Re: [MLUG] backup via cron

[Mike Miller <EMAIL:PROTECTED>]

On Sat, 4 Feb 2006, Daniel Nowlin wrote:

> That was the problem. ^M's.
>
> dvd+rw-format -f /dev/dvd;^M
> growisofs -Z /dev/dvd -J -r /var/ftp/files;^M
>
> What is funny is when I edit the file with nano I can not get rid of the 
> ^M's. Has got to be a setting in nano or putty?


I wouldn't use nano to get rid of them.  I regularly use perl.  The 
UNIX/Linux system typically comes with dos2unix and unix2dos commands 
which I do not like.  I prefer my own perl scripts which are very short. 
The two that I use are shown below.  I call them d2u (short for "DOS to 
UNIX") and u2d (short for "UNIX to DOS"):


-------begin u2d on next line----------
#!/usr/bin/perl -i -p
s/$/\r/
-------end u2d on previous line----------

-------begin d2u on next line----------
#!/usr/bin/perl -i -p
s/\r//g
-------end d2u on previous line----------


Just copy the text to files in your path (probably in /usr/local/bin), use 
chmod to make the files executable...

chmod a+rx /usr/local/bin/d2u
chmod a+rx /usr/local/bin/u2d

...and you're all set.  They handle globbing with wildcards just fine:

d2u *.txt text/*.out

That particular command will delete every carriage return (also known as 
^M or \r) from every file matching the specified pattern.  It will leave 
the newline characters alone.  It does not store a backup of the original 
files.  The commands u2d and d2u are not *exactly* inverse operations, but 
in most real-life uses, they are inverse operations.  If you want to 
understand them fully, which is worthwhile, learn a little more about how 
perl works.

I thank Stephen and Jon for helping me to figure out how to make these 
little scripts a few years ago.

Mike

_______________________________________________
members mailing list
EMAIL:PROTECTED
http://mlug.missouri.edu/mailman/listinfo/members