CODEWITHSUNDEEP
sed
ERH
ERH

Reputation: 23

Replacing part of pattern with sed

I'm looking to replace a space that always comes after a number with a | in the middle of a pattern. There are also similar patterns later in the line sometimes that I do not wish to replace (see first/second lines in the examples).

example:

12315 asdfea 1 1ffesa
45456 asefasef 1 era
12 asfase
4 4aefs

what I need:

12315|asdfea 1 1ffesa
45456|asefasef 1 era
12|asfase
4|4aefs

I have tried this:

 sed 's/\([0-9][ ][a-zA-Z]\)/|/g' file.txt

However this deletes the pattern such that it looks like this:

|sdfea 1 1ffesa
|sefasef 1 era
|sfase
|4aefs

Which is not what I need.

Upvotes: 2

Views: 284

Answers (1)

Sundeep
Sundeep

Reputation: 23667

For given sample input/output,

$ sed 's/ /|/' file.txt
12315|asdfea 1 1ffesa
45456|asefasef 1 era
12|asfase
4|4aefs

By default, only first match will be replaced. g modifier will replace all matches


To replace first matched space between a digit and alphabet (matching depends on locale too)

$ sed 's/\([0-9]\) \([a-zA-Z]\)/\1|\2/' file.txt
12315|asdfea 1 1ffesa
45456|asefasef 1 era
12|asfase
4 4aefs

This uses capture group and backreferences. Note that last line is not modified

Upvotes: 1

Related Questions