sed command replace muti line with reg-CodePudding

I got a text like this:

TOKEN = decrypt_aes(
    "189272123124aqephkiz3")

And I want to change it into:

TOKEN = "189272123124aqephkiz3"

How can I make this? I can do this when its in a single line with following command:enter code here`

sed -i "s/decrypt_aes(\(.*\))/\1/g"

But I don`t no how to do when its in multi line

CodePudding user response：

If your file contains only the two lines you showed, then this might help with GNU sed:

sed -i 'N; s/decrypt_aes(\n *//; s/)//' file

From man sed:

N: Append the next line of input into the pattern space.

CodePudding user response：

This might work for you (GNU sed):

sed 'N;s/decrypt_aes.*\(".*"\).*/\1/;P;D' file

Append the following line and if the combined lines matches decrpty_aes followed by "...", replace the match by the string "...".

N.B. If no match is found, each line will be printed as is.

CodePudding user response：

You can use

sed -i '/.*decrypt_aes($/{N;s/^\([^=]*=\).*\(".*"\))$/\1 \2/}' file

Details:

/.*decrypt_aes($/ - matches a line that ends with decrypt_aes( substring, and if it matches, the block that follows is executed
N - append a newline and the next line to the pattern space
s/^$[^=]*=$.*$".*"$)$/\1 \2/ - replaces
- ^$[^=]*=$.*$".*"$)$ - start of string in the pattern space (^), any zero or more chars other than = and then a = are captured into Group 1 (\1), then any text (.*), then a "..." substring (captured into Group 2 (\2)) and then a ) at the end of string
- \1 \2 is the replacement pattern, which is Group 1 value space Group 2 value.

#!/bin/bash
s='TOKEN = decrypt_aes(
    "189272123124aqephkiz3")'

sed '/.*decrypt_aes($/{N;s/^\([^=]*=\).*\(".*"\))$/\1 \2/}' <<< "$s"

Output:

TOKEN = "189272123124aqephkiz3"