Just found out I had only done a reply and not forwarded to the cygwin ml last time. Sorry about that.
This is a forwarded message From: Paul-Kenji Cahier <[EMAIL PROTECTED]> To: "Peter A. Castro" <[EMAIL PROTECTED]> Date: Wednesday, March 5, 2008, 3:06:21 PM Subject: ZSH - UTF-8 ===8<==============Original message text=============== >>>>??LC_ALL set to C.UTF-8 in all cases. Cygwin freshly installed. >>>>??Also tried LC_ALL=en_US.UTF-8(which shouldnt exist) and it did >>>>??not work either. >>>??Newlib uses "C-UTF-8", not "C.UTF-8" for some reason. I don't see that >>>??newlib understands the dot. Moreover, Cygwin doesn't support utf-8 yet, >>>??so you might have strange effects using utf-8 for filenames. >>>??Corinna >>?(this time text under quote, didnt know it bothered) >>?I'm still failing to get zsh working with utf-8, even >>?with C-UTF-8 as LC_ALL: as specified in my previous message >>?non-ascii bytes get displayed as <00cxx> which the line editor >>?does manage properly. But there is still no way to actually see >>?utf-8 characters displayed, or even their unicode values(ie >>?it's still a single byte editing mode). If someone manages >>?to get it working, please post:) >?Can you send me the specific key sequence you are using to generate this? In this case, "<00c3><00a9>"(which is acute e, or U+00E9). >?Are you using a CMD window or rxvt or an xterm? I tried with CMD/rxvt/putty and the results was always the same. UTF-8 displays well when, for example, doing a cat on a file with utf-8 content(just want to make it clear that it's not a term issue). >? Also, are you using the stock >?zsh 4.3.5 source distro or have you applied the Cygwin patches from >?4.3.4 to it before building? Currently, only with a stock 4.3.5, I'm currently trying to build with the cygwin patch and build system to see if it makes any difference. EDIT: I just did it and it did change the behaviour back to the same as cygwin built 4.3.4: the ZLE displays the utf-8 characters correctly(ie sends the correct bytes) but is still 100% unaware of multi-byte. I tried with both C-UTF-8 and C.UTF-8 and en_US.UTF-8, and none worked properly, I personally . >?I've recently built 4.3.5 for Cygwin and have been testing it before >?releasing it. Send me the key sequences you are using and I'll see if I >?can repro it. Just type any non-ascii character(ie accents, cjk, cyrillic, etc) in an empty prompt then do a backspace(which should erase it) and notice that if you actually press backspace a second time it erases again, when the line should have been empty and not-erased. The reason being that the first time it only erases the second byte of the 2-byte character, and the second time the first byte(instead of erasing both at the same time). >?As Corinna noted, Cygwin doesn't really support utf-8, so this might have >?some bearing on the matter. And, you should know that some multi-byte >?delete issues have been identified on other, supposedly, UTF-8 compliant >?systems, so you aren't the first to hit this. Shouldnt be a problem in this case though, as this is purely for the ZLE(zsh line editor). >>?Also I know about the lack of utf-8 support in cygwin for most >>?of the windows calls(and personally think that the cygwin utf-8 >>?wrapper code patch would be a vast improvement, that is the one >>?from okisoft that got posted to cygwin's ml long ago, though >>?I understand some people dont like how it's done). This should >>?still not be a problem for one to do an "echo éé<backspace>" >>?in his term transparently. >>?(no problematic wide windows calls involved there as far as I know). >>?Paul-Kenji Cahier >>?-- >>?Unsubscribe info: http://cygwin.com/ml/#unsubscribe-simple >>?Problem reports: http://cygwin.com/problems.html >>?Documentation: http://cygwin.com/docs.html >>?FAQ: http://cygwin.com/faq/ ===8<===========End of original message text=========== -- Unsubscribe info: http://cygwin.com/ml/#unsubscribe-simple Problem reports: http://cygwin.com/problems.html Documentation: http://cygwin.com/docs.html FAQ: http://cygwin.com/faq/
