bash ; changing some data inside of a file
 

Hello,
I got some data files which I need to change a little bit.
The data file looks like this:
-0,090 0,000

05BA 0B3A

-719,930 -720,000
-719,655 -719,800
-719,340 -719,600
-719,045 -719,200
-718,740 -718,900
-718,435 -718,600
-718,150 -718,300
...
I have to prepare it for octave.
This means the first four lines should be as a comment (I think - %)
and the rest lines should have no commas but dots.
It should look like:
%-0,090 0,000
%
%05BA 0B3A
%
-719.930 -720.000
-719.655 -719.800
-719.340 -719.600
-719.045 -719.200
-718.740 -718.900
-718.435 -718.600
-718.150 -718.300
...
I try to write a bash script, but I can not find a function that looks for a character in a file.
Can you help me?
regards,
sebastian

You could easily do that with sed.

Note, to save the changes with sed, redirect into a NEW file:

Hello,
thank you very much for the solution with sed; it works!!
the second task is to do it automatically with all 14000 files in one directory changing the names of the files from 0,00 to 0_00 (0,10 to 0_10) and so on...

greetings,
sebastian

Just wrap a loop around the soln above:
one way to change filename

new_file=`echo $file |sed 's/,/_/'`

thank you very much, that's very nice solution...
regards
sebastian

In your for loop, the file will be in a variable. You could run the sed line like
sed '1,4s/^/%/;s/,/./g "$file" >"${file//,/_}"

Using variable substition will be quicker and cleaner the using sed to change the filename.