awk iterating through an array using for loop and printing based on if condition

   bash --version

cat filename.txt|cut -d "|" -f1|tr -d [^:digit:][:punct:]
[:blank:]|tail -10|awk '{a[i++]=$0;}{k=length(a)}END{for (i=k;i>=1;i--)if (5==5){print i " "a[i]" " a[i-1]" "k}}'

10            10
9  18410      10
8 18410 18409 10
7 18409 18408 10
6 18408 18407 10
5 18407 18406 10
4 18406 18405 10
3 18405 18404 10
2 18404 18403 10
1 18403 18402 10


cat filename.txt|cut -d "|" -f1|tr -d [^:digit:][:punct:]   [:blank:]|tail -10|awk '{a[i++]=$0;}{k=length(a)}END{for (i=k;i>=1;i--)if ((a[i]-1)==a[i-1]){print i " "a[i] " "k} else print "not done"}'

not done
not done
not done
not done
not done
not done
not done
not done
not done
not done

trying to get the greatest number x if the preceding number is one lesser than x.

wc -l does not give the row count that I need. I dont need line count.

Column extracted using cut is the primary key in this csv file. File could have unknown number of blank lines at the end.

 cat filename.txt|cut -d "|" -f1|tr -d [^:digit:][:punct:]    [:blank:]|tail -10
18402
18403
18404
18405
18406
18407
18408
18409
18410

Why is this if condition evaluating to take the "else action"? (a[i]-1==a[i-1]) evaluates to false. I dont see why.

Solution

Given the following CSV file (filename.txt):

1a|b|123
2a|b|123
3|4
5.7|4
66
67
10|11
11|aaa

we can determine the number 67 using:

awk -f s.awk filename.txt

where s.awk is:

BEGIN {
    FS="|"
}
NF>=1 {
    q=$1
    gsub(/([^[:digit:]])|([[:blank:][:punct:]])*/,"",q)
    if (q==(prev+1))
        if (q>max)
            max=q
    prev=q
}
END {
    print "Max=" max
}

Update

A one-liner could look like:

awk 'BEGIN { FS="|" } { q=$1; gsub(/([^[:digit:]])|([[:blank:][:punct:]])*/,"",q); if (q==(prev+1)) {if (q>max) max=q} prev=q } END { print "Max=" max }' filename.txt