screen-devel
[Top][All Lists]
Advanced

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: [screen-devel] Bug#882340: screen: please update wcwidth/Unicode dat


From: Thorsten Glaser
Subject: Re: [screen-devel] Bug#882340: screen: please update wcwidth/Unicode data [origin: address@hidden
Date: Sun, 15 Jul 2018 12:13:53 +0000 (UTC)

Hi XTaran,

>this patch has been submitted to Debian with the request to forward it
>to upstream which I hereby do.

thanks!

Note that the patch needs updating for Unicode 11…

>While I generally trust the submitter as well consider him competent
>with regards to charset and especially Unicode encodings, I've not
>much of an idea what the patch actually changes and I'm very reluctant
>to update defacto binary code which I don't really understand and

Indeed, it’s generated.

>I assume one of the upstream devs is more comptent than me with that

I’ve got several programs which need updating their code whenever
a new Unicode version comes out. XTerm in Debian, however, just
uses glibc’s wcwidth data, so updating that makes the system con‐
sistent except for GNU screen. The way encoding.c handles the data
tables is, though, a bit different from the way other applications
(such as XTerm upstream or (Mir)BSD libc) handle them, so I was
not sure how to best deal with that. I’d prefer if GNU screen could
use the system wcwidth data, but I was not confident enough in my
understanding of the code to be able to implement that, hence, this
patch is a mere table update.

I’m not entirely sure how XTerm gets away with using glibc’s
wcwidth() given POSIX locales (but at least, glibc defines
__STDC_ISO_10646__ these days). Perhaps it always operates
in a UTF-8 locale itself. I’m not sure if this can be done
by GNU screen, but considering that it already supports
translating (screen -U, then have a window with :utf8 off)
it might.


The generation script I am using is consistent with GNU glibc’s
as of version 2.27 and current master (Mike FABIAN and I have
been synchronising our respective scripts until the differences
for all actually assigned and representable Unicode characters
went down to 0).

bye,
//mirabilos
-- 
I believe no one can invent an algorithm. One just happens to hit upon it
when God enlightens him. Or only God invents algorithms, we merely copy them.
If you don't believe in God, just consider God as Nature if you won't deny
existence.              -- Coywolf Qi Hunt



reply via email to

[Prev in Thread] Current Thread [Next in Thread]