def find_properties_object(path: List[str], field: str, properties) -> Dict[str, Dict]:
"""
This function is trying to look for a nested "properties" node under the current JSON node to
identify all nested objects.
@param path JSON path traversed so far to arrive to this node
@param field is the current field being considered in the Json Tree
@param properties is the child tree of properties of the current field being searched
"""
result = {}
current_path = path + [field]
current = "_".join(current_path)
if isinstance(properties, str) or isinstance(properties, int):
return {}
else:
if "items" in properties:
return find_properties_object(path, field, properties["items"])
elif "properties" in properties:
# we found a properties object
return {current: properties["properties"]}
elif "type" in properties and is_simple_property(properties["type"]):
# we found a basic type
return {current: {}}
elif isinstance(properties, dict):
for key in properties.keys():
child = find_properties_object(path=current_path, field=key, properties=properties[key])
if child:
result.update(child)
elif isinstance(properties, list):
for item in properties:
child = find_properties_object(path=current_path, field=field, properties=item)
if child:
result.update(child)
return result
def extract_json_column(property_name: str, json_column_name: str, definition: Dict, column_name: str) -> str:
json_path = [property_name]
json_extract = jinja_call(f"json_extract({json_column_name}, {json_path})")
if "type" in definition:
if is_array(definition["type"]):
json_extract = jinja_call(f"json_extract_array({json_column_name}, {json_path})")
elif is_object(definition["type"]):
json_extract = jinja_call(f"json_extract({json_column_name}, {json_path})")
elif is_simple_property(definition["type"]):
json_extract = jinja_call(f"json_extract_scalar({json_column_name}, {json_path})")
return f"{json_extract} as {column_name}"