How to turn this piped command into a single Awk invocation?

Question

I am formatting an input file, here is a sample of it:

                 The First 1,000,000 Primes





         2         3         5         7        11        13        17        19 


        23        29        31        37        41        43        47        53 

And this file, with the xxd command:

00000000: 2020 2020 2020 2020 2020 2020 2020 2020                  
00000010: 2054 6865 2046 6972 7374 2031 2c30 3030   The First 1,000
00000020: 2c30 3030 2050 7269 6d65 7320 2866 726f  ,000 Primes (fro
00000030: 6d20 7072 696d 6573 2e75 746d 2e65 6475  m primes.utm.edu
00000040: 290d 0d0a 0d0d 0a20 2020 2020 2020 2020  )......         
00000050: 3220 2020 2020 2020 2020 3320 2020 2020  2         3     
00000060: 2020 2020 3520 2020 2020 2020 2020 3720      5         7 
00000070: 2020 2020 2020 2031 3120 2020 2020 2020         11       
00000080: 2031 3320 2020 2020 2020 2031 3720 2020   13        17   
00000090: 2020 2020 2031 3920 0d0d 0a20 2020 2020       19 ...     

I am using the command:

awk -v OFS="\n" '{$1=$1}1' inputfile

But I get empty lines that I would like to get rid of:

The
First
1,000,000
Primes

2
3
5

I can get rid of it with a pipe:

awk -v OFS="\n" '{$1=$1}1' inputfile | awk '!/\r/'

Is there a way to achieve the same by calling Awk only once?

Answer 1

Assumptions/Understandings:

• we do not need to maintain the \r characters
• (per comment from) OP only wants to display the prime numbers (one per line)

Reconstituting OP's file from the xxd output:

$ xxd -r <<< '00000000: 2020 2020 2020 2020 2020 2020 2020 2020                  
00000010: 2054 6865 2046 6972 7374 2031 2c30 3030   The First 1,000
00000020: 2c30 3030 2050 7269 6d65 7320 2866 726f  ,000 Primes (fro
00000030: 6d20 7072 696d 6573 2e75 746d 2e65 6475  m primes.utm.edu
00000040: 290d 0d0a 0d0d 0a20 2020 2020 2020 2020  )......         
00000050: 3220 2020 2020 2020 2020 3320 2020 2020  2         3     
00000060: 2020 2020 3520 2020 2020 2020 2020 3720      5         7 
00000070: 2020 2020 2020 2031 3120 2020 2020 2020         11       
00000080: 2031 3320 2020 2020 2020 2031 3720 2020   13        17   
00000090: 2020 2020 2031 3920 0d0d 0a20 2020 2020       19 ...' > inputfile

Tweaking OP's current awk script:

awk -v OFS="\n" '
NR==1 { next }                      # skip 1st line
      { gsub(/\r/,"" )              # strip all "\r" characters from line
        $1=$1
      }
length()                            # if resulting line length > 0 then print to stdout
' inputfile

#### as a one-liner:

awk -v OFS="\n" 'NR==1{next}{gsub(/\r/,""); $1=$1} length()' inputfile

An alternative awk idea:

awk 'NR>1 { for (i=1;i<=NF;i++)     # skip 1st line and loop through all fields
                if ($i+0==$i)       # if the field is a number (prime in this case) then ...
                    print $i        # print on its own line
          }
' inputfile

#### as a one-liner:

awk 'NR>1 {for (i=1;i<=NF;i++) if ($i+0==$i) print $i}' inputfile

These both generates:

2
3
5
7
11
13
17
19

Answer 2

Without expected output it's a guess but is this what you're trying to do?

$ awk -v OFS='\n' '/^[[:space:]0-9]+$/{$1=$1; print}' file
2
3
5
7
11
13
17
19
23
29
31
37
41
43
47
53

or maybe this?

$ awk '/^[[:space:]0-9]+$/{$1=$1; printf "%s%s", sep, $0; sep=OFS} END{print ""}' file
2 3 5 7 11 13 17 19 23 29 31 37 41 43 47 53

They'll both behave the same way with any POSIX awk.