermail/python-dev/2010-December/106353.html [2] http://svn.python.org/view?view=rev&revision=86934 [3] http://bugs.python.org/issue7475

More information about the Python-Dev mailing list