Support of ordered dictionary

[DonaldWL]

YAML does not support

from collections import OrderedDict

This is nice because you can force the order of the data within the YAML file to make it easer for the user to read.

Program:

from collections import OrderedDict

import yaml

MyYaml = OrderedDict()
MyYaml['c'] = 130
MyYaml['a'] = 10
MyYaml['b'] = 20

with open('x.yaml', 'w') as Fh:
yaml.dump(MyYaml, Fh)

Yaml File:

!!python/object/apply:collections.OrderedDict

• • • c
    • 130
  • • a
    • 10
  • • b
    • 20

[nitzmahone]

Ordered dictionaries are the default throughout Python 3.7+ already, and enabling it on older Pythons is pretty easy by changing the default Constructor impl in your loader.

Closing this, since it'll fairly soon be the default on all supported versions of Python anyway, so not much reason to invent a special way to override beyond what already exists.

[DonaldWL]

I am using python 3.9, so I really do not understand you comment. What I am trying to do is have yaml write out the data in the correct order. JSON does this. If I read in a yaml file and then write it out in json its file is in the order of the yaml file. But if I write out a new yaml file the order is what it is, but no the order that the original yaml file was in. Makes it harder for the user to modify the yaml file.

[nitzmahone]

@DonaldWL Oh, you're probably running into the default sort_keys=True on dump()- we're preserving the order internally already using the Python default ordered dict, but dump also sorts keys in mappings by default. Just do yaml.dump(..., sort_keys=False) and it should use the internal order.

[braindevices]

hmm does this supposed to work at all? I get 'cannot represent an object', OrderedDict when doing safe_dump