From Bugzilla Helper:
User-Agent: Mozilla/5.0 (X11; U; Linux i686; ja-JP; rv:1.2.1) Gecko/20030115
Description of problem:
groff-1.18.1-9 adds extra spaces when it formats to show Japanese
Version-Release number of selected component (if applicable):
Steps to Reproduce:
1. man man (or something)
2. all Japanese chars have extra space...
Actual Results: See the screen shot: groff-bad.png
Expected Results: See the screen shot: groff-good.png
Created attachment 89887 [details]
output of groff-1.18.1-8
Created attachment 89888 [details]
output of groff-1.18.1-9
groff-1.18.1-9 also cuts several strings and breaks the Japanese sentence.
(See above groff-bad.png screen shot)
This is the kind of output we're getting from grotty, in hex bytes:
cc be 08 cc be 20 c1 b0 08 c1 b0 0a
----- ^H ----- SP ----- ^H ----- \n
(where '----' indicates a two-byte glyph)
The ^H is for overstriking.
Everything looks fine to me. kterm bug?
Previously it would have looked something like:
cc 08 cc be 08 be 20 c1 08 c1 b0 08 b0 0a
which is incorrect (breaking up multibyte characters)
Perhaps I am missing something, but this really seems to be a groff bug.
While the reported new groff behavior is better than before, there should be
no 0x20 (' ') characters there at all, and the terminal emulator is
correctly displaying the spaces.
Sorry, you're right.
To examine byte output I'm using:
zcat $(man -w rpm)|troff -c -mandoc -Tnippon | head -44 | strace -ewrite grotty
Please try groff-1.18.1-11.
um, don't you need more the one BS to overstrike a two-character cell wide glyph?
I still think that we have a problem here, even with groff-1.18.1-11 installed.
The output of "LANG=ja_JP.UTF-8 man nroff" looks nothing like what is posted in
the bug report (note, this is when running inside kterm). In addition, I'm
still seeing what appear to be extra spaces when running 'man groff'
Please test groff-1.18.1-14 and check if that fixes your problems. Please
test on console, GUI etc and let me know if something is still broken.
Florian La Roche
groff-1.18.1-14 or -15 should have this ok.
Florian La Roche
Confirmed with groff-1.18.1-15
Comment #9: You should try LANG=ja_JP.eucJP
JLESSCHARSET=japanese LANG=ja_JP.eucJP man man
on kterm might be good for QA