On Thu, May 10, 2012 at 10:41:36AM -0500, Bruce Dubbs wrote: > xinglp wrote: > > > > Show them in hex > > mkdir -v aaa|view - -c '%!xxd' > > 0000000: 6d6b 6469 723a 2063 7265 6174 6564 2064 mkdir: created d > > 0000010: 6972 6563 746f 7279 20e2 8098 6161 61e2 irectory ...aaa. > > 0000020: 8099 0a ... > > > > The single quotes is: > > e2 80 98 > > e2 80 99 > > > > The dash character in /usr/share/man/man8/ifdown.8 is 0x2d , > > but use man ifup|view - -c '%!xxd' , the dash character is > > e2 80 90 > > > > Then, I think the "local console" can not display those utf8 character. > > But I remember in ealier time, my lfs works well of show those character. > > > > Now, I fix it by this configure file > > /etc/sysconfig/console > > FONT="lat1-16 -m 8859-1" > > UNICODE=yes > > I am not familiar with non-US locales. I have no idea why man or less would > translate a perfectly good ascii character, like a dash or quote character to > something else. > > -- Bruce I assume it is 'man-db' doing this. When we used the original 'man' with 'groff' it used to do a lot of 'use the _proper_ symbol' changes.
Decoding the byte representations of UTF-8 back to U+nnnn form is tedious - I used to have some notes on how to do it, from when I found characters my font could not render, but it's easiest to look online : www.utf8-chartable.de - in this case, the general punctuation in U+2000 to U+20FF. e2 80 90 U+2010 hyphen e2 80 98 U+2018 left single quote e2 80 99 U+2019 right single quote These are the preferred code points to use in text, any modern UTF-8 console font ought to be able to display them. There are many other U+2nnn glyphs in fairly-common use, my own fonts handle many of them (by mapping multiple code points to a single glyph). I'm not sure if xinglp changed his font or just added UNICODE=yes ? I have some resources at linuxfromscratch in /~ken/fonts : fonts-testing/alphabets-etc.txt (if reading it in firefox, it continues to imagine this is encoded in Western European and has to be told to use Unicode). The mapping of multiple code points to a single glyph can be seen in the sigma tarballs (the .sfm files), or in the .psfu.txt files within the LatGrkCyr tarballs, or in the uni-vga tarball which showed me how to start creating my own fonts. HTH ĸen -- das eine Mal als Tragödie, das andere Mal als Farce -- http://linuxfromscratch.org/mailman/listinfo/lfs-dev FAQ: http://www.linuxfromscratch.org/faq/ Unsubscribe: See the above information page
