How to replace string after colon using sed, awk, regex, yq etc and can run using shell script

gaurav agnihotri 0
Viewed 80

I have a yaml file and in that file I need to replace a specific string that comes after oauth2accesstoken: using a shell script

my current shell script :-

#!/bin/sh

password=$(gcloud auth print-access-token)

password="$password" yq -i '.stringData.creds |= "oauth2accesstoken:" + strenv(password)' /test.yaml

Lets Assume Replace Value is "abcdefg1234"

File :-

kind: Secret
metadata:
  name: cred
type: Opaque
stringData:
  creds: oauth2accesstoken:ya29.a0AJ8fv7hrWsNVlDaXa-j6IUwBRdxt6GDDXMu1234efrtyhAjPWx0uw0174 # dummy key

Output Should be like this:-

kind: Secret
metadata:
  name: cred
type: Opaque
stringData:
  creds: oauth2accesstoken:abcdefg1234

Error:-

Error: unknown command ".stringData.creds |= \"oauth2accesstoken:\" + strenv(password)" for "yq"
Run 'yq --help' for usage.

Note:-Above keys are dummy keys.

awk sed yq regex
3 Answers

With Go yq aka mikefarah/yq, its pretty straightforward to just select the right field and update its value

yq '.stringData.creds |= "oauth2accesstoken:abcdefg1234"' yaml

or pass the value from a shell variable

token="abcdefg1234" yq '.stringData.creds |= "oauth2accesstoken:" + strenv(token)' yaml

To modify the file in-place use the -i flag e.g. yq -i <rest-of-the-code>. Tested on version 4.27.5

Inian0 
sed 's/(oauth2accesstoken:).*/\1:abcdefg1234/g' input.txt

This will search for line with oauth2accesstoken and replace all after it.

Jotne0 
front=$(sed -n '/creds/p' file | cut -d':' -f1-2)
back="abcdefg1234"

echo "${front}:${back}" >> file

cat > ed1 <<EOF
$
-1d
wq
EOF

ed -s file < ed1
rm -v ./ed1
petrus40
Related