This is a sample of data.txt:
"line1":"text1",
"line2":"text2",
"line3":"text3",
"line4":"text4",
"line5":"text5",
"line6":"text6",
"line7":"text7",
"line8":"text8",
"line9":"text9",
"line10":"text10"
This is a very simple data, but I'm trying to do this:
file line `"line7":"text7",\n"line8":"text8",` and three lines before that, and save two variables that I can use in my bash:
content_of_line4='text4'
content_of_lint5='text5'
What I think can be done is something like this:
content_of_line4="$(the_awk_or_grep_command)"
content_of_lint5="$(the_awk_or_grep_command)"
What I tried so far:
grep -E "\"line7\":\"text7",\n\"line8\":\"text8\"," -B3 data.txt
As soon as I add \n\"line8\":\"text8\",, the grep finds nothing. If I remove this, it shows.
I need to use the two variables in my bash commands in the future, so I need to store them.
Your input looks like the "inner part" of a JSON object. If so, you could add the "missing" curly braces, and use jq to extract the fields needed. Example:
content_of_line4="$(jq -Rsr '"{\(.)}" | fromjson.line4' input.txt)"
echo "$content_of_line4"
text4
For better reusability, you can also employ the --arg option to use values from the shell for the field name:
content_of_line5="$(jq -Rsr --arg f "line5" '"{\(.)}" | fromjson[$f]' input.txt)"
echo "$content_of_line5"
text5