UNIX for Dummies Questions & Answers

Hi All,

I have been trying to FTP some data files from Windows directory to a UNIX server. The txt file in the windows contails the following data:
"111~XYZ~1~Contact person’s phone number~COMMENTS~~~~"
but the same line is appearing as
"111~XYZ~1~Contact person^Òs phone number~COMMENTS~~~~" in unix server.

The ' character is appearing as ^Ò in the unix file. Kindly suggest a way to FTP the files w/o distorting the characters.

Thanks,

hi
try to ftp in binary format
or else u can remove this cheracter
open u r file in vi
then follow this
:1,$s/\^Ò//gp
i think this will help u

Hi,

Try the dos2unix command in unix machine

syntax : dos2unix file_name

The dos2unix command isn't there in the UNIX version I am working upon. Nevertheless thanks for the help.

I tried FTPing in the binary format by using type binary on ftp prompt but still the file is being FTPed with the junk character. Also after executing the command :1,$s/\^Ò//gp in vi editor, i am getting the following error "ex: 0602-071 No match exists for the substitute pattern."

dos2unix -- and FTP ASCII mode -- generally only handles line ending convention translations. The file is in a Windows character set; you need to convert it to whatever you are currently using. We can only guess what you would like it to be, and what the original character set is, but have a look at recode and iconv. If you don't have them installed locally, they are available on the Internet. Or if you have a strong stomach, use Windows and "save as" into a sane file format.

For a quick and dirty solution, find out the octal character code of the problematic character (od gives you a basic dumping facility, but it's a bit hard to read -- try piping in just a single line, rather than od:ing the whole file; or if you have xxd or hexdump, you can get more readable output, albeit in hex, not octal) and tr allows you to substitute it with something else.

Code:

vnix$ grep person.s dosfile | head -1 | od -cb
0000000   1   1   1   ~   X   Y   Z   ~   1   ~   C   o   n   t   a   c
        061 061 061 176 130 131 132 176 061 176 103 157 156 164 141 143
0000020   t       p   e   r   s   o   n 222   s       p   h   o   n   e
        164 040 160 145 162 163 157 156 222 163 040 160 150 157 156 145
0000040       n   u   m   b   e   r   ~   C   O   M   M   E   N   T   S
        040 156 165 155 142 145 162 176 103 117 115 115 105 116 124 123
0000060   ~   ~   ~   ~  \n
        176 176 176 176 012
0000065
vnix$ tr '\222' "'" <dosfile > unixfile
vnix$

I'm only using \222 as an example here; the character you want to find is most likely something else.

Don't assume there aren't other "funny" character codes in the file. As a matter of fact, there probably are.

On another note, the vi tip would probably work if you could copy/paste the problematic sequence instead of typing in whatever it is you think you are seeing. The people here who are trying to help might not be using the same character set as you, so just to be unambiguous, it would help to know the character codes of the problematic character(s) in order to help you.