Using Glom on a nested structure, how to I move top level dictionary fields into a list of dictionaries?

Question

This is a question about the usage of Glom (https://github.com/mahmoud/glom/)

I have a dictionary that includes a list of other dictionaries.

{'date': '2020-01-01',
 'location': 'A',
 'items': [
     {'name': 'A', 'id': 'A1'},
     {'name': 'B', 'id': 'B1'},
     {'name': 'C', 'id': 'C1'}
]}

I would like to use Glom to move the outer, global dictionary fields 'date' and 'location' into list of dictionaries for the items.

This is the end result I try to reach

[
 {'name': 'A', 'id': 'A1', 'date': '2020-01-01', 'location': 'A'},
 {'name': 'B', 'id': 'B1', 'date': '2020-01-01', 'location': 'A'},
 {'name': 'C', 'id': 'C1', 'date': '2020-01-01', 'location': 'A'}
]

Alas, when the spec arrives at the 'item' of the dictionary, the other values are not longer accessable and the T object is set to the inner value instead.

from glom import glom, T

def update_dict(x, other_dict):
    x.update({'date': other_dict['date'], 'location': other_dict['location']})  
    return x.copy()

spec = (T, 'items', [(lambda x: update_dict(x, T()))])

data = {'date': '2020-01-01',
       'location': 'A',
       'items': [{'name': 'A', 'id': 'A1'},
                 {'name': 'B', 'id': 'B1'},
                 {'name': 'C', 'id': 'C1'}]}

glom(data, spec)  # print this

returns

[{'name': 'A', 'id': 'A1', 'date': T()['date'], 'location': T()['location']},
 {'name': 'B', 'id': 'B1', 'date': T()['date'], 'location': T()['location']},
 {'name': 'C', 'id': 'C1', 'date': T()['date'], 'location': T()['location']}]

Which is useless.

It's not difficult to update the dictionaries with regular Python code, but is there a way to do this within a Glom spec?

Answer 1

The trick is to pass the target as a global scope as well, this way, the Assign command can access the full target.

from glom import S, glom, Assign, Spec

spec = ('items', 
        [Assign( 'date', Spec(S['date']))], 
        [Assign( 'location', Spec(S['location']))]
       )

target = {'date': '2020-04-01',
     'location': 'A',
     'items': [
         {'name': 'A', 'id': 'A1'},
         {'name': 'B', 'id': 'B1'},
         {'name': 'C', 'id': 'C1'}
    ]}

spec = Spec(('items', [Assign( 'date', Spec(S['date']))], [Assign( 'location', Spec(S['location']))]))

glom(target, spec, scope=target)

Results in

[{'name': 'A', 'id': 'A1', 'date': '2020-04-01', 'location': 'A'},
 {'name': 'B', 'id': 'B1', 'date': '2020-04-01', 'location': 'A'},
 {'name': 'C', 'id': 'C1', 'date': '2020-04-01', 'location': 'A'}]