Shell Programming and Scripting

I have a very 'dirty' bash script that is working and looks like this:

Code:

#!/bin/bash
for i in {1..n}
do
 mv block${i} infile
 ./dnadist <<-EOF
 D
 Y
 EOF
 rm infile
 mv outfile Result${i}
 mv Result${i} infile
 ./neighbor <<-EOF
 Y
 EOF
 rm infile
 rm outfile
 mv outtree Result${i}
done

Most of the times I do not know the value for "n", how can I modify the script in such way that even if I do not know the number of files in that particular folder I will be still able to process them?
Other little thing, I am pretty sure there should be a way to 'clean' the script to make it look "cleaner".
Any help will be greatly appreciate it!

Can you also explain what you're trying to achieve?
There are probably better approaches to do what you want.

Franklin52,
Let me try to explain what I am trying to accomplish. I have hundreds of files named "block1", "block2", etc; containing DNA sequences in Phylip format. I want to calculate the genetic distance for each "block" file using the binary program "dnadist" which uses D (Kimura) and Y (YES) as options. Unfortunately, "dnadist" won't let me use anything as an input but a file named "infile". Thus, I start my script by renaming the files from "blockX" to infile and then feed it into dnadist. The output file that is generated by "dnadist" is bydefault "outfile" and I cannot change the file name either. I just came to realize that I do not need to "temporarily" rename the "outfile" to "ResultX" but instead to "infile", therefore I can delete that part. Finally, I grab the "new" "infile" and process it using "neighbor", using Y (YES) as option, in order to generate the corresponding trees. "neighbor" generates two different files "outfile" and "outtree", I only need the "outtree" file, which I rename as "ResultX", and that' why I just need to delete the "outfile". Then, we process the next file "block2" and the cycle starts again.
As I said, many times I do not know the number of "block" files in my folder, and therefore I would like to modify my script to be able to process whatever number of files, it can range from some few to several hundreds, are present in that given folder.
So, the my script now looks something like this:

Code:

#!/bin/bash
for i in {1..2}
do
 mv block${i} infile
 ./dnadist <<-EOF
 D
 Y
 EOF
 rm infile
 mv outfile infile
 ./neighbor <<-EOF
 Y
 EOF
 rm infile
 rm outfile
 mv outtree Result${i}
done

Thanks!

Can't you do something like:

Code:

n=$(ls block* | wc -l)

while [ $n -ge 1 ]
do
 mv block${i} infile
 ./dnadist <<-EOF
 D
 Y
 EOF
 mv outfile infile
 ./neighbor <<-EOF
 Y
 EOF
 rm outfile
 mv outtree Result${i}
 n=$(( $n - 1 ))
done

I am not quite sure I understand how to modify my script following your suggestion.
You mean something like this?:

Code:

#!/bin/bash
n=$(ls block* | wc -l)
while [ $n -ge 1 ]
do
    mv block${n} infile
    ./dnadist <<-EOF
    D
    Y
    EOF
    rm infile
    mv outfile infile
    ./neighbor <<-EOF
    Y
    EOF
    rm infile
    rm outfile
    sed "s/:[0-9\.-]*//g" outtree > Result${n}
    rm outtree
    n=$(( $n - 1 ))
done

Thanks once again!

Yes, something like that and the rm command is superfluous here:

Code:

rm infile
mv outfile infile

What I am trying to do in that particular step is to delete the original "infile" and replace it with the new "outfile".
Should I only leave it as:

Code:

#!/bin/bash
n=$(ls block* | wc -l)
while [ $n -ge 1 ]
do
 mv block${n} infile
 ./dnadist <<-EOF
 D
 Y
 EOF
 mv outfile infile
 ./neighbor <<-EOF
 Y
 EOF
 rm outfile
 mv outtree Results${n}.dis
 rm infile
 n=$(( $n - 1 ))
done
cat *.dis > intree2
rm *.dis
./treedist <<EOF
2
l
v
y
EOF