Compare orjson.loads() and json.loads() Using Example

Let’s take a look at an example to see the difference in performance between orjson and json. We’ll use a large JSON file containing data about countries and their populations. Let’s create a JSON file called countries.json with the following data:

countries.json:

[
  {"name": "United States", "population": 331002651},
  {"name": "China", "population": 1439323776},
  {"name": "India", "population": 1380004385},
  {"name": "Indonesia", "population": 273523615},
  {"name": "Pakistan", "population": 220892340},
  {"name": "Brazil", "population": 212559417},
  {"name": "Nigeria", "population": 206139589},
  {"name": "Bangladesh", "population": 164689383},
  {"name": "Russia", "population": 145934462},
  {"name": "Mexico", "population": 128932753}
]

In this example, below Python code compares the performance of the standard json library and the optimized orjson library for encoding and decoding JSON data. It first loads JSON data from a file, measures the time taken to decode the data using json.loads(), and then does the same using orjson.loads(). Finally, it checks if the decoded data is identical for both methods.

import json
import orjson
import time

# Load JSON data from file
with open('countries.json') as f:
    data = json.load(f)

# Measure the time taken to decode the JSON data using json.loads()
start_time = time.time()
decoded_data_json = json.loads(json.dumps(data))
end_time = time.time()
print(
    f"Time taken to decode JSON data using json.loads(): {end_time - start_time:.4f} seconds")

# Measure the time taken to decode the JSON data using orjson.loads()
start_time = time.time()
decoded_data_orjson = orjson.loads(orjson.dumps(data))
end_time = time.time()
print(
    f"Time taken to decode JSON data using orjson.loads(): {end_time - start_time:.4f} seconds")

# Check if the decoded data is the same for both methods
print(f"Decoded data is the same: {decoded_data_json == decoded_data_orjson}")

Output:

Time taken to decode JSON data using json.loads(): 0.0002 seconds
Time taken to decode JSON data using orjson.loads(): 0.0002 seconds
Decoded data is the same: True

Which is faster – orjson.loads() vs json.loads()?

As you can see, orjson.loads() is significantly faster than json.loads(). However, it’s important to note that orjson may not be compatible with all Python versions and may not support all features of the json library. Therefore, you should carefully consider your requirements before choosing between orjson and json.

Conclusion

In conlcusion, if you need to deserialize large amounts of JSON data and performance is important, you should consider using orjson. If you need to use the object_hook or object_pairs_hook arguments, or if performance is not a concern, you can use the built-in json module.