S3 plugin doesn't do string interpolation in downstream job

[HungoHung]

Hi all,

I have a drone set like below:
Repo A and Repo B are both enabled in drone and located in github.

Repo A’s .drone.yml:

pipeline:
  build:
    image: hungo/test
    commands:
       - echo $DRONE_TAG
===========================================
It will print the tag correctly if I push a tag to github
===========================================

===========================================
I use @vaijab downstream plugin to pass drone tag to RepoB
===========================================
     trigger:
      image: quay.io/ukhomeofficedigital/drone-trigger:latest
        drone_server:
        drone_token:
        repo: HungoHung/RepoB Calling Repo B
        branch: master
        params: “DRONE_TAG =${DRONE_TAG}”

###################################################

Here is the Repo B’s .drone.yml

workspace:
  base: /opt
  path: src

pipeline:
   build:
     image: hungo/aws-test
     commands:
       - echo “$DRONE_TAG” Here will print out the correct tag
       - ./getBuild.sh Inside this script, echo ${DRONE_TAG} will print out the correct tag

   s3:
     image: plugins/s3
     bucket: drone-build-artifact
     access_key:
     secret_key:
     region: us-west-2
     acl: private
     source: /opt/src/*
     target: /hungotest/_artifact/${DRONE_TAG}
===========================================
I have tried to use $DRONE_TAG or ${DRONE_TAG}, $$DRONE_TAG

  ${DRONE_TAG} will be replaced by empty string
  ex: drone s3 log => target="/hungotest/_artifact/opt/src/.DS_Store"

  $DRONE_TAG will still be $DRONE_TAG
  ex: drone s3 log => target="/hungotest/_artifact/$DRONE_TAG/opt/src/.DS_Store"
===========================================

I have been trying to google and do try&error for few days but still can’t get it through. Could anyone shed some light on this for me? Thank you!

My drone server version is 0.8.0. Thank you
===========================================================
Edit 1:
Not sure if this issue is related to what @bradrydzewski said in this thread

If I understand correctly you are trying to interpolate custom build variables in the yaml. These values are passed to your container as environment variables, but cannot be interpolated in the yaml. If we look at your example … environment: ENV_PARAM=${PARAM1=undefined} The above line would not work because PARAM1 is not available to the yaml pre-process for substitution. It is also unnecessary because PARAM1 is already available to your container as an environment variable. The practice…

But I still don’t have idea how to solve my uploading issue in downstream job based on the passed info(tag).

===========================================================
Edit 2:
I tried to use following command to pass variable directly but it still see ${DRONE_TAG} as empty string
/home/ubuntu # drone build start -p DRONE_TAG=newdx HungoHung/RepoB 36
Starting build HungoHung/RepoB#39

Of course if I put the S3 part in RepoA (Main Job) and test with tag, the following will work like a charm
     source: /opt/src/*
     target: /hungotest/_artifact/${DRONE_TAG}