I have a text file with unknown length. There are two values on each line:
VALUE1[SPACE]VALUE2
Now I have to get another (or the same) file with a new List like:
0.0.0.0/rep/com/bla/blub/VALUE1/VALUE2/VALUE1-VALUE2.zip
...for each line I have in the list with the two values. How can I do this?
This is a good usecase for awk (Updated to take care of the parantheses):
awk -F" " '{ gsub("\\(", "", $1); gsub("\\)", "", $2);print "0.0.0.0/rep/com/bla/blub/"$1"/"$2"/"$1"-"$2".zip"}' test.txt > yournewfile.txt
This will split each line with a space, replace the opening parentheses in your first token $1 and replace the closing parentheses in your second toke $2 and then use values in their respective tokens $1 and $2 in the string you are outputing with print
I assume that VALUE1 and VALUE2 contain no spaces, otherwise the first definition becomes ambigous. With this hypothesis, you can split a line at the first space using cut and compose the filename variable using string interpolation:
cat txt-File | while read VALUE1 VALUE2 _; do
filename="0.0.0.0/rep/com/bla/blub/$VALUE1/$VALUE2/$VALUE1-$VALUE2.zip"
# Do something with filename...
stat "$filename"
done
while read VALUE1 VALUE2 _; do ... done <file would avoid pesky external processes and be somewhat more idiomatic.
cat file | while is not necessary, you can say while ... done < file. Also, read can read more than one argument by saying while read value1 value2 ...
cat file because it makes obvious what the loop is processing, without having to scroll to the bottom. Fixed the other thing.
<file at the front of the loop like you can with a regular single command. It would be clearer.The format of your input data file is (VALUE1 VALUE2).. You first remove the initial (, and latter ). There are many ways to do this, here is one
sed 's/(//g' yourfile | sed 's/).//g'
You can then pass this over to awk. Here $1 corresponds to the first column, and $2 to the second:
sed 's/(//g' yourfile | sed 's/).//g' | awk '{print "0.0.0.0/rep/com/bla/blub/"$1"/"$2"/"$1"-"$2".zip"}'
I would use sed for this:
sed $'s,^[ \t]*\([^ \t]*\)[ \t][ \t]*\([^ \t]*\).*$,0.0.0.0/rep/com/bla/blub/\1/\2/\1-\2.zip,' filename >new-filename
You could modify the original file in place if you prefer by adding -i.bak: sed -i.bak $'s,...,'.
