how to add " to end of specific lines in linux-CodePudding

I have a file as below

"time":"2022-03-01T12:00:25.388Z","duration":0.8255,"error":0
"time":"2022-03-01T12:00:26.809Z","duration":0.29521,"error":0,"state":"pending
"time":"2022-03-01T12:00:29.053Z","duration":2.90688,"error":0,"options":"merge_request.create
"time":"2022-03-01T12:00:00.635Z","duration":0.46049,"error":0
"time":"2022-03-01T12:00:39.351Z","duration":0.76487,"error":0,"state":"failed

I want to add " at the end but only to lines ending as a string

I gave sed -i 's/$/",/' filename but it adds quotes to end of all the lines.

Desired output is

"time":"2022-03-01T12:00:25.388Z","duration":0.8255,"error":0
"time":"2022-03-01T12:00:26.809Z","duration":0.29521,"error":0,"state":"pending"
"time":"2022-301T12:00:29.053Z","duration":2.90688,"error":0,"push_options":"merge_request.create"
"time":"2022-03-01T12:00:00.635Z","duration":0.46049,"error":0
"time":"2022-03-01T12:00:39.351Z","duration":0.76487,"error":0,"state":"failed"

Is there a way to achieve this?

CodePudding user response：

You can use

sed 's/:"[^"]*$/&"/' file > tmp && mv tmp file
sed -i 's/:"[^"]*$/&"/' file # GNU sed

The 's/:"[^"]*$/&"/' command means:

:"[^"]*$ - matches a :" substring and then zero or more chars other than " till end of string
&" - replaces the match with itself and a " char.

See the online demo:

#!/bin/bash
s='"time":"2022-03-01T12:00:25.388Z","duration":0.8255,"error":0
"time":"2022-03-01T12:00:26.809Z","duration":0.29521,"error":0,"state":"pending
"time":"2022-03-01T12:00:29.053Z","duration":2.90688,"error":0,"options":"merge_request.create
"time":"2022-03-01T12:00:00.635Z","duration":0.46049,"error":0
"time":"2022-03-01T12:00:39.351Z","duration":0.76487,"error":0,"state":"failed'
 
sed 's/:"[^"]*$/&"/' <<< "$s"

Output:

"time":"2022-03-01T12:00:25.388Z","duration":0.8255,"error":0
"time":"2022-03-01T12:00:26.809Z","duration":0.29521,"error":0,"state":"pending"
"time":"2022-03-01T12:00:29.053Z","duration":2.90688,"error":0,"options":"merge_request.create"
"time":"2022-03-01T12:00:00.635Z","duration":0.46049,"error":0
"time":"2022-03-01T12:00:39.351Z","duration":0.76487,"error":0,"state":"failed"

CodePudding user response：

Suggesting an awk solution:

awk '/[[:alpha:]]$/{$0=$0"\""}1' input.txt

CodePudding user response：

Using sed

$ sed 's/[a-z]$/&"/' input_file
"time":"2022-03-01T12:00:25.388Z","duration":0.8255,"error":0
"time":"2022-03-01T12:00:26.809Z","duration":0.29521,"error":0,"state":"pending"
"time":"2022-03-01T12:00:29.053Z","duration":2.90688,"error":0,"options":"merge_request.create"
"time":"2022-03-01T12:00:00.635Z","duration":0.46049,"error":0
"time":"2022-03-01T12:00:39.351Z","duration":0.76487,"error":0,"state":"failed"