CODEWITHSUNDEEP
shellawksed
MBA
MBA

Reputation: 41

append lines beginning with space to previous line

I want to append every line in my file that begins with spaces to the previous line.

INPUT FILE

roleOccupant:uid=standard_rhc,ou=tenants,ou=users,dc=e-depot,dc=rad,dc
 =lan
roleOccupant:uid=standard_test,ou=tenants,ou=users,dc=e-depot,dc=rad
 ,dc=lan
roleOccupant:uid=standard_NLHaNA,ou=tenants,ou=users,dc=e-depot,dc=
 rad,dc=lan
roleOccupant:uid=standard_dept,ou=tenants,ou=users,dc=e-depot,dc=rad
 ,dc=lan

OUTPUT FILE

roleOccupant:uid=standard_rhc,ou=tenants,ou=users,dc=e-epot,dc=rad,dc=lan
roleOccupant:uid=standard_test,ou=tenants,ou=users,dc=e-depot,dc=rad,dc=lan
roleOccupant:uid=standard_NLHaNA,ou=tenants,ou=users,dc=e-depot,dc=rad,dc=lan
roleOccupant:uid=standard_dept,ou=tenants,ou=users,dc=e-depot,dc=rad,dc=lan

Thanks in advance

Upvotes: 1

Views: 811

Answers (8)

Ed Morton
Ed Morton

Reputation: 203712

Given the data you posted:

$ awk '{ORS=(sub(/^ /,"")?RS:"")}1' file
roleOccupant:uid=standard_rhc,ou=tenants,ou=users,dc=e-depot,dc=rad,dc=lan
roleOccupant:uid=standard_test,ou=tenants,ou=users,dc=e-depot,dc=rad,dc=lan
roleOccupant:uid=standard_NLHaNA,ou=tenants,ou=users,dc=e-depot,dc=rad,dc=lan
roleOccupant:uid=standard_dept,ou=tenants,ou=users,dc=e-depot,dc=rad,dc=lan

With other data YMMV so hopefully you put some thought into your sample.

Upvotes: 0

potong
potong

Reputation: 58453

This might work for you (GNU sed):

sed ':a;N;s/\n //;ta;P;D' file

Read two lines into the pattern space and if an appended line begins with a space, remove the preceeding newline and the space and repeat. Otherwise, print the first line in the pattern space, delete the first line and repeat.

Upvotes: 0

Rahul Verma
Rahul Verma

Reputation: 3089

Possible solutions using awk

$ awk -v RS= '{gsub(/\n /,"")}1' file
roleOccupant:uid=standard_rhc,ou=tenants,ou=users,dc=e-depot,dc=rad,dc=lan
roleOccupant:uid=standard_test,ou=tenants,ou=users,dc=e-depot,dc=rad,dc=lan
roleOccupant:uid=standard_NLHaNA,ou=tenants,ou=users,dc=e-depot,dc=rad,dc=lan
roleOccupant:uid=standard_dept,ou=tenants,ou=users,dc=e-depot,dc=rad,dc=lan

$ awk -v OFS= -v ORS= '/^ /{$NF=$NF RS}1' file
roleOccupant:uid=standard_rhc,ou=tenants,ou=users,dc=e-depot,dc=rad,dc=lan
roleOccupant:uid=standard_test,ou=tenants,ou=users,dc=e-depot,dc=rad,dc=lan
roleOccupant:uid=standard_NLHaNA,ou=tenants,ou=users,dc=e-depot,dc=rad,dc=lan
roleOccupant:uid=standard_dept,ou=tenants,ou=users,dc=e-depot,dc=rad,dc=lan


