From f5a4fce6f0bc62121cc82dc66aa62b7cd1408c0c Mon Sep 17 00:00:00 2001 From: William Kendrick Date: Tue, 14 Sep 2004 08:00:27 +0000 Subject: [PATCH] Updated docs to correspond to recent 'uppercase' change (attempt ;^) ) --- docs/CHANGES.txt | 3 +++ docs/TODO.txt | 22 ++++++++++++---------- 2 files changed, 15 insertions(+), 10 deletions(-) diff --git a/docs/CHANGES.txt b/docs/CHANGES.txt index 321c02299..023e11643 100644 --- a/docs/CHANGES.txt +++ b/docs/CHANGES.txt @@ -8,6 +8,9 @@ http://www.newbreedsoftware.com/tuxpaint/ 2004.September.13 (0.9.14) [cvs] + * Initial work to get proper uppercase support in languages other + than English. + * Added lockfile support, to prevent Tux Paint from being launched more than once every 30 seconds. (Disable with "--nolockfile" command-line argument.) Thanks to Darci Lindgren for suggestion the feature, and diff --git a/docs/TODO.txt b/docs/TODO.txt index 30b13d5bc..6c5c01c82 100644 --- a/docs/TODO.txt +++ b/docs/TODO.txt @@ -6,7 +6,7 @@ Copyright (c) 2004 by Bill Kendrick bill@newbreedsoftware.com http://www.newbreedsoftware.com/tuxpaint/ -September 7, 2004 +September 14, 2004 HIGH-PRIORITY DOCUMENTATION CHANGES: ------------------------------------ @@ -130,15 +130,17 @@ PLATFORM-SPECIFIC BUGS: HIGH-PRIORITY I18n STUFF: ------------------------- - * Non-ASCII characters (in translations) aren't uppercased when using - '--uppercase'. Example: In the Norwegian Nynorsk ('nn') translation - the top left icon is shown as 'MåLING' instead of 'MÅLING'. - To fix: Convert to wide chars, use 'towupper' instead of 'toupper', - and convert back to UTF-8. 'mbstowcs' and 'wcstombs' unfortunately - depend on on LC_CTYPE, so the encoding they use is locale dependent - (and *not* necessarily UTF-8, which our strings are in). On the other - hand, 'towupper' must be locale-aware for correct conversion (e.g. a - Turkish uppercase 'i' is *not* an 'I'). + * Make sure characters are properly converted when using + '--uppercase' in locales that have uppercase letters. + Currently using 'mbstowcs' and 'wcstombs' to convert between + UFT-8 and wide chars, and using 'towupper' (instead of 'toupper') to + convert from lowercase to uppercase. + + Karl says: "'mbstowcs' and 'wcstombs' unfortunately depend on on LC_CTYPE, + so the encoding they use is locale dependent (and *not* necessarily UTF-8, + which our strings are in). On the other hand, 'towupper' must be + locale-aware for correct conversion (e.g. a Turkish uppercase 'i' is *not* + an 'I')." * Make right-to-left strings not wordwrap with the shortest line at the top!