Google Big Query - 从 JSON 字符串中提取未明确标记为键的键

Question

I have the following JSON string (I added line breaks for visibility, in actual code all of it is squished in one line)

    {"schema":
       {"properties":
          {"key_1":{"label":"key 1","type":"string"},
           "key_2":{"label":"key 2","type":"string"},
           "ley_3":{"label":"key 3","type":"string"},
           "key_4":{"label":"key 4","type":"string"},
           ...
          } 
       }
    }

What I am trying to do is extract all keys and labels associated with the key. I know how to do this when key is explicitly stated in JSON , but in this example key is not explicitly stated.

I followed Google Big Query documentation on working with JSON strings, and here is how far I got:

SELECT json_schema, JSON_EXTRACT(json_schema, "$.schema.properties"), JSON_EXTRACT(json_schema, "$.schema.properties[1]")
FROM schemas

json_schema is a column name in the schemas table.

This gets me to the right direction, but I don't know how to proceed from here. My desired output is (for example), is:

key    value
key_1  key 1
key_2  key 2
key_3  key 3
key_4  key 4

Here is the code to reproduce the example:

SELECT '{"schema":{"properties":{"key_1":{"label":"key 1","type":"string"},"key_2":{"label":"key 2","type":"string"},"key_3":{"label":"key 3","type":"string"},"key 4":{"label":"key_4","type":"string"}}}}' AS json_schema

Answer 1

Consider below approach

select key, 
  json_extract_scalar(regexp_extract(props, r'"' || key || '":({.*?})'), '$.label') value
from your_table, 
unnest([struct(json_extract(json_schema, '$.schema.properties') as props)]),
unnest(`bqutil.fn.json_extract_keys`(props)) key 

       

if applied to sample data in your question

with your_table as (
   select '''
    {"schema":
       {"properties":
          {"key_1":{"label":"key 1","type":"string"},
           "key_2":{"label":"key 2","type":"string"},
           "ley_3":{"label":"key 3","type":"string"},
           "key_4":{"label":"key 4","type":"string"}
          } 
       }
    }
   ''' json_schema
)              

output is