0

I have a file which is the output of several commands piped. Something like this

command1 input.txt| command2 | command3 | input file

The file is tab-separated

After command 3, my input file looks like this

chr6    116732135   116741866   116732135   116732368   116741505   116741866   +   0.79    0.51    0.97    0.77    0.48    0.97    0.02    0.37    'chr6:116732136-116732368:+@chr6:116741506-116741866:+.A.withRI','chr6:116732136-116732368:+@chr6:116741506-116741866:+.B.up_chr6:116732136-116732368:+@chr6:116741506-116741866:+.B.dn'    (0,0):10,(1,0):147,(1,1):1  0:148   (0,0):36,(1,0):161,(1,1):3  0:163,1:1   chr6    +   116732136,116732136 116741866,116741866 gene_id "ENSG00000196911.10"; transcript_id "ENST00000368564.6"; gene_type "protein_coding"; gene_name "KPNA5"; transcript_type "protein_coding"; transcript_name "KPNA5-202"; level 2; protein_id "ENSP00000357552.1"; transcript_support_level "1"; tag "basic"; tag "appris_principal_1"; tag "CCDS"; ccdsid "CCDS5111.1"; havana_gene "OTTHUMG00000015448.4"; havana_transcript "OTTHUMT00000041967.2";
chr6    116732135   116741866   116732135   116732368   116741505   116741866   +   0.79    0.51    0.97    0.77    0.48    0.97    0.02    0.37    'chr6:116732136-116732368:+@chr6:116741506-116741866:+.A.withRI','chr6:116732136-116732368:+@chr6:116741506-116741866:+.B.up_chr6:116732136-116732368:+@chr6:116741506-116741866:+.B.dn'    (0,0):10,(1,0):147,(1,1):1  0:148   (0,0):36,(1,0):161,(1,1):3  0:163,1:1   chr6    +   116732136,116732136 116741866,116741866 gene_id "ENSG00000196911.10"; transcript_id "ENST00000356348.6"; gene_type "protein_coding"; gene_name "KPNA5"; transcript_type "protein_coding"; transcript_name "KPNA5-201"; level 2; protein_id "ENSP00000348704.1"; transcript_support_level "1"; tag "basic"; tag "appris_principal_1"; tag "CCDS"; ccdsid "CCDS5111.1"; havana_gene "OTTHUMG00000015448.4"; havana_transcript "OTTHUMT00000041969.2";

After command 3, I used awk command to split the last column using ; This is the command

command1 input.txt| command2 | command3 | awk -F "\t" -v OFS="\t" '{split($NF,a,";"); $NF=""; print $0,a[1],a[4]}'

I wanted to split on the last field of the file obtained from command3 and then print all the fields except last field and then a[1] and a[4], the split fields, but this adds a tab between columns 1-25 and a[1],a[4]. How can I avoid that?

Thanks

and this is the output

chr6    116732135   116741866   116732135   116732368   116741505   116741866   +   0.79    0.51    0.97    0.77    0.48    0.97    0.02    0.37    'chr6:116732136-116732368:+@chr6:116741506-116741866:+.A.withRI','chr6:116732136-116732368:+@chr6:116741506-116741866:+.B.up_chr6:116732136-116732368:+@chr6:116741506-116741866:+.B.dn'    (0,0):10,(1,0):147,(1,1):1  0:148   (0,0):36,(1,0):161,(1,1):3  0:163,1:1   chr6    +   116732136,116732136 116741866,116741866     gene_id "ENSG00000196911.10"     gene_name "KPNA5"
chr6    116732135   116741866   116732135   116732368   116741505   116741866   +   0.79    0.51    0.97    0.77    0.48    0.97    0.02    0.37    'chr6:116732136-116732368:+@chr6:116741506-116741866:+.A.withRI','chr6:116732136-116732368:+@chr6:116741506-116741866:+.B.up_chr6:116732136-116732368:+@chr6:116741506-116741866:+.B.dn'    (0,0):10,(1,0):147,(1,1):1  0:148   (0,0):36,(1,0):161,(1,1):3  0:163,1:1   chr6    +   116732136,116732136 116741866,116741866     gene_id "ENSG00000196911.10"     gene_name "KPNA5"

1 Answer 1

2

So, given

$ printf 'foo\tbar\ta;b;c;d' | 
    awk -F "\t" -v OFS="\t" '{split($NF,a,";"); $NF=""; print $0,a[1],a[4]}' | cat -A
foo^Ibar^I^Ia^Id$

(where I'm using cat -A to display the tabs as ^I for ease of visualization) you want to eliminate the double tab?

If so, one way would be to decrement NF instead of assigning the empty string to $NF:

$ printf 'foo\tbar\ta;b;c;d' | 
    awk -F "\t" -v OFS="\t" '{split($NF,a,";"); NF--; print $0,a[1],a[4]}' | cat -A
foo^Ibar^Ia^Id$

Another way would be to concatenate the strings instead of printing them as fields - you can do that by removing the , between them:

$ printf 'foo\tbar\ta;b;c;d' | 
    awk -F "\t" -v OFS="\t" '{split($NF,a,";"); $NF=""; print $0 a[1],a[4]}' | cat -A
foo^Ibar^Ia^Id$
1
  • The effect of decrementing NF is strictly speaking "undefined" by the POSIX standard, but it may well work in some implementations of awk.
    – Kusalananda
    Aug 31, 2019 at 18:42

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .