CODEWITHSUNDEEP
sed
qin peter
qin peter

Reputation: 359

how to use \digit in sed regex

I've read sed info. In "3.3 Overview of Regular Expression Syntax".
There is a description:

\digit
    Matches the digit-th \(...\) parenthesized subexpression in the regular expression.
    This is called a back reference. Subexpressions are implicitly numbered by
    counting occurrences of \( left-to-right.

I don't know what it means. Who can give me a example?

Upvotes: 2

Views: 1880

Answers (1)

fedorqui
fedorqui

Reputation: 289495

Sure!

$ echo "23 45" | sed -r 's/^([0-9]*)/---\1---/'
---23--- 45

Graphically:

sed -r 's/^([0-9]*)/---\1---/'
#          ^^^^^^^^    ^^
#          capture -----|
#                   print back

As you see, in a sed expression on the form s/search/replace, you can "capture" a pattern in the search block and then print it back in the replace block by using \1, \2, ... The number is sequential and corresponds to the 1st, 2nd, ... group that has been captured.

$ echo "23 45" | sed -r 's/^([0-9]*) (.)/YEAH \2---\1---/'
YEAH 4---23---5

Upvotes: 5

Related Questions