Nested groups with duplicate keys trivially do not work, even with VDFDict

[GiovanH]

Input file demo.vdf:

"controller_mappings"
{
	"group"
	{
		"id"		"0"
	}
	"group"
	{
		"id"		"1"
	}
}

Test script:

import vdf
with open("demo.vdf") as fp:
    v = vdf.load(fp, mapper=vdf.VDFDict)
    print(type(v))
    print(type(v['controller_mappings']))
    print(vdf.dumps(v, pretty=True))

Expected output:
(demo.vdf)

Actual output:

<class 'vdf.vdict.VDFDict'>
<class 'vdf.vdict.VDFDict'>
"controller_mappings"
{
        "group"
        {
                "id" "0"
                "id" "1"
        }
}

Not only is this incorrect, but no error is thrown: the resulting data structure is silently malformed and passed along.

[rossengeorgiev]

Did you try setting merge_duplicate_keys=False ? The default behaviour is to merge duplicate keys.

See

vdf/vdf/__init__.py

Lines 77 to 79 in d762926

	``merge_duplicate_keys`` when ``True`` will merge multiple KeyValue lists with the
	same key into one instead of overwriting. You can se this to ``False`` if you are
	using ``VDFDict`` and need to preserve the duplicates.

[jericjan]

Since this has been open for quite a while, and I've also come across this. It seems that merge_duplicate_keys only applies when loading a file, not when dumping it. If you want to preserve key order but remove duplicates in a dump, you'll have to use collections.OrderedDict as the mapper when loading the file.