CODEWITHSUNDEEP
bashawksed
Enric Agud Pique
Enric Agud Pique

Reputation: 1105

removing comma and -

How can I clean in bash the following line

-,2,8,--,-,2,6,--,-,-1,2,--,-,0,4,--,-,1,5,--,-,-5,2

in order to get

 2,8,2,6,-1,2,0,4,1,5,-5,2

I only want positive and negative numbers separated by comma

Upvotes: 0

Views: 64

Answers (4)

John Bollinger
John Bollinger

Reputation: 180103

You can run it through this sed script:

sed -e 's/--*,//g' -e 's/,--*$//'

That wipes out every substring consisting of one or more minus signs (-) followed by a comma (,), and, separately, any trailing substring consisting of a comma followed by one or more minus signs.

Upvotes: 2

Ed Morton
Ed Morton

Reputation: 203229

$ awk 'BEGIN{FS=OFS=","} {j=0; for (i=1;i<=NF;i++) if ($i~/^-?[0-9]/) printf "%s%s", (++j>1?OFS:""), $i; print ""}' file 
2,8,2,6,-1,2,0,4,1,5,-5,2

Loop through each field, testing if the current field is a number optionally preceded with -. If it is then print the field preceded by nothing if it's the first matching field on the line or by the output field separator otherwise. When done printing the fields print a newline.

Upvotes: 0

anubhava
anubhava

Reputation: 784998

A gnu-awk solution:

str='-,2,8,--,-,2,6,--,-,-1,2,--,-,0,4,--,-,1,5,--,-,-5,2'
awk -v FPAT='-?[0-9]+' '{for(i=1; i<=NF; i++) printf "%s%s", $i, (i<NF)?",":RS}' <<< "$str"
2,8,2,6,-1,2,0,4,1,5,-5,2

A non-gnu awk solution:

str=$(awk 'BEGIN{RS=ORS=","} /^-?[[:digit:]]+/{print $1}' <<< "$str")

Output:

echo "${str%,}"
2,8,2,6,-1,2,0,4,1,5,-5,2

Upvotes: 1

jyvet
jyvet

Reputation: 2191

try:

egrep -o "\-?[0-9]+" | sed ':a;N;$!ba;s/\n/,/g'

or:

egrep -o "\-?[0-9]+" | xargs | sed 's/ /,/g'

Upvotes: 0

Related Questions