Re: improved charset support.zip

From: "Ingo Brueckl" <ib@wupperonline.de>
Sent: Tuesday, June 01, 2010 8:05 PM
To: <abiword-dev@abisource.com>
Subject: Re: improved charset support.zip

>
> Hello,
>
> I endorse a possibility of having a codepage detection and don't like a fixed
> one either, but I really don't see how this detection from the font name will
> work. On all Windows PCs I know, "Arial" is named "Arial", for example, and
> that's the way the font is stored in the .wri document. Isn't that true for
> Eastern European Windows, too?
>
> And even if there are different font names, a default of 1250 isn't
> advisable, because my "Arial" (not detected through cp_suf_tbl) is most
> certainly 1252.
>
> AFAIK there is no information in the .wri file allowing to decide which
> codepage would be right. So the only way to go seems to be to either make
> this configurable or to detect it from the underlaying system abiword is
> running on. And even this wouldn't allow to open all .wri documents
> correctly, so configuration seems to be safest way.
>
> The patch shouldn't be applied (with this detection method).
>

In Windows 3.1 there are three versions of standard fonts: Arial for 1252, "Arial CE" for 1250 and "Arial Cyr" for 1251. The same true for Courier New and Times New Roman.
As for my patch, the default codepage in it is *1252* which should suffice for most documents. Ideally we should give to user an opportunity to specify his custom (old) font charsets in some kind of configuration file, but I cannot see ways to implement this currently.
I don't think we should force system default codepage on user and my decision seems optimal to me.
Received on Tue Jun 1 17:51:17 2010