[Python-Dev] 2.1 alpha: what about the unicode name database?

[Finn Bock]

[Fredrik Lundh]

> The name database portions of SF task 17335 ("add
> compressed unicode database") were postponed to
> 2.1.
> 
> My current patch replaces the ~450k large ucnhash
> module with a new ~160k large module.  (See earlier
> posts for more info on how the new database works).
> 
> Should I check it in?

[M.-A. Lemburg]

>Since the Unicode character names are probably
>not used for performance sensitive tasks, I suggest to
>checkin the smallest version possible.
>
>If it is too much work to get Finn's version recoded in C
>(presuming it's written in Java), then I'd suggest checking
>in your version until someone comes up with a yet smaller
>edition.

FWIW, I agree the that 160k module should be used. Please, nobody should
use the jython compression as an argument to delay any improvements in
CPython. 

I certainly didn't post because I wanted to complicate your processes. I
just wanted to show off <wink>.

regards,
finn