configparser vs. pyyaml: ini vs. yaml (Benchmarking)

When managing configuration files in Python, we have two popular choices: configparser for INI files and PyYAML for YAML files.

This tutorial compares configparser and PyYAML by benchmarking their parsing speed, memory usage, writing performance, and modification speed.

 

 

Parsing Benchmark Test

You can use the timeit module to benchmark parsing speed between configparser and PyYAML.

import configparser
import yaml
import timeit

ini_data = """
[Database]
host=localhost
port=3306
user=Ahmad
password=secret
"""
yaml_data = """
Database:
  host: localhost
  port: 3306
  user: Ahmad
  password: secret
"""
def parse_ini():
    config = configparser.ConfigParser()
    config.read_string(ini_data)
def parse_yaml():
    config = yaml.safe_load(yaml_data)
ini_time = timeit.timeit(parse_ini, number=10000)
yaml_time = timeit.timeit(parse_yaml, number=10000)
print(f"INI parsing time: {ini_time}")
print(f"YAML parsing time: {yaml_time}")

Output:

INI parsing time: 1.3473019000011845
YAML parsing time: 6.746310700000322

INI files are parsed much faster than YAML files in this benchmark which makes configparser more efficient.

 

Memory Usage

Measuring memory consumption helps determine which library is more memory-efficient during parsing.

import configparser
import yaml
import tracemalloc
ini_data = """
[Database]
host=localhost
port=3306
user=Ahmad
password=secret
"""
yaml_data = """
Database:
  host: localhost
  port: 3306
  user: Ahmad
  password: secret
"""
tracemalloc.start()
config = configparser.ConfigParser()
config.read_string(ini_data)
ini_memory = tracemalloc.get_traced_memory()
tracemalloc.stop()
tracemalloc.start()
config = yaml.safe_load(yaml_data)
yaml_memory = tracemalloc.get_traced_memory()
tracemalloc.stop()
print(f"INI memory usage: {ini_memory[1] - ini_memory[0]} bytes")
print(f"YAML memory usage: {yaml_memory[1] - yaml_memory[0]} bytes")

Output:

INI memory usage: 2350 bytes
YAML memory usage: 10810 bytes

configparser consumes less memory compared to PyYAML.

 

Writing Performance Test

Evaluate how quickly each library can write configuration data to a file.

import configparser
import yaml
import timeit
config_data = {
    'Database': {
        'host': 'localhost',
        'port': 3306,
        'user': 'Sara',
        'password': 'secret'
    }
}
def write_ini():
    config = configparser.ConfigParser()
    config.read_dict(config_data)
    with open('config.ini', 'w') as f:
        config.write(f)
def write_yaml():
    with open('config.yaml', 'w') as f:
        yaml.dump(config_data, f)
ini_time = timeit.timeit(write_ini, number=1000)
yaml_time = timeit.timeit(write_yaml, number=1000)
print(f"INI writing time: {ini_time}")
print(f"YAML writing time: {yaml_time}")

Output:

INI writing time: 0.4919393999989552
YAML writing time: 0.6114596000006713

Writing INI files with configparser is a bit faster than writing YAML files with PyYAML.

 

Modification Speed

You can measure how quickly each library handles modifications to existing configuration data by changing data in both files and measuring the time.

import configparser
import yaml
import timeit

ini_data = """
[Database]
host=localhost
port=3306
user=Ahmad
password=secret
"""
yaml_data = """
Database:
  host: localhost
  port: 3306
  user: Ahmad
  password: secret
"""
def modify_ini():
    config = configparser.ConfigParser()
    config.read_string(ini_data)
    config.set('Database', 'port', '3307')
def modify_yaml():
    config = yaml.safe_load(yaml_data)
    config['Database']['port'] = 3307
ini_time = timeit.timeit(modify_ini, number=10000)
yaml_time = timeit.timeit(modify_yaml, number=10000)
print(f"INI modification time: {ini_time}")
print(f"YAML modification time: {yaml_time}")

Output:

INI modification time: 0.7588212999980897
YAML modification time: 3.409955799997988

Modifying YAML data with PyYAML is much faster than using configparser for INI files.