UNIX for Dummies Questions & Answers

Hi,

My shell script calls a perl script to create an excel and the shell
script emails the excel. This excel file needs to be renamed
to some Kanji name.

I have a flat file that has the required file name in kanji and i extract it
within the shell script and try to rename the file, but the filename shows junk.

the flat file itself has proper data coz the excel gets some kanji from the same flat file and is displayed properly.

fname=`grep FILENAME ${abcd}/etc/flat.cfg`
fname=`echo "${fname}" | awk '{ print substr($0,13,100) }'`
echo ${fname} --> itself displays junk

please help.

You have a kanji displaying terminal setup? Feed the output to od -bc (well, I am too much an octal guy, but something like that) to see what is coming out, compared to what is in the file.

(No Hiragana or Katakana?)

Is this UTF8, UNICODE, EUC, ja : Encodings of Japanese

may be utf8...though i am not sure, how do i find out...?

file -i file.txt
regular file

vi file.txt
\203\214\203|\201[\203g

less file.txt
<83><8C><83>|<81>[<83>g

file created using notepad++
New file > Encoding > Character Sets> Japanese
then copy paste kanji characters from email which i received...

Usually the provider or 'file' or a japanese text viewer can tell you. If it is unicode based like utf8, or any of the other Japanese encoding systems, there are wiki to help decode what is in the file's od output or other binary tools. Since we are talking file rename, I would expect utf8 to be most probable. Even with wide characters, in file names, bytes are bytes. I would expect you have a file system that is utf8 friendly, like the late model windows file windows file system, which I am told is in unicode of some sort under the covers.

Maybe write a little script in notepad++ Japanese, exporting LC_ALL compatibly, and rename the file in there.

Just 8 bytes in the file; how many gliphs did you paste? Shift-JIS looks a good match for 4 glyphs:
http://en.wikipedia.org/wiki/Shift-JIS
http://www.rikai.com/library/kanjita...des.sjis.shtml
http://demo.icu-project.org/icu-bin/...p?conv=ibm-942 for 0x838C 837C 815B 8367 gives: レ ポ ー ト

It says 'Report'!

Now, the trick is to get them into the right set for file names.

Google translate uses their utf8 codes: http://translate.google.com/#ja/en/%...83%BC%E3%83%88

0xE383AC should be the 3 byte utf8 for the first character. http://en.wikipedia.org/wiki/UTF-8#Description

= 0x3000 + 0x00C0 + 0x002C = 30EC = same lower right hook. http://en.wikipedia.org/wiki/Kana

It is katakana, appropriate for talking to 外人 http://translate.google.com/#ja/en/%E5%A4%96%E4%BA%BA

Interestingly, the word Gaijin for foreigner is two kanji.
http://en.wiktionary.org/wiki/%E5%A4%96 out, outside, external, foreign
http://en.wiktionary.org/wiki/%E4%BA%BA person, people, humanity, someone else

I had though those phonemes Gaijin meant http://translate.google.com/#ja/en/%E6%82%AA%E9%AD%94

Japanese Ministry of Educations says literate is about 2K kanji ! Universal Han has 22K or something like.

Google and wiki are your friends!
http://translate.google.com/#ja/en/G...81%99%EF%BC%81