I have a following jsonb data inside the jsonb_col

  {
     "a": { "keyA": 1 , "keyB": 2 }, 
     "b": { "keyA": 1 , "keyB": 3 }, 
     "c": { "keyA": 1 , "keyC": 4 } 
  }

I want to duplicate the key pair for each entry given a new key and a old key

For example: newKey and keyB,

  {
     "a": { "keyA": 1 , "keyB": 2, "newKey": 2 }, 
     "b": { "keyA": 1 , "keyB": 3, "newKey": 3 }, 
     "c": { "keyA": 1 , "keyC": 4 } // do nothing if the object doesn't have the given key.
  }

I know I can do v || '{ "newKey": "value" }' to merge the data, but I actually don’t know how to dynamically grab the value from keyB

That’s my current non-working attempt, since v is invalid as expect

UPDATE the_table
SET jsonb_col = (
  SELECT jsonb_object_agg(
    k, 
    v || '{ "newKey": v -> "keyB" }'
  )
  FROM jsonb_each(jsonb_col) AS dt(k,v)
);

Not sure how to properly use v token to construct the entry and skip the empty case.

The playground for this problem: https://dbfiddle.uk/qp29LGAV