Read variables from file and loop through them in bash - bash

Question

I have a text file containing variables like this:
lane1_pair1="file1"
lane1_pair2="file2"
lane2_pair1="file3"
lane2_pair2="file4"
...
I'd like to loop through the variables and concatenate all of them in a single file. I am applying the loop as:
. variables
for (( n=1; n<=no_lanes; n++ )) {
cat $"lane"${n}_pair1 >> "$sampleID"_cat1.fq
cat $"lane"${n}_pair2 >> "$sampleID"_cat2.fq
}
"$sampleID"_cat1.fq > fq_align_1
"$sampleID"_cat2.fq > fq_align_2
The problem here is that, cat command does not work because instead of replacing the "laneX_pairY" with its value, treat it as a string. I was wondering if anyone here has any idea about this.

Answer 1

As I see it, the problem is you're looping through variables that are named like an array but aren't one. Then you're in the position of trying to evaluate the contents of the variable name created from your other indexing variable, which is a recipe for messy code that breaks easily. I recommend making actual arrays of file names from your variables file:
p1_files=($(grep 'pair1' variables | cut -d '=' -f 2))
p2_files=($(grep 'pair2' variables | cut -d '=' -f 2))
for f in "${p1_files[#]}"; do
cat "${f//\"/}" >> "$sampleID"_cat1.fq
done
for f in "${p2_files[#]}"; do
cat "${f//\"/}" >> "$sampleID"_cat2.fq
done
I'm not sure how you want your fq_align_n variables to be, but you can read the file contents to variables using:
fq_align_1=$(cat "$sampleID"_cat1.fq)
fq_align_2=$(cat "$sampleID"_cat2.fq)
Or just skip the file creation altogether and build the variables incrementally in the loops.