Cases when YAML will not work if not properly declared

Wrong Interpretation

• There are some cases where yaml will not work and can be hard to debug the error
  • Example 1:
    • Suppose you have to store user name in yaml and read it later. One of the user is named “Christopher Null”

      name:
        first_name: Christopher
        second_name: Null
    

  • Example 2
    • Suppose you are storing short form of some countries and one of the country is Norway

      countries:
        - IN
        - SA
        - NO
    

Expected behaviour when yaml is read. Since YAML considers everything as a string.

{ 
	'countries': ['IN', 'SA', 'NO'], 
	'name': 
		{
			'first_name': 'Christopher', 
			'second_name': 'Null'
		}
}

Actual result:

{ 
	'countries': ['IN', 'SA', False], 
	'name': 
		{
			'first_name': 'Christopher', 
			'second_name': None
		}
}

NO is converted to False which was supposed to be short for Norway

Null is converted to None which was supposed to be second name

Data Type Mismatch

• Another example:
  • Suppose you want to store version information of the packages in yaml

    versions:
    	postgres: 9.3
      rabbitmq: 3.18.9
  

  Expectation

    {
    'versions':
    	{
    		'postgres' : '9.3'
    		'rabbitmq' : '3.18.9'
    	}
    }
  

Actual

{
'versions':
	{
		'postgres' : 9.3
		'rabbitmq' : '3.18.9'
	}
}

Since 9.3 is a floating value is is converted as floating value. We need to be extra careful in the parsing logic

Learning

Code snippet

import os
from yaml import safe_load

with open('sample.yaml') as file:
    data = safe_load(file) 

print(data)

countries:
  - IN
  - SA
  - NO

name:
  first_name: Christopher
  second_name: Null

versions:
  postgres: 9.3
  rabbitmq: 3.18.9

Related Reads

• https://ruudvanasseldonk.com/2023/01/11/the-yaml-document-from-hell?ref=architecture-notes