Shell Programming and Scripting

Hi is it possible to change multiple files (~10k) names with out disturbing the data in it. ?

input

Code:

Hynda|cgr10(+):100027702-1000312480|.txt
Hynda|cgr10(+):100027702-1000312483|.txt
Hynda|cgr10(+):100027702-1000312484|.txt
Hynda|cgr10(+):100027702-1000312482|.txt

output

Code:

cgr10_+_100027702_1000312480.txt
cgr10_+_100027702_1000312483.txt
cgr10_+_100027702_1000312484.txt
cgr10_+_100027702_1000312482.txt

Note: Actual input files look like this (with few spaces in them)
Ex: Hynda\|cgr10\(+\)\:100027702-1000312480\|.txt

Probably not the most efficient solution, and maybe better done in sed, but in BASH this is how I would do it:

Code:

for FN in *; do mv $FN "`echo ${FN#*|} | tr -d ':|' | tr '()-' '_'`"; done

${FN#*|} - This removes everything up to and including the first bar.
tr -d ':|' - The above result is piped to tr to delete the colon and remaining bar.
tr '()-' '_'` - And finally that is passed to tr to replace the other characters with underscores.

That generates the new name in the style you described, and is substituted as the destination parameter for mv to rename the file.

sorry forgot to mention another symbol

Code:

Hynda|cgr10(-):100027702-1000312480|.txt

Code:

for x in *.txt
do
    mv $x `echo $x | sed 's/[^\|]*|\([^|]*\)|.*/\1/' | tr -d ':' | sed 's/\([^(]*\)(\(.\))\([^-]*\)-\(.*\)/\1_\2_\3_\4/'`.txt
done

With the possibility of different characters within the parenthesis it is definitely easier to use sed as using just BASH would require different steps. In that case this is the expression I would use:

Code:

for FN in *; do mv $FN "`echo $FN | sed 's/^.*|\(.\+\)(\(.\)):\([0-9]\+\)-\([0-9]\+\)|\(.*\)$/\1_\2_\3_\4\5/'`"; done

That will discard anything up to and including the first bar
Capture the string of characters before the opening parenthesis
Disregard the opening parenthesis
Capture a single character inside the parentheses
Disregard the closing parenthesis and colon.
Capture the numeric string until the dash.
Disregard the dash.
Capture the numeric string until the bar.
Disregard the bar
Capture everything else (the file extension).

Anything that does not exactly match those conditions will remain unchanged and so mv would give a warning that the source and destination are the same, but will leave those filenames unchanged.

Otherwise the captured elements will be combined using underscores (except before the file extension) and be renamed.