Message123275
| Author |
belopolsky |
| Recipients |
belopolsky, eric.smith, ezio.melotti, lemburg, mark.dickinson, skrah |
| Date |
2010年12月03日.17:54:26 |
| SpamBayes Score |
1.1905382e-05 |
| Marked as misclassified |
No |
| Message-id |
<AANLkTine-fsj68FPmGWjNh+9pvhVsFRBEa_zivqx1f2H@mail.gmail.com> |
| In-reply-to |
<1291344829.29.0.413309910585.issue10557@psf.upfronthosting.co.za> |
| Content |
On Thu, Dec 2, 2010 at 9:53 PM, Alexander Belopolsky
<report@bugs.python.org> wrote:
..
> .. The honest reason for the exclusion is that I gave up chasing a bug that only shows
> in full regrtest runs.
I have realized where the problem was. PyUnicode_FromUnicode()
"helpfully" interns single-character Unicode objects in the Latin-1
range. So when TransformDecimal replaces whitespace with ' ', it may
garble cached strings. I think this optimization is a left-over from
the time when unicode did not have interning, but it is never a good
idea to change immutable objects in-place after creation, so I'll fix
this problem in my code. |
|