How to maintain order of insertion of keys when loading file into yaml?

Question

I am trying to use YAML for a python script. The YAML file I have written resembles this:

1:
  name: apple
  price: 5
3:
  name: orange
  price: 6
2:
  name: pear
  price: 2

When I load the YAML file using yaml.load the dictionary is sorted by the keys so it appears in order 1,2,3. How do I maintain the order 1,3,2?

Answer 1

In the YAML specification it is explicitly stated that mapping keys have no order. In a file however they have. If you want a simple way to solve this replace PyYAML with ruamel.yaml (disclaimer: I am the author of that package, which is a superset of PyYAML) and use round_trip_load(), it will give you ordered dictionaries without the hassle of using single mapping item sequence elements that you need for specifying ordered dicts the "official" way.

import ruamel.yaml

yaml_str = """\
1:
  name: apple
  price: 5
3:
  name: orange
  price: 6
2:
  name: pear
  price: 2
"""

data = ruamel.yaml.round_trip_load(yaml_str)
for key in data:
    print(key)

gives

1
3
2

BTW PyYAML doesn't sort by the keys, that ordering is just a side-effect of calculating hashes and inserting integer keys 1, 2 , 3 in python dicts.