Extracting data from a file with a regular expression in sed

Question

How can I use sed to extract f607b9ab-1452-4758-b047-6b620591211b by using cmrepo-222-02 as my reference?

["+--------------------------------------+---------------+----------+------+-------------+", "| ID                                   | Name          | Status   | Size | Attached to |", "+--------------------------------------+---------------+----------+------+-------------+", "| f607b9ab-1452-4758-b047-6b620591211b | cmrepo-222-02 | creating |    1 |             |", "| 5e9b27fd-46b4-4950-9dc1-a838be21aff7 | cmrepo-222-01 | creating |    1 |             |", "+--------------------------------------+---------------+----------+------+-------------+"]

Answer 1

It looks like a json array of strings representing the lines of the output (intended for human consumption, not for post-processing) of some SQL query. So you could use jq to decode the json and awk to attempt to extract data from that formatted table.

<that-file jq -r '.[]' - |
  awk -F '[[:space:]]*[|][[:space:]]*' '$3 == "cmrepo-222-02" {print $2}'

But best would be to do the right SQL query in the first place.

Answer 2

Using sed

$ ref=cmrepo-222-02
$ sed "s/.* \([[:alnum:]-]\+\).*$ref.*/\1/" input_file
f607b9ab-1452-4758-b047-6b620591211b