Strings extraction from text file with sed command

Question

I have a text file which contains some lines as the following:

ASDASD2W 3ASGDD12 SDADFDFDDFDD W11 ACC=PNO23 DFSAEFEA EAEDEWRESAD ASSDRE 
AERREEW2 3122312 SDADDSADADAD W12 ACC=HH34 23SAEFEA EAEDEWRESAD ASEEWEE 
A15ECCCW 3XCXXF12 SDSGTRERRECC W43 ACC=P11 XXFSAEFEA EAEDEWRESAD ASWWWW 
ASDASD2W 3122312 SDAFFFDEEEEE SD3 ACC=PNI22 ABCEFEA EAEDEWRESAD ASWEDSSAD 
...

I have to extract the substring between the '=' character and the following blank space for each line , i.e.

PNO23
HH34
P11
PNI22

I've been using the sed command but cannot figure out how to ignore all characters following the blank space.

Any help?

Ignacio Vazquez-Abrams · Accepted Answer · 2012-06-30 09:31:21Z

2

Use the right tool for the job.

$ awk -F '[= ]+' '{ print $6 }' input.txt
PNO23
HH34
P11
PNI22

answered Jun 30, 2012 at 9:31

Ignacio Vazquez-Abrams

804k160 gold badges1.4k silver badges1.4k bronze badges

Sign up to request clarification or add additional context in comments.

8 Comments

user1492786 Over a year ago

Doubtless awk is a powerful tool and your code will help me a lot but i'm just curious about achieving the same result with sed even if it were harder.

Jo So Over a year ago

Using sed is certainly not harder in this case. "The right tool for the right job" is just wrong in this context. Sed is the right tool. Awk is better for tabular data and quick hacks and calculations. Sed makes regular expressions easy and is more declarative. See my answer.

William Pursell Over a year ago

@JoSo Sed grammar may make it easier to use regular expressions than awk in some situations, but this is not one of them.

Jo So Over a year ago

@WilliamPursell: Please explain. I think s/.*=//; s/ .*// is dead simple.

William Pursell Over a year ago

@JoSo Although s/.*=//; s/ .*// is simple, [= ]+ is much simpler.

|

Jo So · Accepted Answer · 2012-06-30 11:16:04Z

2

Sorry, but have to add another one because I feel the existing answers are just to complicated

sed 's/.*=//; s/ .*//;' inputfile

answered Jun 30, 2012 at 11:16

Jo So

26.9k6 gold badges46 silver badges60 bronze badges

Comments

potong · Accepted Answer · 2012-06-30 15:04:22Z

1

This might work for you:

sed -n 's/.*=\([^ ]*\).*/\1/p' file

or, if you prefer:

sed 's/.*=\([^ ]*\).*/\1/p;d' file

answered Jun 30, 2012 at 15:04

potong

59.3k6 gold badges55 silver badges92 bronze badges

Comments

tripleee · Accepted Answer · 2012-06-30 10:41:43Z

0

Put the string you want to capture in a backreference:

sed 's/.*=\([^ =]*\) .*/\1/'

or do the substitution piecemeal;

sed -e 's/.*=//' -e 's/ .*//'

answered Jun 30, 2012 at 10:41

tripleee

192k37 gold badges318 silver badges369 bronze badges

Comments

Dennis Williamson · Accepted Answer · 2012-06-30 10:42:42Z

0

sed 's/[^=]*=\([^ ]*\) .*/\1/' inputfile

Match all the non-equal-sign characters and an equal sign. Capture a sequence of non-space characters. Match a space and the rest of the line. Substitute the captured string.

answered Jun 30, 2012 at 10:42

Dennis Williamson

364k95 gold badges386 silver badges446 bronze badges

Comments

Sisay Chala · Accepted Answer · 2015-08-09 16:59:05Z

0

A chain of grep can do the trick.

grep -o '[=][a-zA-Z0-9]*' file | grep -o '[a-zA-Z0-9]*'

answered Aug 9, 2015 at 16:59

Sisay Chala

1411 silver badge3 bronze badges

Collectives™ on Stack Overflow

Strings extraction from text file with sed command

6 Answers 6

8 Comments

Comments

Comments

Comments

Comments

Comments

Your Answer

Hot Network Questions

Collectives™ on Stack Overflow

6 Answers 6

8 Comments

Comments

Comments

Comments

Comments

Comments

Your Answer

Sign up or log in

Post as a guest

Related