How can I extract part of the output of a curl command and assign it to a shell variable - part 2?

Question

% Total % Received % Xferd Average Speed Time Time Time Current
Dload Upload Total Spent Left Speed
100 188 100 188 0 0 180 0 0:00:01 0:00:01 --:--:--
180{"result":{"fileId":"3a29ca7a-813a-4b3b-9b9c-
6f9251d2171d","percentComplete":100.0,"status":"complete"},"meta":{"requestId":"*************************************************************************","httpStatus":"200 - OK"}}

Again - I need to extract the highlighted part that follows "fileId":

I tried using the following -

VAR2="$(curl <...> | sed -E -n 's/.*fileId":"([^"]+).+/\2/p')"

But when I run echo $VAR2 I do not get any output

p.s. ideally I want this to happen only when percentComplete:100.0 (but this is not a priority, extracting the fileId is the priority for now)

Is that output one line or many? Please use the formatting tools to format your question, and make sure you format the output as code. — terdon, May 15 '20 at 14:41
Stéphane's answer to your other question could easily be extended to cover this case: https://unix.stackexchange.com/a/586809/116858 — Kusalananda, May 15 '20 at 14:43
I tried using the solution Stephanie provided - did not work for me — Viraj Deshpande, May 15 '20 at 14:47

score 0 · Accepted Answer · answered May 15 '20 at 14:53

0

This should work:

VAR2="$(curl<..> | sed -E -n 's/.*fileId":"([^"]+).+/\1/p')"

The problem with your try is that you used \2 in the sed command, but you only have one regex group. I think you should take a look at the sed info page to understand how sed works.

answered May 15 '20 at 14:53

Francesco

836

3

Or even better, parse the JSON with a JSON parser... – Kusalananda May 15 '20 at 15:00

How can I extract part of the output of a curl command and assign it to a shell variable - part 2?

1 Answers1

Linked