Shell Programming and Scripting

I have files in a folder like ..

citi11082006_1.trn
citi11082006_2.trn
citi11082006_3.trn
...
...
...
citi11082006_13.trn
.,...
citi11082006_nn.trn

Each file will have one string inside and I am builting a master string based out the string in the sub files by merging all into one by script.

I am merging the file into one master file into a string and name should be
citi11082006.trn

my script is like:

Code:

#!/bin/ksh
export TSTAMP=`date +"%d%m%Y"`
cd /export/home/xxx/
count=$(ls citi${TSTAMP}_*.trn | wc -l)
echo "file appending"
count=$(ls citi${TSTAMP}_*.trn | wc -l)
cat citi${TSTAMP}_*.trn | tr -d '\n' > citi${TSTAMP}.trn
chmod 777 citi${TSTAMP}.trn
echo "FILE MERGE SUCCESSFULLY..."
rm -f citi${TSTAMP}_*.trn

There is additional element called END always present in lats file and thus while merging one of builting a string.
This works fine when subfile has 9 in numbers.


This does not work when file have more than 9 in numbers thus unforced to built a wrong string.

Is there any way to bulit a string so that elements of citi11082006_2.trn always come after citi11082006_1.trn and so on...

like wise string citi11082006_13.trn will come after citi11082006_12.trn.

The above is not working in incremental order, as 10 comes after 1 and then 11.
Can any one help me what is best way to get a string in incremental order.
thanks in advance
regards

Here's a thought. Since those are generated sequentially, you might as well try the -t flag for ls.

Try this.

Code:

for file in $(ls -t citi${TSTAMP}_*.trn)
do
cat $file | tr -d '\n' > citi${TSTAMP}.trn
done

for the line

Code:

cat citi${TSTAMP}_*.trn | tr -d '\n' > citi${TSTAMP}.trn

Not tested.

Another way :

Code:

Prefix=citi${TSTAMP}
for file in $(ls ${Prefix}_*.trn | sort -k1.$((${#Prefix}+2))n)
do
   tr -d '\n' < $file >> $Prefix.trn
done

EDIT --
Correction of suffix for filenames (.DAT replaced by .trn)
Correction of redirect operator (> replaced by >>)


Jean-Pierre.

Vino,

I just found the script only display the 9th number file, in the group of 11, any other thoughts

aigles,

I have tried your code also, but it would only showing the last data of last file ,

Code:

Prefix=citi${TSTAMP}
for file in $(ls ${Prefix}_*.DAT | sort -k1.$((${#[Prefix]}+2))n)
do
   tr -d '\n' < $file > $Prefix.trn
done

is there any other thoughts

try with the corrected version of the script (see my previous post)

Jean-Pierre.

i was missed >> in my code.

Thanks a ton...

its works