Find and replace strings in file using regex in sed. Script doesn't work

Question

I have a file, in which I have to find and change strings by specific pattern (phone number). The regex is:

^\+[0-9]{3} \([0-9]{2}\) [0-9]{7}$

When I use it in command:

grep "^\+[0-9]{3} \([0-9]{2}\) [0-9]{7}$" -E filename

It works. But when I try to use it in sed to replace all parenthesis by spaces and add spaces in 13 and 15 position, it doesn't works and I don't have ideas why.

My variants are:

sed '/^\+[0-9]{3} \([0-9]{2}\) [0-9]{7}$/s/[()]//' filename

(only for replacing parenthesis)

sed -e '/^\+[0-9]{3} \([0-9]{2}\) [0-9]{7}$/s/[()]//' -e '/^+[0-9]{2} ([0-9]{2}) [0-9]{7}/s/./& /11;s/./& /14' filename

file structure:

    +380 44 123 45 67
    +380 (44) 1234567
    +350 (56) 1454557
    +330 (76) 1255557
    +380 44 3534 45 67
    +320 (45) 1237887
    +310 (54) 1939997
    adasd
    asdddddddddddd
    sssdad

expected output:

    +380 44 123 45 67
    +380 44 123 45 67
    +350 56 145 45 57
    +330 76 125 55 57
    +380 44 3534 45 67
    +320 45 123 78 87
    +310 54 193 99 97
    adasd
    asdddddddddddd
    sssdad

you need to add -E option to sed command as well, which you use with grep for same purpose of using extended regex — Sundeep
– Sundeep, Commented Oct 5, 2016 at 15:59
@Sundeep thanks for response! but when I write: sed '/^\+[0-9]{3} ([0-9]{2}) [0-9]{7}$/s/[()]//' -E filename it removes only opening parenthesis and I couldn't fix it( — andrew_j
– andrew_j, Commented Oct 5, 2016 at 16:04
well I didn't see if there were issues with regex as well.. would be good if you can add sample input file with expected output so that it can be tested... see stackoverflow.com/help/mcve for more details — Sundeep
– Sundeep, Commented Oct 5, 2016 at 16:05
@Sundeep I've added it! I need to find and replace numbers in format +380 (44) 1234567 to format +380 44 123 45 67 — andrew_j
– andrew_j, Commented Oct 5, 2016 at 16:08

Graham · Accepted Answer · 2017-09-23 16:29:27Z

1

Here's one way to do it:

$ cat ip.txt 
+380 44 123 45 67
+380 (44) 1234567
+350 (56) 1454557
+330 (76) 1255557
+380 44 3534 45 67
+320 (45) 1237887
+310 (54) 1939997
adasd
asdddddddddddd
sssdad

$ sed -E 's/^(\+[0-9]{3}) \(([0-9]{2})\) ([0-9]{3})([0-9]{2})([0-9]{2})$/\1 \2 \3 \4 \5/' ip.txt 
+380 44 123 45 67
+380 44 123 45 67
+350 56 145 45 57
+330 76 125 55 57
+380 44 3534 45 67
+320 45 123 78 87
+310 54 193 99 97
adasd
asdddddddddddd
sssdad

() can be used to surround a pattern so that the matched text inside them can be backreferenced in replacement section
\1 corresponds to first such captured group, \2 to second and so on
To match ( or ) themselves, we need to use escape them like $ and $
So, here the numbers are captured as per required output, excluding the () present in input line so that they are not part of output

edited Sep 23, 2017 at 16:29

Graham

7,86020 gold badges67 silver badges92 bronze badges

answered Oct 5, 2016 at 16:15

Sundeep

23.9k2 gold badges35 silver badges131 bronze badges

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

3 Comments

andrew_j Over a year ago

Thanks! I don't have any idea how,but it works perfectly!

Sundeep Over a year ago

@andrew_j glad to know it works, I will add an explanation and let you know :)

andrew_j Over a year ago

Thank you very much, I've got it!

Smithson · Accepted Answer · 2016-10-05 16:15:37Z

0

Your sed command is wrong. My way:

sed -E 's/^\+[0-9]{3} \([0-9]{2}\) [0-9]{7}$/[()]/'

edited Oct 5, 2016 at 16:15

answered Oct 5, 2016 at 16:08

Smithson

1414 bronze badges

2 Comments

Smithson Over a year ago

I have trouble, -E of couse.

andrew_j Over a year ago

output doens't meet requirements that I've mentioned in question(

Jan Nielsen · Accepted Answer · 2016-10-05 17:06:18Z

0

Use:

sed -e 's|[()]||g' so-tel.txt | sed -E 's|([0-9]{3})([0-9]{2})([0-9]{2})|\1 \2 \3|'

to transform so-tel.txt:

+380 44 123 45 67
+380 (44) 1234567
+350 (56) 1454557
+330 (76) 1255557
+380 44 3534 45 67
+320 (45) 1237887
+310 (54) 1939997
adasd
asdddddddddddd
sssdad

into:

+380 44 123 45 67
+380 44 123 45 67
+350 56 145 45 57
+330 76 125 55 57
+380 44 3534 45 67
+320 45 123 78 87
+310 54 193 99 97
adasd
asdddddddddddd
sssdad

Explanation:

's|[()]||g'

substitute any ( and ) with nothing, globally

's|([0-9]{3})([0-9]{2})([0-9]{2})|\1 \2 \3|'

substitute and capture seven successive digits in lengths 3, 2, and 2, with the captured digit groups separated by a space.

edited Oct 5, 2016 at 17:06

answered Oct 5, 2016 at 16:17

Jan Nielsen

12k16 gold badges75 silver badges136 bronze badges

3 Comments

andrew_j Over a year ago

thanks, but problem is that I also need to add spaces in some positions only in strings which fit by regex

andrew_j Over a year ago

Yes, That's what I neeed!

andrew_j Over a year ago

You can see expected output in question

Collectives™ on Stack Overflow

Find and replace strings in file using regex in sed. Script doesn't work

3 Answers 3

3 Comments

2 Comments

3 Comments

Your Answer

Hot Network Questions

Collectives™ on Stack Overflow

3 Answers 3

3 Comments

2 Comments

3 Comments

Your Answer

Sign up or log in

Post as a guest

Related