$ awk -v ORS= '/^ /{ sub(/ /,""); $NF=$NF RS}1' file
roleOccupant:uid=standard_rhc,ou=tenants,ou=users,dc=e-depot,dc=rad,dc=lan
roleOccupant:uid=standard_test,ou=tenants,ou=users,dc=e-depot,dc=rad,dc=lan
roleOccupant:uid=standard_NLHaNA,ou=tenants,ou=users,dc=e-depot,dc=rad,dc=lan
roleOccupant:uid=standard_dept,ou=tenants,ou=users,dc=e-depot,dc=rad,dc=lan

$ awk '/^[^ ]/{ORS=""} /^ /{OFS=""; $1=$1; ORS=RS}1' file
roleOccupant:uid=standard_rhc,ou=tenants,ou=users,dc=e-depot,dc=rad,dc=lan
roleOccupant:uid=standard_test,ou=tenants,ou=users,dc=e-depot,dc=rad,dc=lan
roleOccupant:uid=standard_NLHaNA,ou=tenants,ou=users,dc=e-depot,dc=rad,dc=lan
roleOccupant:uid=standard_dept,ou=tenants,ou=users,dc=e-depot,dc=rad,dc=lan

I would recommend 1st, 2nd and 3rd soluion.
4th one is just more verbose version of 2nd one.

There are more but I think this would be sufficient.

Upvotes: 1

hek2mgl
hek2mgl

Reputation: 158060

With sed:

sed ':a;N;$!ba;s/\n //g' file

Better explained in multiline version:

# Define a label called 'a'
:a
# Read the next line of input and append it to the
# internal pattern buffer
N
# $ means 'the last line of input'
# ! negates the match
# ba (b)ranches back to label 'a'
# Means:
# Unless we've reached the end of the input file
# read the file line by line into the pattern buffer 
$!ba
# Once the end of the input file is reached we substitute
# every occurrence of newline+space by an empty string
s/\n //g
# sed will print the result of the substitution by default

PS: Depending on your version of sed you may need to use:

s/
 //g

instead of

s/\n //g

Meaning you need to use a literal newline if \n isn't supported

And btw, if your input file only contains pairs of two lines where one line has to be appended to the previous, and only that(!), then it could be as as simple as:

sed 'N;s/\n //' file

This would also not require to buffer the whole file in memory.

Upvotes: 1

karakfa
karakfa

Reputation: 67497

awk to the rescue!

$ awk '{if(sub(/^ /,p,$0)) print; else p=$0}' file

roleOccupant:uid=standard_rhc,ou=tenants,ou=users,dc=e-depot,dc=rad,dc=lan
roleOccupant:uid=standard_test,ou=tenants,ou=users,dc=e-depot,dc=rad,dc=lan
roleOccupant:uid=standard_NLHaNA,ou=tenants,ou=users,dc=e-depot,dc=rad,dc=lan
roleOccupant:uid=standard_dept,ou=tenants,ou=users,dc=e-depot,dc=rad,dc=lan

Upvotes: 0

William Pursell
William Pursell

Reputation: 212298

perl -0pe 's/\n +//g' input-file

Upvotes: 0

RomanPerekhrest
RomanPerekhrest

Reputation: 92854

Another awk approach:

awk '{ printf "%s%s",$1,(/^ /? ORS:"") }' file

The output:

roleOccupant:uid=standard_rhc,ou=tenants,ou=users,dc=e-depot,dc=rad,dc=lan
roleOccupant:uid=standard_test,ou=tenants,ou=users,dc=e-depot,dc=rad,dc=lan
roleOccupant:uid=standard_NLHaNA,ou=tenants,ou=users,dc=e-depot,dc=rad,dc=lan
roleOccupant:uid=standard_dept,ou=tenants,ou=users,dc=e-depot,dc=rad,dc=lan

Upvotes: 0

RavinderSingh13
RavinderSingh13

Reputation: 133545

Following awk may help you in same.

awk '{sub(/^ +/,"");printf("%s%s",$0~/^role/?(NR==1?$0:RS $0):$0,$0)} END{print ""}'   Input_file

Upvotes: 0

Related Questions