CODEWITHSUNDEEP
bashwhile-loop
Spencer K
Spencer K

Reputation: 73

While loop not stopping after condition is met (bash/linux)

I want to run a script that searches if each snp (list of snps are contained in the variable $snplist) is in all GWAS cohorts (all cohorts are in separate files ending with *renamed_snp_search.txt). If a snp is in all the cohorts then the snp goes into a log file and I want the loop to terminate after 10 snps are found. I thought that redefining the $total_snp variable toward the end of the while loop would help with this but it appears that the loop just keeps going after using sample data.

touch snp_search.log
total_snp=$(cat snp_search.log | wc -l)
files=(*renamed_snp_search.txt)
count_files=${#files[@]}
while [ "$total_snp" -lt 10 ] ; do
    for snp in $snplist ; do
    count=$(grep -wl "${snp}" *snp_search.txt | wc -l)
            if ((count == count_files)) ; then
            echo "$snp was found in all $count_files files" >> ${date}_snp_search.log
            total_snp=$(cat snp_search.log | wc -l)
            fi
     done
done

Upvotes: 1

Views: 575

Answers (1)

Gordon Davisson
Gordon Davisson

Reputation: 125708

You're misunderstanding the logical structure of the two loops you have; the while [ "$total_snp" -lt 10 ] loop and the for snp in $snplist loop. The condition on the while loop is only tested at the beginning of each time through that loop, so it will not interrupt the for loop if the condition is met partway through that loop.

Essentially, the execution process is like this:

  1. Check to see whether $total_snp is less than 10; it is, so run the while loop's contents:
  2. Run the for loop, search files for each item in $snplist
  3. Check to see whether $total_snp is less than 10; if it is run the while loop's contents again, otherwise exit the loop.

...so if there are 10 or more snps that're found in all files, it won't notice that it's found enough until it's run through the entire snp list.

(On the other hand, suppose there were only 7 snps that were found in all files. In that case, it'd search for all snps, find the 7 matches, check to see whether it'd found 10 yet, and since it hadn't it'd run the for loop again and find and log the same 7 matches again. After which $total_snp would be 14, so it would finally exit the while loop.)

What you want to do instead is to break out of the for loop if $total_snp reaches 10 as that loop runs. So remove the while loop, and add a break condition inside the for loop:

for snp in $snplist ; do
count=$(grep -wl "${snp}" *snp_search.txt | wc -l)
    if ((count == count_files)) ; then
        echo "$snp was found in all $count_files files" >> ${date}_snp_search.log
        total_snp=$(cat snp_search.log | wc -l)
        if [ "$total_snp" -ge 10 ]; then
            break    # Break out of the `for` loop, we found enough
        fi
    fi
done

Upvotes: 1

Related Questions