Search Results

No. With my script you need the END clause to print the trailing <newline>; with spacebar's script you need the END clause to print the entire last line.
Until you hit EOF, you don't know that...

Try this:
awk '/^1_/{if (x)print x;x="";}{x=(!x)?$0"|":x$0;}END{print x;}'

FILENAME is a built-in variable but it would not be useful here, as it contains the name of the current input file, in this case "input.tab" ..

Try


awk '
BEGIN{OFS=FS="|"}
{
if(NF>max) max=NF
for(i=1;i<=NF;i++)
{
start=match(a[i],$i)
if(start && substr(a[i],(start+RLENGTH),1) ~ /^[,]*$/)
continue
if(a[i])
...

Try this:

awk -F\| 'BEGIN{OFS=FS} {start=match(a[$1],$2);if(start && substr(a[$1],(start+RLENGTH),1) ~ /^[,]*$/) next; if(a[$1]) a[$1]=a[$1]","$2; else a[$1]=$2} END{for(i in a) print i,a[i]}' file

Try:
#!/usr/bin/awk -f

BEGIN{FS=OFS="|"}

function convert (n, v,w,x,y,r1,r10,r100,r1000){
split("I II III IV V VI VII VIII IX",r1," ")
split("X XX XXX XL L LX LXX LXXX XC",r10," ")
...

I guess you have your reasons, but I don't see the point of an array of sequential numbers.


awk '
BEGIN {
for( i = 1; i <= 10000; i++ )
numbers[i] = i;
}

You would need to do something like this to shift the fields:
awk '{for(i=1; i<=NF; i++) if($i=="") for(j=i;j<=NF;j++)$j=$(j+1)}1' FS=\| OFS=\| infile
output:
1|2|3|4|5|6||
But that still leaves...

Carlo@CRM-MT4 ~
$ echo "||AAAA 1 2.; AAAA 3.|AAAA 5 6.||" | awk '{print gensub (/([^[:blank:]]+[[:blank:]]+[[:digit:]]+[[:blank:]]+[[:digit:]]+)[^.;\|]*/,"\\1","g")}'
||AAAA 1 2.; AAAA 3.|AAAA 5...

awk -F'|' '{for(i=3; i<=NF; ++i) print $1FS$2FS$i}'

Well, they worked on your original example :p.

EDIT:
Try:
Carlo@CRM-MT4 ~
$ echo "AAAA 1 2 f45 iop 234-RT.; AAAA 123 456 gf kuygb 2iuygb.; AAAA 678 89 wer4k lh8." |
> awk '{print gensub...

awk -F'|' '{print $0"|"$2}' inputfile

There is one closing brace missing. It may be that your input file differs from your sample, for example, when it contains some tabs instead of spaces. What happens when you cut and paste your...

This will check for the largest length in every column and will space accordingly.

perl -F'\t' -lane '
BEGIN {
open I, "< inputfile";
for (<I>) {
chomp;
@x = split...

nawk -F"\t" 'BEGIN{b=0}{for(i=1;i<=NF;i++){a=length($i);if(a>b)b=a;}}END{print b}' input.txt
the above awk will give the longest field in your file.

after that you can pass that value to awk...

Yes, you can start with perl -F'\t' -lane '.....

perl -lane '@x=split /,/,$F[1]; print "$F[0]\t$_\t$F[2]" for(@x)' inputfile

Try: awk -F"\t" -vOFS="\t" '{n=split($2,a,",");for (i=1;i<=n;i++) print $1,a[i],$3}' file

the join command concatenates, so:
print join ("\t", $-[1]+1 . "-" . $+[1] , @F )

Try:
perl -F\\t -lane 'if ($F[5] =~ s/.*?([13]{1,5}tomatoes[13]{1,5}).*/$1/p) {print join ("\t", @F, $-[1]+1 . "-" . $+[1] )}' infile

You are selecting on the sixth column instead of the second column in the example of the starting post, so then you need to replace F[1] with F[5] and are these fields tab separated? Slight...

awk splits the line into fields based on whitespaces by default. So $2 is not "bb11 toma toes 111b bbbbb" but just "bb11". Tell awk to split fields based on tab '\t' and then $2 would be what you...

In the MAC OSX, can you try the below sed


sed "s/[[:<:]]$a[[:>:]]/$b/g"

if you use single quote in the sed '

then, it will not evaluate the $a and $b values.

so, it search for $a and replce it to $b

ls -1