JMESPath join output and strip unwanted chars

I have the following json;

[
  {
    "compartment-id": "CompartmentID-123",
    "defined-tags": {},
    "display-name": "Test-123",
    "freeform-tags": {},
    "id": "ID-ABC",
    "kms-key-id": "",
    "lifecycle-state": "ACTIVE",
  }
]

I am looking to join the id and display-name params into a comma seperate string with " and [] stripped, like so:

ID-ABC,Test-123

Closest I've managed to get so far is:

oci fs file-system list -c $compart --availability-domain $ad --query 'data[].[id,"display-name"][]' | tr -d '" '
[
ID-ABC,
Test-123
]

Wondering if there's a cleaner way of doing this all within JMESPath without piping output to tr, jq, sed or awk etc etc

UPDATE based on input from β.εηοιτ.βε

So close...

oci fs file-system list -c $compart --availability-domain $ad3 --query 'data[0].join(',', [id, "display-name"])'

Returns

ParseError: invalid token: Parse error at column 13, token "," (COMMA), for expression:
"data[0].join(,, [id, "display-name"])"

However playing around with quotes, best I can get is by using;

oci fs file-system list -c $compart --availability-domain $ad3 --query "data[0].join(',', [id, 'display-name'])"
Private key passphrase:
"ID-ABC,display-name"

I'm beginning to wonder if there's something wrong with my local settings or shell, whereby it's getting confused by the quotes marks?

Solution

Since you do have an array, you will first need to target the first element of the array, to get rid of it. Since an array is 0-based, this can easily be achieved via:

data[0]

which gives

{
  "compartment-id": "CompartmentID-123",
  "defined-tags": {},
  "display-name": "Test-123",
  "freeform-tags": {},
  "id": "ID-ABC",
  "kms-key-id": "",
  "lifecycle-state": "ACTIVE"
}

Then, in order to create the string you are looking for, you can use the join function, that accepts, as parameter, a glue character and array.

You can get the array you look for pretty easily, using filters and multiselect lists:

data[0].[id, "display-name"]

which gives

[
  "ID-ABC",
  "Test-123"
]

Now we just need to apply the join function on top of all this:

data[0].join(',', [id, "display-name"])

which finally gives:

"ID-ABC,Test-123"

Additional notes on quoting in JMESPath:

```
'some-string'
```
is a raw literal string and won't be interpreted
```
`some-string`
```
is another form of raw literal string and won't be interpreted
```
"some-string"
```
is a literal expression and will be interpreted

So:

data[0].join(',', [id, "display-name"])

and

data[0].join(`,`, [id, "display-name"])

are two strictly equals queries.

While

data[0].join(',', [id, 'display-name'])

has a totally different meaning and would end you with the string display-name being the second element of your array, so it will result in

"ID-ABC,display-name"

All of this based on the JSON structure:

{
   "data":[
      {
         "compartment-id":"CompartmentID-123",
         "defined-tags":{},
         "display-name":"Test-123",
         "freeform-tags":{},
         "id":"ID-ABC",
         "kms-key-id":"",
         "lifecycle-state":"ACTIVE"
      }
   ]
}