That's exactly why Lithuanian keymap with true Unicode symbol names is not recognized, while keymap with ISO-8859-1 equivalents is. I am CCing it to Dmitry Timoshkov (sorry if you got this already). And I think better place for this discussion is [EMAIL PROTECTED] > -----Original Message----- > From: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED]]On > Behalf Of Alexey Morozov > Sent: Sunday, December 17, 2000 6:17 PM > To: [EMAIL PROTECTED] > Subject: WINE keyboard handling > > > I just looked onto WINE keyboard handling in windows/x11drv/keyboard.c. > Hell, it seems to be senseless, gentlemen :-), at least if we use XKB > (well we do :-)). > > Look: TSXLookupString gets us a two-byte keysym code. Perfectly! Why don't > just handle "bad cases" (such as those keys that generate an X event but > don't generate an Windows event) and translate this two-byte code to its > corresponding Unicode cousin?? Currently we translate this two-byte code > into single char (loosing its national group property) and then translate > it into two byte Unicode sequence w/ guessed translation page :-(. > > As far as I understand all we need is just a big one-to-one translation > table (or maybe a function that acts as such translator). It seems we even > have a similar function called 'MultiByteToWideChar'. Am I right? > > One may ask: why to make such changes to existing code? > > The answers are simple. > > First, existing code doesn't work properly. It assumes that we have the > only alternative group switched by Mode_Switch. It's a wrong assumption. > Mode_Switch usage is optional, there are ISO_Next_Group, ISO_Prev_Group > pseudo-keys and even user defined "functions" are possible (XKB setup can be > arbitrary complicated :-)) as I just read on the Ivan Pascal's XKB pages > (http://www.tsu.ru/~pascal/other/xkb, on Russian, I'm sorry, you probably > may use http://www.translate.ru to translate them into your language)) > > Current WINE keyboard implementation doesn't work properly w/ these > switchers. > > Another reason to start hacking is multilingual support. Currently XKB can > handle up to four independant keys groups so it's possible to have more > than two languages simultaneously switching between them either in cycle > or by pressing specially defined key [sequences]. Advanced applications > (like those that work w/ Unicode) can simultaneously have, say, English, > Russian, Greek and Japanese text in a single document. Why don't use this > _instant_ ability and reimplement the wheel? > > Yours, > Alexey Morozov. > _______________________________________________ > wine-users mailing list > [EMAIL PROTECTED] > http://www.winehq.com/mailman/listinfo/wine-users
