Search Results

Not sure if this works for you: replace gsub (/[ "}]*/, _) with gsub (/[ \r"}]*/, _). Be aware that the space in the search pattern is a <space> plus a <TAB> char!

Which is an indicator for an Non-*nix-text, probably created by an MS application. Try - if available - dos2unix to convert your input file. Other means have been published in these fora as well.

Not sure I understand your problem. If the original file still exists, tryawk '/gender/ {gsub (/[ "}]*/, _); T = $2 $6; sub ("data", _, T); print "echo mv " T " " $6}' FS="[:,]" file | sh
mv...

Howsoever, tryawk '/gender/ {gsub (/[ "}]*/, _); T = $2 $6; sub ("data", _, T); print "echo mv " $6 " " T}' FS="[:,]" file | sh
mv data1 f1
mv data2 f2
mv data3 m3

Remove the "echo" if...

Any attempt/idea/thought from your side?

I used files from dir1 as reference since you said same number of files with exactly same names, $i is file from dir1 and $file is file from dir2 to understand clearly, better you insert below line...

[akshay@localhost tmp]$ cat test.sh
#!/usr/bin/env bash

dir1="/tmp/test1/*.txt"
dir2="/tmp/test2"
output="/tmp/new"


mkdir -p $output

for i in $dir1; do

filename=${i##*/}
...

# set input FieldSeparator (FS) to ,
# set the value to be be subtracted to 1 and assigned it to the internal awk variable s
awk -F, -v s=1 '
# for each record/line, execute block within {}
{...

awk -F, -v s=1 '{for (i=1;i<=NF;i++) if($i>=s) $i-=s}1' OFS=, myFile

With that input, we're seeing the output you seem to want.
Please post the output from the command:
od -bc file
where file is the input file you fed to either of the awk scripts provided by...

Both proposals deliver the desired output, not the errorneous one you posted. Are you sure there's no hidden control chars in the input file? Did yo run the codes exactly as posted?

Bit simplified version of vgersh99's solution

$ awk -F, '{for(i=1; i<=NF; i++)$i=split($i,a," ")}1' OFS="," file
1,2,3
1,3,1

awk -F, '{for(i=1;i<=NF;i++) printf("%d%s", split($i,a, " "), (i==NF)?ORS:OFS)}' OFS=',' myFile

Your script appears to have issues to em.

What follows is what your script does, which is probably not what you intended.

mkdir result/$doc/02-data
-> create a new empty...

Hi.

I like the solution of radoulov. I also like to break tasks down into manageable steps. One way to do that is to use code that packages up common tasks. A source for that is the missing...

If Perl is acceptable:
perl -F, -lane'
push @d, [@F];
END {
$, = ",";
for $i (0..$#F) {
push @o, [sort { $a <=> $b } map $d[$_][$i], 0..$#d]
}
for $i (0..$#d) {
...

Hi A-V!

I forgot to use -d, with paste command you can remove tr -s '\t' ',', and replace

final_cmd="paste $arg | tr -s '\t' ','"

with

final_cmd="paste -d, $arg" to simplify code.

Try :

#!/bin/ksh

# Loop through files
for file in *.csv; do

# counting fields from first line and assuming fields in remaining lines are equal to first line
field=$(awk -F, 'NR==1{print...

Try :

$ cat f
1,5,7
3,4,9
2,6,8

$ paste <(cut -d, -f1 f | sort) <(cut -d, -f2 f | sort) <(cut -d, -f3 f | sort) | tr -s '\t' ','
1,4,7
2,5,8
3,6,9

This looks like you're just combining both files and sorting them.
So if file x.x is one of the files and y.y the other, then,...

cat x.x y.y | sort
Please clarify.

If you're sure you just want to delete the 1st three bytes of a file:
dd ibs=1 skip=3 if=filename of=filename.new

Like RudiC says, it appears to be UTF-8 encoding with the first three bytes being the BOM 0xEF 0xBB 0xBF. So you could try removing those using:
tail -c+4 file > file.new

or

If you have GNU...

Try using the iconv program. That is the way I ever translate character sets. For example if I have utf-16 and I want ascii it is just:
iconv -f utf-16 -t ascii < input.file

Read the man page...

Apply these changes:
END {
printf "\t"
for ( k in P )
printf "%s\t", k
printf "\tCount\n"

for (...

The program will not work if you feed one input file at a time using a for loop.

You could pass them all at once:

awk '
-- code --
' accept-all.txt People/*